Plant Genetic Resources: Characterization and Utilization

Special issue on the 2nd International Symposium on “Genomics of Plant Genetic Resources”, 24–27 April 2010, Bologna, Italy

Guest editors Roberto Tuberosa Andreas Graner Rajeev K. Varshney Plant Genetic Resources Characterization and Utilization journals.cambridge.org/pgr

Publishing, Production, Marketing and publication. Replacement copies supplied after this date will be Subscription Sales Office: chargeable. Cambridge University Press The Edinburgh Building US POSTMASTERS: please send address corrections Plant Genetic Shaftesbury Road Resources USA address. Cambridge CB2 8RU Information for Authors: UK Please email manuscript (with any accompanying figures or tables) For Customers in North America: to the Journal Administrator Faye Kalloniatis at Cambridge University Press [email protected] Journals Fullfillment Dept Notes for Authors are available on the internet at 100 Brook Hill Drive journals.cambridge.org/pgr West Nyack 10994-2133 USA Offprints: The corresponding author of an accepted paper will receive a pdf offprint. Additional offprints are available for a fee Publisher: Katy Christomanou and should be ordered at proof stage. No page charges or Plant Genetic Resources: Characterization and Utilization is an inter- submission charges are levied by this journal. national journal published tri-annually by Cambridge University Press in April, August and December on behalf of NIAB. The online edition Copyright: q NIAB 2011. All rights reserved: permission for repro- is available at journals.cambridge.org/pgr. duction of any part of the journal (text, figures, tables or other matter) in any form (on paper, microfiche or electronically) should Special sales and supplements: be sought directly from CUP or a licence permitting restricted copying This Journal accepts advertising and inserts. We also provide bulk obtained from the Copyright Licensing Agency, Tottenham Court reprints of suitable papers to meet teaching or promotional require- Road, London WIP 9HE, UK, or in the USA from the Central ments. The Journal also publishes proceedings on behalf of academic Clearance Center, 27 Congress Street, Salem MA 01970. and corporate sponsors. Please contact Katy Christomanou at the Cambridge address above. Disclaimer: The information contained herein, including any expression of opinion and any projection or forecast, has been obtained Subscription information: from or is based upon sources believed by us to be reliable, but is The subscription rates for Volume 9, 2011 (4 issues) are: not guaranteed as to accuracy or completeness. The information is Institutional subscription supplied without obligation and on the understanding that any Internet/Print Package £321.00/$616.00 (Americas only) person who acts upon it or otherwise changes his/her position in Internet only £252.00/$486.00 (Americas only) reliance thereon does so entirely at his/her own risk. Print only £297.00/$535.00 (Americas only) Any supplements to this Journal published in the course of the CUP uk_journals_customerservice does not accept responsibility for annual volume are normally supplied to subscribers at no extra charge. any trade advertisement included in this publication. Printed by Latimer Trend, Plymouth, UK. Back volumes are available. Please contact CUP Publishing for further information. This journal issue has been printed on FSC-certified paper and cover board. FSC is an independent, non-governmental, not-for-profit Claims for non-receipt of journal issues will be considered on organization established to promote the responsible management of their merit and only if the claim is received within six months of the world’s forests. Please see www.fsc.org for information. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization 9(2) July 2011 ISSN 1479-2621

Plant Genetic Resources Characterization and Utilization

Contents

Genomics of plant genetic resources: an introduction Roberto Tuberosa, Andreas Graner and Rajeev K. Varshney ...... 151

Genomics of plant genetic resources: past, present and future Kyujung Van, Dong Hyun Kim, Jin Hee Shin and Suk-Ha Lee ...... 155

Genomic tools for the analysis of genetic diversity J. Antoni Rafalski ...... 159

Long terminal repeat retrotransposon Jeli provides multiple genetic markers for common wheat (Triticum aestivum) Nataliya V. Melnikova, Fedor A. Konovalov and Alexander M. Kudryavtsev ...... 163

HRM technology for the identiﬁcation and characterization of INDEL and SNP mutations in genes involved in drought and salt tolerance of durum wheat Linda Mondini, Miloudi M. Nachit, Enrico Porceddu and Mario A. Pagnotta ...... 166

Starch metabolism mutants in barley: A TILLING approach Riccardo Bovina, Valentina Talame`, Salvi Silvio, Maria Corinna Sanguineti, Paolo Trost, Francesca Sparla and Roberto Tuberosa ...... 170

Collection of mutants for functional genomics in the legume Medicago truncatula O. Calderini, M. Carelli, F. Panara, E. Biazzi, C. Scotti, A. Tava, A. Porceddu and S. Arcioni ...... 174

Polymorphisms in intron 1 of carrot AOX2b – a useful tool to develop a functional marker? He´lia Cardoso, Maria Doroteia Campos, Thomas Nothnagel and Birgit Arnholdt-Schmitt ...... 177

Next-Gen sequencing of the transcriptome of triticale Y. Xu, C. Badea, F. Tran, M. Frick, D. Schneiderman, L. Robert, L. Harris, D. Thomas, N. Tinker, D. Gaudet and A. Laroche ...... 181

A comparison of population types used for QTL mapping in Arabidopsis thaliana Joost J. B. Keurentjes, Glenda Willems, Fred van Eeuwijk, Magnus Nordborg and Maarten Koornneef ...... 185

Molecular characterization of the Latvian apple (Malus) genetic resource collection based on SSR markers and scab resistance gene Vf analysis Gunars Lacis, Irita Kota, Laila Ikase and Dainis Rungis ...... 189

Molecular adaptation of the chloroplast matK gene in Nymphaea tetragona, a critically rare and endangered plant of India Jeremy Dkhar, Suman Kumaria and Pramod Tandon ...... 193 The genetic make-up of the European landraces of the common bean S. A. Angioi, D. Rau, L. Nanni, E. Bellucci, R. Papa and G. Attene ...... 197

Investigation of genetic diversity in Russian collections of raspberry and blue honeysuckle Didier Lamoureux, Artem Sorokin, Isabelle Lefe`vre, Sergey Alexanian, Pablo Eyzaguirre and Jean-Franc¸ois Hausman ...... 202

Morpho-agronomic characterization and variation of indigo precursors in woad (Isatis tinctoria L.) accessions Luı´s Rocha, Carlos Carvalho, Sandra Martins, Fernando Braga and Valdemar Carnide ...... 206

Genetic diversity in woad (Isatis tinctoria L.) accessions detected by ISSR markers Luı´s Rocha, Sandra Martins, Valdemar Carnide, Fernando Braga and Carlos Carvalho ...... 210

Genetic diversity among Italian melon inodorus (Cucumis melo L.) germplasm revealed by ISSR analysis and agronomic traits S. Sestili, A. Giardini and N. Ficcadenti ...... 214

Analysis of genetic diversity in Citrus Franc¸ois Luro, Julia Gatto, Gilles Costantino and Olivier Pailly ...... 218

Diversity of seed storage protein patterns of Slovak accessions in jointed goatgrass (Aegilops cylindrica Host.) Edita Gregova´, Pavol Hauptvogel, Rene´ Hauptvogel, Ga´bor Vo¨ro¨sva´ry and Ga´bor Ma´lna´si Csizmadia ...... 222

Assessment of genetic diversity among Sri Lankan rice varieties by AFLP markers Gowri Rajkumar, Jagathpriya Weerasena, Kumudu Fernando and Athula Liyanage ...... 224

Molecular and morphological diversity in Japanese rice germplasm Fa´tima Bosetti, Maria Imaculada Zucchi and Jose´ Baldin Pinheiro ...... 229

Molecular characterization of the European rice collection in view of association mapping Brigitte Courtois, Raffaella Greco, Gianluca Bruschi, Julien Frouin, Nourollah Ahmadi, Gae¨tan Droc, Chantal Hamelin, Manuel Ruiz, Jean-Charles Evrard, Dimitrios Katsantonis, Margarida Oliveira, Sonia Negrao, Stefano Cavigiolo, Elisabetta Lupotto and Pietro Piffanelli ...... 233

Screening of barley germplasm for resistance to root lesion nematodes Shiveta Sharma, Shailendra Sharma, Tobias Keil, Eberhard Laubach and Christian Jung ...... 236

Allelic variation at the EF-G locus among northern Moroccan six-rowed barleys Takahide Baba, Ken-ichi Tanno, Masahiko Furusho and Takao Komatsuda ...... 240

Comparison of genomic and EST-derived SSR markers in phylogenetic analysis of wheat Agata Gadaleta, Angelica Giancaspro, Silvana Zacheo, Domenica Nigro, Stefania Lucia Giove, Pasqualina Colasuonno and Antonio Blanco ...... 243

Exploring the genetic diversity of the DRF1 gene in durum wheat and its wild relatives Domenico Di Bianco, Karthikeyan Thiyagarajan, Arianna Latini, Cristina Cantale, Fabio Felici and Patrizia Galefﬁ ...... 247

Allele variation in loci for adaptive response in Bulgarian wheat cultivars and landraces and its effect on heading date Stanislav Kolev, Dimitar Vassilev, Kostadin Kostov and Elena Todorovska ...... 251 Diversity of seed storage proteins in common wheat (Triticum aestivum L.) Zuzana Sˇramkova´, Edita Gregova´, Svetlana Sˇlikova´ and Ernest Sˇturdı´k ...... 256

Seed longevity in oilseed rape (Brassica napus L.) – genetic variation and QTL mapping Manuela Nagel, Maria Rosenhauer, Evelin Willner, Rod J. Snowdon, Wolfgang Friedt and Andreas Bo¨rner . . 260

Genetic variation at ﬂowering time loci in wild and cultivated barley James Cockram, Huw Hones and Donal M. O’Sullivan ...... 264

Cold-modulated expression of genes encoding for key enzymes of the sugar metabolism in spring and autumn cvs. of Beta vulgaris L. D. Paciﬁco, C. Onofri and G. Mandolino ...... 268

Study of symptoms and gene expression in four Pinus species after pinewood nematode infection Albina R. Franco, Carla Santos, Mariana Roriz, Rui Rodrigues, Marta R. M. Lima and Marta W. Vasconcelos ...... 272

Development and application of EST-SSRs for diversity analysis in Ethiopian grass pea M. Ponnaiah, E. Shiferaw, M. E. Pe` and E. Porceddu ...... 276

A novel genetic framework for studying response to artiﬁcial selection Randall J. Wisser, Peter J. Balint-Kurti and James B. Holland ...... 281

Molecular basis of cytoplasmic male sterility in beets: an overview Tetsuo Mikami, Masayuki P. Yamamoto, Hiroaki Matsuhira, Kazuyoshi Kitazaki and Tomohiko Kubo ..... 284

Agronomic and molecular analysis of heterosis in alfalfa C. Scotti, M. Carelli, O. Calderini, F. Panara, P. Gaudenzi, E. Biazzi, S. May, N. Graham, F. Paolocci and S. Arcioni ...... 288

Mapping QTLs for yield components and chlorophyll a ﬂuorescence parameters in wheat under three levels of water availability Ilona Czyczyło-Mysza, Izabela Marcin´ ska, Edyta Skrzypek, Małgorzata Chrupek, Stanisław Grzesiak, Tomasz Hura, Stefan Stojałowski, Beata Mys´ko´w, Paweł Milczarski and Steve Quarrie ...... 291

Identifying QTLs for cold-induced resistance to Microdochium nivale in winter triticale Magdalena Szechyn´ ska-Hebda, Maria We˛dzony, Mirosław Tyrka, Gabriela Gołe˛biowska, Małgorzata Chrupek, Ilona Czyczyło-Mysza, Ewa Dubas, Iwona Z˙ ur and Elz˙ bieta Golemiec ...... 296

Use of EcoTILLING to identify natural allelic variants of rice candidate genes involved in salinity tolerance S. Negra˜o, C. Almadanim, I. Pires, K. L. McNally and M. M. Oliveira ...... 300

Allele mining in the gene pool of wild Solanum species for homologues of late blight resistance gene RB/Rpi-blb1 Artem Pankin, Ekaterina Sokolova, Elena Rogozina, Maria Kuznetsova, Kenneth Deahl, Richard Jones and Emil Khavkin ...... 305

SCAR markers of the R-genes and germplasm of wild Solanum species for breeding late blight-resistant potato cultivars Ekaterina Sokolova, Artem Pankin, Maria Beketova, Maria Kuznetsova, Svetlana Spiglazova, Elena Rogozina, Isol’da Yashina and Emil Khavkin ...... 309

Exploitation of nuclear and cytoplasm variability in Hordeum chilense for wheat breeding Cristina Rodrı´guez-Sua´rez, Marıá J. Gimeńez, Marıá C. Ramı´rez, Azahara C. Martıń, Natalia Gutierrez, Carmen M. A´ vila, Antonio Martıń and Sergio G. Atienza ...... 313 Improvement of crop protection against greenbug using the worldwide sorghum germplasm collection and genomics-based approaches Yinghua Huang ...... 317

An overlooked cause of seed degradation and its implications in the efﬁcient exploitation of plant genetic resources Dionysia A. Fasoula ...... 321

Cultivated and wild Solanum species as potential sources for health-promoting quality traits Christina B. Wegener and Gisela Jansen ...... 324

Iron biofortiﬁcation of maize grain Owen A. Hoekenga, Mercy G. Lung’aho, Elad Tako, Leon V. Kochian and Raymond P. Glahn ...... 327

Polymorphism of waxy proteins in Spanish hulled wheats C. Guzma´n, L. Caballero, M. V. Gutierrez and J. B. Alvarez ...... 330

Molecular characterization of the Glu-Ay gene from Triticum urartu for its potential use in quality wheat breeding M. V. Gutie´rrez, C. Guzma´n, L. M. Martı´n and J. B. Alvarez ...... 334

Protein disulphide isomerase promoter sequence analysis of Triticum urartu, Aegilops speltoides and Aegilops tauschii Arun Prabhu Dhanapal, Mario Ciafﬁ, Enrico Porceddu and Elisa d’Aloisio ...... 338

Protein disulphide isomerase family in bread wheat (Triticum aestivum L.): genomic structure, synteny conservation and phylogenetic analysis E. d’Aloisio, A. R. Paolacci, A. P. Dhanapal, O. A. Tanzarella, E. Porceddu and M. Ciafﬁ ...... 342

Protein disulphide isomerase family in bread wheat (Triticum aestivum L.): protein structure and expression analysis A. R. Paolacci, M. Ciafﬁ, A. P. Dhanapal, O. A. Tanzarella, E. Porceddu and E. d’Aloisio ...... 347

Deployment of either a whole or dissected wild nuclear genome into the wheat gene pool meets the breeding challenges posed by the sustainable farming systems Ciro De Pace, Marina Pasquini, Patrizia Vaccino, Marco Bizzarri, Francesca Nocente, Maria Corbellini, Maria Eugenia Caceres, Pier Giorgio Cionini, Doriano Vittori and Gyula Vida ...... 352

Identiﬁcation of root morphology mutants in barley Riccardo Bovina, Valentina Talame`, Matteo Ferri, Roberto Tuberosa, Beata Chmielewska, Iwona Szarejko and Maria Corinna Sanguineti ...... 357 q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 151–154 ISSN 1479-2621 doi:10.1017/S1479262111000700

Genomics of plant genetic resources: an introduction

Roberto Tuberosa1*, Andreas Graner2 and Rajeev K. Varshney3 1Department of Agroenvironmental Science and Technology, Viale Fanin 44, 40127 Bologna, Italy, 2Leibniz Institute of Plant Genetics and Crop Plant Research (IPK), Corrensstraße 3, D-06466 Gatersleben, Germany and 3International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru, Hyderabad 502 324, Andhra Pradesh, India

This special issue of Plant Genetic Resources has inaccessible until recently, is now receiving increasing assembled 52 short articles selected among the over 350 attention in view of its importance in the regulation of oral and poster communications presented during the quantitative trait expression. Structural variability (e.g. 2nd International Symposium on Genomics of Plant Gen- copy number variations and/or indels) in what, as a etic Resources (GPGR2; www.GPGR2.com) held in reflection of our past ignorance, was often referred to Bologna, Italy, from April 24 to 27, 2010. The second as ‘junk DNA’ has instead been shown to be an important edition of GPGR2, co-organized by Bioversity International, driver of phenotypic variability (Magalhaes et al., 2007; the Leibniz Institute of Plant Genetics and Crop Plant Salvi et al., 2007). Operationally, the new paradigm has Research and the University of Bologna, followed the been set by next generation sequencing (NGS) and bioin- first edition organized in 2005 by the Chinese Academy formatics, quickly adopted as the gold standard required of Agricultural Science in Beijing, China. to deliver an exhaustive, accurate characterization of The overall objective of GPGR2 was to critically evalu- DNA variation, the discovery of single nucleotide poly- ate how the latest advances in genomics platforms and morphisms (SNPs) in massive number (Akhunov et al., resources have enhanced our capacity to investigate 2009; Varshney et al., 2009) and the analysis of synteny plant genetic resources and harness their potential for (Bolot et al., 2009). The cost of sequencing has already improving crop productivity and quality. The unifying fallen dramatically and keeps dropping, thus allowing picture that emerges from the articles collected in this for the direct analysis of large sets of accessions at a frac- issue shows the increasingly pivotal role of genomics tion of the cost of such an operation just a few short years for characterizing germplasm collections, best managing ago. NGS is also becoming the preferred choice for tran- genebanks, elucidating plant functions and identifying scriptome profiling (Forrest and Carninci, 2009; Tamura superior alleles at key loci for the selection of improved and Yonemaru, 2010). Unlike microarray platforms genotypes. In this brief introduction, we present an over- (Gupta et al., 2008; Pietsch et al., 2009), NGS offers the view of the main topics covered and have included distinct advantage of being able to report on changes additional references to provide some further reading across the entire transcriptome, including in rare tran- opportunities to the interested reader who wishes for a scripts. The power and benefits of NGS are particularly more comprehensive overview of the merits and limi- evident in species such as apple, potato or maize that tations of genomics-based approaches. suffer from low linkage disequilibrium while enjoying a The first group of articles (from page 155–184) offers a high level of polymorphism, two features which require glimpse of the tools, platforms and resources currently a highly detailed analysis at the DNA level to identify available to investigate the structural and functional haplotype diversity in germplasm collections. A level of diversity present in both the coding and non-coding genetic resolution sufficient to validate candidate genes regions of the plant genome. This complexity, largely and, in some cases, even identify causal polymorphisms can be attained by association mapping, an approach increasingly adopted to dissect the genetic basis of target traits (Ersoz et al., 2007; Rafalski, 2010). Genome- * Corresponding author. E-mail: [email protected] wide association mapping greatly benefits from the 152 R. Tuberosa et al. utilization of genome-wide SNP genotyping (Waugh et al., been summarized in two recent articles (Glaszmann 2009). Compared with other classes of molecular markers, et al., 2010; Varshney et al., 2010) and some examples SNPs are amenable to high-throughput automation at a are also presented in this special issue. The adoption of relatively low cost (Edwards et al., 2009). Although genomics-assisted breeding has considerably enhanced SNPs are already routinely utilized in a number of the effectiveness of breeding programmes and the important crops (e.g. rice, barley and maize; McNally response to selection (Varshney et al., 2005; Varshney et al., 2009; Hayden et al., 2010; Yan et al., 2010), con- and Tuberosa, 2007; Xu, 2010). Marker-assisted selection siderable work is still required to establish suitable (MAS) is now routinely included in many breeding pro- platforms in polyploid species such as wheat, in view grammes, particularly for traits controlled by major loci of the additional difficulties caused by the presence of (Ejeta and Knoll, 2007; Gupta and Langridge, 2010). This homoeologous loci (Ganal and Ro¨der, 2007; Berard notwithstanding, MAS for complex quantitative traits et al., 2009; Trebbi et al., 2011). (e.g. drought tolerance) remains a highly challenging The second group of articles (from page 185–280) undertaking, mainly because so much of the variation deals with one of the most difficult challenges faced for these traits is under the control of many genes, by genebank managers, namely the characterization of where the contribution of each is too small to allow their germplasm collections. Over the past two decades, mol- ready identification and so justify the implementation of ecular profiling has greatly improved the accuracy of an MAS breeding strategy (Collins et al., 2008). Nonethe- this characterization (Glaszmann et al., 2010), particu- less, some major loci which affect yield per se (i.e. not larly in the so-called ‘orphan’ species, which have linked to phenology) across a broad range of environ- been largely neglected also due to the lack of the ments have been described (Maccaferri et al., 2008; means to properly investigate their biodiversity Yadav et al., 2011). When major loci that affect organ (Varshney et al., 2010). An accurate characterization at growth (e.g. leaf size) and yield components (e.g. seed the genomic level is a fundamental step required for number and seed weight) are identified, modelling the (1) a more cost-efficient management of germplasm effects of the relevant quantitative trait loci (QTLs) in collections, both in situ and ex situ, (2) understanding response to environmental cues should provide a highly phylogenetic relationships among species (Bolot et al., effective approach for predicting yield performance 2009), (3) the assembly of core collections suitable for across different environmental conditions (Parent et al., association mapping studies (Maccaferri et al., 2011) 2010; Tardieu and Tuberosa, 2010). From a breeding and (4) assessing genetic similarity among accessions standpoint, an interesting alternative to phenotypic selec- sharing common ancestors (Maccaferri et al., 2007). tion for improving yield per se is provided by genome- The third group of articles (from page 281–360) pre- wide selection, which is increasingly being adopted for sents examples on how genomics-based approaches the improvement of those major crops where SNP can provide information useful for crop improvement platforms allow for a cost-effective, high-throughput programmes by providing the breeder with effective profiling of large populations (Bernardo, 2009). indirect selection schemes. In particular, a number of Over the past decade, genomics has ushered in novel articles (from page 324–351) deal with the improvement approaches which have produced a quantum leap in our of crop quality and nutritional value, a topic of great ability to characterize and utilize plant genetic resources interest in countries where malnutrition – the so-called (Tuberosa et al., 2002; Varshney et al., 2005; Ribaut et al., ‘hidden hunger’ – is rampant as a consequence of an 2010; Yano and Tuberosa, 2009; Langridge and Fleury, unbalanced diet. To address this problem, programmes 2011). Overall, molecular profiling suggests that the diver- such as HarvestPlus (www.harvestplus.org) are exploit- sity stored in genebanks has been only marginally tapped ing both natural and artificially induced variation to into so far. Cloning of the key genes which underlie increase the iron, zinc and provitamin A content in agronomically valuable traits will allow breeders to mine crops. For the genomics-based improvement of crops genebank collections much more effectively for novel for resistance to abiotic and biotic stresses, the generation alleles (Salvi and Tuberosa, 2007). Cloning will also pro- challenge programme (GCP; www.generationcp.org) has vide perfect MAS markers and the opportunity to identify developed a valuable molecular marker toolkit which novel alleles via TILLING in mutant collections (Talame` provides easy access to existing information on molecular et al., 2008; Sestili et al., 2010). Importantly, when access markers used in breeding programmes. Additionally, the to genes is hampered by low recombination, genetic GCP’s genotyping support service offers cost-efficient engineering will facilitate the transfer of these cloned genotyping services, both for fingerprinting and the anal- genes, especially when sourced from wild relatives. In ysis of genetic diversity, as well as for molecular breed- this regard, recent progress regarding site-specific recom- ing. The most significant results achieved by the GCP bination may open the door to further improve plant per- through the application of genomics approaches have formance through the replacement of distinct alleles. Genomics of plant genetic resources 153 Meeting the challenges posed by climate change and Feuillet C, Langridge P and Waugh R (2008) Cereal breeding the fast increasing demand for food, feed, fibre and fuel takes a walk on the wild side. Trends in Genetics 24: 24–23. will require an acceleration of the rate of crop improve- Forrest ARR and Carninci P (2009) Whole genome transcriptome ment, which in some key crops (e.g. wheat) has wor- analysis. RNA Biology 6: 107–112. riedly started to slow down (Tester and Langridge, Ganal MW and Ro¨der MS (2007) Microsatellite and SNP markers 2010). Achieving higher gains from selection will require in wheat breeding. In: Varshney RK and Tuberosa R (eds) enlarging the pool of genetic resources and exploiting Genomics-assisted Crop Improvement, Vol. 2: Genomics wild relatives of crops (Feuillet et al., 2008; Kovach and Applications in Crops. Houten: Springer, pp. 1–24. Glaszmann JC, Kilian B, Upadhyaya HD and Varshney RK (2010) McCouch, 2008) to identify superior alleles not yet uti- Accessing genetic diversity for crop improvement. Current lized in the cultivated gene pool. This brief introduction Opinion in Plant Biology 13: 167–173. and the articles of this special issue clearly show that Gupta PK and Langridge P (2010) Marker-assisted wheat breed- genomics of plant genetic resources is having a tangible ing: present status and future possibilities. Molecular impact on the way genebanks are being managed and Breeding 26: 145–161. how germplasm collections are being exploited to Gupta PK, Rustgi S and Mir RR (2008) Array-based high-throughput DNA markers for crop improvement. Heredity 101: 5–18. improve crop performance. Hayden MJ, Tabone TL, Nguyen TM, Coventry S, Keiper FJ, Fox RL, Chalmers KJ, Mather DE and Eglinton JK (2010) An informative set of SNP markers for molecular characterisation of Australian barley germplasm. Crop and Pasture Acknowledgements Science 61: 70–83. Kovach MJ and McCouch SR (2008) Leveraging natural diversity: The GPGR2 organizers thank the sponsors (for a back through the bottleneck. Current Opinion in Plant complete list, see the congress website at www.GPGR2. Biology 11: 193–200. com/sponsors.html) for their generous financial support Langridge P and Fleury D (2011) Making the most of ‘omics’ for crop breeding. Trends in Biotechnology 29: 33–40. and all participants for their contributions to the scientific Maccaferri M, Sanguineti MC, Xie C, Smith JSC and Tuberosa R programme of the Congress. (2007) Relationships among durum wheat accessions. II. A comparison of molecular and pedigree information. Genome 50: 385–399. References Maccaferri M, Sanguineti MC, Corneti S, Ortega JLA, Ben Salem M, Bort J, DeAmbrogio E, del Moral LFG, Demontis A, Akhunov E, Nicolet C and Dvorak J (2009) Single nucleotide El-Ahmed A, et al. (2008) Quantitative trait loci for grain polymorphism genotyping in polyploid wheat with the yield and adaptation of durum wheat (Triticum durum Illumina GoldenGate assay. Theoretical and Applied Desf.) across a wide range of water availability. Genetics Genetics 119: 507–517. 178: 489–511. Berard A, Le Paslier MC, Dardevet M, Exbrayat-Vinson F, Maccaferri M, Sanguineti MC, Demontis A, El-Ahmed A, Garcia Bonnin I, Cenci A, Haudry A, Brunel D and Ravel C del Moral L, Maalouf F, Nachit M, Nserallah N, Ouabbou (2009) High-throughput single nucleotide polymorphism H, Rhouma S, Royo C, Villegas D and Tuberosa R (2011) genotyping in wheat. (Triticum spp.). Plant Biotechnology Association mapping in durum wheat grown across a Journal 7: 364–374. broad range of water regimes. Journal of Experimental Bernardo R (2009) Genomewide selection for rapid intro- Botany 62: 409–438. gression of exotic germplasm in maize. Crop Science 49: Magalhaes JV, Liu J, Guimaraes CT, Lana UGP, Alves VMC, Wang 419–425. YH, Schaffert RE, Hoekenga OA, Pineros MA, Shaff JE, et al. BolotS,AbroukM,Masood-QuraishiU,SteinN,MessingJ,FeuilletC (2007) A gene in the multidrug and toxic compound and Salse J (2009) The ‘inner circle’ of the cereal genomes. extrusion (MATE) family confers aluminum tolerance in Current Opinion in Plant Biology 12: 119–125. sorghum. Nature Genetics 39: 1156–1161. Collins NC, Tardieu F and Tuberosa R (2008) Quantitative trait McNally KL, Childs KL, Bohnert R, Davidson RM, Zhao K, Ulat loci and crop performance under abiotic stress: where do VJ, Zeller G, Clark RM, Hoen DR, Bureau TE, Stokowski we stand? Plant Physiology 147: 469–486. Edwards KJ, Reid AL, Coghill JA, Berry ST and Barker GLA R, Ballinger DG, Frazer KA, Cox DR, Padhukasahasram B, (2009) Multiplex single nucleotide polymorphism (SNP)- Bustamante CD, Detlef Weigel D, Mackill DJ, Bruskiewich based genotyping in allohexaploid wheat using padlock RM, Ra¨tsch G, Buell CR, Leung H and Leach JE (2009) probes. Plant Biotechnology Journal 7: 375–390. Genomewide SNP variation reveals relationships among Ejeta G and Knoll JE (2007) Marker-assisted selection in landraces and modern varieties of rice. Proceedings of sorghum. In: Varshney RK and Tuberosa R (eds) the National Academy of Sciences of the United States of Genomics-assisted Crop Improvement, vol. 2. Houten: America 106: 12273–12278. Springer, pp. 187–206. Parent B, Suard B, Serraj R and Tardieu F (2010) Rice leaf growth Ersoz ES, Yu J and Buckler ES (2007) Applications of linkage dis- and water potential are resilient to evaporative demand equilibrium and association mapping. In: Varshney RK and and soil water deficit once the effects of root system are Tuberosa R (eds) Genomics-assisted Crop Improvement, neutralized. Plant Cell and Environment 33: 1256–1267. Vol. 1: Genomics Approaches and Platforms. Houten: Pietsch C, Sreenivasulu N, Wobus U and Roder MS (2009) Link- Springer, pp. 97–120. age mapping of putative regulator genes of barley grain 154 R. Tuberosa et al. development characterized by expression profiling. BMC Trebbi D, Maccaferri M, de Heer P, Sørensen A, van der Vossen Plant Biology 9: 4. E, Andries G, Giuliani S, Salvi S, Sanguineti MC, Massi A Rafalski JA (2010) Association genetics in crop improvement. and Tuberosa R (2011) High-throughput SNP discovery Current Opinion in Plant Biology 13: 174–180. and genotyping in durum wheat (Triticum durum Desf.). Ribaut JM, de Vicente MC and Delannay X (2010) Molecular Theoretical Applied Genetics. DOI:10.1007/s00122-011-1607-7. breeding in developing countries: challenges and perspec- Tuberosa R, Gill BS and Quarrie SA (2002) Cereal genomics: tives. Current Opinion in Plant Biology 13: 213–218. ushering in a brave new world. Plant Molecular Biology Salvi S and Tuberosa R (2007) Cloning QTLs in plants. In: Varsh- 48: 445–449. ney RK and Tuberosa R (eds) Genomics-assisted Crop Varshney RK and Tuberosa R (2007) Genomics-assisted Crop Improvement, Vol. 1: Genomics Approaches and Platforms. Improvement, Vol. 1: Genomics Approaches and Platforms, Houten: Springer, pp. 207–226. p. 386 and Vol. 2: Genomics Applications in Crops, p. 516. Salvi S, Sponza G, Morgante M, Tomes D, Niu X, Fengler KA, Houten: Springer. Meeley R, Ananiev EV, Svitashev S, Bruggemann E, Li B, Varshney RK, Glaszmann J-C, Leung H and Ribaut JM (2010) Hainey CF, Radovic S, Zaina G, Rafalski JA, Tingey SV, More genomic resources for less-studied crops. Trends Miao GH, Phillips RL and Tuberosa R (2007) Conserved Biotechnology 28: 452–460. noncoding genomic sequences associated with a flower- Varshney RK, Graner A and Sorrells ME (2005) Genomics- ing-time quantitative trait locus in maize. Proceedings of assisted breeding for crop improvement. Trends in Plant the National Academy of Sciences of the United States of Science 10: 621–630. America 104: 11376–11381. Varshney RK, Nayak SN, May GD and Jackson SA (2009) Next- Sestili F, Botticella E, Bedo Z, Phillips A and Lafiandra D (2010) generation sequencing technologies and their implications Production of novel allelic variation for genes involved in for crop genetics and breeding. Trends in Biotechnology starch biosynthesis through mutagenesis. Molecular Breed- ing 25: 145–154. 27: 522–530. Talame` V, Bovina R, Sanguineti MC, Tuberosa R, Lundqvist U Waugh R, Jannink JL, Muehlbauer GJ and Ramsay L (2009) The and Salvi S (2008) TILLMore, a resource for the discovery emergence of whole genome association scans in barley. of chemically induced mutants in barley. Plant Biotechnol- Current Opinion in Plant Biology 12: 218–222. ogy Journal 6: 477–485. Xu Y (2010) Molecular Plant Breeding. Wallingford: CABI. Tamura K and Yonemaru J (2010) Next-generation sequencing Yadav RS, Sehgal D and Vadez V (2011) Using genetic mapping for comparative transcriptomics of perennial ryegrass and genomics approaches in understanding and improving (Lolium perenne L.) and meadow fescue (Festuca pratensis drought tolerance in pearl millet. Journal of Experimental Huds.) during cold acclimation. Grassland Science 56: Botany 62: 397–408. 230–239. Yan JB, Yang XH, Shah T, Sanchez-Villeda H, Li JS, Warburton Tardieu F and Tuberosa R (2010) Dissection and modelling of M, Zhou Y, Crouch JH and Xu Y (2010) High-throughput abiotic stress tolerance in plants. Current Opinion in SNP genotyping with the GoldenGate assay in maize. Plant Biology 13: 206–212. Molecular Breeding 25: 441–451. Tester M and Langridge P (2010) Breeding technologies to Yano M and Tuberosa R (2009) Genome studies and molecular increase crop production in a changing world. Science genetics-from sequence to crops: genomics comes of age. 327: 818–822. Current Opinion in Plant Biology 12: 103–106. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 155–158 ISSN 1479-2621 doi:10.1017/S1479262111000098

Genomics of plant genetic resources: past, present and future

Kyujung Van1, Dong Hyun Kim1, Jin Hee Shin1 and Suk-Ha Lee1,2* 1Department of Plant Science and Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul 151-921, Korea and 2Plant Genomics and Breeding Institute, Seoul National University, Seoul 151-921, Korea

Abstract Plant genetic resources (PGR) include cultivars, landraces, wild species closely related to cultivated varieties, breeder’s elite lines and mutants. The loss of genetic diversity caused by the practice of agriculture and the availability of genetic information has resulted in a great effort dedicated to the collection of PGR. Prior to the advent of molecular proﬁling, accessions in germplasm collections were examined based on morphology. The development of molecular techniques now allows a more accurate analysis of large collections. Next-generation sequencing (NGS) with de novo assembly and resequencing has already provided a substantial amount of information, which warrants the coordination of existing databases and their integration into genebanks. Thus, the integration and coordination of genomic data into genebanks is very important and requires an international effort. From the determination of phenotypic traits to the application of NGS to whole genomes, every aspect of genomics will have a great impact not only on PGR conservation, but also on plant breeding programmes.

Keywords: genomics; germplasm collection; next-generation sequencing; plant genetic resources

Introduction Early impact of genomics on PGR

Plant genetic resources (PGR) began to establish around The advent of agriculture made possible by domestication 1993 as a consequence of growing concerns about bio- greatly affected the diversity of crops (Gepts, 2006). The diversity, its conservation and genetic erosion. Although voyages of Christopher Columbus marked the earliest the rate of population growth is slowing down, global recorded acquisition of new plant resources, and, ever food production is still a major challenge for the future since, collected plants have been conserved in botanical of mankind (Hoisington et al., 1999; Hammer, 2003; gardens and herbaria (Short, 2003). The rediscovery of Gepts, 2006). Therefore, securing PGR for future gener- Mendel’s law in the early 20th century helped the dramatic ations has become a priority not only in developing increase in agricultural productivity, although the overall countries but also in the entire world. The development genetic diversity decreased as a result of modern agri- and application of molecular techniques and genomics cultural practices. Fearing genetic erosion, the world have dramatically improved the characterization and community increased the effort to better evaluate PGR in deployment of PGR. This review surveys the past and genebanks (Hoisington et al., 1999). current status of the application of genomics to the The characterization of PGR by comparisons of plant PGR characterization and discusses future directions. morphology, such as yield, colour, texture, taste, etc., is the simplest and easiest approach (Gilbert et al., 1999; Hoisington et al., 1999). In addition to these qualitative/ quantitative phenotypic traits, pedigree analysis and * Corresponding author. E-mail: [email protected] geographical distribution are also helpful for measuring 156 K. Van et al. genetic diversity (Hammer, 2003). A renewed impetus haplotypes (Hamblin et al., 2007; Yan et al., 2010). towards PGR characterization was made possible by the This multiplexed genotyping technology facilitates the development of modern molecular techniques. effective examination and selection of germplasms by unravelling novel and potentially agronomically useful alleles (Tanksley and McCouch, 1997). The QTL mapping Current status of plant genomics of soybean rust was successfully conducted by SNP genotyping using the GoldenGate assay (Hyten et al., 2009). The genetic diversity of major crops has been declining These NGS technologies and the massively developed through domestication and the introduction of modern genome-wide markers are also applied for the construc- plant breeding (Tanksley and McCouch, 1997; Hyten tion of high-density maps and genetic diversity analysis et al., 2009). To prevent the genetic vulnerability of (Gupta et al., 2008). crops and to preserve valuable genetic resources, it needs to collect, preserve, examine and utilize germplasm effectively. The concept of the core set was proposed to Future directions minimize replicates and ensure the representation of the maximum genetic diversity of the entire germplasm A wealth of genetic resources in Arabidopsis and other collection (Frankel, 1984; Brown, 1989; van Hintum, model species have promoted great advances in plant 1999). Phenotyping was the traditional criteria for science. Furthermore, whole genome sequencing pro- germplasm evaluation; however, currently, these evalu- jects involving more than 20 plants will be completed ations are changed to genotyping by molecular markers in the near future (Gupta et al., 2008). With the improve- (Tanksley and McCouch, 1997). ment in sequencing techniques, more genetic resources, Genetic markers are powerful tools for genetic map- including the sequences, will be available in the future. ping, and molecular markers are highly polymorphic, Second (next) generation sequencers – Illumina’s GA, easily detected and unaffected by the environment Roche’s 454 and Applied Biosystems’ SOLiD – have gen- (Andersen and Lubberstedt, 2003). Various molecular erated large amounts of short DNA sequence reads. markers have been developed, such as restriction These have been updated to produce longer read lengths fragment length polymorphisms, randomly amplified and greater amounts of sequence reads. Currently, sev- polymorphic DNA, simple sequence repeats, amplified eral companies are attempting to introduce a new fragment length polymorphisms and single nucleotide sequencing machine, which will be called third gener- polymorphisms (SNPs) (Gupta et al., 2001), which are ation sequencing (Rusk, 2009). Helicos Biosciences used for the construction of genetic and physical maps. developed a true single molecule sequencer that These markers are applied in plant breeding for quanti- sequenced the virus M13 genome by an amplification- tative trait loci (QTLs) mapping, map-based cloning, free method (Harris et al., 2008). Pacific Biosciences marker-assisted selection, etc. (Moose and Mumm, 2008). developed a single molecular real-time sequencing Previously, genome-sequencing projects depended on machine, based on an assessment of the temporal order Sanger sequencing methods. Recently, introduction of of incorporation of fluorescently labelled nucleotides, next-generation sequencing (NGS) technologies into which can produce reads longer than 1 kb (Eid et al., plant breeding programmes has enabled the acquisition 2009). Oxford Nanopore’s sequencer is designed to of high-throughput sequence data inexpensively in a avoid amplification or labelling by detecting a direct elec- short time (Morozova and Marra, 2008). However, the trical signal (Clarke et al., 2009). Despite dramatic de novo assembly of plant genomes using NGS with improvements in sequencing speed and capacity, third short-read length is not yet adequate because most generation sequencers will not completely replace the plant genomes are large and harbour long repeat previous sequencing methods. Frequent use by research- sequences (Varshney et al., 2009). Thus, NGS technol- ers will likely reveal not only the benefits but also the ogies are applied for the resequencing of species for limitations of these new techniques. Similar to the use which a complete reference genome sequence exists of second generation sequencers together with ABI and are actively used for high-throughput genotyping 3730, new sequencers will also be used with earlier of up to a million SNP markers in Arabidopsis and several technologies. polyploidy crops (Rostoks et al., 2006; Weber et al., 2007; Until several years ago, whole genome plant sequen- Hyten et al., 2008; Akhunov et al., 2009; Yan et al., 2010). cing projects were limited to model species. However, Genome-wide SNP genotyping is a powerful tool for de novo sequencing and assembly are now easier due association mapping and evolutionary studies (Akhunov to longer reads and lower costs, which in the past few et al., 2009). Furthermore, SNP markers can be used years has allowed for much greater sequencing depth. more effectively when combined with genotypes and In addition to de novo genome sequencing, the whole Genomics of plant genetic resources 157 genome sequence variations in 1001 accessions of Dixon J, Foquet M, Gaertner A, Hardenbol P, Heiner C, Arabidopsis were analyzed in 2008 (Weigel and Mott, Hester K, Holden D, Kearns G, Kong X, Kuse R, Lacroix Y, Lin S, Lundquist P, Ma C, Marks P, Maxham M, 2009). Furthermore, in rice, a high-throughput method Murphy D, Park I, Pham T, Phillips M, Roy J, Sebra R, for genotyping recombinant populations was developed Shen G, Sorenson J, Tomaney A, Travers K, Trulson M, (Huang et al., 2009). Third generation sequencers could Vieceli J, Wegener J, Wu D, Yang A, Zaccarin D, Zhao P, be also used for detection of sequence variation, associ- Zhong F, Korlach J and Turner S (2009) Real-time DNA ations between important agronomic traits and gene sequencing from single polymerase molecules. Science identification in regulatory networks by ChIP-chip and 323: 133–138. Frankel PH (1984) Genetic perspective of germplasm ChIP-seq protocols. conservation. In: Arber W, Llimensee K, Peacock WJ and Presently, bioinformatics is the major bottleneck for a Starlinger P (eds) Genetic manipulations: Impact on Man more complete exploitation of the information of genetic and Society. Cambridge: Cambridge University Press, resources that is rapidly accumulating. The integration pp. 161–170. and organization of the available genomic resources to Gepts P (2006) Plant genetic resources conservation and utilization: the accomplishments and future of a societal facilitate their use by researchers are therefore important. insurance policy. Crop Science 46: 2278–2292. It could be a similar concept to that of an ‘omic space’ Gilbert JE, Lewis RV, Wilkinson MJ and Caligari PDS (1999) comprising a comprehensive omic planes (Toyoda Developing an appropriate strategy to assess genetic varia- and Wada, 2004). Several integrated databases, such bility in plant germplasm collections. Theoretical and as the arabidopsis information resource (Arabidopsis), Applied Genetics 98: 1125–1131. Gupta PK, Roy JK and Prasad M (2001) Single nucleotide Gramene (rice) and SoyBase (soybean), provide genetic polymorphisms: a new paradigm for molecular marker maps, genomic sequences, gene predictions, expressed technology and DNA polymorphism detection with sequence tags, marker data, QTLs, repetitive sequences, emphasis on their use in plants. Current Science India etc. One of the most significant contributions of the 80: 524–535. comprehensive genomic resources is that it provides a Gupta PK, Rustgi S and Mir RR (2008) Array-based high-through- benefit to researchers who want to start new experiments put DNA markers for crop improvement. Heredity 101: 5–18. or compare related information. Hamblin MT, Warburton ML and Buckler ES (2007) Empirical comparison of simple sequence repeats and single nucleotide polymorphisms in assessment of maize diversity and relatedness. Public Library of Science ONE 2: e1367. Acknowledgements Hammer K (2003) A paradigm shift in the discipline of plant genetic resources. Genetic Resources and Crop Evolution This work was supported by a grant from the BioGreen 50: 3–10. 21 Project (code no. 20080401034010), Rural Develop- Harris TD, Buzby PR, Babcock H, Beer E, Bowers J, Braslavsky I, ment Administration, the Republic of Korea. S.-H. Lee is Causey M, Colonell J, DiMeo J, Efcavitch JW, Giladi E, Gill J, Healy J, Jarosz M, Lapen D, Moulton K, Quake SR, grateful for the Senior Visiting Fellowship provided by Steinmann K, Thayer E, Tyurina A, Ward R, Weiss H and the Institute of Advanced Studies at the University of Xie Z (2008) Single-molecule DNA sequencing of a viral Bologna, Italy. genome. Science 320: 106–109. Hoisington D, Khairallah M, Reeves T, Ribaut J-M, Skovmand B, Taba S and Warburton M (1999) Plant genetic resources: what can they contribute toward increased crop pro- References ductivity? Proceeding of the National Academy of Sciences USA 96: 5937–5943. Akhunov E, Nicolet C and Dvorak J (2009) Single nucleotide Huang X, Feng Q, Qian Q, Zhao Q, Wang L, Wang A, Guan J, polymorphism genotyping in polyploid wheat with the Fan D, Weng Q, Huang T, Dong G, Sang T and Han B Illumina GoldenGate assay. Theoretical and Applied (2009) High-throughput genotyping by whole-genome Genetics 119: 507–517. resequencing. Genome Research 19: 1068–1076. Andersen JR and Lubberstedt T (2003) Functional markers in Hyten DL, Song Q, Choi IY, Yoon MS, Specht JE, Matukumalli plants. Trends in Plant Science 8: 554–560. LK, Nelson RL, Shoemaker RC, Young ND and Cregan PB Brown AHD (1989) The case for core sets. In: Brown AHD, (2008) High-throughput genotyping with the GoldenGate Frankel OH, Marshall DR and Williams JT (eds) The assay in the complex genome of soybean. Theoretical Use of Plant Genetic Resources. Cambridge: Cambridge and Applied Genetics 116: 945–952. University Press, pp. 136–155. Hyten DL, Smith JR, Frederick RD, Tucker ML, Song Q and Clarke J, Wu H-C, Jayasinghe L, Patel A, Reid S and Bayley H Cregan PB (2009) Bulked segregant analysis using the (2009) Continuous base identification for single-molecule GoldenGate assay to locate the Rpp3 locus that confers nanopore DNA sequencing. Nature Nanotechnology 4: resistance to soybean rust in soybean. Crop Science 49: 265–270. 265–271. Eid J, Fehr A, Gray J, Luong K, Lyle J, Otto G, Peluso P, Rank D, Moose SP and Mumm RH (2008) Molecular plant breeding as Baybayan P, Bettman B, Bibillo A, Bjornson K, Chaudhuri the foundation for 21st century crop improvement. Plant B, Christians F, Cicero R, Clark S, Dalal R, deWinter A, Physiology 147: 969–977. 158 K. Van et al. Morozova O and Marra MA (2008) Applications of next- van Hintum TJL (1999) The general methodology for creating generation sequencing technologies in functional geno- a core collection. In: Johnsons RC and Hodgkin T (eds) mics. Genomics 92: 255–264. Core Set for Today and Tomorrow. Rome: International Rostoks N, Ramsay L, MacKenzie K, Cardle L, Bhat PR, Roose Plant Genetic Resources Institute (IPGRI), pp. 10–17. ML, Svensson JT, Stein N, Varshney RK, Marshall DF, Varshney RK, Nayak SN, May GD and Jackson SA (2009) Next- Graner A, Close TJ and Waugh R (2006) Recent history of generation sequencing technologies and their implications artificial outcrossing facilitates whole-genome association for crop genetics and breeding. Trends in Biotechnology mapping in elite inbred crop varieties. Proceedings of the 27: 522–530. National Academy of Sciences USA 103: 18656–18661. Weber APM, Weber KL, Carr K, Wilkerson C and Ohlrogge JB Rusk N (2009) Cheap third-generation sequencing. Nature (2007) Sampling the Arabidopsis transcriptome with Methods 6: 244–245. massively parallel pyrosequencing. Plant Physiology 144: Short P (2003) In Pursuit of Plants. Portland: Timber Press. 32–42. Tanksley SD and McCouch SR (1997) Seed banks and molecular Weigel D and Mott R (2009) The 1001 genomes project for maps: unlocking genetic potential from the wild. Science Arabidopsis thaliana. Genome Biology 10: 107. 277: 1063–1066. Yan JB, Yang XH, Shah T, Sanchez-Villeda H, Li JS, Warburton Toyoda T and Wada A (2004) Omic space: coordinate-based M, Zhou Y, Crouch JH and Xu YB (2010) High-throughput integration and analysis of genomic phenomic interactions. SNP genotyping with the GoldenGate assay in maize. Bioinformatics 20: 1759–1765. Molecular Breeding 25: 441–451. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 159–162 ISSN 1479-2621 doi:10.1017/S1479262111000104

Genomic tools for the analysis of genetic diversity

J. Antoni Rafalski* DuPont Agricultural Biotechnology Group and Pioneer Hi-Bred International, Wilmington, DE 19880-0353, USA

Abstract We now understand that many different types of DNA structural polymorphisms contribute to functional diversity of plant genomes, including single nucleotide polymorphisms, insertions of retrotransposons and DNA transposons, including Helitrons carrying pseudogenes, and other types of insertion–deletion polymorphisms, many of which may contribute to the phenotype by affecting gene expression through a variety of mechanisms including those involving non-coding RNAs. These polymorphisms can now be probed with tools such as array comparative genomic hybridization and, most comprehensively, genomic sequencing. Rapid developments in next generation sequencing will soon make genomic sequencing of germplasm collections a reality. This will help eliminate an important difﬁculty in the estimation of genetic relationships between accessions caused by ascertainment bias. Also, it has now become obvious that epigenetic differences, such as cytosine methylation, also contribute to the heritable phenotype, although detailed understanding of their transgenerational stability in crop species is lacking. The degree of linkage disequilibrium of epialleles with DNA sequence polymorphisms has important implications to the analysis of genetic diversity. Epigenetic marks in complete linkage disequilibrium (LD) with DNA polymorphisms do not add additional diversity information. However, epialleles in partial or low LD with DNA sequence alleles constitute another layer of genetic information that should not be neglected in germplasm analysis, especially if they exhibit transgenerational stability.

Keywords: comparative genomic hybridization; epigenetic; haplotype; next generation sequencing; single nucleotide polymorphism

Introduction chain reaction. Development of single nucleotide polymorphism (SNP)-based markers brought a new level Thirty years ago Botstein et al. (1980) introduced the of resolution to the analysis of genetic diversity and for method of constructing genetic maps with DNA markers, most applications superseded other genetic marker cat- known as restriction fragment length polymorphisms egories. More recently, DNA sequencing of partial or com- (RFLP). This development revolutionized genetic mapping plete genomes from multiple individuals has expanded and the analysis of diversity. Subsequent methodological our understanding of the range of intraspecific genetic advances, such as development of simple sequence variation encountered in higher plants (Fu and Dooner, repeat markers (SSRs), random amplification of poly- 2002; Yang and Bennetzen, 2009). With the rapid decline morphic DNA (RAPD) (Williams et al., 1990) and amplified in the cost of DNA sequencing and new technological fragment length polymorphisms (AFLP) (Zabeau and Voss, developments, it is certain that genome sequencing of 1993), were enabled by the development of polymerase germplasm collection will become accessible, eliminating biases present in existing genotyping methodologies, although it will also impose a significant data analysis overhead, necessitating increased investment in bioinfor- * Corresponding author. E-mail: [email protected] matics. The proposed 1001 Arabidopsis genomes project 160 J. A. Rafalski (http://1001genomes.org/about.html) is a sign of things to come. Beyond DNA sequence, there is a renewed interest in the epigenetic marks, such as cytosine methylation, dec- orating DNA and chromatin, and potentially influencing the phenotype. We have discussed the impact of these developments on the analysis of genetic diversity.

Intraspeciﬁc diversity and the phenotype

Genomic sequencing of diverse genotypes in several Teosinte accessions plant species demonstrated that in addition to SNPs and SSR polymorphisms, extensive intraspecific differences include large insertions/deletions frequently composed of highly repetitive sequences such as retrotransposons and DNA transposons (Wang and Dooner, 2006), and in some cases also genes (Belo´ et al., 2009; Springer et al., 2009). For example, the complement of disease resistance Maize genes may differ between accessions (Chin et al., 2001; Fig. 1. An example of ascertainment bias. SNP markers Yahiaoui et al., 2009). Sequences that do not code for ascertained in elite maize inbred collection were used to proteins may nevertheless affect the phenotype, by sup- fingerprint a set of maize and teosinte lines. Genetic dis- plying enhancers or promoters to nearby genes, or tances between maize lines appear much lengthened with respect to those between teosinte accessions, which are fore- code for small RNAs, which affect expression of other shortened. Using unbiased genotyping method eliminates genes by a variety of mechanisms (Chen, 2009). Pseudo- this disparity. Data courtesy of Stan Luck (Pioneer Hi-Bred). genes, which in maize are frequently generated by Heli- tron transposons, are sometimes transcribed in sense or in turn, some alleles common in adapted material may antisense direction, also affecting gene expression phe- be rare in non-elite accessions. As a result, genetic notype (Yang and Bennetzen, 2009). distances determined in the ascertainment population If these types of polymorphisms are in linkage may be lengthened in comparison with those in the disequilibrium with genetic markers used for germplasm non-ascertained population (Fig. 1). It is difficult to ident- characterization (predominantly SNPs and SSRs), then no ify a priori an appropriate collection of germplasm for additional information other than marker genotype is ascertainment (marker discovery), given unbalanced needed to reflect correctly the underlying genetic relation- representation of different types of germplasm in many ships of accessions. However, if linkage disequilibrium collections. Perhaps, the most appropriate unbiased (LD) between markers for germplasm fingerprinting and methodology for germplasm fingerprinting is genotyping genic or non-genic large indel polymorphisms breaks by genomic sequencing. down rapidly, direct genotyping of these differences may The sequencing technology is rapidly approaching be necessary by DNA sequencing or other methods such the stage where it will become a cost-effective tool for as array comparative genomic hybridization (Belo´ et al., genotyping (Edwards and Batley, 2009; Varshney et al., 2009; Springer et al., 2009). This is likely to occur in the 2009). A number of accessions will be simultaneously case of variants, which occurred recently on the back- sequenced in each lane of the instrument, after appropri- ground of pre-existing haplotype pattern. ate encoding. Depending on the size of the genome, An important issue not always appreciated in the germ- some form of reduced representation analysis (Yuan et al., plasm analysis context is the prevalence of ascertainment 2003) will probably be necessary to focus the effort on bias, which occurs when polymorphic loci are identified non-repetitive fraction of the genome. (ascertained) in one collection of germplasm, but used to evaluate diversity in another set (Clark et al., 2005). For example, a collection of SNP loci identified in a set Perspective on epigenotyping of germplasm of cultivated lines will not correctly represent polymorphic loci present in unadapted accessions, leading It is well established that epigenetic variation encoded by to incorrect estimates of genetic distances in the latter DNA base modifications such as 5-methylcytidine affects set of germplasm. Many polymorphic loci in the non- phenotype in animals and plants (Peaston and Whitelaw, adapted accessions will not be represented in the SNP 2006; Henderson and Jacobsen, 2007; Chandler and collection developed from adapted germplasm, and, Alleman, 2008). Some of the epialleles in plants are Genomic tools for genetic diversity analysis 161 remarkably stable and affect important plant character- detected by array comparative genome hybridization. istics (Cubas et al., 1999). It is therefore reasonable to Theoretical and Applied Genetics 120: 355–357. Botstein D, White RL, Skolnick MH and Davis RW (1980) propose that a complete characterization of a germplasm Construction of a genetic map in man using restriction accession or a breeding stock should involve not only the fragment length polymorphisms. American Journal of description of the genotype but also of the epigenotype. Human Genetics 32: 314–331. It has recently been demonstrated that recursive selection Chandler V and Alleman M (2008) Paramutation: epigenetic for a yield component in canola results in plants that instructions passed across generations. Genetics 178: are genetically identical but can be distinguished by 1839–1844. Chen X (2009) Small RNAs and their roles in plant development. DNA methylation differences and exhibit significant Annual Review of Cell and Development Biology 25: 21–44. differences in yield (Hauben et al., 2009). The tools for Chin DB, Arroyo-Garcia R, Ochoa OE, Kesseli RV, Lavelle DO comprehensive epigenotyping are available and involve and Michelmore RW (2001) Recombination and spon- chemical deamination of m5C to U followed by DNA taneous mutation at the major cluster of resistance genes sequencing, enabling single base resolution across the in lettuce (Lactuca sativa). Genetucs 157: 831–849. Clark AG, Hubisz MJ, Bustamante CD, Williamson SH and whole genome, albeit at considerable expense (Lister and Nielsen R (2005) Ascertainment bias in studies of human Ecker, 2009; Lister et al., 2009; Wang et al., 2009). The high genome-wide polymorphism. Genome Research 15: throughput sequencing technology, especially rapidly 1496–1502. developing single molecule sequencing (Edwards Cubas P, Vincent C and Coen E (1999) An epigenetic mutation and Batley, 2009), promises to enable comprehensive responsible for natural variation in floral symmetry. Nature 401: 157–161. epigenotyping of germplasm collections in the coming Edwards D and Batley J (2009) Plant genome sequencing: years. Currently, several options exist for epigenotyping applications for crop improvement. Plant Biotechnol of a subset of the genome, for example by excluding Journal 8: 2–9. repetitive fraction of the genome (Peterson et al., 2002) Flusberg BA, Webster DR, Lee JH, Travers KJ, Olivares EC, or capturing specific sequences of interest (Hodges Clark TA, Korlach J and Turner SW (2010) Direct detection et al., 2009). of DNA methylation during single-molecule, real-time sequencing. Nature Methods 7: 461–465. Fu H and Dooner HK (2002) Intraspecific violation of genetic colinearity and its implications in maize. Proceedings of Conclusions the National Academy of Sciences USA 99: 9573–9578. Hauben M, Haesendonckx B, Standaert E, Van Der Kelen K, Rapid technological developments are changing our Azmi A, Akpo H, Van Breusegem F, Guisez Y, Bots M, understanding of genetic diversity, by allowing increas- Lambert B, Laga B and De Block M (2009) Energy use ingly dense genotyping and identification of types of efficiency is characterized by an epigenetic component genetic polymorphisms that were previously not easily that can be directed through artificial selection to increase yield. Proceedings of the National Academy of Sciences accessible to molecular analysis. In the next few years, USA 106: 20109–20114. another step change will occur with the availability of Henderson IR and Jacobsen SE (2007) Epigenetic inheritance in inexpensive genomic sequencing and development of plants. Nature 447: 418–424. tools for direct probing of epigenetic layer of information Hodges E, Smith AD, Kendall J, Xuan Z, Ravi K, Rooks M, (Flusberg et al., 2010). These developments will further Zhang MQ, Ye K, Bhattacharjee A, Brizuela L, McCombie WR, Wigler M, Hannon GJ and Hicks JB (2009) High enable the understanding of relationship between haplo- definition profiling of mammalian DNA methylation by type defined at the sequence level and phenotypic array capture and single molecule bisulfite sequencing. expression, through the use of association mapping and Genome Research 19: 1593–1605. genome prediction techniques. To fully exploit these Lister R and Ecker JR (2009) Finding the fifth base: genome-wide developments, we need to better understand the extent sequencing of cytosine methylation. Genome Research 19: of linkage disequilibrium in the germplasm of interest. 959–966. Lister R, Pelizzola M, Dowen RH, Hawkins RD, Hon G, Tonti-Filippini J, Nery JR, Lee L, Ye Z, Ngo Q-M, Edsall L, Antosiewicz-Bourget J, Stewart R, Ruotti V, Millar AH, Acknowledgements Thomson JA, Ren B and Ecker JR (2009) Human DNA methylomes at base resolution show widespread epi- I appreciate many discussions with Scott Tingey and with genomic differences. Nature 462: 315–322. Peaston AE and Whitelaw E (2006) Epigenetics and phenotypic all of my professional colleagues at DuPont/Pioneer variation in mammals. Mammalian Genome 17: 365–374. Hi-Bred Int. Peterson DG, Wessler SR and Paterson AH (2002) Efficient capture of unique sequences from eukaryotic genomes. References Trends in Genetics 18: 547–550. Springer NM, Ying K, Fu Y, Ji T, Yeh C-T, Jia Y, Wu W, Richmond Belo´ A, Beatty MK, Hondred D, Fengler KA, Li B and Rafalski A T, Kitzman J, Rosenbaum H, Iniguez AL, Barbazuk WB, (2009) Allelic genome structural variations in maize Jeddeloh JA, Nettleton D and Schnable PS (2009) 162 J. A. Rafalski Maize inbreds exhibit high levels of copy number Williams JGK, Kubelik AR, Livak KJ, Rafalski JA and Tingey SV variation (CNV) and presence/absence variation (PAV) in (1990) DNA polymorphisms amplified by arbitrary primers genome content. PLoS Genetics 5(11): e1000734. are useful as genetic-markers. Nucleic Acids Research 18: doi:10.1371/journal.pgen.1000734. 6531–6535. Varshney RK, Nayak SN, May GD and Jackson SA (2009) Yahiaoui N, Kaur N and Keller B (2009) Independent evolution of Next-generation sequencing technologies and their functional Pm3 resistance genes in wild tetraploid wheat and implications for crop genetics and breeding. Trends in domesticated bread wheat. Plant Journal 57: 846–856. Biotechnology 27: 522–530. Yang L and Bennetzen JL (2009) Distribution, diversity, Wang Q and Dooner HK (2006) Remarkable variation in maize evolution, and survival of Helitrons in the maize genome. genome structure inferred from haplotype diversity at the Proceedings of the National Academy of Sciences USA bz locus. Proceedings of the National Academy of Sciences 106: 19922–19927. USA 103: 17644–17649. Yuan Y, SanMiguel PJ and Bennetzen JL (2003) High-Cot sequence Wang X, Elling AA, Li X, Li N, Peng Z, He G, Sun H, Qi Y, Liu XS analysis of the maize genome. Plant Journal 34: 249–255. and Deng XW (2009) Genome-wide and organ-specific Zabeau M and Voss P (1993) Selective restriction fragment landscapes of epigenetic modifications and their relation- amplification: a general method for DNA fingerprinting. ships to mRNA and small RNA transcriptomes in maize. European Patent Application 92402629.7 (publication no. Plant Cell 21: 1053–1069. 0 534 858 A1). q NIAB [2011]. This is a work of the U.S. Government Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 163–165 and is not subject to copyright protection in the United States. doi:10.1017/S1479262111000487 ISSN 1479-2621

Long terminal repeat retrotransposon Jeli provides multiple genetic markers for common wheat (Triticum aestivum)

Nataliya V. Melnikova*, Fedor A. Konovalov and Alexander M. Kudryavtsev Vavilov Institute of General Genetics, Moscow, Russia

Abstract The recombinant inbred line mapping population Opata85 £ Synthetic W7984 was used to map Jeli long terminal repeat retrotransposon insertion sites in the hexaploid wheat genome. Sequence-speciﬁc ampliﬁed polymorphism technique was applied to reveal Jeli insertions. Jeli was found to provide multiple genetic markers for common wheat. Our marker system revealed A-genome Jeli insertions, and therefore can be used for targeted analysis of the A genome.

Keywords: long terminal repeat retrotransposons; molecular mapping; sequence-speciﬁc ampliﬁed polymorphism; Triticum aestivum L.

Introduction Materials and methods

Retrotransposons, also called class I transposable The recombinant inbred line (RIL) mapping population elements, are mobile genetic elements that undergo repli- Opata85 £ Synthetic W7984 was used to map Jeli cative transposition through reverse transcription of RNA LTR retrotransposon (gypsy-like family) insertions in intermediates (Kumar and Bennetzen, 1999). Retrotran- common wheat (Triticum aestivum L.) genome. Total sposons are the major components of cereal genomes DNA was extracted from individual plants by using that are largely located in repetitive DNA (Barakat et al., hexadecyltrimethylammonium bromide protocol (Torres 1997). Polymorphic retrotransposon insertions can be et al., 1993) with minor modification. SSAP analysis used as a molecular marker system for genetic studies was performed, as described by Konovalov et al. (2010). in plant species (Ellis et al., 1998; Queen et al., 2004). Six primer combinations were used to amplify the One of the most popular transposon-based marker DNA sequences between the Jeli LTRs and Taq I methods is the sequence-specific amplified polymorph- restriction sites. The LTR primer sequence 50-CCC- ism (SSAP) technique (Waugh et al., 1997) that was TAGGAACATAGCTTCATCA-30 was based on the Jeli designed to analyse the insertion polymorphism of high sequence TREP3458 obtained from the International copy number long terminal repeat (LTR) retrotranspo- Triticeae Mapping Initiative Triticeae Repeat Sequence sons in plant genomes. SSAP techniques can be powerful Database (Wicker et al., 2002). To reduce the number experimental genomic tools that can be applied to of amplification products, two to three selective bases molecular mapping, marker-assisted selection, diversity were added to the 30 ends of the LTR primers (AC, analysis and evolutionary studies. AG, CTG, GAC, GGA and TC). The PCR reaction products were separated by electrophoresis in a polyacrylamide sequencing gel using a 38 £ 50 cm Bio-Rad SequiGen GT cell and visualised using silver staining. The polymorphic SSAP markers were mapped within * Corresponding author. E-mail: [email protected] a framework of known restriction fragment length 164 N. V. Melnikova et al. polymorphism markers (http://www.graingenes.org/). of 10.8. The polymorphism level (proportion of poly- The marker presence/absence data were analyzed morphic bands in the total number of bands) for Jeli using MAPMAKER 3.0 (Lander et al., 1987). A logarithm observed in our study was 0.18, which was close to of odds score threshold of 5.0 was used to detect the polymorphism level of the markers based on the genetic linkage. Markers that were not assigned into retrotransposons Tar1 and TAGERMINA, but higher linkage groups were discarded. than that of the markers based on Thv19 and BARE- 1/Wis-2-1A elements (Queen et al., 2004). Seventeen of the 65 markers could not be assigned into linkage groups, and therefore, were discarded; thus, 48 markers Results and discussion were placed on the chromosome maps. The number of SSAP markers mapped on each chromosome varied Polymorphic Jeli insertions represented by SSAP bands from zero (chromosomes 1D, 4D, 5D and 7D) to six were scored in a RIL mapping population Opata85 £ (chromosome 2A) markers/chromosome. Synthetic W7984 (Fig. 1). The total number of SSAP The mapping experiments demonstrated approxi- bands for the six primer combinations was 316 with a mately the same general distribution of Jeli markers mean of 52.7 bands/primer combination. The number between the three wheat genomes as the nulli-tetrasomic of polymorphic bands was 65, ranging from 5 (selective analysis had revealed earlier (Konovalov et al., 2010): bases CTG) to 18 (selective bases AG) with a mean 30 insertion sites in the A genome (63%); 14 in the B genome (14%) and 4 in the D genome (4%). The A-genome preference of the Jeli polymorphic insertions was revealed by our marker system. Such an unequal MOS123456789101112131415161718192021 distribution of the retrotransposon-based markers could be explained by a burst of Jeli amplification in the diploid A-genome donor (Triticum urartu Thum. ex Gandil.) with lower retrotransposition activity in the B and D genome donors (Konovalov et al., 2010) that led to a higher copy number of Jeli itself in the A genome; another alternative involves possible differences in the LTR sequences of the Jeli lineages from the A, B and D genomes that could result in preferable primer annealing at the Jeli copies from the polyploid wheat genome A. Tight clustering of Jeli markers was observed in some cases (chromosomes 1B, 2A, 4A, 5A and 6A). Some clustering of BARE-1/Wis-2-1A markers on the linkage map has been shown (Queen et al., 2004). SSAP marker clustering may be because of cases wherein one polymorphic change is scored as two markers (e.g. if an single-nucleotide polymorphism at a restriction site changes the SSAP band length) or by clustering of the retrotransposons themselves, but a more detailed investigation is needed to confirm this hypothesis. The SSAP system based on the Jeli retrotransposon provides multiple genetic markers for common wheat. Our marker system preferably revealed A-genome Jeli insertions, and therefore can be used for targeted analysis of the A genome in evolutionary studies, genetic mapping, polymorphism screening and marker-assisted selection. The number of Jeli insertion sites that can be revealed by different primer/restriction enzyme combinations can be estimated as being at least several Fig. 1. Jeli-based SSAP marker (LTR primer with selective bases AG) segregation in an Opata85 £ Synthetic W-7984 hundreds. It may also be a promising A-genome identi- mapping population. M, 10-bp DNA ladder; O, Opata85; fication tag when used as a probe in fluorescent in situ S, Synthetic W7984. hybridization analyses. Jeli retrotransposons as genetic markers for wheat 165 References computer package for constructing primary genetic linkage maps of experimental and natural populations. Genomics Barakat A, Carels N and Bernardi G (1997) The distribution of 1: 174–181. genes in the genomes of Gramineae. The Proceedings of Queen RA, Gribbon BM, James C, Jack P and Flavell AJ (2004) National Academy of Sciences USA 94: 6857–6861. Retrotransposon-based molecular markers for linkage and Ellis TH, Poyser SJ, Knox MR, Vershinin AV and Ambrose MJ genetic diversity analysis in wheat. Molecular Genetics (1998) Polymorphism of insertion sites of Ty1-copia class and Genomics 271: 91–97. retrotransposons and its use for linkage and diversity anal- Torres AM, Weeden NF and Martin A (1993) Linkage among iso- ysis in pea. Molecular and General Genetics 260: 9–19. zyme, RFLP and RAPD markers in Vicia faba. Theoretical Konovalov FA, Goncharov NP, Goryunova S, Shaturova A, and Applied Genetics 85: 937–945. Proshlyakova T and Kudryavtsev A (2010) Molecular mar- Waugh R, McLean K, Flavell AJ, Pearce SR, Kumar A, Thomas BB kers based on LTR retrotransposons BARE-1 and Jeli uncover and Powell W (1997) Genetic distribution of Bare-1-like different strata of evolutionary relationships in diploid retrotransposable elements in the barley genome revealed wheats. Molecular Genetics and Genomics 283: 551–563. by sequence-specific amplification polymorphisms Kumar A and Bennetzen JL (1999) Plant retrotransposons. (S-SAP). Molecular and General Genetics 253: 687–694. Annual Review of Genetics 33: 479–532. Wicker T, Matthews DE and Keller B (2002) TREP: a database for Lander ES, Green P, Abrahamson J, Barlow A, Daly MJ, Lincoln Triticeae repetitive elements. Trends in Plant Science 7: SE and Newburg L (1987) MAPMAKER: an interactive 561–562. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 166–169 ISSN 1479-2621 doi:10.1017/S1479262111000517

HRM technology for the identiﬁcation and characterization of INDEL and SNP mutations in genes involved in drought and salt tolerance of durum wheat

Linda Mondini1, Miloudi M. Nachit2, Enrico Porceddu1 and Mario A. Pagnotta1* 1Department of Agrobiology and Agrochemistry, University of Tuscia, Via S.C. de Lellis, 01100 Viterbo, Italy and 2ICARDA, PO Box 5466, Aleppo, Syria

Abstract WRKY transcription factors are one of the largest families of transcriptional regulators and form an integral part of signalling webs which modulate many plant processes, such as abiotic stress tolerance. In the present paper, an innovative method has been applied to identify novel WRKY-1 alleles involved in the responses to salt and drought stresses in Triticum durum. This technique involves scanning for sequencing variations in cDNA-derived PCR amplicons, using high-resolution melting (HRM) followed by direct Sanger sequencing of only those amplicons which were predicted to carry nucleotide changes. HRM represents a novel advance in detection of single-nucleotide polymorphisms (SNPs) by measuring temperature-induced strand separation of short PCR amplicons. The use of this approach is still limited in the field of plant biology. Here, HRM analysis has been applied to the discovery and genotyping of durum wheat SNPs. Specific primers have been designed, starting at multi-alignment of WRKY-1-conserved portions. The PCR amplicons, containing single SNPs, produce distinctive HRM profiles, and by sequencing the PCR products identified, SNPs have been characterized and validated. The results showed that all the revealed SNPs are located on salt-tolerant varieties, confirming their value in breeding activities.

Keywords: durum wheat; high-resolution melting; salt tolerance; single-nucleotide polymorphisms

Introduction divided into two groups: the effector molecules and the regulatory molecules (Hasegawa and Bressan, 2000). In Environmental stresses, such as salinity, are the cause of particular, WRKY factors are one of the largest families great losses in crop yields every year all over the globe of transcriptional regulators in plants and form integral (Boyer, 1982). Therefore, salt-responsive or salt-tolerance parts of signalling webs that modulate many plant pro- mechanisms have been intensively studied from biologi- cesses, including the responses to abiotic stresses, such cal and genetic perspectives. Signal transduction path- as drought, cold and salt stresses. The WRKY family is ways regulate reactive oxygen species, damage repair among the ten largest families of transcription factors in and ion homeostasis, while maintaining a low Naþ/Kþ higher plants and is found throughout the green lineage ratio is important for plant survival under salt stress (Rushton et al., 1996). Recent studies have confirmed that (Kader et al., 2006). Molecules that have a function WRKY proteins often act as repressors as well as acti- in the adaptation to environmental stresses can be vators and that the members of the family play a role in both repression and depression of important plant processes (Ishiguro and Nakamura, 1994). Furthermore, it may be that a single WRKY transcription factor is * Corresponding author. E-mail: [email protected] involved in regulating several seemingly disparate HRM for identifying INDEL and SNP mutations 167 processes. The defining feature of WRKY transcription the gene in wild relatives of the species; specific primers factors is their DNA-binding domain, called WRKY after have been designed in order to obtain amplicons not the most unvaried WRKY amino-acid sequence at the N longer than 100 bp analysed by HMR technique. terminus (Rushton et al., 1996). The WRKY domain consists of a four-stranded b-sheet, with the zinc coordinating Cys/His residues forming a zinc-binding pocket, enabling Materials and methods access to the major DNA groove and contacting with the DNA. High-resolution melting (HRM) analysis has been Four durum wheat varieties, Cham I (moderately salt developed to detect single-nucleotide polymorphisms tolerant), Jennah Khetifa (salt tolerant), Belikh 2 (moder- (SNPs) in small PCR amplicons because it is an easy and ately salt tolerant) and Trinakria (salt susceptible) were low-cost method. Up to now, it has been principally used for SNPs discovery and validation in the WRKY-1 used for scanning mutations of large multiple exon gene. Plants were grown with a nutritive solution for genes to identify disease-related mutations in humans 7 d, after which NaCl was added at different concen- (Kennerson et al., 2007). In the area of plant biology, the trations: 0 M (as a negative control), 0.75 and 1.5 M. use of this technique is still limited and has been applied RNA were then extracted from leaf and root material exclusively for constructing linkage map (Chagne´ et al., using TRIzolw reagent (Invitrogen) following the manu- 2007; Croxford et al., 2008). This technique measures tem- facturer’s protocol. The cDNA was synthesized using perature-induced strand separation, and is therefore able oligo(dT)-primers and the SuperScripte III reverse tran- to detect variations as small as one base difference scriptase (Invitrogen), according to the manufacturer’s between samples. The evolution of the HRM technique instructions; the reactions were subsequently treated from traditional melting curve analysis has been with RNaseH (Invitrogen). The cDNA obtained was implemented through the invention of new-generation used in HRM amplifications. Primers employed in HRM intercalating dyes which can saturate-bind to the double- analysis were designed utilizing the software Primer3, stranded DNA without PCR inhibition. In this study, we starting from a multi-alignment of the sequences, which have searched for the presence of SNPs in durum wheat characterize the conserved domain of the WRKY-1 gene varieties with different degrees of tolerance to salt stress. close to WRKY site, in 14 wild relative species. Two Conserved portions of WRKY-1 gene were selected primer pairs were designed to cover a conserved through a multi-alignment of homologue sequences of domain of 250 bp. PCR amplifications were performed

25 Cham I (Nacl 0 M)

15 J. Ketifa (Nacl 1.5 M) Fluorescence 10

81.5 82.0 82.5 83.0 83.5 84.0 84.5 85.0 85.5 86.0 86.5 87.0 87.5 88.0 88.5 89.0 89.5 90.0 Temperature (°C) Fig. 1. WRKY-1 melting curve profiles. 168 L. Mondini et al. on a Rotor-Gene 6000 realtime PCR Thermocycler revealed that eight amplicons showed polymorphic (Corbett Research, Australia). For data quality control, melting curves when assayed against non-treated samples PCR amplification was analysed through the assessment (negative controls), while the remaining 28 were of the threshold cycle (Ct) value, endpoint fluorescence monomorphic (Fig. 1). Nine amplicons, in two replicates, level and the amplification efficiency. The melting data showing both mono- and polymorphic melting curves, were normalized by adjusting the beginning and end and analysed on normalised melting curves and through fluorescence signals to the same level. HRM curve anal- different plots, were chosen, purified and sequenced. ysis was performed using the HRM analysis module. Among these, two SNPs were found in salt-tolerant Different plots of the melting data were visualized by lines Cham I and J. Ketifa both in leaves and roots. The selecting a genotype for comparison and negative first SNPs present in an A/T transversion were located first-derivative melting curves were produced from the close (only one amino acid distant) to the WRKY fluorescence versus temperature plots. Representative domain. The A/T transversions are commonly considered genotypes were chosen for sequencing, using an ABI to be the SNPs variations most difficult to resolve by melt- 3130xl sequencing platform. Sequences obtained were ing analysis. Despite this, the current study showed the blasted with the corresponding sequences present in amplicon with the A/T transversion was distinctly differ- genes databases through BLAST2 Sequences program entiated (Fig. 1). Moreover, this SNP was present in the and multi-aligned to identify mutations. salt-tolerant line, Cham I, even in the NaCl-untreated condition. A second SNP, consisting of a G/C transversion, was located on the other resistant line, J. Ketifa, treated Results and discussions with the maximum concentration (1.5 M) of NaCl. Also, this SNP was located close (only two amino acid distant, A total of 14 sequences relative to a conserved WRKY-1 adjacent to the first SNP) to the WRKY domain. This SNP domain were multi-aligned for primer design. There also showed a distinctive melting curve profile (Fig. 1). were 36 amplicons corresponding to the cDNA extracted Both of the identified SNPs created two corresponding from leaves and roots of the different varieties treated amino acid substitutions, Q/L for the first SNP and K/N with the different salt concentrations. HRM analysis for the second (Q ¼ Gln ¼ Glutamine, L ¼ Leu ¼ Leucine,

Fig. 2. Amino acid and nucleotide sequences of WRKY-1 with WRKY domain (grey box) and the two different SNPs found for Cham I (round box) and J. Ketifa (squarely box). HRM for identifying INDEL and SNP mutations 169 K ¼ Lys ¼ Lysine, N ¼ Asn ¼ Asparagine), both close to RP and Gardiner SE (2007) Mapping a candidate gene the WRKY domain (Fig. 2). The presence of these two (MdMYB10) for red flesh and foliage colour in apple. altered amino acids so close to WRKY may cause a func- BMC Genomics 8: 212. Croxford AE, Rogers T, Caligari PDS and Whilkinson MJ tional modification in the biochemistry of the WRKY-1 (2008) High-resolution melt analysis to identify and map promoter (considering that this protein site is involved sequence-tagged site anchor points onto linkage maps: in DNA binding) that could explain and contribute to a white lupin (Lupinus albus) map as an exemplar. New the resistance of these two durum wheat lines to salt Phytologist 180: 594–607. stress. The presence of the first SNP in the Cham I line, Hasegawa PM and Bressan RE (2000) Plant cellular and mole- in the untreated condition, may be explained by cular responses to high salinity. Annual Review of Plant Physiology and Plant Molecular Biology 51: 463–499. WRKY-1 being part of one of the largest families of tran- Ishiguro S and Nakamura K (1994) Characterization of a cDNA scription factors. These factors are involved in regulating encoding a novel DNA-binding protein, SPF1, that recog- several seemingly disparate processes where some pro- nizes SP8 sequences in the 50 upstream regions of genes teins act as repressors as well activators (Rushton et al., coding for sporamin and beta-amylase from sweet potato. 2010). Furthermore, understanding the different mechan- Molecular and General Genetics 244: 563–571. isms of WRKY transcription factors is straightforward, as Kader MA, Seidel T, Golldack D and Lindberg S (2006) Expressions of OsHKT1, OsHKT2, and OsVHA are diffe- reported by Ishiguro and Nakamura (1994) and Rushton rentially regulated under NaCl stress in salt-sensitive and et al. (2010). Nevertheless, the localization of both SNPs salt-tolerant rice (Oryza sativa L.) cultivars. Journal of in both the resistant lines confirms their value for future Experimental Botany 57: 4257–4268. breeding activities. In addition, this work confirms that Kennerson L, Warburton T, Nelis E, Brewer M, Polly P, De HRM represents a feasible means for SNP detection and Jonghe P, Timmerman V and Nicholson GA (2007) genotyping, being a simple, accurate, high-throughput Mutation scanning the GJB1 gene with high-resolution and low-cost method. melting analysis: implication for mutation scanning of genes for Charcot-Marie-Tooth disease. Clinical Chemistry 53: 349–352. Rushton PJ, Torres JT, Parniske M, Wernert P, Hahlbrock K References and Somssich IE (1996) Interaction of elicitor-induced DNA binding proteins with elicitor response elements in Boyer JS (1982) Plant productivity and environment. Science the promoters of parsley PR1 genes. EMBO Journal 15: 218: 443–448. 5690–5700. Chagne´ D, Carlisle CM, Blond C, Volz RK, Whitworth CJ, Rushton PJ, Somssich IE, Ringler P and Shen QJ (2010) WRKY Oraguzie NC, Crowhurst RN, Allan AC, Espley RV, Hellens transcription factors. Trends in Plant Science: 15: 247–258. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 170–173 ISSN 1479-2621 doi:10.1017/S147926211100044X

Starch metabolism mutants in barley: A TILLING approach

Riccardo Bovina1, Valentina Talame`1, Salvi Silvio1, Maria Corinna Sanguineti1, Paolo Trost2, Francesca Sparla2 and Roberto Tuberosa1* 1Department of Agroenvironmental Science and Technology, University of Bologna, Bologna, Italy and 2Department of Experimental Evolutionary Biology, University of Bologna, Bologna, Italy

Abstract In this study, the targetting-induced local lesions in genomes approach was used to identify mutants for genes related to starch metabolism in barley. Starch is the major reserve of plants and serves as primary carbohydrate component in human and livestock diets and has also numerous industrial applications. Mutants for biosynthetic or regulatory genes of starch metabolism often produce starch granules with abnormal morphological and molecular features that could be of interest for technological applications. We report the identiﬁcation of 29 mutations in ﬁve starch-related barley genes (Bmy1, GBSSI, LDA1, SSI and SSII) through the molecular screening of TILLMore, a sodium azide-mutagenized population. Almost all the mutations detected were CG–TA transitions and several (c. 60%) implied a change in amino-acid sequence and therefore possible phenotypic effects. Four mutants showed non-sense or splice-junction alterations, which could drastically affect the protein function.

Keywords: Hordeum vulgare; reverse-genetics; starch; TILLING

Introduction technology is strong. Although the genetic and physiological bases of starch biosynthesis in plants are well Starch is the major reserve of plants and the major source known, the regulatory machinery controlling the for- of food calories for humans, as well as an important raw mation of the complex and ordered structure of the material for the food and processing industries (James starch granule is still not fully understood. Mechanisms et al., 2003). The major component of starch granules is of post-translational regulation are likely to play a major amylopectin, which forms partially crystalline structures, role in starch metabolism (Michalska et al., 2009; Valerio while amylose constitutes the amorphous portion of the et al., 2011; Zeeman et al., 2010). Mutants for biosynthetic granule (Hizukuri, 1996; Lemke et al., 2004). The differ- or regulatory genes of starch metabolism often produce ent molecular features of starch polymers (i.e. chain starch granules with abnormal morphological and mol- length, frequency of branching, abundance of amylose, ecular features (Sehnke et al., 2001; Asano et al., 2002). etc.) influence both the morphology of the granule and We describe the utilization of TILLMore (http://www. the technological properties of starch as a raw material distagenomics.unibo.it/TILLMore/), a barley targeting- or foodstuff (Jobling, 2004). The European starch induced local lesions in genomes (TILLING) resource market is substantial and the interest of both the scientific (Talame` et al., 2008) to identify new alleles involved in community and industry in starch biosynthesis and starch biosynthesis and degradation in seeds. TILLING has already been successfully applied to identify starch mutants in wheat (Slade et al., 2005). The long-term goal of this research is the identification of barley mutants * Corresponding author. E-mail: [email protected] with starch granules of peculiar morphological, structural Starch metabolism mutants in barley 171 and molecular features eventually leading to novel tech- and Primer3 (http://frodo.wi.mit.edu/cgi-bin/primer3/ nological properties. primer3_www.cgi). The CODDLE program is able to identify regions where point mutations are most likely to result in deleterious effects on the gene’s function (Till et al., Materials and methods 2003). The following primer sequences were used for PCR and sequencing of the population and the putative For the TILLING screening, five genes involved in starch mutants: metabolism were selected based also on the genomic HvBMY1_For, TTTGCCTTCCGGGAGACCATGT; sequence availability in barley cv. ‘Morex’, provided by HvBMY1_Rev, CGCGTTTTCGGATGCCACATTT; Dr. Edward Schiefelbein (University of Minnesota, St. Paul, HvGBSSI_For, GAGCACCCAGCCACCCACACA; MN, USA). The genes chosen for the analysis are: b-amylase HvGBSSI_Rev, CTGCAGCATACGCCCAGACCA; 1 (HvBMY1 accession no. EF175470), granule-bound starch HvLDA1_For, CTCGTGTGCAGCTGACGGGAAA; synthase I (HvGBSSI accession no. AB089162), limit dextri- HvLDA1_Rev, GTGCCATCGTGGGCGCTGTAAT; nase 1 (HvLDA1 accession no. AF122050), starch synthase HvSSI_For, TGTCGCGTTCCCCATTCTGATA; I (HvSSI accession no. AF234163) and starch synthase II HvSSI_Rev, TGGCATGGCTACAGTTCACCAAGC; (HvSSII accession no. AY133250). Primers were designed HvSSII_For, CCGATTCGATGTATGCCGGCAAT; with Codons Optimized to Discover Deleterious LEsions HvSSII_Rev, CCAGATCGGAATCAGCGTCTCA. (CODDLE; http://www.proweb.org/coddle), a tool faci- The TILLING analyses were implemented using the litating the selection of gene regions for TILLING purposes, procedure described by McCallum et al. (2000); DNA

Table 1. Details of the mutants identiﬁed in the ﬁve genes analyzed

Nucleotidic Effect Aminoacid Gene Plant code substitution Position of mutation substitution PSSMa SIFTb HvBMY1 1513 C to T Exon Silent Y to Y – – 2253 G to A Exon Missense D to N 10.2 0.30 2682 G to A Exon Missense E to K – – HvGBSSI 94 A to T Intron – – – – 570 T to A Intron – – – – 1090 G to A Exon Missense G to E 20.5 0.00 2209 C to A Intron – – – – 2733 A to T Intron – – – – 5214 G to C Intron – – – – HvLDA1 250 G to A Exon Missense R to K – – 905 G to A Exon Missense V to I – – 1020 C to T Exon Missense S to F 8.5 0.02 1139 C to T Exon Missense T to I – – 1317 G to A Intron – – – – 1550 C to T Exon Missense P to S – – 1696 C to T Intron – – – – HvSSI 662 G to A Exon Non-sense W to SCc –– 877 C to T Intron – – – – 1089 G to A Splice jun. – – – – 1132 C to T Exon Missense T to I – – 1284 G to A Exon Missense G to E – – 1808 G to A Intron – – – – 1963 C to T Intron – – – – 2822 G to A Splice jun. – – – – 5758 G to A Intron – – – – 5850 G to A Exon Missense G to D – – HvSSII 1039 G to A Exon Missense G to R – – 1517 G to A Exon Non-sense W to SC –– 2273 C to T Exon Silent L to L – – jun, junction. a,b PSSM and SIFT are values used to rank the missense mutations on the basis of their probability of affecting protein function. PSSM and SIFT can be calculated only if the missense mutation occurs in a conserved domain of the protein. Mutations are considered to be deleterious for PSSM values above 10 and SIFT values below 0.10. c Stop codon. 172 R. Bovina et al. samples of eight M3 lines were pooled and subjected to alleles, causing changes in one of the aminoacids of the gene-specific PCR amplification using properly designed protein. In four cases, non-sense alleles (two truncation labelled-primers (MWG-Biotech). The PCR reaction and mutations and two splice junction mutations) were cycling were performed as described in Colbert et al. identified. All the non-sense mutations occurred in (2001). The PCR products were then digested with a com- starch synthase I and II, two genes with a crucial role mercial endonuclease, the Surveyorw Mutation Detection in the elongation of the amylopectin chains. Severe Kit (Transgenomics, Omaha, NE, USA), according to the mutations in these genes are expected to drastically manufacturer’s directions. The digested PCR products reduce the content of amylopectin, hence conferring a were analyzed using a detection method based on clear phenotype (Umemoto et al., 2002; Fujita et al., denaturing electrophoretic gels (LI-COR-4200; LI-COR 2006; Sestili et al., 2009). As to the missense mutations, Biosciences; Lincoln, NE, USA). The final validation of identified for all the genes analyzed, bioinformatic tools the results was performed by sequencing using an were applied to estimate the impact of mutations on Applied Biosystems’ 377 DNA Analyzer (Applied Bio- protein function. In particular, PARSESNP and SIFT pro- systems, Forster City, USA). Finally, the sequences were grams were utilized to identify the mutations that more analyzed with the PARSESNP (Taylor and Greene, 2003) likely will have a deleterious effect on protein function. and sorting intolerant from tolerant (SIFT) (Ng and In our case, the mutations GBSSI 1090 and BMY1 2253 Henikoff, 2003) programs. showed position specific scoring matrix (PSSM) values of 20.5 and 10.2, respectively (mutations are considered to be deleterious for PSSM values above 10). The Results and discussion application of the SIFT algorithm predicted a possible deleterious effect for the mutations GBSSI 1090 (SIFT The molecular screening was achieved on five genes 0.0) and LDA1 1020 (SIFT 0.02). Mutations are predic- involved in starch metabolism, using a cell-based hetero- ted to be deleterious for SIFT values below 0.05 or even duplex assay, coupled with gel electrophoresis on DNA below 0.10. Since all other missense mutations were pre- sequencers. A total number of 4906 DNA samples from dicted to be located outside conserved domains, PSSM individual M3 plants were screened. The analyses ident- and SIFT values could not be calculated. ified an allelic series for each of the genes examined In conclusion, we detected at least one interesting with a total number of 29 mutations and an average of allele for all the five genes analyzed in our study. These c. five mutations/gene (Table 1). The estimated mutation findings provide valuable genetic materials for studies density was of one mutation/520 kb screened, which on the regulation of starch biosynthesis in barley and compares well with what was previously reported by for applications in mutation breeding. Talame` et al. (2008) on the same collection. The value of the mutation density was computed by dividing the total number of identified mutations by the number of base pairs screened and corrected, considering the effec- Acknowledgements tive screened window. In fact, a limitation of the TILLING procedure is that mutations can escape identification The financial support of the University of Bologna (Stra- when present in the terminal 80 bp of both ends of the tegic project ‘Starchitecture’) is gratefully acknowledged. amplicon as a result of PCR priming and electrophoresis artifacts. In our case, a correction on the effective screening window was applied by subtracting 160 bp from the References length of each amplicon (Greene et al., 2003). Almost all the mutations detected were G/C to A/T Asano T, Kunieda N, Omura Y, Ibe H, Kawasaki T, Takano M, transitions. Since a previous study proposed that NaN3 Sato M, Furuhashi H, Mujin T, Takaiwa F, Wu C, Tada Y, causes mutations of transition type (Olsen et al., Satozawa T, Sakamoto M and Shimada H (2002) Rice 1993) and because almost all of our mutations were SPK, a calmodulin-like domain protein kinase, is required for storage product accumulation during seed develop- G/C to A/T transitions, the possibility that the poly- ment. The Plant Cell 14: 619–628. morphisms identified in TILLMore are naturally occur- Colbert T, Till BJ, Tompa R, Reynolds S, Steine MN, Yeung AT, ring as a result of seed contamination of our starting McCallum CM, Comai L and Henikoff S (2001) High- ‘Morex’ seed stock can be ruled out. throughput screening for induced point mutations. Plant Among the 29 alleles, 13 silent mutations occurred in Physiology 126: 480–484. Fujita N, Yoshida M, Asakura N, Ohdan T, Miyao A, Hirochika H non-coding regions or affected the third base of a and Nakamura Y (2006) Function and characterization of codon which does not change the aminoacid encoded starch synthase I using mutants in rice. Plant Physiology by that codon; 12 mutations were classified as missense 140: 1070–1084. Starch metabolism mutants in barley 173 Greene EA, Codomo CA, Taylor NE, Henikoff JG, Till BJ, Sestili F, Botticella E, Bedo Z, Phillips A and Lafiandra D (2009) Reynolds SH, Enns LC, Burtner C, Johnson JE, Odden AR, Production of novel allelic variation for genes involved in Comai L and Henikoff S (2003) Spectrum of chemically starch biosynthesis through mutagenesis. Molecular Breed- induced mutations from a large-scale reverse-genetic ing 25: 145–154. screen in Arabidopsis. Genetics 164: 731–740. Slade AJ, Fuerstenberg SI, Loeffler D, Steine MN and Facciotti D Hizukuri S (1996) Starch: analytical aspects. In: Eliasson AC (ed.) (2005) A reverse genetic, nontransgenic approach to wheat Carbohydrates in Food. New York: Dekker, pp. 347–429. crop improvement by TILLING. Nature Biotechnology 23: James MG, Denyerz K and Myers AM (2003) Starch synthesis in 75–81. the cereal endosperm. Current Opinion in Plant Biology 6: Talame` V, Bovina R, Sanguineti MC, Tuberosa R, Lundqvist U 215–222. and Salvi S (2008) TILLMore, a resource for the discovery Jobling S (2004) Improving starch for food and industrial applications. Current Opinion in Plant Biology 7: 210–218. of chemically induced mutants in barley. Plant Biotechnol- Lemke H, Burghammer M, Flot D, Ro¨ssle M and Riekel C (2004) ogy Journal 6: 477–485. Structural processes during starch granules hydration by Taylor NE and Greene EA (2003) PARSESNP: A tool for the anal- synchrotron radiation microdiffraction. Biomacromolecules ysis of nucleotide polymorphisms. Nucleic Acids Research 5: 1316–1324. 31: 3808–3811. McCallum CM, Comai L, Greene EA and Henikoff S (2000) Tar- Till BJ, Reynolds SH, Greene EA, Codomo CA, Enns LC, Johnson geting Induced Local Lesion IN Genomes (TILLING) for JE, Burtner C, Odden AR, Young K, Taylor NE, Henikoff JG, plant functional genomics. Plant Physiology 123: 439–442. Comai L and Henikoff S (2003) Large-scale discovery of Michalska J, Zauber H, Buchanan BB, Cejudo FJ and Geigenber- induced point mutations with high-throughput TILLING. ger P (2009) NTRC links built-in thioredoxin to light and Genome Research 13: 524–530. sucrose in regulating starch synthesis in chloroplasts Umemoto T, Yano M, Satoh H, Shomura A and Nakamura Y and amyloplasts. Proceedings of the National Academy of (2002) Mapping of a gene responsible for the difference Sciences USA 106: 9908–9913. in amylopectin structure between japonica-type and Ng PC and Henikoff S (2003) SIFT: Predicting amino acid indica-type rice varieties. Theoretical and Applied Genetics changes that affect protein function. Nucleic Acids 104: 1–8. Research 31: 3812–3814. Valerio C, Costa A, Marri L, Issakidis-Bourguet E, Pupillo P, Olsen O, Wang X and von Wettstein D (1993) Sodium azide mutagenesis: preferential generation of AT/GC transition Trost P and Sparla F (2011) Thioredoxin-regulated {beta}- in the barley Ant18 gene. Proceedings of the National amylase (BAM1) triggers diurnal starch degradation in Academy of Sciences USA 90: 8043–8047. guard cells, and in mesophyll cells under osmotic stress. Sehnke PC, Chung HJ, Wu K and Ferl RJ (2001) Regulation of Journal Experimental Botany 62: 545–555. starch accumulation by granule-associated plant 14-3-3 Zeeman SC, Kossmann J and Smith AM (2010) Starch: its metab- proteins. Proceedings of the National Academy of Sciences olism, evolution, and biotechnological modification in USA 98: 765–770. plants. Annual Review of Plant Biology 61: 209–234. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 174–176 ISSN 1479-2621 doi:10.1017/S1479262111000165

Collection of mutants for functional genomics in the legume Medicago truncatula

O. Calderini1*, M. Carelli2, F. Panara1, E. Biazzi2, C. Scotti2,A.Tava2, A. Porceddu3 and S. Arcioni1 1CNR – Istituto di Genetica Vegetale, Perugia, Italy, 2C.R.A. – Centro per le Produzioni Foraggere e Lattiero-Casearie, Lodi, Italy and 3Dipartimento Di Scienze Agronomiche e Genetica Vegetale Agraria, Universita` degli Studi di Sassari, Sassari, Italy

Abstract We have established mutant collections of the model species Medicago truncatula according to current protocols. In particular, we used a transposon (Tnt1) tagging method and an ethyl methanesulfonate (EMS) mutagenesis approach (TILLING). The collections were subjected to both forward and reverse genetics screenings, and several mutants were isolated that affect plant traits (e.g. shoot, root developments, ﬂower morphology, etc.) and also biosynthetic pathways of secondary compounds (saponins and tannins). Genes responsible for some of the mutations were cloned and further characterized.

Keywords: Medicago truncatula; mutants; Transposon tagging; TILLING

Introduction transposons such as Tnt1. Several public resources are being developed both in Europe and in the USA, and Legume species are a major source of protein for both some are available for public screening (Tadege et al., human and animal nutrition. Proteins from legumes are 2009). In the frame of an Italian functional genomics highly sustainable compared with those obtained from project ‘Post-genomics of forage legumes’ (MIUR-FIRB), other sources as, among other reasons, legume plants we developed three complementary mutant collections have positive impact on soil fertility due to their ability to that can be an integration to the resources already avail- ﬁx nitrogen via rhizobia symbioses. Functional genomics able at the international level (Porceddu et al., 2008); we in model species such as Medicago truncatula offers the have also produced several hundreds of transposon possibility to speed up the knowledge of the genetic con- tagged lines as part of an European collection supported trol of traits of agronomic interest and translate such by the EU FP6 ‘Grain Legumes Integrated Project’. knowledge to crop legumes via different means (Cannon et al., 2009). M. truncatula was chosen as a model for legumes because it is diploid, with relatively small Materials and methods genome (450–550 Mb), self-fertile, relatively easy to trans- form, and it has a short generation time. Mutants in Production of the collections M. truncatula are currently obtained by three main strategies: (1) chemical mutagenesis mainly based on EMS Transposon tagging (Tnt1) and the subsequent reverse genetic screening TILLING; A starter line from the R108-1 genotype harbouring (2) deletion mutagenesis based on fast neutron bombard- three Tnt1 insertions was produced via Agrobacterium ment and g-rays; (3) insertional mutagenesis based on transformation with the construct Tnk23. Transposition was induced in vitro according to d’Erfurth et al. (2003). Mutant lines were also obtained in the Jemalong background using starter lines as reported in Iantcheva * Corresponding author. E-mail: [email protected] et al. (2009). Mutant collections for functional genomics 175 (a) (b) (c)

Fig. 1. Phenotypic variation among Tnt1 lines: (a) lack of anthocyanins in the leaves, (b) no tannin staining in glandular trichomes and (c) seeds.

Activation tagging (FSTs) were recovered from 13 R108 mutant plants. Transfer-deoxyribonucleic acid (T-DNA) transformed BLAST analysis of the FSTs showed that at least 47% lines were obtained in the background R108-1 with the of insertions are inside genes, 27% are in sequenced pSKI074 vector (Weigel et al., 2000). but not yet annotated M. truncatula BAC clones and 25% do not show similarity with any sequence in the TILLING database. This preliminary molecular analysis of M. truncatula seeds from cv. Jemalong genotype the Tnt1 insertion sites confirms the data from Tadege 2HA10–9–3 were treated with 0.15% EMS to generate a et al. (2008); in fact in a larger scale experiment, these mutant collection. M2 seeds were collected from 2281 authors reported an overall FSTs match with M1 individuals together with 65 tester plants from the M. truncatula sequences in the database of 78.6, 60.2% control treatment (0% EMS). DNA was extracted from of which having high homology with a range of known almost all the M2 families, and M3 seed was collected; genes. The rate of transposition in R108 was as high about 1900 plants are present in the final EMS mutant as reported in the literature (,80%). Differently in the collection (DNA and M3 seed). Jemalong background, we noticed a lower transposition efficiency (60%), which was also reported in Iantcheva et al. (2009). Results and discussion Tnt1 R0 lines were visually screened during multipli- cation, and several mutants were identified concerning Tnt1 tagging leaf, flower and root morphology. A forward screening was conducted on the content of condensed tannins in We have produced approximately 1000 R0 lines in the the leaves, which was also extended to 2000 lines from R108 background and 1000 lines in the Jemalong back- the collection at the Samuel Roberts Noble Foundation ground; all the lines are stored as legumes/seeds. In a (Tadege et al., 2008). Even in this case, several mutants small scale experiment, 96 flanking sequence tags were identified, namely two main phenotypic classes

Mutant (a)15 (b) Pi treatment Po treatment 12

9 Po treatment Pi treatment 6 mU/g root FW

0 Mutant wt wt Fig. 2. Phytase activity in the mutant MtPHY1598: (a) enzymatic assay (ﬁve root bulks of ten plants/treatment); (b) histochem- ical assay on intact roots. FW, fresh weight; Pi, inorganic phosphorous; Po, organic phosphorous; wt, wild type. 176 O. Calderini et al. could be detected: plants lacking both anthocyanins genes. The genetic information achieved can be easily and tannins and plants lacking anthocyanins but pro- translated to alfalfa, a closely related crop to M. trunca- ducing tannins (Fig. 1). For several of the mutants tula and one of the major forage crop worldwide and mentioned, it was possible to isolate a candidate gene possibly used in breeding programmes. based on FST recovery and segregation analysis.

Activation tagging References

A small population (128) of activation tagged lines was Ane` JM, Zhu H and Frugoli J (2008) Recent advances in produced in R108 as an initial test of transformation. Medicago truncatula genomics. International Journal of The population was screened for the presence of haemo- Plant Genomics 2008: 256597. Balestrazzi A, Gonfalonieri M, Odoardi M, Ressegotti V, Allegro lytic saponins in the leaves, and a loss-of-function mutant G, Tava A and Carbonera D (2004) A trypsin inhibitor was identified. The corresponding gene was cloned and cDNA from a novel source, snail medic (Medicago characterized (Carelli et al., unpublished). scutellata L.): cloning and functional expression in response to wounding, herbivore, jasmonic and salicylic acid. Plant Science 167: 337–346. TILLING Cannon SB, May GD and Jackson SA (2009) Three sequenced legume genomes and many crop species: rich opportunities for translational genomics. Plant Physiology 151: The TILLING collection is based on the initial treatment 970–977. of M. truncatula seeds from cv. Jemalong genotype d’Erfurth I, Cosson V, Eschstruth A, Lucas H, Kondorosi A and 2HA10–9–3 with 0.15% EMS. The M2 generation (1658 Ratet P (2003) Efficient transposition of the Tnt1 tobacco families, represented by 1–5 plants/family for a total of retrotransposon in the model legume Medicago truncatula. Plant Journal 34: 95–106. 2560 plants) and 34 tester families were grown in a Iantcheva A, Chabaud M, Cosson V, Barascud M, Schutz B, cold greenhouse and phenotypically screened for identi- Primard-Brisset C, Durand P, Barker DG, Vlahova M and fying mutants. DNA was extracted from almost all the M2 Ratet P (2009) Osmotic shock improves Tnt1 transposition families, and M3 seed has been collected; about 1900 frequency in Medicago truncatula cv Jemalong during in plants are present in the final EMS mutant collection vitro regeneration. Plant Cell Reports 28: 1563–1572. Porceddu A, Panara F, Calderini O, Molinari L, Taviani P, Lanfa- (DNA and M3 seed). TILLING analysis was performed loni L, Scotti C, Carelli M, Scaramelli L, Bruschi G, Cosson in collaboration with the Genomic Platform of the V, Ratet P, de Larembergue H, Duc G, Piano E and Arcioni Parco Tecnologico Padano (Lodi, Italy); the estimated S (2008) An Italian functional genomic resource for rate of mutation was one mutation/Kbp/400 plants simi- Medicago truncatula. BMC Research Notes 1: 129. lar to the collection generated by the R. Cook laboratory Tadege M, Wen J, He J, Tu H, Kwak Y, Eschstruth A, Cayrel A, Endre G, Zhao PX, Chabaud M, Ratet P and Mysore KS as reported by Ane` et al. (2008). It is worth of mention (2008) Large-scale insertional mutagenesis using the Tnt1 that two alleles were recovered for the candidate gene retrotransposon in the model legume Medicago truncatula. related to the absence of haemolytic saponin phenotype; The Plant Journal 54: 335–347. the mentioned mutants lack haemolytic saponins in the Tadege M, Wang TL, Wen J, Ratet P and Mysore KS (2009) leaves (Carelli et al., unpublished). Two more genes Mutagenesis and beyond(Tools for understanding legume were analysed by TILLING: the extracellular phytase biology. Plant Physiology 151: 978–984. Weigel D, Ahn JH, Bla´zquez MA, Borevitz JO, Christensen SK, MtPHY1 (Xiao et al., 2005) and the trypsin inhibitor Fankhauser C, Ferrańdiz C, Kardailsky I, Malancharuvil MsTI (Balestrazzi et al., 2004). Interesting results were EJ, Neff MM, Nguyen JT, Sato S, Wang ZY, Xia Y, Dixon obtained with an MtPHY mutant (MtPHY1598) that RA, Harrison MJ, Lamb CJ, Yanofsky MF and Chory J showed an enhanced phytase activity when grown on (2000) Activation tagging in Arabidopsis. Plant Physiology organic phosphorus (Fig. 2). 122: 1003–1014. Xiao K, Harrison MJ and Wang ZY (2005) Transgenic expression The collections reported in the present paper have of a novel M. truncatula phytase gene results in improved proved to be efficient in isolating mutants related to acquisition of organic phosphorus by Arabidopsis. Planta relevant agricultural traits and in cloning the underlying 222: 27–36. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 177–180 ISSN 1479-2621 doi:10.1017/S1479262111000591

Polymorphisms in intron 1 of carrot AOX2b – a useful tool to develop a functional marker?

He´lia Cardoso1, Maria Doroteia Campos1, Thomas Nothnagel2 and Birgit Arnholdt-Schmitt1* 1EU Marie Curie Chair – Laboratory of Molecular Biology, ICAAM, University of E´vora, Apartado 94, 7002-554 E´vora, Portugal and 2Federal Center of Breeding Research on Cultivated Plants, Institute for Breeding Research on Horticultural and Fruit Crops, Neuer Weg 22/23, D-06484 Quedlinburg, Germany

Abstract Alternative oxidase (AOX) has been proposed as a promising functional marker candidate for multiple plant stress behaviour. The present paper describes natural polymorphism in AOX2b of Daucus carota L. (DcAOX2b). Exon-primed intron crossing-PCR (EPIC-PCR) revealed length variation (intron length polymorphisms, ILPs) in intron 1. Six fragment patterns were identiﬁed in 40 genotypes. However, no more than two fragments were found per genotype, suggesting the presence of two alleles. The ILPs were able to discriminate between single plant genotypes in cv. Rotin and to distinguish individual wild carrot plants. The repetitive pattern of intron 1 length variation allows the grouping of genotypes for functional analysis in future studies. Sequence analysis in intron 1 of polymorphic but also of obviously identical PCR-fragments revealed underlying high levels of sequence polymorphisms between alleles and genotypes. Variation was due to repetitive insertion/deletion (InDel) events and single-nucleotide polymorphisms (SNPs). The results suggest that high AOX2b gene diversity in D. carota may be a source of functional markers for agronomic traits related to environmental stress responses.

Keywords: Daucus carota; exon-primed intron crossing-PCR; insertion/deletion; intron length polymorphism; miRNA; single-nucleotide polymorphism

Introduction in intron sequences (Wang et al., 2005). These intronic polymorphisms are becoming important genetic markers Until the 1970s, eukaryotic introns were considered to be for genetic research and breeding in major crop species genetically inert sequences with no function in the genome (Bi et al., 2006; Braglia et al., 2010). and not subject to evolutionary selection. They were The recent identiﬁcation of intron variation in carrot known as dubbed selﬁsh or junk DNA (Orgel and AOX2a (Cardoso et al., 2009) suggested the investigation Crick, 1980). However, recent evidences show that plant of intron variation in other carrot AOX genes (Campos introns play an active role in the control of gene et al., 2009; Costa et al., 2009a). Alternative oxidase expression (Gianı` et al., 2003; Fiume et al., 2004) and (AOX) is an inner mitochondrial membrane protein evolution. Insertions/deletions (InDels) and single- involved in stress acclimation and adaptation of diverse nucleotide polymorphisms (SNPs) are the two main nature (e.g. McDonald and Vanlerberghe, 2006; Plaxton types of DNA polymorphism (Nasu et al., 2002) found and Podesta´, 2006). We describe here the intron variation in carrot AOX2b (DcAOX2b) among genotypes, with a view to potential applications in carrot breeding (Arnholdt-Schmitt et al., 2006; Arnholdt-Schmitt, 2009; * Corresponding author. E-mail: [email protected] Polidoros et al., 2009). 178 H. Cardoso et al.

Table 1. Oligonucleotides applied in the present work

Ta Te Aim Olinucleotides sequence Enzyme/kit (oC) (min) ILPs in intron 1 Fw: 50-TGGGGACAAGGATGATGAG-30 Ready-to-go PCR beads 55 1 Rev: 50-CCCTTAGGTTTATGGTGTTTATTC-30 (GE Healthcare, Little Chalfont, England) ILPs in intron 2 Fw: 50-TGATCTGAATAAACACCATAAACC-30 Ready-to-go PCR beads 55 1 Rev: 50-GAAGAAAGCATTGAAGAAAACTCC-30 (GE Healthcare) ILPs in intron 3 Fw: 50-TTTAGCCGTGCAGGGAGTTTTCTT-30 Ready-to-go PCR beads 60 1 Rev: 50-GCATCCCTCAGTTCCTTTCCTTCA-30 (GE Healthcare) Sequence Fw: 50-TGCATGCGTCCTTCCTTATTTTTC-30 Phusione high-ﬁdelity DNA polymerase 55 2 polymorphisms Rev: 50-GCTCTGCTGTGATTTTCTGGAC-30 (Finnzymes Oy, Espoo, Finland) o Fw, forward; Rev, reverse; Ta, annealing temperature; Te, extension time at 72 C for 35 cycles.

Materials and methods PCR-fragment pattern related to the single 1 kb (b) band was found more frequently than other patterns In a search for intron length polymorphisms (ILPs) in the (Fig. 1). No more than two fragments were observed in three introns of DcAOX2b (see gene structure in Campos the same individual plant, consistent with biallelic et al., 2009), exon-primed intron crossing-PCR (EPIC- nature in a diploid species. Breviario et al. (2007), PCR) was performed using 10 ng DNA of 40 Daucus using ILPs as molecular markers, described a link carota L. genotypes (14 of cv. Rotin and 26 of wild between the number of amplified bands and ploidy carrot). Specific primers were designed in the exons level of the taxon. ILPs were also reported in the other flanking each intron (Table 1). To avoid the occurrence member of the AOX2-subfamily in D. carota (Cardoso of heteroduplex fragments, electrophoresis was per- et al., 2009) and in the AOX1-subfamily of other plant formed in denaturing conditions in 3-(N-morpholino) species (Naydenov et al., 2005; Costa et al., 2009b; Fer- propanesulphonic acid buffer (BDH, Haasrode, Belgium) reira et al., 2009). supplied with 2% formaldehyde (Sigma-Aldrich, Stein- The length of inton 1 of DcAOX2b was reported as heim, Germany). 822 bp (Cardoso et al., 2009). The present extended anal- For analysis of intron variation in the DcAOX2b ysis reveals that intron 1 has at least three length variants sequence, the gene was isolated from six genotypes of of 1019 (a), 822 (b) and 557 bp (c). cv. Rotin using a forward primer located in the 50-untrans- The existence of SNPs and InDels was confirmed by lated region (UTR) and a reverse in the 30UTR (Table 1). sequencing. Both types of polymorphisms are known For sequence analysis, EditSeq and the ClustalW algorithm of Megalign (Lasergene 7; GATC Biotech AG, Kon- stanz, Germany) were applied. Putative miRNA Heterozygous Homozygous precursors (pre-miRNAs) located in intron sequences MMab ac bc a b c MM were searched and validated as described in Cardoso bp bp et al. (2009). 1500 1500

1000 1000

Results and discussion 500 500

EPIC-PCR fragments from intron 2 and intron 3 show Fig. 1. Agarose gel showing the six-band patterns identified lengths consistently around 430 or 400 bp in all carrot by EPIC-PCR for intron 1 of DcAOX2b. From the homozy- genotypes analyzed. Sequence analysis yielded a size gote genotypes, 40% demonstrated fragment b (seven geno- for intron 2 of 91 bp and for intron 3 of 85 bp. All types of cv. Rotin and nine of wild genotypes), fragment a was detected once in Rotin and also in wild genotypes, intron 2 sequences were identical. In intron 3 sequences, fragment c was not detected in cv. Rotin, but was present four putative SNPs were identified. in four genotypes of wild carrot. From the heterozygous ILP was found in intron 1 in the genome of individual genotypes, the combination of fragments bc was more fre- plants and between plant genotypes. Six polymorphic quent than other combinations (22%: two genotypes in cv. patterns were identified: three single-band patterns Rotin and seven in the wild genotypes), the combination ab amplified in three genotypes of cv. Rotin and wild (homozygous) named fragments a, b and c, and three carrots and ac appeared once in cv. Rotin and twice in wild double-band patterns (heterozygous) named ab, ac carrot plants. MM: molecular marker (O’Range Ruler and bc (Fig. 1). In cv. Rotin and in wild carrot, the 100bp þ 500bp Ladder; Fermentas, Ontario, Canada). Polymorphisms in intron 1 of carrot AOX2b 179 as two of the major driving forces in genome evolution References (e.g. Gregory, 2004). Twelve putative SNPs were identified between the intron 1 sequences of fragment a and Arnholdt-Schmitt (2009) Alternative oxidase (AOX) and stress b (the most similar fragments). When fragment c was tolerance – approaching a scientific hypothesis. Physiologia included in the alignment, 64 additional putative SNPs Plantarum 137: 314–315. Arnholdt-Schmitt B, Costa JH and Fernandes de Melo D (2006) were identified. Four of them were specific to fragment AOX – a functional marker for efficient cell reprogram- a (427:T/C, 611:A/C, 852:T/C and 1061:T/C). ming under stress? Trends in Plant Science 11: 281–287. Three InDels were identified by alignment of frag- Bi McMullenMD IV, Sanchez-Villeda H, Schroedor S, Gardiner J, ments a and b. Integrating the 557 bp intron sequence Polacco M, Soderlund C, Wing R, Fang Z and Coe EH into the alignment allowed the identification of nine (2006) Single-nucleotide polymorphisms and insertion– deletions for genetic markers and anchoring the maize new InDels. In total, 12 InDels were identified, three fingerprint contig physical map. Genomics, Molecular larger than 50 bp, six with a size between 5 and 50 bp Genetics and Biotechnology 46: 12–21. and three shorter than 5 bp. This observation is not in Braglia L, Manca A, Mastromauro F and Breviario D (2010) full agreement with Wang et al. (2005), who reported cTBP: a successful intron length polymorphisms (ILP)- that the most frequent InDels (72.6%) had a size of based genotyping method targeted to well defined exper- ,5 bp, followed (23.5%) by 5–50 bp and very few imental needs. Diversity 2: 572–582. Breviario D, Baird W, Sangoi S, Hilu K, Blumetti P and Gianı` S (3.9%) longer than 50 bp. (2007) High polymorphism and resolution in targeted fin- The involvement of introns in the regulation of gene gerprinting with combined b-tubulin introns. Molecular expression can be due to the coding of regulatory elements Breeding 20: 249–259. such as miRNAs, recently associated with plant stress Campos MD, Cardoso HG, Linke B, Costa JH, Fernandes de responses (Chiou et al., 2006) and development Melo D, Justo L, Frederico AM and Arnholdt-Schmitt B (Wang et al., 2004). In DcAOX2b intron 1 sequences, five (2009) Differential expression and co-regulation of carrot AOX genes (Daucus carota). Physiologia Plantarum 137: pre-miRNAs were predicted: two exclusively in fragment 578–591. a, one in fragment b and two common to both. No pre- Cardoso HG, Campos MD, Costa AR, Campos MC, Nothnagel T miRNA could be predicted for the short fragment c. and Arnholdt-Schmitt B (2009) Carrot alternative oxidase The data obtained for the orthologous gene AOX2b in gene AOX2a demonstrates allelic and genotypic poly- Vigna unguiculata point to its function in modulating morphisms in intron 3. Physiologia Plantarum 137: abiotic stress response (Costa et al., 2007). McCabe et al. 592–608. Chiou TJ, Aung K, Lin SI, Wu CC, Chiang SF and Su CL (2006) (1998) reported an enhancement of AOX2b expression Regulation of phosphate homeostasis by MicroRNA in during the ageing of soybean cotyledons concomitant Arabidopsis. Plant Cell 18: 412–421. to a reduction of AOX2a transcripts and a low level of Costa JH, Jolivet Y, Hasenfratz-Sauder M-P, Orellano EG, Lima AOX1 expression. MGS, Dizengremel P and Fernandes de Melo D (2007) The polymorphisms identified by us in DcAOX2b are Alternative oxidase regulation in roots of Vigna unguiculata cultivars differing in drought/salt tolerance. Journal of special interest due to their location in the first 0 of Plant Physiology 164: 718–727. intron. Introns that are most proximate to the 5 end of Costa JH, Cardoso HC, Campos MD, Zavattieri A, Frederico AM, a gene were observed to exert a more pronounced Fernandes de Melo D and Arnholdt-Schmitt B (2009a) effect on gene expression (Rose et al., 2008). In future D. carota L. – an old model for cell reprogramming experiments, genotypes with homozygous and heterozy- gains new importance through a novel expansion pattern gous polymorphic intron 1 sequences will be explored of AOX genes. Plant Physiology and Biochemistry 47: 753–759. for an association to diverse abiotic stress phenotypes Costa JH, Fernandes de Melo D, Gouveia Z, Cardoso HG, Peixe and differential expression of miRNA. A and Arnholdt-Schmitt B (2009b) The alternative oxidase family of Vitis vinifera reveals an attractive model for genomic design. Physiologia Plantarum 137: 553–565. Ferreira A, Cardoso HG, Macedo ES, Breviario D and Arnholdt- Acknowledgements Schmitt B (2009) Intron polymorphism pattern in AOX1b of wild St John’s wort (Hypericum perforatum) allows discrimination between individual plants. Physiologia This work was supported by the European Commission Plantarum 137: 520–531. through a Marie Curie Chair for Birgit Arnholdt-Schmitt Fiume E, Christou P, Gianı` S and Breviario D (2004) Introns are and by the Foundation for Science and Technology key elements of rice tubulin expression. Planta 218: (FCT) through the COMPETE programme and for scho- 693–703. ` larship to M. Doroteia Campos (SFRH/BI/15991/2006). Gianı S, Morello L, Bardini M and Breviario D (2003) Tubulin intron sequences: multi-functional tools. Cell Biology Inter- The authors are grateful to James C. Nelson from the national 27: 203–205. Kansas University in USA for comments on the manu- Gregory TR (2004) Insertion–deletion bases and the evolution script and language revision. of genome size. Gene 423: 15–34. 180 H. Cardoso et al. McCabe TC, Finnegan PM, Millar H, Day DA and Whelan J Orgel LE and Crick FH (1980) Selfish DNA: the ultimate parasite. (1998) Differential expression of alternative oxidase Nature 284: 604–607. genes in soybean cotyledons during postgerminative Plaxton WC and Podesta´ FE (2006) The functional organisation development. Plant Physiology 118: 675–682. and control of plant respiration. Critical Reviews in Plant McDonald AE and Vanlerberghe GC (2006) The organization Sciences 25: 159–198. and control of plant mitochondrial metabolism. In: Plaxton Polidoros AN, Mylona PV and Arnholdt-Schmitt B (2009) Aox WC and McManus MT (eds) Control of Primary Metabolism gene structure, transcript variation and expression in in Plants. Annual Plant Reviews 22: 290–324. plants. Physiologia Plantarum 137: 342–353. Nasu S, Suzuki J, Ohta R, Hasegawa K, Yui R, Kitazawa N, Rose AB, Elfersi T, Parra G and Korf I (2008) Promoter-proximal Monna L and Minobe Y (2002) Search for and analysis of introns in Arabidopsis thaliana are enriched in dispersed single nucleotide polymorphisms (SNPs) in rice (Oryza signals that elevate gene expression. Plant Cell 20: sativa, Oryza rufipogon) and establishment of SNP mar- 543–551. kers. DNA Research 9: 163–171. Wang XJ, Reyes JL, Chua NH and Gaasterland T (2004) Predic- Naydenov N, Takumi S, Sugie A, Ogihara Y, Atanassov A and tion and identification of Arabidopsis thaliana microRNAs Nakamura C (2005) Structural diversity of the wheat and their mRNA targets. Genome Biology 5: R65–R6515. nuclear gene WAOX1a encoding mitochondrial alternative Wang X, Zhao X, Zhu J and Wu W (2005) Genome-wide inves- oxidase, a single unique enzyme in the cyanide-resistant tigation of intron length polymorphisms and their potential alternative pathway. Biotechnology and Biotechnological as molecular markers in rice (Oryza sativa L.). DNA Equipment 19: 48–56. Research 12: 417–427. q Her Majesty the Queen in Right of Canada, as Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 181–184 represented by the Minister of Agriculture and doi:10.1017/S1479262111000281 Agri-Food Canada [2011] ISSN 1479-2621

Next-Gen sequencing of the transcriptome of triticale

Y. Xu 1, C. Badea1,F.Tran2, M. Frick1, D. Schneiderman2, L. Robert2, L. Harris2, D. Thomas1, N. Tinker2, D. Gaudet1 and A. Laroche1* 1Agriculture and Agri-Food Canada, Research Centre, 5403 1st Avenue S., Lethbridge, AB, Canada T1J 4B1 and 2Agriculture and Agri-Food Canada, ECORC, 960 Carling Avenue, Ottawa, ON, Canada K1A 0C6

Abstract Triticale possesses favourable agronomic attributes originating from both its wheat and rye progenitors, including high grain and biomass yields. Triticale, primarily used as animal feed in North America, is an excellent candidate for production of industrial bio-products. Little is known about the coordination of gene expression of rye and wheat genomes in this intergeneric hybrid, but significant DNA losses from the parental genomes have been reported. To clarify the regulation of gene expression in triticale, we carried out 454 sequencing of cDNAs obtained from root, leaf, stem and floral tissues in different lines of triticale and rye exhibiting different phenotypes and assembled reads into contigs. Related to the data assembly were the absence of reference genomes and the paucity of rye sequences in GenBank or other public databases. Consequently, we have sequenced cDNA libraries from roots, seedlings, leaves, floral tissues and immature seeds to facilitate the identification of triticale sequences originating from rye. To further characterize the wheat-derived cDNAs, we also developed a database close to 25,000 non-redundant full-length wheat coding sequence genes, based on existing databases and contigs that were verified against protein sequences from the grass genomes of Brachypodium distachyon, rice, sorghum and maize.

Keywords: next-generation sequencing; transcriptome; Triticale

Introduction Next-generation sequencing provides an efficient tool to address biological questions at the transcriptomic Triticale ( £ Triticosecale Wittm.) is an intergeneric hybrid and genomic levels. RNA-Seq for transcriptomics (Wang between wheat species (Triticum ssp. AA, AABB and et al., 2009) has been successfully employed in crop AABBDD) and rye (Secale cereale, RR). Triticale possesses plants (Zenoni et al., 2010; Zhang et al., 2010). favourable agronomic attributes originating from both its We are reporting on the utilisation of the Roche 454 wheat and rye progenitors, including high grain and sequencing technology to investigate the transcriptome biomass yields. Although significant DNA losses from the of triticale and rye in different tissues and at different parental genomes have been reported (Ma and Gustafson, developmental stages. This preliminary analysis reports 2008), little is known about the coordination of gene on gene expression and identification of triticale- and expression of rye and wheat genomes in this intergeneric rye-assembled genes and on the development of an hybrid. So far, public databases (e.g. GenBank) are reporting enhanced full-length (FL) cDNA database of wheat to about 14,000 Expressed Sequence Tag (EST) or DNA facilitate our analysis. sequences of rye and a few dozen sequences for triticale. Materials and methods

Tissues including leaf, root, stem and different reproduc- * Corresponding author. E-mail: [email protected] tive organs (stigma, anthers, pollen and immature heads) 182 Y. Xu et al. from genotypes of rye (Prima and Vacaria) and triticale whereas 19,142 (75.1%) Brachypodium proteins were (AC Certa, hollow stemmed; Triticale 797 and Triticale recognized by rye contigs. The triticale and rye sequences 1308, both solid stemmed) sampled at different stages of displayed homology to 15,904 Brachypodium proteins, development were used in this study. AC Certa seedling while triticale- and rye-assembled sequences identified tissues were also exposed to water stress. 2011 and 3777 specific proteins, respectively. These results RNA-Seq was carried out. Briefly, total RNA clearly show that a substantial proportion of rye sequences (Trizol (Invitrogen) followed by Qiagen RNeasy midi are not expressed in triticale. purification) was extracted from the different tissues. We also combined all the rye and triticale 454 reads PolyAþ mRNA using Poly (A) Puriste Kit (Ambion, Inc) together and conducted a second de novo assembly was purified from 0.6 mg of total RNA followed by cDNA utilising the same parameters used for the individual library synthesis. Five micrograms of double stranded rye and triticale assembly. When we parsed the contig cDNA was utilized for 454 sequencing. Eleven triticale and makeup, we found that most contigs were made up of eight rye libraries were sequenced using the 454 GS FLX sequences from only one species, triticale or rye (Fig. 1), Titanium (Roche) technology. Different publically available thus clearly indicating the genome origin of the majority and proprietary software programs were used in the of triticale transcripts. analysis of 454 sequencing results and included: BLAST Sequence variation between rye (RR) and triticale (ftp://ftp.ncbi.nih.gov/blast/; Altschul et al., 1997), TGICL (AABBRR) was observed as exemplified in Fig. 2. As indi- (http://compbio.dfci.harvard.edu/tgi/software/; Pertea et al., cated by the arrows, two diagnostic nucleotides between 2003), CAP3 (Huang and Madan, 1999), SeqClean (http:// rye and triticale were identified. In triticale, five and poss- compbio.dfci.harvard.edu/tgi/software/), OrthoMCL 1.4 ibly six reads out of the 14 sequences were clearly related (http://www.orthomcl.org/cgi-bin/OrthoMclWeb.cgi), CD- to the rye genome. This can also provide information on HIT-EST (http://weizhong-lab.ucsd.edu/cdhit_suite/cgi-bin/ relative expression of homeologous sequences from index.cgi), BioPerl module Bio::SeqIO and Bio::SearchIO wheat and rye in triticale when applied to datasets from (http://www.bioperl.org/wiki/Main_Page) and DNASTAR each individual cDNA library. SeqMan and NGEN (DNASTAR, Madison, WI, USA) for de novo assembly. Enhancing the wheat FLcDNA dataset for analysis of triticale and other small grain cereals Results and discussion In order to compare transcriptional units of rye (RR), Assembling sequences reflects the genome triticale (AABBRR) and common wheat (AABBDD), we divergence of triticale and rye built an enhanced reference FLcDNA dataset starting

Sequencing results from eleven triticale libraries generated 3,310,375 reads with an average length 320 bp, while results from eight rye libraries yielded 3,124,641 10,000 reads with an average length 345 bp. De novo assembly of these two datasets with parameters set at 92% identity and minimum match of 30 nt yielded 162,686 contigs 5000

from triticale, which was 35% higher than the number Frequency of contigs obtained from rye, 120,416. Interestingly, the number of long contigs above 2 kb from rye, 4657, was almost double than those from triticale, 2774. Given the similar number of input sequences for assembly of rye 0 and triticale, it was not surprising to identify a smaller 0 20406080100 number of contigs above 2 kb in triticale due to the Triticale nucleotides (%) divergence between the three subgenomes (AABBRR) Fig. 1. Distribution of sequences originating from triticale and of triticale (Chalupska et al., 2008) and the stringent rye libraries in the assembly of contigs. For any given contig parameters used for both the rye and triticale contigs on the x-axis, we have evaluated the contribution of the assembly. number of sequences from rye (Nr) and triticale (Nt) using þ £ To further investigate expressed sequence divergence the following formula Nt/(Nt Nr) 100. The value at ‘0’ represents contigs with 100% rye sequences, while the value between triticale and rye, we compared these contigs to at ‘100’ represents contigs with 100% triticale sequences. the Brachypodium proteome using BLASTX. Triticale This could be used to estimate the divergence between contigs matched 17,915 (70.3%) Brachypodium proteins, triticale and rye sequences within each individual contig. Next-Gen sequencing of triticale 183

1 2 Rye 3 4 5 6 7 8 9 10 11 12 Triticale 13 14 15 16 17 18 19 Fig. 2. Nucleotide variation to distinguish ESTs originating from rye and wheat genomes in triticale. The upper panel represents the nucleotide composition from rye, while the lower panel represents the nucleotide composition detected in triticale. The sequence corresponding to rye can be found in triticale in lines 8, 9, 15, 16, and 17. with 1,067,304 wheat EST sequences available in This dataset will also be very valuable for any wheat GenBank. Firstly, we optimized the assembly pipeline RNA-Seq project. by introducing a ‘noise site correction’ step for the indi- We have assembled de novo rye and triticale 454 vidual ESTs and an iterative assembly step based on the datasets, distinguished rye from wheat sequences and widely used EST assembly program TGICL (Pertea et al., developed an enhanced nr wheat FLcDNA dataset of 2003). More specifically, we used a first round CAP3 almost 24,800 distinct elements. assembly with overlap of 40 nt and per cent identity of 80% to remove the ‘noise’ nt from the aligned sequences. Then, a second round of CAP3 assembly, still with an Acknowledgements overlap of 40 nt but with a per cent identity increased to 95%. Under these conditions, the assembly of over 1 Funding from the AAFC-ABIP program to The Cellulosic million wheat EST dataset yielded a total of 79,034 con- Biofuel Network (159) and the Canadian Triticale Bio- sensus sequences. The FLcDNA prediction pipeline refinery Initiative (227) projects and from Alberta guided by four known grass genomes (Brachypodium, Energy–Genome Alberta is greatly appreciated. rice, maize and sorghum) identified 21,756 FL wheat cDNAs. The 21,002 publically available FLcDNAs from References TriFLDB (11,877; Mochida et al., 2009) and GenBank (9125) were combined and treated with ‘Cd-hit-est’ to Altschul SF, Madden TL, Scha¨ffer AA, Zhang J, Zhang Z, Miller W delete the redundancy at 95% identical level to yield and Lipman DJ (1997) Gapped BLAST and PSI-BLAST: 12,715 non-redundant (nr) sequences. This number was a new generation of protein database search programs. almost doubled to 24,789 nr wheat FLcDNAs when our Nucleic Acids Research 25: 3389–3402. FLcDNA-EST dataset was amalgamated under the same Chalupska D, Lee HY, Faris JD, Evrard A, Chalhoub B, Haselkorn R and Gornicki P (2008) Acc homoeoloci and parameters. A comparison to the Brachypodium pro- the evolution of wheat genomes. The Proceeding of the teome suggests that the expanded nr wheat FLcDNA National Academy of Sciences USA 105: 9691–9696. dataset covers the majority of the proteins of this refer- Huang X and Madan A (1999) CAP3: a DNA sequence assembly ence species. program. Genome Research 9: 868–877. To evaluate the usefulness of the nr wheat FLcDNA Ma XF and Gustafson JP (2008) Allopolyploidization-accommo- dated genomic sequence changes in triticale. Annals of dataset, we mapped 384,470 triticale 454 ESTs using Botany 101: 825–832. MegaBLAST. More than 66% of our 454 reads could be Mochida K, Yoshida T, Sakurai T, Ogihara Y and Shinozaki K aligned to our enhanced nr wheat FLcDNAs dataset. (2009) TriFLDB: a database of clustered full-length coding 184 Y. Xu et al. sequences from Triticeae with applications to comparative Zenoni S, Ferrarini A, Giacomelli E, Xumerle L, Fasoli M, grass genomics. Plant Physiology 150: 1135–1146. Malerba G, Bellin D, Pezzotti M and Delledonne M (2010) Pertea G, Huang X, Liang F, Antonescu V, Sultana R, Characterization of transcriptional complexity during Karamycheva S, Lee Y, White J, Cheung F, Parvizi B, Tsai berry development in Vitis vinifera using RNA-Seq. Plant J and Quackenbush J (2003) TIGR gene indices clustering Physiology 152: 1787–1795. tools (TGICL): a software system for fast clustering of Zhang G, Guo G, Hu X, Zhang Y, Li Q, Li R, Zhuang R, Lu Z, He large EST datasets. Bioinformatics 19: 651–652. Z, Fang X, Chen L, Tian W, Tao Y, Kristiansen K, Zhang X, Wang Z, Gerstein M and Snyder M (2009) RNA-Seq: a revolu- Li S, Yang H, Wang J and Wang J (2010) Deep RNA sequen- tionary tool for transcriptomics. Nature Reviews Genetics cing at single base-pair resolution reveals high complexity 10: 57–63. of the rice transcriptome. Genome Research 20: 646–654. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 185–188 ISSN 1479-2621 doi:10.1017/S1479262111000086

A comparison of population types used for QTL mapping in Arabidopsis thaliana

Joost J. B. Keurentjes1,2, Glenda Willems3,4, Fred van Eeuwijk1,2, Magnus Nordborg3,5 and Maarten Koornneef1,4* 1Wageningen University, Wageningen, The Netherlands, 2Centre for Biosystems Genomics, Wageningen, The Netherlands, 3University of Southern California, Los Angeles, CA, USA, 4Max Planck Institute for Plant Breeding Research, Cologne, Germany and 5Gregor Mendel Institute, Vienna, Austria

Abstract In Arabidopsis, a variety of mapping populations have been used for the detection of quantitative trait loci (QTLs) responsible for natural variation. In this study, we present an overview of the advantages and disadvantages of the different types of populations used. To do this, we compare the results of both experimental and natural populations for the commonly analysed trait flowering time. It is expected that genome wide association (GWA) mapping will be an increasingly important tool for QTL mapping because of the high allelic richness and mapping resolution in natural populations. In Arabidopsis, GWA mapping becomes ever more facilitated by the increasing availability of re-sequenced genomes of many accessions. However, specifically designed mapping populations such as recombinant inbred lines and near isogenic lines will remain important. The high QTL detection power of such experimental populations can identify spurious GWA associations, and their unique genomic structure is superior for investigating the role of low-frequency alleles. Future QTL studies will therefore benefit from a combined approach of GWA and classical linkage analysis.

Keywords: Arabidopsis thaliana; ﬂowering time; natural variation; quantitative trait loci mapping

Introduction mapping populations have also the advantage that experiments analysing different traits and conditions can be The analysis of the loci controlling genetic variation for compared on the basis of QTL map positions. This has quantitative traits is performed by so-called quantitative already led to the identification of new pleiotropic func- trait loci (QTL) analysis, where trait values are compared tions of known genes (Koornneef et al., 2004). In this between molecular marker genotypes, which are linked study, we will describe some of the properties and to different alleles of causal genes. For this type of characteristics of a number of immortal populations analysis, large populations with accurate and dense used in Arabidopsis thaliana. genetic marker maps as well as accurate phenotyping are required. The latter can be achieved when one can work with homozygous lines, which can be analysed in replications. Such materials can be derived by repeated Types of mapping populations used in Arabidopsis selfing of initially heterozygous individuals or by the use of naturally occurring inbred genotypes. Immortal The quantitative variation in complex traits and the control by heritable factors are broadly acknowledged in Arabidopsis (Alonso-Blanco and Koornneef, 2000). This observation, together with the ease to generate * Corresponding author. E-mail: [email protected] segregating material and the availability of genomic and 186 Joost J. B. Keurentjes et al. genetic resources, led to a manifold of QTL studies same goal, which is often referred too as ‘Mendelizing’ aiming at the detection of the genes underlying this a QTL. Apart from the inaccuracy of the map positions, natural variation (reviewed in Alonso-Blanco et al., 2009). biparental populations have as disadvantage that they Although genetic analysis can be performed on any allow only the analysis of genetic variation present type of segregating population, this review focuses on between the two parents. collections of homozygous lines. Widely used are recom- The solution for both mapping accuracy and the binant inbred lines (RILs) derived by single seed descent analysis of more variation is the use of genome wide from F2 individuals, which are the progeny of a hybrid association (GWA) mapping (Nordborg and Weigel, between two distinct homozygous (inbred) lines. The 2008). This method uses (in practice a subset of) all variants largest disadvantage of inbred lines, compared with available and aims to associate trait differences with natural (outbreeding) populations, is the limited number specific genotypes. Due to the fast LD decay in natural of effective meiotic recombination events. To overcome Arabidopsis accessions, often less than 10 Kbp (Kim this limitation, large population sizes are needed to et al., 2007), this procedure requires dense genetic maps sample enough meiotic events. Alternatively, one can (Weigel and Mott, 2009). Although a powerful approach, artificially increase the number of meiotic events and a number of factors limit the successful use of GWA. thereby the resolution to map loci by additional rounds One is the often present population structure resulting in of selective outcrossing. Several RIL populations have false positive associations. Compensating for the effects been derived after inter-crossing F2 plants and later gen- of population structure often also removes true positives. erations (Balasubramanian et al., 2009). However, even These so-called false negatives (Brachi et al., 2010) do after repeated inter-crossing, RIL populations still suffer show up and are therefore much better studied in from slowly decaying linkage disequilibrium (LD), experimental populations. Another important limitation which together with their often limited population size is that low-frequency alleles, even when having strong makes that identified QTLs have rather large confidence effects, remain undetected. An example of this is the intervals. Resolution can be increased by using large CRY2 allele from the Cvi accession that results in a major populations, or when QTL analyses of different popu- QTL in crosses (El-Assal et al., 2001; Brachi et al., 2010) lations can be combined as was recently demonstrated but is not detected in GWA analyses (Atwell et al., 2010). for the analysis of seed dormancy in six RIL populations Intermediate mapping populations that combine the (Bentsink et al., 2010). advantages of both approaches consist of experimental Studying only one QTL at the time avoids compli- crossings with increasing numbers of parents, sizes of cations of the segregation of multiple loci (e.g. epistasis). populations and recombination events. One solution is This can be achieved by the construction of introgression the combined analysis of multiple populations as was or near isogenic lines (NILs), in which small chromoso- demonstrated by nested association mapping in maize, mal regions from a donor parent are introduced in the where 25 RIL populations were generated by crossing genetic background of a recurrent parent (Koumproglou diverse parents with a common reference (Buckler et al., 2002; Keurentjes et al., 2007; Torjek et al., 2008). et al., 2009). In Arabidopsis, this was demonstrated at a NILs, although often providing less accurate map pos- smaller scale, using Ler (Bentsink et al., 2010) or Col itions, allow the detection of minor QTLs that cannot (Brachi et al., 2010) as a common parent, respectively. be detected in RILs (Keurentjes et al., 2007). The use of The use of multiple parents in different combinations is heterogeneous inbred line families selected from early another alternative, and two of such multiple parent com- generation inbred lines that are still heterozygous in the binations are now described for Arabidopsis. One is the region of interest (Tuinstra et al., 1997) achieves the so-called MAGIC population developed by Kover et al.

Table 1. QTL analyses of ﬂowering time in studies using different types of mapping populations

Type Parents Size QTL R 2 References NIL Ler £ Cvi 92 5 83 Keurentjes et al. (2007) RIL Ler £ Div.a 120–164 3–8 68–85 El-Lithy et al. (2006); Keurentjes et al. (2007) RIL Col £ Div.a 223–456 2–7 31–70 Brachi et al. (2010) MAGIC 19 Inter-crossed 527 4 64 Kover et al. (2009) AMPRIL 8 Inter-crossed 550 4 88 Huang et al. (2011) GWA Wild accessions 96–256 .20b ,50b Atwell et al. (2010); Brachi et al. (2010) R 2, the total explained variance. a Data of diverse populations using a common reference accession (Ler or Col). b Number of detected QTLs and total explained variance in GWA analyses depends strongly on the applied methods and threshold levels. QTL mapping in Arabidopsis thaliana 187 (2009). This population is derived by inter-crossing Jones JD, Michael T, Nemri A, Roux F, Salt DE, Tang C, 19 parents for several generations, after which RILs were Todesco M, Traw MB, Weigel D, Marjoram P, Borevitz JO, Bergelson J and Nordborg M (2010) Genome-wide associ- developed. The other, the so-called AMPRIL population ation study of 107 phenotypes in Arabidopsis thaliana (Paulo et al., 2008; Huang et al., 2011), was derived inbred lines. Nature 465: 627–631. from inter-crossing four pairwise hybrids from eight Balasubramanian S, Schwartz C, Singh A, Warthmann N, founder lines, followed by RIL construction via single Kim MC, Maloof JN, Loudet O, Trainer GT, Dabi T, Borevitz JO, Chory J and Weigel D (2009) QTL mapping seed descent up to the F5. in new Arabidopsis thaliana advanced intercross- recombinant inbred lines. Public Library of Science ONE 4: e4318. QTL analysis of flowering time Bentsink L, Hanson J, Hanhart CJ, Blankestijn-de Vries H, Coltrane C, Keizer P, El-Lithy M, Alonso-Blanco C, de Table 1 summarizes the results of different population Andres MT, Reymond M, van Eeuwijk F, Smeekens S and Koornneef M (2010) Natural variation for seed dormancy types. Many QTLs were identified at similar positions and in Arabidopsis is regulated by additive genetic and mole- most likely represent the same loci. Only when field cular pathways. Proceedings of the National Academy of conditions were employed predominantly different, Sciences USA 107: 4264–4269. QTLs were detected (Brachi et al., 2010), indicating the Brachi B, Faure N, Horton M, Flahauw E, Vazquez A, Nordborg importance of genotype £ environment interactions. M, Bergelson J, Cuguen J and Roux F (2010) Linkage and When considering multiple parent populations, the association mapping of Arabidopsis thaliana flowering time in nature. Public Library of Science Genetics 6: number of QTLs was surprisingly low in the MAGIC and e1000940. AMPRIL populations (Kover et al., 2009; Huang et al., Buckler ES, Holland JB, Bradbury PJ, Acharya CB, Brown PJ, 2010). Partially this can be explained by the reduced Browne C, Ersoz E, Flint-Garcia S, Garcia A, Glaubitz JC, power due to the multiple alleles that segregate at multiple Goodman MM, Harjes C, Guill K, Kroon DE, Larsson S, loci in multi-parent populations. A recent GWA study Lepak NK, Li H, Mitchell SE, Pressoir G, Peiffer JA, Rosas MO, Rocheford TR, Romay MC, Romero S, Salvo S, Sanchez included flowering time analyses from various experi- Villeda H, da Silva HS, Sun Q, Tian F, Upadyayula N, Ware ments (Atwell et al., 2010). The study showed the D, Yates H, Yu J, Zhang Z, Kresovich S and McMullen MD complication of population structure that is relatively (2009) The genetic architecture of maize flowering time. strong for flowering time, as well as the known issues of Science 325: 714–718. significance thresholds and false positives. By combining El-Assal S, Alonso-Blanco C, Peeters AJ, Raz V and Koornneef M the RIL and GWA approaches in the same experiment, (2001) A QTL for flowering time in Arabidopsis reveals a novel allele of CRY2. Nature Genetics 29: 435–440. these alleged false positives could be reduced and clear El-Lithy ME, Bentsink L, Hanhart CJ, Ruys GJ, Rovito D, candidates could be indicated (Brachi et al., 2010). Broekhof JL, van der Poel HJ, van Eijk MJ, Vreugdenhil D The recent developments in GWA mapping together with and Koornneef M (2006) New Arabidopsis recombinant the relatively ease to generate experimental mapping inbred line populations genotyped using SNPWave and populations, either as F /F and/or the efficient develop- their use for mapping flowering-time quantitative trait 2 3 loci. Genetics 172:1867–1876. ment of doubled haploids using centromere-mediated Huang X, Paulo MJ, Boer M, Effgen S, Keizer P, Koornneef M genome elimination (Ravi and Chan, 2010), will continue and van Eeuwijk F (2011) Analysis of natural allelic to contribute to our understanding of complex traits. variation in Arabidopsis, using a new multi-parent recom- We therefore argue that multiple types of populations binant inbred line population. Proceedings of the National and approaches are needed for a thorough understanding Academy of Sciences USA in press. Keurentjes JJB, Bentsink L, Alonso-Blanco C, Hanhart CJ, of the complex genetic architecture of quantitative traits. Blankestijn-De Vries H, Effgen S, Vreugdenhil D and Koornneef M (2007) Development of a near-isogenic line population of Arabidopsis thaliana and comparison of mapping power with a recombinant inbred line popu- References lation. Genetics 175: 891–905. Kim S, Plagnol V, Hu TT, Toomajian C, Clark RM, Ossowski S, Alonso-Blanco C and Koornneef M (2000) Naturally occurring Ecker JR, Weigel D and Nordborg M (2007) Recombination variation in Arabidopsis: an underexploited resource for and linkage disequilibrium in Arabidopsis thaliana. Nature plant genetics. Trends in Plant Science 5: 22–29. Genetics 39: 1151–1155. Alonso-Blanco C, Aarts MG, Bentsink L, Keurentjes JJB, Koornneef M, Alonso-Blanco C and Vreugdenhil D (2004) Natu- Reymond M, Vreugdenhil D and Koornneef M (2009) What rally occurring genetic variation in Arabidopsis thaliana. has natural variation taught us about plant development, Annual Review of Plant Physiology and Plant Molecular physiology, and adaptation? Plant Cell 21: 1877–1896. Biology 55: 141–172. Atwell S, Huang YS, Vilhjalmsson BJ, Willems G, Horton M, Li Y, Koumproglou R, Wilkes TM, Townson P, Wang XY, Beynon J, Meng D, Platt A, Tarone AM, Hu TT, Jiang R, Muliyati NW, Pooni HS, Newbury HJ and Kearsey MJ (2002) STAIRS: Zhang X, Amer MA, Baxter I, Brachi B, Chory J, Dean C, a new genetic resource for functional genomic studies of Debieu M, de Meaux J, Ecker JR, Faure N, Kniskern JM, Arabidopsis. Plant Journal 31: 355–364. 188 Joost J. B. Keurentjes et al. Kover PX, Valdar W, Trakalo J, Scarcelli N, Ehrenreich IM, Ravi M and Chan SW (2010) Haploid plants produced by centro- Purugganan MD, Durrant C and Mott R (2009) A multi- mere-mediated genome elimination. Nature 464: 615–618. parent advanced generation inter-cross to fine-map Torjek O, Meyer RC, Zehnsdorf M, Teltow M, Strompen G, quantitative traits in Arabidopsis thaliana. Public Library Witucka-Wall H, Blacha A and Altmann T (2008) of Science Genetics 5: e1000551. Construction and analysis of 2 reciprocal Arabidopsis intro- Nordborg M and Weigel D (2008) Next-generation genetics in gression line populations. Journal of Heredity 99: 396–406. plants. Nature 456: 720–723. Tuinstra MR, Ejeta G and Goldsbrough PB (1997) Hetero- Paulo M, Boer M, Huang X, Koornneef M and van Eeuwijk F geneous inbred family (HIF) analysis: a method for (2008) A mixed model QTL analysis for a complex cross developing near-isogenic lines that differ at quantitative population consisting of a half diallel of two-way hybrids trait loci. Theoretical and Applied Genetics 95: 1005–1011. in Arabidopsis thaliana: analysis of simulated data. Weigel D and Mott R (2009) The 1001 genomes project for Euphytica 161: 107–114. Arabidopsis thaliana. Genome Biology 10: 107. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 189–192 ISSN 1479-2621 doi:10.1017/S1479262111000384

Molecular characterization of the Latvian apple (Malus) genetic resource collection based on SSR markers and scab resistance gene Vf analysis

Gunars Lacis1*, Irita Kota1, Laila Ikase1 and Dainis Rungis2 1Latvia State Institute of Fruit-Growing, Graudu Str.1, Dobele, Latvia and 2Genetic Resource Centre, Latvian State Forestry Research Institute “Silava” Rigas Str. 111, Salaspils, Latvia

Abstract Apple cultivars are an integral part of the Latvian State Program for Preservation and Sustain- able Use of Genetic Resources for Food and Agriculture. Latvian apple (n ¼ 109) cultivars of local origin, nominated as National Plant Genetic Resources, were analyzed with eight simple sequence repeat (SSR) markers recommended by the European Cooperative Programme for Plant Genetic Resources (ECPGR) Malus/Pyrus working group and a marker for scab resistance gene (Vf). All SSR loci exhibited a high level of polymorphism – 12 to 39 alleles,

18.75 in average, with high observed heterozygosity (Ho) ranging from 0.64 to 0.89 and a mean of 0.78. The gene diversity (polymorphism information content value) varied from 0.79 to 0.90, with an average of 0.86. All cultivars could be distinguished with the tested set of SSR loci, with a high frequency of rare (38%) or unique (26%) alleles. For the Vf gene, all three possible genotypes (VfVf, Vfvf and vfvf) were detected for 1, 6 and 102 varieties, respectively. The analyzed germplasm showed high genetic diversity, particularly landraces, while the modern cultivars clustered mainly into two groups. Cluster analysis did not reveal a clear pattern with well-deﬁned groups, but conﬁrmed some relationships based on known or putative pedigrees, as well as suggesting the possible parentage of some cultivars.

Keywords: genetic diversity; germplasm; Malus spp.; microsatellite

Introduction and elite hybrids originated and selected in Latvia, with valuable quantitative or qualitative traits; cultivars of Fruit crop genetic resources are very important both for local origin (landraces), including foreign cultivars with research and breeding purposes. Apple (Malus sp.) acces- a long history of cultivation in Latvia. sions are an integral part of the Latvian State Program for Evaluation and characterization of apple genetic Preservation and Sustainable Use of Genetic Resources resources is an important task for further maintenance for Food and Agriculture. The Latvia State Institute of and utilization activities. Since characterization by Fruit-Growing (LSIFG) pome fruit collection includes phenotype has serious limitations due to environmental ,700 apple accessions, of which 109 accessions belong influence, limited data exchange and comparison with to the 1st priority group of national genetic resources. other researchers, the LSIFG has introduced molecular This genetic resources group includes cultivars produced marker technologies for genetic resource characterization from Latvian breeding programmes, as well as clones and breeding. Apple genotyping was performed using SSR markers recommended by the European Cooperative Programme for Plant Genetic Resources (ECPGR) Malus/ Pyrus working group as well as gene-specific molecular * Corresponding author. E-mail: [email protected] markers (e.g. Vf resistance gene marker). 190 G. Lacis et al. Materials and methods genotyping showed high presence of rare or unique alleles, 38 and 26% respectively (Table 1). All cultivars in Total DNA was isolated from young leaves using a Geno- the collection could be distinguished with the tested set mic DNA Purification Kit (Fermentas, Lithuania). of SSR loci. Internal relationships of apple cultivars were PCR were performed in a 20 ml reaction mixture with evaluated using SSR data-based dendrogram created 25 ng DNA, 2 mM each primer, 200 mM of each nucleotide, applying Nei and Li genetic distances and Neighbour- 1X PCR buffer and 0.05 U/ml REDTaqw DNA Polymerase Joining tree construction method. Cluster analysis did (Sigma, USA) per reaction, in the ep gradient thermal not reveal a clear pattern of clustering with well-defined cycler (Eppendorf, Germany). The PCR conditions were variety groups, but confirmed some known relationships the same as already described in the methodology for based on known or putative pedigree. The analyzed SSR markers (Guilford et al., 1997; Gianfranceschi et al., Latvian apple germplasm showed high genetic diversity, 1998) and Vf gene detection (Vejl et al., 2003). particularly landraces, while the cultivars developed PCR products were first checked on 1% agarose gels in within the modern breeding program clustered mainly 1£ Tris–acetate–ethylenediaminetetraacetic acid buffer into two groups. The analyzed plant material also and visualized by staining with ethidium bromide to included two widely grown cultivars ‘McIntosh’ and test for the presence of PCR products. For SSR geno- ‘Prima’, which allowed for checking the accuracy of typing, the same PCR products were subsequently allele scoring and comparing the results of this study analyzed on an ABI PRISMw 3100 Genetic Analyzer with those presented in previously published studies. (Applied Biosystems, USA) and genotyped using Gene- A significant part of the cultivars grown in Latvia Mapperw Software v4.0 (Applied Biosystems). include in their pedigree the cultivars derived from Genetic parameters were calculated using the compu- ‘McIntosh’ (‘Cortland’, ‘Lobo’, ‘Melba’ etc.), local cultivars ter program GENALEX 6.1 (Peakall and Smouse, 2006). ‘Sı¯polin¸sˇ’, ‘Baltais Dzidrais’ (‘Yellow Transparent’), ‘Anto- Potential population structure was analyzed and phylo- novka’, ‘Re¯veles Bumbiera¯bele’ (‘Revaler Birnapfel’) genetic tree was constructed based on Nei’s genetic or ‘Cukurin¸sˇ’ (‘Korobovka’). The obtained dendrogram distance (Nei, 1973) and Neighbour-Joining clustering did not always show clear relationships according to method using MEGA version 4 (Tamura et al., 2007). known pedigrees or morphological features, but helped to confirm some known ones or to cast doubt upon reported pedigrees, as well as suggesting possible pre- Results and discussion viously unknown parentages (Fig. 1). For example, the Daugmale clone of ‘Baltais Dzidrais’ comes out as a poss- The 109 apple varieties nominated as National Plant Gen- ible seedling, not as a clone. The same may be true for etic Resources were analyzed using a set of eight selected the two supposed clones of the old cultivar ‘Nicˇnera SSR markers (recommended by the ECPGR Malus/Pyrus Zemen¸u’ (‘Nitschners Erdbeerapfel’). Also, ‘Robezˇnieku working group) and a marker for scab resistance gene Sı¯polin¸sˇ’ is too far distanced from ‘Sı¯polin¸sˇ’tobea (Vf). All SSR loci exhibited a very high level of poly- clone. On the other side, our results confirmed the morphism – 12 to 39 alleles, 18.75 in average, with genetic difference between Dobele and Pu¯re clones of high observed heterozygosity (Ho) ranging from 0.64 to ‘A¯dama¯bele’, which was already previously doubted. 0.89 and a mean of 0.78. The gene diversity (polymorph- The dendrogram confirmed also the supposed origin of ism information content (PIC) value) for tested loci varied cv. ‘Sa¯ritsa Agra¯’ from ‘Baltais Dzidrais’ and cv. ‘Vigo’ from 0.79 to 0.90, with an average of 0.86. Results of from ‘Sı¯polin¸sˇ’.

Table 1. Genotyping results for the Latvian apple accessions using SSR markers

No. of homozygous No. of No. of Discrimination Allele length Locus plants alleles Ho PIC value genotypes power range (bp) CH02d08 19 17 0.826 0.849 37 0.958 168–258 CH04c07 32 13 0.706 0.871 46 0.963 94–140 CH01e12 28 39 0.725 0.837 71 0.975 89–223 CH01h01 36 12 0.670 0.789 28 0.935 102–130 CH02c09 13 14 0.881 0.876 36 0.961 219–259 CH02c06 12 16 0.890 0.890 59 0.980 202–267 CH01f02 14 17 0.872 0.903 53 0.982 62–224 NZ05g8 39 22 0.642 0.877 49 0.964 68–246 Average 24.1 18.8 0.777 0.862 47.4 0.965 Molecular characterization of the Latvian apple 191

Fig. 1. Genetic relatedness dendrogram of Latvian apple genetic resources accessions. Group of origin: W, ‘McIntosh’; X, ‘Sı¯polin¸sˇ’; X/W, ‘McIntosh’ and ‘Sı¯polin¸sˇ’, A, ‘Antonovka’; B, ‘Baltais Dzidrais’ (‘Yellow Transparent’); K, ‘Re¯veles Bumbiera¯bele’ (‘Revaler Birnapfel’); O, ‘Cukurin¸sˇ’ (‘Korobovka’); S, ‘Delicious’.

The origin of landraces and amateur cultivars While the claimed pedigree of ‘Korta’ is ‘Cukurin¸sˇ’ £ usually is either unknown or unclear. Our results, ‘Re¯veles Bumbiera¯bele’, our study shows that the together with known pedigree data, helped to suggest second parent could be ‘Baltais Dzidrais’ instead. The the possible parentage of several local cultivars, e.g. close linking between ‘Ella’ and ‘Ventspiliete’ creates for ‘Celmin¸u Dzeltenais’ (‘Ananasrenette’ £ ‘Dzeltenais doubt about the precision of data about their parentage Dzidrais’?), ‘Vilka Rozˇa¯bele’ (‘Suislepp’ £ ‘Nicˇnera (‘Anoka’ £ ‘Golden Delicious’ and ‘James Grieve’ £ Zemen¸u’?), ‘Ugunda’ and ‘Lı¯cı¯ˇsu Ziemas’ as both derived ‘Serinka’, respectively). The phylogenetic analysis also from ‘Chernoguz’, ‘Lizuma Ziemas’ as derived from the supported the previous doubts about the parentage of ancient landrace ‘Zak¸apurni’, established a close link the modern cultivar ‘Ligita’ (‘Alkmene’ £ ‘Bogatyr’?); between cultivars ‘Velte’ (‘James Grieve’ £ pollen mix) its pollen parent may be a scab-immune cultivar with and ‘Aivarin¸sˇ’ (unknown), etc. ‘McIntosh’ in the pedigree. 192 G. Lacis et al. The grouping of Malus sylvestris (sample collected Acknowledgements at Slı¯tere Nature Reserve) with several ‘McIntosh’-type cultivars was surprising, as well as the similarity of This research was supported by the grant of the Latvian red-leaved (f. niedzwetskyana) local foundling ‘Carni- Council of Science and project co-funded by EU ‘Scientific kava’ with cv. ‘Ra¯ja’ (‘Cortland’ £ ‘Serinka’), and the capacity building in fruit-growing, forestry and infor- close relationship between the highly unsimilar land- mation technology sectors, providing research on races ‘Majoru Saldais’ and ‘Vasaras Citronu’. More pre- environmental-friendly growing strategies, product devel- cise results may be obtained when other known or opment and introduction aided by computer technol- possible parent cultivars are included in the study; for ogies’, no. 2009/0228/1DP/1.1.1.2.0/09/APIA/VIAA/035. example, ‘Antonovka’, ‘Revaler Birnapfel’, ‘Suislepp’ and ‘Treboux’. For the Vf gene, all three possible genotypes (VfVf, Vfvf and vfvf) were detected for 1, 6 and 102 cultivars, References respectively. Cultivar ‘Prima’ was used as a positive control for apple scab resistance gene (Vf) detection. Gianfranceschi L, Seglias N, Tarchini R, Komjanc M and Cultivars with VfVf and Vfvf genotypes have been sele- Gessler C (1998) Simple sequence repeats for the genetic cted as valuable sources for further apple scab resistance analysis of apple. Theoretical and Applied Genetics 96: breeding. All cultivars where the dominant Vf allele 1069–1076. was found have been produced as a result of the Guilford P, Prakash S, Zhu JM, Rikkerink E, Gardiner S, Bassett H and Forster R (1997) Microsatellites in Malus £ domestica modern breeding programme. (apple): abundance, polymorphism and cultivar identi- Application of eight well-tested microsatellite markers fication. Theoretical and Applied Genetics 94: 249–254. as well as Vf gene-specific markers on Latvian Malus Nei M (1973) Analysis of gene diversity in subdivided popu- genetic resources showed that: lations. The Proceedings of the National Academy of Sciences USA 70: 3321–3323. (1) The tested microsatellite markers showed suitability Peakall R and Smouse PE (2006) GENALEX6: genetic analysis in Excel. Population genetic software for teaching and for evaluation of genetic diversity and relatedness research. Molecular Ecology Notes 6: 288–295. of Latvian apple genetic resources, and ensured Tamura K, Dudley J, Nei M and Kumar S (2007) MEGA4: discrimination of all accessions including varieties Molecular Evolutionary Genetics Analysis (MEGA) software sharing the same pedigree. version 4.0. Molecular Biology and Evolution 24: (2) Application of eight tested microsatellite markers 1596–1599. Vejl P, Skupinova S, Blazˇek J, Sedla´k P, Bardova´ M, Drahosˇova showed suitability for confirmation or prediction of H, Blazˇkova H and Milec Z (2003) PCR markers of apple possible parentage of apple varieties. resistance to scab (Venturia inaequalis CKE.) controlled (3) Old local Latvian apple varieties do not contain the by Vf gene in Czech apple breeding. Plant, Soil and Vf resistance gene. Environment 49: 427–432. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 193–196 ISSN 1479-2621 doi:10.1017/S1479262111000396

Molecular adaptation of the chloroplast matK gene in Nymphaea tetragona, a critically rare and endangered plant of India

Jeremy Dkhar, Suman Kumaria* and Pramod Tandon Plant Biotechnology Laboratory, Centre for Advanced Studies in Botany, North Eastern Hill University, Shillong 793022, India

Abstract Sustainable utilization of plant genetic resources for food and agriculture has been increasingly discussed at both national and international forums. Besides exploitation, conservation of plant genetic resources has become an integral part of these discussions. Conservation aims at maintaining the diversity of living organisms, their habitat and the interrelationship between organisms and their environment. For achieving such goals, appropriate conservation strategies have to be adopted. Determining the genetic makeup of a particular plant species is of critical importance when planning a suitable conservation strategy. In this study, we sequenced the chloroplast trnK intron, matK and rbcL gene aimed at understanding the rarity of Nymphaea tetragona, a critically rare and endangered plant of India found at only one location. We extended our investigation to other Nymphaea species such as N. nouchali, N. pubescens and N. rubra that are commonly available throughout India. Interestingly, matK gene of N. tetragona revealed high number of non-synonymous substitutions. Molecular evolutionary analysis indicated that three of these sites may be under mild selective pressures. Such adaptive changes at the DNA and protein sequence level of matK gene may have been associated with the colonization of N. tetragona, suggesting that it could have migrated from China.

Keywords: matK; molecular adaptation; molecular evolutionary analysis; Nymphaea tetragona

Introduction Nymphaea viz. N. nouchali, N. pubescens, N. rubra and N. tetragona based on internal transcribed spacers Nymphaea tetragona represents one of the naturally (ITS), trnK intron and matK gene was reported (Dkhar occurring species found in India (Mitra, 1990). Its distri- et al., 2010). Molecular evidence suggested probable bution is restricted to one particular location found at misidentification of one specimen of N. nouchali and Nongkrem, Meghalaya, India (258280N–918520E). It is con- N. tetragona. Furthermore, a comparison of the genetic sidered to be rare and endangered and has been included closeness among the Indian, Chinese and Russian material for recovery programmes (Ganeshaiah, 2005). Besides of N. tetragona based on the ITS region was also evaluated. N. tetragona, there are an additional nine species (both The results indicated a close relationship between the wild and cultivated) of the genus Nymphaea occurring Indian and the Chinese representatives, which is further in India (Mitra, 1990). However, traditional classification supported by the morphological similarities shared of the genus in India has received unconvincing treat- between them (eFloras, 2008). ments with some names inaccurately used (Cook, An interesting observation made from the previous 1996). To vindicate Cook’s remarks, molecular taxonomic study was the relatively higher number of nonsynon- revision of four Indian representatives of the genus ymous substitutions as compared to synonymous substitutions detected in the matK gene of N. tetragona. The higher rate of nonsynonymous substitutions is usually inferred as an indication of selective pressures acting on * Corresponding author. E-mail: [email protected] protein-coding sequences. The chloroplast matK gene, 194 J. Dkhar et al. coding for the maturase enzyme, has been reported to Results and discussion evolve under positive selection in some lineages of land plants (Hao et al., 2010). Hao et al. (2010) showed Positive selection in matK gene that several regions (N-terminal region, RT domain and domain X) of matK experience molecular adaptation, Maximum likelihood analysis of the combined datasets which fine-tunes maturase performance. yielded a phylogenetic tree with log likelihood value In the present study, to gain insight into whether an of 2 10 921.95 (Fig. 1). The reconstructed phylogenetic adaptive evolutionary process is linked to the coloniza- tree was identical in topology to those obtained in a tion of new habitats by N. tetragona, we tested whether study by Lo¨hne et al. (2007). there was selective pressure in matK using phylogenetic For detecting positive selection using codon substi- and maximum likelihood analyses of codon and branch tution models, three LRTs were performed comparing site substitution models. This estimation was extended M1A (nearly neutral) with M2A (positive selection), M7 to members of the order Nymphaeales, not included in (b) with M8 (b and vs $ 1), and M8A (b and vs ¼ 1) the study of Hao et al. (2010), representing a lineage at with M8 (b and vs $ 1). Because positive selection oper- the base of the angiospermic tree. ates more strongly at some sites (Aguileta et al., 2009), the distribution of sites influenced by selection for all three regions (N-terminal region, RT domain and Materials and methods domain X) of the matK gene was evaluated. For N-terminal region, model M2A indicated that 13.2% of the sites Taxon sampling and nucleotide sequence data are under positive selection (v ¼ 1.778). In model M8, 13.0% of the sites are under selective pressure Sequence data of matK gene and trnK intron of four wild (v ¼ 1.791). For RT domain, model M8 estimated an v Indian representatives of the genus Nymphaea, viz. value of 46.57 for 0.01% of codon sites. No positive sites N. nouchali, N. pubescens, N. rubra and N. tetragona were detected for domain X, a relatively more conserved was taken from our own previous study (Dkhar et al., and important functionally, as it determines the maturase 2010). The rbcL gene was amplified and sequenced follow- activity of matK proteins (Hausner et al., 2006). Two ing the method adopted in another study (Dkhar et al., LRTs, M7-M8 and M8A-M8, were significant at the 0.01 in press). In addition, barring Nymphaea, nucleotide and 0.05 significance level, respectively. Comparing M8 sequence data of members of the order Nymphaeales against the null model M8A is more robust than M1A (Barclaya longifolia, Euryale ferox, Nuphar advena, versus M2A or M7 versus M8 (Swanson et al., 2003). Ondinea purpurea and Victoria cruziana), Amborella trichopoda and Austrobaileya scandens, identified as the BEB analysis Positive sites for foreground most basal angiosperms, were retrieved from GenBank. lineages 100 N. pubescens (N. tetragona) Prob (w>1) N. rubra 1 93 E 0.607 86 2 132 H 0.527 V. cruziana 3 150 F 0.582 Molecular evolutionary analysis 94 E. ferox N. nouchali JD 07 100 100 In order to conduct phylogenetic maximum likelihood 69 N. nouchali JD 06 analysis of selection in chloroplast genes, a phylogenetic O. purpurea 89 99 tree was reconstructed based on the chloroplast trnK N. nouchali JD 02 GE QH LF 72 N. tetragona intron, matK and rbcL genes using maximum likelihood 123 method implemented in Phylip (Felsentein, 1989). B. longifolia 100 Molecular adaptation tests on the chloroplast matK N. advena codon sites were performed using maximum likelihood Brasenia schreberi models and programs included in PAML ver. 4.3 (Yang, 100 Cabomba caroliniana 2007). We performed five site-specific codon substitution A. scandens Outgroup models: null models for testing positive selection (M1A, A. trichopoda M7 and M8A) and models allowing for positive selection Fig. 1. Maximum likelihood tree of combined chloroplast (M2A and M8). Furthermore, to provide evidence DNA dataset (3859 bp) using Phylip. Three positive sites whether positive selection is linked to the colonization with probability for foreground (N. tetragona) estimated of N. tetragona, the branch site models (model A with v using branch-site models of PAML are indicated. Amino- acid replacements at these sites are also indicated at the fixed or estimated) implemented in PAML ver 4.3 were branch (G: Glycine, E: Glutamic acid, Q: Glutamine, used. The likelihood ratio test (LRT) was used to compare H: Histidine, L: Leucine, F: Phenylalanine). Numbers at these alternative models. nodes indicate bootstrap values. Prob, probability. Adaptation of matK gene in Nymphaea tetragona 195

Table 1. Estimation of log-likelihood (LnL) values and parameters for each region of the matK gene using branch-site models implemented in PAML selecting N. tetragona as foreground branch.

N-terminal region RT domain Domain X Model LnL Parameters LnL Parameters LnL Parameters Foreground: N. tetragona Model A 23148.00 p0 ¼ 0.73 633, 2229.30 p0 ¼ 1.00 000, 21250.61 p0 ¼ 0.67 267, (v2 ¼ 1 ﬁxed) v0 ¼ 18 596 v0 ¼ 0.32 206 v0 ¼ 0.11 716 p1 ¼ 0.25 529, p1 ¼ 0.00 000, p1 ¼ 0.32 733, v1 ¼ 1.00 000 v1 ¼ 1.00 000 v1 ¼ 1.00 000 p2 þ p3 0.00 839, p2 þ p3 ¼ 0.00 000, p2 þ p3 ¼ 0.00 000, v2 ¼ 1.00 000 v2 ¼ 1.00 000 v2 ¼ 1.00 000 Model A 23147.80 p0 ¼ 0.73 321, 2229.30 p0 ¼ 1.00 000, 21250.61 p0 ¼ 0.67 267, (v2 ¼ estimated) v0 ¼ 0.18 697 v0 ¼ 0.32 206 v0 ¼ 0.11 716 p1 ¼ 0.24 589, p1 ¼ 0.00 000, p1 ¼ 0.32 733, v1 ¼ 1.00 000 v1 ¼ 1.00 000 v1 ¼ 1.00 000 p2 þ p3 ¼ 0.0209, p2 þ p3 ¼ 0.00 000, p2 þ p3 ¼ 0.00 000, v2 ¼ 23.09 944 v2 ¼ 1.17 397 v2 ¼ 1.00 000 p, proportion of codon sites with v; v,dN/dS.

Similarly, branch-site model A (v2 ¼ 1 ﬁxed) with critically rare and endangered plant of India. Such adaptive branch-site model A (v2 estimated) were compared. changes at the molecular level may have been associated For N-terminal region, model A (v2 estimated) for with the adaptation of N. tetragona to varying ecological N. tetragona branch indicated that 2.09% of the sites are conditions. So, the threat to the existence of this plant under selective pressure with v2 ¼ 23.099 (Table 1). taxon is primarily anthropogenic, brought about through Bayes Empirical Bayes (BEB) analysis showed three large-scale cultivation at the vicinity of the location. positive sites with probability value of 0.607, 0.527 and Conservation of N. tetragona can be targeted at controlling 0.582, respectively. such activities or the probable translocation to another site.

Selective pressure during colonization of N. tetragona Acknowledgements

N. tetragona is well distributed globally, found through- The authors acknowledge financial support from out China, Japan, Finland and Russia; but its location in Department of Biotechnology, Ministry of Science and India is restricted to one particular population. Our pre- Technology, India (file no. BT/PR-7055/BCE/08/437/2006). vious study indicated that the Indian plant is closely related to plants found in China (Dkhar et al., 2010). Molecular evolutionary analysis suggested that the plant References taxon migrated from China, and the migratory processes involved might have brought about some changes at both Aguileta G, Refre´gier G, Yockteng R, Fournier E and Giraud T the DNA and protein sequence levels. This adaptive (2009) Rapidly evolving genes in pathogens: methods for response would have rendered slight advantage for the detecting positive selection and examples among fungi, bacteria, viruses and protests. Infection, Genetics and plant to survive in new habitat. Such adaptive changes Evolution 9: 656–670. at the DNA and protein sequence levels of rbcL gene Cook CDK (1996) Aquatic and Wetland Plants of India. Oxford: have been reported in Schidea (Kapralov and Filatov, Oxford University Press. 2006) and Potamogeton (Iida et al., 2009). Dkhar J, Kumaria S, Rama Rao S and Tandon P (2010) Molecular phylogenetics and taxonomic reassessment of four Indian representatives of the genus Nymphaea. Aquatic Botany 93: 135–139. Implications for conservation of N. tetragona Dkhar J, Kumaria S and Tandon P Molecular evidence suggests that an Indian plant called Nymphaea alba var. rubra is a Determining the genetic makeup of a particular plant hybrid originating from N. alba and N. odorata. Annales Botanici Fennici (in press). species is of critical importance when planning a suitable eFloras (2008) Published on the Internet http://www.efloras.org conservation strategy. Here, we provided evidence for Missouri Botanical Garden, St. Louis, MO and Harvard Uni- molecular adaptation of matK gene in N. tetragona, a versity Herbaria, Cambridge, MA. Accessed 16 March 2010 196 J. Dkhar et al. Felsentein J (1989) Phylip-phylogeny inference package Kapralov MV and Filatov DA (2006) Molecular adaptation during (version 3.2). Cladistics 5: 164–166. adaptive radiation in the Hawaiian endemic genus Ganeshaiah KN (2005) Recovery of endangered and threatened Schiedea. PLoS ONE 1: e8. doi:10.1371/journal.pone0000008 species. Developing a national priority list of plants and Lo¨hne C, Borsch T and Wiersema JH (2007) Phylogenetic anal- insects. Current Science 89: 599–600. ysis of Nymphaeales using fast-evolving and noncoding Hao DC, Chen SL and Xiao PG (2010) Molecular evolution and chloroplast markers. Botanical Journal of the Linnean positive Darwinian selection of the chloroplast maturase Society 154: 141–163. matK. Journal of Plant Research 123: 241–247. Mitra RL (1990) Nymphaeaceae. In: Nayar MP, Thothathri K and Hausner G, Olson R, Simon D, Johnson I, Sanders ER, Karol KG, Sanjappa M (eds) Fascicles of flora of India, Fascicles 20. McCourt RM and Zimmerly S (2006) Origin and evolution of the chloroplast trnK (matK) intron: a model for evol- Howrah: Botanical Survey of India, pp. 11–25. ution of group II intron RNA structures. Molecular Biology Swanson WJ, Nielsen R and Yang Q (2003) Pervasive adaptive and Evolution 23: 380–391. evolution in mammalian fertilization proteins. Molecular Iida S, Miyagi A, Aoki S, Ito M, Kadono Y and Kosuge K (2009) Biology and Evolution 20: 18–20. Molecular adaptation of rbcL in the heterophyllous Yang Z (2007) PAML4: a program package for phylogenetic aquatic plant Potamogeton. PLoS ONE 4: e4633. doi:10.1371/ analysis by maximum likelihood. Molecular Biology and journal.pone0004633 Evolution 24: 1586–1591. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 197–201 ISSN 1479-2621 doi:10.1017/S1479262111000190

The genetic make-up of the European landraces of the common bean

S. A. Angioi1, D. Rau1, L. Nanni2, E. Bellucci2, R. Papa2,3 and G. Attene1,4* 1Dipartimento di Scienze Agronomiche e Genetica Vegetale Agraria, Universita` degli Studi di Sassari, Sassari, Italy, 2Dipartimento di Scienze Ambientali e delle Produzioni Vegetali, Universita` Politecnica delle Marche, Ancona, Italy, 3CRA-CER Council for Agricultural Research – Cereal Research Centre, S.S. 16, Km 675, 71122 Foggia, Italy and 4Centro per la Conservazione e Valorizzazione della Biodiversita` Vegetale, Universita` degli Studi di Sassari, Surigheddu, Alghero, Italy

Abstract Here, we present a brief overview of the main studies conducted on the common bean (Phaseolus vulgaris L.) in Europe and other countries outside its centres of origin. We focus on the proportions of the Andean and Mesoamerican gene pools, and on the inter-gene pool hybridization events. In Europe, for chloroplast microsatellites, 67% of European germplasm is of Andean origin. Within Europe, interesting trends have been seen; indeed, the majority of the Andean type is found in the three macro-areas of the Iberian Peninsula, Italy and central-northern Europe, while, in eastern and south-eastern Europe, the proportion of the Mesoamerican type increased. On a local scale, the contribution of the Mesoamerican type is always low. On other continents, various situations are seen using different markers: in China and Brazil, the Mesoamerican gene pool prevails, while in an African sample, overall, both gene pools are equally represented, with differences in individual countries. The frequency of European bean genotypes deriving from at least one hybridization event was 44% with an uneven distribution. Interestingly, hybrids tend to have intermediate seed size in comparison with ‘pure’ Andean or Mesoamerican types. On other continents, very few hybrids are found, probably because of the different marker systems used.

Keywords: Introgression; Mesoamerican and Andean gene pools; Phaseolus vulgaris

Introduction seed-storage proteins) and by various marker systems (Beebe et al., 2001; Papa and Gepts, 2003; Kwak and The common bean (Phaseolus vulgaris L.) is of great Gepts, 2009; Rossi et al., 2009). The Mesoamerican agronomic interest worldwide, and represents 50% of types are small or medium seeded, with phaseolins grain legumes for direct human consumption (McClean S or B; while the Andean are large seeded, with phaseo- et al., 2004). Domestication of P. vulgaris occurred lins T, C, H and A (Gepts et al., 1986; Singh et al., independently in the Mesoamerican and Andean areas, 1991). Based on the occurrence of a strong bottleneck which gave rise to two highly differentiated gene in the Andean wild population, Rossi et al. (2009) pools (Gepts and Debouck, 1991; Gepts, 1998). The recently suggested a Mesoamerican origin of the two gene pools can be distinguished according to mor- common bean. phological traits (Gepts et al., 1986), phaseolins (major The common bean was introduced into Europe after Columbus’s voyages. It was distributed widely in all parts of Europe, where many landraces and varieties evolved, as they were grown to provide dry seeds or * Corresponding author. E-mail: [email protected] fresh pods (Zeven, 1997). 198 S. A. Angioi et al. Here, we present a brief overview regarding the level f f e and structure of the genetic diversity of the European b d common bean, and we compare this with information Ethiopia, g provided for other continents. Hybrids (%)

Composition of the European common bean ) H Overall, outside the domestication centres of the Data obtained through com- common bean (Table 1), the proportions of the Andean d and Mesoamerican gene pools vary considerably across different countries and continents. In Europe, studies have been conducted on different scales: continental, macro-areas, single country and local (within country). The first studies on large collections of the European common bean were carried out by using phaseolins. These demonstrated that the germplasm arose from both of the American gene pools, As defined by Tautz 1998. with a higher frequency of Andean types (76–66%) c (Gepts and Bliss, 1988; Lioi, 1989). This prevalence (76%) was confirmed in a large collection that included, for the first time, central European countries (Logozzo et al., 2007). Angioi et al. (2010) used chloroplast microsatellites Intermediate genotypes identified in the neighbour-joining tree. 6721 33 79 0.36 0.30 0.34 0.33 44 4 (cpSSRs), two nuclear markers and morphological f Gene pool composition (%) Genetic diversity ( seed traits to analyze a large part of the collection of Andean Mesoamerican Andean Mesoamerican Logozzo et al. (2007), while adding new accessions from countries previously less represented (e.g. France). Also for this analysis, the prevalence (67%) of European germplasm was of Andean origin (Fig. 1). Within Europe, an interesting trend was seen, as the c Andean type was in the majority in three macro-areas: a the Iberian Peninsula, Italy and central-northern

Europe. In contrast, in eastern and south-eastern Molecular/ biochemical marker Europe, the proportion of the Mesoamerican type increased (Fig. 1). This was supported by other studies in the Iberian Peninsula (Rodin˜o et al., 2003; Ocampo et al., 2005), and at a country level in Greece (46%; Lioi, 1989) and Bulgaria (79%; Svetleva et al., 2006) (see Papa et al., 2006, for review). In Italy, on a local scale, the prevalence of the

Andean type has been conﬁrmed. The contribution of Data obtained through comparison with nuclear STS and phaseolins. b the Mesoamerican gene pool varied from 5% in Sardi- nian landraces (Angioi et al., 2009) to 29% in the (2010) 307 cpSSR (2008) 299 nuSSR 25 75 0.42 0.37 5 (2009)(2009) 99 89 nuSSR nuSSR 28 73 72 27 0.44 0.45 0.59 0.59 1 0 (2010) 279 nuSSR Marche region (Sicard et al., 2005), according to chlor- (2010) 355 nuSSR 40 60 – – 10 Data obtained through comparison with morphological data. e

oplast and nuclear markers, and 12 and 13% in Abruzzo et al. et al. et al. et al. et al. and Basilicata, respectively (Piergiovanni et al., 2000a, et al. b), according to phaseolins. References Sample size Looking at other continents, the Mesoamerican gene Gepts and Bliss (1988) 111 phaseolin 81 19 – – – (2001), for a review. pool prevailed in a Chinese collection (Zhang et al., g Gene pool compositions (% of Andean and Mesoamerican), gene diversities (Nei, 1978) and % hybrids in European sample, compared with countries on

2008) and a Brazilian collection (Burle et al., 2010) et al. (Table 1 and Fig. 1). The Mesoamerican predominance in Brazil is surprising, given its close proximity to Provan Table 1. other continents BrazilChinaEthiopiaKenya Burle East Africa Zhang Asfaw a Asfaw Rwanda Blair Europe Angioi parison with phaseolins. the Andes. Multiple introductions of Mesoamerican Kenya, Uganda, Burundi, Rwanda, Democratic Republic of Congo (former Zaire), Tanzania, Malawi and Zambia. Genetic make-up of the European common bean 199 Europe* A

33% 67%

China** F A A 25% 75% A A A

C 29% 71%

Ethiopia** Rwanda** D 40% 60% Kenya** Brazil** C E 27% 21% 73% 79% Overall Africa

East Africa*** 50% 50% B 19% 81%

Fig. 1. Distribution map of the Andean and Mesoamerican gene pools in Europe and on other continents, analyzed with molecular markers. In the pie charts: white, Andean gene pool; black, Mesoamerican gene pool. (A) Europe (sample size, n ¼ 307) and the Iberian Peninsula (53), Italy (32), central-northern Europe (74), eastern Europe (69), south-eastern Europe (79); Angioi et al. (2010). (B) east Africa (111), Gepts and Bliss (1988); (C) Ethiopia (99) and Kenya (89), Asfaw et al. (2009); (D) Rwanda (355), Blair et al. (2010); (E) Brazil (279), Burle et al. (2010); (F) China (299), Zhang et al. (2008). *cpSSRs, **nuSSRs, ***phaseolins. The overall African pie chart is obtained pooling together the data provided in Table 1.

germplasm and similarities in climate and soil between macro-areas and/or homogenizing selection (anthropic Brazil and Mesoamerica might have had a considerable and natural) has been suggested (Angioi et al., 2010). impact in establishing this pattern (Burle et al., 2010). The role of selection is suggested by the findings of In Africa, overall, both gene pools are equally rep- Logozzo et al. (2007) in European landraces, where resented (Fig. 1), although at the single country level, accessions with Mesoamerican phaseolin had signifi- contrasting situations are seen. In Ethiopia (Asfaw cantly larger seed size than individuals from America in et al., 2009) and Rwanda (Blair et al., 2010), the Mesoa- the same phaseolin class. merican types predominate, and vice versa in Kenya (Asfaw et al., 2009) and East Africa (Gepts and Bliss, 1988) (Table 1 and Fig. 1). This suggested that the gen- Introgression between the gene pools etic divergence in bean landraces might be due to the original differences in the introduced germplasm from Introgression is an event that arises from hybridization the centres of origin (Asfaw et al., 2009), combined among gene pools, as spontaneous outcrossing in with differences in pest resistance (e.g. Mesoamerican farmer fields followed by selection for adaptation to pro- types resistant to root rot in Rwanda) and production duction niches and uses. Comparing chloroplast data (Mesoamerican genotypes have the highest yields). with nuclear (phaseolin and sequence-tagged sites, STS) Finally, gene flow between Ethiopia and Kenya was and morphological data, Angioi et al. (2010) estimated moderate, probably due to different farmer preferences that a high proportion of the European bean germplasm according to ecological adaptation, cooking values and (44%) derived from at least one hybridization event using market orientation (Asfaw et al., 2009). This is in contrast a maximum-likelihood approach. Although hybrids are to Europe. Indeed, in Europe, high gene flow among present everywhere, they show uneven distributions, 200 S. A. Angioi et al. with high frequencies in central Europe and low frequen- Angioi SA, Rau D, Attene G, Nanni L, Bellucci E, Logozzo G, cies in the Iberian Peninsula and Italy. A comparison of Negri V, Spagnoletti Zeuli PL and Papa R (2010) Beans in Europe: origin and structure of the European landraces of chloroplast data with nuclear and morphological data is Phaseolus vulgaris L. Theoretical and Applied Genetics a reliable method to identify the hybrids, as they tend 121: 829–843 doi: 10.1007/s00122-010-1353-2. to have intermediate seed size with respect to ‘pure’ Asfaw A, Blair MW and Almekinders C (2009) Genetic diversity Andean or Mesoamerican, with Andean £ Mesoamerican and population structure of common bean (Phaseolus seeds smaller than ‘pure’ Andean, and Mesoamerican £ vulgaris L) landraces from the East African highlands. Andean seeds larger than ‘pure’ Mesoamerican. This Theoretical and Applied Genetics 120: 1–12. Beebe S, Rengifo J, Gaitan E, Duque MC and Tohme J (2001) method was also applied at local scales in Italy, in the Diversity and origin of Andean landraces of common Marche region (12% hybrids; Sicard et al., 2005) and bean. Crop Science 41: 854–862. Sardinia (4%; Angioi et al., 2009). Blair MW, Gonza´lez LF, Kimani M and Butare L (2010) Genetic Other studies have analyzed hybridization among gene diversity, inter-gene pool introgression and nutritional pools, but they found few hybrids: 4% in Brazil, compar- quality of common beans (Phaseolus vulgaris L.) from Cen- tral Africa. Theoretical and Applied Genetics 121: 237–248. ing nuclear SSRs (nuSSRs) to phaseolins (Burle et al., Burle ML, Fonseca JR, Kami JA and Gepts P (2010) Microsatellite 2010), and from 1 to 10% in Ethiopia and Rwanda diversity and genetic structure among common bean (Asfaw et al., 2009; Blair et al., 2010), considering (Phaseolus vulgaris L.) landraces in Brazil, a secondary individuals intermediate among gene pools in the neigh- center of diversity. Theoretical and Applied Genetics 121: bour-joining tree (Table 1). The differences in hybrid 801–813 doi: 101007/s00122-010-1350-5. Gepts P (1998) Origin and evolution of common bean: past frequency can be explained by different marker systems events and recent trends. Horticultural Science 33: used (chloroplast and nuclear) and the definition of 1121–1130. hybrids as recent (Asfaw et al., 2009; Blair et al., 2010) Gepts P and Bliss FA (1988) Dissemination pathways of versus old generation hybrids (Angioi et al., 2010). common bean (Phaseolus vulgaris, Fabaceae) deduced Another explanation might be that, as seen in Brazil, from phaseolin electrophoretic variability. II Europe and the frequency of the two gene pools are very different, Africa. Economic Botany 42: 86–104. Gepts P and Debouck DG (1991) Origin, domestication and or that in some environments, there is no flowering syn- evolution of common bean, Phaseolus vulgaris. In: chronization between Andean and Mesoamerican types Schoonhoven A and Voysest O (eds) Common Beans: (Asfaw et al., 2009). In the Chinese sample, Zhang et al. Research for Crop improvement. CAB International (2008) found 5% hybrids, noting that average seed weight Wallingford, UK, pp 4–54. of the Andean types was lower than that of the American Gepts P, Osborne TC, Rashka K and Bliss FA (1986) Electrophor- etic analysis of phaseolin protein variability in wild forms Andean beans, with the opposite for the Mesoamerican and landraces of the common bean Phaseolus vulgaris L.: Chinese bean. evidence for two centers of domestications. Economic The existence of the high frequency of inter-gene pool Botany 40: 451–468. hybridization in Europe might have had a significant Johnson WC and Gepts P (1999) Segregation for performance in impact on the structure of the genetic and genotypic recombinant inbred populations resulting from inter-gene diversity in the nuclear genome. This is consistent pool crosses of common bean (Phaseolus vulgaris L.). Euphytica 106: 5–56. with the breakdown of geographical isolation between Johnson WC and Gepts P (2002) The role of epistasis in control- the two gene pools (Angioi et al., 2010). Moreover, the ling seed yield and other agronomic traits in an Andean £ European hybrids appear to be of great importance Mesoamerican cross of common bean (Phaseolus vulgaris for breeding that aims to recombine Andean and L.). Euphytica 125: 69–79. Mesoamerican traits (Johnson and Gepts, 1999, 2002). Kwak M and Gepts P (2009) Structure of genetic diversity in the two major gene pools of common bean (Phaseolus vulgaris L., Fabaceae). Theoretical and Applied Genetics 118: 979–992. Lioi L (1989) Geographical variation of phaseolin patterns in an old world collection of Phaseolus vulgaris. Seed Science Acknowledgements and Technology 17: 317–324. Logozzo G, Donnoli R, Macaluso L, Papa R, Knupffer H and This study was partially supported by the Italian Govern- Spagnoletti Zeuli PL (2007) Analysis of the contribution ment (MIUR) grant no. 20083PFSXA PRIN 2008. of Mesoamerican and Andean gene pools to European common bean (Phaseolus vulgaris L.) germplasm and strategies to establish a core collection. Genetic Resources and Crop Evolution 54: 1763–1779. References McClean PE, Lee RK and Miklas PN (2004) Sequence diversity analysis of dihydroflavonol 4-reductase intron 1 in Angioi SA, Desiderio F, Rau D, Bitocchi E, Attene G and Papa R common bean. Genome 47: 266–280. (2009) Development and use of chloroplast microsatellites Nei M (1978) Estimation of average heterozygosity and genetic in Phaseolus spp. and other legumes. Plant Biology 11: distance from a small number of individuals. Genetics 89: 598–612. 583–590. Genetic make-up of the European common bean 201 Papa R and Gepts P (2003) Asymmetry of gene flow and differ- Rossi M, Bitocchi E, Bellucci E, Nanni L, Rau D, Attene G and ential geographical structure of molecular diversity in wild Papa R (2009) Linkage disequilibrium and population and domesticated common bean (Phaseolus vulgaris L.) structure in wild and domesticated populations of Phaseo- from Mesoamerica. Theoretical and Applied Genetics 106: lus vulgaris L. Evolutionary Applications 2: 504–522. 239–250. Sicard D, Nanni L, Porfiri O, Bulfon D and Papa R (2005) Papa R, Nanni L, Sicard D, Rau D and Attene G (2006) The Genetic diversity of Phaseolus vulgaris L and P. coccineus evolution of genetic diversity in Phaseolus vulgaris L. L. landraces in central Italy. Plant Breeding 124: 464–472. In: Motley TJ, Zerega N and Cross H (eds) New Approaches Singh SP, Nodari R and Gepts P (1991) Genetic diversity in to the Origins, Evolution and Conservation of Crops. cultivated common bean. I. Allozymes. Crop Science 31: Darwin’s Harvest. New York: Columbia University Press. 19–23. Piergiovanni AR, Cerbino D and Brandi M (2000a) The common Svetleva D, Pereira G, Carlier J, Cabrita L, Leitaõ J and Genchev bean populations from Basilicata (southern Italy). An D (2006) Molecular characterization of Phaseolus vulgaris evaluation of their variation. Genetic Resources and Crop L. genotypes included in Bulgarian collection by ISSR and Evolution 47: 489–495. AFLPe analyses. Scientia Horticulturae 109: 198–206. Piergiovanni AR, Taranto G and Pignone D (2000b) Diversity Tautz D (1998) Hypervariability of simple sequences as a among common bean populations from the Abruzzo general source for polymorphic DNA markers. Nucleic region (central Italy): a preliminary inquiry. Genetic Acids Research 17: 6463–6471. Resources and Crop Evolution 47: 467–470. Zeven AC (1997) The introduction of the common bean Provan J, Powell W and Hollingsworth PM (2001) Chloroplast (Phaseolus vulgaris L.) into Western Europe and the phe- microsatellites: new tools for studies in plant ecology notypic variation of dry bean collected in the Netherlands and evolution. Trends in Ecology and Evolution 16: in 1946. Euphytica 94: 319–328. 142–147. Zhang X, Blair MW and Wang S (2008) Genetic diversity of Rodinõ AP, Santalla M, De Ron AM and Singh SP (2003) A core Chinese common bean (Phaseolus vulgaris L.) landraces collection of common bean from the Iberian peninsula. assessed with simple sequence repeats markers. Theoretical Euphytica 131: 165–175. and Applied Genetics 117: 629–640. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 202–205 ISSN 1479-2621 doi:10.1017/S1479262111000323

Investigation of genetic diversity in Russian collections of raspberry and blue honeysuckle

Didier Lamoureux1, Artem Sorokin2, Isabelle Lefe`vre1, Sergey Alexanian2, Pablo Eyzaguirre3 and Jean-Franc¸ois Hausman1* 1Department of Environment and Agro-biotechnologies, Centre de Recherche Public – Gabriel Lippmann, Belvaux, Luxembourg, 2Department of Fruit Crops Genetic Resources, N.I. Vavilov Institute of Plant Industry, St. Petersburg, Russian Federation and 3Regional Ofﬁce for Europe, Bioversity International, Rome, Italy

Abstract The N.I. Vavilov Institute of Plant Industry (VIR) holds and maintains collections of various crop plants among the largest and oldest worldwide. Among them, small berry trees have gained attention because of their potential for human health. Small berries, usually containing various valuable compounds such as vitamins or antioxidants in significant quantities, could be used for easily improving the human diet. Subsets of VIR collections of raspberry (Rubus idaeus L.) and blue honeysuckle (Lonicera caerulea L.) were investigated for genetic diversity. Ninety-five raspberry accessions were genotyped with eight nuclear simple-sequence repeat (microsatellite) markers. Results indicated a fair level of genetic diversity, but also a structure of three main groups in the collection. Blue honeysuckle accessions were genotyped with five intersimple-sequence repeat markers, yielding more than 1100 polymorphic fragments across the 194 accessions. Statistical analysis of these data showed that the subspecies level was key in explaining blue honeysuckle diversity. This study shows that the collections constitute important resources that could be used for either direct consumption goals or breeding of new cultivars. Results may also be used to establish recommendations for efficient conservation of these genetic resources.

Keywords: genotyping; germplasm collections; Lonicera caerulea L.; molecular markers; Rubus idaeus L.

Introduction represented in this article. The VIR Rubus collection is rich with about 250 accessions. VIR started to gather The N.I. Vavilov Institute of Plant Industry (VIR), Lonicera accessions in the middle of the 20th century. established in 1894, is the lead institution in the Russian The collection now includes 420 accessions and it is the Federation with the overall mandate for the conservation most comprehensive Lonicera collection in the world, of plant genetic resources for agriculture and forestry. especially when it comes to wild species. At present, VIR holds one of the world’s oldest and The red raspberry (Rubus idaeus L.) is a well-known largest collection of crop plants and their wild relatives fruit tree from the Rosaceae family. It is widely that consists of 320,000 accessions belonging to 2532 cultured and almost 460,000 tons of raspberries were species (http://www.vir.nw.ru). Berries are, of course, produced in 2008 (source: The Food and Agricultural Organization Statistical Database). Although raspberry has been bred for decades, it is only recently that new molecular tools were applied to this species (Graham * Corresponding author. E-mail: [email protected] et al., 2004). Genetic diversity in raspberry and honeysuckle 203 The genus Lonicera L. belongs to the Caprifoliaceae Material and methods family, and comprises about 200 different species. Among them, only a few species, usually named blue Dried young leaves of 95 Rubus accessions and 194 honeysuckle, have edible berries. All of them belong Lonicera accessions were submitted to standard DNA to subsection Caerulea Rehd., but their taxonomy is isolation procedure with cetyl trimethylammonium difficult because of different interpretations of species, bromide. Similar quantities of DNA were PCR-amplified subspecies and varieties (Naugzˇemys et al., 2007). Blue with published Rubus microsatellite-flanking primers honeysuckle (Lonicera caerulea L.) originated from or anonymous microsatellite primers for Lonicera the Baltic region to China, Japan and Russian Far East. (Supplementary Table S1, available online only at http:// Plants present exceptional cold hardiness; they fre- journals.cambridge.org). Fluorescent-tagged amplified quently can tolerate 2 408C, while flowers can stand fragments were separated on a capillary electrophoresis spring frost as low as 2 88C (Plekhanova, 2000). 3130 Genetic Analyzer (Applied Biosystems Inc., Foster Depending on the variety, ovoid to cylindrical berries City, CA, USA). Resulting peaks were calculated using may reach 1.5 g, and are dark blue to purple with a the GeneMapperw v4.0 software (Applied Biosystems). blue wax coating (Plekhanova, 2000). The taste may Genotyping results were analyzed using Structure software vary from sweet to sour or bitter. Blue honeysuckle is (Falush et al., 2007) and DARwin5 software (Perrier and known as edible berry in Russia, China, Japan, and Jacquemoud-Collet, 2006). has recently been introduced in North America (Chaova- nalikit et al., 2004). The goal of this study was to investigate the genetic Results and discussion diversity available in the VIR collections of raspberry and blue honeysuckle through molecular analysis of Raspberry nuclear DNA. This work, by improving the global level of characterization of the collections, was expected to The eight chosen simple-sequence repeat markers were help (1) in the management of the collections and used for genotyping on the 95 accessions set, which (2) in promoting the use of these valuable resources for included two accessions of wild species (R. odoratus human nutrition and health purposes. and R. nutcanus) in addition to the 93 R. idaeus

0 0.5 Fig. 1. Neighbour-Joining tree of raspberry accessions, indicated by main contribution from three hypothetical ancestral populations. Stars indicate wild species. 204 D. Lamoureux et al. accessions. It is noticeable that 28 accessions could not These results highlight the complex structure of the be unambiguously distinguished from the others. The collection, reflecting the multiple possibilities of cross- polymorphic information content (PIC) was calculated hybridization of raspberry with related species. With the 2 for each marker as PIC ¼ 1 2 S( pi ), where pi is the help of more molecular markers, it would probably be frequency of the ith allele from the sampled accessions. possible to distinguish between all the investigated The mean PIC value for the eight markers was 0.626, accessions. However, additional data such as passport indicating a fair level of genetic diversity in the sampled data, morphological data or biochemical data are collection subset. obviously required to address the issue of the structure The structure of the collection subset was investigated of the collection. using Structure2.2 software (Falush et al., 2007). Computations did not point at a clear number of subpopulations: the probabilities for a number K of Blue honeysuckle subpopulations were not statistically different from K ¼ 2 to 4, although the highest probability was obtained The five used intersimple-sequence repeat markers for K ¼ 3 (data not shown). yielded several thousands of scorable fragments with Based on the same dataset, a Neighbour-Joining tree our 194 accessions subset. In order to ensure accuracy was calculated to represent diversity (Fig. 1; detailed results and reproducibility, we restricted to the top 10% of are available on request to the corresponding author and most-intense fragments; that resulted in 1158 scored frag- will be available soon on VIR website). The tree was ments. Every accession was unambiguously distinguished coloured for each accession, according to the main con- from the others. tribution of the hypothetical subpopulations above for A Neighbour-Joining tree was calculated from the K ¼ 3. In that case, three relatively homogenous clusters genotyping data. The analysis of the tree reveals that (top left, top right and bottom, Fig. 1) can be distinguished. Lonicera accessions are mainly grouped by subspecies However, a fourth group (top middle) appears as a mix (Fig. 2). The cluster with a majority of accessions of the of the three other groups. The wild species accessions, diploid species L. boczkarnikovae and L. edulis is quite although included in one group, are obviously quite distant specific The remaining of the tree is made of accessions of from the other accessions. L. caerulea subspecies, all of them being reputed tetraploids.

L. caerulea (4x) Altaica Kamtschatica Pallasii Venulosa Stenantha Altaica Diploids Altaica x kamtschatica Venulosa x kamtschatica Pallasii Stenantha Emphyllocalyx L. boczkarnikovae (2x) L. edulis (2x) Venulosa

Kamtschatica 1

Kamtschatica 2

Emphyllocalyx 0 0.2 Fig. 2. Neighbour-Joining tree of blue honeysuckle accessions indicated by subspecies, with delimitations of clusters as dashed lines. Genetic diversity in raspberry and honeysuckle 205 These tetraploid accessions can be divided into distinct Acknowledgements clusters. The first group harbours only accessions of L. caerulea subsp. emphyllocalyx originating from The authors are grateful to Nadezhda Tikhonova and Japan. This relatively isolated geographical origin may Laurent Solinhac for excellent technical assistance. This explain this grouping. A second cluster is mostly consti- work was financially supported by the Ministry of tuted with accessions of L. caerulea spp. altaica, pallasii Finance, Luxembourg. and stenantha. This finding confirms other results suggesting that altaica and pallasii subspecies could be identical (Avena, 1971; Kuklina, 1985). In our study, References this could even concern stenantha. Most of our accessions were L. caerulea subsp. Avena MA (1971) O samostojatel’nosti Lonicera baltica Pojark. kamtschatica, but interestingly, these accessions do not Botanicheskie sady Pribaltiki 27: 71–78. cluster together on the Neighbour-Joining tree. Instead, Chaovanalikit A, Thompson MM and Wrolstad RE (2004) they split up into two clearly different groups (Fig. 2). Characterization and quantification of anthocyanins and This phenomenon, similarly observed in a previous polyphenolics in blue honeysuckle (Lonicera caerulea L.). Journal of Agricultural and Food Chemistry 52: 848–852. ˇ report (Naugzemys et al., 2007), may have implications Falush D, Stephens M and Pritchard JK (2007) Inference of in Lonicera taxonomy (Lamoureux et al., in preparation). population structure using multilocus genotype data: domi- Finally, the L. caerulea subsp. venulosa accessions raise nant markers and null alleles. Molecular Ecology Notes 7: some questions. Although the main part of them is 574–578. grouped into one cluster, a few accessions are spread Graham J, Smith K, MacKenzie K, Jorgenson L, Hackett C and Powell W (2004) The construction of a genetic linkage throughout the entire tree in various clusters, thus ques- map of red raspberry (Rubus idaeus subsp. idaeus) based tioning the homogeneity of this subspecies. on AFLPs, genomic-SSR and EST-SSR markers. Theoretical The six clusters described above may provide a basis in and Applied Genetics 109: 740–749. defining the heterotic groups, useful for the breeding of Kuklina AG (1985) Populjacionnaja izmenchivost’ zhimolosti new elite cultivars. goluboj v Sibiri. Bjulletin GBS 136: 24–27. Naugzˇemys D, Zˇilinskaite˙ S, Denkovskij J, Patamsyte˙ J, Literskis J As shown above, the VIR collection of raspberry and Zˇvingila D (2007) RAPD based study of genetic contains a significant amount of genetic diversity, but variation and relationships among Lonicera germplasm the elucidation of the fine structure of the collection accessions. Biologija 53: 34–39. obviously requires more data. As an uncommon fruit Perrier X and Jacquemoud-Collet JP (2006) DARwin software. crop, the blue honeysuckle collection represents a http://darwin.cirad.fr/darwin Plekhanova MN (2000) Blue honeysuckle (Lonicera caerulea L.) – resource of tremendous importance. Subspecies proved a new commercial berry crop for temperate climate: to be a key level for diversity in edible Lonicera, with genetic resources and breeding. Acta Horticulturae 538: possible implications on taxonomy and breeding. 159–164. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 206–209 ISSN 1479-2621 doi:10.1017/S1479262111000499

Morpho-agronomic characterization and variation of indigo precursors in woad (Isatis tinctoria L.) accessions

Luı´s Rocha1, Carlos Carvalho1,2, Sandra Martins1, Fernando Braga3 and Valdemar Carnide1,2* 1Department of Genetics and Biotechnology, University of Tra´s-os-Montes and Alto Douro, 5001-801 Vila Real, Portugal, 2Centre of Genomics and Biotechnology/Institute for Biotechnology and Bioengineering, University of Tra´s-os-Montes and Alto Douro, 5001- 801 Vila Real, Portugal and 3Centre of Chemistry of Vila Real, University of Tra´s-os-Montes and Alto Douro, 5001-801 Vila Real, Portugal

Abstract Woad (Isatis tinctoria L.) was the most important source of natural blue indigo, a pigment used mainly for dyestuff until the beginning of the 20th century, when the increased use of synthetic dyes by the industry lead to a decrease in the interest of natural dyes and to the abandonment of dye crops. The aim of this study was to characterize, according to morphological and agronomical traits, 11 woad accessions from different countries and to quantify the indigo precursor content by high-performance liquid chromatograph-diode array detection (HPLC-DAD). Qualitative traits revealed a low variability and a great variation was observed in quantitative traits. Principal component analysis (PCA) diagrams divided the accessions into four groups, primarily according to their geographic origin. The ﬁrst three components of the PCA accounted for 76.8% of the total variation. Reciprocal interactions between indigo precursors and fresh leaf weight were compared through genetic diversity, with signiﬁcant differences in isatan B (0.5–5.1 g/kg), indoxyl (0.3–2.0 g/kg) and residual indigo (0.3–0.5 g/kg). This information, together with genetic data, can be used to assist local farmers to re-introduce Isatis species in the European agricultural system, not only indicating the higher indigo yielding genotypes, but also the most suitable harvest time.

Keywords: indigo; isatan B; Isatis tinctoria L.; morphological traits

Introduction Woad (Isatis tinctoria L.) is among the plant-derived dyestuffs which produce the blue pigment indigo. Both natural pigments obtained from dye plants and Woad belongs to the Brassicaceae family and its centre synthetic dyes can be used for dyeing purposes. The of origin is Eurasia, but it was widely diffused throughout increasing interest in obtaining raw materials from Europe (Spataro et al., 2007). renewable sources, the control of toxic effluents and the Ancient Egyptians used woad to dye cloth (Barber, classification of many synthetic colourants as toxic when 1991) and ancient Britons and Celts used it to colour in contact with the skin are the main reasons for a refreshed their faces and bodies (Carr, 2005). As the colour blue interest in natural dyes by the dyestuff industry. created was well suited to use with madder for cheaper purples and it was also used to dye fabric black, woad cultivation was widespread throughout the medieval Europe (Hofenk de Graaff et al., 2004; Cardon, 2007). * Corresponding author. E-mail: [email protected] The woad and dye production were highly profitable in Variation of indigo precursors in woad 207 Europe from the 12th to the 17th centuries (Oberthu¨r industry shows that this can be an interesting crop in et al., 2004), but during the 16th century, its importance the near future. began to decrease with the import of blue dye derived The aims of this study were to evaluate the agronomic from Indigofera tinctoria L. from the Far East, since it characteristics and to quantify the indigo precursors in a was cheaper (Galletti et al., 2006). From the 19th century collection of woad accessions from different countries. onwards, woad almost disappeared from the European agriculture due to the development of synthetic indigo. The indigo molecule is not produced in significant Material and methods amounts directly by woad. Instead, it is formed during the extraction process from the precursors accumulated Ten woad accessions from nine countries (UK, Spain in the leaves. During extraction, water-soluble isatan B (SP), Belgium (BL), Germany (GR), Austria (AU), Italy (indoxyl-5-ketogluconate) is hydrolysed via fermentation (IT), Poland (PL), Kazakhstan (KZ) and Portugal (Portugal- to produce indoxyl, which further reacts with another Madeira (PT-M) and Portugal-Coimbra (PT-C)) were indoxyl molecule, producing indigo (Gilbert et al., 2004). evaluated in Vila Real (418170N; 78440W; 460 m above The indole-derived compounds produced in the the sea level), Portugal. The seeds were sown in jiffy leaves, roots and seeds have anti-bacterial, anti-fever pots in August and transplanted to the field in September and anti-swelling properties and the essential oils pro- 2008. From each accession, ten plants were characterized duced in the seeds can be used in soap and body according to 5 qualitative and 15 quantitative traits. creams (Frećhard et al., 2001; Chen et al., 2002; Oberthu¨r The pigment content was also evaluated in the plants, et al., 2005; Galletti et al., 2006). by HPLC-DAD, regarding each main woad precursor. While the studies on the assessment of morphological The identification of these precursors was done in com- diversity revealed a great variation among the Isatis parison with Maugard et al., (2001) chromatograms. An species (Kizil, 2006) and among plants and accessions analysis of variance (ANOVA) of quantitative data and a (Spataro and Negri, 2008), quantitative analysis show principal component analysis (PCA) were performed. that isatin B, indican and indigo amounts vary significantly throughout the year, being higher in spring and summer. The concentration also varies between young and older leaves. The younger leaves have a higher con- Results and discussion centration of precursors (Maugard et al., 2001). The renewed interest showed by the dyestuff industry ANOVA showed significant differences (P , 0.05) among in natural dyes, the potential cultivations of woad in mar- accessions for all traits. The rosette leaves accessions ginal lands and the use by the medicine and cosmetic from PT-M and UK had the longest (14.00 and

BL 1.0

UK OF NRL

HOF 0.5 FAS QS DSE HEF AU DSE, days to scape emission PTM LRL ES, emergence of seeds 0.0 FAS, final appearance of seeds HEF, height at the end of flowering PTC IT SP STW HOF, height at the onset of flowering LRL, length of rosette leaves KZ NRL, number of rosette leaves GR SW TSW OF, onset of flowering –0.5 WRL ES QS, quantity of seed SL SL, silique length PL STW, silique total weight SW, silique width TSW, thousand seed weight WRL, width of rosette leaves –1.0 –1.0 –0.5 0.0 0.5 1.0 Fig. 1. PCA for ten accessions estimated with morphologic traits. 208 L. Rocha et al.

Table 1. Amount of indigo and indigo precursors in ten populations of Isatis tinctoria

Indican Isatan Isatan Indoxyl Indirrubin Indigo Accessions (g/kg) B (g/kg) C (g/kg) (g/kg) (g/kg) (g/kg) AU 2.116 1.478 0.205 1.346 0.099 0.422 UK 0.932 2.605 0.278 2.047 0.095 0.504 IT 0.894 0.699 0.186 0.327 0.518 0.420 BL 1.533 3.793 0.147 0.494 0.106 0.304 PL 1.288 3.533 0.206 0.837 0.066 0.346 KZ 1.551 2.394 0.204 0.469 0.137 0.501 PT-M 0.604 2.366 0.238 0.454 0.303 0.331 PT-C 0.791 3.804 0.178 0.257 0.128 0.375 SP 0.538 0.525 0.243 0.700 – 0.476 GR 0.906 5.059 – 0.365 0.136 0.312

14.06 cm) and widest (4.06 and 3.50 cm) rosette leaves, are shown in Fig. 1. PC1 was highly correlated with but had the lower number (28.8 and 21.2, respectively). the length of rosette leaves, plant height, number of The tallest height of the plants at the beginning of seeds/plant, total seed weight and 1000 seeds weight; flowering (68.5 cm) as well as at the end of flowering PC2 was highly correlated with the number of day (84.9 and 87.0 cm, respectively) was obtained in the onset of flowering, number of rosette leaves and silique accessions from PT-C and KZ. length and PC3 was highly correlated with the width of The Portuguese accession from Madeira was the ear- rosette leaves and number of days from seed germina- liest to flower, while the accession from SP was the tion to onset of scape. latest, with a 23 d difference. The longest period of flow- Chemical analysis (Table 1) showed that the amount of ering was registered in the accession from AU (23 d), isatan B present in the leaves of woad is higher in acces- while the shortest one was observed in the accession sions from GR, BL and PT-C, ranging from 3.8 to 5.1 g/kg from SP (9 d). fresh weight. Reciprocal interactions between indoxyl, The accession from SP had the highest average number isatan B and indican precursors and fresh leaf weight of seeds/plant (3423), but a low total seed weight were compared and significant differences (P , 0.05) (14.75 g) as well as a low 1000 seeds weight (4.54 g), were observed. Despite previous observations of while the accession of PT-M had a high average Angelini et al. (2007) of a indican:isatan B ratio of 1:5 number of seeds/plant (2837), the highest total weight in woad, in our case, this ratio was 1:2 in the same (21.00 g) and the highest 1000 seeds weight (7.53 g). species. The presence of indoxyl, isoindigo, cis-indirrubin High variability for morphological traits was also and indigo is an indication of the physiological state of observed among woad accessions by Spataro and Negri the leaf. While the levels of indoxyl and indigo are lower (2008) and between wild and cultivated woad species than those of indican and isatan B, their presence is due by Kizil (2006). In other dye species like weld and to the oxidation of their indigo precursors. madder, high diversity in morphological and agronomic Natural dyes are important in modern dyeing pro- traits were also observed (Angelini et al., 1997; Angelini cedures as they can contribute to an increasing sustain- et al., 2003; Gaspar et al., 2009; Baghalian et al., 2010). able textile dyeing, in what concerns, for example, The first three main components of the PCA separ- water, chemicals and energy consumption. ated the accessions into four groups. One group In conclusion, we can say that this study revealed the includes the four accessions from South Europe high variability present in woad accessions, concerning (PT-M, PT-C, SP and IT); the second group contains the morphological and indigo precursors contents. This the accessions from Central and Eastern European variability shows that there is a great potential for the countries (GR, PL and KZ); the accessions from UK, improvement of woad, regarding both agronomic and BL and AU form two different groups (Fig. 1). quality characteristics. A good correlation between the clusters and the geographic origin of the accessions was also observed by Spataro and Negri (2008) for woad accessions from different European and Asian countries. Acknowledgements The first three components of the PCA accounted for 41.84, 22.74 and 12.18% of the total variation, respect- The authors thank the Institute of Plant Genetics and ively. The eigenvectors for the 14 morphological traits Crop Plant Research (IPK), Gatersleben, Germany; the Variation of indigo precursors in woad 209 Royal Botanic Garden of Madrid; the Botanic Garden of Frećhard A, Fabre N, Pean C, Montant S, Fauvel MT, Rollin P and Coimbra; the National Botanic Garden of BL; the Germ- Fouraste´ I (2001) Novel indole-type glucosinolates from woad (Isatis tinctoria L.). Tetrahedron Letters 42: plasm Bank of Hungary and Dr. David Hill of University 9015–9017. of Bristol for providing the woad material. This work was Galletti S, Barillari J, Jori R and Venturi G (2006) Glucobrassicin supported by FCT project POCTI/AGR/56087/2004. enhancement in woad (Isatis tinctoria L.) leaves by chemical and physical treatments. Journal of the Science of Food and Agriculture 86: 1833–1838. References Gaspar H, Moiteiro C, Tukman A, Coutinho J and Carnide V (2009) Influence of soil fertility on dye flavonoids production in weld (Reseda luteola L.) accessions from Angelini LG, Pistelli L, Belloni P, Bertoli A and Panconesi S Portugal. Journal of Separation Science 32: 4234–4240. (1997) Rubia tintorum a source of natural dyes: agronomic Gilbert KG, Maule HG, Rudolph B, Lewis M, Vandenburg H, evaluation, quantitative analysis of alizarin and industrial Sales E, Tozzi S and Cooke DT (2004) Quantitative analysis assays. Industrial Crops and Production 6: 303–311. of indigo and indigo precursors in leaves of Isatis spp. and Angelini LG, Bertoli A, Rolamdelli S and Pistelli L (2003) Polygonum tinctorium. Biotechnology Progress 20: Agronomic potential of Reseda luteola L. as a new crop 1289–1292. for natural dyes in textile production. Industrial Crops Hofenk de Graaff JH, Roelofs WG and van Bommel M (2004) and Production 17: 199–207. Angelini LG, Tozzi S and Di Nasso NN (2007) Differences in leaf The Colourful Past: Origins, Chemistry and Identification yield and indigo precursors production in woad (Isatis of Natural Dyestuffs. UK: Archetype Books, p. 396. tinctoria L.) and Chinese woad (Isatis indigotica Fort.) gen- Kizil S (2006) Morphological and agronomical characteristics of otypes. Field Crops Research 101: 285–295. some wild and cultivated Isatis species. Journal of Central Baghalian K, Maghsodi M and Naghvi MR (2010) Genetic diver- European Agriculture 7: 479–484. sity of Iranian madder (Rubia tinctorium) populations Maugard T, Enaud E, Choisy P and Legoy MD (2001) Identifi- based on agro-morphological traits, phytochemical content cation of an indigo precursor from leaves of Isatis tinctoria and RAPD markers. Industrial Crops and Products 31: (Woad). Phytochemistry 58: 897–904. 557–562. Oberthu¨r C, Graf H and Hamburger M (2004) The content of Barber EJ (1991) Prehistoric Textiles. Princeton, NJ: Princeton indigo precursors in Isatis tinctoria leaves – a comparative University Press. study of selected accessions and post-harvest treatments. Cardon D (2007) Natural Dyes: Sources, Tradition, Technology Phytochemistry 65: 3261–3268. and Science. Paris: Archetype Publications. Oberthu¨r C, Ja¨ggi R and Hamburger M (2005) HPLC based Carr G (2005) Woad tattooing and identity in later Iron Age and activity profiling for 5-lipoxygenase inhibitory activity in early Roman Britain. Oxford Journal of Archaeology 24: Isatis tinctoria leaf extracts. Fitoterapia 76: 324–332. 273–292. Spataro G and Negri V (2008) Adaptability and variation in Isatis Chen X, Howard OM, Yang X, Wang L, Oppenheim JJ and tinctoria L.: a new crop for Europe. Euphytica 163: 89–102. Krakaner T (2002) Effects of Shuanghuanglian and Spataro G, Taviani P and Negri V (2007) Genetic variation and Qingkailing, two multi-components of traditional Chinese population structure in a Eurasian collection of Isatis medicinal preparations, on human leukocyte function. tinctoria L. Genetic Resources and Crop Evolution 54: Life Science 70: 2897–2913. 573–584. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 210–213 ISSN 1479-2621 doi:10.1017/S1479262111000463

Genetic diversity in woad (Isatis tinctoria L.) accessions detected by ISSR markers

Luı´s Rocha1, Sandra Martins1, Valdemar Carnide1,2, Fernando Braga3 and Carlos Carvalho1,2* 1Department of Genetics and Biotechnology, University of Tra´s-os-Montes and Alto Douro, 5001-801 Vila Real, Portugal, 2Centre of Genomics and Biotechnology/Institute for Biotechnology and Bioengineering, University of Tra´s-os-Montes and Alto Douro, 5001-801 Vila Real, Portugal and 3Centre of Chemistry of Vila Real, University of Tra´s-os-Montes and Alto Douro, 5001-801 Vila Real, Portugal

Abstract Woad (Isatis tinctoria L.) was introduced in Europe in ancient times to produce indigo, a natural blue pigment used mainly for dyestuff. This species was cultivated in Portugal until the beginning of the 20th century, especially in the inner North and South. A set of nine inter-simple sequence repeat (ISSR) markers generated 177 reproducible fragments, of which 171 were polymorphic. The mean number of fragments/accession was 111, ranging between 100 (Portugal-Coimbra) and 124 (Poland). The total polymorphism observed was 0.3272, the average polymorphism was 0.1784 and the gene differentiation between accessions was 0.4546. Polymorphism ranged between 53.8% (Austria) and 73.1% (Belgium). The genetic relationship among woad accessions was obtained with unweighted pair group method with arithmetic mean dendrogram based on a molecular marker, clearly clustering the woad accessions according to their geographic origin. The genetic diversity observed in this collection shows that there is a considerable potential for its improvement and that ISSR could be used to evaluate intra- and inter-accession similarities in I. tinctoria species.

Keywords: genetic relationships; inter-simple sequence repeat-PCR markers; Isatis tinctoria L.

Introduction of origin of this plant appears to be Central Asia (Spataro et al., 2007). It produces indigo, a natural Painting and colour used along human history relied on blue pigment used mainly for dyestuff. Woad is also natural dyes, giving rise to the textile industry in used for medical purposes and in cosmetic industries Europe, an industry that decreased with the develop- manufacturing soaps and body creams (Spataro and ment of synthetic dyes in the 20th century. Today Negri, 2008). The three main indigo precursors, isatan there is an increasing interest in natural products from A, isatan B and indican, are present in the leaves renewable sources, such as natural dyes, mainly from (Gilbert and Cooke, 2001; Oberthur et al., 2004). dye plants, due to the fact that they are friendly to Various landraces of woad present differ in their pheno- the environment. Woad or glastum (Isatis tinctoria L.) types and amounts of indigo precursors (Stoker et al., is a dye plant that belongs to the Cruciferae family 1998). The evaluation of genetic diversity could (Brassicaceae), tetraploide (2n ¼ 4x ¼ 28), outbreeding promote the efficient use of genetic variations in breed- and biennial (Darlington and Wyle, 1955). The centre ing programmes (Paterson et al., 1991). Inter-simple sequence repeat (ISSR) markers have been used, in Brassicaceae, to study genetic relationships among species of the genus Diplotaxis (Martıń and Sańchez- * Corresponding author. E-mail: [email protected] Ye´lamo, 2000); distinction of cultivated and wild on Genetic diversity in woad accessions 211

B. rapa (Andersen et al., 2009); cultivar identification Taq-PCR Master mix (Quiagen) and 0.5 mM of each and genetic diversity analysis of B. oleracea (Lu et al., primer. Amplification was performed in an Biometra 2009); and to detect genotoxic effect of heavy metals on One II thermocycle with an initial denaturation at 948C Eruca sativa (L.) (Al-Qurainy, 2010). They are based on for 5 min and 44 cycles at 948C for 30 s, 528C for 45 s, amplification of DNA sequence present at an amplifiable 728C for 2 min, followed by 10 min at 728C. Amplified distance between two identical microsatellite adjacent products were separated in 1.8% agarose gel in 1 £ tris- regions oriented in opposite direction, using a single borate-EDTA buffer at 120 V. The gels were stained with microsatellite primer (Zietkiewicz et al., 1994). This type 0.5 mg/ml ethidium bromide solution and visualized by of marker has advantages similar to random amplified illumination under an ultra violet light. polymorphic DNAs (RAPDs) and simple sequence repeats, The ISSR fragments were scored manually as present since they are polymorphic, have good reproducibility, (1) or absent (0). Cluster analyses, using simple matching cheap in cost, require no prior knowledge regarding the (SM) coefficient and unweighted pair group method with sequences of the genome, require only a small amount arithmetic mean (UPGMA), were performed with the of DNA and are fast in reacting (Zietkiewicz et al., 1994; NTSYSpc-2.0 g software. Rakoczy-Trojanowska and Bolibok, 2004). However, ISSRs are more reproducible than the RAPD; they use longer primers and hybridize at higher temperatures Result and discussion (Pasqualone et al., 2001). The aim of this work was to evaluate the genetic diver- The nine ISSR primers selected yielded very clear identifi- sity and similarity levels among and within accessions able bands and informative patterns in the eleven of woad from Europe and Central Asia by using ISSR accessions. This set of primers produced fragments with markers. This information, combined with data from sizes between 250 and 2500 bp and generated a total agronomic and phytochemical content, can contribute of 177 reliable fragments, of which 171 (96.6%) were to the prediction of the potential of this crop in subsis- polymorphic (Table 1), with an average of 19.7 poly- tence agriculture and as an alternative to the production morphic fragments/primer. Spataro et al. (2007), in a of synthetic dyes. I. tinctoria Eurasian collection, using amplified fragment length polymorphism (AFLP) and selective amplified microsatellite polymorphic locus, found similar results, Material and methods

Table 1. Total number of bands and percentage of poly- Plant material morphism detected in each accession and primer

Eleven accessions of woad from nine different European Total bands Polymorphic bands (%) countries – UK, Spain (SP), Belgium (BL), Germany (GR), Accessions Austria (AU), Italy (IT), Poland (PL), Portugal SP 109 69.7 (Portugal-Madeira (PT-M) and Portugal-Coimbra (PT-C) UK 104 68.3 and one from Central Asia – Kazakhstan (KZ) – were PT-M 120 69.2 studied. The seeds were sown in jiffy pots in August PT-C 100 59.0 GR 107 72.9 and transplanted to the field in September 2008. AU 106 53.8 KZ 111 65.8 PL 124 64.5 DNA extraction and ISSR analyses BL 119 73.1 IT 104 61.5 HU 112 59.8 Young woad leaves were collected from nine individual Mean 111 65.4 plants, frozen in liquid nitrogen and stored at 2808C. Primers DNA was extracted using the DNeasyTM Plant Mini Kit 809 20 95.0 (Qiagen) and its concentration and purity were evaluated 811 21 95.2 by Nanodrop ND-1000. 817 17 100 826 19 94.7 Initially, 28 primers of the University of British Colum- 847 23 100 bia (UBC) were tested, of which nine primers (UBC 809, 850 21 95.2 UBC 811, UBC 817, UBC 826, UBC 847, UBC 850, UBC 880 23 100 880, UBC 888 and UBC 889) were chosen, due to 888 16 87.5 the polymorphism and higher efficiency amplification. 889 17 100 Total 177 96.6 Each PCR reaction contained 60 ng of DNA, 10 ml 212 L. Rocha et al. 95.63 and 96.44% of polymorphic fragments, respectively. The dendrogram based on UPGMA analysis grouped However, Gilbert et al. (2002) found 86.9% of poly- the 11 accessions into three major clusters with SM simi- morphic fragments in three species of Isatis (I. glauca, larity coefficient, ranging from 0.63 to 0.94 (Fig. 1). I. tinctoria and I. indigotica) with AFLP, results that The accessions from PL, BL and Hungary (HU) are could be attributed to the molecular marker (ISSR) and grouped in Cluster I. Cluster II comprises seven acces- to the different species used. sions, and within this cluster, we can find two sub- The 11 accessions showed an average of 111 frag- clusters. Subcluster IIa comprises the accessions from ments, ranging from 100 in the accession from PT-C, to UK, SP, PT-M and PT-C, while subcluster IIb comprises 124 in the accession from PL. The accession from AU the accessions from AU, GR and KZ. Cluster III is revealed the lowest percentage of polymorphism formed only by the accession from IT, which shows (53.8%), while the accession from BL showed the highest less similarity with the other accessions. percentage of polymorphism (73.1%). For the assembly The analysis of molecular variance showed that the of accessions, the average of polymorphism was 65.4% percentage of variation within accessions (58%) was (Table 1). Gilbert et al. (2002) found a higher polymorph- higher than the variation among accessions (42%). The ism (86.8%) in a collection of five landraces in the same first three components in the principal coordinate ana- species. The primer UBC 8882000 generated a specific lysis (PCoA) that resulted from the Nei genetic distance fragment for the Italian accession. Seven plants, of the matrix accounted for 61.5% of the total variations nine studied, from the UK accession, also presented (22.4, 20.9 and 18.3% for the first, second and third com- one specific allele with primer UBC 817600. Such specific ponents respectively). The PCoA grouped the accessions bands could be useful for identifying accessions. Two into three clusters, which are in accordance with the unique alleles were found in accessions from PT-M, BL clusters obtained with UPGMA. The total variation in and UK with primers UBC 8172000, UBC 8501800 and PCoA (61.5%) was lower than the principal component UBC 888600, respectively. analysis observed with the morpho-physiological traits

I BL

SP IIa PTM

PTC II AU

GR IIb

IT III

0.63 0.71 0.79 0.86 0.94 Coefficient Fig. 1. Dendrogram of genetic similarity relationship based on total 99 plants data using UPGMA. Genetic diversity in woad accessions 213 (79%, not published), and the graphic projection of the Darlington CD and Wyle AP (1955) Chromosome Atlas of accessions was not similar. Flowering Plants. London: George Allen & Unwin Ltd, p. 38. Gilbert KG and Cooke DT (2001) Dyes from plants: past The detection of high levels of polymorphism makes usage, present understanding and potential. Plant Growth ISSR a powerful tool for assessing genetic diversity in Regulation 34: 57–69. woad, as it is fast, simple and efficient. None of the Gilbert KG, Garton S, Karam MA, Arnold GM, Karp A, Edwards individual plants were genetically identical according to KJ, Cooke DT and Barker JHA (2002) A high degree of the ISSR analysis, indicating that the level of resolution genetic diversity is revealed in Isatis spp. (dyer’s woad) in our study was sufficient to distinguish all genotypes. by amplified fragment length polymorphism (AFLP). This study, together with the morphological and Theoretical and Applied Genetics 104: 1150–1156. chemical characterization, could contribute to the selec- Lu X, Liu L, Gong Y, Zhao L, Song X and Zhu X (2009) Cultivar identification and genetic diversity analysis of broccoli and tion of woad accessions for breeding programmes, in its related species with RAPD and ISSR markers. Scientia order to have plants with higher number of leaves Horticulturae 122: 645–648. and/or bigger leaves and higher content of indigo Martıń JP and Sańchez-Ye´lamo MD (2000) Genetic relationships precursors, becoming an economically viable crop in the among species of the genus Diplotaxis (Brassicaceae) coming years. using inter-simple sequence repeat markers. Theoretical and Applied Genetics 101: 1234–1241. Oberthur C, Graf H and Hamburger M (2004) The content of Acknowledgements indigo precursors in Isatis tinctoria leaves – a comparative study of selected accessions and post-harvest treatments. Phytochemistry 65: 3261–3268. The authors thank the Institute of Plant Genetics and Pasqualone A, Caponio F and Blanco A (2001) Inter-simple Crop Plant Research (IPK), Gatersleben, Germany; the sequence repeat DNA markers for identification of drupas Royal Botanic Garden of Madrid; the Botanic Garden from different Olea europaea L. cultivars. European Food of Coimbra; the National Botanic Garden of Belgium; Research and Technology 213: 240–243. Germplasm Bank of Hungary and Dr. David Hill from Paterson AH, Tanksley SD and Sorrells ME (1991) DNA markers University of Bristol for providing the woad material. in plant improvement. Advances in Agronomy 46: 39–90. Rakoczy-Trojanowska M and Bolibok H (2004) Characteristics This work was supported by FCT project POCTI/AGR/ and a comparison of three classes of microsatellite-based 56087/2004. markers and their application in plants. Cellular and Molecular Biology Letters 9: 221–238. References Spataro G and Negri V (2008) Adaptability and variation in Isatis tinctoria L.: a new crop for Europe. Euphytica 163: 89–102. Spataro G, Taviani P and Negri V (2007) Genetic variation and Al-Qurainy F (2010) Application of inter simple sequence repeat (ISSR marker) to detect genotoxic effect of heavy metals on population structure in a Eurasian collection of Isatis tinc- Eruca sativa (L.). African Journal of Biotechnology 9: toria L. Genetic Resources and Crop Evolution 54: 573–584. 467–474. Stoker KG, Cooke DT and Hill DJ (1998) Influence of light on Andersen NS, Poulsen G, Andersen BA, Kiær LP, D’Hertefeldt T, natural indigo production from woad (Isatis tinctoria). Wilkinson MJ and Jørgensen RB (2009) Processes affecting Plant Growth Regulation 25: 181–185. genetic structure and conservation: a case study of wild Zietkiewicz E, Rafalski A and Labuda D (1994) Genome finger- and cultivatedc Brassica rapa. Genetic Resources and printing by simple sequence repeat (SSR) anchored Crop Evolution 56: 189–200. polymerase chain reaction. Genomics 20: 176–183. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 214–217 ISSN 1479-2621 doi:10.1017/S1479262111000335

Genetic diversity among Italian melon inodorus (Cucumis melo L.) germplasm revealed by ISSR analysis and agronomic traits

S. Sestili*, A. Giardini and N. Ficcadenti Agricultural Research Council, Horticultural Crop Research Unit, Via Salaria 1, 63030 Monsampolo del Tronto (AP), Italy

Abstract The genetic relationships among 13 melon inodorus populations that were collected in southern Italy were assessed using 100 inter-simple-sequence repeat (ISSR) primers and 15 morphological traits. The dihaploid line Nad-1 and the cultivar Charentais-T, both of which belong to the botanical variety cantalupensis, were used as reference accessions in the molecular analysis. A total of 358 polymorphic bands were obtained from 39 of the 100 ISSR primers used, and 15 phenotypic traits were scored and used for genetic-similarity calculations and cluster analysis. The resulting dendrograms based on the ISSR and phenotypic data allowed almost all of the melon genotypes to be distinguished on the basis of the skin colour of the fruits. Mantel’s test revealed a good correlation between the morphological and molecular data in their ability to detect genetic relationships among melon ecotypes (r ¼ 0.50, P ¼ 0.99). The data obtained conﬁrm the effectiveness of this approach, and open new perspectives to reveal possible molecular associations with the phenotypic traits analysed.

Keywords: Cucumis melo; inodorus; inter-simple-sequence repeat; landraces; phenotypic traits

Introduction Both molecular markers and agronomic traits are commonly used in studies of the genetic relationships The melon (Cucumis melo L.) inodorus landraces are among melon genotypes (Garcia et al., 1998; Staub commonly known as ‘winter melons’, as they can survive et al., 2004; Dhillon et al., 2007). Inter-simple-sequence through the winter. They are traditionally cultivated in the repeat (ISSR) markers represent a successful tool for the Mediterranean area, and are an important horticultural assessment of genetic diversity in melon because these crop in Italy, with particular relevance for the economy are known to be highly informative and can be used to of the southern regions of Italy (Ficcadenti et al., 2007). detect polymorphism in genotypes with both wide and A difficulty in the cultivation of these melon genotypes narrow genetic backgrounds (Danin-Poleg et al., 1998; is that they are often identically named in the same culti- Stepanski et al., 1999; Sestili et al., 2004, 2008). vation areas, which results in homonyms and synonyms In the present study, we estimated the genetic diversity that produce confusion for the recognition of the specific among Italian inodorus melon populations using ISSR populations (Sestili et al., 2008). Therefore, it is impe- markers and phenotypic traits. rative to preserve the genetic variability of these local varieties as part of the conservation of the genetic resources of Italian melons, and for their further use in Materials and methods breeding programme (Ficcadenti et al., 2007, 2010). Thirteen melon inodorus landraces were used. The dihaploid line Nad-1 and the cultivar Charentais-T were used as reference accessions in the molecular analysis. All of * Corresponding author. E-mail: [email protected] the plants were grown in an open field according to Genetic diversity among Italian melon inodorus 215 a randomized block design, with two replicates and six Each polymorphic ISSR band was scored as either present plants/replicate. (1) or absent (0) for all of the genotypes, and the binary The phenotypic characterization was carried out on matrix obtained was used to calculate the genetic simi- three fruits randomly harvested from each population. larity coefficient for each pair of accessions (Nei and Li, Fifteen morphological-agronomic traits were recorded: 1979). The robustness of the nodes in the molecular fruit weight, fruits/plant, total yield, fruit length, fruit tree was tested by bootstrap analysis with 1000 replicates, width, fruit shape rind thickness, flesh thickness, placenta using the TREECON software (Van de Peer, 1994). The length, placenta width, skin colour, skin texture, flesh levels of correlation between the ISSR and agronomic colour, taste and Brix. distance matrices were determined using the Mantel test The morphological data were subjected to two-way with 1000 permutations (Mantel, 1967). ANOVA (Statistica 7 software), and the mean comparisons were calculated by Duncan’s test (P ¼ 0.05). Genomic DNA was extracted from young melon leaves Results and discussion according to Levi and Thomas (1999). One-hundred ISSR primers were used (set no. 9; UBC, Vancouver, Canada). Similarities among the inodorus accessions were estab- The PCR analysis was performed according to Gupta lished. The binary data matrix that was obtained from et al. (1994). PCR products were separated through the phenotypic data was used to generate the dendro- 1.2% agarose gels and photographed by the Kodak 1D gram (UPGMA), which grouped the genotypes that 3.6 documentation system. A 100-bp ladder was used share similar agronomic features. The cluster analysis for the molecular weight standards. revealed two main clades, in which the genotypes are All of the statistical analyses were performed using the grouped on the basis of the skin colour of their fruits NTSYS-pc software, version 2.20 (Rohlf, 2006). The mor- (Fig. 1). In particular, the first clade is distinguished by phological data were first standardized according to two groups: one with all of the green genotypes and Lopez-Sese` et al. (2003), and the resulting binary data the other represented by only two accessions with a matrix was used to calculate the Euclidean distances. yellow skin colour. The second main clade grouped all

Napoletano Giallo

Gialletto Napolet

Napoletano Verde

Tendral Tardivo

Rognoso Capuano

Verde Quasi Rotondo

Purceddu

Rugosodi Cosenza

Giallodi Paceco

Giallo Sfilato

Cartucciaru

Alsia Fascista

Alsia

0.38 0.62 0.87 1.11 1.35 Coefficient Fig. 1. UPGMA cluster analysis on the basis of the phenotypic data. 216 S. Sestili et al.

44 Napoletano Giallo Rugosodi Cosenza 24 A 60 Giallodi Paceco 36 Giallo Sfilato

Gialletto Napoletano 88 41 Napoletano Verde 28 Tendral Tardivo B 32 94 Purceddu 27 Rognoso Capuano

Cartucciaru 64 Verde Quasi Rotondo

100 Alsia Alsia Fascista

100 Nad-1 Charentais-T

0.46 0.59 0.73 0.86 1.00 Fig. 2. UPGMA cluster analysis on the basis of the molecular data. of the remaining yellow genotypes. The Verde Quasi traits analysed. Such analyses of plant diversity using Rotondo accession was included in the green cluster precise morphological and molecular evaluations of although the skin colour of its fruit is green during grow- regional collections are useful for germplasm curators ing period and yellow at physiological maturity. Alsia was and plant geneticists, as they help to define accessions an unclustered accession, as expected, since the fruit according to geographical regions, and they provide shape is very similar to the watermelon morphology. solid historical reference data for future genetic studies The cophenetic correlation coefficient of this cluster that are aimed at assessing genetic erosion, exploring analysis was r ¼ 0.87 (P ¼ 1). genetic potential and site conservation priorities. A total of 358 polymorphic bands obtained from 39 of This study has confirmed the usefulness of these ISSR the 100 ISSR primers were used for the genetic-similarity molecular markers to distinguish among genotypes that calculations. The cluster analysis (UPGMA) grouped are characterized by a narrow genetic background, and the inodorus melon accessions into two separate main it opens new perspectives towards conservation of the groups on the basis of the skin colour of the fruit, in Italian melon genetic resources and their further use in agreement with the morphological analysis, and with the breeding programmes. exception of the yellow genotype Cartucciaru (Fig. 2). Furthermore, the Verde Quasi Rotondo was a distinct group, in contrast to what was observed with the morphological data analysis. Alsia and Alsia Fascista were References grouped together in a separate group and were highly similar (bootstrap 100). The dihaploid line Nad-1 and the Danin-Poleg Y, Tzuri G, Reis N and Katzir N (1998) Application genotype Charentais-T were an outgroup, as expected of inter-SSR markers in melon (Cucumis melo L.). Cucurbit Genetics Cooperative Report 21: 25–28. (Fig. 2). The cophenetic correlation computed for the Dhillon NPS, Ranjana R, Singh K, Eduardo I, Monforte AJ, molecular cluster analysis was r ¼ 0.91, with P ¼ 1. The Pitrat M, Dhillon NK and Singh PP (2007) Diversity distance matrices obtained from the phenotypic and among landraces of Indian snapmelon (Cucumis melo molecular data were further compared with Mantel’s test, var. momordica). Genetic Resources and Crop Evolution which revealed a good correlation (r ¼ 0.50; P ¼ 0.99). 54: 1267–1283. Ficcadenti N, Sestili S, Luongo L, Campanelli G, Rosa A, Ribeca The results obtained confirm the efficacy of this C, Ferrari V, Maestrelli A, Genna A and Belisario A (2007) approach and open new perspectives to reveal the Recupero e valorizzazione di germoplasma di melone possible molecular associations with the phenotypic inodorus (Cucumis melo L.). Italus Hortus 14: 58–65. Genetic diversity among Italian melon inodorus 217 Ficcadenti N, Giardini A, Sestili S, Lo Scalzo R, Palermo ML, Rohlf FJ (2006) A comment on “phylogenetic correction”. Tumbarello B, Monteleone G, Bono M and Bongiovı` C Evolution 60: 1509–1515. (2010) Varieta` migliorate e nuovi ibridi rilanciano il Sestili S, Minollini A, Luciani M, Campanelli G and Ficcadenti N melone d’inverno. L’Informatore Agrario 9: 58–61. (2004) Molecular characterization of several melon geno- Garcia E, Jamilena M, Alvarez JI, Arnedo T, Oliver JL and Lozano types (Cucumis melo L.) by using molecular markers R (1998) Genetic relationships among melon breeding to identify resistance genes to Fusarium oxysporum f. sp. lines revealed by RAPD markers and agronomic traits. melonis. 15–18 September, XLVIII SIGA, Lecce, p. 182 Theoretical and Applied Genetics 96: 878–885. Sestili S, Daniele A, Rosa A, Ferrari V, Belisario A and Ficcadenti Gupta M, Chyi Y-S, Romero-Severson J and Owen JL (1994) Amplification of DNA markers from evolutionarily diverse N (2008) Molecular characterization of different Italian genomes using single primers of simple-sequence repeats. inodorus melon populations based on ISSR molecular Theoretical and Applied Genetics 89: 998–1006. markers and preliminary SSR analysis. Proceedings of IX Levi A and Thomas CE (1999) An improved procedure for iso- Eucarpia Meeting on Genetics and Breeding of Cucur- lation of high quality DNA from watermelon and melon bitaceae, 21–24 May, Avignon, France, pp. 307–311. leaves. Cucurbit Genetics Cooperative Report 22: 41–42. Staub JE, Lopez-Sese` AI and Fanourakis N (2004) Diversity Lopez-Sese´ AI, Staub JE and Gomez-Guillamon ML (2003) among melon landraces (Cucumis melo L.) from Greece Genetic analysis of Spanish melon (Cucumis melo L.) and their genetic relationships with other melon germ- germplasm using a standardized molecular-marker array plasm of diverse origins. Euphytica 136: 151–166. and geographically diverse reference accessions. Theoretical Stepanski A, Kovalski I and Perl-Treves R (1999) Intra-specific and Applied Genetics 108: 41–52. classification of melons (Cucumis melo L.) in view of Mantel M (1967) The detection of disease clustering and a generalized regression approach. Cancer Research 27: their phenotypic and molecular variation. Plant Systematics 209–220. and Evolution 217: 313–332. Nei M and Li W-H (1979) Mathematical model for studying Van de Peer Y (1994) TREECON for windows: a software genetic variation in terms of restriction endonucleases. package for the construction and drawing of evolutionary The Proceedings of the National Academy of Sciences USA trees for the Microsoft Windows environment. Computer 76: 5269–5273. Applications in the Biosciences 10: 569–570. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 218–221 ISSN 1479-2621 doi:10.1017/S1479262111000189

Analysis of genetic diversity in Citrus

Franc¸ois Luro*, Julia Gatto, Gilles Costantino and Olivier Pailly Unite´ de Recherche 1103 Ge´ne´tique et Ecophysiologie de la Qualite´ des Agrumes (GEQA) station de recherche INRA, 20230 San Giuliano, Corse, France

Abstract Sugar and acidity levels are the main criteria of general fruit quality and for citrus juices pulp, in particular. The constituents of the acidity (organic acids) and the sweetness (glucose, fructose and sucrose) and the genes involved in their regulation have seldom been used to explore Citrus genetic diversity. We evaluated the juice composition of primary metabolic components for 87 varieties belonging to the eight major Citrus species grown under the same environmental and cultivation conditions by HPLC. We investigated the sequence polymorphism of nine candidate genes encoding for key enzymes of sugars and organic acids metabolic pathways by single strand conformation polymorphism (SSCP). Whatever the biochemical or molecular analyses, the observed structure of Citrus diversity was organized around three groups corresponding to the ancestral species (mandarin, pummelo and citron). As expected, the secondary species were closely related to their putative ancestors except for Citrus aurantium. Biochemical diversity was strongly correlated to molecular SSCP diversity at the genus level but not at the intraspeciﬁc level. Compared with other molecular marker types, higher diversity has been observed with SSCP technology, which makes it suitable for future quantitative trait loci mapping approach on gene polymorphism in citrus pulp acidity and sweetness regulation.

Keywords: acidity; Citrus spp; genetic diversity; single strand conformation polymorphism; sweetness

Introduction their hybrids, acidity decreases during fruit maturation and thus determines the most favourable time of harvest, Citron (C. medica), mandarin (C. reticulata) and suitable for commercial fruit quality (Sinclair, 1984). In pummelo (C. maxima) are considered to be modern terms of genetic evolution, acidity and sweetness had cultivated types most similar to the ancestors (Barret probably played a determinant role in the natural selec- and Rhodes, 1976; Nicolosi et al., 2000; Luro et al., tion and dissemination of these species during their 2001). Economically important types (orange, grapefruit, history. Unfortunately, their impact on natural selection lemon and lime) are believed to have originated from is hard to verify and remains only a hypothesis suggested one or more generations of hybridization between by the extension of pulp acid character on citrus varieties. these ancestral genera. In terms of composition and Nevertheless, some ‘acidless’ varieties exist, due to spon- commercial assessment of fruit maturity, sweetness and taneous mutations, characterized by a very low acidity acidity are considered as major components of any and a lack in citric acid (Bogin and Wallace, 1966; citrus (Ting and Attaway, 1971; Tucker, 1993). Organic Canel et al., 1995). More often, they are considered as acids and sugars vary according to species, varieties, and useful biological models to display the cellular mechan- also to environmental conditions (e.g. climate, irrigation, isms and genes expression involved in acidity and sweet- etc.) and fruit maturation (Bain, 1958; Sinclair, 1984; ness regulation (Albertini et al., 2006; Cercos et al., 2006; Marsh et al., 2003). In mandarin, sweet orange and Talon and Gmitter, 2008). The general behaviours in terms of primary metabolic contents and evolution for different cultivars are largely described in the literature. Nevertheless, these characteristics have never been * Corresponding author. E-mail: [email protected] investigated for a large panel of species representing Genetic diversity in Citrus 219 the Citrus genus simultaneously in a standard time and taxonomic group was represented by several genotypes environmental conditions. The objective of this work was to evaluate intra and interspeciﬁc diversity: 18 for to investigate the variation of pulp sweetness and acidity mandarin (Citrus reticulata Blanco), 9 for pummelo between and within major species for citrus history (C. maxima (Burm) Merr.), 9 for citron (C. medica L.), (ancestral species) and for citrus industry (cultivars) and 12 for orange (C. sinensis (L.) Osb.), 8 for grapefruit then to explore its relationship within Citrus phylogeny. (C. paradisi Macf.), 7 for lemon (C. limon (L.) Burm.), The polymorphism of gene sequences involved in the 11 for limes (C. aurantifolia (Christm.) Swing), 9 for primary metabolic pathway was explored by single sour orange (C. aurantium L.), 1 for combava (C. hystrix strand conformation polymorphism (SSCP) approach to D.C.) and 5 for various putative lemon hybrids. Lemon, ﬁnd genetic markers related to this biochemical diversity. orange, sour orange, lime and citron acidless mutants have also been included in the sample design.

Materials and methods HPLC analysis

Materials Each fruit juice was diluted tenfold for sugar analysis and twofold for organic acids analysis and centrifuged at Fruit and leaves were sampled from 87 citrus varieties 2250 g for 10 min. The supernatant was filtered through growing in “Institut National de la Recherche Agronomique- 0.45 mm acetate cellulose membrane filter. The separation Centre International de Recherche Agronomique et of organic acids and sugars was achieved by method de Developpement” citrus germplasm (San Giuliano, described in Albertini et al. (2006), using an analytical Corsica, France), in the first week of February. Each HPLC unit (Series 200; Perkin-Elmer, France).

0.3 Citron Pummelo group group 0.2

0.1

0 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.13 0.2 0.3 0.4 0.5

–0.1

–0.2 Mandarin

Axis 2 (42.35%) Lime group Lemon Acidic hybrid –0.3 Combava Citron Mandarin –0.4 Orange Sour orange –0.5 Pummelo Ascorbic Grapefruit Citric Oxalic Glucose –0.6 Malic Fructose Axis 1 (19%) 0 1

Saccharose

Succinic

0 Fig. 1. PCA plot of citrus species based on the pulp concentrations of soluble sugars and organic acids and the contribution of each component to the diversity represented on the two first axes of PCA (at right and lower part of the plot); ‘acidless’ mutants have been removed from the citrus samples. 220 F. Luro et al. SSCP analysis pummelo and citron groups. Mandarin is the most diversified group characterized by a high variation of succinic Primers were designed using Primer3 software for nine and citric acid amounts. Moreover, sucrose content is a genes involved in primary metabolic pathway: vacuolar valuable criterion to distinguish mandarins from all acid invertase (AB074885), mitochondrial malic enzyme other taxi. (CB417399), aconitase (AF073507), phosphoenolpyru- SSCP technique of DNA fragments from genes involved vate carboxylase (EF058158), isocitrate dehydrogenase in the metabolic pathway of sugars and organic acids (AF176669), vacoluar citric acid transporter (EF028327), allowed the detection of diversity between varieties phosphofructokinase (AF095520), phosphoenolpyruvate undetectable at the amplified fragment-size level. Then, carboxykinase (Csi5808) and malate dehydrogenase these polymorphisms were supposed to be related to (DQ901430). DNA extraction from citrus leaves was single nucleotide variations and not insertions or deletions. carried out according to Doyle and Doyle (1987). PCR The general organization of Citrus diversity obtained reactions (mixture and amplification conditions) were with the SSCP data is quite similar to that observed performed in a ‘Mastercycler gradient’ thermocycler with biochemical criteria (Fig. 2). The three major groups (Eppendorf) according to Luro et al. (2008). SSCP ana- are maintained. However, some cultivated species such lyses of amplified fragments were displayed according as sour oranges are associated with mandarins, whereas, to the method of Markoff et al. (1997). phylogenetically, they have been suggested to derive from a pummelo and mandarin crossing. This is also the case for sweet oranges, associated with pummelo group Statistical analysis while they are supposed to derive from a mandarin backcross following an initial pummelo £ mandarin cross. DARwin5 v. 4.0 (Perrier et al., 2003) was used to examine The second major information resulting from this analysis the molecular genetic diversity. From SSCP data, a dissim- was the high level of diversity between varieties of each ilarity matrix was established using Dice’s distance (Dice, species. This is a common situation for ancestral species 1945) and a dendrogram was elaborated by Hierarchical that have evolved by intersexual crosses. But molecular and Ascendant classification method. A principal com- diversity between varieties resulting from somatic ponent analysis (PCA) was performed with R software for HPLC data.

0 0.1 Results and discussion Mandarins

The general organization of Citrus diversity based on the amount of primary metabolic compounds is presented in Sour oranges the PCA (Fig. 1) It is organized around three groups where the main ancestor species are citrons (associating Oranges lemons, lemon hybrids, limes and combava), pummelos (associating oranges and grapefruits) and mandarins (without any other species). This representation totally Pummelos agreed with the Citrus diversity and supposed phylogenic relationships established with molecular markers Grapefruits (Nicolosi et al., 2000; Luro et al.; 2001; Barkley et al., 2006). Lemon may have originated from a cross between Limes and lemon citron and sour orange, orange from a cross between hybrids pummelo and mandarin, and grapefruit from a cross between orange and pummelo. These hypotheses were Lemons confirmed with our analysis of sugars and organic acids composition of citrus pulp. Sample 3, representing Clem- entine, is linked to pummelo group in total accord with Citrons its mandarin £ orange origin. One major deviation from the putative phylogeny was observed for sour orange Fig. 2. £ Hierarchical and Ascendant classification tree origin supposed to be a mandarin pummelo cross, showing the relationships of Citrus species accession as which is not supported by our analysis since sour determined by Darwin using a distance matrix calculated orange varieties have an intermediate position between from the proportion of shared alleles scored on SSCP gels. Genetic diversity in Citrus 221 mutations, such as for oranges, lemons or grapefruits, juice cells. Journal of American Society of Horticultural is very rare or not detected previously by neutral markers Science 120: 510–514. Cercos M, Soler G, Iglesias DJ, Gadea J, Forment J and Talon M (Luro et al., 2001; Barkley et al., 2006). Four genotypes (2006) Global analysis of gene expression during develop- are detected for oranges and two for both lemons ment and ripening of citrus fruit flesh: a proposed mechan- and grapefruits. ‘Acidless’ mutant profiles were not ism for citric acid utilization. Plant Molecular Biology 62: distinguishable from the other, suggesting independency 513–527. between molecular polymorphism and biochemical Dice LR (1945) Measures of the amount of ecologic association between species. Ecology 26: 297–302. variations. Doyle JJ and Doyle JL (1987) Isolation of DNA from fresh plant The overall biochemical diversity of Citrus reinforced tissue. Focus 12: 13–15. the idea that Citrus diversity is distributed among three Luro F, Rist D and Ollitrault P (2001) Evaluation of genetic ancestral species suggesting a speciation of fruit pulp relationships in Citrus genus by means of sequence sweetness and acidity prior to secondary species genesis. tagged microsatellites. Acta Horticultarae 546: 537–542. Luro F, Costantino G, Argout X, Froelicher Y, Terol J, Talon M, SSCP approach revealed a polymorphism apparently Wincker P, Ollitrault P and Morillon R (2008) Transferability neutral against acidity and sweetness regulation but suit- of the EST-SSRs developed on Nules clementine (Citrus able for further genetic studies such as quantitative trait clementina Hort ex Tan) to other Citrus species and their loci mapping and gene expression profiling. effectiveness for genetic mapping. BMC Genomics 9: 287. Markoff A, Savov A, Vladimirov V, Bogdanova N, Kremensky I and Ganev V (1997) Optimization of single-strand conformation polymorphism analysis in the presence of polyethylene glycol. Clinical Chemistry 43: 30–33. Marsh KB, Richardson AC and Erner Y (2003) Effect of environ- References mental conditions and horticultural practices on citric acid content In: International Society of Citriculture (ed.). Albertini MV, Carcouet E, Pailly O, Gambotti C, Luro F and Berti Proceedings of the 9th International Citriculture Congress, L (2006) Changes in organic acids and sugars during early Orlando, pp. 640–643. stages of development of acidic and acidless citrus fruit. Nicolosi E, Deng ZN, Gentile A, La Malfa S, Continella G and Journal of Agricultural and Food Chemistry 54: Tribulato E (2000) Citrus phylogeny and genetic origin of 8335–8339. important species as investigated by molecular markers. Bain JM (1958) Morphological, anatomical and physiological Theoritical and Applied Genetics 100: 1155–1166. changes in the developing fruit of the Valencia orange Perrier X, Flori A and Bonnot F (2003) Data analysis methods. Citrus sinensis (L.) Osbeck. Ausralian Journal of Botany In: Hamon P, Seguin M, Perrier X and Glaszmann JC 6: 1–24. (eds) Genetic Diversity of Cultivated Tropical Plants. Barkley NA, Roose ML, Krueger RR and Federici CT (2006) Montpellier: Enfield Science Publishers, pp. 43–76. Assessing genetic diversity and population structure in a Sinclair WB (1984) Organic acids of lemon fruits. In: The Citrus germplasm collection utilizing simple sequence Regents of the University of California (ed.) The Biochem- repeat markers (SSRs). Theoritical and Applied Genetics istry and Physiology of the Lemon and Other Citrus 112: 1519–1531. Fruits. Oakland, CA: University of California, pp. 109–156. Barret HC and Rhodes AM (1976) A numerical taxonomic study Talon M and Gmitter F Jr (2008) Citrus genomics. International of affinity relationships in cultivated Citrus and its close Journal of Plant Genomics. 2008: 528361. relatives. Systematic Botany 1: 105–136. Ting SV and Attaway JA (1971) Citrus fruits. In: Hulme AC (ed.) Bogin E and Wallace A (1966) Organic acid synthesis and The Biochemistry of Fruits and their Products. London: accumulation in sweet and sour lemon fruits. Journal of Academic Press, pp. 107–169. American Society of Horticultural Science 89: 182–194. Tucker GA (1993) Introduction. In: Seymour G, Taylor J and Canel C, Bailey-Serres JN and Roose ML (1995) In vitro [14C] Tucker G (eds) Biochemistry of Fruit Ripening. London: citrate uptake by tonoplast vesicles of acidless Citrus Chapman and Hall, pp. 1–37. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 222–223 ISSN 1479-2621 doi:10.1017/S1479262111000128

Diversity of seed storage protein patterns of Slovak accessions in jointed goatgrass (Aegilops cylindrica Host.)

Edita Gregova´1*, Pavol Hauptvogel1, Rene´ Hauptvogel1, Ga´bor Vo¨ro¨sva´ry2 and Ga´bor Ma´lna´si Csizmadia2 1Plant Production Research Center, Piestany, Slovak Republic and 2Central Agricultural Ofﬁce, Research Centre for Agrobotany, Ta´pio´szele, Hungary

Abstract Variations in seed storage protein patterns were investigated for six accessions of jointed goatgrass (Aegilops cylindrica) populations collected from Slovakia within the framework of the bilateral Co-operation in Science and Technology between the Slovak Republic and Hungary. The study covered populations collected from the southwestern (localities: Sered and Dunajska´ Streda), southern (localities: Chlaba and Kamenica nad Hronom) and southeastern (localities: Cierna nad Tisou and Dobra) parts of Slovakia. Analysis of proﬁles of seed storage proteins – glutenins and gliadins – was carried out using acid polyacrylamide gel electrophoresis and sodium dodecyl sulphate polyacrylamide gel electrophoresis. All accessions have a uniform three-band high molecular weight glutenin pattern with CxCyDy subunit composition. The highest variations in gliadin bands among the populations were observed from Cierna nad Tisou. There were small differences among the populations from Chlaba and Dobra. The lowest variations were in populations from Sered, Dunajska Streda and Kamenica nad Hronom. The present investigation showed that these jointed goatgrass populations are valuable genetic resources for wheat crop improvement programmes.

Keywords: Aegilops cylindrica Host.; diversity; gliadins; glutenins

Introduction locus consists of two tightly linked but not always expressed genes (Payne et al., 1981). Within the loci, The Aegilops genus contains species closely related recombination is rare. In diploid species, both subunits to wheat. Electrophoretic analyses of seed storage pro- are expressed, whereas in polyploid species such as teins – glutenins and gliadins – have proven very wheat and jointed goatgrass, there is silencing of one or useful in evaluating and characterizing of jointed goat- more subunits (Payne et al., 1981; Galili and Feldman, grass (Aegilops cylindrica Host.) accessions. Jointed 1983). Jointed goatgrass has a three-band pattern due to goatgrass has C and D genomes. High molecular non-expression of the Glu-1Dx subunit (Johnson, 1967). weight-glutenin subunits (HMW-GS) are controlled by genes at two complex loci – Glu-1C and Glu-1D – located on the long arm of the group 1 chromosome. Materials and methods Different allelic subunits at each Glu-1 locus have been identified by sodium dodecyl sulphate polyacrylamide We analysed six accessions of A. cylindrica populations gel electrophoresis (SDS-PAGE). Each Glu-1 complex collected from the Slovak Republic. HMW-GS were extracted from randomly selected single seeds, and protein was extracted from the non-embryo half of the seed. Extractions, electrophoretic separation of glutenin * Corresponding author. E-mail: [email protected] and detection procedures used were in accordance to Variations in seed storage protein patterns 223 International Seed Testing Association (ISTA) standard under acid conditions. Traditionally, the electrophoreto- procedure for SDS-PAGE, and glutenin subunits were grams of gliadins are divided into a, b, g and v zones, identified following the catalogue of HMW-GS alleles with proteins found in each zone being grouped into (Payne and Lawrence, 1983). Gliadin in the genotypes separate subclasses. The v gliadins have a low cysteine was examined using acid PAGE according to the standard content, differing from the other groups of gliadins, which ISTA reference method (Bushuk and Zillman, 1978). are cysteine-rich (Shewry and Tatham, 1990). We identified g, d and v zones of gliadins in the jointed goatgrass populations, but not a. The highest variation in gliadin bands among the populations was from Cierna nad Results and discussion Tisou, with five different compositions. There were small differences among the populations from Chlaba The x- and y-type subunits of the respective C and D and Dobra, with three different gliadin compositions. genome band contributions to the HMW glutenin pattern The lowest variations were from Sered and Dunajska for jointed goatgrass were identified by comparison with Streda populations, with only two different compositions. the band pattern of Triticum aestivum cv. Chinese The present investigation showed that the jointed Spring. The Glu-1 subunit assignment system developed goatgrass populations collected from Slovakia were by Payne and Lawrence (1983) was used to identify valuable genetic resources for wheat crop improvement the relative band positions for the Glu-1A, Glu-1B and programmes. Glu-1D subunits. Glu-1C alleles are not encompassed in this assignment system. Jointed goatgrass has a three-band HMW glutenin pattern with a CxCyDy subunit composition and is in agreement with the assignments Acknowledgements determined by Wan et al. (2000). Glu-1Dx subunit was This work was supported by the Slovak Research and not expressed in all tested material. This low variation Development Agency under the contract no. APVV- for the Glu-1 genetic marker was consistent with other 0770-07. findings of low genetic diversity in jointed goatgrass (Okuno et al., 1998). Gliadin electrophoresis (Fig. 1) showed a higher level of polymorphism than glutenin References and therefore should be better for use in population identification. Gliadins are a highly polymorphic and Bushuk W and Zillman RR (1978) Wheat cultivar identification biochemically unusual class of proteins characterized by gliadin electrophoreogram. I. Apparatus, method, and nomenclature. Canada Journal of Plant Science 58: by very complex electrophoretograms when separated 505–515. Galili G and Feldman M (1983) Genetic control of endosperm proteins in wheat 2. Variation in high-molecular-weight glutenin and gliadin subunits of Triticum aestivum. Theore- tical and Applied Genetics 66: 77–86. Johnson BL (1967) Confirmation of the genome donors of Aegilops cylindrica. Nature 216: 859–862. Okuno KK, Ebana B, Noov B and Yoshida H (1998) Genetic diversity of central Asian and north Caucasian Aegilops species as revealed by RAPD markers. Genetic Resources and Crop Evolution 45: 389–394. Payne PI and Lawrence GJ (1983) Catalogue of alleles from the complex loci, Glu-A1, Glu-B1 and Glu-D1 which code for high molecular-weight subunits of glutenin in hexaploid wheat. Cereal Research Communications 11: 29–35. Payne PI, Holt LM and Law CN (1981) Structural and genetic studies on the high-molecular-weight subunits of wheat glutenin. Theoretical and Applied Genetics 60: 229–236. Shewry PR and Tatham AS (1990) The prolamin proteins of 12345678910 11 12 13 14 151617181920 cereal seeds: structure and evolution. Biochemical Journal Fig. 1. Gliadin pattern of Slovak accessions in jointed 267: 1–12. goatgrass (A. cylindrica Host.): lanes 1–3, populations from Wan YK, Liu D, Wang D and Shewry PR (2000) High-molecular- Chlaba; lanes 4, 5, populations from Sered; lanes 6, 7, weight glutenin subunits in the Cylindropyrum and populations from Dunajska Streda; lines 8–10, populations Vertebrata section of the Aegilops genus and identification from Kamenica nad Hronom; lanes 11–17, populations of subunits related to those encoded by the Dx alleles of from Cierna nad Tisou; lanes 19–20, populations from common wheat. Theoretical and Applied Genetics 101: Dobra. 879–884. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 224–228 ISSN 1479-2621 doi:10.1017/S1479262111000256

Assessment of genetic diversity among Sri Lankan rice varieties by AFLP markers

Gowri Rajkumar1, Jagathpriya Weerasena1*, Kumudu Fernando2 and Athula Liyanage3 1Institute of Biochemistry Molecular Biology and Biotechnology, University of Colombo, Colombo, Sri Lanka, 2Agricultural Biotechnology Centre, Faculty of Agriculture, University of Peradeniya, Peradeniya, Sri Lanka and 3Plant Genetic Resources Centre, Gannoruwa, Peradeniya, Sri Lanka

Abstract Sri Lanka has a valuable repository of germplasm collection due to the availability of a large number of different traditional and improved rice varieties. Molecular techniques can increase the effectiveness of traditional technologies in assessing genetic diversity. Amplified fragment length polymorphism (AFLP) was used to evaluate the genetic diversity among rice varieties available in the germplasm collection of Plant Genetic Resources Centre, Sri Lanka. AFLP analysis of rice varieties using ten different primer combinations yielded a total of 772 polymorphic bands (98.4%). Genetic similarities were estimated using Jaccard’s (J) similarity coefficient. Unweighted pair group method with arithmetic mean (UPGMA)-based dendrogram was constructed. Genetic similarities varied from 0.073 to 0.565. Cluster analysis by genetic similarity divided the accessions into four main groups. The Cophenetic correlation with r ¼ 0.781 indicated high confidence of AFLP data to group the varieties in UPGMA clusters. Principal component analysis further confirmed the patterns obtained by the cluster analysis. The results revealed very high genetic diversity at molecular level among the Sri Lankan rice varieties used in this study.

Keywords: AFLP; dendrogram; genetic diversity; rice; UPGMA

Introduction method used to assess a large number of traits without prior sequence knowledge. Therefore, AFLP is widely Rice varieties grown in Sri Lanka from ancient to middle of accepted as an effective tool for identifying genomic differ- the last century are known as traditional varieties. High ences (Loh et al., 1999). Furthermore, fluorescent labelling yielding new improved varieties have been produced in fluorescent AFLP (FAFLP) replaces the radioactive label- from crosses between traditional and exotic genotypes. ling and increases the throughput by enabling automated Almost all rice varieties grown today are new improved var- detection and scoring of fragments generated in AFLP. ieties. However, traditional genotypes are still conserved The objective of this study was to assess the genetic in situ and/or ex situ in Genebanks. As these locally diversity among different rice accessions available in adapted genotypes may contain traits/genes of economic the Genebank of Sri Lanka by FAFLP markers. importance, their characterization is essential to use the available genetic diversity. AFLP technique is a robust, reliable and highly informative DNA fingerprinting Materials and methods DNA extraction

Seeds of rice accessions (Table 1) were obtained from * Corresponding author. E-mail: [email protected] the Genebank of the Plant genetic Resources Center, Assessment of genetic diversity by AFLP markers 225

Table 1. List of traditional rice varieties (TR), cultivated varieties (CV), wild rice species (WS) and hybrid variety (HV) used in this study

Accession Accession Accession no. Name Type no. Name Type no. Name Type 5642 Pokkali TR 6276 Mudukiriel TR 8919 Bg357 CV 3142 Molaga samba TR 7799 SuduheenatiIcpy 19 TR 8923 Bg359 CV 3445 Yakada wee TR 4724 Goda al wee TR 11630 At306 CV 3470 Hatada wee TR 2056 Gonabaru TR 10935 Bg363 CV 3689 Kiriyal TR 3147 Kaluheenati TR 6311 Bw267-3 CV 3707 Heenati TR 3598 Bala mawee TR 8920 Bg360 CV 3725 Siviru wee TR 3355 Suduheenatilcpy 15 TR 10325 At362 CV 3131 Dahanala TR 6248 Kombili TR 8682 Bg403 CV 3136 Pachchaiperumal TR 6233 Kalu kura wee TR 2835 Bg450 CV 3143 Sulai 27614 TR 4278 Kohumawi B11 TR 10591 Ld356 CV 3145 Podiwi A8 TR 2199 Suwada samba TR 8540 Bg304 CV 3150 Vellaiperunel TR 3177 Madael TR 9103 At353 CV 3151 Vellai Ilankalayan TR 3724 Goda heenati TR 9102 At303 CV 3156 Oddavalan TR 8499 Hondarawala TR 7182 Bg352 CV 3183 Hathiel TR 6188 Heen dik wee TR 10446 Bw361 CV 3484 Kuruluthudu wi TR 4992 Rathu heenati TR 5310 Bw351 CV 3133 Murunga 137 TR 3751 Chembavala samba 1 TR 3520 Bg11-11 CV 3444 Dikwee TR 4614 Chembavala samba 2 TR 2830 Bg379-2 CV 3478 Kahata wee TR 2079 Murunga TR 2836 Bg380 CV 4832 Hatapanduru wee TR 3495 Murungakayan 104 TR 9478 Bg358 CV 4909 Niyan wee TR 3192 Thahanala TR 10030 Oryza WS rhizomatis 5630 Peillianel 26081 TR 2714 Gambada samba TR 5138 Oryza WS granulata 6169 Dewaradderi 26081 TR 2846 Bg745 CV 10199 Oryza nivara WS 6283 Murunga kayan TR 2840 Bg300 CV 10031 Oryza WS eichingeri 3727 Kottamalli TR 2837 Bg350 CV 10096 Oryza ruﬁpogon WS 5569 Red leaf variety TR 4010 Bg90-2 CV 6290 Taichung Native 1 HV 2340 Wedaheenati TR 4017 Bg34-5 CV Hygroryza aristata

Sri Lanka. They were planted in pots filled with soil collec- or FAM) EcoR1 primers and unlabelled Mse1 primers. ted from a paddy field and were grown in a green house. Ten different primer combinations were used for selec- Tender leaves were harvested after 2 weeks and stored at tive amplification. The products were purified by ethanol 2808C. DNA was extracted from leaves according to the precipitation followed by washing with 70% ethanol. method described by Chen and Ronald (1999). Concen- The dried pellets were re-suspended in 5 ml of water. trations of DNA were estimated, and concentrations of Finally, 2.5 ml of re-suspended sample was mixed with all DNA samples were adjusted approximately to 300 ng/ml. ET 550-ROX size standard (GE Healthcare Life Sciences, Amersham Place, Little Chalfont, Buckinghamshire, HP7 9NA, UK) and deionized formamide according to the manufacturer’s instructions and denatured at 958C for AFLP analysis 2 min. Fragments were resolved using capillary electrophoresis on MegaBACE 1000 automated DNA sequencer AFLP analysis was carried out according to Vos et al. (1995) (GE Healthcare Life Sciences). AFLP fragment analysis with modifications. DNA was digested with enzymes was performed with Genetic Profiler software 2.2 EcoR1 and Mse1at378C for 3 h and 30 min. Oligo- (GE Healthcare Life Sciences). nucleotide adapters were ligated to digested DNA by incubating DNA, and adapters with T4 DNA ligase at 378C for overnight. Then, digested/ligated DNA was pre-amplified with pre-amplification primers. The pre- Data analysis amplified products were diluted 20 times with sterile distilled water and used for selective amplification. The electropherograms in the range of 30–550 bp were Selective amplification reactions were performed using analyzed by Genetic Profiler software version 2.2 (GE pre-amplified DNA and fluorescently labelled (HEX, TMR Healthcare Life Sciences). Each fragment size was treated 226 G. Rajkumar et al. as a unique character and converted to binary data cluster analysis by unweighted pair group method with (present, 1; absent, 0). Jaccard’s (1901) similarity coeffi- arithmetic mean (UPGMA) (Sneath and Sokal, 1973), cients (J) were calculated, and similarity coefficients’ and the dendrogram was constructed by Multivariate Stat- matrix was constructed. The matrix was used for istical Package (MSVP 3.1; Kovach, 1998). The confidence

Hygroryza aristata Cluster I At303 Bw363 O. granulata Suduheenati1cpy 15 Pokkali Chembavala samba 1 Murunga Rathu heenati Gambada samba Hondarawala Chembavala samba 2 Goda heenati Heen dik wee At353 Madael Thahanala Suwada samba O. eichingeri Bg358 Bala mawee Goda al wee Cluster II O. rufipogon Gonabaru Kohumawi B11 Kalu kura wee Bg304 Kombili Bg403 Bw351 O. nivara Bg352 O. rhizomatis Kaluheenati Ld356 Suduheenatilcpy 19 At362 Bg360 Bw267-3 Mudukiriel Wedaheenati Bg380 Podiwi A 8 Bg90-2 Bg34-5 Hathiel Oddavalan Kuruluthudu wi Vellai Ilankalayan Vellaiperunel Bg300 Bg350 Sulai 27614 Bg745 Bg379-2 Pachchaiperumal Dahanala Bg450 BW361 Hatapanduru wee Kahata wee Cluster III Niyan wee Dikwee At306 Bg359 Bg357 Murunga kayan Dewaradderi 26081 Peillianel 26081 Murungakeyan 104 Murunga 137 Red leaf variety Kottamalli Kiriyal Heenati Hatada wee Yakada wee Siviru wee Bg11-11 Molaga samba TN 1 Cluster IV 1 0.04 0.2 0.36 0.52

Av J = 0.186 Fig. 1. The UPGMA dendrogram showing genetic diversity among Sri Lankan rice varieties. Assessment of genetic diversity by AFLP markers 227 of the UPGMA clusters was assessed by Mantel (1967) on the Gower’s general similarity coefficient (Gower test to calculate the Cophenetic correlation coefficient and Legendre, 1986) also confirmed the distribution of (r) as described by Abhijit et al. (2004). Cophenetic clusters in UPGMA analysis (data not shown). Scatter correlation coefficient is a comparison between the diagram of first three coordinates (PCo1, PCo2 and dendrogram and the similarity matrix. PCo3) revealed three well-defined groups. All wild species were found within the same group. Rice varieties with Indian origin were only found in two other groups, Results and discussion while all other verities were scattered among these groups. Ten primer combinations generated 784 fragments. Of The genetic diversity of rice varieties revealed by this which 772 were polymorphic (98.4%) and 12 (1.6%) study is useful in categorizing the accessions and were monomorphic. The number of amplified products preventing duplications in core collection in Genebanks. generated by each primer pair ranged from 41 to 171 In addition, this information at molecular level can be with an average of 78.4 fragments. integrated to devise strategies for ex situ and in situ J showed that the genetic similarity varied from 0.083 genetic conservation, their utilization and exchange of to 0.565. The traditional rice Dikwee and At353 showed genetic material. the lowest similarity (0.073), while Mudukiriel and Bw267-3 showed highest genetic similarity (0.565). The UPGMA dendrogram (Fig. 1) separated the acces- Conclusion sions into four main clusters at the similarity coefficient of 0.186. AFLP analysis of the rice varieties revealed significantly Clusters I and IV contained only one accession, while high genetic diversity among the Sri Lankan rice germ- clusters II and III comprised 37 and 38 accessions, plasm used in this study and degree of genetic resem- respectively. A modern rice variety (Taichung Native 1 blance/distance of each one of those varieties. This (TN-1)), Hygroryza aristata and five wild species were genetic diversity data will provide more direct and also included in this study to find out the reliability of reliable genetic information for selecting suitable parents data and analysis. H. aristata previously classified in rice breeding programmes. Furthermore, the infor- under genus Oryza but now separated as a different mation given here may be useful in the management of genus was introduced in the analysis as an outlier. in situ and ex situ preservations of rice germplasm. H. aristata was found to be the most divergent line (cluster I) by separating from the rest at the similarity coefficient of 0.130. All accessions in other clusters belong to genus Oryza. Cluster IV encloses only TN-1, Acknowledgements which is the first Indica variety carrying the Dee- geo-woo-gen gene (semi-dwarfing gene), and TN-1 is The authors greatly appreciate the financial support given the most susceptible variety for pest and diseases and by the National Research Council of Sri Lanka grant no. photoperiod-insensitive and known to be different from 05-61. Technical assistance of Ms Anoma Jayasoma is rest of the varieties used. These results confirmed the gratefully acknowledged. reliability of data and methods of analyses. Wild species Oryza nivara, Oryza rufipogon, Oryza References rhizomatis, Oryza granulata and Oryza eichingeri showed clear separation from other accessions repre- Abhijit SD, Kamlesh J, Jose´ MG, Vyankatesh JP, Milind SP, senting cluster II. Most of the traditional rice varieties Dilip RR and Yogesh SS (2004) Comparison of 16S rRNA known to be introduced from India (Molaga Samba, gene sequences of genus Methanobrevibacter. BMC Micro- Pachchaiperumal, Vellaiperunel, Vellai Illankalayan, biology 4: 20. Peillianel and Murunga) were in cluster III. These Chen DH and Ronald PC (1999) A rapid DNA mini preparation method suitable for AFLP and other PCR applications. plants had been cultivated in northern part of Sri Lanka Plant Molecular Biology Reporter 17: 53–57. during 1930s. Gower JC and Legendre P (1986) Metric and Euclidean proper- The Cophenetic correlation coefficient (r) from the ties of dissimilarity coefficients. Journal of Classification 3: comparison between the dendrogram and the similarity 5–48. ´ matrix was 0.781. The high value of Cophenetic corre- Jaccard P (1901) Etude comparative de la distribuition florale- dans une portion des Alpes et des Jura. Bull Soc Vandoise lation coefficient (r) indicates that the UPGMA dendro- Sci Nat 37: 547–579. gram represents similarity data accurately. Principal Kovach W (1998) MVSP-A multivariate statistical package. component analysis of those 81 rice accessions based Version 3.1. Pentraeth, Wales: Kovach Computing Services. 228 G. Rajkumar et al. Loh JP, Kiew R, Kee A, Gan LH and Gan YY (1999) Amplified Sneath PHA and Sokal RR (1973) Numerical Taxonomy: The fragment length polymorphism (AFLP) provides molecular Principles and Practice of Numerical Classification. San markers for the identification of Caladium bicolor cultivars. Francisco, CA: Freeman. Annals of Botany 84: 155–161. Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Hornes M, Mantel NA (1967) The detection of disease clustering and a Fritjers A, Pot J, Peleman J, Kuiper M and Zabeau M (1995) generalized regression approach. Cancer Research. 27: AFLP: a new technique for DNA fingerprinting. Nucleic 209–220. Acid Research 23: 4407–4414. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 229–232 ISSN 1479-2621 doi:10.1017/S1479262111000360

Molecular and morphological diversity in Japanese rice germplasm

Fa´tima Bosetti1*, Maria Imaculada Zucchi2 and Jose´ Baldin Pinheiro1* 1Luiz de Queiroz College of Agriculture, University of Sa˜o Paulo – ESALQ/USP, Piracicaba, Brazil and 2Agronomic Institute of Campinas – IAC, Campinas, Brazil

Abstract Germplasm molecular and phenotypic characterization is instrumental to its utilization and to genetic variability incorporation into rice breeding programmes. The diversity within 192 Japanese rice accessions was analysed for 22 agro-morphological traits and 24 single sequence repeat markers. A total of 181 alleles were detected, 38 of which were exclusive. The number of alleles/marker ranged from 2 to 16, with an average of 7.54 alleles/locus and the He value ranged from 0.01 to 0.82, with an average of 0.46. The accessions showed diversity at molecular and phenotypic level and few showed also good agronomic performance. Tocher’s method applied on a total-dissimilarity matrix was used to determine cluster formation of 13 diversity groups. Most of the accessions (81%) were clustered within a group, whereas eight accessions (Kyuushuu, Eika Ine, Ishiwari Mochi, Col/Fukui/1965, Ookuma Nishiki, Suzume Shirazu, Iwate Ryoon and Toga) did not cluster with other accessions.

Keywords: germplasm; molecular markers; morphological traits; multivariate analyses; Oryza sativa

Introduction that best represent the entire population or gene collection with the minimum loss of genetic diversity (Crossa Cultivated rice (Oryza sativa L.) is the most important and Franco, 2004). In addition to the descriptors rec- food crop feeding more than half of the world’s popu- ommended for each species, the molecular characteriz- lation. The use of rice genetic resources available at gen- ation of accessions becomes increasingly common ebanks is important to incorporate genetic variability in practice, but studies that integrate molecular and agro- rice breeding programmes that can potentially generate nomic data are less frequent. Several authors have new cultivars with broadened genetic basis and allow called attention to the need for a joint treatment of infor- new and useful allelic combinations (McCouch, 2005). mation coming from these two different sources (Franco While it is important to collect and preserve genetic et al., 2001; Bramardi et al., 2005) variation in seed banks, these activities are not sufﬁcient The purpose of this study was to identify the diversity to ensure the future productivity of agriculture. The among 192 Japanese rice accessions source using multi- establishment and maintenance of seed banks must be variate data from morphological traits and molecular coupled with the ability to actively utilize the materials markers. The utilization of all variables together showed in those collections (Tanksley and McCouch, 1997). a greater power to discriminate genotypes and enabled In genetic resource conservation and plant breeding, to summarize relationship among accessions. multivariate data on continuous and categorical traits are collected with the objective of selecting genotypes

Materials and methods

* Corresponding authors. E-mail: [email protected]; The diversity within 192 Japanese rice accessions (Sup- [email protected] plementary Table S1, available online only at http:// 230 F. Bosetti et al. journals.cambridge.org) was analysed by investigating 22 (RM165, RM145, RM 143, RM131, RM146, RM190, agro-morphological traits and 24 single sequence repeat RM180, RM149, RM160, RM184, RM181 and RM155). (SSR) markers. Experiments, designed as augmented Amplification products were visualized on polyacryl- blocks, were conducted in the experimental field of the amide denaturing gels followed by silver staining. Department of Genetics – ESALQ/USP, in SP, Brazil, Genetic divergence among the accessions was quantified during the agricultural years of 2007/2008 and 2008/ using Mahalanobis’s distance for quantitative variables, 2009, respectively. The accessions were characterized Jaccard’s arithmetic complement for qualitative variables with rice descriptors indicated by IBPGR-IRRI Rice Advi- and Rogers-W distance for SSR markers. Four matrices of sory Committee (1980) and Bioversity International dissimilarities were obtained (two Mahalanobis’s distance (2007). The descriptors used were: matrices, one for each agricultural year; one Jaccard’s distance matrix and one Rogers-W’s distance matrix). (1) Quantitative variables: number of days to heading; A total-dissimilarity matrix, which is a sum of elements of culm, number and length (cm); flag leaf, length the four matrices, was generated using ‘Genes’ software and width (cm); plant, length (cm); panicle, length (Cruz, 2008). Tocher’s method applied to total-dissimilarity (cm); number of days to maturity, 100-grain weight matrix was used to determine cluster formation. (g); yield (g/experimental plot); grain, length (mm), width (mm) and length–width ratio; (2) Qualitative variables: leaf blade, pubescence; auricle: colour; flag leaf, orientation; panicle, type, orientation of main axis, orientation of branches and Results secondary branching; awns, distribution and colour; lemma, colour of apiculus. Genetic diversity among 192 Japanese rice accessions was identified by simultaneously investigating morpho- Total DNA was isolated from young rice leaves by the logical and molecular data. Morphological data were cetyl trimethylammonium bromide method. SSR markers collected in experiments conducted in the field during were chosen by their polymorphic information content two agricultural years. The traits ‘plant: length, culm: and by their distribution considering the twelve rice length, panicle: length, number of days to heading, chromosomes. The primers used were: genomic SSRs flag leaf: length and width, number of days to maturity (RM9, RM207, RM55, RM335, RM334, RM204, RM11, and yield’ presented more contribution for the variabil- RM38, RM257, RM304, RM229 and RM247) and expre- ity among the accessions during the first year, accord- ssed sequence tags-simple sequence repeats (EST-SSRs) ing to canonical variable analysis (data not shown)

80 259J

134J 40 268J

y 0

191J 197J –20

–40 113J Cluster I Cluster II 245J 99J Cluster III –60 Cluster IV Cluster V Unclustered accessions –80 –200 20406080 100 120 140 160 180 200 220 x

Fig. 1. Patterns of relationships among 192 rice accessions and 13 Tocher’s clusters revealed by two-dimensional projection of the distances. Diversity in Japanese rice germplasm 231 and were evaluated during the second year to consider environmental effects in the variability. From 24 SSR markers, a total of 181 alleles were detected, 38 of which were exclusive. They were observed in 26 accessions and for 16 of the 24 analyzed

loci. He ranged from 0.01 to 0.82, with an average of 0.46 and the number of alleles/marker ranged from 2 (for RM 143 and RM 146 e RM 184) to 16 (for RM 257), with an average of 7.54 alleles/locus. The average number of alleles/locus was higher than that observed by Lu et al. (2005) for 115 U.S. rice cultivars and 30 ancestral accessions introduced from Asia (5.15 alleles/ locus) and by Yu et al. (2003) in parental lines used by the International Rice Molecular Breeding Program (6.25 alleles/locus). The number of alleles/locus was lower than the number observed in genetic diversity studies in Brazilian traditional rice cultivars (Brondani et al., 2006) and in traditional and improved Indonesian rice germplasm (Thomson et al., 2007), which exhibited an average of 14.7 and 13 alleles/locus, respectively. Thirteen diversity groups were determined by Tocher’s method (Fig. 1 and Supplementary Table S2, available online only at http://journals.cambridge.org). The majority of the accessions (81%) were clustered to the same group,

Means of accessions not clustered with other ones while eight accessions (Kyuushuu, Eika Ine, Ishiwari Mochi, Col/Fukui/1965, Ookuma Nishiki, Suzume Shirazu, Iwate Ryoon and Toga) were not clustered with other accessions, giving rise to clusters with a single accession (clusters VI to XIII). Fourteen accessions were grouped in cluster II, nine in cluster III and two in clusters IV and V. The inter-cluster distance ranged from 8.19 (between cluster I and X) to 11.53 between cluster IV (Hitachi Nishiki and Nourin Mochi 6) and X (Ookuma Nishiki). Accessions of cluster IV have longer cycle, shorter grains and lesser yield than Ookuma Nishiki, which was the most similar accession to cluster I accessions.

CV (%) 113J 99J 245J 191J 268JUngrouped 134J accessions 259J 197J have speciﬁc traits that prevent them from clustering with the remaining groups. Means of quantitative variables of these accessions are presented

2 in Table 1. h Accession Kyuushuu exhibits exclusive alleles for ﬁve loci and longer cycle, but it is distinct from the other late accessions that are, in general, taller than Kyuushuu. ) and coefﬁcient of variation (CV) of 13 quantitative variables of 192 Japanese rice accessions and means of eight rice accessions not

2 Exclusive alleles for loci RM9 and RM229 were detected h in the accession Col/Fukui/1965 which is a late and tall accession. Eika Ine is also a tall accession, but it differs from other tall accessions due to its precocity and for being the only accession exhibiting exclusive allele for the expressed loco RM160. Exclusive alleles were absent from the accessions Iwate Ryoon, Ishiwari

Mean, heritability ( Mochi, Ookuma Nishiki, Suzume Shirazu and Toga. As observed for accession Eika Ine, the accession Iwate Ryoon is tall, has long grains and exhibits precocity, con- Table 1. Traits Mean clustered with other ones by Tocher’s method Days to headingCulm: length (cm)Plant: length (cm)Panicle: length (cm)Flag leaf: length (cm)Flag leaf: 93 width 77.87 (cm)Days to maturity 98.40 20.60Grain yield (kg/plot) 28.52Culm: 81.80 number 93.22100-grain 80.76 weight 59.52 1.71 (g)Grain: 66.43 length (mm) 124Grain: 5.89 width 0.67 (mm) 76.75Grain: 2.69 length/width 5.23 6.62 2.98 8.97 34.11 5.46 93.52 65.67 8.30 107 5.87 89.74trasting 24.07 90.95 3.42 2.45 23.5 23.54 33.64 89.58 89.43 1.89 113.5 94.32 24.07 1.17 95.11 to 3.46 95 12.95 36.15 the 133 82.45 2.39 0.40 99.36accession 16.91 1.88 2.84 1.80 24.17 2.29 125 5.75 97.58 119.2 130 0.62 8.95 21.67 2.56 31.36 1.53Toga, 3.50 3.39 79.40 132 6.58 100.4 21.06 8.84 0.12 156 which 22.83 3.37 1.88 73.93 2.63 3.23 3.72 92.49 18.56 91is 0.33 9.82 28.35 158 about 89.71 3.12 1.30 3.15 109.2 3.09 19.50 5.98 28.01 0.69 8.7530 91 85.57 138d 3.21 105.1 2.09 2.73 19.60 earlier. 2.87 6.72 24.40 8.13 0.60 121 84 3.33 1.86 2.45 2.59 5.05 0.48 8.66 119 3.13 1.68 2.76 90 2.71 4.05 0.49 7.84 3.35 128 2.38 3.33 4.95 9.14 3.31 2.77 232 F. Bosetti et al. Ishiwari Mochi is a late accession (one of the latest of Crossa J and Franco J (2004) Statistical methods for classifying the bank) and has longer grains distinct from most genotypes. Euphytica 137: 19–37. ´ accessions, characterized by short grains. Traits linked Cruz CD (2008) Programa Genes – Diversidade Genetica. Viçosa: Editora UFV, p. 278. to flag leaf characterize the accessions Ookuma Nishiki Franco J, Crossa J, Ribaut JM and Betran J (2001) A method for and Suzume Shirazu. While Ookuma Nishiki exhibits a combining molecular markers and phenotypic attributes narrow flag leaf, Suzume Shirazu presents a large flag leaf. for classifying plant genotypes. Theoretical and Applied The accessions showed diversity at phenotypic and Genetics 103: 944–952. molecular level and some of them also showed suitable IBPGR-IRRI Rice Advisory Committee (1980) Descriptors for rice agronomic performance and can be used to broaden (Oryza sativa L.) p. 21. the genetic base of breeding programmes. Lu H, Redus MA, Coburn JR, Rutger JN, Mccouch SR and Tai TH (2005) Population structure and breeding patterns of 145 U.S. rice cultivars based on SSR marker analysis. Crop Science 45: 66–76. McCouch SR (2005) Diversifying selection in plant breeding. References Plos Biology 2: 1507–1512. Tanksley SD and McCouch SR (1997) Seed banks and molecular Bioversity International, IRRI and WARDA (2007) Descriptors maps: unlocking genetic potential from the wild. Science for wild and cultivated rice (Oryza spp.) Bioversity 277: 1063–1066. International, Rome, Italy; International Rice Research Insti- Thomson MJ, Septiningsih EM, Suwardjo F, Santoso TJ, Silitonga tute, Los Banõs, Philippines; WARDA, Africa Rice Center, TS and Mccouch SR (2007) Genetic diversity analysis of Cotonou, Benin. traditional and improved Indonesian rice (Oryza sativa L.) ´ Bramardi SJ, Bernet GP, Asıns MJ and Carbonell EA germplasm using microsatellite markers. Theoretical and (2005) Simultaneous agronomic and molecular charac- Applied Genetics 114: 559–568. terization of genotypes via the generalized procrustes Yu SB, Xu WJ, Vijayakumar CHM, Ali J, Fu BY, Xu JL, Jiang YZ, analysis: an application to cucumber. Crop Science 45: 1603–1609. Marghirang R, Domingo J, Aquino C, Virmani SS and Li ZK Brondani C, Borba TCO, Brunes T, Rangel PHN and Brondani (2003) Molecular diversity and multilocus organization of RPV (2006) Determination of genetic variability of the parental lines used in the International Rice Molecular traditional varieties of Brazilian rice using microsatellite Breeding Program. Theoretical and Applied Genetics 108: markers. Genetics and Molecular Biology 29: 676–684. 131–140. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 233–235 ISSN 1479-2621 doi:10.1017/S147926211100027X

Molecular characterization of the European rice collection in view of association mapping

Brigitte Courtois1*, Raffaella Greco2, Gianluca Bruschi2, Julien Frouin1, Nourollah Ahmadi1, Gae¨tan Droc1, Chantal Hamelin1, Manuel Ruiz1, Jean-Charles Evrard1, Dimitrios Katsantonis3, Margarida Oliveira4, Sonia Negrao4, Stefano Cavigiolo5, Elisabetta Lupotto5 and Pietro Piffanelli2 1CIRAD, 34098 Montpellier, France, 2Parco Technologico Padano, 26900 Lodi, Italy, 3NAGREF, 57001 Thermi-Thessaloniki, Greece, 4ITQB, 2784-505 Oeiras, Portugal and 5CRA-RIS, 13100 Vercelli, Italy

Abstract In South Europe, rice is grown as an irrigated crop in river deltas where it plays an important role in soil desalinization. Specific varieties are needed for these tough conditions. We analyzed the genetic structure of a set of 305 varieties coming from the European Rice Germplasm Collection (ERGC) with 90 single nucleotide polymorphisms and compared it with a reference set representative of the diversity of Oryza sativa (mini-Germplasm Bank (GB)). These accessions had been characterized for their grain type and growth cycle duration. The polymorphism information contents of the ERGC were lower than those of the mini-GB, indicating a narrower genetic basis. Indeed, almost all ERGC accessions belong to the japonica group. Within the japonica group, both a dendrogram and a Bayesian clustering identified two major clusters. The first cluster encompassed tropical japonicas and American varieties from USA and Argentina characterized by long and narrow grains and medium to long duration. On a finer level, tropical japonicas appear separated from the other accessions. The second cluster is composed of European varieties mostly early or medium in duration and Asian temperate accessions, with a subgrouping based on grain format. A set of 200 accessions was composed for association mapping studies on traits such as salt tolerance.

Keywords: association mapping; genetic diversity; O. sativa; single nucleotide polymorphisms; temperate japonica

Introduction grown mostly in river deltas, in areas ecologically fragile, where it plays an important environmental role by limit- Grown in the South of Europe since the 15th century, ing salinity problems through field flooding. Rice is at the rice occupies presently approximately 400,000 ha for a northern limit of its natural cultivation zone and suffers of production of 2.7 Mt in 2008 (http://faostat.fao.org). short cropping seasons (May to September) and low tem- The main producers are Italy (224,000 ha) and Spain peratures at both extremities of the cycle. (96,000 ha) and, to a lesser extent, Greece (31,000 ha), To face these constraints, specific rice varieties are Portugal (25,000 ha) and France (16,000 ha) with an aver- needed. The varieties grown in Europe are mainly but age yield of 6.5 t/ha. All the rice area is irrigated. Rice is not exclusively temperate japonica types. In the framework of two European projects (RESGEN 1999–2002 and EURIGEN 2006–2010), public breeding institutions decided to pool their working collections and character- * Corresponding author. E-mail: [email protected] ize the resulting European Rice Germplasm Collection 234 B. Courtois et al.

(ERGC) of 455 accessions on phenotypic and molecular The number of sub-populations, K, was assessed using bases. Part of the entries originates from Asia or from the software Structure. We run 50,000 burn-ins, 500,000 other temperate areas. Most of the collection, however, iterations and 10 replications/K value, with K varying is composed of cultivars created by European breeding from 2 to 8. An accession was classified into a sub- programmes, which are original genetic combinations population when its component from this population potentially useful for the Mediterranean and the overall was above 75%. temperate areas. European breeding programmes target salinity tolerance, blast resistance, grain quality and adaptation to Results water-saving strategies. The understanding of the genetic basis of these traits has made large progress through the The genotyping was successful for the 90 markers with use of mapping populations for quantitative trait loci only 0.39% missing data. The heterozygosity rate was detection. The decision to undertake association studies low (0.19%) as expected for accessions conducted as that bring a better resolution without the constraint of pure lines and represented by a unique plant. The aver- having to develop mapping populations motivated the age polymorphism information content (PIC) value of teams to define a common association panel from the the ERGC was 0.11 against 0.19 for the mini-GB with 40 ERGC and establish marker-assisted breeding strategies. and 3% of the markers with a PIC below 0.05 respect- The objective of this study was to characterize a set of ively, indicating a narrower genetic basis for the ERGC. 305 accessions from the ERGC with single nucleotide The NJ tree assembling ERGC and mini-GB accessions polymorphism (SNP) markers and extract a representa- shows the bipolar structure characteristic of O. sativa tive panel for association studies. with indica (in red) and aus/boro (in yellow) accessions clearly separated from sadri/basmati (in green) and japonica (in blue) accessions (Supplementary Figs. S1 and S2, Materials and methods available online only at http://journals.cambridge.org). Most of the ERGC accessions (in black) clustered with The 305 accessions included in the genotyped collection the japonica accessions. The exceptions were limited to are listed in Supplementary Table S1 (available online three accessions from Asia, such as Zhenshan 97 from only at http://journals.cambridge.org). Accessions were China, which clustered with the indica group and one classified according to their growth cycle (early, medium accession, Fragrance from Italy, derived from a basmati £ or late) and grain type (round, medium, long and large japonica cross which clustered with the sadri/basmati called ‘long A’ or long and narrow called ‘long B’) based group. on data obtained through several field experiments and Among the japonica group, two subgroups could be will soon be available on http://eurigen.cirad.fr. A set distinguished. The first subgroup encompassed most of of 62 additional accessions representative of the four the tropical japonicas and accessions from America, main rice varietal groups (indica, japonica, aus and which could be further subdivided into accessions from basmati) extracted from a core collection known as mini- USA and Argentina. Almost all EGRC accessions belong- Germplasm Bank (GB) (Glaszmann et al., 1995) was ing to this subgroup have long B grains and medium to genotyped at the same time to be able to position the late duration. The second subgroup of accessions clus- ERGC accessions in comparison with this reference set. tered almost all accessions from Europe together with A set of 90 SNPs was extracted from the Oryza SNP the temperate japonicas accessions from the reference database (www.oryzasnp.org) on the basis of (1) their set. No specific pattern linked to geographic origin polymorphism among the japonica accessions of the could be detected within this subgroup. However, a database; (2) a good coverage of the genome with a mini- trend to sub-grouping by grain type and, to a lesser mum distance of 1 Mb between SNPs (Supplementary extent, by duration, was noticed. The upper part of the Table S2, available online only at http://journals. cluster corresponds to varieties with round to medium cambridge.org). DNA was extracted from one plant/ grain. The intermediate part corresponds to a mixture accession. The SNPs were genotyped using an Illumina of grain types. The lower part corresponds to accessions Veracode assay. Appropriate controls (duplicated acces- with mostly long A grains. Most varieties were early or sions and artificial heterozygotes) were included in the medium in duration with a specific cluster for the very plates. The allele calling, performed with Bead Studio, early Eastern Europe varieties. was manually controlled for each SNP. The results of Structure confirmed almost perfectly the A weighted neighbour-joining (NJ) tree was built based split between American and European accessions for on a dissimilarity matrix using a Sokal and Michener K ¼ 2 (Supplementary Table S1, available online only at index with DARwin software (http://darwin.cirad.fr). http://journals.cambridge.org). For K ¼ 3, the American Molecular characterization of European rices 235 accessions stayed together while the subgroup with DARwin based on allelic combinations. Because of the long and narrow grains separated from the European overlapping between the genetic and phenotypic struc- accessions. Higher values of K were unlikely. ture, association mapping is not likely to be successful for traits such as duration and grain type but other important phenotypic traits such as salinity tolerance do Discussion not show the same organization (Ahmadi et al., in preparation) and will be amenable to association mapping. We have characterized 305 ERGC accessions and shown, as expected, that most of them were japonica, with a narrow genetic basis, since many of them were derived Acknowledgements from the others by hybridization. A first subgroup was composed of American accessions and tropical japonicas EURIGEN (‘Genotyping for the Conservation and which constituted three separate clusters at a finer Valorization of European Rice Germplasm’) is a project level. The proximity between US accessions, except co-financed by the European Commission – DG Agricul- those from California, the Argentinian accessions and ture and Rural Development within the AGRI GEN tropical japonica has already been demonstrated RES programme for conservation, characterisation, (Mackill, 1995; Giarrocco et al., 2007), but the fact they collection and utilization of genetic resources in represented three separate clusters show that modern agriculture. varieties have specific allelic combinations, which is a new information. The European accessions which represented a distinct subgroup were organized mainly by grain type and growth cycle duration. Both traits References have been the target of strong selection: the grain type, Giarrocco LE, Marassi MA and Salerno GL (2007) Assessment of because it determines the marketability of the varieties the genetic diversity in Argentine rice cultivars with SSR with a lasting importance of round and medium grains, markers. Crop Science 47: 853–860. often needed for local preparations (e.g. risotto or Glaszmann J-C, Mew T, Hibino H, Kim CK, Mew TI, paella), and the growth cycle duration, because of the Vera Cruz CH, Notteghem J-L and Bonman JM (1995) need for adaption to European latitudes. Molecular variation as a diverse source of disease resistance in cultivated rice. Rice Genetics III. Los Banõs: IRRI, From these data, we composed an association pp. 460–466. panel of 200 accessions (see passport data) using a Mackill DJ (1995) Classifying japonica rice cultivars with RAPD linkage disequilibrium minimization method available in markers. Crop Science 35: 889–894. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 236–239 ISSN 1479-2621 doi:10.1017/S1479262111000293

Screening of barley germplasm for resistance to root lesion nematodes

Shiveta Sharma1*, Shailendra Sharma1, Tobias Keil1†, Eberhard Laubach2 and Christian Jung1 1Plant Breeding Institute, Christian-Albrechts University of Kiel, Olshausenstrasse 40, 24098 Kiel, Germany and 2Nordsaat Saatzucht GmbH, Hofweg 8, 23899 Segrahn, Germany

Abstract Root lesion nematodes of the genus Pratylenchus are important pests in crop cultivation that cause severe damage to crops throughout the world. P. neglectus is one of the most important members of this genus. The present study aimed to select barley accessions with resistance to P. neglectus in a greenhouse resistance test and to detect resistance quantitative trait loci (QTLs). Infection rates have been found to vary greatly among different barley accessions; however, immunity could not be found. An existing Igri £ Franka doubled-haploid mapping population was used to map resistance genes after artiﬁcial inoculation with P. neglectus under controlled environment. QTLs were found with a likelihood of odds score between 2.71 and 6.35 and explaining phenotypic variation of 8 to 16%.

Keywords: Hordeum vulgare; pest resistance; quantitative trait loci; root lesion nematodes

Introduction RLN have been identified as major pests in wheat cultivation (Taylor et al., 2000; Ogbonnaya et al., 2008). Exten- Root lesion nematodes (RLN) of the genus Pratylenchus sive work has been carried out in Australia to map are significant pests in crop cultivation throughout the quantitative trait loci (QTLs) for RLN resistance using world. They are polyphagous in nature and feed on different mapping populations in wheat. Williams et al. several crops of economic importance like cereals, coffee, (2002) mapped the P. neglectus resistance locus Rlnn1 corn, banana, legumes, potato, peanut and many fruits. in the Australian wheat cultivar Excalibur using a combi- They are migratory endoparasites and cause severe root nation of bulked segregant analysis and genetic mapping. damage on a wide range of crops while feeding mainly In addition, several other RLN resistance QTLs were in the cortical parenchyma. Some of the commonly mapped by studying the inheritance of RLN resistance observed symptoms of infected plants are (1) massive in wheat. Zwart et al. (2010) mapped four QTLs for plant tissue necrosis, (2) sloughing of cortical and epider- P. thornei and P. neglectus resistance in a doubled-hap- mal cells, (3) retarded development of lateral roots in loid (DH) population developed from a cross between terms of length and number and (4) fewer root hairs the synthetic hexaploid wheat line CPI133872 and (Taylor et al., 1999; Vanstone and Russ, 2001). Till the bread wheat Janz., designated as QRlnt.lrc-6D.1, today, 68 species of this genus are known. In Australia, QRlnt.lrc-6D.2, QRlnn.lrc-6D.1 and QRlnn.lrc-4D.1 on linkage groups 6DL, 6DS and 4DS. In the northern parts of Germany, enormous yield loss has been reported in winter barley, which was caused by P. neglectus, P. crenatus, P. fallax and P. penetrans (Hesselbarth, 2006). * Corresponding author. E-mail: [email protected] †Present address: Syngenta Seeds GmbH, Alte Reeser Straße 95, In the first phase of this project, screening of a 47 533 Kleve, Germany large collection of barley germplasm was carried out. Root lesion nematode resistance in barley 237 Five hundred and sixty-five barley accessions encom- accessions (129), Asian-African barley accessions (213) passing cultivated (Hordeum vulgare) and wild species and Australian barley accessions (5). As a result of the (H. spontaneum) were screened for resistance against large number of plants, the 565 accessions had to be P. neglectus (Keil et al., 2009). In the second part, a tested in three different experiments. Each experiment biparental DH mapping population was used to map contained a representative number of accessions from QTLs. Greenhouse tests are comparatively costly and each geographical region. For each experiment, about much time consuming. Therefore, a molecular marker 186 to 190 accessions were tested in a completely approach shall be implemented to identify favourable randomized block design. The accessions of each QTLs which reduce nematode infection rates. Results experiment were tested as single plant in six replications. shall be the basis for establishing a marker test to replace A validation experiment was carried out in the following expensive and time-consuming greenhouse test. way: 5% of the most resistant accessions of each experiment, regarding their overall rank, were selected and tested with six repeats under the conditions described Material and methods above. All the experiments were conducted in the glasshouse and in the climate chamber with 238C day and A total number of 565 barley accessions were selected for 188C night temperature. Seeds were germinated on wet screening with P. neglectus, comprising of 375 winter, 30 filter paper at 268C for 1 d in the dark. Each seedling spring barley lines (H. vulgare ssp. vulgare) and 160 wild was placed in a 20 cm3 tube (12 cm (H) height £ 2cm species accessions (H. vulgare ssp. spontaneum). As a (B) thickness 2.3 mm) filled with steam-sterilized sand. result of limited glasshouse facilities, it was impossible Seedlings were inoculated with 400 P. neglectus juveniles to test all accessions at a time. The plant material was in 1 ml water medium after 10 d. Further steps were fol- grouped to its geographical origin (four groups): Euro- lowed as described in Keil et al. (2009). After 12 weeks, pean barley accessions (218), North American barley the plants were uprooted and nematodes were extracted

(a) (b) 7.0 4.0

5.6 3.2

2.5 4.2 2.4 0 0 LOD LOD 2.8 2.5 1.6

1.4 0.8

0.0 0.0 0.0 17.0 33.0 50.0 66.0 83.0 100.0 116.0 133.0 149.0 166.0 0.0 22.0 43.0 65.0 87.0 109.0 130.0 152.0 174.0 195.0 217.0 Chromosome 3 - (3H) cM Chromosome 5 - (5H) cM (c) (d) 4.0 3.0

2.5 3.2 2.4

2.5 2.4 1.8 0 0 LOD LOD 1.6 1.2

0.8 0.6

0.0 0.0 0.0 17.0 33.0 50.0 67.0 84.0 100.0 117.0 134.0 150.0 167.0 0.0 19.0 37.0 56.0 75.0 94.0 112.0 131.0 150.0 168.0 187.0 Chromosome 6 - (6H) cM Chromosome 7 - (7H) cM Fig. 1. (a–d) Chromosomal location of QTL for P. neglectus resistance in the barley DH population Igri £ Franka. 238 S. Sharma et al. from the sand and the roots were chopped using a barley pathogens such as scald (Rhynchosporium secalis, Baermann funnel placed in a misting chamber for 5 d. Rrs) (Graner and Tekauz, 1996; Garvin et al., 2000), Nematode suspension were collected in a bottle and Pyrenophora teres (net type blotch disease; Graner and placed at 58C before counting. For QTL mapping, 126 Tekauz, 1996), and barley yellow dwarf virus (Ryd2;

F1 anther-derived DH lines from a cross Igri £ Franka Collins et al. 1996). The data provide clear evidence of were used (Graner et al., 1991). A molecular map was a polygenic inheritance of RLN resistance in barley with constructed by integrating diversity array technology major QTL having a big impact on infection rates. The (DarT) markers into the already available Igri £ Franka tightly linked markers ﬂanking the QTLs will be turned map (Stein et al., 2007). QTL analysis was carried out into diagnostic markers for marker-assisted selection of by composite interval mapping using the program QTL resistant plants from segregating offspring. Cartographer V2.5 (Wang et al., 2010).

Results and discussion Acknowledgements

A representative collection of 565 cultivated and wild The authors gratefully acknowledge Ba¨rbel Wohnsen, barley accessions was tested for P. neglectus resistance Cay Kruse and Ines Schu¨tt for their excellent technical in three different experiments. The mean number of assistance during the greenhouse resistance tests. This nematodes/plant in the three experiments was 3335, work was financially supported by the BLE grant no. 2920 and 1546 (P ¼ 0.0001), respectively. Fifty per cent 28-1-41.030-06 and the Nordsaat GmbH (Bo¨hnshausen, of the accessions had 1564 to 2988, 5% had below 982 Germany). and another 5% had above 5048 nematodes/plant. There was no barley accession without any nematode infection, i.e. showing immunity to P. neglectus. The References number of nematodes/plant ranged from 350 to 12,000. In a verification experiment with 32 barley accessions, Collins NC, Paltridge NG, Ford CM and Symons RH (1996) The Yd2 gene for barley yellow dwarf virus resistance maps five accessions namely ‘BCB-39’, ‘AC Queens’, ‘BYDV close to the centromere on the long arm of barley chromo- 17’, ‘AC Legend’ and ‘Beysehir’ were identified as moder- some 3. Theoretical and Applied Genetics 92: 858–864. ately resistant. Among the ten least susceptible accessions, Garvin DF, Brown AHD, Raman H and Read BJ (2000) Genetic three Turkish and only one German accessions were mapping of the barley Rrs14 scald resistance gene with found (Keil et al., 2009). In general, German accessions RFLP, isozyme and seed storage protein markers. Plant Breeding 119: 193–196. had a tendency for high susceptibility to P. neglectus Graner A and Tekauz A (1996) RFLP mapping in barley of a infection, reflecting the lack of selection pressure. In dominant gene conferring resistance to scald (Rhynchos- Turkey, breeders have selected for resistance towards porium secalis). Theoretical and Applied Genetics 93: Pratylenchus species over decades due to a high infection 421–425. pressure. On the other hand, in Germany, nematode Graner A, Jahoor A, Schondelmaier J, Siedler H, Pillen K, Fischbeck G, Wenzel G and Herrmann RG (1991) resistance was not a breeding aim because this problem Construction of an RFLP map of barley. Theoretical and arose only recently mainly due to narrow crop rotations. Applied Genetics 83: 250–256. To unravel the genetics of RLN resistance, an Igri £ Hesselbarth C (2006) Freilebende Wurzelnematoden. Probleme Franka DH population was tested under greenhouse in engen “Getreide-Raps-Dauergru¨n Fruchtfolgen” in and climate chamber conditions. The means of nematode Schleswig-Holstein. Getreide Magazin 2: 118–123. Keil T, Laubach E, Sharma S and Jung C (2009) Screening for counts of both parents, Igri (861) and Franka (1591), resistance in the primary and secondary gene pool of were significantly different (P , 0.05). Among the DHs, barley against the root lesion nematode Pratylenchus a high phenotypic variation was observed for P. neglectus neglectus. Plant Breeding 128: 436–442. infection. Transgressive segregants were also observed in Ogbonnaya FC, Imtiaz M, Bariana HS, Mclean M, Shankar MM, the population which indicates that favourable alleles are Hollaway GJ, Trethowan RM, Lagudah ES and van Ginkel M (2008) Mining synthetic hexaploids for multiple disease dispersed between both parental lines. For mapping, resistance to improve bread wheat. Australian Journal of DArT markers and previously mapped restriction frag- Agricultural Research 59: 421–431. ment length polymorphisms were used. Five QTLs were Stein N, Prasad M, Uwe S, Thiel T, Zhang H, Wolf M, Kota R, mapped with a likelihood of odds score between 2.71 Varshney RK, Perovic D, Grosse I and Graner A (2007) and 6.35 and explaining phenotypic variation of 8 to A 1000 loci transcript map of the barley genome – new anchoring points for integrative grass genomics. 16%, as shown in the Fig. 1(a–d). Some QTL positions Theoretical and Applied Genetics 114: 823–839. are coincident with previously identified QTLs or major Taylor SP, Vanstone VA, Ware AH, Mckay AC, Szot D and Russ genes conferring resistance to a diverse spectrum of MH (1999) Measuring yield loss in cereals caused by root Root lesion nematode resistance in barley 239 lesion nematodes (Pratylenchus neglectus and P. thornei) Wang S, Basten CJ and Zeng ZB (2010) Windows QTL Cartogra- with and without nematicide. Australian Journal of pher V2.5. Program in Statistical Genetics, North Carolina Agricultural Research 50: 617–622. State University. Taylor SP, Hollaway GJ and Hunt CH (2000) Effect of field crops Williams J, Taylor P, Bogacki P, Pallotta M, Bariana S and on population densities of Pratylenchus neglectus and Wallwork H (2002) Mapping of the root lesion nematode P. thornei in Southeastern Australia; part 1: P. neglectus. (Pratylenchus neglectus) resistance gene Rlnn1 in wheat. Journal of Nematology 32: 591–599. Theoretical and Applied Genetics 104: 874–879. Vanstone VA and Russ MH (2001) Ability of weeds to host Zwart RS, Thompson JP, Milgate AW, Bansal UK, Williamson PM, the root lesion nematodes Pratylenchus neglectus and Raman H and Bariana HS (2010) QTL mapping of multiple P. thornei – I. Grass weeds. Australian Journal of Plant foliar disease and root-lesion nematode resistances in Pathol ogy 30: 245–250. wheat. Molecular Breeding 26: 107–124. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 240–242 ISSN 1479-2621 doi:10.1017/S1479262111000451

Allelic variation at the EF-G locus among northern Moroccan six-rowed barleys

Takahide Baba1,2, Ken-ichi Tanno1,3, Masahiko Furusho2 and Takao Komatsuda1* 1National Institute of Agrobiological Sciences, Tsukuba 305-8602, Japan, 2Fukuoka Agricultural Research Center, Chikushino 818-8549, Japan and 3Faculty of Agriculture, Yamaguchi University, Yamaguchi 753-8511, Japan

Abstract A germplasm panel of 52 six-rowed barley landraces from northern Morocco was analysed by a Cleaved Ampliﬁed Polymorphic Sequences (CAPS) assay of a fragment of the elongation factor G (EF-G) gene. Forty-nine of these accessions carried allele A, and the other three carried allele D. The latter all originated from a narrow region close to the border with Algeria, whereas the former were represented across the whole collection area. Since six-rowed D allele carriers are present in North Africa, along with both two-rowed cultivated and wild barleys, it is likely that the European six-rowed barley varieties carrying the D allele have Moroccan parentage.

Keywords: elongation factor G; Hordeum vulgare; Morocco; row type

Introduction Polymorphic Sequences (CAPS) assay) are generally correlated with, respectively, the six- and two-rowed The row type of the barley (Hordeum vulgare) spike is type, although some recombinant genotypes existed controlled by the vrs1 locus located on chromosome (Komatsuda et al., 1999a). When the assay was applied 2H. The vrs1 locus encodes a HD-ZIP I type transcription to a collection of 65 cultivated barley accessions, Tanno factor specific to barley (Komatsuda et al., 2007), which et al. (1999) uncovered a third allele D of EF-G, which may interact with the other HD-ZIP I transcription was restricted to six-rowed types. A wider survey of 464 factor encoded by HvHox2, which is a gene highly accessions showed that the A allele is distributed world- conserved in cereal species (Sakuma et al., 2010). The wide, whereas the D allele only occurs in southern six-rowed spike is restricted to cultivated barley (subsp. Europe (Tanno et al., 2002). The D allele has, however, vulgare), and is believed to have evolved from the two- also been identified in one two-rowed cultivated barley rowed type (Komatsuda et al., 2007) around the time of (cv. Palmella Blue) from an unidentified north African the crop’s domestication (c. 6500 BCE) (Harlan, 1995). country and all the seven Moroccan wild barley applied The cDNA sequence cMWG699 (Graner et al., 1991), (subsp. spontaneum) from Morocco. Molina-Cano et al. which is a fragment of an elongation factor gene EF-G, (1987) have suggested that Moroccan wild barley was a maps within 0.1 cM of the vrs1 locus (Komatsuda et al., possible progenitor of southern European cultivated 1999a), and has been applied for phylogenetic studies barleys, and given the probability that the six-rowed in the Hordeum and related species (Komatsuda et al., spike evolved from the two-rowed one, it is likely that 1999b, 2009; Blattner, 2009). The two co-dominant EF-G the D allele originated in Moroccan subsp. spontaneun. alleles (A and K, defined by a Taq I Cleaved Amplified If this supposition is correct, then the expectation is that the allele should still be represented among Moroccan six-rowed barleys. Thus, here we set out to screen a collection of six-rowed Moroccan barleys for their allelic * Corresponding author. E-mail: [email protected] constitution at the EF-G CAPS marker. Allelic variation at the EF-G locus in barley 241 Materials and methods narrow area around the city of Ojuda (Fig. 1), which lies close to the border with Algeria, at an altitude of A germplasm panel of 52 six-rowed landraces collected 1250–1560 m (Supplementary Table S1, available online from northern Morocco (Kuwabara et al., 1988) was only at http://journals.cambridge.org). The second assembled (Supplementary Table S1, available online dCAPS analysis did not reveal any polymorphism only at http://journals.cambridge.org). Their collection among the A allele accessions, so no A–D intermediate sites ranged from sea level to an altitude of .2000 m (as found in a Chinese cultivar by Komatsuda et al., (Fig. 1). Four seeds from each accession were grown to 1999b) was present. isolate a single pure line to provide a source of DNA, which was extracted from leaves of 9-d-old seedlings Discussion following the protocol described by Komatsuda et al. (1998). This DNA provided the template for PCR amplifi- The EF-G fragment has offered a means of exploring the cation of the EF-G fragment, based on the primer pair evolution of the six-rowed ear. The distribution of alleles T7-3 and T3-3 (Tanno et al., 1999). The resulting ampli- within this sequence has suggested that the A and D cons were Taq I restricted as described by Komatsuda alleles arose prior to the appearance of the six-rowed et al. (1998). A further Derived Cleaved Amplified Poly- ear, thus implying that a two-rowed barley carrying the morphic Sequences (dCAPS) assay was based on the A allele was the ancestor of the six-rowed barleys carry- primer pair T7-3 and T3-4 (5’-AACTCTGAGAATAAAATG- ing the A allele, and likewise a D allele two-rowed type GCTAGCG), using Hha I restriction to recognize the was the ancestor the current D allele six-rowed types additional nucleotide polymorphism between the A and (Tanno et al., 1999, 2002). The hypothesis that the the D allele (Tanno et al., 1999). six-rowed D allele carriers evolved from a Moroccan two-rowed D allele individual, before spreading toward Results southern Europe, requires that six-rowed D allele carriers can be located in Morocco. The EF-G amplicon resolved as a single fragment of In that event, we found three such accessions among a size , 490 bp in all 52 accessions. Its digestion with collection of 52 six-rowed northern Moroccan landraces. Taq I produced two restriction patterns, corresponding to The result confirms the plausibility that the progenitor of the A and D alleles (Tanno et al., 1999). Forty-nine of six-rowed D allele carriers is a wild barley from Morocco. these accessions carried allele A, and the other three D allele wild barleys are distributed throughout Algeria, carried allele D (Supplementary Table S1, available online Tunisia, Egypt and Spain (T. Komatsuda, unpublished data). only at http://journals.cambridge.org). The A allele car- The result also supports the suggestion made by Molina-Cano riers were distributed across the whole collection et al. (1987) that, on the basis of some isozyme and mor- region, whereas the D allele ones originated from a phological data, a number of Spanish six-rowed barleys were descended from Moroccan subsp. spontaneum types. Tanno et al. (1999) have shown that the EF-G fragment A and D alleles reflect four nucleotide substitutions, and that the age of the divergence between these alleles is in the range 138,000–830,000 years. The evolution of the six-rowed spike appears to have occurred independently in an A allele and a D allele carrier (Komatsuda et al., 2007; Saisho et al., 2009). Here, we have reported that the D allele is present in the Moroccan population of six- rowed (cultivated) barley, while Tanno et al. (2002) showed that it is also present in Moroccan subsp. spontaneum. Thus, the likelihood is that the lineage of six- rowed barley D allele carriers involved a Moroccan subsp. spontaneum followed by a Moroccan two-rowed vulgare.

References

Fig. 1. Geographical distribution of the EF-G/Taq I poly- Blattner FR (2009) Progress in phylogenetic analysis and a new morphism across northern Morocco. Closed circle, A allele; infrageneric classiﬁcation of the barley genus Hordeum open circle, D allele. (Poaceae: Triticeae). Breeding Science 59: 471–480. 242 T. Baba et al. Graner A, Jahoor A, Schondelmaier J, Siedler H, Pillen K, Kuwabara T, Furusho M and Miyagawa S (1988) Collaborative Fischbeck G, Wenzel G and Herrmann RG (1991) Construc- exploration of wheat and barley in Morocco and the tion of an RFLP map of barley. Theoretical and Applied Syrian Arab Republic 1987. Shokutanho. Japan: National Genetics 83: 250–256. Institute of Agrobiological Resources, pp. 108–130. Harlan JR (1995) Barley. In: Smartt J and Simmonds NW (eds) Molina-Cano JL, Fra-Mon P, Salcedo G, Aragoncillo C, Roca de Evolution of Crop Plants. 2nd ed. Harlow: Longman, Togores F and Gracia-Olmedo F (1987) Morocco as a pp. 140–147. possible domestication center for barley: biochemical and Komatsuda T, Pourkheirandish M, He C, Azhaguvel P, Kanamori agromorphological evidence. H, Perovic D, Stein N, Graner A, Wicker T, Tagiri A, Theoretical and Applied Lundqvist U, Fujimura T, Matsuoka M, Matsumoto T and Genetics 73: 531–536. Yano M (2007) Six-rowed barley originated from a Saisho D, Pourkheirandish M, Kanamori K, Matsumoto T and mutation in a homeodomain-leucine zipper I–class Komatsuda T (2009) Allelic variation of row type gene homeobox gene. The Proceedings of National Academy Vrs1 in barley and implication of the functional divergence. of Sciences USA 104: 1424–1429. Breeding Science 59: 621–628. Komatsuda T, Nakamura I, Takaiwa F and Oka S (1998) Sakuma S, Pourkheirandish M, Matsumoto T, Koba T and Development of STS markers closely linked to the vrs1 Komatsuda T (2010) Duplication of a well-conserved locus in barley, Hordeum vulgare. Genome 41: 680–685. homeodomain-leucine zipper transcription factor gene in Komatsuda T, Li W, Takaiwa F and Oka S (1999a) High resol- barley generates a copy with more speciﬁc functions. ution map around the vrs1 locus controlling two- and Functional and Integrative Genomics 10: 123–133. six-rowed spikes in barley, Hordeum vulgare. Genome Tanno K, Takaiwa F, Oka S and Komatsuda T (1999) A nucleo- 42: 248–253. tide sequence linked to the vrs1 locus for studies of differ- Komatsuda T, Tanno K, Salomon B, Bryngelsson T and Bothmer entiation in cultivated barley (Hordeum vulgare L.). RV (1999b) Phylogeny in the genus Hordeum based on nucleotide sequences closely linked to the vrs1 locus Hereditas 130: 77–82. (row number of spikelets). Genome 42: 973–981. Tanno K, Taketa S, Takeda K and Komatsuda T (2002) A DNA Komatsuda T, Salomon B and Bothmer RV (2009) Evolutionary marker closely linked to the vrs1 locus (row-type gene) process of Hordeum brachyantherum 6 £ and related indicates multiple origins of six-rowed cultivated barley tetraploid species revealed by nuclear DNA sequences. (Hordeum vulgare L.). Theoretical and Applied Genetics Breeding Science 59: 611–616. 104: 54–60. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 243–246 ISSN 1479-2621 doi:10.1017/S147926211100030X

Comparison of genomic and EST-derived SSR markers in phylogenetic analysis of wheat

Agata Gadaleta*, Angelica Giancaspro, Silvana Zacheo, Domenica Nigro, Stefania Lucia Giove, Pasqualina Colasuonno and Antonio Blanco Department of Environmental and Agro-Forestry Biology and Chemistry, University of Bari, Via Amendola, 165/A, 70126 Bari, Italy

Abstract Microsatellite markers (simple sequence repeats, SSRs) are used for a wide range of crop genetic and breeding applications, including genetic diversity assessment, phylogenetic analysis, genotypic proﬁling and marker-assisted selection. Genomic SSR (gSSR) have attracted more attention because of abundance in plant genome, reproducibility, high level of polymorphism and codominant inheritance. Recently, the availability of data for expressed sequence tags (EST), has given more emphasis to EST-derived SSRs, which belong to the transcribed regions of DNA, and are expected to be more conserved and have a higher transferability rate across species than gSSR markers. In the present study, several gSSR and EST-SSR markers were investigated for their transferability and level of DNA polymorphism in different ancestral tetraploid and diploid Triticum and Aegilops species. The same gSSR and EST-SSR markers were also evaluated for their applicability in the phylogenetic analysis of wheat. Both gSSR and EST-SSR markers showed differences for the average transferability rate and the number of alleles/ locus. Phylogenetic trees based on gSSR and EST-SSR markers were in accordance with phylogenetic relations based on cytogenetic and molecular analyses.

Keywords: expressed sequence tags-simple sequence repeats; genomic simple sequence repeats; phylogenetic analysis; transferability; wheat

Introduction SSRs, which belong to the transcribed regions of DNA and are expected to be more conserved and have a Molecular markers are used for a wide range of purposes higher rate of transferability across species than genomic in crop genetics and breeding, including genetic linkage SSR markers (Rudd, 2003). The objectives of the present and comparative mapping, positional cloning, genetic work were to test the transferability and polymorphisms diversity assessment, phylogenetic analysis, genotypic of gSSR and EST-SSR markers in Triticum and Aegilops proﬁling, quantitative trait loci and marker-assisted selec- species closely related to cultivated wheats, and to test tion. In recent years, genomic microsatellites (or simple their applicability for wheat phylogenetic analyses. The sequence repeats, gSSR) have attracted more attention use of polymorphic SSR markers either for the character- because of abundance in plant genome, reproducibility, ization and evaluation of germplasm or for phylogenetic high level of polymorphism and codominant inheritance analysis of wheat was also discussed. (Nicot et al., 2004). The recent wide availability of data for expressed sequence tags (ESTs) increased EST-derived Materials and methods

Twelve species or sub-species were analyzed: Triticum * Corresponding author. E-mail: [email protected] aestivum (AABBDD), T. turgidum ssp. durum (AABB), 244 A. Gadaleta et al.

T. timopheevii (AAGG); T. turgidum ssp. dicoccoides null alleles or failed PCR amplifications. The results of (AABB), T. monococcum, T. urartu, Aegilops squarrosa, DNA amplification from the genotypes of the five Aegilops A. speltoides (S), A. bicornis, A. longissima, A. sharonen- species indicated that the polymorphism was genotype sis and A. searsii. A total of 20 genotypes, including dependent. The mean number of alleles among the species ancestral tetraploid and diploid Triticum and Aegilops was 1.71 (ranging from 1.90 for A. speltoides to 3.43 for T. species, were finally used to assess SSR applicability for aestivum) for gSSR markers and 1.88 (ranging from 1.26 phylogenetic analysis of wheat. for A. searsii to 1.84 for T. aestivum) for EST-SSR markers. SSR and EST-SSR primer sequences developed by La Polymorphism was relatively higher in the source species Rota et al. (2005), annealing temperature and expected than in related species. As expected, polymorphism of PCR product size are reported in the web site http:// EST-SSR markers was lower than that of gSSR markers, wheat.pw.usda.gov. likely because the higher level of conservation of DNA DNA amplifications were carried out as described by sequences belonging to the transcribed region of the Gadaleta et al. (2009). genome, as previously reported (Nicot et al., 2004). Results of DNA amplification from the genotypes of the Aegilops species indicated that the rate of transferability was genotype dependent. Cross-species transferability Results and discussion of EST-SSRs was observed in 670 out of 1115 combinations (60.1%), whereas gSSRs gave amplified products Primer pairs of 79 gSSRs and 61 EST-SSRs were tested in 20 in 845 out of 1407 combinations (60.0%). As expected, genotypes belonging to the 12 species or subspecies of transferability was higher in the Triticum species with the Triticum–Aegilops complex under the same PCR respect to Aegilops species (Fig. 1(b)), as also reported conditions as originally applied for amplification in wheat. by Sourdille et al. (2001). The 79 gSSR and 61 EST-SSR markers were polymorphic DICE genetic similarity coefficients were used to across the 12 species or subspecies of the Triticum– prepare dendrograms using the UPGMA method. Aegilops complex (Fig. 1(a)). Among the 2800 data points The dendrograms (Fig. 2) based on gSSR and EST-SSR (140 SSR £ 20 genotypes), c. 10% were missing data, true bands were not significantly different. Each species was separated and the genotypes tested for the Aegilops species were always grouped together. Phylogenetic (a) gSSR EST-SSR trees were consistent with cytotaxonomical and molecu- 4.0 ) n 3.5 lar data on species relationships in the Triticum–Aegilops 3.0 complex. Cluster analysis showed that: 2.5 2.0 1.5 (1) Triticum species were separated from Aegilops 1.0 0.5 species; Alleles/primer pair ( 0.0 2 1 2 2 3 1 2 3 1 1 2 3 1 (2) T. monococcum was clustered with T. urartu, both species having a common A genome;

T. urartu (3) tetraploid and hexaploid species of formed T. t.durum Triticum A. searsii A. searsii T. aestivum A. bicornis A. bicornis A. bicornis A. squarrosa T. timopheevii a close group; A. speltoides A. speltoides A. speltoides A. longissima A. longissima T. t.dicoccoides T. monococcum A. sharonensis A. sharonensis A. sharonensis (4) close clustering of A. bicornis, A. longissima and

(b) 100 A. sharonensis is consistent with cytotaxonomical 90 gSSR EST-SSR data; A. searsii was less clustered with the above species; 80 70 (5) within the section Sitopsis, separation of A. speltoides 60 from the remaining four species (A. bicornis, 50 A. longissima, A. sharonensis and A. searsii) was 40 30 consistent with the Eig’s classiﬁcation on morpho- Polimorphism (%) 20 logical traits; 10 (6) A. squarrosa seemed to be closer to the B genome 0 1 2 3 1 2 1 2 1 2 3 1 2 3 donor species in the EST-SSR dendrogram and to the A genome donor species in the genomic T. urartu

T. t.durum A. searsii A. searsii T. aestivum A. bicornis A. bicornis A. bicornis dendrogram. A. squarrosa

T. timopheevii A. speltoides A. speltoides A. speltoides A. longissima A. longissima T. t.dicoccoides T. monococcum A. sharonensis A. sharonensis A. sharonensis Fig. 1. Average number of alleles/primer pair of wheat gSSR and EST-SSR markers (a) and their transferability in 20 Thus, we can conclude that wheat EST-SSR accessions of Triticum and Aegilops species (b). markers show a high transferability across a range of SSR markers in phylogenetic analysis of wheat 245 A. speltoides 1 (a) A. speltoides 2 A. speltoides 3 A. searsii 1 A. searsii 2 A. bicornis 1 A. bicornis 2 A. bicornis 3 A. sharonensis 3 A. sharonensis 1 A. sharonensis 2 A. longissima 1 A. longissima 2 A. squarrosa T. urartu T. monococcum T. timopheevii T. t.dicoccoides T. t.durum T. aestivum 0.00 0.25 0.50 0.75 1.00 Coefficient (b) A. searsii 1 A. searsii 2 A. bicornis 1 A. bicornis 2 A. bicornis 3 A. sharonensis 1 A. sharonensis 2 A. sharonensis 3 A. longissima 1 A. longissima 2 A. squarrosa A. speltoides 1 A. speltoides 2 A. speltoides 3 T. urartu T. monococcum T. t.dicoccoides T. t.durum T. aestivum T. timopheevii 0.00 0.25 0.50 0.75 1.00 Coefficient Fig. 2. Dendograms based on (a) genomic SSR and (b) EST-SSR markers.

species. This transferability makes them a powerful Acknowledgements tool to work on orphan wild species, where less effort has been devoted to develop genomic resources such This research was supported by a grant from Ministero as molecular markers. Wild species are an important dell’Universita` e della Ricerca, Italy, project ‘AGROGEN’. source of both abiotic and biotic resistances, and molecular markers are precious tools to use and reduce References the linkage drag derived from the introgressions of genes (location and size of the introgression) from Gadaleta A, Giancaspro A, Giove SL, Zacheo S, Mangini G, these species. Simeone R, Signorile A and Blanco A (2009) Genetic and 246 A. Gadaleta et al. physical mapping of new EST-derived SSRs on the A and B repeat (SSR) markers from wheat expressed sequence tags genome chromosomes of wheat. Theoretical and Applied (ESTs). Theoretical and Applied Genetics 109: 800–805. Genetics 118: 1015–1025. Rudd S (2003) Expressed sequence tags: alternative or comp- La Rota M, Kantety RV, Yu JK and Sorrells ME (2005) Non- lement to whole genome sequences? Trends in Plant random distribution and frequencies of genomic and Science 8: 321–329. EST-derived microsatellite markers in rice, wheat, and Sourdille P, Tavaud M, Charmet G and Bernard M (2001) Trans- barley. BMC Genomics 6: 23–30. ferability of wheat microsatellites to diploid Triticeae Nicot N, Chiquet V, Gandon B, Amilhat L, Legeai F, Leroy P, species carrying the A, B and D genomes. Theoretical Bernard M and Sourdille P (2004) Study of simple sequence and Applied Genetics 103: 346–352. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 247–250 ISSN 1479-2621 doi:10.1017/S1479262111000311

Exploring the genetic diversity of the DRF1 gene in durum wheat and its wild relatives

Domenico Di Bianco1,2, Karthikeyan Thiyagarajan1,2, Arianna Latini1†, Cristina Cantale1, Fabio Felici1 and Patrizia Galefﬁ1* 1ENEA Casaccia (UTAGRI-GEN), Rome, Italy and 2Scuola Superiore Sant’Anna, Pisa, Italy

Abstract A drought-related gene belonging to the Dehydration Responsive Element Binding protein (DREB) family has been reported and characterized in durum wheat. Unlike other DREB-homologous genes, it consists of four exons and three introns and produces three transcripts by an alternative splicing mechanism. The gene sequence was analysed in a number of varieties/ lines/accessions of durum wheat, triticale and in wheat genome donors, Aegilops speltoides, A. tauschii and Triticum urartu, in order to evaluate the variability and to detect other interesting molecular features. Herewith, some results are presented. In the exon 1, a single sequence repeat codifying for a stretch of alanine residues variable in length (from 3 to 7), was identiﬁed. A novel non-autonomous transposon was identiﬁed, encompassing the intron1–intron3 region and this was characterized in detail. Part of the exon 4, containing the APetala2 (AP2) domain, responsible for DNA recognition and binding, was isolated and sequenced in a collection of Aegilops species and A. speltoides accessions from the Fertile Crescent, a region characterized by a high wheat biodiversity. Although the 338 bp-long analysed sequences did not reveal substantial differences in the polymorphic patterns, using a geographic subdivision with three clusters (east, centre and west), they completely separated Aegilops from the A. speltoides genus.

Keywords: AP2 domain; biodiversity; DREB-family; single sequence repeat; transposon

Introduction accessions of durum wheat, triticale, wheat genome donors (A. speltoides, A. tauschii and Triticum urartu) TdDRF1 (T. durum dehydration responsive factor 1), a as well as in other related plants. DREB2-related gene that was isolated and characterized in durum wheat (Latini et al., 2007), belongs to the AP2 gene family and is highly homologous to the barley HvDRF1 gene (Xue and Loveridge, 2004) and to Material and methods the bread wheat wdreb2 gene (Egawa et al., 2006). Eight T. durum genotypes (total 67 accessions), a few These genes share a complex gene structure, consisting accessions of T. urartu, A. tauschii and triticale, as well of four exons and three introns, and produce three as 69 accessions of A. speltoides v speltoides, A. speltoides transcript variants by means of alternative splicing. v ligustica and other Aegilops were grown in the green- In the present investigation, the different parts of the house and used for DNA extraction. gene have been systematically analysed in several The amplified fragments (PCR thermal cycle reaction: initial denaturation 948C for 5 min, 948C for 1 min, 558C for 1 min, 728C for 1 min and 30 s and final extension at 728C for 7 min, then at 48C for storage) were gel- * Corresponding author. E-mail: galeffi@enea.it w w † purified, cloned (pCR II-TOPO vector by Invitrogen, Present address: Canadian Forest Service, Que´bec City, Quebec, Canada. USA) and sequenced (ABI 3730 DNA analyzer; Applied D. D. Bianco and K. Thiyagarajan contributed equally to this work. Biosystems, USA) following standard procedures. 248 D. D. Bianco et al. An exon 1 region of 150 bp at the beginning of protein at the N terminus of exon 1, position 10, and codes for a CoDing Sequence (CDS), was sequenced in 100 genotypes stretch of alanine residues variable in length (from 3 (E1 For: 50-AAGTCGACGCGGCGAA 30;SSR1Rev:50-CCG- to 7). Another study on SSR loci in expressed sequences GGATCTCGAAGGGTG 30). suggested the potential correlation between mutational An exon 4 part, from 400 bp to 1 kb long, including the events within SSR repeats and the evolutionary relation- whole AP2 domain coding region, was sequenced in 168 ships across taxa (Rossetto et al., 2002). It is worthwhile samples of Aegilops (E4 FOR: 50-ATGATCCACAGGGT- pointing out that the shortest SSRs belong mainly to the GCAA 30; E4 Rev: 50-GGTCCACCATTTGATCTTCATT 30). ancestor of Triticum and to Brachypodium distachyon. Six full-length DRF1 gene sequences, namely FJ858188, Variability was also observed, resulting from a single FJ858187, FJ843102, EU089819, EU197052 and GU017675, nucleotide mutation (ALA to SER or THR), exclusively at were submitted to Genbank and used for transposon the beginning or at the end of the stretch. The ALA stretch identification and analysis. is present also in Leymus chinensis and B. distachyon, Several computational tools were used for phylo- suggesting that this SSR represents a shared feature of genetic and molecular evolutionary analyses (Rozas the Pooideae subfamily (Poaceae family), and is not pre- et al., 2003; Excoffier et al., 2005; Huson and Bryant, sent in other subfamilies such as Ehrhartoideae (Oryza 2006; Tamura et al., 2007), sequence analyses as well as sativa) and Panicoideae (Zea mays). A double mutation identification of repeats and palindromes (Kurz et al., changing ALA to LEU was observed in L. chinensis, while 2001; Kohany et al., 2006), secondary structure prediction a single change, GLU to VAL, was found in B. distachyon. and modelling (for details see Supplementary Table S1, Further investigations are needed to better understand available online only at http://journals.cambridge.org). the possible linkage of this SSR and its variability, both at genomic and transcriptomic level, in relationship with maps and phenotypic traits. Results and discussion

The various analyses provided interesting insights into Exons 2–3 the DRF1 gene which are outlined below, separately, for each exon. The six available full-length gene sequences from T. durum and Aegilops were analysed in order to identify possible transposable elements. Repeats, palindromes, Exon 1 tandem and inverted repeats were investigated. All the identiﬁed elements, namely a terminal inverted repeat Simple sequence repeat (SSR) genetic polymorphism (TIR-32 bp), a target site duplication (TSD-2 bp), an occurs due to variations in the number of repeated units, internal TSD (ITSD-4 bp), direct and reverse repeats and probably due, in turn, to slippage during DNA replication a variable number of tandem repeats revealed the pre- (Levinson and Gutman, 1987; Taylor et al., 1999). An SSR sence of a transposable element in the DRF1 gene was identiﬁed in the DRF1 gene (Fig. 1). This SSR is located which had never been described before.

cDNA from many durum cultivars, as Ciccio, and some pseudogene EAAAAAAAP 34% forms from A. speltoides ESAAAAALP 21% cDNA from pseudogene form from Ciccio durum cultivar

EAAAAAAAP cDNA from L. chinensis cDNA from many durum cultivars, as Cannizzo, and some gDNA from 16% EAAAAAATP T. urartu accessions EAAAAAAP 26% cDNA from many durum cultivars as Karalis, Duilio, Creso and Colosseo

ETAAAAAP 4% cDNA from Triticale Pollmer

EAAAAAP 4% gDNA from A. speltoides and A. tauschii accessions

EAAAAP 2% gDNA from A. speltoides

EAAATP 12% gDNA from many T. urartu accessions

VAAAP gDNA from B. distachyon

Fig. 1. Sequences and frequencies of the ALA-stretch in T. durum, A. speltoides, T. urartu, A. tauschii and in two related plants. (Brachypodium d.: Bd2:29505646.29515646, http://www.brachybase.org/blast/; Leymus c.: GenBank accession EU999998.1). Diversity of DRF1 gene in durum wheat 249 Thus, a novel transposable element, approximately Exon 4 1.5 kb in length, was identified between intron 1 and intron 3, including exon 2 and exon 3. Since no sequence Even if most of the variations intra- and interspecies are for transposase was found in the transposon identified, selectively neutral (Nei, 1987), there is an increasing it was classified as a non-autonomous transposon. interest in detecting genes, or genomic regions, that Furthermore, the presence of nested TIR-18bp was also affect the fitness of the organisms (Nielsen, 2005). As observed, the sequences of which are highly homo- the exon 4 sequence of DRF1 gene contains the region logous to the longest ones (Supplementary Fig. S1, codifying the AP2 domain, responsible for DNA recog- available online only at http://journals.cambridge.org). nition and binding, 168 sequences (137 from A. speltoides The new transposon is now included in Repbase and 31 from other Aegilops species) were analysed (see Report at http://www.girinst.org/2009/vol9/issue3/ to evaluate variability (Supplementary Fig. S2, available AsDRF1.html; Thiyagarajan et al., 2009). online only at http://journals.cambridge.org). All analysed genes showed the above observations, Overall, 279 sites were analyzed (excluding gaps/ with minor mismatch/deletion, mainly located at missing data) for all 168 sequences; 67 polymorphic 30 Terminal Inverted Repeat (TIR) that appears to be sites were identified (25 singleton variable sites, two var- the most variable part of the transposable element. iants; 35 parsimony informative sites, two variants; and The transposon is inserted into a poor CG (about 40%) seven parsimony informative sites, three variants). Out region and its core (exon 2 þ intron 2 þ exon 3 and few of 67 polymorphic sites, five were located in the AP2 bps of intron 1 and intron 3 at flanking sides) appears to domain and their effect was investigated in a structural be very similar (85% identity, 372 bp fragment, E ¼ 10269) model. The AP2 local geometry was not affected (data to a sequence from B. distachyon, suggesting that the not shown). transposon might have played a vital role in moving To investigate the evolutionary relationships among these exons during evolution of the Pooideae subfamily. the 168 sequences, a minimum spanning tree was built. Furthermore, it is tempting to hypothesize its possible Two main groups were observed, lineage I and lineage involvement in the mechanism of the alternative splicing II, separated by 21 mutation steps (Fig. 2). The neigh- regulation of the DRF1 gene. bour-joining tree, estimated using a Kimura-2 parameter

10.0 223_1

213_1 222_1 219_1 216_1 215_1

236_1 235_1 225_1 2065_3 237_1 2065_4 202_1217_1 2055_5 214_1 212_1 220_1 33_2 27_2 226_1 22_1 228_1 25_1 Lineage I 13_10 22_9 207_1 32_1 221_1 3_1 16_2 13_1 33_1 17_42114_2 205_1 211_1 49_5 12_9 13_8 269_1 16_10 206_1 234_1 49_4 229_1 230_1 232_1 32_4 413_8 231_1 18_9 227_1 233_1 413_2 201_1 413_1 348_4_2 18_2 32_4_1 13_7 348_10 1969_2 13_9 13_7_2 1795_3 8_1 24_8 14_6 1085_2 33_3_1 22_7 1_1 32_3 39_2 2_1 9_1 31_9 22_4 21_9 1085_1 16_9 31_8 33_3_2 348_7 17_2 16_7 16_5 15_9_2 Lineage II 18_1 2357_5 16_8 24_5 28_9 20_2 5_1 19_1 20_5 17_6 11_4 32_5 19_1 28_5 20_7 39_4 1085_4 11_9 32_2 24_1 28_10 33_5 24_4 31_7 21_4 26_6 31_1 348_9 33_4 39_5 1955_3 18_3_2 26_4 1_2 9_2 19_9 26_10 20_10 348_4_1 1085_3 16_4 15_2 49_2 27_3 39_1 28_8 31_2 18_3_1 15_10 38_9 30_3 21_8 12_10 12_2 21_6 11_2 28_8 38_8 38_10 11_7 24_6 21_5 12_7 12_5 20_6 15_9_1 15_6 30_2 12_1 15_7 49_1 30_5 21_7 19_2 19_6 27_6 19_10 19_3 19_7

Fig. 2. Minimum spanning tree of 168 genotypes of Aegilops calculated by an exon 4 region of the DRF1 gene. A. speltoides ssp ligustica is shown in light grey ( ), A. speltoides ssp speltoides is shown in black ( ), A. speltoides is shown in black ( ) and other Aegilops are shown in dark grey ( ). 250 D. D. Bianco et al. model, clustered in the same way, a 70% bootstrap value References (data not shown). Lineage I is a complex network, with two main groups, Egawa C, Kobayashi F, Ishibashi M, Nakamura T, Nakamura C ‘star-like’ pattern, including all except three A. speltoides and Takumi S (2006) Differential regulation of transcript accumulation and alternative splicing of a DREB2 homolog accessions. These three genotypes, classified as under abiotic stress conditions in common wheat. Genes A. speltoides (2065_3, 2065_4 and 2065_5), cluster in the Genetics and Systems 81: 77–91. lineage II, together with all Aegilops genotypes. Possibly, Excoffier L, Laval G and Schneider S (2005) Arlequin ver 3.0: an a reconsideration of the taxonomic assignment could be integrated software package for population genetics data proposed, even if more loci have to be analysed before analysis. Evolutionary Bioinformatics 1: 47–50. Huson DH and Bryant D (2006) Application of phylogenetic reaching a final conclusion. networks in evolutionary studies. Molecular Biology and A geographic subdivision was applied to investigate Evolution 23: 254–267. geographic frequency patterns, but the design, based Kohany O, Gentles AJ, Hankus L and Jurka J (2006) Annotation, on three clusters (east, centre and west) was unable to submission and screening of repetitive elements in Repbase: reveal substantial differences in polymorphic patterns, Repbase Submitter and Censor. BMC Bioinformatics 7: 474. Kurtz S, Choudhuri JV, Ohlebusch E, Schleiermacher C, Stoye J thus suggesting a neutral selection. and Giegerich R (2001) REPuter: the manifold applications A design based on taxonomy was also tested; the varia- of repeat analysis on a genomic scale. Nucleic Acids bility observed in the analysed DNA region was unable Research 29: 4633–4642. to distinguish ssp speltoides from ssp ligustica, while it Latini A, Rasi C, Sperandei M, Cantale C, Iannetta M, Dettori M, was able to distinguish Aegilops from A. speltoides (data Ammar K and Galeffi P (2007) Identification of a DREB-related gene in Triticum durum and its expression under water stress not shown). Analysis of molecular variance (AMOVA) conditions. Annals of Applied Biology 150: 187–195. was carried out testing the genetic structure (two Levinson G and Gutman GA (1987) Slipped-strand mispairing: a groups and four populations). The F-statistic was major mechanism for DNA sequence evolution. Molecular shown to be quite significant (F-statistic, i.e. specific Biology and Evolution 4: 203–221. Nei M (1987) Molecular Evolutionary Genetics. New York: population index (Fst) P value ¼ 0.00) and a substantial Columbia University Press. variation between and species A. speltoides Aegilops Nielsen R (2005) Molecular signatures of natural selection. was found, accounting for 62.34% of the total variation, Annual Review of Genetics 39: 197–218. even if not significant, P . 0.1. Due to missing data, Rossetto M, McNally J and Henry RJ (2002) Evaluating the poten- locus-by-locus AMOVA provided a better estimate and, tial of SSR flanking regions for examining taxonomic indeed, the data were significant (54.7%, P ¼ 0.00). relationships in the Vitaceae. Theoretical and Applied Genetics 104: 61–66. In conclusion, the detailed analysis of the sequence Rozas J, Sanchez-Delbarrio JC, Messeguer X and Rozas R (2003) of the DRF1 gene and its variability intra- and DnaSP, DNA polymorphism analyses by the coalescent and inter-species highlighted some interesting features char- other methods. Bioinformatics 19: 2496–2497. acterizing the complex structure of this gene, clearly Tamura K, Dudley J, Nei M and Kumar S (2007) MEGA4: contributing to our overall knowledge and thus Molecular Evolutionary genetics analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24: 1596–1599. opening new perspectives regarding the evolution and Taylor JS, Durkin JMH and Breden F (1999) The death of a micro- regulation of this gene. satellite: a phylogenetic perspective on microsatellite inter- ruptions. Molecular Biology and Evolution 16: 567–572. Thiyagarajan K, Latini A, Galeffi P and Porceddu E (2009) A non-autonomous DNA transposon inserted in the first Acknowledgements and third intron of the AsDRF1 gene. Repbase Reports 9(3): 726, ISSN# 1534-830X. Xue G and Loveridge CW (2004) HvDRF1 is involved in abscisic Authors thank Dr. K. Ammar (CIMMYT) for valuable acid-mediated gene regulation in barley and produces two suggestions during this investigation and Mrs. Marian forms of AP2 transcriptional activators, interacting preferably Boreham for the revision of English language. with a CT-rich element. The Plant Journal 37: 326–339. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 251–255 ISSN 1479-2621 doi:10.1017/S1479262111000475

Allele variation in loci for adaptive response in Bulgarian wheat cultivars and landraces and its effect on heading date

Stanislav Kolev1, Dimitar Vassilev1, Kostadin Kostov2 and Elena Todorovska1* 1AgroBioInstitute, 8 Dragan Tsankov, Str., 1164 Soﬁa, Bulgaria and 2Dobroudzha Agricultural Institute, 9200 G. Toshevo, Bulgaria

Abstract Allele composition at the major growth habit (Ppd-D1, Vrn-1, Rht-1 and Rht8) loci was determined in 52 Bulgarian bread wheat cultivars and landraces, using recently developed diagnostic molecular markers. The study showed that Bulgarian wheat germplasm varies for photoperiod, vernalization and height-reducing genes. The photoperiod-sensitive allele (Ppd-D1b) was the most frequent one in the old cultivars and landraces (90.9%), while the photoperiod-insensitive allele (Ppd-D1a) showed the highest frequency in the modern cultivars (96.71%). The alleles conferring winter growth habit (vrn-A1, vrn-B1 and vrn-D1) were more common in both the old (72.7%) and the modern (93.3%) wheat genotypes. The spring allele Vrn-A1c was not detected in Bulgarian germplasm, while the spring allele Vrn-B1 was found only in the old genotypes (13.6%). The semi-dwarfing allele Rht-B1b was observed in several modern cultivars. Seven allele variants were found in the microsatellite locus Xgwm261, closely located to the Rht8 gene. Among them, alleles of 164, 212 and 216 bp length were specific for the old genotypes studied, while alleles of 192 and 202 bp length were specific for the modern ones. The allele combination Rht-B1b//192 or 202 bp allele (Xgwm261 locus)//Ppd-D1a//vrn-A1/vrn–B1/vrn-D1 was detected in most of the early-heading modern cultivars. Our study emphasizes on the plasticity of the adaptive response of bread wheat cultivars sown in Bulgaria, as well as on the effect of variation for major growth habit on some yield and reproductive characteristics.

Keywords: adaptive response genes (Ppd-D1; Vrn-A1; Vrn-B1; Vrn-D; Rht8 and Rht-1); allele composition; molecular markers; wheat (T. aestivum L.)

Introduction deeper understanding of the effect of allele variation on the agronomically important traits as well as to more The recently developed allele-specific DNA markers for effective management of selection activities during the defining the allele variation in major loci controlling breeding process. Our previous works on the identifi- growth habit (vernalization, photoperiod response and cation, distribution and effect of the semi-dwarfing plant stature) (Korzun et al., 1998; Ellis et al., 2002; Yan Rht alleles on some agronomic traits in Bulgarian bread et al., 2004; Fu et al., 2005; Beales et al., 2007) have wheat showed that the allele 192 bp at the microsatellite allowed more efficient characterization of wheat germ- locus Xgwm 261, which is linked to Rht8 gene (2DS), plasm. These markers have also contributed to the prevailed in the cultivars developed during the last 50 years (Ganeva et al., 2005; Zheleva et al., 2006). This study was motivated by the lack of more complete information concerning the allele variation at the major * Corresponding author. E-mail: [email protected] loci for growth habit and their effect on maturity and 252 S. Kolev et al. other productive agronomic traits in Bulgarian bread day length, were detected using primers according to wheat varieties. The aim was to define and to assess the Beales et al. (2007). Following Beales et al. (2007), the allele variation and distribution at loci for growth habit photoperiod-insensitive allele was labelled as Ppd-D1a. genes (Vrn, Ppd and Rht) in Bulgarian germplasm collec- The alternative allele, inferring photoperiod sensitivity, tions from the Institute of Plant Genetic Resources was designated Ppd-D1b by analogy. (IPGR), Sadovo and Dobroudzha Agricultural Institute (DAI), G. Toshevo, Bulgaria. The obtained results could also prove useful for planning and implementation of mol- Vrn-A1 ecular markers in practical wheat-breeding programmes. The dominant spring alleles (Vrn-1Aa and Vrn-1Ab) were identified utilizing genome-specific primers Material and methods (VRN1AF and VRN1-INT1R) described by Yan et al. (2004), by variation in the promoter region of the Plant material Vrn-A1 locus. The primer combination Intr1/A/F2/Intr1/ A/R3 was used to differentiate the dominant spring A total of 52 old and modern Bulgarian wheat genotypes allele VrnA1c from vrn-A1 (the recessive winter allele). from the genebank collections of the IPGR, Sadovo and DAI, Bulgaria were used in this study. The modern cultivars were developed in the period 1960–2008. Vrn-D1

Microsatellite analysis Vrn-D1 intron-1 alleles were detected using Intr1/D/F, Intr1/D//R3 and Intr1/D/R4 primers (Fu et al., 2005). Microsatellite analysis of the Xgwm261 locus was per- The primer pair Intr1/D/F and Intr1/D//R3 was used for formed as described by Ro¨der et al. (1998), using Cy50- the ampliﬁcation of the dominant spring allele Vrn-D1, labelled forward and unlabelled reverse primers. Fragment while the primer pair Intr1/D/F and Intr1/D/R4 was analysis was carried out on AFL Express II sequencer used for the ampliﬁcation of the recessive winter allele (Amersham Biosciences). The size of the fragments was vrn-D1. determined with the program Allele locator, version 1.03.

Rht-1 Ppd-D1 The primer pairs BF/WR1 and BF/MR1 (Ellis et al., 2002) The alleles of the pseudo-response regulator Ppd-D1 were used for the ampliﬁcation of the alleles Rht-B1a (2D), which determine the sensitivity or insensitivity to (wild type) and Rht-B1b (dwarf), respectively.

Old cultivars and landraces 120 Modern cultivars 100

Allele frequency (%) 20

vrn-A1 vrn-B1

216 bp 212 bp 202 bp 192 bp 174 bp 164 bp vrn-D1 Vrn-A1 Vrn-B1 Vrn-D1

Rht-B1a Rht-B1b Ppd-D1a Ppd-D1b 164/174 bp Heterozygote Rht-B1 Xgwm261 Ppd-D1 Vrn-A1 Vrn-B1 Vrn-D1 Loci Fig. 1. Distribution of alleles for major growth habit genes among the old and modern Bulgarian wheat genotypes. leevraini oifraatv response adaptive for loci in variation Allele

Table 1. Rht-1, Xgwm261(Rht8), Ppd-D1 and Vrn-1 allele composition in the modern Bulgarian bread wheat cultivars (T. aestivum L.) and their heading datesa

Average Cultivars Ppd-D1 locus Vrn-A1 locus Vrn-B1 locus Vrn-D1 locus heading and breeding Rht1 locus Xgwm261 (Ppd-D1a; (Vrn-A1a; Vrn-A1b; (Vrn-B1; (Vrn-D1; date (days after lines (Rht-B1a, Rht-B1b) locus (bp) Ppd-D1b) Vrn-A1c; vrn-A1) vrn-B1) vrn-D1) 1st of May) Roussalka Rht-B1b 192 Ppd-D1a vrn-A1 vrn-B1 Vrn-D1 8.8 Prostor Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 10.4 Katya Rht-B1a/Rht-B1b 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 10.4 Galateya Rht-B1b 202 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 10.4 Kristal Rht-B1b 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 10.6 Pliska Rht-B1b 202 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 10.8 Enola Rht-B1b 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 11.4 Karat Rht-B1a 202 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 11.4 Sadovo 772 Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 12 Kaloyan1 Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 14.4 Zlatina2 Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 14.4 Zora Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 14.4 Momchil Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 14.4 Ideal Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 14.6 Progress Rht-B1a/Rht-B1b 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 14.6 Aglika Rht-B1b 202 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 14.6 Yantar Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 14.8 Slaveya Rht-B1b 202 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 14.8 Diamant Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 15 Kristi Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 15.2 Laska Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 Vrn-D1 15.2 Svilena Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 15.4 Karat Rht-B1a 174 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 15.6 Dobrud-zhanka Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 15.6 Sadovo 552 Rht-B1a 174/192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 15.6 Rekviem Rht-B1a/Rht-B1b 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 15.6 Elitsa Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 16.2 Bezostaya 1 Rht-B1a 192 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 20 Lozen 6 Rht-B1a 174 Ppd-D1a vrn-A1 vrn-B1 vrn-D1 20 Gladiator 113 Rht-B1b 174 Ppd-D1b Vrn-A1a vrn-B1 vrn-D1 25 a The data were recorded during 2005–2009. 253 254 S. Kolev et al. Results and discussion and other agronomic traits was studied in the modern Bulgarian bread wheat cultivars developed after 1960. Distribution of the alleles of the major growth habit In these cultivars, the dominating allele is 192 bp at genes (Ppd-D1, Vrn-1, Rht-1 and Rht8) within locus Xgwm261 (2D), which is considered to be a diag- Bulgarian bread wheat genotypes (Triticum nostic for the presence of Rht8 gene. aestivum L.) Here, it was confirmed that some of the commercial Bulgarian cultivars carried the Rht8 gene alone or in com- The assessment of Vrn-1(5A/5B/5D), Ppd-D1(2DS), bination with the Rht-1 gene. Similar results have been Rht-1(4B) and Rht8(2D) allele variation in 52 old and reported by Ganeva et al. (2005), using gibberellin test modern Bulgarian wheat cultivars and landraces and isogenic lines of cv. ‘Mercia’. The combination of showed that the recessive (vrn) alleles of Vrn-1 genes different Rht genes in Bulgarian cultivars could be con- responsible for winter growth habit prevailed in both sidered as successful, as these cultivars have found exten- the old and the modern Bulgarian genotypes (72.7 and sive use in the breeding practice. In the last few years, the 93.3%, respectively). lands sown with the cultivars ‘Enola’ (Rht-B1b þ Rht8) The dominant alleles Vrn-A1a and Vrn-A1b, which and ‘Aglika’ (Rht-B1a þ 202 bp allele at the Xgwm 261 determine the spring growth habit, were found only in locus) comprise 19 and 10%, respectively, of the total two cultivars, the modern one – ‘Gladiator 113’ and the wheat-sown areas in 2009, followed by ‘Sadovo1’ old one – ‘Karnobatska ranozreika’. A combination of (8.5%), ‘Pobeda’ (6.1%) and ‘Milena’ (3.7%) (Table 1). two dominant alleles Vrn-A1b and Vrn-B1 was identified The photoperiod insensitivity of the modern Bulgarian only in cv. ‘Karnobatska ranozreika’. Alleles Vrn-A1a and wheat cultivars is due to the presence of the Ppd-D1a Vrn-A1b separately or in combination with other Vrn-1 allele, which has been introduced from the main alleles have been observed in high proportion in spring donors of Rht8 (192 bp allele at the Xgwm261 locus) in wheat genotypes from Argentina, Canada, USA and Bulgarian wheat such as the Russian cultivars ‘Bezostaya China (Yan et al., 2004; Fu et al., 2005; Santra et al., 1’, ‘Skorospelka’ and ‘Kavkaz’. 2009). Similar to the results in some other studies All modern Bulgarian wheat cultivars excluding ‘Laska’ (Fu et al., 2005; Iqbal et al., 2007), the allele Vrn-A1c (vrn-A1/vrn-B1/Vrn-D1) and ‘Gladiator 113’ (Vrn-A1a/ was not found in Bulgarian wheat germplasm. On the vrn-B1/vrn-D1) were of the winter type (vrn-A1/vrn- other hand, the spring allele Vrn-B1 was found only in B1/vrn-D1). Ten allele combinations for growth habit the old Bulgarian cultivars and landraces, while Vrn-D1 and plant stature were detected in the old cultivars and allele was detected in a few old and in one modern landraces. In contrast, the major combination of alleles cultivar. The photoperiod sensitive allele (Ppd-D1b) pre- of the photoperiod sensitivity (Ppd-D1) and the Vrn- vailed in the old cultivars (90.9%). A significant increase 1(5A/5B/5D) genes in the modern Bulgarian wheat in the frequency of the photoperiod-insensitive allele gene pool Ppd-D1a//vrn-A1/vrn–B1/vrn-D1 amounted (Ppd-D1a) was observed in the modern Bulgarian culti- to 93.1%. The results showed that most of the early-heading vars (96.7%) (Fig. 1). The same tendency has also been cultivars possessed the allelic combination Rht-B1b// observed for the modern Chinese wheat (Yang et al., 192 or 202 bp allele at the Xgwm261 locus//Ppd-D1a// 2009). In our study, the allele Rht-1a (tall type) was vrn-A1/vrn–B1/vrn-D1 (Table 1). found in both old cultivars and landraces as well as in Additional studies on the allele variation in other genes the modern ones. The semi-dwarfing allele (Rht-1b) affecting the photoperiod sensitivity in wheat (Ppd-B1 was detected in several modern cultivars. This may be and Ppd-A1), the vernalization response (Vrn-B3 and related to the greater use of CIMMYT germplasm in Vrn-D4) and the earliness per se gene(s) would general- Bulgarian breeding in the previous years (Table 1). At ize to a great extent our knowledge of their effect on the Xgwm261 (2D) locus located close to the Rht8 maturity and other agronomic traits. gene, the alleles of 164, 212 and 216 bp length were Assessment of the effects of these genes on the head- found only in the old cultivars and landraces. ing date and grain yield in diverse Bulgarian production environments is also a task of current priority.

Allele combinations of the major growth habit genes (Ppd-D1, Vrn-1, Rht8 and Rht-1) in the modern Bulgarian bread wheat and their References influence on some agronomic traits Beales J, Turner A, Griffiths S, Snape JW and Laurie DA (2007) A pseudo response regulator is misexpressed in the photo- The contribution of the allelic composition at the Vrn, period insensitive Ppd-D1a mutant of wheat (Triticum Ppd and Rht loci to the variation of the heading date aestivum L.). Theoretical and Applied Genetics 115: 721–733. Allele variation in loci for adaptive response 255 Ellis MH, Spielmeyer W, Gale KR, Rebetzke GJ and Richards RA Ro¨der MS, Korzun V, Wendehake K, Plaschke J, Tixier MH, (2002) Perfect markers for the Rht-B1b and Rht-D1b dwarf- Leroy P and Ganal MW (1998) A microsatellite map of ing genes in wheat. Theoretical and Applied Genetics 105: wheat. Genetics 149: 2007–2023. 1038–1042. Santra DK, Santra M, Allian RE, Campbell KG and Kidwell KK Fu D, Szu¨cs P, Yan L, Helguera M, Skinner JS, von Zitzewitz J, (2009) Genetic and molecular characterization of vernaliza- Hayes PM and Dubcovsky J (2005) Large deletions within tion genes Vrn-A1, Vrn-B1, and Vrn-D1 in spring wheat the first intron in VRN-1 are associated with spring germplasm from the Pacific Northwest Region of the growth habit in barley and wheat. Molecular Genetics USA. Plant Breeding 128: 576–584. and Genomics 273: 54–65. Yan L, Helguera M, Kato K, Fukuyama S, Sherman J and Ganeva G, Korzun V, Landjeva S, Tsenov N and Atanassova M Dubcovsky J (2004) Allelic variation at the VRN-1 promoter (2005) Identification, distribution and effect on agronomic region in polyploid wheat. Theoretical and Applied traits of the semi-dwarfing Rht alleles in Bulgarian Genetics 109: 1677–1686. common wheat cultivars. Euphytica 145: 305–315. Yang FP, Zhang XK, Xia XC, Laurie DA, Yang WX and He ZH Iqbal M, Navabi A, Yang RC, Salmon DF and Spaner D (2007) (2009) Distribution of the photoperiod insensitive Ppd-D1a Molecular characterization of vernalization response allele in Chinese wheat cultivars. Euphytica 165: 445–452. genes in Canadian spring wheat. Genome 50: 511–516. Zheleva D, Todorovska E, Jacquemin J-M, Atanassov A, Christov Korzun V, Roder MS, Ganal MW, Worland AJ and Law CN (1998) N, Panayotov I and Tsenov N (2006) Allele distribution at Genetic analysis of the dwarfing gene (Rht8) in wheat. microsatellite locus Xgwm261 marking the dwarfing gene Part I. Molecular mapping of Rht8 on the short arm of Rht8 in hexaploid wheat from Bulgarian and Belgian gene chromosome 2D of bread wheat (Triticum aestivum L.). bank collections and its application breeding programs. Theoretical and Applied Genetics 96: 1104–1109. Biotechnology and Biotechnological Equipment 20: 45–56. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 256–259 ISSN 1479-2621 doi:10.1017/S1479262111000554

Diversity of seed storage proteins in common wheat (Triticum aestivum L.)

Zuzana Sˇramkova´1, Edita Gregova´2*, Svetlana Sˇlikova´2 and Ernest Sˇturdı´k3 1Institute of Biochemistry, Nutrition and Health Protection, Slovak University of Technology, Bratislava, Slovakia, 2Plant Production Research Center, Plant Production Research Institute, Piesˇtany, Slovakia and 3Department of Biotechnology, Faculty of Natural Sciences, University of Ss. Cyril and Methodius, Trnava, Slovakia

Abstract The objective of our study was to determine the composition of high-molecular weight- glutenin subunits (HMW-GS) in 120 cultivars of common wheat (Triticum aestivum L.). Fourteen alleles and 34 allelic compositions were detected using sodium dodecyl sulphate- polyacrylamide gel electrophoresis. The most frequent HMW-GS alleles at the Glu-A1, Glu-B1 and Glu-D1 loci were null (57.1%), 7 þ 9 (43.3%) and 5 þ 10 (61.9%), respectively. However, low-frequency HMW-GS alleles were also observed, such as 13 þ 16, 20, 21, 7 and 18, encoded by the Glu-B1 locus, and 4 þ 12, encoded by the Glu-D1 locus. The wheat–rye 1BL.1RS translocation was identiﬁed in 25 cultivars, using acid polyacrylamide gel electrophoresis. The Glu-score varied greatly, and some lines reached the maximum value of 10.

Keywords: high-molecular weight-glutenin subunits; quality; sodium dodecyl sulphate-polyacrylamide gel electrophoresis; Triticum aestivum L; wheat

Introduction long arms of homologous group-one chromosomes. These loci encode an x-type subunit of higher molecular Common wheat (Triticum aestivum L.) is the most widely weight and a y-type subunit of lower molecular weight. It grown wheat and it is one of the major food crops uti- was demonstrated that good bread-making quality is lized all over the world. Identification and characteriz- highly associated with the presence of specific HMW- ation of the genetic resources of wheat are highly GS (Payne et al., 1987). Since then, a number of novel desirable because the available information can be GS have been identified and characterized. These novel applied in further research and breeding. The improve- alleles may be a potential genetic resource for the cre- ment of wheat for use of the end-product has become ation of wheat lines with improved quality (Ga´lova´ an important part of wheat breeding programmes world- et al., 2002; Dotlacˇil et al., 2003; Ren et al., 2008). Data- wide. It has been documented that quality is determined bases that contain information on wheat genotypes/culti- principally by the molecular structure of the seed storage vars and their HMW-GS compositions may offer breeders proteins, glutenins and gliadins (Bushuk, 1998). Thus, the prospect of further advancement by selecting wheat not only protein quantity but also their quality is of cultivars with beneficial HMW-GS. The objective of our great importance. High-molecular weight-glutenin sub- study was to determine the composition of HMW-GS in units (HMW-GS) play a major role in this regard. The 120 wheat cultivars (T. aestivum L.). HMW-GS are encoded by genes at three Glu-1 loci (Glu-A1, Glu-B1 and Glu-D1), which are located on the Materials and methods

A total of 120 hexaploid wheat cultivars (T. aestivum L.), * Corresponding author. E-mail: [email protected] provided by the Gene Bank of the Slovak Republic in Diversity of seed storage proteins in wheat 257

Table 1. The frequencies of HMW-GS coded at the Glu-1 loci (Draper, 1987). The bread-making quality of the grain was expressed as a score (Glu-score and Rye-score) Locus HMW-GS Frequency (%) derived from the presence or absence of specific high- Glu-A1 Null 57.1 molecular weight-glutenins or -gliadins. 1 14.3 2* 28.6 Glu-B1 7 þ 9 43.3 7 þ 8 18.1 Results and discussion 6 þ 8 18.1 þ 17 18 7.1 Fourteen alleles and 34 allelic compositions were 7 9.4 20 1.6 detected in the set of 120 cultivars. The most frequent 13 þ 16 0.8 HMW-GS patterns were null, 7 þ 9 and 5 þ 10, which 18 0.8 were present in 21 cultivars. Ten of the wheat accessions 21 0.8 analyzed (8.3%) were observed to be heterogeneous in þ Glu-D1 5 10 61.9 their glutenin profiles. Five cultivars were heterogeneous 2 þ 12 37.3 4 þ 12 0.8 at one locus, two cultivars at two loci and another three varieties were heterogeneous at all three Glu-1 loci. Three alleles were detected at the Glu-A1 locus. The most abundant allele was Glu-A1c, encoding the null sub- Piesˇˇtany, were analyzed in this study. The wheat samples unit (57.1%, Table 1). It was found mostly in cultivars bred originated from 11 countries: Slovakia (SVK, 12), Czech in SVK, CZE, YUG and GBR (Table 2). In contrast, the Republic (CZE, 12), France (FRA, 12), Austria (AUT, 12), alleles Glu-A1b (subunit 2*) and Glu-A1a (subunit 1) Italy (ITA, 12), Hungary (HUN, 12), former Yugoslavia were present in only 14 and 29% of the genotypes, respect- (YUG, 12), Ukraine (UKR, 6), Russia (RUS, 6), Great Brit- ively. These alleles were observed mainly in Hungarian, ain (GBR, 12) and the USA (12). The HMW-GS were Italian, Austrian and American wheat samples (Table 2). extracted from randomly selected single seeds. Protein A high frequency of alleles encoding subunits 1 and 2* in extraction, electrophoretic separation of glutenins and wheat cultivars from the USA was observed also by Shan the detection procedures used were in accordance with et al. (2007). These subunits (1, 2*) have a positive the International Seed Testing Association (ISTA) stan- impact on the rheological properties of dough, so they dard procedure for use of sodium dodecyl sulphate- are highly desirable in cultivars used for bread-making. polyacrylamide gel electrophoresis (SDS-PAGE) (Wrigley, A total of nine HMW-GS were identified at the Glu-B1 1992). HMW-GS were identified by following the catalo- locus. Subunit 7 þ 9 was the predominant one (43.3%, gue of HMW-GS alleles (Payne and Lawrence, 1983). In Table 1); it was abundant mostly in wheat bred in HUN, addition, the presence or absence of the secalin block SVK, YUG, RUS and UKR. The subunit 6 þ 8, which is (1BL.1RS translocation) in the genotypes was examined associated with poor bread-making quality, was highly fre- using acid polyacrylamide gel electrophoresis (A-PAGE), quent in British and Austrian wheat samples (Table 2). according to the standard ISTA reference method Low-frequency HMW-GS, such as 13 þ 16, 20, 21, 7 and

Table 2. The frequencies (%) of HMW-GS at the Glu-1 loci according to the country of origin

Locus HMW-GS HUN YUG CZE AUT SVK FRA USA UKR RUS ITA GBR Glu-A1 Null 33 75 75 42 100 62 43 50 62 23 67 1 25 8 – 16 – 15 21 17 13 31 8 2* 42 17 25 42 – 23 36 33 25 46 25 Glu-B1 7 þ 9676933426718405757298 7 þ 8 25 8 42 8 25 18 13 29 – 22 8 6 þ 8 – 15 17 50 8 9 27 – – – 59 17 þ 18 – 8 8 – – 9 – – – 21 25 7 – – – – – 46 20 – 29 14 – 20 8 – – – – – – – – 7 – 13 þ 16 – – – – – – – – 14 – – 18 – – – – – – – – – 7 – 21 – – – – – – – 14 – – – Glu-D1 5 þ 10 83 42 92 67 67 50 63 100 83 43 25 2 þ 12 17 58 8 33 33 50 31 – 17 57 75 4 þ 12 – – – – – – 6 – – – – 258 Z. Sˇramkova´ et al. 18, which are encoded by alleles at the Glu-B1 locus were Acknowledgements also detected (Table 1). An unpaired subunit 7 was present mainly in French cultivars and also in wheat samples orig- This research was supported by the Slovak Research and inating from USA, RUS and ITA (Table 2). The rare subunit Development Agency under the contracts APVV LPP- 21 was found in the cultivar Mironovskaja ulucsennaja 025107 and VMSP-P-0055-09. (UKR). Subunit 18, which is expressed commonly coupled with subunit 17, was observed as an unpaired subunit in the Italian cultivar Glutinoso. Redaelli et al. (1997) observed expression of unpaired subunit 6 in some culti- References vars of T. aestivum L. They assumed that it was derived from the HMW-GS pair 6 þ 8 (because it had the same Bushuk W (1998) Wheat breeding for end-product use. Euphytica 100: 137–145. mobility as subunit 6 of this pair), but that the expression Dotlacˇil L, Bradova´ J, Hermuth J and Stehno Z (2003) Diversity of only subunit 6 was caused by a switch-off mutation in of HMW-Glu alleles in landraces and cultivars of winter the coding sequence or in the regulatory region of the wheat (Triticum aestivum), spelt (Triticum spelta) and gene for subunit 8. Yuan et al. (2009) characterized emmer (Triticum dicoccoides). In: Pogna NE, Romano M, two inactive y-type genes for Glu-B1 in T. aestivum Pogna EA and Galterio G (eds) Proceedings of the 10th International Wheat Genetics Symposium. 1–6 September ssp. yunnanese and ssp. tibetanum. They indicated that 2003, Paestum, Italy, pp. 475–477. the silenced 1By genes were 1By9 and that the silencing Draper SR (1987) ISTA variety committee report of the working was caused by deletions of nucleotide A, which resulted group for biochemical tests for cultivar identification in frameshift mutations responsible for the presence of 1983–1986. Seed Science and Technology 15: 431–434. premature stop codons. Ga´lova´ Z, Michalı´k I, Knoblochova´ H and Gregova´ E (2002) Variation in HMW glutenin subunits of different species Most of the cultivars (61.9%, Table 1) were found to of wheat. Rostlinna Vy´roba 48: 15–19. express the subunit pair 5 þ 10, which is encoded by Gobaa S, Brabant C, Kleijer G and Stamp P (2008) Effect of the allele Glu-D1d at the Glu-D1 locus. It has been shown 1BL.1RS translocation and of the Glu-B3 variation on fif- that varieties that contain the pair 5 þ 10 can form stron- teen quality tests in a doubled haploid population of ger dough than those that contain subunits 2 þ 12 wheat (Triticum aestivum L.). Journal of Cereal Science 48: 598–603. (Payne, 1987; Liang et al., 2010). Thus, wheat carrying Graybosch RA (2001) Uneasy unions: quality effects of rye the pair 2 þ 12 can be utilized in the production of chromatin transfers to wheat. Journal of Cereal Science cookies, crackers or pasta rather than in bread-making. 33: 3–16. In this study, subunit pair 2 þ 12, encoded by allele Li Y, Huang Ch, Sui X, Fan Q, Li G and Chu X (2009) Genetic Glu-D1a, was observed frequently in cultivars bred in variation of wheat glutenin subunits between landraces and varieties and their contributions to wheat quality ITA and GBR (Table 2). A high frequency of subunit improvement in China. Euphytica 169: 159–168. pair 2 þ 12 has been published for wheat originating in Liang D, Tang J, Pena RJ, Singh R, He X, Shen X, Yao D, Xia X Asian countries, such as China and Japan (Nakamura, and He Z (2010) Characterization of CIMMYT bread wheats 2001; Li et al., 2009), where wheat is used predominately for high- and low-molecular weight glutenin subunits and in the production of noodles. Our analysis revealed also other quality-related genes with SDS-PAGE, RP-HPLC and þ molecular markers. Euphytica 172: 235–250. the presence of subunit 4 12, encoded by allele c at the Nakamura H (2001) Genetic diversity of high-molecular-weight Glu-D1 locus, in one cultivar that originated in the USA. glutenin subunit compositions in landraces of hexaploid The wheat–rye 1BL.1RS translocation was identified in wheat from Japan. Euphytica 120: 227–234. 25 cultivars by use of A-PAGE. No cultivars that originated Payne PI (1987) Genetics of wheat storage proteins and the in USA and FRA were found to carry this translocation. effect of allelic variation on bread-making quality. Annual Review of Plant Physiology 38: 141–153. The presence of the 1BL.1RS translocation has a negative Payne PI and Lawrence GJ (1983) Catalogue of alleles for the effect on the rheology of dough; it produces a sticky complex gene loci, Glu-A1, Glu-B1, Glu-D1, which code dough with a lack of overmixing, and low loaf volumes for high molecular weight subunits of glutenin in hexa- (Graybosch, 2001; Gobaa et al., 2008; Zheng et al., 2009). ploid wheat. Cereal Research Communications 11: 29–35. The Glu-score and Rye-score varied greatly; some Payne PI, Nightigale MA, Krattinger AF and Holt LM (1987) The relationship between the HMW glutenin subunit samples reached the maximum value of 10 (cultivars composition and the breadmaking quality of British- GK Kapos, GK David, Vlasta, Baltimor, Luzanovka, grown wheat varieties. Journal of the Science of Food Maverick and Smuggler). These cultivars are supposed and Agriculture 40: 51–65. to have superior bread-making quality. The highest aver- Redaelli R, Ng KW and Pogna NE (1997) Allelic variation at the age value of Glu-score was achieved by Hungarian culti- storage protein loci of 55 US-grown white wheats. Plant Breeding 116: 429–436. vars (8.1). In contrast, the HMW-GS composition of Ren Y, Wang T, Xu Z-B, Yang Z-J and Ren Z-L (2008) Molecular British wheat samples resulted in an average value of characterization of a novel HMW-GS 1Dx50 associated with only 5.9. good bread making quality (Triticum aestivum L.) and the Diversity of seed storage proteins in wheat 259 study of its unique inheritance. Genetic Resources and Crop Yuan Z-W, Chen Q-J, Zhang L-Q, Yan Z-H, Zheng Y-L and Evolution 55: 585–592. Liu D-C (2009) Molecular characterization of two silenced Shan X, Clayshulte SR, Haley SD and Byrne PF (2007) Variation y-type genes for Glu-B1 in Triticum aestivum ssp. for glutenin and waxy alleles in the US hard winter wheat yunnanese and ssp. tibetanum. Journal of Integrative germplasm. Journal of Cereal Science 45: 199–208. Plant Biology 51: 93–99. Wrigley CW (1992) Identification of cereal varieties by gel Zheng S, Byrne PF, Bai G, Shan X, Reid SD, Haley SD and electrophoresis of the grain proteins. In: Linskens HF and Seabourn BW (2009) Association analysis reveals effects of Jackson JF (eds) Seed Analysis. Heidelberg: Springer- wheat glutenin alleles and rye translocations on dough- Verlag, pp. 17–41. mixing properties. Journal of Cereal Science 50: 283–290. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 260–263 ISSN 1479-2621 doi:10.1017/S1479262111000372

Seed longevity in oilseed rape (Brassica napus L.) – genetic variation and QTL mapping

Manuela Nagel1, Maria Rosenhauer1, Evelin Willner2, Rod J. Snowdon3, Wolfgang Friedt3 and Andreas Bo¨rner1* 1Leibniz Institute for Plant Genetics and Crop Plant Research (IPK), Corrensstraße 3, Gatersleben, Germany, 2Leibniz Institute for Plant Genetics and Crop Plant Research, Satellite Collections North, Inselstraße 9, Malchow/Poel, Germany and 3Department of Plant Breeding, Justus Liebig University, Heinrich-Buff-Ring 26–32, Giessen, Germany

Abstract Although oilseed rape has become one of the most important oil crops in Europe, little is known regarding the viability of its seed under conditions of long-term storage. We report here an examination of oilseed rape seed longevity performed on a set of 42 accessions housed at the German ex situ genebank at IPK, Gatersleben. A comparison of germination between the accessions stored for 26 years showed that viability was in part genetically determined, since it ranged between 42 and 98%. An attempt was made to deﬁne the genetic basis of viability by subjecting a mapping population of doubled haploids to three artiﬁcial ageing treatments. Quantitative trait loci (QTL) were detected on six chromosomes: N6, N7, N8, N15, N16 and N18. The chromosomal locations of these QTL were compared with their syntenic regions in Arabidopsis thaliana in order to explore what genes might underlie genetic variation for longevity.

Keywords: Brassica napus L; ex situ genebank; quantitative trait loci; seed longevity; synteny

Introduction the growing dominance of transgenic material in the soil seed bank, while at the same time, there is a continu- Originally developed as a source of oil for energy pur- ing loss of non-transgenic materials stored in genebanks, poses, oilseed rape (Brassica napus L.) has now risen as a result of ageing. Seeds from Brassica spp. typically to become one of the most important ﬁeld crops in lose ,50% of their viability with 7.3 years of storage at Europe, Canada, China and Australia. Its oil is now 208C and 50% relative humidity (RH) (Nagel and used in many applications, including biodiesel, culinary Bo¨rner, 2010), and within 23 years under standard low oil and livestock feed. Due to its remarkable increase in temperature (2188C) storage conditions (Walters et al., production, canola has become the focus of much breed- 2005a). Evidence that seed viability in some crop species ing and molecular genetics in recent years (Friedt and is in part genetically determined (Nagel et al., 2010) has Snowdon, 2010). prompted the present study of intra-speciﬁc variation in Until recently, the longevity of B. napus seed has been B. napus seed longevity. considered to be of little importance. However, as particularly the spring-type canola crop becomes transgenic-based, some concern has been expressed about Materials and methods

We tested the seed viability of 42 accessions of B. napus * Corresponding author. E-mail: [email protected] ssp. napus var. napus f. biennis, which had been Seed longevity in oilseed rape 261 B. napus L. above, with a final count after 7 d. AA2: this was identical 100 to AA1, except that the temperature was 448C. AA3: following Hay et al. (2008), in which two replicates of 100 80 seeds/line were placed inside a sealable box containing 60 a 8.7 M LiCl solution which ensured that the RH at 208C stabilized at 47% within 14 d. Artificial ageing was then 40 initiated by changing the solution (7.1 M LiCl), and main- Viability (%) taining the temperature at 458C (to give an RH of 60%) for 20 17 d. This was followed by the same germination test as above. A measure of relative viability for each method 0 1983 1990 1993 2009 was calculated by dividing the viability of the aged Year of testing seeds by their initial viability. The data were subjected Fig. 1. Mean viability of B. napus genebank accessions over to a quantitative trait loci (QTL) analysis, using QGENE different test years. The bold line indicates mean viability, software (Nelson, 1997) and a genetic map of, in total, and standard deviations over the years are shown. 161 markers comprising amplified fragment length polymorphism and simple-sequence repeat markers (Snowdon et al., unpublished data). A permutation test multiplied in 1983 and have been maintained in the was used to set the appropriate logarithm of odds interim at 7 ^ 38C and 6 ^ 2% seed moisture content. (LOD) ratio threshold for each treatment. Three replicates of 50 seeds/accession were placed on moistened filter paper and germination was monitored after 28 d. Their current viability was compared with Results and discussion historic data collected in 1983, 1990 and 1993 by using arithmetic means and standard deviations. The decay in viability of the genebank accessions over 26 A set of artificial ageing protocols was applied to a years of storage is shown in Fig. 1. The mean proportion population of 153 doubled haploid lines of the winter oil- of viable seeds fell from 94.7% in 1983 to 92.9, 79.1 and seed rape YE2-DH mapping population (Badani et al., 65.7% after 7, 10 and 26 years of storage, respectively. 2006) to explore the genetic basis of seed longevity. The associated standard deviations were 5.4% (1983) The protocols were: AA1, following Hampton and 4.6% (1990), 12.5% (1993) and 14.2% (2009). Despite TeKrony (1995), in which two replicates of 50 seeds having been grown simultaneously and subjected to the each were sealed in glass jars containing 200 ml deionised same post-harvest and storage conditions, the accessions water to raise the RH above 99%. After holding the jars nevertheless displayed variation with respect to seed via- for 48 h at 428C, a germination test was conducted as bility, as also occurs in barley, wheat, sorghum, rye and

(a) Brightened: yellow-seeded 1012-98 parent (b) QTL Treatment Marker LOD score R 2 value Darkened: black-seeded Express 617 parent 1 AA2E3350_328 1.53** 0.049 AA1 (44°C, 99%RH, 2 d) 2 AA2E3249_469 1.25* 0.040 QTL 6 AA2 (42°C, 99%RH, 2 d) E3862_73 3 AA2E3259_325 1.19* 0.038 AA3 (45°C, 60%RH, 17 d) 4 AA3E3261_205 1.90* 0.056 N08 mr000166 5 AA1mr000166 0.87* 0.028 QTL 5 5 AA3mr000166 1.24* 0.039 E3259_325 6 AA1E3862_73 0.88* 0.027 QTL 3 E3249_469 QTL 2 7 AA2Ra2A04_A 1.30* 0.041 E3350_328 E3261_205 QTL 1 QTL 4 8 AA1E3861_108 0.98* 0.030 N06 N07 9 AA2Na12D03 1.13* 0.037 10 AA2E3349_351 1.29* 0.042 BnTT1_RF1 10 AA2E3249_491 1.24* 0.040 11 AA1BnTT1_RF2 1.09* 0.062 QTL 12 12 AA1BnTT1_RF1 1.05* 0.061 BnTT1_RF2 E3861_108 QTL 11 13 AA3Na10C01_C 2.24** 0.067 N18a QTL 8 BnCD11SSR 13 AA3E3261_263 2.02* 0.060 E3349_351 Na10C01_B QTL 7 13 AA3Na10C01_B 1.85* 0.055 Na10C01_C QTL 14 13 AA3E3247_121 1.34* 0.040 Ra2A04_A QTL 10 E3249_491 E3261_263 QTL 13 E3247_121 13 AA3E3259_141 1.27* 0.038 QTL 9 Na12D03 E3259_141 14 AA3BnCD11SSR 2.68*** 0.081 , , N15 N16 N18b * ** *** Significant at P = 0.05, P = 0.01 and P = 0.001, respectiverly. Fig. 2. QTL interval mapping in the B. napus YE2-DH population detected a range of loci with effects on seed longevity after artificial ageing. The choice of artificial ageing protocol (AA1, AA2 and AA3) applied was found to be important. The 14 QTL map to seven different chromosomes (a). Individual QTL effects are tabulated (b). 262 M. Nagel et al. linseed (Nagel et al., 2009, 2010). We therefore concluded compromised in its synthesis, tend to show poor seed that there is a genotypic component involved in the longevity (Clerkx et al., 2004a). This demonstrates that determination of seed viability. ABA plays a role in maintaining seed viability during When the doubled haploid lines from the YE2-DH storage. Certain chemical and/or physical properties of population were tested using the three artificial ageing the seed-coat also affect germination rate after storage, methods, 13 significant QTL affecting seed longevity since the seed of both structural and pigmentation mutants were identified (Fig. 2). Most of the QTL were method- tends to deteriorate faster than that of their wild-type specific. AA1 produced five QTL, mapping to chromo- progenitor (Debeaujon et al., 2000; Clerkx et al., 2004a). somes N7, N8, N15 and N18a, while AA2 generated six QTL on N6, N15 and N16. By using AA3, four QTL on chromosomes N7 and N18b were found with in particular highest explained phenotypic variation in QTL 14: References R 2 ¼ 0.081. The only QTL common to AA1 and AA3 was QTL5, on chromosome N7. Badani AG, Snowdon RJ, Baetzel R, Lipsa FD, Wittkop B, The reliability of artificial ageing as a surrogate for Horn R, De Haro A, Font R, Lu¨hs W and Friedt W (2006) long-term storage has been repeatedly discussed in the Co-localisation of a partially dominant gene for yellow seed colour with a major QTL influencing acid detergent literature (Delouche and Baskin, 1973; Priestley and fibre (ADF) content in different crosses of oilseed rape Leopold, 1979; McDonald, 1999; Freitas et al., 2006), (Brassica napus). Genome 49: 1499–1509. leading to a number of mutually inconsistent con- Bentsink L, Alonso-Blanco C, Vreugdenhil D, Tesnier K, Groot clusions. Rajjou et al. (2008) suggested that a controlled SPC and Koornneef M (2000) Genetic analysis of seed- deterioration test protocol, as described by Tesnier et al. soluble oligosaccharides in relation to seed storability of Arabidopsis. Plant Physiology 124: 1595–1604. (2002), mimics many of the molecular and biochemical Clerkx EJM, Blankestijn-De Vries H, Ruys GJ, Groot SPC and events experienced during seed ageing. On the other Koornneef M (2004a) Genetic differences in seed longevity hand, in this study, we showed that even a modest of various Arabidopsis mutants. Physiologia Plantarum increase in ageing temperature (from 42 to 448C) can 121: 448–461. have a major effect on the expression of relevant genes. Clerkx EJM, El-Lithy ME, Vierling E, Ruys GJ, Blankestijin-De Vries H, Groot SPC, Vreugdenhil D and Koornneef M The QTL identified in AA1 and AA2 mapped to distinct (2004b) Analysis of natural allelic variation of Arabidopsis regions from those detected following AA3, which seed germination and seed longevity traits between the compared with AA1 and AA2 operates on lower accessions Landsberg erecta and Shakdara, using a new seed moisture contents in a dry state, according to recombinant inbred line population. Plant Physiology Walters et al. (2005b). 135: 432–443. Debeaujon I, Leon-Kloosterziel KM and Koornneef M (2000) The Brassica consensus map of Parkin et al. (2005) Influence of the testa on seed dormancy, germination, and was used to estimate potential locations of orthologous longevity in Arabidopsis. Plant Physiology 122: 403–413. loci in the Arabidopsis thaliana genome with respect to Delouche JC and Baskin CC (1973) Accelerated aging tech- the B. napus QTL. Potential links with B. napus QTL niques for predicting the relative storability of seed lots. were found for the upper part of the A. thaliana chromo- Seed Science and Technology 1: 427–452. Freitas RA, Dias DCFS, Oliveira MGA, Dias LAS and Jose IC somes AtC1 and AtC3, and near the lower ends of (2006) Physiological and biochemical changes in naturally AtC1 and AtC5. The genetic basis of seed longevity in and artificially aged cotton seeds. Seed Science and A. thaliana has been ascribed to genes mapping in the Technology 34: 253–264. upper parts of chromosomes AtC1 and AtC3 (Bentsink Friedt W and Snowdon RJ (2010) Oilseed rape. In: Vollmann J et al., 2000; Clerkx et al., 2004b). Other QTL related to and Rajan J (eds) Oil crops. Handbook of Plant Breeding, vol. 4. NY: Springer-Verlag, pp. 91–126. germination in the presence of salinity or heat stress, as Hampton JG and TeKrony DM (eds) (1995) Handbook of well as some controlling the rate of germination, are Vigour Test Methods.Zu¨rich: International Seed Testing also known in the region of AtC1, which suggests the Association. possibility that tolerance to stress represents an aspect Hay FR, Adams J, Manger K and Probert R (2008) The use of of seed longevity (Clerkx et al., 2004b). Nevertheless, non-saturated lithium chloride solutions for experimental control of seed water content. Seed Science and Technology many of the genes involved in the stress response are 36: 737–746. also participants in the oxidative stress response. An McDonald MB (1999) Seed deterioration: physiology, repair and example of this connection has been provided by assessment. Seed Science and Technology 27: 177–237. Thorlby et al. (1999), who detected the expression of Nagel M and Bo¨rner A (2010) The longevity of crop seeds stored genes in the region of AtC1 related to tolerance of oxi- under ambient conditions. Seed Science Research 20: 1–12. Nagel M, Vogel H, Landjeva S, Buck-Sorlin G, Lohwasser U, dative stress, as well as of some involved in abscisic Scholz U and Bo¨rner A (2009) Seed conservation in acid (ABA) biosynthesis and perception. A. thaliana ex-situ genebanks – genetic studies on longevity in mutants which have lost sensitivity to ABA, or are barley. Euphytica 170: 1–10. Seed longevity in oilseed rape 263 Nagel M, Abdur Rehman Arif M, Rosenhauer M and Bo¨rner A Arabidopsis: a comparison between artificial and natural (2010) Longevity of seeds – intraspecific differences in aging protocols. Plant Physiology 148: 620–641. the Gatersleben genebank collections. Tagungsband der Tesnier K, Strookman-Donkers HM, Van Pijlen JG, Van der 60. Tagung der Vereinigung der Pflanzenzu¨chter und Geest AHM, Bino RJ and Groot SPC (2002) A controlled ¨ Saatgutkaufleute Osterreichs. Raumberg-Gumpenstein deterioration test for Arabidopsis thaliana reveals genetic (Austria), pp. 179–181. variation in seed quality. Seed Science and Technology 30: Nelson JC (1997) QGene: software for marker-based genomic 149–165. analysis and breeding. Molecular Breeding 3: 239–245. Thorlby G, Veale E, Butcher K and Warren G (1999) Map Parkin IAP, Gulden SM, Sharpe AG, Lukens L, Trick M, Osborn positions of SFR genes in relation to other freezing-related TC and Lydiate DJ (2005) Segmental structure of the Brassica napus genome based on comparative analysis genes of Arabidopsis thaliana. Plant Journal 17: 445–452. with Arabidopsis thaliana. Genetics 171: 765–781. Walters C, Wheeler LM and Grotenhuis JM (2005a) Longevity of Priestley DA and Leopold AC (1979) Absence of lipid oxidation seeds stored in a genebank: species characteristics. Seed during accelerated aging of soybean seeds. Plant Science Research 15: 1–20. Physiology 63: 726–729. Walters C, Hill LM and Wheeler LJ (2005b) Dying while dry: kin- Rajjou L, Lovigny Y, Groot SPC, Belghaz M, Job C and Job D etics and mechanisms of deterioration in desiccated organ- (2008) Proteome-wide characterization of seed aging in isms. Integrative and Comparative Biology 45: 751–758. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 264–267 ISSN 1479-2621 doi:10.1017/S1479262111000505

Genetic variation at ﬂowering time loci in wild and cultivated barley

James Cockram*, Huw Hones and Donal M. O’Sullivan National Institute of Agricultural Botany, Huntington Road, Cambridge CB3 0LE, UK

Abstract The worldwide spread of barley cultivation required adaptation to agricultural environments far distant from those found in its centre of domestication. An important component of this adaptation is the timing of flowering, achieved predominantly in response to day length and temperature. Here, we use a collection of cultivars, landraces and wild barley accessions to investigate the origins and distribution of allelic diversity at four major flowering time loci, mutations at which have been under selection during the spread of barley cultivation into Europe. Our findings suggest that while mutant alleles at the PPD-H1 and PPD-H2 photoperiod loci occurred pre-domestication, the mutant vernalization non-responsive alleles utilized in landraces and cultivars at the VRN-H1 and VRN-H2 loci occurred post-domestication. The transition from wild to cultivated barley is associated with a doubling in the number of observed multi-locus flowering-time haplotypes, suggesting that the resulting phenotypic variation has aided adaptation to cultivation in the diverse ecogeographic locations encountered. Despite the importance of early-flowering alleles during the domestication of barley in Europe, we show that novel VRN alleles associated with early flowering in wild barley have been lost in domesticates, highlighting the potential of wild germplasm as a source of novel allelic variation for agronomic traits.

Keywords: barley; ﬂowering time; germplasm; photoperiod; vernalization

Introduction barley flowering time have been identified (Laurie et al., 1995). Wild-type alleles at the vernalization response loci Cultivated barley (Hordeum vulgare ssp. vulgare) was (VRN-H1 and VRN-H2) delay flowering in the absence of domesticated from its wild progenitor (H. vulgare vernalization, thereby maintaining vegetative growth in ssp. spontaneum) ,12,000 years ago within the Fertile wild barley and autumn-sown domesticates during the Crescent (Zohary and Hopf, 2000), a region which winter months, and preventing damage to sensitive spans Israel, Jordan and parts of Turkey. Barley is now floral organs (Cockram et al., 2007a). Selection for verna- one the world’s most important grain crops, and is lization-non-sensitive alleles at these loci resulted in cultivated in temperate environments throughout the spring-sown domesticates which lack a vernalization world. One of the major genetic adaptations to the response, allowing avoidance of the comparatively cold novel agricultural environments encountered during winters encountered in northern Europe. Of the two the post-domestication spread of cultivation was the major photoperiod response loci, PPD-H1 modulates modulation of flowering time, which is largely controlled flowering in response to long-day (LD) photoperiods, in response to day length (photoperiod) and low tem- while PPD-H2 controls flowering in response to short- perature (vernalization). Four major genes controlling days (SDs). Mutations at PPD-H1 resulted in photoperiod-non-sensitive alleles that remove the promotion of flowering in response to LDs (Turner et al., 2005), extending the growing period and allowing domesticates * Corresponding author. E-mail: [email protected] to take advantage of the long cool and wet summers of Variation at flowering time loci in barley 265 north Europe. Deletions within the PPD-H2 candidate Results gene HvFT3 are associated with a delay in flowering under SDs (Faure et al., 2007), which helps to maintain We sourced a collection of 315 barley germplasm acces- the vegetative stage in autumn-sown types until the sions with associated collection site geodata (longitude onset of spring, even after the vernalization requirement and latitude) from national genebanks (Fig. 1). Of has been met. Here, we use recently developed genetic these, 120 represent wild barley accessions, sampled markers diagnostic for allelic state at these major flower- from throughout their natural range. The remaining 195 ing time loci to explore diversity in wild and domesti- accessions are landraces (cultivated, locally adapted bar- cated barley. leys that pre-date systematic plant breeding), from across north-western Europe. Overlaying biome data illustrates that the western spread of barley domestication into the Methods Mediterranean basin was into environments broadly similar to those encountered in the centre of domestication Genomic DNA was extracted as described by Jones (Fig. 1). However, the spread into north and north- et al. (2008). The PCR-based marker assaying for allelic western Europe resulted in cultivation in regions with variation at VRN-H1 is described by Cockram et al. very different seasonal conditions and biomes. Genotypic (2009). VRN-H2 genotyping was performed using the analysis of VRN-H1 intron 1 InDel variation finds that, assay described by Karsai et al. (2005). The HvFT3 while 98% of wild barley accessions are predicted to candidate gene at the PPD-H2 locus was genotyped possess winter alleles, three accessions (all of which are as described by Faure et al. (2007). PPD-H1 genotypic located within Israel) carry spring alleles (Fig. 2). Of data was sourced from Jones et al. (2008). Sequen- these, the amplicons obtained from two accessions cing was performed as described by Cockram et al. suggest they possess that same allele, found at low (2008) and contig formation was conducted using the frequency within landrace accessions from the western VectorNTI package (Eppendorf). Geodata were plotted Mediterranean. Sequencing the VRN-H1 amplicon obtain- using ArcGIS v.10 (ESRI). For clarity, we use the verna- ed in the third accession finds that it possesses an intron 1 lization locus nomenclature described by Dubcovsky deletion not previously identified in cultivated barley et al. (1998). (Cockram et al., 2007b,c) (Supplementary Fig. S1,

Subspecies Biome description Mediterranean forests, woodlands and scrub Temperate conifer forests Spontaneum Boreal forests/Taiga Montane grasslands and shrublands Temperate grasslands, savannas and shrublands Vulgare Deserts and xeric shrublands Temperate broadleaf and mixed forests Fig. 1. Biome descriptions, overlaid with the geographic locations of germplasm collection sites for the landrace and wild barley accessions utilized in this study. 266 J. Cockram et al.

Multi-locus haplotype Spontaneum, WSG1 Vulgare, SSG0 Vulgare, SWG0 Vulgare, WSG0 Vulgare, WWG0 Spontaneum, SSG1 Spontaneum, WWG0 Vulgare, SSG1 Vulgare, SWG1 Vulgare, WSG1 Vulgare, WWG1 Spontaneum, SWG0 Spontaneum, WWG1 Vulgare, SST0 Vulgare, SWT0 Vulgare, WST0 Vulgare, WWT0 Spontaneum, SWG1 Spontaneum, WWT1 Vulgare, SST1 Vulgare, SWT1 Vulgare, WST1 Vulgare, WWT1 Fig. 2. Predicted allelic combinations at the major barley ﬂowering time loci in landraces and wild barley. Allelic status predicted by VRN-H1 and VRN-H2 genotypes is indicated by S (spring) and W (winter). Allelic status at the diagnostic PPD-H1 single-nucleotide polymorphism described by Turner et al. (2005): G, LD-photoperiod responsive (wild-type); T, LD-photoperiod non-responsive. PPD-H2:1,HvFT3 present (SD-photoperiod responsive, wild-type); 0, HvFT3 absent (SD-photoperiod non-responsive).

available online only at http://journals.cambridge.org). Conclusions PCR assays for the presence/absence of the three candidate genes at VRN-H2 (ZCCT-Ha, ZCCT-Hb and The modulation of flowering time during the post- ZCCT-Hc) shows that the deletion of all three genes domestication spread of barley cultivation allowed flexi- associated with spring vrn-H2 alleles in landrace and bility in the timing of sowing and harvesting, helping to modern European cultivars (Cockram et al., 2007b) is maximise the yield. Our findings support the assumption not found in wild barley. However, additional deletions that this flexibility was conferred both by the selection are observed: individual deletions of ZCCCT-Ha (ten for mutations at major flowering time genes and the accessions), ZCCT-Hb (one accession) and ZCCT-Hc utilization of multiple allelic combinations at these loci. (one accession), as well as a single example of a We find that the mutated photoperiod-non-responsive double deletion (ZCCT-Ha and ZCCT-Hb). Analysis of ppd-H1 and photoperiod-responsive ppd-H2 alleles origi- the PPD-H2 candidate gene HvFT3 shows that the nated pre-domestication. In contrast, the spring Vrn-H1 mutated allele that delays flowering under SDs occurs and vrn-H2 alleles found in modern European varieties at low frequency across the mid-to-western range of (Cockram et al., 2007b) are absent in wild barley. However, H. vulgare ssp. spontaneum. Overlaying genotypic data we identify rare instances of novel VRN-H1 and VRN-H2 for allelic status at PPD-H1 (sourced form Jones et al., deletions, predictive of spring alleles. Deletions within 2008) shows the presence of the mutant photoperiod VRN-H1 intron 1 result in the loss of vernalization require- non-responsive allele in wild barley from the east of its ment, and it is thought that the size and location of these natural range. Simultaneous analysis of allelic status at deletions may have a quantitative effect on flowering time all four flowering time loci finds an increase in haplotype (Szu˝cs et al., 2007). The identification of a novel deletion number from 7 in wild barley, to 16 in landraces. in ssp. spontaneum demonstrates that unexplored allelic Variation at flowering time loci in barley 267 variation exists within wild germplasm. The presence Cockram J, Chiapparino E, Taylor SA, Stamati K, Donini P, Laurie of identical 6 bp repeat motifs flanking this deletion DA and O’Sullivan DM (2007b) Haplotype analysis of vernalization loci in European barley germplasm reveals suggests that it was formed by illegitimate recombination novel VRN-H1 alleles and a predominant winter VRN-H1/ (IR) following double strand break (DSB) repair (Puchta, VRN-H2 multi-locus haplotype. Theoretical and Applied 2005), as previously observed in other VRN1 deletions in Genetics 115: 993–1001. barley and wheat (Cockram et al., 2007c). As DSBs that Cockram J, Mackay IJ and O’Sullivan DM (2007c) The role of occur anywhere within the intron have the potential to double-stranded break repair in the creation of phenotypic result in large deletions after IR, the creation of novel diversity at cereal VRN1 loci. Genetics 177: 1–5. Cockram J, White J, Leigh FJ, Lea V, Mackay IJ, Laurie DA, spring Vrn-H1 alleles is predicted to be more frequent Powell W and O’Sullivan DM (2008) Association mapping than the mutation of single DNA bases at a precise location of partitioning loci in barley. BMC Genetics 9: 16. within a gene. The predicted increased frequency of allele Cockram J, Norris C and O’Sullivan DM (2009) PCR-based mar- conversion supports our hypothesis that the spring Vrn-H1 kers diagnostic for spring and winter seasonal growth habit alleles observed in modern European cultivars arose in barley. Crop Science 49: 403–410. Dubcovsky J, Lijavetzky D, Appendino L and Tranquilli G (1998) post-domestication. Comparative RFLP mapping of Triticum monococcum As was the case with VRN-H1, our data suggest the genes controlling vernalization requirement. Theoretical triple ZCCT deletion associated with spring vrn-H2 alleles and Applied Genetics 97: 968–975. in all European germplasm surveyed to date (Cockram Faure S, Higgins J, Turner A and Laurie DA (2007) The FLOWE- et al., 2007b) occurred post-domestication. The predic- RING LOCUS-T-like family in barley (Hordeum vulgare). Genetics 176: 599–609. tion of single and double ZCCT deletions in wild barley Jones H, Leigh FJ, Mackay I, Bower MA, Smith LMJ, Charles MP, indicates this genomic region may be prone to structural Jones G, Jones MK, Brown TA and Powell W (2008) Popu- rearrangements. Genetic analysis of populations con- lation-based resequencing reveals that the flowering time structed from accessions differing for ZCCT copy adaptation of cultivated barley originated east of the Fertile number could help determine the relative function/ Crescent. Molecular Biology and Evolution 25: 2211–2219. phenotypic strength of each copy. Studies in human Karsai I, Szu˝cs P, Me´sza´ros K, Filichkina T, Hayes PM, Skinner JS, Lańg L and Bedo¨ Z (2005) The Vrn-H2 locus is a major and animal genetics are increasingly finding copy determinant of flowering time in a facultative £ winter number variation (CNV) to be an important component growth habit barley (Hordeum vulgare L.) mapping popu- of functional genetic variation. Our findings suggest lation. Theoretical and Applied Genetics 110: 1458–1466. that mining germplasm for CNV could help resolve the Laurie DA, Pratchett N, Bezant JH and Snape JW (1995) RFLP contributions of duplicated genes in such cases. We mapping of five major genes and eight quantitative trait loci controlling flowering time in a winter £ spring barley hope to build on the preliminary analysis of barley (Hordeum vulgare L.) cross. Genome 38: 575–585. germplasm presented here, with the aim of investigating Puchta H (2005) The repair of double-strand breaks in plants: the routes of cultivation into Europe, and the basis of mechanisms and consequences for genome evolution. genetic adaptation to local agricultural environments. Journal of Experimental Botany 56: 1–14. Szu˝cs P, Skinner J, karsai I, Cuesta-Marcos A, Haggard KG, Corey AE, Chenn THH and Hayes PM (2007) Validation of the VRN-H2/VRN-H1 epistatic model in barley reveals that intron length variation at VRN-H1 may account for a References continuum of vernalization sensitivity. Molecular Genetics and Genomics 277: 249–261. Cockram J, Jones H, Leigh FJ, O’Sullivan D, Powell W, Laurie DA Turner A, Beales J, Faure S, Dunford RP and Laurie DA (2005) and Greenland AJ (2007a) Control of flowering time in The pseudo-response regulator Ppd-H1 provides adap- temperate cereals: genes, domestication and sustainable tation to photoperiod in barley. Science 310: 1031–1034. productivity. Journal of Experimental Botany 58: Zohary D and Hopf M (2000) Domestication of Plants in the Old 1231–1244. World. Oxford: Oxford University Press. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 268–271 ISSN 1479-2621 doi:10.1017/S1479262111000438

Cold-modulated expression of genes encoding for key enzymes of the sugar metabolism in spring and autumn cvs. of Beta vulgaris L.

D. Paciﬁco, C. Onofri and G. Mandolino* CRA-CIN (Centro di ricerca per le Colture Industriali), via di Corticella, 133, 40128 Bologna, Italy

Abstract An integrated approach based on the use of bioinformatics and gene expression analysis tools was carried out to evaluate the organ-specific transcription modulation of nine genes relevant to sugar metabolism of Beta vulgaris L. plantlets of the autumn cv. Franca and spring cv. Bianca, in response to low-temperature (LT) treatments. Different growth cycles imply different plant capability to adapt to the environment that includes variations in gene expression of key metabolic enzymes. The transcriptional response was evaluated by quantitative PCR analysis before, during and after the LT treatments. The results were correlated with the LT-induced electrolyte leakage measure and the carbohydrate content. Stress-induced transcript level alterations were detected in the two cultivars, suggesting a modulation of sucrose synthesis and carbohydrate partitioning. Cold stress induced deep changes in the autumn cultivar, especially in fructose-1,6-biphosphatase gene expression, irrespective of temperature or exposure time. These differential features of expression profiles constitute first clues on the molecular basis of the differential LT response of sugarbeet autumn and spring cultivars.

Keywords: Beta vulgaris; cold stress; real-time PCR; sucrose metabolism; sugarbeet

Introduction Many different sugarbeet varieties are available for cultivation in different environments, and with specific Cold stress is a significant cause of crop losses in adaptation to different sowing times, as the two commer- European and North American agriculture, conditioning cial varieties employed in this study: cv. Bianca, selected crop quality and production. The early sowing of sugar- for spring sowing and characterized by high germin- beet (Beta vulgaris L.), to improve root production and ability and tolerance to cercospora, and cv. Franca, escape drought periods at maturity, and the autumn selected for autumn sowing and tolerant to bolting. sowing, a practice adopted in Southern Europe, expose The productivity in sugarbeet is a complex trait and young plantlets to freezing temperatures (below 08C) depends on sucrose content (SC) and yield (SY). The quan- and to the risk of severe crop and quality losses. Never- titative trait-corrected SY, comprehensive of both SC theless, the response to low temperature (LT) changes and SY, is the result of the activity of different enzymes: according to the genotype, suggesting the existence of sucrose synthases (SBSS1 and SBSS2), sucrose phosphate genetic variability among the different varieties. synthases (SPS1 and SPS2), fructose bisphosphate aldolase (FBPald), fructose bisphosphatase (FBPase) and choline monooxygenase (CMO) (Schneider et al., 2002). With the aim to understand the adaptation of sugarbeet * Corresponding author. E-mail: [email protected] crop to the environment, the response to cold stress of Cold-modulated expression of genes 269 C) two different cultivars of B. vulgaris, selected for early 8 2 2 and late sowing respectively, was investigated employing h) and molecular, physiological and metabolic approaches, focusing specifically on metabolic pathways involved and in the control of sucrose biosynthesis/degradation and distribution in the different tissues of the plant. C, 7 d Calibrator C, 7 d Calibrator C, 7 d Target C, 7 d Target C, 7 d Target C, 7 d Target 8 8 8 8 8 8 # 4 C/17 C/17 C/17 C/17 C/17 Materials and methods C/17 8 8 8 8 8 8

Seedlings of diploid sugarbeet (spring cv. Bianca and autumn cv. Franca) were hydroponically grown; the LT treatments, chosen accordingly to the electrolyte leakage (EL) test results, and the samplings are reported in C, 24 h 23 C, 24 h 23 C, 24 h 23 C, 24 h 23 C, 24 h 23 C, 24 h 23 8 8 8 8 8 Table 1. Each experiment was replicated twice. The 8 # 3 speciﬁc intron-spanning primers and probes (TaqMan; C/17 C/17 C/17 C/17 C/17 C/17 8 8 8 8 8 ‘Assay by Design’) were obtained interrogating the institute 8 for genomic resources B. vulgaris Gene Index database, by the selection of tentative consensus functionally Sampling times annotated with high signiﬁcance by BLAST X (similarity

above 90%; Supplementary Table S1, available online C/8 h for three nights). 8 Temperature and exposition times C, 5 h 23 only at http://journals.cambridge.org). C, 5 h 23 8 8 # C/5 h 23 C/5 h 23 2 8 8 C/5 h 23 Changes in the relative quantification of mRNA levels of C/5 h 23 8 8 2 2 C/17 C/17 2 2 8 nine target genes were analyzed by a two-step real-time 8 PCR analysis (ABI PRISM 7000; Applied Biosystems). For their relative quantification, the geometric mean of the most suitable reference genes (rRNA18S and tubulin, data not shown) was used, in order to compensate for C, 3 h 23 small expression fluctuations, according to Nicot et al. C, 3 h 23 8 8 # C/3 h C/3 h 1 8 8 C/3 h 0 (2005). The fold-change in gene expression was calculated C/3 h 0 8 8 2 2 C/17 C/17 2 2 8 by relative quantification method of comparative 8 Ct(22DDCt), according to Livak and Schmittgen (2001) and checked for statistical significance by P value and standard deviation of DDCt, according to Yuan et al. (2006). Glucose, fructose and SC in leaves and roots of control and treated plantlets, pot-grown in controlled conditions for 2 months, were determined by HPLC-evaporative light scattering detector (ELSD). C/8 h, three nights 23 C/8 h, three nights 0 C/8 h, three nights 8 8 Results and discussion 8

The response to cold stress tests showed that LT injury – measured as EL of excised segments of leaf – increased 1 C 24 d No 23 C24d 6 C 24 d No 0 C24d 6 C24d No in the range from 0 to 288C for both cvs., and that 228C C24d 6 8 8 8 8 8 was a temperature at which about 25–40% of tissue 8 C/17 C/17 C/17 C/17 C/17 damage could be observed. Besides, cv. Franca showed a C/17 8 8 8 8 8 8 lower electrolyte release in the solution in the range between 22 and 248C (Supplementary Fig. S1, available online only at http://journals.cambridge.org), making the two cvs. a good model to study LT response. Experimental schedule Expression level of nine target genes (Supplementary Table S1, available online only at http://journals. Plantlets of sugar beet (spring cv. Bianca and autumn cv. Franca) were hydroponically grown for 4 weeks in controlled conditions. The LT treatments (0 were applied forafter 5 h LT treatment and (24 during h and each 7 experiment d). The (named experiments from II, I IV and to VI VI), included the a samples short acclimation (leaves period and (6 roots) were collected four times (arrows): during (at 3 and 5 Table 1. 1 I23 Experiment Growth period Acclimation II 23 III 23 IV 23 V23 cambridge.org) in untreated tissues (leaves and roots) VI 23 270 D. Paciﬁco et al. (a) (b) BBFF 792 Leaves Roots Leaves Roots 900 SBSS1 SBSS1 111114 600 SBSS2 SBSS2 1 792 1 104 300 SPS1 1541 4 29 SPS2 1281 6 24 20 FBPase 71122

Fold-change 11 FBPald 1212 10 1 1 PDHkin 2151 0 Leaves Roots Cotyledons CMO 91127 DREB2A 1 500 1 200

Fig. 1. (a) Transcript levels of the two sucrose synthase isoforms (SBSS1 and SBSS2) in leaves, roots and cotyledons of control sugarbeet plants. Expression was normalized on the geometric mean of rRNA 18S and tubulin, using leaf as calibrator. (b) Fold-change of all the target genes analyzed in Bianca (B) and Franca (F) in leaves and roots of untreated plantlets; in each case, the transcript level of the organ in which the expression was lower has been used as calibrator. of both cvs. are reported in Fig. 1(c). For the two sucrose while in cv. Franca, it was found that the cold stress, irrespec- synthase genes, low transcriptional levels were observed tive of temperature or exposure time, induced deep and in the leaves of both cvs., while in roots, especially SBSS2 fixed changes, mainly in fructose 1,6 bisphosphatase (Sup- was strongly transcribed, as also reported by Hesse and plementary Fig. S4, available online only at http://journals. Willmitzer (1996); Fig. 1(a, b). cambridge.org) expression, suggesting a key role of the The induction of the two osmotic stress marker genes enzyme in sucrose metabolism regulation and outlining a (CMO and DREB2A) proved the transcriptional response cultivar-dependent response specificity. of sugarbeet organs to LT treatments used in the experi- The hardening response, triggered by gradual expo- mental system (Supplementary Figs. S2 and S3, available sition to low but not injuring temperatures, has never online only at http://journals.cambridge.org): CMO tran- been observed in sugarbeet. We registered however, scription showed a strong increase in roots of cv. a differential behaviour in the two cultivars examined Bianca, in the late phase, after the exposition to 228C, upon night exposure to 68C, followed by a further suggesting an accumulation of the osmoprotector glycine- exposure to 0 and 228C. betaine, in order to maintain the osmotic potential inside Here, we detected stress-induced alterations in gene the root cells, as observed also by Pestsova et al. (2008). expression, suggesting an induction at the transcriptional The time course of modulation of the genes coding for level of sucrose synthesis and a reduction of carbo- enzymes of carbohydrate metabolism upon exposure to hydrate partitioning during and after cold stress. These LT revealed a different behaviour in the two sugarbeet differential features of LT-response expression profiles cultivars in terms of response onset and gene induc- in cultivars selected for different climates and suscepti- tion/repression. The differences were detectable mainly bility to cold deserve further analysis, and were pre- in the early phase, after 3–5 h from stress application, viously unknown. The above-described changes are, to when the LT stress was applied as a shock (Supplemen- some extent, similar to those described in Arabidopsis tary Figs. S2 and S3, available online only at http:// thaliana leaves undergoing LT stress (Guy et al., 2008), journals.cambridge.org). Especially, in the leaves of suggesting a general conservation of the mechanism in Bianca, SBSS1 and SPS1, transcription responded dycotyledons. Actually, consistent expression variations promptly, though transiently, to the exposure to LT, in the majority of the analyzed genes have been reported suggesting their implication in a early response to cold mainly in cv. Franca, suggesting these gene expression stress; in Franca cv., SPS1 was mainly modulated after levels a possible factor contributing to its suitability for an acclimation at 68C followed by exposure to acute tem- autumn sowing. perature stress, differently from Bianca. The most import- The different expression profiles detected for both cvs. ant step of sucrose synthesis/degradation is regulated by reflected a different behaviour, evident also in the differ- fructose 1,6 biphosphatase controlling the resynthesis of ent tolerance of leaf and root to LT, suggesting a superior glucose from sucrose (Nielsen et al., 2004) and fructose performance of cv. Franca; moreover, this autumnal cul- biphosphate aldolase catalyzing theconversion of dihydrox- tivar showed, in all conditions analyzed, a decreased yaceton-phosphate and glyceraldehyde 3-phosphate in expression of FBPase, especially in roots, suggesting fructose 1,6-bisphosphate. Their expression profiles did that triose-phosphate pool was directed towards the not change significantly in roots and leaves of cv. Bianca, glycolytic pathway, i.e. energy production, rather than Cold-modulated expression of genes 271 towards glucose biosynthesis and consequential sucrose Livak KJ and Schmittgen TD (2001) Analysis of relative gene accumulation, to better cope with stress conditions. expression data using real-time quantitative PCR and the 22DDCt method. Methods 25: 402–408. HPLC-ELSD analysis showed that glucose and fructose Nicot N, Hausman J-F, Hoffmann L and Evers D (2005) House- contents of leaves and rootlets did not change after LT keeping gene selection for real-time PCR normalization treatments in cv. Bianca compared with the control in potato during biotic and abiotic stress. Journal of plants (Supplementary Fig. S5, available online only at Experimental Botany 56: 2907–2914. http://journals.cambridge.org), suggesting that early Nielsen TH, Rung JH and Villadsen D (2004) Fructose-2, cold exposure of plantlets does not affect sucrose pro- 6-bisphosphate: a traffic signal in plant metabolism. duction in sugarbeet. Trends in Plant Science 9: 556–563. Pestsova E, Meinhard J, Menze A, Fischer U, Windhovel A and Westhoff P (2008) Transcript profiles uncover temporal and stress-induced changes of metabolic pathways in germinat- References ing sugar beets seeds. BMC Plant Biology 8: 122–142. Schneider K, Scha¨fer-Preg R, Borchardt DC, and Salamini F Guy C, Kaplan F, Kopka J, Selbig J and Hincha DK (2008) (2002) Mapping QTLs for sucrose content, yield and quality Metabolomics of temperature stress. Physiologia Plantarum in a sugar beet population fingerprinted by EST-related 132: 220–235. markers. Theoretical and Applied Genetics 104: 1107–1113. Hesse H and Willmitzer L (1996) Expression analysis of a Yuan JS, Reed A, Chen F and Stewart CN Jr (2006) Statis- sucrose synthase gene from sugar beet (Beta vulgaris L.). tical analysis of real-time PCR data. BMC Bioinformatics Plant Molecular Biology 30: 863–872. 7: 85. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 272–275 ISSN 1479-2621 doi:10.1017/S1479262111000062

Study of symptoms and gene expression in four Pinus species after pinewood nematode infection

Albina R. Franco, Carla Santos, Mariana Roriz, Rui Rodrigues, Marta R. M. Lima and Marta W. Vasconcelos* Escola Superior de Biotecnologia, Universidade Cato´lica Portuguesa, Rua Dr. Anto´nio Bernardino de Almeida, 4200-072 Porto, Portugal

Abstract Pine wilt disease, caused by the pinewood nematode Bursaphelenchus xylophilus (Steiner and Buhrer) Nickle, is originating severe infections in pine trees. The disease is detected when external symptoms appear (e.g. needle chlorosis), but trees could remain asymptomatic for long periods and serve as a long-term host. The primary goal of this study was to assess the effect of inoculation with an avirulent isolate of B. xylophilus (C14-5) on different Pinus spp. seedlings (P. sylvestris, P. nigra, P. pinea and P. pinaster). At the same time, seedlings were also inoculated with a virulent strain, HF, in order to compare the phenotypic and genomic results of the two types of inoculations. The effect of inoculation was determined in terms of expression of various Pinus genes potentially involved in the response to the disease.The results suggest that P. pinea and P. nigra are more resistant to infection by the nematode than P. sylvestris and P. pinaster. The phenotypic and genetic differences were more marked among P. pinea and P. pinaster.

Keywords: Bursaphelenchus xylophilus; genetic expression; Pinus spp

Introduction involved genes are regulated in trees with differential disease resistance after attack by nematodes with varying Recently, pine wilt disease (PWD), caused by the pine- degrees of virulence (Kosaka et al., 2001; Kuroda et al., wood nematode (PWN) Bursaphelenchus xylophilus 2004). Thus, a targeted gene expression approach was (Steiner and Buhrer) Nickle, has become a major threat taken in order to investigate the infection mechanisms in to the European forests, with an estimated mortality risk commercially important pine species, namely P. sylvestris, of 50%. In Portugal, PWN was ﬁrst detected in 1999 which is the most threatened species in northern/central (Mota et al., 1999), and in 2008, the entire continental Europe, P. nigra and P. pinaster, that are being affected Portugal was demarcated as PWN-infested. Its insect in the central/southern areas and P. pinea that is thought vector is Monochamus galloprovincialis and, once to be resistant to the infection (OEPP/EPPO, 2001). infected, most plants cease resin production and show Symptoms of infection with a virulent strain of PWN (HF, the symptoms of needle chlorosis. Infection usually isolated from Setu´bal, Portugal) and with an avirulent becomes a fatal condition in just a few months strain (C14-5, described by Takehushi et al., 2006) were (Fukuda, 1997). However, it is not known how the also monitored and evaluated.

Materials and methods * Corresponding author. E-mail: [email protected] This project was supported by National Forest Authority, Agriculture Seeds of P. pinaster, P. pinea, P. sylvestris and P. nigra Ministry and Rural and Fisheries Development. were sterilized, germinated in 1% water agar (Agar no. 1, Gene expression study in Pinus species 273

Table 1. Development of symptoms in P. pinaster manufacturer’s instructions. Gene expression was (PP), P. pinea (PPi), P. nigra (Pni), P. sylvestris (Psy), determined using 100 ng of RNA, with the conditions and before (T0), and 10 (T10) and 20 d (T20) after inocu- program presented in Supplementary Table S1 (available lation with sterile water (H2O), avirulent strain (A) and virulent strain (HF) online only at http://journals.cambridge.org), using an MJ Mini Gradient Thermal Cycler (Bio-Rad Laboratories, Incubation PA, USA). The gene 18S was used as internal control. time (d) Scanning electron microscope was used to examine the morphology of P. pinea and P. pinaster stems. Treatment conditions T0 T10 T20 Thin, manual cuts were made with a scalpel. Each PP H2OIIIIsample was attached to a support with double-sided PP A I II III PP HF I III IV duct tape and placed in a desiccator until the samples PPi H2OIIIwere dehydrated. Samples were analysed following PPi A I II II the user manual of SEM JSM5600LV, operating at 20 kV. PPi HF I I II Pni H2OIII Pni A I II II Pni HF I II II Results and discussion Psy H2OIIII Psy A I II II Symptoms of disease were only detected 10 d after Psy HF I II III inoculation. P. pinea and P. nigra seem less susceptible I, Healthy plant; II, partial needle discolouration; III, to the infection since their symptoms did not develop partial needle discolouration, necrosis and reduction beyond stage II; P. pinaster appears to be the most in the resin production; IV, total discolouration, necro- susceptible species, as some seedlings died 20 d after sis and seedling death. the inoculation. Also, inoculation with the avirulent nematode resulted in seedlings with some degree of Lab M) and incubated for 2 weeks at 258C, with a needle discolouration. (Table 1) These differences of photoperiod of 8 h light–16 h dark. Once germinated, infection may be due to the blocking of the vascular seedlings were individually incubated vertically for system with resin produced by cells from radial 4 months under the same conditions described earlier parenchyma, which are damaged by the nematode, and supplied with 10 ml of nutritive solution Murashige resulting in needle chlorosis and plant death some and Skoog basal medium (Sigma). time after infection (Jones et al., 2008). B. xylophilus strains were grown on barley seeds with Anatomical differences among genotypes might be Botrytis cinerea at 268C, in the dark, and extracted using on the basis of differential resistance to PWN, therefore the Baermann funnel technique. A total of 20 seedlings of P. pinaster and P. pinea were examined with scanning each pine species were inoculated with 500 avirulent or electron microscope (SEM). Visual inspection of the virulent nematodes in a 100 ml sterile water suspension resulting photographs (Fig. 1) indicates that the round (Asai and Futai, 2002). shape of the stem is better maintained in P. pinea To evaluate the genetic expression, samples were than in P. pinaster after manual cross-sectioning. This taken at 0, 10 and 20 d after inoculation and stored at may be due to higher lignin content in P. pinea and 2 808C. Total RNA was extracted according to Le Provost may be related to increased resistance to PWD in this et al. (2007) and puriﬁed with Turbo DNA-free kit (Applied species. Lignin has a recognized role in plant defence, Biosystems, Foster city, CA, USA), according to the and constitutive lignin has already been related with

(a) RD (b) C E X PH E RD PH C X P P

20 kV×35500 µm19 31 15Pa 20 kV ×35 500 µm 19 31 16Pa

Fig. 1. SEM imaging of P. Pinea (a) and P. pinaster (b) stem cross-sections showing morphological differences between the two species. E, epidermis; C, cortex; RD, resin duct; PH, phloem; X, xylem; P, pith. 274 A. R. Franco et al. defensive mechanisms against nematodes in other http://journals.cambridge.org) were not specifically species (Fogain and Gowen, 1996). Experiments regar- designed for the species in question. Thus, it could ding lignin quantification are ongoing. Furthermore, explain the absence of genes that normally would be the diameter in cortical resin ducts seems larger in expressed. Also, plant mechanisms vary during the time P. pinaster stem. This may contribute to the suscepti- of a day, so the time when the sampling was made bility of this genotype to PWD, since PWNs progress may also have influenced the results. inside the plant through resin ducts (Fukuda, 1997). This is the first report of inoculations with virulent and Increased number and diameter of resin ducts have avirulent B. xylophilus strains in various pine species, and already been associated with PWD susceptibility though infection mechanisms of both PWN were not (Kawaguchi, 2006). clear, this study suggests that inoculation with virulent Biotic and abiotic factors stimulate the plant’s defence nematode can trigger a phased systemic response that response, diminishing the negative impacts of the patho- differs from the avirulent strain. However, it is necessary genic attack. The genes of interest tested in this work to identify other factors that may be responsible for the were found to be associated to osmotic stress, oxireductive plant defence when it is attacked by the different processes and cell death, among others, which are import- pathogens. ant in the defence response of P. densiflora (Japanese Red Pine) against the nematode (Shin et al., 2009). Pathogenesis-related proteins 4 expression was detected in all treatments (Supplementary Figs. S1–S4, available Acknowledgements online only at http://journals.cambridge.org). PR proteins are induced as response to pathogen attacks (Osmond The authors would like to thank the help of Dr. Manuel et al., 2001) and can be factors of hypersensitive response Mota (Universidade de E´vora, Portugal) for providing to nematode infection (Meins and Ahl, 1989; Shin et al., the HF nematode strain, and Dr. Hajime Kosaka 2009). ATTRX1, a protective gene against oxidative stress, (Kyushu Research Center Forestry and Forest Products was also expressed in all treatments and pine tree species Research Institute, Kumamoto, Japan) and Dr. Mitsutera (Supplementary Figs. S1–S4, available online only at Aikida (Forestry and Forest Products Research Institute, http://journals.cambridge.org), suggesting that different Tsukuba, Japan) for providing the C14-5 nematode strain. types of defences may be activated. There is a reported relationship between metallothionein being expressed in the presence of intensive oxidative stress (Mir et al., 2004). References Ethylene is an important component of conifer response against pathogens (Miller et al., 2005), hence it can induce Asai E and Futai K (2002) Promotion of the population growth of cell defence. In P. pinaster (Supplementary Fig. S1, avail- pinewood nematode in 4-month-old. Journal of Forestry able online only at http://journals.cambridge.org), the Research 7: 113–116. expression of MAT2/SAM2 (ethylene production) was OEPP/EPPO (2001) Bursaphelenchus xylophilus. PM 7/4 (1) only detected at the end of 20 d, which in the case of viru- OEPP/EPPO Bulletin 31: 61–69. Fogain R and Gowen SR (1996) Investigations on possible lent B. xylophilus treatment corresponded to plant death; mechanisms of resistance to nematodes in Musa. Euphytica in P. pinea (Supplementary Fig. S2, available online 92: 375–381. only at http://journals.cambridge.org), the inoculations Fukuda K (1997) Physiological process of the symptom devel- with both avirulent and virulent nematodes originated opment and resistance mechanism in pine wilt disease. the same type of response as water-inoculated Journal of Forest Research 2: 171–181. Jones J, Moens M, Mota M, Li H and Kikuchi I (2008) Bursaphe- plants (control). On the other hand, in seedlings of lenchus xylophilus: opportunities in comparative genomics P. nigra (Supplementary Fig. S3, available online only at and molecular host–parasite interactions. Molecular Plant http://journals.cambridge.org), MAT2/SAM2 and SHEPERD Pathology 9: 357–368. (water reduction) gene expression was only verified in Kawaguchi E (2006) Relationship between the anatomical nematode-inoculated plants. This can be explained by characteristics of cortical resin canals and migration of Bursaphelenchus xylophilus in stem cuttings of Pinus the increasing release of volatile compound production, thunbergii seedlings. Journal of the Japanese Forest Society which alters water transportation (Jones et al., 2008). 88: 240–244. Finally, P. sylvestris seedlings (Supplementary Fig. S4, avail- Kosaka H, Aikawa T, Ogura N, Tabata K and Kiyohara T (2001) able online only at http://journals.cambridge.org) demon- Pine wilt disease caused by the pine wood nematode: the strated that virulent strains of B. xylophilus did not cause induced resistance of pine trees by the avirulent isolates of nematode. European Journal of Plant Pathology 107: any MAT2/SAM2 gene expression. 667–675. It must be noted that some of the primers Kuroda K (2004) Inhibiting factors of symptom development in used (Supplementary Table S2, available online only at several Japanese red pine (Pinus densiflora) families Gene expression study in Pinus species 275 selected as resistant to one wilt. Journal of Forestry from cork tissue responds to oxidative stress. Journal of Research 9: 217–224. Experimental Botany 55: 2483–2493. Le Provost G, Herrera R, Paiva J, Chaumeil P, Salin FF and Mota MM, Braasch H, Bravo MA, Penas AC, Burgermeister W, Plomion C (2007) A micromethod for high throughput Metge K and Sousa E (1999) First report of Bursaphelenchus RNA extraction in forest trees. Biological Research 40: xylophilus in Portugal and in Europe. Nematology 1: 291–297. 727–734. Meins F and Ahl F (1989) Induction of chitinase and [beta]-1, Osmond R, Hrmova J, Fontaine F, Imberty A and Fincher G 3-glucanase in tobacoo plants infected with Pseudomonas (2001) Binding interactions between barley thaumatin-like tabaci and Phytophythora parasitica var. nicotianae. proteins and (1,3)-b-d-glucans. European Journal of Plant Physiology 61: 155–161. Biochemistry 268: 4190–4199. Miller B, Madilao L, Ralph S and Bohlmann J (2005) Insect- Shin H, Lee H, Woo K-S, Noh E-W, Koo Y-B and Lee K-J (2009) induced conifer defense, White pine weevil and methyl Identification of genes upregualted by pinewood nema- jasmonate induce traumatic resinosis, de novo formed tode inoculation in Japanese red pine. Tree Physiology 29: volatile emissions, and accumulation of terpenoid synthase 411–421. and putative octadecanoid pathway transcripts in Sitka Takehushi Y, Kanzaki N and Futai K (2006) How different is spruce. Plant Physiology 137: 369–382. induced resistance against the pine wood nematode, Mir G, Demenech G, Huguet G, Guo W, Goldsbrough P, Atrian Bursaphelenchus xylophilus, by two avirulent microbes? S, Molinas M (2004) A plant type 2 metallothionein (MT) Nematology 8: 435–442. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 276–280 ISSN 1479-2621 doi:10.1017/S1479262111000426

Development and application of EST-SSRs for diversity analysis in Ethiopian grass pea

M. Ponnaiah1*, E. Shiferaw1,M.E.Pe`1 and E. Porceddu2 1Scuola Superiore Sant’Anna, Piazza Martiri della Liberta`, 33-56127 Pisa, Italy and 2DABAC, University of Tuscia, Via S. C. De Lellis, 01100 Viterbo, Italy

Abstract To date, very limited molecular knowledge and molecular tools are available on grass pea (Lathyrus sativus L.). Herein, we present a genetic diversity analysis of 20 grass pea accessions collected from various regions of Ethiopia by means of seven newly developed Lathyrus expressed sequence tag-derived simple sequence repeat (EST-SSR) markers and four cross- transferable EST-SSRs derived from Medicago truncatula L. Forty-ﬁve alleles were detected among all the analyzed accessions. The number of alleles/locus ranged from two to seven with an average of four alleles/locus. The observed heterozygosity (Ho) ranged from 0.320 to 0.504, while the expected heterozygosity (He) ranged from 0.354 to 0.470. FST values estimated by analysis of molecular variance were 0.01, 0.15 and 0.84 for among regions, among accessions and within accessions, respectively, indicating that most of the variation (84%) resides within accession. The model-based cluster analysis grouped the accessions into three clusters grouping accessions, irrespective of their collection regions.

Keywords: expressed sequence tag-derived simple sequence repeat; genetic diversity; Lathyrus

Introduction Here, we present the ﬁrst assessment of genetic variability and genetic structure among Ethiopian grass Grass pea (Lathyrus sativus L.) is a legume crop widely pea accessions using EST-SSRs. cultivated in several arid and semi-arid countries, especially in Ethiopia, covering 9% of total pulse growing area (CSA, 2007). Material and methods DNA markers, particularly simple sequence repeats (SSRs) have proven to be very powerful tools in a variety Plant material of genetic studies in plants (Katti et al., 2001). The presence of SSR in gene coding regions and their availability A total of 240 plants, representing 20 grass pea accessions of extensive full and partial cDNA/expressed sequence from different regions of Ethiopia, were analyzed. The tag (EST) sequences for many plant species provide an GenElute Plant Genomic DNA Miniprep Kit (Sigma- effective way to develop SSR markers directly from EST Aldrich, St. Louis, MO, USA) was used to isolate genomic (EST-SSR) (Eujayl et al., 2004; Holton et al., 2002; Kantety DNA from 2-3-week-old leaves. et al., 2002). This strategy allows to overcome crucial limitations associated to the development of SSR markers, especially in crops for which molecular information is EST-SSR marker development scant, such as grass pea. Nineteen new EST-SSRs were designed from the 65 L. sativus ESTs deposited from public database (http://www.ncbi. nlM.nih.gov/dbEST), using Batchprimer3 software (http:// * Corresponding author. E-mail: [email protected] probes.pw.usda.gov/cgi-bin/batchprimer3/batchprimer3.cgi). EST-SSRs for diversity analysis in grass pea 277

In addition, 24 EST-SSR from M. truncatula, which proven seven of the 19 Lathyrus EST and four of the 24 Medicago to be transferable to other legume species (Gutierrez et al., ESTs to be polymorphic and those 11 markers were 2005) were selected to be used in grass pea. utilized for the variation analysis. From the 11 polymorphic EST-SSRs, a total of 45 alleles were detected in the 240 individual plants genotyped. PCR reaction and fragment analysis The number of alleles/locus ranged from two (locus 942) to seven (locus MtBA32F05) and averaged four. PCR was performed in a ﬁnal reaction volume of 15 ml Polymorphism information content (PIC) ranged from containing 30 ng genomic DNA, 5 £ PCR buffer, 0.2 mM 0.184 (Ls942) to 0.776 (MtBA32F05), with a mean value each of dNTPs, 0.5 unit GoTaqw polymerase (Promega), of 0.416 (Table 1). The most informative markers were 0.3 ml each of forward and reverse primers and 0.02 mM MtBA32F05 and MtBA10B02 with PIC value of 0.776 labelled M13 primer (6-FAM/VIC/PET/NED) (Schuelke, and 0.639, respectively. Rare alleles (frequencies ,0.05) 2000). Amplicons were analyzed using an ABI3130xl were observed in all markers except marker Ls932, genetic analyzer (Applied Biosystems). the highest being Ls074. Among the 45 alleles detected, rare alleles represent 35% of the alleles found in this analysis. The correlation coefﬁcient between gene diversity Data analysis (GD) and the number of alleles was high, r ¼ 0.825 (P , 0.05). Allele size was determined as base pairs using GeneMap- perw Software v3.7 (Applied Biosystems). The allelic data were subjected to diversity analysis within and among the accessions using PowerMarker 3.25 (Liu and Muse, 2005). Diversity among and within accessions MICRO-CHECKER 2.2.1 (Van Oosterhout et al., 2004) was used to check for potential genotyping errors, such as Allele frequencies were re-adjusted within populations allelic dropouts, stuttering or null alleles. Analysis of to account for null alleles and diversity analysis was molecular variance (AMOVA) and other population statistics performed using on the adjusted data. Average value were measured using GenAlEx 6.1 (Peakall and Smouse, of effective number of alleles/locus, percentage of poly-

2006) and population structure was examined using morphic loci, Ho and He, Shannon’s information index STRUCTURE 2.3.1 (Pritchard et al., 2000; Falush et al.,2003). (I) were 1.96, 95.5%, 0.404, 0.419 and 0.704, respectively; therefore, our data show that moderately high diversity exists among the accessions under the study. Results Accessions were grouped as seven populations based on their collection site to measure the diversity among Microsatellite validation and variability regions. Average of effective number of alleles/locus, percentage of polymorphic loci, I, Ho and He were Polymorphism screening was performed in ﬁve randomly 2.09, 97.4%, 0.760, 0.390 and 0.430, respectively. Regions chosen grass pea accessions. The screening revealed ‘Gojam, Welo and Gonder’ showed higher values in

Table 1. Characteristics of the EST-SSR markers used

No. of Size Allele size (bp) Major a Marker name Repeat motif alleles range (bp) (high frequency) AF (%) GD H PIC Ls989 (GT)8 4 157–165 159 45 0.628 0.297 0.551 Ls074 (TC)7 5 160–168 162 67 0.477 0.430 0.413 Ls617 (GTTG)3 3 164–172 168 54 0.500 0.375 0.379 Ls576 (ATG)5 3 173–179 176 86 0.237 0.150 0.212 Ls744 (TTC)4 4 132–147 144 80 0.325 0.391 0.279 Ls848 (CAT)4 4 158–173 170 80 0.321 0.378 0.274 Ls942 (CCAA)3 2 137–145 137 88 0.205 0.223 0.184 MtBA10B02 (TGG)7 5 338–368 344 39 0.695 0.435 0.639 MtBA52F10 (TC)17 5 117–125 119 55 0.556 0.300 0.470 MtBB52E08 (AG)7 3 79–91 81 58 0.504 0.630 0.399 MtBA32F05 (AG)5 7 162–176 172 29 0.804 0.646 0.776 Mean 4 0.477 0.387 0.416 AF, frequency of alleles; H, heterozygosity. a Calculated over a set of 240 individual plants. 278 M. Ponnaiah et al. diversity measures, whereas the ‘Arsi and Hararge’ region a relatively cheap way of developing sequence-based exhibited lower levels of diversity. markers (Gupta et al., 2003; Ellis and Burke, 2007). Thirty-seven per cent (7 out of the 19 newly designed Lathyrus EST-SSRs) of newly designed EST-SSR mar- Genetic structure kers were polymorphic between the accessions under the study. Seventeen per cent (4 out of 24 Medicago Ls989 locus showed null alleles in most of the accessions EST-SSRs) were polymorphic under the study. Diversity and it was excluded from further analysis. STRUCTURE analysis among the analyzed accessions showed the was run for K ¼ 1–10 based on the distribution of presence of a moderate level of diversity. Our analysis remaining 41 alleles at ten EST-SSR loci among the 240 also confirmed the transferability of this type of markers plants. STRUCTURE simulation produced the highest K also to related species, as demonstrated from the suc- value at K ¼ 3. cessful utilization of Medicago EST-SSRs. On the other STRUCTURE revealed also that cluster I is composed hand, we also demonstrated the successful transferability of individuals from Northern regions (Tigray, Gojam, of Lathyrus EST-SSRs to related species such as ground- Gonder and Welo). Cluster II comprised of individuals nut and green peas (data not shown). from all the growing regions, and cluster III consisted High levels of heterozygosity were observed in of individuals primarily from Shewa and Gojam, and a Gojam, Gonder and Welo regions. Accessions from few representatives from Welo and Gondar. None of Gonder also showed high number of different alleles. the clusters had individuals exclusively from one region Tadesse and Bekele (2003) reported the presence of a only (Fig. 1). significant variation among grass pea accessions from Ethiopia, based on morphological data. Their study showed a higher variability in accessions from Gondar AMOVA and Tigray regions. The lowest diversity estimates using EST-SSR markers were observed in Arsi and AMOVA showed that the within-accession diversity Hararge regions. This might be due to the limited explained most of the variation (84%). The mean Fpt sampling, since these two regions were represented value (analogous to FST), 0.15, indicated the presence by only one accession each, but it could also be due of moderate level of differentiation among the acces- to the actual low level of diversity present, since sions, and a low level of differentiation (1%) among grass pea is not common in these two regions. regions (Supplementary Table S1, available online only Uneven distribution of alleles was observed among at http://journals.cambridge.org). the analyzed samples, as revealed by the number of rare alleles (frequency #0.05), which accounted for 35% of the total number of alleles detected. Discussion Population genetic structure across the analyzed accessions identified three groups in which individuals Development of SSRs from EST databases has shown to are clustered independently of their collection region, be a feasible option for obtaining high-quality nuclear and it also showed admixture among accessions. This markers. For under-studied crops, this method is relatively low genetic differentiation among regions

1.00

0.80

0.60

0.40

0.20

0.00 Tigray, Gojam, Gonder, Welo All growing region Shewa and Gojam

Fig. 1. Estimated population structure of grass pea landraces from Ethiopia. Summary plot of estimates by Q (estimated membership coefficients for each individual in the three clusters) as inferred from Structure. EST-SSRs for diversity analysis in grass pea 279 could be interpreted by gene flow due to movement of Ethiopia, where this crop is often the only choice seeds – seed exchange among farmers being a mechan- available to harness in adverse cultural conditions. ism used to enhance diversity of local germplasm and Modern breeding strategies for this important but neg- avoid crop failure. This results in an increase in distri- lected crop might take advantage of the information bution of alleles among different population, irrespec- provided herewith. tive of geographical distance (Louette et al., 1997). Grass pea reproductive biology might also have contributed to increase within-population variation. In fact, although the floral biology of grass pea favours self- References pollination (Campbell, 1997; Yadav and Bejiga, 2006), there are records of substantial outcrossing, which is CSA (2007) Agricultural sample survey 2006/2007. Report on area and production of crops volume I. Ethiopia: Central dependent on environmental and/or genetic factors Statistics Authority. (Chowdhury and Slinkard, 1997; Gutie´rrez-Marcos et al., Campbell CG (1997) Grass pea. Lathyrus sativus L. Promoting 2006). Our result showed the existence of a significant the conservation and use of underutilized and neglected level of variation among accessions, though most of crops. 18. Rome: Institute of Plant Genetics and Crop the variation was present within populations. The Plant Research, Gatersleben/International Plant Genetic Resources Institute. current Ethiopian ex-situ collection of grass pea predo- Chowdhury MA and Slinkard AE (1997) Natural outcrossing minantly includes samples from the Shewa (45%) in grasspea. J Hered 88: 154–156. region. Based on present results, it might be useful to Chowdhury MA and Slinkard AE (2000) Genetic diversity in increase representative samples from other regions. grass pea (Lathyrus sativus L.). Genetic Resources and Due to the establishment of second-generation high- Crop Evolution 47: 163–169. Ellis JR and Burke JM (2007) EST-SSRs as a resource for popu- throughput sequencing technologies, it is expected lation genetic analyses. Heredity 99: 125–132. that, in the near future, the number of EST sequences Eujayl I, Sledge MK, Wang L, May GD, Chekhovskiy K, from Lathyrus deposited at the public database will sig- Zwonitzer JC and Mian MAR (2004) Medicago truncatula nificantly increase. This could facilitate the development EST-SSRs reveal cross-species genetic markers for and application of a large set of molecular markers, Medicago spp. Theoretical and Applied Genetics 108: 414–422. such as EST-SSRs, which proved to be very informative Falush D, Stephens M and Pritchard JK (2003) Inference of and easy to develop. Due to modern genomic revolu- population structure using multilocus genotype data: tion, it is expected that the number of EST sequences linked loci and correlated allele frequencies. Genetics getting deposited at the public database will increase 164: 1567–1587. exponentially and hence an increase in utilization of Gupta PK, Rustgi S, Sharma S, Singh R, Kumar N and Balyan HS (2003) Transferable EST-SSR markers for the study of markers like EST-SSRs is highly recommended, as this polymorphism and genetic diversity in bread wheat. study showed that the new EST-SSRs developed for Molecular Genetics and Genomics 270: 315–323. the first time for grass pea are useful tools for the Gutie´rrez-Marcos JF, Vaquero F, Saénz de Miera LE and Vences FJ genetic diversity analysis in the species. (2006) High genetic diversity in a world-wide collection In our study, Gojam, Gonder and Welo regions had of Lathyrus sativus L. revealed by isozymatic analysis. Plant Genetic Resources 4: 159–171. higher level of diversity compared with the others. This Gutierrez MV, Vaz Patto MC, Huguet T, Cubero JI, Moreno MT is consistent with the results of Tadesse and Bekele and Torres AM (2005) Cross-species amplification of (2003), whose studies of grass pea accessions from Ethio- Medicago truncatula microsatellites across three major pia were based on morphological data. Most of the vari- pulse crops. Theoretical and Applied Genetics 110: ations were due to the differences between individuals 1210–1217. Holton TA, Christopher JT, McClure L, Harker N and Henry RJ within accession (84%), with moderate high level of (2002) Identification and mapping of polymorphic SSR population differentiation (mean FST ¼ 0.15, P , 0.001). markers from expressed gene sequences of barley and This may be due to farmer’s habit to mix seeds from wheat. Molecular Breeding 9: 63–71. different sources before sowing as a mean to avoid Kantety R, La Rota VM, Matthews DE and Sorrells ME crop failure. The results are similar to those reported by (2002) Data mining for simple sequence repeats in expressed sequence tags from barley, maize, rice, sor- Chowdhury and Slinkard (2000), who mentioned that ghum and wheat. Plant and Molecular Biology 48: 90.7% of the variability was due to within-region diver- 501–510. sity. Our results showed the existence of moderate Katti MV, Ranjekar PK and Gupta VS (2001) Differential genetic variability in grass pea populations of Ethiopia distribution of simple sequence repeats in eukaryotic mostly within accession. genome sequences. Molecular Biology and Evolution 18: 1161–1167. Although still based on a limited number of molecu- Liu K and Muse SV (2005) PowerMarker: an integrated analysis lar markers, our study represents the most accurate environment for genetic marker analysis. Bioinformatics description to date of L. sativus germplasm from 21: 2128–2129. 280 M. Ponnaiah et al. Louette D, Charrier A and Berthaud J (1997) In situ conservation Tadesse W and Bekele E (2003) Phenotypic diversity of of maize in Mexico: genetic diversity and maize seed Ethiopian grass pea (Lathyrus sativus L.) in relation to management in a traditional community. Economic Botany geographical regions and altitudinal range. Genetic 51: 20–38. Resources and Crop Evolution 50: 497–505. Peakall R and Smouse PE (2006) GENALEX6: genetic analysis Van Oosterhout C, Hutchinson WF, Wills DPM and Shipley P in Excel. Population genetic software for teaching and (2004) MICRO-CHECKER: software for identifying and research. Molecular Ecology Notes 6: 288–295. correcting genotyping errors in microsatellite data. Pritchard JK, Stephens M and Donnelly P (2000) Inference of population structure using multilocus genotype data. Molecular Ecology Notes 4: 535–538. Genetics 155: 945–959. Yadav SS and Bejiga G (2006) In: Brink M and Belay G (eds) Schuelke M (2000) An economic method for the fluorescent Lathyrus sativus L. Record from Protabase. Wageningen: labeling of PCR fragments. Nature Biotechnology 18: PROTA (Plant Resources of Tropical Africa), http:// 233–234. database.prota.org/search.htm q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 281–283 ISSN 1479-2621 doi:10.1017/S1479262111000359

A novel genetic framework for studying response to artiﬁcial selection

Randall J. Wisser1*, Peter J. Balint-Kurti2,3 and James B. Holland3,4 1Department of Plant and Soil Sciences, University of Delaware, Newark, DE, USA, 2Department of Plant Pathology, North Carolina State University, Raleigh, NC, USA, 3United States Department of Agriculture-Agricultural Research Service, Plant Science Research Unit, Raleigh, NC, USA and 4Department of Crop Science, North Carolina State University, Raleigh, NC, USA

Abstract Response to selection is fundamental to plant breeding. To gain insight into the genetic basis of response to selection, we propose a new experimental genetic framework allowing for the identification of trait-specific genomic loci underlying population improvement and the characterization of allelic frequency responses at those loci. This is achieved by employing a sampling scheme for recurrently selected populations that allows for the simultaneous application of genetic association mapping and analysis of allelic frequency change across generations of selection. The combined method unites advantages of the two approaches, permitting the estimation of trait-specific allelic effects by association mapping and the detection of rare favourable alleles by their significant enrichment over generations of selection. Our aim is to develop a framework applicable for many crop species in order to gain a broader and deeper understanding of the genetic architecture of response to artificial selection.

Keywords: adaptation; association mapping; plant breeding, quantitative trait; selection

Introduction Genetic association mapping

Insights into the genetic basis of response to artificial Genetic association mapping is now in widespread use selection lag, despite substantial advances in our under- in plants (for review, see Zhu et al., 2008) and has led standing of sequence diversity. Relative to the abundant to the identification of quantitative trait loci and genes literature on the genetic mapping of loci conditioning underlying standing variation (e.g. Brown et al., 2008; trait variation, few studies have investigated the geno- Harjes et al., 2008). It has also been applied to study mic response or genetic architecture of generational the genetic architecture associated with long-term artifi- improvement in crop species in which artificial selection cial selection for kernel oil composition in maize is paramount. We are maintaining a collection of cita- (Laurie et al., 2004). tions for studies that have investigated the genetic Association mapping works by identifying significant architecture of response to artificial selection, accessible correlations between allelic and trait variation; however, at http://www2.udel.edu/wisserlab/RAS-citations or upon methods are required to control for non-independence request. Herein, a novel genetic framework is proposed or relatedness among samples comprising many associ- for examining the genetic architecture of response to ation mapping panels (e.g. Yu et al., 2006). The primary artificial selection. advantages of association mapping are that pre-existing lines from within or across breeding programmes may be used without the need to develop new genetic stocks, multiple alleles can be examined at once and, in some cases, very high-resolution mapping can be achieved * Corresponding author. E-mail: [email protected] because of rapid decay in linkage disequilibrium (such 282 R. J. Wisser et al. as in maize [Remington et al., 2001]). The trade-off for Combining genetic association mapping with high-resolution mapping panels is that, in the absence of analysis of allelic frequency change good candidate genes, extremely large numbers of markers across the genome or whole genome sequences are In an effort to understand response to artificial selection required to identify associations. Other disadvantages to and the adaptation of exotic germplasm populations to association mapping are that correction for relatedness, new environments, we have designed and are implement- required to reduce false positive rates, can lead to a signifi- ing a framework (Fig. 1) that is expected to counterbalance cant loss in power, that rare but potentially important the opposing limitations and exploit the complementary alleles cannot be detected and that, for association advantages of association mapping and analysis of allelic panels including very diverse (some unadapted) germ- frequency change. Recurrently selected populations rep- plasm, phenological variation can confound variation in resent germplasm that has been selected and adapted to the trait(s) of interest. an environment of interest. Such populations thus offer the potential of identifying relevant alleles on which to exert selection in a breeding programme. In recurrently selected populations, initially, rare but favourable alleles Analysis of allelic frequency change increase in frequency from generation to generation while common but unfavourable alleles reduce in fre- The most common method used to study the genetic quency. Thus, across generations, allelic frequencies at architecture of response to artificial selection has been loci underlying selection response would be relatively through the analysis of shifts in allelic frequencies. Fre- balanced compared with allelic frequencies in any single quencies are estimated from samples taken from different generation. Although perfect balance is unlikely due to generations of selection, and for each locus assayed, the transient nature of populations undergoing improve- a statistic is used to test for significant departures from ment, this multigenerational sampling strategy will gener- genetic drift expectations or directional change. Signi- ally overcome a major limitation of association mapping, ficant departures are taken as evidence of selection; it is which is the detection of low-frequency variation. In commonly assumed that the influence of migration and turn, the application of association mapping to genetically mutation is minimal. characterize artificial selection response overcomes the Initial molecular genetic studies seeking insight into major limitation of analysis of allelic frequency change, artificial selection response were conducted in maize, i.e. the inability to resolve trait-specific locus associations. using handfuls of allozyme loci (Brown, 1971; Brown The ultimate outcome of combining these two approaches and Allard, 1971; Stuber and Moll, 1972). These studies set the stage for the use of more efficient marker techno- Phenotype logies and improved statistical methodologies that have allowed for genome-wide scale analyses and the identi- Self-pollinate/topcros fication of loci presumably underlying response to selection (e.g. Labate et al., 1999; De Koeyer et al., 2001; G G G Coque and Gallais, 2006; Wisser et al., 2008). 0 i n There are advantages and limitations to studying Analysis recurrently selected populations. A key advantage is Collect tissue samples that the direct products of a breeding population can be studied in which selection has sorted out the most Genotype favourable and unfavourable alleles. Relatively rare but Fig. 1. A genetic framework for studying the genetic archi- important alleles are enriched by selection and detect- tecture of response to artificial selection. The figure depicts able by the analysis of allelic frequency shifts. The pri- a sampling scheme in which random samples of individuals mary limitation of this approach is that allelic effects are taken from multiple generations of a recurrently on specific traits cannot often be estimated since selec- selected population (G0 representing the base population, Gi representing any intermediate generation and Gn repre- tion is not typically conducted or achieved for a single senting the last generation). Tissue is collected from the trait. Development of the base population used to initiate sample of individuals that are also self-pollinated or selection and the intensity of artificial selection can topcrossed. Genotyping is conducted on the sample of indi- impact the structure of linkage disequilibrium, which, viduals and phenotyping is conducted on the sample of in turn, can affect the mapping resolution for loci associ- families or topcrosses. In the analysis phase, association mapping is conducted using genotype data and family or ated with selection. The statistical power of detecting topcross mean performance data. With the same genotype loci through the analysis of allelic frequency shifts also data, allelic frequency shifts can be examined at the quanti- needs to be addressed. tative trait loci identified by association mapping. Studying response to artificial selection 283 is a powerful approach to gain information about the map References locations of genomic loci associated with trait-specific ‘response’ variation and insights into the population Breseghello F and Sorrells ME (2006) Association analysis as a genetic basis for phenotypic change. strategy for improvement of quantitative traits in plants. Crop Science 46: 1323–1330. The application of association genetic mapping to Brown AHD (1971) Isozyme variation under selection in Zea any type of population requires consideration of genetic mays. Nature 232: 570–571. disequilibrium. For most recurrently selected populations, Brown AHD and Allard RW (1971) Effect of reciprocal recurrent random mating is conducted among selection units. selection for yield on isozyme polymorphisms in maize Therefore, gametic phase disequilibrium present in the (Zea mays L.). Crop Science 11: 888–893. initial generation or arising due to selection is expected Brown PJ, Rooney WL, Franks C and Kresovich S (2008) Efficient mapping of plant height quantitative trait loci to decay in each generation. This allows unlinked loci to in a sorghum association population with introgressed be independently associated with specific traits, despite dwarfing genes. Genetics 180: 629–637. the fact that selection was exerted; however, family-level Coque M and Gallais A (2006) Genomic regions involved in response structure may need to be accounted for, if the sample of to grain yield selection at high and low nitrogen fertilization selection units is small or the chance of sampling siblings in maize. Theoretical and Applied Genetics 112: 1205–1220. De Koeyer DL, Phillips RL and Stuthman DD (2001) Allelic shifts is high. The level of resolution of association mapping and quantitative trait loci in a recurrent selection popu- and allelic frequency tests depends on the extent of lation of oat. Crop Science 41: 1228–1234. disequilibrium at linked loci. Recurrently selected popu- Harjes CE, Rocheford TR, Bai L, Brutnell TP, Kandianis CB, lations are expected to have lower levels of linkage Sowinski SG, Stapleton AE, Vallabhaneni R, Williams M, disequilibrium than biparental mapping populations Wurtzel ET, Yan JB and Buckler ES (2008) Natural genetic variation in lycopene epsilon cyclase tapped for maize and may be as low as diverse germplasm collections, if biofortification. Science 319: 330–333. sufficient generations of random mating were involved in Labate JA, Lamkey KR, Lee M and Woodman WL (1999) Tem- population development (Breseghello and Sorrells, 2006). poral changes in allele frequencies in two reciprocally In the proposed framework, genotyping is conducted selected maize populations. Theoretical and Applied on random samples of individuals taken from different Genetics 99: 1166–1178. Laurie CC, Chasalow SD, LeDeaux JR, McCarroll R, Bush D, generations of selection. The same individuals are self- Hauge B, Lai CQ, Clark D, Rocheford TR and Dudley JW pollinated or crossed with a common tester to produce (2004) The genetic architecture of response to long-term entries that are evaluated in replicated trials. Entry artificial selection for oil concentration in the maize means are used to represent the phenotypes of the kernel. Genetics 168: 2141–2155. parental individuals from which genotype data were pro- Remington DL, Thornsberry JM, Matsuoka Y, Wilson LM, Whitt SR, Doeblay J, Kresovich S, Goodman MM and Buckler ES duced, and association mapping is conducted by fitting (2001) Structure of linkage disequilibrium and phenotypic the genotype data to the phenotype data. With the associations in the maize genome. Proceedings of the same genotype data, estimates of allelic frequencies are National Academy of Sciences USA 98: 11479–11484. obtained and examined across generations of selection, Stuber CW and Moll RH (1972) Frequency changes in isozyme particularly at genomic loci identified via association alleles in a selection experiment for grain yield in maize mapping. The two approaches provide complementary (Zea mays L.). Crop Science 12: 337–340. Wisser RJ, Murray SC, Kolkman JM, Ceballos H and Nelson RJ tests to examine loci that are responsive to selection. (2008) Selection mapping of loci for quantitative disease resistance in a diverse maize population. Genetics 180: 583–599. Yu JM, Pressoir G, Briggs WH, Bi IV, Yamasaki M, Doebley JF, Acknowledgements McMullen MD, Gaut BS, Nielsen DM, Holland JB, Kresovich S and Buckler ES (2006) A unified mixed-model method for association mapping that accounts for multiple levels of This research was supported by the USDA National Insti- relatedness. Nature Genetics 38: 203–208. tute of Food and Agriculture project 2007-35301-18133/ Zhu C, Gore M, Buckler ES and Yu J (2008) Status and prospects of 19859. association mapping in plants. The Plant Genome 1: 5–20. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 284–287 ISSN 1479-2621 doi:10.1017/S1479262111000177

Molecular basis of cytoplasmic male sterility in beets: an overview

Tetsuo Mikami*, Masayuki P. Yamamoto†, Hiroaki Matsuhira‡, Kazuyoshi Kitazaki and Tomohiko Kubo Laboratory of Genetic Engineering, Faculty of Agriculture, Hokkaido University, Sapporo 060-8589, Japan

Abstract Sugarbeet cultivars are almost exclusively hybrids, which are produced using the sole source of cytoplasmic male sterility (CMS), the so-called Owen CMS. Several alternative sources of CMS have been described. One of these, I-12CMS(3), was derived from wild beets collected in Pakistan, and another CMS source, GCMS, has a cytoplasmic origin in wild sea beets from France. During the past decade, male sterility-associated mitochondrial genes have been identiﬁed in these three CMS systems. Moreover, the recent development of a variety of DNA markers has permitted the genetic mapping of nuclear restorer-of-fertility genes for both Owen and GCMS. This review focuses on the mechanism of CMS in beets.

Keywords: beets; cytoplasmic male sterility; fertility restoration

Introduction of the cytoplasm. During the past decade, male sterility- associated mitochondrial genes have been reported in The identification of cytoplasmic male sterility (CMS) and three CMS sources in beets. We present in this study a maintaining genotypes was a major step in the success of brief overview of recent advances in our understanding hybrid breeding programmes of sugarbeet. The first of the mechanism of CMS in beets. report of CMS in this crop was made by F. V. Owen, who found male sterile plants in an old cultivar, ‘US1’ (Owen, 1945). He described that male sterility resulted Owen CMS-associated gene from the interaction of two recessive nuclear genes Because of its importance for breeding, the molecular (x and z) with the sterilizing cytoplasm and identified basis of Owen CMS has been extensively investigated. the maintainer plants, which had an xxzz genotype and Kubo et al. (2000) and Satoh et al. (2004) determined normal fertile cytoplasm. Hybrid seed production in the entire nucleotide sequences of the mitochondrial sugarbeet has relied entirely on this single CMS source, genomes from normal fertile and Owen CMS sugarbeet the so-called Owen CMS (Bosemark, 2006). Up to now, plants. In-depth sequence comparison of the two mito- additional sources of CMS have been discovered in wild chondrial genomes, together with a mitochondrial pro- beet populations (e.g. Mikami et al., 1985; Halldeń tein assay (Yamamoto et al., 2005), indicated that the et al., 1988; Bosemark, 1998; Touzet et al., 2004) and 50 leader sequence of atp6 (designated preSatp6) encodes might offer the opportunity to broaden the genetic base a variant 39 kDa protein that is possibly related to Owen CMS (Table 1). The 39 kDa preSATP6 is closely associated with the mitochondrial membrane and assembles into an Owen CMS-specific protein complex. Interestingly, this * Corresponding author. E-mail: [email protected] characteristic is shared by the sterility-related proteins †Present address: Department of Biology, Faculty of Science, identified in CMS-T maize and Ogura-CMS radish: University of Toyama, Toyama 930-8555, Japan. ‡Present address: National Agricultural Research Center for URF13 and ORF138 (Grelon et al., 1994; Krishnasamy Hokkaido Region, Sapporo 062-8555, Japan. and Makaroff, 1994; Wise et al., 1999). Cytoplasmic male sterility in beets 285 Other CMS sources genes (IV) RfG2 Rf A second source of CMS, called I-12CMS(3), was derived

Rf2 from wild beets collected in Pakistan (Mikami et al., 1985). The I-12CMS(3) and Owen cytoplasms show a (VIII) and

(IV) different patterns of male sterility maintenance and (III) and

gene male fertility restoration when crossed with the same Rf Rf1 More than two RfG1 R1H pollen parents (Mikami et al., 1985). What is worthy of special mention is that preSATP6 is missing in the sterile anthers carrying the I-12CMS(3) cytoplasm, which instead express a CMS-correlated protein of 12 kDa (Yamamoto et al., 2008). This 12 kDa protein proved to be encoded by an unusual mitochondrial gene, designated orf129 (Table 1). The translation product of orf129 was found to be primarily present in the matrix and loosely associated with the inner mitochondrial membrane, a feature

ﬂower buds and roots ﬂower buds and roots shared by the PCF protein involved in petunia CMS CMS-associated protein 39 kDa preSATP6 in 12 kDa ORF129 in 31 kDa COX2 in leaves (Nivison et al., 1994). Further implicating orf129 with CMS is the observation that ORF129 causes pollen disruption in transgenic tobacco plants when targeted to the mitochondria (Yamamoto et al., 2008). cox2 Another interesting case is GCMS, which has a cytoplasmic origin in wild sea beets from France. Ducos et al. (2001) reported that GCMS is associated with a

preSatp6 orf129 mitochondrial cox2 gene lacking eight highly conserved, C-terminal amino acids, and that cytochrome oxidase activity is decreased by 50% in vegetative tissues (Table 1). The GCMS can be classiﬁed as a loss-of-function mutant, in contrast to Owen CMS and I-12CMS(3), which are considered to be gain-of-function mutants. Beets thus appear to maintain distinct CMS cytoplasm, each capable of conferring male sterility by an apparently different mechanism. microspore degeneration microspore degeneration Microsporogenesis breakdown in Owen CMS

In Owen CMS plants, male meiosis proceeds normally until the tetrad stage, after which the microspores degen- erate either during tetrad formation or immediately after microspore liberation from the tetrads. Concurrently, the anther tapetum shows marked symptoms of abnorm- ality. The most common irregularity is extensive vacuo- lation and enlargement (hypertrophy) of the tapetal cells, accompanied by mitochondrial disorganization (Kaul, 1988; Majewska-Sawka et al., 1993). At this stage of development, demand for energy or particular mito- Wild beets collected in France Microspore degeneration Truncated Wild beets collected in France Microspore degeneration Not determinedchondrial Not determined biosynthetic products may be markedly high, so the impairment of mitochondrial function is devastating (Budar and Berthome´, 2007). Consistent with this

CMS sources in beets hypothesis is the observation that plants expressing an antisense mitochondrial pyruvate dehydrogenase subunit gene in tapetal cells are male sterile (Yui et al., 2003). Chromosomal location is given in parentheses. I-12CMS(3) Wild beets collected in Pakistan Tapetal hypertrophy; Cytoplasm Origin CMS phenotype Causal gene Table 1. Owen Cultivar ‘US-1’a Tapetal hypertrophy; G H Male sterile transgenic plants, in common with Owen 286 T. Mikami et al. CMS plants, exhibited poor development of tapetal is highly sensitive to mitochondrial dysfunction. It is mitochondria and aberrant formation of microspore therefore interesting to define a range of anther-specific walls. These results led us to speculate that the synthesis genes that may act downstream of preSatp6 or orf129 of preSATP6 protein in the tapetum of Owen CMS plants and that are responsible for aberrant biochemical and results in dysfunctional mitochondria, which in turn physiological processes leading to defective microsporo- could cause pollen abortion. genesis, in response to mitochondrial dysfunction (Matsuhira et al., 2007). Another important clue to the CMS mechanism will certainly come from identification Restorer-of-fertility (Rf) genes of the Rf gene products. This research is underway in our laboratory. Hagihara et al. (2005a) found that male fertility restoration in Owen CMS by a Japanese breeding line is controlled by a single dominant gene (designated Rf1). They constructed a regional map encompassing the Rf1 References locus (Hagihara et al., 2005b). The mapping data also provided a clear indication of the location of the Rf1 Bosemark NO (1998) Genetic diversity for male sterility in wild and cultivated beets. In: Frese L, Panella L, Srivastava HM locus on chromosome III, and the locus most likely cor- and Lange W (eds) International Beta Genetic Resources responds to the X gene described previously by Owen Network. A report on the 4th International Beta Genetic (Pillen et al., 1993; Table 1). The Rf1 locus could be Resources Workshop and World Beta Network Conference further narrowed to the 250 kb region, which was delim- held at the Aegean Agricultural Research Institute, Izmir, ited by two DNA markers (Hagihara et al., 2005b). Turkey, 28 February–3 March 1996. International Crop Network Series 12, International Plant Genetic Resources A second Rf locus (Rf2, Z) of Owen CMS has been Institute, Rome, pp. 44–56. located on chromosome IV (Schondelmaier and Jung, Bosemark NO (2006) Genetics and breeding. In: Draycott AP 1997; Hjerdin-Panagopoulos et al., 2002; Table 1). In (ed.) Sugar Beet. Oxford: Blackwell Publishing, pp. 50–88. GCMS, at least two genes were reported to restore Boudry P, Mo¨rchen M, Saumitou-Laprade P, Vernet PH and fertility, and one (RfG1) was mapped on chromosome Van Dijk H (1993) The origin and evolution of weed beets: consequences for the breeding and release of VIII and was not co-located with either Rf1 or Rf2 herbicide-resistant transgenic sugar beets. Theoretical (Touzet et al., 2004; Table 1). This is consistent with the and Applied Genetics 87: 471–478. conclusion that the causal mechanism of male sterility Budar F and Berthome´ R (2007) Cytoplasmic male sterilities and in Owen CMS is different from that in GCMS. mitochondrial gene mutations in plants. In: Logan DC (ed.) An additional source of CMS (designated HCMS) from Plant Mitochondria. Oxford: Blackwell Publishing, pp. 278–307. wild beets has been described by Boudry et al. (1993), Ducos E, Touzet P and Boutry M (2001) The male sterile G but its sterilizing gene remains unclear. Genetic analysis cytoplasm of wild beet displays modified mitochondrial of the fertility restoration of HCMS led to the conclusion respiratory complexes. The Plant Journal 26: 171–180. that a single dominant gene (R1H) is involved (Laporte Grelon M, Budar F, Bonhomme S and Pelletier G (1994) Ogura et al., 1998). Intriguingly, R1H is located on chromosome cytoplasmic male-sterility (CMS)-associated orf138 is translated into a mitochondrial membrane polypeptide in IV and is linked to the gene for monogerm seed character male-sterile Brassica cybrids. Molecular and General (Table 1). Several authors have reported weak linkage Genetics 243: 540–547. between the monogermy locus and the Rf2 gene associ- Hagihara E, Itchoda N, Habu Y, Iida S, Mikami T and Kubo T ated with Owen CMS (Hogaboam, 1957; Roundy and (2005a) Molecular mapping of a fertility restorer gene for Theurer, 1974). The question inevitably arises whether Owen cytoplasmic male sterility in sugar beet. Theoretical and Applied Genetics 111: 250–255. the same mitochondrial gene ( preSatp6) is responsible Hagihara E, Matsuhira H, Ueda M, Mikami T and Kubo T for both Owen CMS and HCMS or whether these two (2005b) Sugar beet BAC library construction and assembly types have a different CMS mechanism but share a of a contig spanning Rf1, a restorer-of-fertility gene for common Rf gene. This is worthy of further investigation. Owen cytoplasmic male sterility. Molecular Genetics and Genomics 274: 316–323. Halldeń C, Bryngelsson T and Bosemark NO (1988) Two new types of cytoplasmic male sterility found in wild Beta Future prospects beets. Theoretical and Applied Genetics 75: 561–568. Hjerdin-Panagopoulos A, Kraft T, Rading IM, Tuvesson S and How does the accumulation of CMS gene products, such Nilsson NO (2002) Three QTL regions for restoration of as preSATP6 and ORF129, in anther tissue lead to failed Owen CMS in sugar beet. Crop Science 42: 540–544. Hogaboam GT (1957) Factors influencing phenotypic pollen development? This issue has hardly been expression of cytoplasmic male sterility in the sugar beet addressed. As mentioned earlier, it has been shown that (Beta vulgaris L.). Journal of the American Society of sexual reproduction, and pollen production in particular, Sugar Beet Technologists 9: 457–465. Cytoplasmic male sterility in beets 287 Kaul MLH (1988) Male Sterility in Higher Plants. Berlin: Pillen K, Steinru¨cken G, Herrmann RG and Jung C (1993) An Springer-Verlag. extended linkage map of sugar beet (Beta vulgaris L.) Krishnasamy S and Makaroff CA (1994) Organ-specific reduction including nine putative lethal genes and the restorer gene in the abundance of a mitochondrial protein accompanies X. Plant Breeding 111: 265–272. fertility restoration in cytoplasmic male-sterile radish. Roundy TE and Theurer JC (1974) Linkage and inheritance Plant Molecular Biology 26: 935–946. studies involving an annual pollen restorer and other gen- Kubo T, Nishizawa S, Sugawara A, Itchoda N, Estiati A and etic characters in sugarbeets. Crop Science 14: 230–232. Mikami T (2000) The complete nucleotide sequence of Satoh M, Kubo T, Nishizawa S, Estiati A, Itchoda N and Mikami T the mitochondrial genome of sugar beet (Beta vulgaris L.) (2004) The cytoplasmic male-sterile type and normal type reveals a novel gene for tRNACys (GCA). Nucleic Acids mitochondrial genomes of sugar beet share the same Research 28: 2571–2576. complement of genes of known function but differ in the Laporte V, Merdinoglu D, Saumitou-Laprade P, Butterlin G, content of expressed ORFs. Molecular Genetics and Vernet P and Cuguen J (1998) Identification and mapping Genomics 272: 247–256. of RAPD and RFLP markers linked to a fertility restorer Schondelmaier J and Jung C (1997) Chromosomal assignment of gene for a new source of cytoplasmic male sterility in the nine linkage groups of sugar beet (Beta vulgaris L.) Beta vulgaris ssp. maritima. Theoretical and Applied using primary trisomics. Theoretical and Applied Genetics Genetics 96: 989–996. 95: 590–596. Majewska-Sawka A, Rodriguez-Garcia MI, Nakashima H and Touzet P, Hueber N, Bu¨rkholz A, Barnes S and Cuguen J (2004) Jassen B (1993) Ultrastructural expression of cytoplasmic Genetic analysis of male fertility restoration in wild male sterility in sugar beet (Beta vulgaris L.). Sexual cytoplasmic male sterility G of beet. Theoretical and Plant Reproduction 6: 22–32. Applied Genetics 109: 240–247. Matsuhira H, Shinada H, Yui-Kurino R, Hamato N, Umeda M, Wise RP, Gobelman-Werner K, Pei D, Dill CL and Schnable PS Mikami T and Kubo T (2007) An anther-specific lipid (1999) Mitochondrial transcript processing and restoration transfer protein gene in sugar beet: its expression is of male fertility in T-cytoplasm maize. The Journal of strongly reduced in male-sterile plants with Owen Heredity 90: 380–385. cytoplasm. Physiologia Plantarum 129: 407–414. Yamamoto MP, Kubo T and Mikami T (2005) The 50-leader Mikami T, Kishima Y, Sugiura M and Kinoshita T (1985) sequence of sugar beet mitochondrial atp6 encodes a novel Organelle genome diversity in sugar beet with normal polypeptide that is characteristic of Owen cytoplasmic male and different sources of male sterile cytoplasms. Theoreti- sterility. Molecular Genetics and Genomics 273: 342–349. cal and Applied Genetics 71: 166–171. Yamamoto MP, Shinada H, Onodera Y, Komaki C, Mikami T and Nivison H, Sutton CA, Wilson RK and Hanson MR (1994) Kubo T (2008) A male sterility-associated mitochondrial Sequencing, processing, and localization of the petunia protein in wild beets causes pollen disruption in transgenic CMS-associated mitochondrial protein. The Plant Journal plants. The Plant Journal 54: 1027–1036. 5: 613–623. Yui R, Iketani S, Mikami T and Kubo T (2003) Antisense Owen FV (1945) Cytoplasmically inherited male-sterility in inhibition of mitochondrial pyruvate dehydrogenase E1a sugar beets. Journal of Agricultural Research 71: subunit in anther tapetum causes male sterility. The Plant 423–440. Journal 34: 57–66. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 288–290 ISSN 1479-2621 doi:10.1017/S1479262111000621

Agronomic and molecular analysis of heterosis in alfalfa

C. Scotti1*, M. Carelli1, O. Calderini2, F. Panara2, P. Gaudenzi1, E. Biazzi1,S.May3, N. Graham3, F. Paolocci2 and S. Arcioni2 1CRA-FLC Centro per le Produzioni Foraggere e Lattiero-Casearie, Lodi, Italy, 2CNR-Istituto di Genetica Vegetale, Perugia, Italy and 3School of Biosciences, University of Nottingham, Sutton Bonington Campus, Loughborough LE12 5RD, UK

Abstract Double ‘free-hybrids’ (DH) in alfalfa were obtained by crossing in a diallelic scheme, six multiplied simple hybrids (SH) derived from four partly inbred (S2) lines. Analysis of the specific combining ability demonstrated that the main source of variation was for dry matter yield (DMY) in DHs and supported heterosis values of DHs versus the best parent of an average þ45% (ranging from þ5toþ76%). Investigation at the molecular level was carried out by analysis of simple sequence repeat markers on the six parental SHs and 15 DH progenies and by comparison of gene expression profiles using microarrays of a single DH line to its parental lines. The variation of heterozygosity estimates of the DHs explained a small part (about 20%) of their variation in DMY, while the number of alleles was significantly related to DM performance (r¼0.61; P , 0.05). The microarray analysis identified genes with both significant additive and non-additive levels of expression in the hybrid compared with the parents. The majority of the variation in gene expression was additive (87%), but among the genes with a non-additive pattern of expression, the greater proportion of probe sets (86%) fell outside the parental range. Gene ontology analysis of these genes revealed the presence of a number of terms related to metabolism and genetic information processing.

Keywords: free hybrids; Medicago sativa; simple sequence repeat markers; transcriptome

Introduction using two molecular approaches: ﬁrst, by estimating the genetic diversity and the heterozygosity levels using Double ‘free-hybrids’ (DH, obtained in the absence of simple sequence repeat (SSR) molecular markers and, male sterility mechanisms and consequently with a same second, by analyzing the variation in gene expression theoretical probability of within- and among-population using microarrays. crossings) in alfalfa, an autotetraploid, allogamous and perennial forage crop, were obtained using a variety construction process developed at the Lodi Institute (Rotili Materials and methods et al., 1999; see Supplementary Fig. S1, available online only at http://journals.cambridge.org). Plant material The aim of this study was to analyze the heterosis effect in alfalfa by comparing the DH and their respective S2 families originating from eight unrelated and geo- parents (multiplied simple hybrids (SHs) or 2S2Syn3) graphically distant populations, selected for dry matter production (DMY) during selﬁng, were crossed manually without emasculation to produce SH in a diallelic scheme. One single plant/family was used for each * Corresponding author. E-mail: [email protected] cross (seven plants/family in total). The six SHs from Analysis of heterosis in alfalfa 289

3% 28% 4% 34% Biosystems, USA). Heterozygosity was estimated by means of the average number of alleles (peaks)/locus.

69% 62% Microarray analysis Diallel A Diallel B SCA (specific combining ability) The heterotic DH B2 £ B5 line was chosen for transcrip- GCA (general combining ability) tome analysis. Both the DH forms, obtained using either one or four plants/parent, were grown and displayed a Error similar heterotic behaviour. The microarray analysis was Fig. 1. Diallel analysis: variance partitioning for DMY in performed on the ﬁrst form. The transcriptome of the diallel crosses. DH B2 £ B5 was compared with the two parents

(2S2Syn3) using the Affymetrix Medicago Genome Array the four parents with the highest general combining (Affymetrix, USA). RNA was extracted from leaves of ability (GCA) effects were multiplied upto the Syn3 plants at early flowering stage, each parent of the DH generation and finally crossed in a diallelic scheme to B2 £ B5 being a single plant; the RNA of ten heterotic produce DH. DHs were produced in different years, by progenies of the DH was bulked in one sample. Hybrid- two procedures: using a single multiplied SH plant ization signals were analysed using the Genespring selected for vigour, for each parent (five plants/family GX11 software (Agilent technologies, USA) using in total) or crossing four plants/parental SH (20 plants/ robust multichip average (RMA) prenormalization family in total). The DHs, and their respective parents, algorithm. Subsequently per-gene normalization was per- obtained with the second procedure were used for SSR formed by standardizing the probe set signals to the analysis. median value for all arrays. Gene expression was classi- £ Two diallelic crosses (diallel A and B), 6 6, were fied as additive/non-additive according to Hochholdinger obtained using the above process and the resulting and Hoecker (2007). Genes having a hybrid expression DHs were analyzed for DMY in greenhouse conditions value similar to the parental average (fold change ,1.5, for 2 years (ten harvests) together with their parents 22 t-test P # 0.05) were considered as additive. Non-addi- (150 plants/family at the density of 250 plants/m ). All tive genes were classified in functional categories using the 15 DHs of diallel A were analyzed using the SSR the GeneBins software (Goffard and Weiller, 2007). All approach (ten plants/DH family, five vigorous and five the data have been donated to the GEO database weak; 150 plants in total). (http://www.ncbi.nih.gov/geo/, with accession ID GSE25034).

Molecular analysis Results and discussion Sixty-three microsatellite (SSR) markers derived from both M. truncatula and M. sativa were analyzed by Diallelic analysis indicated speciﬁc combining ability multiplex PCR reactions and capillary electrophoresis (SCA) as the main source of variation for DMY in the in an automated detection system (ABI 310; Applied DHs (Fig. 1). This important deviation from additivity

(a)1% (b) 1% 6% P 5% Low P High P

Additive Below low parent Above high parent Above high parent High parent Low parent 1128 174 154 917 Below low parent

87%

Cutoff Fig. 2. Transcriptome analysis. (a) Proportions of genes falling in the additive/non-additive category. (b) Classification of genes assigned to the non-additive category. 290 C. Scotti et al. expressed by SCA supported heterosis values versus the unclassified and without homologue, 44.1 and 13.9%, best parent of þ50% in the DH B2 £ B5 and an average respectively) were metabolism (carbohydrate 10.5%; of þ45% (ranging from þ5toþ76%) in the 15 DHs of amino acid 7.3%; cofactors and vitamins 6.3%; lipid diallel A. The diallel hybridization of six SHs, from the 5.6%; nucleotide 4.9%; secondary metabolites 4.8%) and diallel crossing of four S2 lines, produced unrelated genetic information processing (signal transduction (3 out of 15) and related (12 out of 15) DHs, based on 7.2%; translation 6.0%; folding, sorting and degradation the presence or absence of a constituent in common 6.1%). Equivalent additive gene expression has been between the two parental SHs. The genetic similarity observed in other studies of heterotic hybrids in different between parents, estimated by Dice coefficient based species (maize, Arabidopsis; reviewed by Hochholdinger on 63 SSR markers, was consistent with the kinship and Hoecker, 2007). Interestingly, Li et al. (2009), study- degree of parents. In diallel A, the related parents ing free-hybrids between a Medicago falcata and two showed an average similarity of 0.61, significantly M. sativa populations, reported that probe sets exhibiting higher than the value of 0.56 found for the three unre- expression levels different from midparent value (non- lated DHs. In diallel B, the genetic similarity between additive expression) had a higher proportion displaying the related parental SHs B2 and B5 was 0.64. Parental over/under-dominance in heterotic (interspecific hybrids, genetic diversity showed a significant relationship with i.e. falcata £ sativa) than in non-heterotic (intraspecific, DMY (r ¼ 0.59; P , 0.05), heterosis versus the best i.e. sativa £ sativa) hybrids. Both experiments indicate parent (r ¼ 0.70; P , 0.01) and SCA effects (r ¼ 0.76; that genes with non-additive expression, and in particular P , 0.005) of DH progenies in diallel A. those exhibiting over/under dominance, play a role in Parental diversity was the basis of the significant yield heterosis in alfalfa. It is worth underlining that, in heterozygosity recovery in DHs with respect to the an autotetraploid allogamous species such as alfalfa and parents, 2.16 alleles/locus versus 2.11 in parents in in the free-hybrid construction process used, the parental diallel A (average of 15 DHs and their respective parents) multiplied SHs are not homozygous as the maize inbred and 2.14 versus 2.04 in B2 £ B5. However, the variation lines, but have, on average, 2.11 (diallel A) alleles/ of heterozygosity estimates explained only a small part locus. As a consequence, parental performance, either (,20%) of the variation in DMY of the DHs (r ¼ 0.45; agronomic (DMY) or as gene expression, is in part P . 0.05, in diallel A), while the number of alleles based on interactions between the two different alleles based on all the 63 SSR loci (allelic richness) was signi- and deviations from additivity (SCA in the case of DMY ficantly related to DM performance (r ¼ 0.61; P , 0.05). and over/under dominance in the case of gene It should be feasible to relate SCA values, found in the expression) reflect mainly higher order interactions at diallelic analysis of DHs to different genetic parameters single loci and/or epistatic phenomena. including allelic interactions, estimated by heterozygosity A combination of agronomic and advanced molecular level/locus; non-allelic interactions, estimated by allelic approaches is likely to have great potential in highlighting richness on the whole set of SSR loci studied; positive the genetic basis of heterosis in autotetraploids species. complementation of specific interactions, estimated by the study of differential gene expression of a DH and the respective parents. Expression analysis of the heterotic hybrid B2 £ B5 revealed that 87% of the probe sets were expressed in References an additive manner (Fig. 2(a)). Of the genes expressed Goffard N and Weiller G (2007) GeneBins: a database for classi- non-additively, the majority (86%) were outside the par- fying gene expression data, with application to plant ental range, i.e. either above the high parent or below genome arrays. BMC Bioinformatics 8: 47. the low parent expression levels (Fig. 2(b)). To confirm Hochholdinger F and Hoecker N (2007) Towards the molecular these results, gene expression values were validated for basis of heterosis. Trends in Plant Science 12: 427–432. 14 out of 17 selected genes using quantitative-PCR Li X, Wei Y, Nettleton D and Brummer C (2009) Comparative (data not shown). gene expression profiles between heterotic and non- heterotic hybrids of tetraploid Medicago sativa. BMC To further investigate the function of these gene sets, Plant Biology 9: 107. gene ontology analysis was performed. Analysis of the Rotili P, Gnocchi G, Scotti C and Zannone L (1999) Some aspects genes expressed outside the parental range showed of breeding methodology in alfalfa. www.naaic.org/TAG/ that the highest represented categories (excluding TAGpapers/rotili/rotili.html q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 291–295 ISSN 1479-2621 doi:10.1017/S1479262111000207

Mapping QTLs for yield components and chlorophyll a ﬂuorescence parameters in wheat under three levels of water availability

Ilona Czyczyło-Mysza1*, Izabela Marcin´ska1, Edyta Skrzypek1, Małgorzata Chrupek1,2, Stanisław Grzesiak1, Tomasz Hura1, Stefan Stojałowski3, Beata Mys´ko´w3, Paweł Milczarski3 and Steve Quarrie4 1The F. Go´rski Institute of Plant Physiology, Polish Academy of Sciences, Poland, 2Rzeszo´w University of Technology, Rzeszo´w, Poland, 3West Pomeranian University of Technology in Szczecin, Poland and 4Institute for Research on Environment and Sustainability, Newcastle University, UK

Abstract Drought is one of the major factors limiting wheat yield in many developing countries worldwide. Parameters of chlorophyll a fluorescence kinetics under drought stress conditions have been used to characterize dehydration tolerance in wheat. In the present study, a set of 94 doubled haploid lines obtained from Chinese Spring £ SQ1 (CSDH), mapped with 450 markers, was evaluated for yield (grain dry weight/main stem ear), number of grains/main stem ear (NG) and chlorophyll a fluorescence parameters (FC) under moderate and severe drought stress, and compared with results for well-watered plants. quantitative trait loci (QTLs) were identified using Windows QTLCartographer version 2.5 software and the results were analysed using single-marker analysis (SMA) and composite interval mapping (CIM). Analysis using SMA and CIM showed mostly similar QTLs for all traits, though more QTLs were identified by SMA than by CIM. The genetic control of yield, NG and FC varied considerably between drought-stressed and non-stressed plants. Although no major QTL co-locations were found for yield and FC using

CIM, the co-location of QTLs for NG, yield and Fv/Fm in drought-stressed plants was observed on chromosome 5A using SMA.

Keywords: chlorophyll a ﬂuorescence parameters; drought; quantitative trait loci; Triticum aestivum L; yield

Introduction very sensitive to water stress (Lu and Zhang, 1998). Changes in the biochemical reaction of PSII can be Drought is one of the major factors limiting wheat yields sensitively characterized by chlorophyll fluorescence as in many developing countries worldwide. Drought-resist- a practical guide to study the relationship between ant plants maintain physiological processes like photo- activity of the photosynthetic apparatus and drought tol- synthesis at lower water potentials (Pugnaire et al., erance in wheat (Lu and Zhang, 1998; Inoue et al., 2004; 1999). Photosystem II (PSII) has been proven to be Tambussi et al., 2005; Zhang et al., 2005). Mapping quantitative trait loci (QTLs) for yield and chlorophyll a fluorescence parameters will provide information on their genetic relationships, by identifying genes that * Corresponding author. E-mail: [email protected] control them, and thereby provide opportunities to 292 I. Czyczyło-Mysza et al. improve crop yield by increasing photosynthetic were tested for each line during 1-year experiment. QTLs capacity. The aim of this study was to identify the QTLs were identified using Windows QTLCartographer version controlling yield components and chlorophyll fluor- 2.5 (Wang et al., 2007) and the results were analysed escence as physiological factors associated with drought using single-marker analysis (SMA) and composite stress in wheat. interval mapping (CIM). A QTL was accepted when the likelihood odds ratio (LOD) score was greater than 3.

Materials and methods Results and discussion The mapping population used in this study Chinese Spring, SQ1, Double, Haploids (CSDH) consisted of 94 QTL analysis identified several genomic regions involved doubled haploid lines (DHLs) generated from the cross in regulating yield and chlorophyll fluorescence par- between hexaploid wheat (Triticum aestivum L.) geno- ameters under three water regimes (Tables 1 and 2). QTL types Chinese Spring (CS) and SQ1 (a high abscisic acid analysis using SMA and CIM showed mostly similar results breeding line), according to Quarrie et al. (2005). The for all traits, though more QTLs were identified by SMA CSDH genetic map presented in Quarrie et al. (2005) (Table 1) than by CIM (Table 2). The genetic control of was used for QTL analysis of yield components: yield, NG and FC varied considerably between drought- number of grains/main stem ear (NG), grain dry stressed and non-stressed plants. Both methods identified weight/main stem ear (YIELD) and chlorophyll fluo- a major QTL for NG under MD on chromosome 6D rescence parameters: Fv/F0 (maximum primary yield of (LOD ¼ 3.3) and QTLs under SD on chromosomes 2B 2 photochemistry of PSII), Fv/Fm (maximum quantum (two QTLs) and 5D, with R of 11.2, 17.0, 17.7 and yield of PSII), performance index as an essential indicator 12.0%, respectively. QTLs for yield were identified on of sample vitality (P.I.) and electron transport flux/excited chromosomes 5A and 6B only under MD. SMA identified cross section, at t ¼ 0 (ET0/CS). These traits were eva- additional major QTLs on 5A for YIELD and NG for both luated under moderate (MD) and severe (SD) drought drought stress conditions. QTLs for NG by SMA were also stress and compared with well-watered plants (WW). localized on chromosome 4A under MD and 3B under SD. Plants were grown in pots in a rain-out shelter. MD and The level of chlorophyll a fluorescence varied con- SD drought conditions were imposed for 4 weeks siderably amongst drought treatments. QTL mapping of during the late vegetative stage by maintaining soil FC using CIM identified 20 additive QTLs: 8 QTLs water contents at 60 and 40% of WW water contents. detected under WW, 1 QTL under MD and 13 QTLs Chlorophyll fluorescence parameters were measured on under SD for the four FC traits (Table 2). QTLs for the last day of drought stress conditions. Three plants Fv/F0, which provides an estimate of leaf photosynthetic

Table 1. Summary of QTLs for yield components and chlorophyll a ﬂuorescence parameters of DHLs under three water regimes identiﬁed using SMA

Chromosomal locations Water regime Traits WW MD SD NG 2A*, 2B***, 3D*, 5D*, 6D*, 7A*, 7B* 2B*, 2D*, 4A**, 5A**, 5B*, 1A*, 2B***, 2D*, 3B**, 4B*, 5A**, 5D*, 6A*, 6D***, 7A* 5D*, 6D**, 7A*, 7B**, 7D* YIELD 2D*, 4B***, 4D*, 5B*, 5D*, 6B* 2D*, 4A*, 4B**, 5A***, 5B**, 2B*, 2D*, 3A*, 3B*, 4B**, 5A**, 6B***, 6D*, 7A* 5D*, 6B** Fv/F0 1B*, 2A**, 2B**, 2D**, 3A**, 3B*, 1B*, 2A*, 2D**, 3A*, 3B*, 3D*, 1B**, 2A***, 2D***, 4A**, 5A*, 5B*, 4B**, 5B*, 6B*** 4A*, 4B*, 6A*, 7A**, 7D** 6B**, 7B**, 7D* Fv/Fm 1B*, 2A**, 2B**, 2D**, 3A**, 1B*, 2A*, 2D**, 3A*, 3B*, 1B**, 2A***, 2D***, 3D*, 4A**, 4D*, 3B*, 4B**, 4D**, 5D*, 6B***, 7D* 4A*, 4B*, 6A*, 7A**, 7D** 5A**, 5B*, 6B**, 7A*, 7B**, 7D* P.I. 1A*, 1B**, 1D*, 2A**, 2B**, 2D**, 2B*, 2D*, 3B*, 3D*, 4B*, 1B*, 2A****, 2B*, 2D**, 3A*, 3B*, 4A*, 3A**, 3B*, 4A*, 4B**, 4D***, 5A*, 5A*, 5D*, 6B*, 6D*, 7D* 4B**, 4D*, 5A*, 5B*, 7B**, 7D* 5D**, 6B***, 6D**, 7A*, 7D* ET0/CS 1A*, 1B**, 2A*, 2B***, 2D**, 3B*, 3B*, 3D**, 5B*, 5D*, 6D*, 7D* 1B*, 1D*, 2A***, 2B*, 2D*, 3A*, 4B**, 4B**, 4D***, 5A**, 5B*, 5D**, 4D**, 5B*, 6D*, 7B** 6B***, 6D**, 7A* *,**,*** and **** Signiﬁcance at 0.05, 0.01, 0.001 and 0.0001 probability level, respectively. Mapping QTLs in wheat 293

Table 2. Main characteristics of significant QTLs for yield components and chlorophyll a fluorescence parameters of DHLs under three water regimes identified using CIM

Traits Water regime QTL Marker Peak LOD R 2 (%) Add NG MD QGne.csdh-6D cfd49 2 3.3 11.2 22.590 SD QGne.csdh-2B-1 barc128a 121 4.1 17.0 2.580 SD QGne.csdh-2B-2 barc101a 128 5.0 17.7 2.650 SD QGne.csdh-5D cfd7 17 3.1 12.0 22.250 YIELD MD QGwe.csdh-5A psr120.1 86 3.4 10.0 0.100 MD QGwe.csdh-6B-1 wmc397 68 4.5 14.1 20.120 MD QGwe.csdh-6B-2 GS1 79 5.0 15.5 20.120 Fv/F0 WW QFv /F0.csdh-6B m71p78.3 71 4.4 16.7 20.098 SD QFv /F0.csdh-1B psr2019.2 91 3.8 12.6 0.080 SD QFv /F0.csdh-2A-1 m85p65.4 89 5.2 16.6 0.091 SD QFv /F0.csdh-2A-2 psr575.1 102 6.3 22.2 0.105 Fv/Fm WW QFv /Fm.csdh-6B m87p78.5aYT 73 4.3 16.2 20.003 SD QFv /Fm.csdh-1B psr2019.2 91 3.7 11.9 0.002 SD QFv /Fm.csdh-2A psr575.1 102 5.4 18.6 0.003 SD QFv /Fm.csdh-4A dupw004a 107 3.2 9.5 0.002 P.I. WW QP.I.csdh-2A psr575.1 73 3.3 9.1 0.200 WW QP.I.csdh-4D psp3103 0 5.2 15.5 0.314 WW QP.I.csdh-6B m71p78.3 71 3.5 10.5 20.221 WW QP.I.csdh-7D psp3094.2 173 4.0 14.0 20.257 MD QP.I.csdh-7D wmc488a 146 3.6 16.6 20.210 SD QP.I.csdh-1B psr325.2 99 3.3 9.8 0.152 SD QP.I.csdh-2A psr575.1 98 5.5 17.5 0.203 SD QP.I.csdh-4B psp3163 51 3.4 10.2 20.153 ET0/CS WW QET0 /CS.csdh-2D psr331.1 129 4.0 11.6 8.121 WW QET0 /CS.csdh-6B m71p78.3 71 4 14.1 28.072 SD QET0 /CS.csdh-2A psr575.1 98 6.4 22.8 6.086 SD QET0 /CS.csdh-4B blt801 47 3.7 11.9 24.440 Peak, position of QTL peak from ﬁrst marker in cMorgans; R 2 (%), % of phenotypic variance explained by the QTL; Add, additive effect of the CS allele.

capacity, were found under WW on 6B and under SD on localization of QTLs by SMA for NG and all FC para- 1B and 2A (two QTLs), explaining 16.7 and 12.6, 16.6, meters were identified on chromosome 7B (P ¼ 0.01) 22.2% of the phenotypic variation, respectively. QTLs under SD. For both droughts, in comparison to WW, for other fluorescence parameters under WW were ident- co-localization of QTLs for Fv/F0 and Fv/Fm was ified: Fv/Fm on chromosome 6B; P.I. on chromosomes 2A, observed also on chromosomes 4A, 7A and 7D (Table 1). 4D, 6B, 7D and ET0/CS on chromosomes 2D and 6B, The presence of QTLs for Fv/F0 on chromosome 7D explaining phenotypic variation ranging from 9.1 to was also identified in wheat drought studies of Yang

16.7%. QTLs for Fv/Fm, which is related to the drought et al. (2007). Notably, using CIM, no QTLs for yield tolerance of photosynthesis, were identiﬁed on chromo- and FC parameters were coincident for any of the somes 1B, 2A and 4A under SD, explaining pheno- water regimes used here. However, the presence of typic variation ranging from 9.5 to 18.6%. Three QTLs coincident QTLs for NG, YIELD and Fv/Fm was controlling P.I. under SD were located on chromosomes observed by SMA on chromosome 5A (P ¼ 0.01).

1B, 2A and 4B. Two QTLs for ET0/CS located on Till now, molecular markers have enabled identification chromosomes 2A and 4B (Table 1) explained 22.8 and of many QTLs that are involved in the expression of 11.9% of the phenotypic variation under SD, respectively. agronomically important traits of wheat, such as yield Only one QTL under MD was detected for P.I.: chromo- and its components (e.g. Bo¨rner et al., 2002; Campbell some 7D. CIM of all FC parameters showed major loci et al., 2003; Kumar et al., 2006, 2007; Kirigwi et al., 2007; (LOD . 5.0) under SD on chromosome 2A (near Maccaferri et al., 2008, 2011). The genetic dissection of psr575.1), with the increasing allele from CS. These 2A water stress resistance in the CSDH mapping population QTLs were confirmed by SMA under MD and, more has previously been carried out by Quarrie et al. (2005, significantly, under SD. 2006) and Czyczyło-Mysza (unpublished). SMA identified additional major QTLs on 2B for all Relationships between physiological traits have been FC parameters under WW. In comparison to WW, investigated in many studies through QTL mapping 294 I. Czyczyło-Mysza et al.

(e.g. Lebreton et al., 1995; Quarrie et al., 1997; for grain yield in wheat under drought. Molecular Breeding Thumma et al., 2001). Chlorophyll fluorescence has 20: 401–413. Kumar N, Kulwal PL, Gaur A, Tyagi AK, Khurana JP, great potential as a high-throughput screen for Khurana P, Balyan HS and Gupta PK (2006) QTL analysis photosynthetic potential and therefore yield potential. for grain weight in common wheat. Euphytica 151: Nonetheless, evidence for its association with genetic 135–144. gains in yield in crops is lacking, and Foulkes et al. Kumar N, Kulwal PL, Balyan HS and Gupta PK (2007) QTL (2009) suggest that, it would be difficult to screen for mapping for yield contributing traits in two mapping populations of bread wheat. 19: this trait in any crop breeding programmes. Here, we Molecular Breeding 163–177. have used parameters of chlorophyll a fluorescence Lebreton C, Lazic-Jancic V, Steed A, Pekic S and Quarrie SA kinetics under drought stress conditions to characterize (1995) Identification of QTL for drought responses in dehydration tolerance in wheat. Until now, only a few maize and their use in testing causal relationships between studies have mapped loci associated with FC para- traits. Journal of Experimental Botany 46: 853–865. meters in wheat (Yang et al., 2007; Liang et al., Liang Y, Zhang K, Zhao L, Liu B, Meng Q, Tian J and Zhao S (2010) Identification of chromosome regions conferring 2010). Our identification of QTLs for FC parameters dry matter accumulation and photosynthesis in wheat will help improve understanding of the genetic basis (Triticum aestivum). Euphytica 171: 145–156. of photosynthetic processes in plants. Lu CM and Zhang JH (1998) Effects of water stress on photo- The present preliminary study has shown that yield synthesis, chlorophyll fluorescence and photoinhibition in and FC parameters are suitable quantitative traits for wheat plants. Australian Journal of Plant Physiology 25: 883–892. identification of QTLs that could be involved in the regu- Maccaferri M, Sanguineti MC, Natoli E, Araus-Ortega JL, Ben lation of drought resistance of wheat. We plan further Salem M, Bort J, Chenenaoui S, Deambrogio E, Garcia QTL analyses of these parameters under drought con- Del Moral L, Demontis A, El-Ahmed A, Maalouf F, Machlab ditions in wheat to confirm the weak associations H, Moragues M, Motawai J, Nachit M, Nserallah N, between the genetic control of fluorescence parameters Ouabbou H, Royo C and Tuberosa R (2008) Quantitative and yield identified here. trait loci for grain yield and adaptation of durum wheat (Triticum durum Desf.) across a wide range of water availability. Genetics 178: 489–511. Maccaferri M, Sanguineti MC, Garcia Del Moral L, Demontis A, El- Ahmed A, Maalouf F, Moragues M, Nachit M, Nserallah N, Acknowledgements Royo C and Tuberosa R (2011) Association mapping in durum wheat grown across a broad range of water regimes The present research study was funded by COST FA0604 and yield potential. Journal of Experimental Botany 62: 409–438. (479/N-COST/2009/0). Pugnaire FI, Serrano L and Pardos J (1999) Constraints by water stress on plant growth. In: Pessarakli M (ed.) Handbook of Plant and Crop Stress. New York: Marcel Dekker, pp. 271–283. References Quarrie SA, Laurie DA, Zhu J, Lebreton C, Semikhodskii A, Steed A, Witsenboer H, Calestani C and Dodig D (1997) QTL Bo¨rner A, Schumann E, Furste A, Coster H, Leithold B, analysis to study the association between leaf size and Roder MS and Weber WE (2002) Mapping quantitative abscisic acid accumulation in droughted rice leaves and trait loci determining agronomic important characters in comparisons across cereals. Plant and Molecular Biology hexaploid wheat. Theoretical and Applied Genetics 105: 35: 155–165. 921–936. Quarrie SA, Steed A, Calestani C, Semikhodskii A, Lebreton C, Campbell BT, Baenziger PS, Gill KS, Eskridge KM, Budak H, Chinoy C, Steele N, Pljevljakusic D, Waterman W, Weyen Erayman M, Dweikat I and Yen Y (2003) Identification of J, Schondelmaier J, Habash DZ, Farmer P, Saker L, Clarkson QTLs and environmental interactions associated with agro- DT, Abugalieva A, Yessimbekova M, Turuspekov Y, nomic traits on chromosome 3A wheat. Crop Science 43: Abugalieva S, Tuberosa R, Sanguineti M-C, Hollington PA, 1493–1505. Aragues R, Royo A and Dodig D (2005) A high density Foulkes MJ, Reynolds MP and Bradley RS (2009) Genetic genetic map of hexaploid wheat (Triticum aestivum L.) improvement of grain crops: yield potential. In: Sandras from the cross Chinese Spring £ SQ1 and its use to com- VO and Calderini DF (eds) Crop Physiology: Applications pare QTLs for grain yield across a range of environments. for Genetic Improvement and Agronomy. New York: Theoretical and Applied Genetics 110: 865–880. Elsevier, pp. 355–385. Quarrie SA, Pekic Quarrie S, Radosevic R, Rancic D, Kaminska Inoue T, Inanaga S, Sugimoto Y and El Siddig K (2004) Contri- A, Barnes JD, Leverington M, Ceoloni C and Dodig D bution of pre-anthesis assimilates and current photosyn- (2006) Dissecting a wheat QTL for yield present in a thesis to grain yield, and their relationships to drought range of environments: from the QTL to candidate genes. resistance in wheat cultivars grown under different soil Journal of Experimental Botany 57: 2627–2637. moisture. Photosynetica 42: 99–104. Tambussi EA, Nogue´s S and Araus JL (2005) Ear of durum wheat Kirigwi FM, Van Ginkel M, Brown-Guedira G, Gill BS, Paulsen under water stress: water relations and photosynthetic GM and Fritz AK (2007) Markers associated with a QTL metabolism. Planta 221: 446–458. Mapping QTLs in wheat 295 Thumma BR, Naidu BP, Chandra A, Cameron DF, Bahnisch LM Yang DL, Jing RL, Chang XP and Li W (2007) Quantitative trait and Liu C (2001) Identification of causal relationships loci mapping for chlorophyll fluorescence and associated among traits related to drought resistance in Stylosanthes traits in wheat (Triticum aestivum). Journal of Integrative scabra using QTL analysis. Journal of Experimental Plant Biology 49: 646–654. Botany 52: 203–214. Zhang YJ, Zhao CJ, Liu LY, Wang JH and Wang RC (2005) Wang S, Basten CJ and Zeng ZB (2007) Windows QTL Carto- Chlorophyll fluorescence detected passively by difference grapher, new version. Statistical Genetics, North Carolina reflectance spectra of wheat (Triticum aestivum L.) leaf. State University. Journal of Integrative Plant Biology 47: 1228–1235. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 296–299 ISSN 1479-2621 doi:10.1017/S1479262111000268

Identifying QTLs for cold-induced resistance to Microdochium nivale in winter triticale

Magdalena Szechyn´ska-Hebda1, Maria We˛dzony1,2*, Mirosław Tyrka3, Gabriela Gołe˛biowska1,2, Małgorzata Chrupek1,3, Ilona Czyczyło-Mysza1, Ewa Dubas1, Iwona Z˙ur1 and Elz˙bieta Golemiec1 1Institute of Plant Physiology, Polish Academy of Sciences, Niezapominajek 21, Krakow 30-239, Poland, 2Pedagogical University of Krakow, Podchorazych 2, Krakow 30-084, Poland and 3Rzeszow University of Technology, Wincentego Pola 2, 35-959 Rzeszo´w, Poland

Abstract Snow mould caused by Microdochium nivale (Fr.) Samuels & Hallett is the most widespread seedling disease in winter cereals. Due to the complexity of the resistance mechanisms, a poorly understood genetic background and strong interaction with winter weather conditions, it is difﬁcult to assess the resistance of triticale cultivars via conventional inoculation methods. Genetic resistance is the most economical and environmental friendly way to control M. nivale infection; therefore, the objective of this study was to detect the quantitative trait loci (QTLs) associated with resistance components of winter triticale in a mapping population derived from a cross of the ‘Modus’ (partly resistant) and ‘SaKa 3006’ (sensitive) varieties. High-resolution mapping was conducted by using 1518 molecular markers (diversity arrays technology, simple sequence repeat and ampliﬁed fragment length polymorphism). Partial resistance components assessed in this study, i.e. candidate QTLs, were detected on chromosomes 1B, 2A, 3A, 3B, 5A, 5B, 6A, 6B and 7B, whereas QTLs describing overall seedling vitality in non-infected control plants were located on chromosomes 1B, 2B, 3A, 5A, 7B and 7R.

Keywords: disease resistance; Microdochium nivale; quantitative trait loci; seedling vitality; triticale

Introduction resistance of the best triticale cultivars is far from satisfac- tory. According to Nakajima and Abe (1996), Ergon et al. Microdochium nivale (Fr.) Samuels & Hallett is the most (1998), Tronsmo et al. (2001), and from our own experi- widespread snow mould fungus. It is a pathogen-caused ments (Gołe˛biowska and We˛dzony, 2009), cold-hardening disease which develops under high humidity and low is the factor switching on cereal defence responses to temperature, and results in major damage to economically M. nivale infection. The mechanisms of plant resistance important winter cereals and grasses. It is the source of induced by cold have not yet been recognised and seedling blight, stem rot, leaf blotch and a part of the the genotypes differ in their ability to obtain resistance disease complex called ‘fusarium head blight’. Genetic expression (Ho¨mmo, 1996; Tronsmo, 1994; Pulli et al., resistance is the most economical and environmental 1996). Thus, it is important to identify the genetic back- friendly way to control M. nivale infection. Triticale has ground and the markers that can be exploited to increase been suggested as a source of disease resistance genes natural resistance via marker-assisted selection. Moreover, since it is a bridging species for the transfer of such genes the most agronomically important characters in cereals, to wheat and rye (Kuleung et al., 2004). However, even including resistance, are classified as polygenic or quantitative (Reszka et al., 2007; Carter et al., 2009). Therefore, we present the results of studies in which we tried to identify QTLs related to seedling survival and its vitality * Corresponding author. E-mail: [email protected] while recovering from snow mould infection in triticale. Identifying QTLs for cold-induced resistance 297 The studies exploit the first triticale marker map (paper Results and discussion in preparation) and the mapping doubled haploids (DH) population being developed at SPB Institute (Germany) Genetic maps are the fundamental tools that identify the by Dr. Eva Bauer. features of phenotypes that are linked to specific genetic loci, including those that influence QTLs. Over 40 maps with at least 300 markers have been published for differ- Materials and methods ent Triticeae populations; however, to date, only a few rye and one partial triticale map with more than 300 The mapping population consisting of 90 DH lines was markers have been produced (Gonzalez et al., 2005; derived by the maize method (We˛dzony et al., 1998; Lehmensiek et al., 2009). Therefore, reports identifying We˛dzony, 2003) from triticale ( £ Triticosecale Wittmack) and mapping QTLs in triticale are limited (Reszka et al.,

F1 generation of ‘Modus’ (partially resistant) £ ‘SaKa3006’ 2007; Hura et al., 2009). In our research, the population (sensitive) (Fig. 1). of 90 DH lines derived from the F1 hybrid of triticale Seedlings were planted in multipots in a randomised ‘Modus’ and ‘SaKa3006’ was mapped with 1518 markers complete block design with three replicates. Plants selected from a total of 1670 wheat and rye candidate grew in a chamber at 16/128C, 10/12 h light/darkness for markers: simple sequence repeat, amplified fragment 2 weeks. Then they were subjected to a pre-hardening length polymorphism and diversity arrays technology period (128C for 7 d) and hardening (28C for 28 d) in the markers. The map spans a distance covering a total of same light regime. Hardened seedlings were inoculated 2197 cM genomic regions with an average interval with soil-borne mycelium of M. nivale (monosporal isolate of 1.44 cM between adjacent marker loci. The highest no. 38z/5a/01). Then the plants were covered with moist marker density was established in rye chromosomes 4R, paper and black plastic bags to imitate a snow cover and 5R, 6R and 7R. to keep high humidity in darkness at 28C. The control Resistance to M. nivale showed a wide variation range plants were treated the same way, except for the infection. within the mapping population, 23% of lines had lower The covers were removed after 21 d of incubation and the survival capability than resistant Magnat, 43% lines did seedlings were moved for 2 weeks to optimal conditions not differ from Magnat and 20% lines fully regenerated for recovery. Five traits describing seedling survival and after infection. However, CIM approaches, performed for vigour were selected for analysis (Table 1). 14 wheat and 7 rye chromosomes, revealed that reaction QTLs were identified using Windows QTLCartographer to snow mould in this population is not determined by 2.5 and the results were analysed using composite inter- ‘plant survival after infection’ because the QTLs were not val mapping (CIM). Threshold logarithm of the odds identified. In the case of plants regenerated after infec- (LOD) scores were calculated based on 1000 permu- tion, two QTLs were detected on chromosomes 2A and tations. A QTL was accepted when the LOD score was 3B (around the Xbarc212 and wPt-6802 regions, respecti- greater than 3. vely) at a similar position for both traits concerning the

120 Modus

100 Saka3006

40 Survival of DH lines (%)

Fig. 1. Survival of doubled haploids (DH) lines selected from a triticale mapping population after infection with M. nivale. Parents are indicated with the arrows. 298 M. Szechyn´ska-Hebda et al.

Table 1. Results of the CIM analysis – detected main QTLs for the studied traits

Position Trait QTL Chromosome Marker QTL (cM)a LOD R 2 (%) Addb No. of regenerated leaves/planted 1 2A Xbarc212 5.2 3.66 11.44 0.3457 plant – infected 2 3B wPt-6802 25.9 3.61 11.29 20.3458 3 3B wPt-4625 31.5 3.03 9.61 20.3170 4 5B wPt-9872 32.9 3.17 10.93 0.3343 No. of regenerated leaves/ planted 1 1B rPt-399959 87.8 7.82 27.54 0.4384 plant – control 2 7R rPt-399750 11.3 3.85 11.61 20.2862 Plant survival control (%) 1 2B rPt-509666 56.7 3.42 12.57 20.0121 No. of regenerated leaves/recovered 1 2A Xbarc212 5.2 6.27 20.35 0.3455 plants – infected 2 3B wPt-6802 25.9 4.58 14.22 20.3324 No. of regenerated leaves/recovered 1 5A Xwmc713 0 4.18 12.51 20.2900 plants – control 2 1B rPt-399959 83.8 7.49 25.78 0.4226 3 7B wPt-3723 84.1 3.36 10.01 0.2567 Height of regenerated plants – infected 1 5A rPt-389308 7.9 6.59 21.02 0.6645 Height of regenerated plants – control 1 5A rPt-389308 7.9 3.60 10.9 0.5601 2 7B wPt-9880 116.4 3.20 9.61 0.5251 Dry mass regenerated leaves – infected 1 6A Xgwm1009 55.6 3.40 12.48 20.0175 2 6A rPt-507997 64.3 5.75 20.08 20.0255 Dry mass regenerated leaves – control 1 3A Xcfa2193 106.9 4.54 16.47 0.0122 R 2 (%), % of phenotypic variance explained by the QTL; Add, additive effects of QTLs expressed in the trait unit. a Position of QTL peak from the first marker in cMorgans. b A positive value means that the allele from the Modus increases the value of the trait. number of regenerated leaves (traits 1 and 4, Table 1), values (about 3 for control and 6.6 for infected plants), with LOD from 3.03 to 6.27. Other QTLs for trait 1 were mapped on wheat chromosome 5A. Another QTL were discovered on chromosome 3B (near wPt-4625) for the trait ‘height of regenerated plant-control’, related and 5B (near wPt-9872), which explained 9.61 and rather to seedling morphology than disease resistance, 10.93% of the phenotypic variation. The QTLs on was located on chromosome 7B. Unlike what was chromosomes 3B and 5B were consistent with reports reported in rye (Bo¨rner et al., 2000), we detected one of resistance to Stagonospora nodorum blotch disease QTL for height on chromosome 5R, which confirmed in triticale and wheat (Reszka et al., 2007). The nearby that the results obtained depend on the population. region between the loci Xgwm533.1 and Xgwm493 on Zhang et al. (2008) constructed a linkage map for the chromosome 3B was shown to harbour a QTL related DH lines of the Huapei3/Yumai57 population and to Fusarium head blight resistance in wheat cultivars they detected four additive QTLs and five pairs of epi- Ning894037 and Wangshuibai (Jia et al., 2005; Shen static effects for plant height on chromosomes 3A, 4B, et al., 2006). Carter et al. (2009) mapped the putative 4D, 5A, 6A, 7B and 7D. However, within the population QTLs on 3B wheat chromosome, which is significantly of Hanxuan10/Lumai14, three additive QTLs signifi- associated with resistance to stripe rust. These coinci- cantly affecting plant height were found on chromo- dences suggest the possibility that resistance to certain somes 1B, 4D and 5B, and three pairs of epistatic fungal diseases may be controlled by the same resistance QTLs were found on chromosomes 1B–1B, 2A–2D, loci in wheat and triticale. 2D–5B (Wang et al., 2010). Candidate QTLs for other QTL distribution differed for the control plants. assessed components that could be related to resistance One QTL for traits that related the number of leaves in this study were detected on chromosomes 1B, 2A, was mapped on chromosome 1B with R 2 values of 2B, 3A, 6A, 6B and 7B (Table 1). 27.5 (trait 2, Table 1) and 25.8% (trait 5, Table 1). The pre- In conclusion, the results were consistent with the sence of one QTL on chromosome 7R (near rPt-399750) observation that plant resistance to M. nivale is polygenic for trait 2 and two QTLs on chromosomes 5A and 7B for in nature. Principal component analysis showed that trait 5 were also identified. 95.7% of the variation can be explained by the interaction QTLs controlling variation in ‘height of regenerated of five independent factors. Finally, although the triticale plant’ (traits 6 and 7), each showing significant LOD are hexaploids composed of genomes from durum Identifying QTLs for cold-induced resistance 299 wheat (AABB) and rye (RR), it seems that seedling vitality Lehmensiek A, Bovill W, Wenzl P, Langridge P and Appels R is mainly controlled by the wheat genome. (2009) Genetic mapping in the Triticeae. Genetics and Genomics of the Triticeae, vol. 7. Heidelberg/London/ New York: Springer, Dordrecht, pp. 201–235. Nakajima T and Abe J (1996) Environmental factors affecting Acknowledgements expression of resistance to pink snow mould caused by Microdochium nivale in winter wheat. Canadian Journal This work was financed by project COST/254/2006. of Botany 74: 1783–1788. Pulli S, Hjortesholm K, Larsen A, Gudleifsson B, Larsson S, M. S.-H. is grateful for financial support from projects Kristiansson B, Ho¨mmo LM, Tronsmo AM, Ruuth P and MERG-CT-2007-207350 and 595/N-COST/2009/0. Kristensson K (1996) Development and Evaluation of Laboratory Testing Methods for Winter Hardiness Breeding, vol. 32. SLU, Alnarp: Nordic Gene Bank, pp. 1–68. References Reszka E, Song Q, Arseniuk E, Cregan PB and Ueng PP (2007) The QTL controlling partial resistance to Stagonospora nodorum blotch disease in winter triticale Bogo. Plant Bo¨rner V, Korzun AV, Voylokov AJ, Worland W and Weber E Pathology Bulletin 16: 161–167. (2000) Genetic mapping of quantitative trait loci in rye Shen X, Francki MG and Ohm HW (2006) A resistance-like gene (Secale cereale L.). Euphytica 116: 203–209. identified by EST mapping and its association with a QTL Carter AH, Chen XM, Garland-Campbell K and Kidwell KK (2009) Identifying QTL for high-temperature adult-plant controlling Fusarium head light infection on wheat resistance to stripe rust (Puccinia striiformis f. sp. tritici) chromosome 3BS. Genome 49: 631–635. in the spring wheat (Triticum aestivum L.) cultivar Tronsmo AM (1994) Effect of different cold hardening regimes ‘Louise’. Theoretical and Applied Genetics 119: 1119–1128. on resistance to freezing on snow mould infection in Ergon A, Klemsdal S and Tronsmo AM (1998) Interactions timothy varieties of different origin. In: Doerffling K, between cold hardening and Microdochium nivale infec- Brettschneider B, Tantau H and Pithan K (eds) Crop tion on expression of pathogenesis-related genes in Adaptation to Cool Climates. Brussels: European Commis- winter wheat. Physiological and Molecular Plant Pathology sion (ECSP-EEC-EAEC), pp. 83–89. 53: 301–310. Tronsmo AM, Hsiang T, Okuyama H and Nakajima T (2001) Low Gołe˛biowska G and We˛dzony M (2009) Cold-hardening of temperature diseases caused by Microdochium nivale. In: winter triticale ( £ Triticosecale Wittm.) results in increased Iriki N, Gaudet DA, Tronsmo AM, Matsumoto N, Yoshida resistance to pink snow mould Microdochium nivale M and Nishimune A (eds) Low Temperature Plant Microbe (Fr., Samuels & Hallett) and genotype-dependent Interactions Under Snow. Sapporo: Hokkaido National chlorophyll fluorescence modulations. Acta Physiologiae Agricultural Experimental Station, pp. 75–86. Plantarum 31: 12–19. Wang Z, Wu X, Ren Q, Chang X, Li R and Jing R (2010) QTL Gonzalez JM, Muniz LM and Jouve N (2005) Mapping of QTLs mapping for developmental behavior of plant height in for androgenic response based on a molecular genetic wheat (Triticum aestivum L.). Euphytica 174: 447–458. map of (Tritocosecale Wittmack. Genome 48: 999–1009. We˛dzony M (2003) Protocol for doubled haploid production Ho¨mmo L (1996) Effect of hardening and dehardening on snow in hexaploid Triticale ( £ Triticosecale Wittm.) by crossing mould (Microdochium nivale) resistance. Nord Jordbruks- it with maize. In: Maluszynski M, Kasha KJ, Forster BP forskning 78: 89–92. and Szarejko I (eds) Doubled Haploid Production in Crop Hura T, Hura K and Grzesiak S (2009) Physiological and Plants – A Manual. ISBN 1-4020-1544-5. Dordrecht/ biochemical parameters for identification of QTLs control- Boston/London: Kluwer Academic Publishers, pp. 129–134. ling the winter triticale drought tolerance at the seedling We˛dzony M, Marcin´ska I, Ponitka A, S´lusarkiewicz-Jarzina A and stage. Plant Physiology and Biochemistry 47: 210–214. Wozna J (1998) Production of doubled haploids in triticale Jia G, Chen P, Qin G, Bai G, Wang X, Wang S, Zhou B, Zhang S ( £ Triticosecale Wittm.) by means of crosses with maize and Liu D (2005) QTLs for Fusarium head blight response (Zea mays L.) using Picloram and Dicamba. Plant Breeding in a wheat DH population of Wangshuibai/Alondra‘s’. 117: 211–215. Euphytica 146: 183–191. Zhang K, Tian J, Zhao L and Wang S (2008) Mapping QTLs with Kuleung C, Baenziger PS and Dweikat I (2004) Transferability of epistatic effects and QTL environment interactions for plant SSR markers among wheat, rye, and triticale. Theoretical height using a doubled haploid population in cultivated and Applied Genetics 108: 1147–1150. wheat. Journal of Genetics and Genomics 35: 119–127. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 300–304 ISSN 1479-2621 doi:10.1017/S1479262111000566

Use of EcoTILLING to identify natural allelic variants of rice candidate genes involved in salinity tolerance

S. Negra˜o1, C. Almadanim1, I. Pires1, K. L. McNally2 and M. M. Oliveira1* 1ITQB (Instituto de Tecnologia Quı´mica e Biolo´gica), IBET (Instituto de Biologia Experimental e Tecnolo´gica), Universidade Nova de Lisboa, Avenida da Repu´blica, 2780-157 Oeiras, Portugal and 2International Rice Research Institute, DAPO Box 7777, Metro Manila, Philippines

Abstract Rice is a salt-sensitive species with enormous genetic variation for salt tolerance hidden in its germplasm pool. The EcoTILLING technique allows us to assign haplotypes, thus reducing the number of accessions to be sequenced, becoming a cost-effective, time-saving and high- throughput method, ideal to be used in laboratories with limited ﬁnancial resources. Aiming to ﬁnd alleles associated with salinity tolerance, we are currently using the EcoTILLING technique to detect single nucleotide polymorphisms (SNPs) and small indels across 375 germplasm accessions representing the diversity available in domesticated rice. We are targeting several genes known to be involved in salt stress signal transduction (OsCPK17) or tolerance mechanisms (SalT). So far, we found a total of 15 and 23 representative SNPs or indels in OsCPK17 and SalT, respectively. These natural allelic variants are mostly located in 30-untranslated region, thus opening a new path for studying their potential contribution to the regulation of gene expression and possible role in salt tolerance.

Keywords: EcoTILLING; genetic variability; rice; salt tolerance

Introduction to food production, particularly to mitigate the effects of increasing salinization and climate change. Soil salinity existed long before humans and agriculture; In rice, differences among genotypes account for its however, the problem has been aggravated by agricul- response towards salinity (Zeng, 2005). The importance tural practices such as irrigation (Zhu, 2001). It is esti- of genes hidden in the primary rice gene pool mated that about 20% of irrigated agricultural land (including low-yielding ancestors and traditional land- throughout the world is adversely affected by salinity. races) to enhance rice performance in stress conditions Plants differ greatly in their tolerance to salinity, with was previously illustrated (Ali et al., 2006). Since the rice (Oryza sativa L.) being the most sensitive cereal. most common forms of genetic variation within natural Rice is the primary food source for over half of the populations are single nucleotide polymorphisms world’s population and has been the subject of countless (SNPs) and small insertions and deletions (indels), breeding programmes to increase its productivity and a further step can be achieved by investigating such tolerance to both biotic and abiotic stresses. Analysis of variation in key-responsive genes from diverse germ- the molecular mechanisms underlying salinity tolerance plasm (Raghavan et al., 2007). While we are now in is being undertaken to provide practical contributions the ‘next-generation sequencing’ era, many research programmes have limited funding and need more cost-effective, high-throughput techniques. The EcoTIL- LING method allows SNPs and indels discovery and the delineation of haplotypes at loci of interest (Comai * Corresponding author. E-mail: [email protected] et al., 2004). We are using EcoTILLING to explore the Use of EcoTILLING in rice varieties 301 natural variability existing in rice germplasm at key region of SalT (forward – ACCACTCAACACCGGTAGGA- genes related to salinity tolerance. Although this large- CACT; reverse – GCAGATTAAACTGGGCTCCTCTGA), scale analysis is still ongoing, we have already obtained corresponding to 979 and 825 bp products, respectively. promising results showing high genetic variability in The polymerase chain reaction (PCR) was performed rice germplasm at salinity tolerance loci and the poten- in 14 ml final volume, using 3.5 ng of total DNA as tem- tial to deliver superior alleles for breeding programmes plate, 0.4 U/reaction of Taq DNA polymerase (Promega, targeting salinity tolerance. Madison, WI, USA). The PCR cycling conditions were set at 958C for 3 min, followed by 35 cycles of 958C for 20 s, 61–648C (depending on primer specificity) for 30 s, Materials and methods 728C for 30 s and a final extension of 7 min at 728C. The PCR products were denatured at 998C for 10 min and We used 375 genotypes that represent the diversity pre- renatured initially at 708C for 20 s followed by 69 cycles sent in O. sativa, which were selected from the set geno- with a temperature decrease by 0.38C per cycle. Celery typed by single sequence repeats in the ‘Generation juice extract (CJE) was produced by the technique of Challenge Program’ with a population structure similar Till et al. (2004). The mismatch cleavage (CJE digestion to that of Garris et al. (2005). Deoxyribonucleic acid optimized for each primer) and EcoTILLING analysis in (DNA) was extracted from leaf tissue according to agarose gel were performed according to Raghavan Fulton et al. (1995). DNA from all samples was brought et al. (2007). to a concentration of 0.5 ng/ml. For EcoTILLING, DNA from each genotype was contrasted with either ‘IR64’ (against indica, admixed and Aus accessions) or Results and discussion ‘Nipponbare’ (against japonica, aromatic and admixed) separately, in a 1:1 ratio. Rice cultivars with different salt sensitivities have been Target genes were SalT (LOC_Os01g24710) and studied using transcriptomic approaches (Walia et al., OsCPK17 (LOC_Os07g06740). Primers were designed 2005, 2007; Kumari et al., 2009; Senadheera et al., (based on the ‘Nipponbare’ genome sequence) with 2009), revealing highly significant differences in gene Primer3 software to amplify part of an intron and the regulatory mechanisms between genotypes. A clear 30-untranslated region (UTR) of OsCPK17 (forward – example of the importance of genotypic differences AATTGGAGGTTGGGCCATAG; reverse – TGTGAGGTG- between varieties towards salt stress was given by the GAAGAAGCAAAC) and the second exon and the 30-UTR allelic differences found in OsHKT8 gene between two different genotypes (Ren et al., 2005). The six 12345678910M nucleotide substitutions in the coding region leading to four amino acid changes present in ‘Nona Bokra’ enhanced the overall Naþ transport activity (Ren et al., 2005), showing the importance of discovering superior alleles in salt stress-related genes. We are presently evaluating the haplotype groups of OsCPK17 and SalT genes by EcoTILLING. OsCPK17 encodes a calcium-dependent protein kinase, and its promoter contains cis elements responsive to various stress stimuli. Wan et al. (2007) showed that OsCPK17 is down-regulated by salt, cold and drought, indicating its importance in stress signal transduction in rice. The SalT gene was first isolated and characterized from the roots of rice plants treated with salt (Claes et al., 1990) and co- AB E AABBAAD localizes with the Saltol Quantitative Trait Loci. Although several studies have already been performed (Claes et al., Fig. 1. Analysis of the EcoTILLING digestion patterns in 1990; Zhang and Blumwald, 2001; de Souza et al., 2003), agarose gel for OsCPK17 in different rice varieties. The the function of SalT is not clearly understood. varieties shown were contrasted with ‘IR64’ and created four In our EcoTILLING assay, we amplified c. 1 kb targets haplotype groups (A, B, E and D) according to CJE-cleaved of each gene using specific primers. If the amplicons products. 1, PATNAI 23; 2, CODE NO. 31293; 3, TOS7564; 4, KHAO PON; 5, MALLIGAI (KOTTAMALLI differ in sequence content between the reference and SAMBA); 6, TUNGHWANPEI; 7, DE ABRIL; 8, LAGEADO; target germplasm, heteroduplex mismatch molecules 9, BALA; 10, SINNA SITHIRA KALI; M, molecular marker. will occur on re-annealing. Digestion with CJE 302

Table 1. Sequence and position of the SNPs and indels found in OsCPK17 and SalT genes

OsCPK17 position relative to the beginning of locus Os07g06740 Haplotype 4170 4181 4203 4238 4288 4417 4479 4547 4589 4593 4644 4699 4705 4731 4735

A C A (CT)7 C (CT)5 T T G (T)10 C (GT)4 AAAGA B C A (CT)9 C (CT)12 T C C (T)9 C (GT)4 AAAGA C C A (CT)8 C (CT)5 T T G (T)10 C (GT)4 AAAGA D C T (CT)7 A (CT)9 T T G (T)10 C (GT)4 AGGAA E C A (CT)5 C (CT)7 C C G (T)9 – GT GC (GT)2 AAAGA F C A (CT)8 C (CT)8 T C G (T)11 C (GT)4 GAAGT G C A (CT)9 C (CT)11 T C C (T)9 C (GT)4 AAAGA H A A (CT)10 C (CT)12 T C C (T)9 C (GT)4 AAAGA I C A (CT)9 C (CT)12 T C C (T)9 C (GT)3 AAAGA J C A (CT)7 C (CT)9 T T G (T)9 C (GT)4 AAAGA

SalT position relative to the beginning of Locus Os01g24710 Haplotype 1108 1159 1292 1307 1360 1365 1387 1391 1395 1408 1413 1425 1476 1480 1530 ATA(AT)2 TC ACAATGTTGT C T TTA C G- C BCA(AT)2 G T G T G GTTC– T C – A TA C CTC(AT)2 TC ACAATGTTGT C T – C G- C D C A AT G T G T G GTTC– T C – A TA T ECA(AT)2 G T G T G GTTC– T C – A TA C FCA(AT)2 G T G T G GTTC– T C – A TA C

Haplotype 1537 1550 1573 1683 1685 1711 1742 1803 AGAGGTTAG BGAAGACGA CAAGATTAG DGCAGACGA EGAAGACGA FGAAGACGG .Negra S. ˜ o tal. et Use of EcoTILLING in rice varieties 303 endonuclease reveals bands other than full-length Elisabeth B. Naredo (IRRI) for the laboratory support, products indicating SNPs or/and indels. By comparing and Isabel Abreu (IBMC, Portugal) for the protein the digestion patterns, haplotypes can be assigned, as analysis and help with the CJE production. Nelson shown in Fig. 1. The number, position and type of Saibo is also acknowledged for the critical review SNPs in the haplotypes were validated by sequencing of this manuscript. the PCR amplicons. The analysis of the EcoTLLING digestion patterns identified ten haplotypes for OsCPK17 and six haplotypes for SalT (Table 1). For OsCPK17, haplotypes A References and G (which include the reference types ‘IR64’ and ‘Nipponbare’, respectively) are more frequent, whereas Ali AJ, Xu JL, Ismail AM, Fu BY, Vijaykumar CHM, Gao YM, Domingo J, Maghirang R, Yu SB, Gregorio G, Yanaghihara E, F, I and J are very rare (each represented by one S, Cohen M, Carmen B, Mackill D and Li ZK (2006) Hidden accession). The observed mismatches were explained diversity for abiotic and biotic stress tolerances in the by 15 SNPs or indels with most of the indels being primary gene pool of rice revealed by a large backcross detected in repetitive sequences: (CT), (T) or (GT). breeding program. Field Crops Research 97: 66–76. All polymorphisms detected in OsCPK17 do not seem Claes B, Dekeyser R, Villarroel R, Vandenbulcke M, Bauw G, 0 Vanmontagu M and Caplan A (1990) Characterization of a to interfere with 3 -splice sites. As of now, we have rice gene showing organ-specific expression in response found 23 SNPs or indels in the SalT gene. Haplotype to salt stress and drought. Plant Cell 2: 19–27. A is the most frequent and includes both reference Comai L, Young K, Till BJ, Reynolds SH, Greene EA, Codomo CA, genotypes ‘IR64’ and ‘Nipponbare’. The transition T/C Enns LC, Johnson JE, Burtner C, Odden AR and Henikoff S (position 1108) is a silent mutation, whereas the (2004) Efficient discovery of DNA polymorphisms in natural populations by Ecotilling. Plant Journal 37: 778–786. transition A/C (position 1159) leads to a glutamate to de Souza GA, Ferreira BS, Dias JM, Queiroz KS, Branco AT, aspartate change. Since these two amino acids belong Bressan-Smith RE, Oliveira JG and Garcia AB (2003) to the same group, this change may have little or no Accumulation of SALT protein in rice plants as a response consequence on protein structure/function. The remain- to environmental stresses. Plant Science 164: 623–628. ing SNPs and indels were located in the 30-UTR region. Fulton TM, Chunwongse J and Tanksley SD (1995) Microprep protocol for extraction of DNA from tomato and other These variations may be of significance since it is herbaceous plants. 13: 0 Plant Molecular Biology Reporter now recognized that 3 -UTRs play crucial roles in 207–209. the post-transcriptional regulation of gene expression Garris AJ, Tai TH, Coburn J, Kresovich S and McCouch S (2005) through the modulation of nucleocytoplasmic messen- Genetic structure and diversity in Oryza sativa L. Genetics ger ribonucleic acid (mRNA) transport, regulation of 169: 1631–1638. Kumari S, Panjabi V, Kushwaha H, Sopory S, Singla-Pareek S and mRNA polyadenylation and translation, translation effi- Pareek A (2009) Transcriptome map for seedling stage ciency, sub-cellular localization and messenger stability specific salinity stress response indicates a specific set of (Mignone et al., 2005). genes as candidate for saline tolerance in Oryza sativa L. Our large-scale analysis is also targeting and analys- Functional and Integrative Genomics 9: 109–123. ing other regions of SalT and OsCPK17, as well as of Mignone F, Grillo G, Licciulli F, Iacono M, Liuni S, Kersey PJ, Duarte J, Saccone C and Pesole G (2005) UTRdb and UTR- other genes important in salinity tolerance. We will site: a collection of sequences and regulatory motifs of the further assess the relevance of the DNA variations by untranslated regions of eukaryotic mRNAs. Nucleic Acids testing association with salinity tolerance phenotypes. Research 33: D141–D146. Eventually, we hope to uncover novel superior alleles Raghavan C, Naredo MEB, Wang HH, Atienza G, Liu B, Qiu FL, that can be used in rice-breeding programmes for salt McNally KL and Leung H (2007) Rapid method for detecting SNPs on agarose gels and its application in candidate stress tolerance. gene mapping. Molecular Breeding 19: 87–101. Ren ZH, Gao JP, Li LG, Cai XL, Huang W, Chao DY, Zhu MZ, Wang ZY, Luan S and Lin HX (2005) A rice quantitative Acknowledgements trait locus for salt tolerance encodes a sodium transporter. Nature Genetics 37: 1141–1146. Senadheera P, Singh RK and Maathuis FJM (2009) Differentially This work was financially supported by Fundaçaõ para expressed membrane transporters in rice roots may con- a Cieˆncia e a Tecnologia (Portugal) through the project tribute to cultivar dependent salt tolerance. Journal of FCT- PTDC/AGR-GPL/70920/2006. Sońia Negraõ grate- Experimental Botany 60: 2553–2563. fully acknowledges FCT-Portugal for the financial Till BJ, Burtner C, Comai L and Henikoff S (2004) Mismatch clea- support (SFRH/BPD/34593/2007). All the plant material vage by single-strand specific nucleases. Nucleic Acids Research 32: 2632–2641. was kindly provided by the International Rice Gene- Walia H, Wilson C, Condamine P, Liu X, Ismail AM, Zeng LH, bank, located at the International Rice Research Wanamaker SI, Mandal J, Xu J, Cui XP and Close TJ Institute (IRRI, The Philippines). We also thank Ma. (2005) Comparative transcriptional profiling of two 304 S. Negraõ et al. contrasting rice genotypes under salinity stress during the Zeng LH (2005) Exploration of relationships between physio- vegetative growth stage. Plant Physiology 139: 822–835. logical parameters and growth performance of rice Walia H, Wilson C, Zeng LH, Ismail AM, Condamine P and Close (Oryza sativa L.) seedlings under salinity stress using TJ (2007) Genome-wide transcriptional analysis of salinity multivariate analysis. Plant and Soil 268: 51–59. stressed japonica and indica rice genotypes during panicle Zhang HX and Blumwald E (2001) Transgenic salt-tolerant initiation stage. Plant Molecular Biology 63: 609–623. tomato plants accumulate salt in foliage but not in fruit. Wan B, Lin Y and Mou T (2007) Expression of rice Ca2þ-depen- Nature Biotechnology 19: 765–768. dent protein kinases (CDPKs) genes under different Zhu JK (2001) Plant salt tolerance. Trends in Plant Science 6: environmental stresses. FEBS Letters 581: 1179–1189. 66–71. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 305–308 ISSN 1479-2621 doi:10.1017/S1479262111000414

Allele mining in the gene pool of wild Solanum species for homologues of late blight resistance gene RB/Rpi-blb1

Artem Pankin1*, Ekaterina Sokolova1, Elena Rogozina2, Maria Kuznetsova3, Kenneth Deahl4, Richard Jones4 and Emil Khavkin1 1Institute of Agricultural Biotechnology, Moscow, Russia, 2Institute of Plant Industry, St. Petersburg, Russia, 3Institute of Phytopathology, Bol’shiye Vyazemy, Russia and 4USDA-ARS, Agriculture Research Center, Beltsville, MD, USA

Abstract A coiled coil-nucleotide binding site-leucine-rich repeat gene RB/Rpi-blb1 isolated from Solanum bulbocastanum confers broad-spectrum resistance to Phytophthora infestans and is currently employed in potato breeding for durable late blight (LB) resistance. RB homologues were reported in several Solanum species; some of them retained defence function. Here, we report additional evidence on RB-like sequences in 21 Solanum species of the section Petota. The panel of Solanum species was screened with three RB-related PCR markers. RB-like sequences were found in every tested Solanum accession, suggesting universal distribution of RB structural homologues among Solanum genomes, while locus-speciﬁc RB-629 was found only in 15 species. Phylogenetic analysis of RB-629 sequences suggested a highly conservative pattern of polymorphisms that was neither species- nor series-speciﬁc. Apparently, duplication and evolution of RB-like loci preceded Solanum speciation. Marker presence and particular haplotypes were not immediately associated with high LB resistance.

Keywords: late blight resistance; Phytophthora infestans; potato; R genes; Solanum spp.

Introduction varieties, but resistance was supposedly defeated in the field by rapidly evolving P. infestans races (Fry, Late blight (LB, pathogen Phytophthora infestans) resist- 2008). Several genes of LB resistance were mapped on ance mediated by the R genes is one of the integral the linkage groups of various wild Solanum species elements of plant immune system (Dangl and Jones, (Hein et al., 2009). A cluster of four resistance gene 2001). Cultivated potato (Solanum tuberosum) lacks analogues (RGAs) located on chromosome 8 of S. bulbo- R genes active against P. infestans, apparently due to castanum was cloned and RGA2 (Rpi-blb1/RB) conferred the vegetative propagation excluding natural selection resistance to LB in both transient and stable expression for functional R loci under the recurrent pathogen attacks. systems (Song et al., 2003; van der Vossen et al., 2003). On the other hand, wild Solanum species inhabiting Potato transformation with RB homologues isolated regions with the most diverse populations of P. infestans from S. bulbocastanum (Rpi-bt1), S. stoloniferum (sensu acquired numerous R loci functional against LB and are Spooner et al., 2004; Rpi-sto1, Rpi-pta1) and S. verruco- essential genetic resources for potato breeding. sum (RB ver) confirmed specificity of these genes against A set of 11 R genes was identified in the Mexican broad spectrum of P. infestans races (Liu and Halterman, species S. demissum and introgressed into potato 2006; Vleeshouwers et al., 2008; Oosumi et al., 2009). However, recently, P. infestans races lacking Avr effectors compatible with RB ligand and thus virulent on potato plants transformed with RB have been identified * Corresponding author. E-mail: [email protected] (Champouret et al., 2009; Fo¨rch et al., 2010; Halterman 306 A. Pankin et al. et al., 2010). Pyramiding broad-spectrum resistance genes Lasergene 6.0 (DNAStar) and ExPASy Translate tool from various sources with different specificity to patho- (http://www.expasy.org). Cluster analysis was performed gen races in potato genome is probably a more effective using Maximum likelihood algorithm with 1000 bootstrap approach to durable LB resistance of potato cultivars replicates implemented in Phylip 3.69 (Felsenstein, (Tan et al., 2010). 1989). LB resistance was assessed using a modified In the present study, we followed an effective and detached leaf assay (Filippov et al., 2004; Kuznetsova efficient allele mining approach (Wang et al., 2008) to and Rogozina, unpublished data). analyse distribution and diversity of RB-like candidate resistance genes in germplasm of the wild Solanum species section Petota. Conservative patterns of polymorphisms were specific for paralogous RB-like loci Results and discussion rather than for Solanum species, thus suggesting that RB homologues duplicated and diverged before Solanum Based on their structural polymorphisms, the functionally speciation. active RB-like loci can be arranged into three distinct groups: RB-group (RB, Rpi-blb1, Rpi-sto1 and Rpi-pta1), RB ver-group and Rpi-bt1-group. It is noteworthy that Materials and methods exonic regions of RB-like loci retained over 90% homology, while introns diverged dramatically after dupli- Genomic DNA was isolated from 139 accessions cation of the RB-like loci. Apparently, these groups representing 21 wild Solanum species (Supplementary represent orthologous loci which emerged from different Table S1, available online only at http://journals.cambridge. RB-like paralogues duplicated in ancient Solanum org), using AxyPrepe Multisource Genomic DNA Mini- species and independently acquired defence function prep Kit. To amplify RB-like homologues, we designed against LB under the selective pressure of the pathogen universal RB-1223 and locus-specific RB-629 PCR primers invasion events after Solanum speciation. (Table 1 and Supplementary Fig. S1, available online only In order to investigate the distribution of RB-like genes at http://journals.cambridge.org) and optimised them in the wild Solanum germplasm, three sequence charac- using OligoCalc (Kibbe, 2007). We also modified the terised amplified region (SCAR) markers were designed: allele-specific PCR primers 1 and 10 recognising a func- RB-1223 tagging all three groups of RB-like loci, RB-629 tional allele of bulbocastanum RB (Colton et al., 2006; specific for RB-group only and allele-specific RB-226 RB-226) to increase reaction specificity. The amplification (Supplementary Fig. S1, available online only at reactions contained 1 mlof10£ PCR buffer, 100–150 ng http://journals.cambridge.org). Marker RB-1223 was of genomic DNA, 1 ml 2.5 mM dNTP, 10 pmol each of two used to screen 22 accessions representing 13 species primers, 1 U of either Pfu (cloning; Fermentas) or Taq (S. avilesii, S. bulbocastanum, S. ehrenbergii, S. demissum, (screening; Syntol) DNA polymerase and sterile water S. hjertingii, S. hougasii, S. iopetalum, S. microdontum, to a volume of 10 ml, and were run in an MJ PTC-200 ther- S. pinnatisectum, S. polyadenium, S. polytrichon, mocycler (Bio-Rad). PCR products were separated S. stenophyllidium, S. stoloniferum and S. verrucosum); this by electrophoresis in 1.5% (w/v) agarose and stained marker was universally present in every tested accession, with ethidium bromide. Amplified fragments were suggesting ubiquitous distribution of the RB-like loci in Sola- cloned using InsTAclonee and CloneJETe PCR Cloning num genomes. RB-1223 was present in several copies (one to Kits (Fermentas) and sequenced using BigDyew three discernable bands/accession) and greatly varied in size Terminator v3.1 Cycle Sequencing Kit and ABI 3730 (,800–1300 bp). Sequencing experiments revealed that DNA Analyzer (Applied Biosystems). DNA sequences polymorphic bands in various Solanum species corre- were analysed using BLAST 2.2.23 (Altschul et al., 1990), sponded to paralogous and orthologous RB-like loci.

Table 1. PCR primers for ampliﬁcation of the RB-related SCAR markers

Markers Primers PCR thermal proﬁles RB-1223 F 50-atggctgaagctttcattcaagttctg-30 3 min at 948C; 35 cycles 45 s at 948C, 45 s R50-caagtattgggaggactgaaaggt-30 at 658C, 1 min 20 s at 728C; 15 min at 728C RB-629 F 50-gaatcaaattatccaccccaacttttaaat-30 3 min at 948C; 35 cycles 45 s at 948C, 45 s R50-caagtattgggaggactgaaaggt-30 at 658C, 1 min 20 s at 728C; 15 min at 728C RB-226 F 50-cacgagtgcccttttctgac-30 3 min at 948C; 35 cycles 30 s at 948C, 30 s R50-ttcaattgtgttgcgcactag-30 at 608C, 1 min at 728C; 15 min at 728C Allele mining for the RB-like late blight resistance genes 307

Observed variation in size was mainly due to the polymorph- both in resistant and susceptible Solanum accessions, isms in introns (Pankin et al., unpublished data). including S. bulbocastanum, and therefore cannot be The panel of the 134 accessions of 19 Solanum species universally used to discern the active RB allele even in was screened with RB-629 and RB-226 markers. RB-629 S. bulbocastanum accessions. was present in 54% of accessions representing 15 species, RB-629 was cloned from 16 accessions representing 12 whereas allele-specific RB-226 was found only in 7% of Solanum species (Supplementary Table S1, available accessions from five species (Supplementary Table S1, online only at http://journals.cambridge.org). Phyloge- available online only at http://journals.cambridge.org). netic analysis of RB-629 sequences produced four distinct Our data suggest wider distribution of RB-group loci in clusters: cluster 1 of bulbocastanum-like haplotypes, Solanum germplasm than reported earlier (Wang et al., cluster 2 comprising pseudogenes except one pinnatisec- 2008; Lokossou et al., 2010). RB-226 was also found tum RB-629 (pnt2), cluster 3 specific for S. polytrichon

MR MS S MS Cluster 3 S hou hjt2_2 plt2 bst_1 hjt1 R S MS plt1 pld1_3 MS hjt2_5 hjt2_6 100 Cluster 2 S fen_2 MS R hjt2_1 60.9 MR pld2 pnt2

57.9 Cluster 4 94.3 jam2 51 R R pnt3_2

cph_2 S sph R

R pld1_2 cph_1 R 91.2 fen_1 hjt2_3 Cluster 1 MS S

+MS jam1 67.1

+R blb

MS R pnt1 pld1_1 ehr +MS

hjt2_4 pnt3_1 MS bst_2 sto dms R S +R R Series (Hawkes, 1990) - Demissa - Longipedicellata - Pinnatisecta - Polyadenia - Bulbocastana

Fig. 1. Phylogenetic analysis (maximum likelihood) of the RB fragments (RB-629). þ , presence of allele-specific RB-226. Resistance ranks are as follows: open symbols: S, susceptible, MS, moderately susceptible; closed symbols: MR, moderately resistant, R, resistant. Bootstrapping was performed with 1000 replicates, and values higher than 50% are shown at the nodes. Cluster 1, Rpi-blb/RB-like haplotypes; cluster 2, pseudogenes; cluster 3, S. polytrichon-specific haplotypes; cluster 4, other haplotypes. Sequence abbreviations with underscore tag are either allelic or homeologous variants of RB-629. For the list of abbreviations and sequences, refer to Supplementary Table S1 (available online only at http://journals. cambridge.org). 308 A. Pankin et al. and cluster 4 combining other RB-group sequences with Fry W (2008) Phytophthora infestans: the plant (and R gene) open reading frame (Fig. 1). The described pattern of destroyer. Molecular Plant Pathology 9: 385–402. Halterman DA, Chen Y, Sopee J, Berduo-Sandoval J and polymorphisms was neither species- nor series-specific; Sańchez-Pe´rez A (2010) Competition between Phytoph- thus the observed diversity of RB-group loci emerged thora infestans effectors leads to increased aggressiveness before Solanum speciation and probably is not linked on plants containing broad-spectrum late blight resistance. to allopolyploidisation in Solanum species. Apparently, PLoS ONE 5: e10536. each cluster combines allelic variants of RB orthologues, Hawkes JG (1990) The potato: evolution, biodiversity and whereas inter-cluster polymorphisms are indicative of genetic resources. London: Belhaven Press, 259 pp. Hein I, Birch P, Danan S, Lefebvre V, Achieng Odeny D, different RB loci. Despite the defence function against Gebhardt C, Trognitz F and Bryan G (2009) Progress in LB unequivocally demonstrated in complementation mapping and cloning qualitative and quantitative resistance experiments with RB genes, the presence and poly- against Phytophthora infestans in potato and its wild rela- morphisms of RB sequences in various Solanum species tives. Potato Research 52: 215–227. were not immediately associated with higher LB resist- Kibbe WA (2007) OligoCalc: an online oligonucleotide proper- ance. Redundant copies of RB-like paralogues apparently ties calculator. Nucleic Acids Research 35(2): W43–W46. Liu Z and Halterman D (2006) Identification and characteriz- serve as a backup pool essential to the adaptive evolution ation of RB-orthologous genes from the late blight resistant of R gene-related pathogen recognition when Solanum wild potato species Solanum verrucosum. Physiological species respond to novel races of pathogen. and Molecular Plant Pathology 69: 230–239. Lokossou AA, Rietman H, Wang M, Krenek P, van der Schoot H, Henken B, Hoekstra R, Vleeshouwers VG, van der Vossen EA, Visser RG, Jacobsen E and Vosman B (2010) Diversity, Acknowledgements distribution, and evolution of Solanum bulbocastanum late blight resistance genes. Molecular Plant-Microbe We thank all colleagues who generously provided Sola- Interactions 23: 1206–1216. Oosumi T, Rockhold D, Maccree M, Deahl K, McCue K and Bel- num germplasm used in this study and the anonymous knap W (2009) Gene Rpi-bt1 from Solanum bulbocasta- reviewer for constructive criticisms. The study was sup- num confers resistance to late blight in transgenic ported by the ISTC-USDA-ARS project 3714p. potatoes. American Journal of Potato Research 86: 456–465. Song J, Bradeen JM, Naess SK, Raasch JA, Wielgus SM, References Haberlach GT, Liu J, Kuang H, Austin-Phillips S, Buell CR, Helgeson JP and Jiang J (2003) Gene RB cloned from Solanum bulbocastanum confers broad spectrum resist- Altschul SF, Gish W, Miller W, Myers EW and Lipman DJ (1990) ance to potato late blight. Proceedings of the National Basic local alignment search tool. Journal of Molecular Academy of Sciences USA 100: 9128–9133. Biology 215: 403–410. Spooner DM, van den Berg RG, Rodrigues A, Bamberg JB, Champouret N, Bouwmeester K, Rietman H, van der Lee T, Hijmans RJ and Lara-Cabrera SI (2004) Wild Potatoes Maliepaard C, Heupink A, van de Vondervoort PJI, (Solanum section Petota; Solanaceae) of North and Jacobsen E, Visser RGF, van der Vossen EAG, Govers F Central America. Ann Arbor, MI: American Society of and Vleeshouwers VGAA (2009) Phytophthora infestans Plant Taxonomists. isolates lacking class I IpiO variants are virulent on Tan M, Hutten R, Visser R and van Eck H (2010) The effect Rpi-blb1 potato. Molecular Plant-Microbe Interactions 22: of pyramiding Phytophthora infestans resistance genes 1535–1545. R and R in potato. Theoretical and Applied Colton LM, Groza HI, Wielgus SM and Jiang J (2006) Marker- Pi-mcd1 Pi-ber assisted selection for the broad-spectrum potato late Genetics 121: 117–125. blight resistance conferred by gene RB derived from a van der Vossen E, Sikkema A, Hekkert BL, Gros J, Stevens P, wild potato species. Crop Science 46: 589–594. Muskens M, Wouters D, Pereira A, Stiekema W and Dangl JL and Jones JDG (2001) Plant pathogens and integrated Allefs S (2003) An ancient R gene from the wild potato defence responses to infection. Nature 411: 826–833. species Solanum bulbocastanum confers broad-spectrum Felsenstein J (1989) PHYLIP – phylogeny inference package resistance to Phytophthora infestans in cultivated potato (version 3.2). Cladistics 5: 164–166. and tomato. Plant Journal 36: 867–882. Filippov AV, Gurevich BI, Kozlovsky BE, Kuznetsova MA, Vleeshouwers VGAA, Rietman H, Krenek P, Champouret N, Rogozhin AN, Spiglazova SY, Smetanina TI and Smirnov Young C, Oh S-K, Wang M, Bouwmeester K, Vosman B, AN (2004) A rapid method for evaluation of partial potato Visser RGF, Jacobsen E, Govers F, Kamoun S and Vossen resistance to late blight and of aggressiveness of EAGV (2008) Effector genomics accelerates discovery and Phytophthora infestans isolates originating from different functional profiling of potato disease resistance and regions. Plant Breeding and Seed Science 50: 29–41. Phytophthora infestans avirulence genes. PLoS ONE 3: e2875. Fo¨rch M, van den Bosch T, van Bekkum P, Evenhuis B, Vossen J Wang M, Allefs S, van den Berg R, Vleeshouwers V, van der and Kessel G (2010) Monitoring the Dutch P. infestans Vossen E and Vosman B (2008) Allele mining in Solanum: population for virulence against new R genes. In: Schepers conserved homologues of Rpi-blb1 are identified in Sola- HTAM (ed.) PPO Special Report no. 14. Wageningen: DLO num stoloniferum. Theoretical and Applied Genetics 116: Foundation, pp. 45–50. 933–943. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 309–312 ISSN 1479-2621 doi:10.1017/S1479262111000347

SCAR markers of the R-genes and germplasm of wild Solanum species for breeding late blight-resistant potato cultivars

Ekaterina Sokolova1†, Artem Pankin1†, Maria Beketova1,2, Maria Kuznetsova3, Svetlana Spiglazova3, Elena Rogozina4, Isol’da Yashina2 and Emil Khavkin1* 1Institute of Agricultural Biotechnology, Moscow, Russia, 2Institute of Potato Husbandry, Korenevo, Russia, 3Institute of Phytopathology, Bol’shiye Vyazemy, Russia and 4Vavilov Institute of Plant Industry, St.-Petersburg, Russia

Abstract New races of Phytophthora infestans rapidly defeat potato late blight (LB) resistance based on Solanum demissum germplasm, and breeders search for new sources of durable LB resistance. We developed and verified six sequence characterized amplified region markers recognizing the race-specific genes R1 and R3 of S. demissum and the broad-spectrum resistance gene RB of S. bulbocastanum and the germplasms of these species and used them to screen 209 accessions of 21 wild Solanum species. In addition to S. demissum, homologues of R1 and R3 were found in several species of series Demissa, Longipedicellata and diploid Tuberosa; R3 homologues were also detected in S. bulbocastanum, S. cardiophyllum and S. ehrenbergii. The RB homologues were found in a wider range of Solanum species. The markers of R1 and R3 genes reliably discerned between germplasms of S. tuberosum ssp. tuberosum and wild sources of LB resistance. Following introgression, the species-specific markers of demissum and bulbocastanum germplasm were rapidly lost, whereas the markers of R1 and R3 genes lasted through several meiotic generations and were maintained at high frequencies in modern potato cultivars. The presence of these markers in demissoid potato cultivars was significantly associated with LB resistance, presuming that both genes contribute to overall defence response.

Keywords: allele mining; CC-NBS-LRR genes; late blight resistance; Solanum spp.

Introduction higher LB resistance than the genotypes lacking R-genes (for review see Stewart et al., 2003). The hopes to Late blight (LB) caused by Phytophthora infestans breed for durable LB resistance using broad-specificity (Mont.) de Bary is among the most devastating potato genes, such as RB of S. bulbocastanum Dunal, seem diseases. The resistance genes R1–R11 from the wild to vanish with the discovery of P. infestans races species Solanum demissum Lindl. previously deployed overcoming LB resistance in RB-transformed potato for breeding LB-resistant potato cultivars are rapidly (Champouret, 2010; Halterman et al., 2010; for early evi- defeated by new pathogen races; nevertheless, many dence of S. bulbocastanum susceptibility to P. infestans potato cultivars comprising these R-genes maintain see Budin, 2002). Another option to obtain varieties with durable resistance is pyramiding and stacking several R-genes from wild Solanum species other than S. demissum (Haverkort et al., 2009; Verzaux, 2010). * Corresponding author. E-mail: [email protected] Several already characterized R-genes for LB resistance †These authors contributed equally to this work. from diverse wild Solanum species encode coiled 310 E. Sokolova et al. coil-nucleotide binding site-leucine rich repeat kinases. the Institute of Potato Husbandry, Korenevo, Russia. Once such R-genes are characterized, their orthologues Microtubers and seeds of wild Solanum species were found in other species using an allele-mining approach obtained from the collections of the Institute of Plant (Hein et al., 2009) would help expand the range of Industry, Centre for Genetic Resources, the Netherlands candidate genes for breeding for LB resistance. and the United States Potato Genebank, NRSP-6, We developed and verified six sequence characterized Sturgeon Bay, WI. amplified region (SCAR) markers recognizing R1 and R3 Standard protocols were employed for genomic DNA of S. demissum and RB of S. bulbocastanum and the isolation from plant leaves, PCR analysis, and cloning germplasms of these species, and used them to screen and identifying genome fragments. Specific primers 21 Solanum species sect. Petota Dumort. The structural for SCAR markers (Supplementary Table S1, available homologues of all three genes were found far beyond online only at http://journals.cambridge.org) were desi- the taxa where they had been initially discovered. gned following multiple alignment of the prototype gene sequences and anonymous genome fragments with their structural homologues from the NCBI Genbank using the programs BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi) Materials and methods and Vector NTI Suite 8 package (Invitrogen). The data were processed with the Mann–Whitney U test (1947) Tubers of potato varieties came from the collections of using SPSS Statistics 17.0 software (http://www.spss. the Institute of Plant Industry, St. Petersburg, Russia and com). For phylogenetic analysis, Maximum Likelihood

Table 1. Frequencies of SCAR markers for the R-genes and germplasms of Solanum

Total number of accessions comprising the particular marker Total number Series and species of accessions R1-1205 R3-1380 Sdms-523 Ssto-449 RB-629 Sblb-509 Tetraploid Tuberosa: S. tuberosum 17000 0 0 0 ssp. tuberosa (Chilotanum varieties and old cultivars free of demissum and stoloniferum germplasm) Demissoid potato accessions 144 57 55 0 50 (of 60) 0 0 Diploid Tuberosa S. avilesii 2 0 0 ND 0 ND 0 S. berthaultii 6 1 0 ND 1 ND 0 S. microdontum 5 1 1 ND 0 ND 0 S. verrucosum 7 0 0 0 1 (of 3) 0 0 Bulbocastana S. bulbocastanum 23 0 6 0 0 11 (of 18) 15 (of 18) Demissa S. brachycarpum 2 110 1 0 0 S. demissum 39 16 6 13 (of 13) 24 (of 29) 4 (of 13) 0 S. hougasii 3 110 1 1 0 S. iopetalum 1 100 1 0 0 Longipedicellata S. fendleri 4 000 2 3 0 S. hjertingii 6 000 4 4 0 S. polytrichon 10110 4 8 0 S. papita 4 010 4 2 0 S. stoloniferum 48 9 8 0 31 (of 39) 17 (of 19) 0 Pinnatisecta/Cardiophylla S. brachistotrichum 3 000 0 3 0 S. cardiophyllum 6 020 0 3 0 S. ehrenbergii 13020 0 4 0 S. jamesii 10000 0 4 0 S. pinnatisectum 8 000 0 7 0 S. stenophyllidium 3 000 1 2 0 Polyadenia S. polyadenium 6 000 1 2 0 ND, not determined. SCAR markers for potato late blight resistance 311 trees were constructed using the software PHYLIP 3.2 75 hou plt (Felsenstein, 1989). The marker R1-1205 was modified 76 from R1-1400 (Gebhardt et al., 2004) for more consistent 67 dms* scoring. Other markers were developed in this laboratory. 78 sto 1 98 sto 2 R3 group 100 sto 3 Results and discussion 69 52 blb 1 100 cph 1 All markers reliably discriminated S. tuberosum L. ssp. cph 2 tuberosum represented by potato varieties free of sto 4 wild Solanum germplasm from wild Solanum species 100 tub 1 currently employed in potato introgression breeding 56 blb 2 (Table 1). The markers for demissum R1 and R3, R3-like group 59 tub 2 bulbocastanum RB and stoloniferum Ssto-449 (the latter tub 3 corresponds to a fragment flanking R1 in S. demissum) passed through as many as six to eight crosses into lyc I2 group the modern potato cultivars. When verified by cloning Fig. 1. Maximum Likelihood tree comparing the R3-1380 and sequencing, the marker R1-1205 in S. polytrichon sequences from wild Solanum species (in boxes) with the Rydb. and S. stoloniferum Schltdl. and Bouchet and the functional gene R3, gene I2 and R3 inactive homologues. marker R3-1380 in S. bulbocastanum, S. cardiophyllum Bootstrap values (in percent) are shown at the nodes. Sequences are listed with the numbers of Solanum accessions Lindl., S. hougasii Correll, S. polytrichon and S. stoloni- and NCBI Genbank nucleotide accessions (italicized). ferum shared 98–99% identity with the corresponding S. bulbocastanum: blb 1, PI275198,HM124855;blb2, regions in the prototype genes. Full-length homologues PI243508, HQ731036; S. cardiophyllum: cph 1, VIR24375, of R3 from S. bulbocastanum and S. stoloniferum HM124857; cph 2, VIR21301; S. demissum: dms*, the (GenBank accession nos. HQ731036 and HQ731037, functional gene R3a from AY849382; S. hougasii:hou,VIR 8818, HM124858; S. lycopersicum:lyc,I2, AF118127; respectively) shared 92 and 98% identity with the proto- S. polytrichon:plt,VIR24463,HM124859; S. stoloniferum: type R3a gene (AY849382) from S. demissum (Sokolova sto 1, VIR23652; sto 2, GLKS588, FJ175386; sto 3, PI365401, et al., in preparation). The phylogenetic analysis of HQ731037; sto 4, CGN18348, FJ175388; S. tuberosum: tub 1, the R3-1380 fragments from wild Solanum species I2GA-SH23-3, AY849384; tub 2, I2GA-SH-23-1, AY849383; sequenced in our laboratory as compared to those of tub 3, I2GA-SH194-2, AY849385. the demissum R3 gene, tomato I2 gene and non- functional R3-like genes from the NCBI Genbank to emphasise the possibility that R3-1380 also recog- demonstrated that the R3-1380 sequences belonged nises the R5-R11 sequences allelic to R3 (Hein et al., to the same cluster as R3a (Fig. 1), presuming that 2009). The marker RB-629 was found in over half of this marker apparently represented the active gene. the accessions representing all wild Solanum series Sblb-509 was maintained for at least three generations, under study (Table 1; for more details, see Pankin while Sdms-523 designed from the fragment of ribo- et al., this issue). somal internal transcribed spacer was lost already after The results of the marker analysis even when sup- two crosses. ported by sequencing do not immediately prove that Only the markers Sblb-509 and Sdms-523 were species- newly screened Solanum species comprise the promising specific. In addition to S. demissum and S. stoloniferum, orthologues of the R1, R3 and RB genes. Nonetheless, the markers R1-1205, R3-1380 and Ssto-449 were found such evidence helps focus on prospective targets for in other species of the series Demissa and Longipedi- wider and deeper allele mining and functional analysis. cellata and even beyond these series. The presence However, in one case, we can relate the presence of of several R-genes in S. stoloniferum was first shown the markers of demissum R1 and R3 to their function. by phytopathological methods (for review see Gru¨nwald Using the Mann—Whitney U test, we demonstrated and Flier, 2005), and recently R3 sequences were isolated that LB resistance of demissoid potato cultivars compris- from S. stoloniferum by allele mining (Champouret, ing the markers R1-1205 and R3-1380 significantly 2010). The presence of the R1-1400 marker in S. stoloni- exceeded the resistance of the demissoid cultivars ferum was confirmed by Gebhardt et al. (2004), and devoid of these markers. The difference was especially now we have verified it by sequencing. Of special inter- great when the cultivars comprising R1 and R3 were est is the presence of R3-1380 in S. bulbocastanum, compared with varieties free of wild Solanum germplasm S. cardiophyllum, S. ehrenbergii (Bitter) Rydb. and (Supplementary Table S2, available online only at http:// S. microdontum Bitter (Table 1). However, we want journals.cambridge.org). 312 E. Sokolova et al. Conclusions Gebhardt C, Ballvora A, Walkemeier B, Oberhagemann P and Schu¨ler K (2004) Assessing genetic potential in germplasm We developed and validated several markers of the collections of crop plants by marker-trait association: a case study for potatoes with quantitative variation of resistance R-genes and germplasm of S. bulbocastanum, S. demissum to late blight and maturity type. Molecular Breeding 13: and S. stoloniferum which discriminated between cultivated 93–102. S. tuberosum and many wild species. These markers can Gru¨nwald N and Flier WG (2005) The biology of Phytophthora be employed for germplasm characterization in genetic infestans at its center of origin. Annual Review of Phyto- collections and for monitoring the potato populations pathology 43: 171–190. segregating after crosses. The presence of the R1 and R3 Halterman DA, Chen Y, Sopee J, Berduo-Sandoval J and genes introgressed from S. demissum,asevidencedby Sańchez-Pe´rez A (2010) Competition between Phytophthora the marker analysis, significantly improved LB resistance infestans effectors leads to increased aggressiveness on plants containing broad-spectrum late blight resistance. PLoS of potato cultivars. ONE 5: e10536. Haverkort AJ, Struik PC, Visser RGF and Jacobsen E (2009) Applied biotechnology to combat late blight in potato Acknowledgements caused by Phytophthora infestans. Potato Research 52: 249–264. Hein I, Birch PRJ, Danan S, Lefebvre V, Odeny DA, Gebhardt C, The authors thank all colleagues who provided Solanum Trognitz F and Bryan GJ (2009) Progress in mapping accessions and the pedigree information. This study was and cloning qualitative and quantitative resistance against supported by the ISTC-USDA-ARS project 3714p. Phytophthora infestans in potato and its wild relatives. Potato Research 52: 215–227. Mann HB and Whitney DR (1947) On a test of whether one References of two random variables is stochastically larger than the other. Annals of Mathematical Statistics 18: 50–60. Stewart HE, Bradshaw JE and Pande B (2003) The effect of Budin KZ (2002) Genetic foci of Solanum species, Petota Dumort, resistant to Phytophthora infestans (Mont.) De the presence of R-genes for resistance to late blight Bary. Genetic Resources and Crop Evolution 49: 229–235. (Phytophthora infestans) of potato (Solanum tuberosum) Champouret N (2010) Functional genomics of Phytophthora on the underlying level of field resistance. Plant Pathology infestans effectors and Solanum resistance genes. PhD 52: 193–198. Thesis, Wageningen University, p. 162. Verzaux E (2010) Resistance and susceptibility to late blight Felsenstein J (1989) PHYLIP - Phylogeny Inference Package in Solanum: gene mapping, cloning and stacking. PhD (Version 3.2). Cladistics 5: 164–166. Thesis, Wageningen University, p. 144. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 313–316 ISSN 1479-2621 doi:10.1017/S1479262111000402

Exploitation of nuclear and cytoplasm variability in Hordeum chilense for wheat breeding

Cristina Rodrı´guez-Sua´rez1, Marıá J. Gimeńez1, Marıá C. Ramı´rez1, Azahara C. Martıń1,2, Natalia Gutierrez2, Carmen M. A´vila2, Antonio Martıń1 and Sergio G. Atienza1* 1Plant Breeding Department, IAS-CSIC, Apdo. 4084, E-14080, Co´rdoba, Spain and 2A´ rea Mejora y Biotecnologıá, IFAPA-Centro Alameda del Obispo, Co´rdoba, Spain

Abstract Hordeum chilense Roem. et Schultz. is a diploid wild barley native to Chile and Argentina. The high crossability of this species with other members of the Triticeae tribe promoted the development of the new species £ Tritordeum Ascherson et Graebner. Hexaploid tritordeum was developed from the hybrid derived from the cross between H. chilense (used as female parent) and durum wheat. The interest of H. chilense is based on the presence of traits potentially useful for wheat breeding, including high endosperm carotenoid content, septoria tritici blotch resistance and abiotic stress tolerance. Besides, the variability at cytoplasm level is also important in this species. The development of common wheat–H. chilense alloplasmic lines (nucleus from wheat and cytoplasm from H. chilense) results in fertile or male sterile genotypes, depending on the accession donating the cytoplasm. Furthermore, these alloplasmic lines constitute an ideal system for deepening our knowledge on nuclear–cytoplasm interactions. In conclusion, H. chilense is an interesting source of variability for wheat breeding.

Keywords: cytoplasm; Hordeum chilense; plant breeding; plasmon; wild relatives

Wild relatives for crop breeding: potential of (at both morphological and molecular levels) distributed nuclear variability in Hordeum chilense Roem. into two main groups plus an intermediate group, as et Schultz revealed by molecular markers (Vaz Patto et al., 2001; Castillo et al., 2010). The high compatibility of H. chilense The progressive narrowing of the genetic base in crops with the genomes of Triticum species gives rise to fertile (Tanksley and McCouch, 1997; Warburton et al., 2006) and stable amphiploids and allows the transfer of traits has promoted a renewed interest in wild relatives such to wheat (Martin et al., 1998), such as resistance to sep- as Hordeum vulgare ssp. spontaneum (Matus et al., toria tritici, abiotic stress tolerance or endosperm storage 2003; Inostroza et al., 2009) or other distant relatives proteins (Martin et al., 1999; Atienza et al., 2002), but such as Aegilops tauschii, donor of the D genome probably the main interest of this species is its potential of common wheat (van Ginkel and Ogbonnaya, 2007); for increasing carotenoid content (Alvarez et al., 1999; or H. chilense Roem. et Schultz. (Atienza et al., 2000; Atienza et al., 2004; Atienza et al., 2005a; Atienza et al., Atienza et al., 2005b; Martin et al., 2008b). The wild 2007b). The phytoene synthase 1 from H. chilense is a barley H. chilense shows a wide range of variation good candidate gene for the improvement of carotenoid content (Atienza et al., 2007a), and, therefore, the cloning and characterization of this gene offer new possibilities for wheat breeding (Rodriguez-Suarez et al., * Corresponding author. E-mail: [email protected] 2010). Similarly, the development of H. chilense durum 314 C. Rodrı´guez-Sua´rez et al. wheat chromosome substitution lines will be useful for The development of H. chilense–common wheat evaluating the substitution effect of durum wheat by alloplasmic lines gives rise to two types of lines: H. chilense genes for carotenoid content. The use of barley male-sterile when the line H1 is used as cytoplasm expressed sequence tag (EST) markers (Hagras et al.,2005a; donor, or fully fertile when other H. chilense lines are Hagras et al., 2005b; Nasuda et al., 2005) has proven used. Accordingly, a research line is being developed very useful for physical mapping in H. chilense (Atienza to investigate the potential of this new CMS source, et al., 2007c; Said and Cabrera, 2009; Cherif-Mouaki et al., designated msH1, to produce hybrid wheat. The male 2011). Besides, the development of the genetic linkage sterile line does not show any floral or developmental map using ESTs, conserved orthologous set (Bolot et al., abnormalities, but reduced height and delayed heading 2009) and H. chilense-specific diversity arrays technology (Martin et al., 2008b). Fertility restoration is obtained markers will allow the establishment of precise relation- when chromosome 6HchS from H. chilense line H1 is ships between H. chilense and related species genomes, added (Martin et al., 2008b). Further research allowed thus providing more efficient tools for the use of this wild the obtaining of a fertile euplasmic line carrying the barley in wheat breeding. translocation T6HchS·6DL (Martin et al., 2009). How- ever, a single dose of this translocation is insufficient for fertility restoration, which suggests the presence of Cytoplasm 3 nuclear variability in H. chilense– one or more inhibitors of fertility genes in chromosome wheat interactions 6DL (Martin et al., 2009). More recently, a highly fertile line with 42 chromosomes plus an extra acrocentric The nuclear genome has a predominating role for chromosome has been obtained (Martin et al., 2010), the inheritance of most plant traits Nevertheless, cyto- whose long arm is the 1HchS chromosome, as demon- plasmic factors and cytoplasm £ nucleus interactions strated by molecular markers and fluorescent in situ are also important and still largely unexplored. Genetic hybridization. It seems that this chromosome originated information of eukaryotic organisms is divided into a from a deletion of the distal part of chromosome 1HchL nuclear genome in the nucleus and organelle genomes and that the restorer gene is located on the retained (sometimes referred to as plasmon) in the cytoplasm. segment from the 1HchL (Martin et al., 2010). The diso- Since the cytoplasm is maternally inherited in Triticeae mic addition of this acrocentric chromosome is fully species (Kihara, 1951), the best way to investigate fertile and thus constitutes an additional source of nuclear–cytoplasm interactions is by developing allo- restoration for wheat hybrid production based on plasmic lines, i.e. lines with the same nucleus but msH1 system. cytoplasms from different species. On the other hand, fully fertile alloplasmic lines H. chilense–wheat alloplasmic lines have been devel- were also obtained (Atienza et al., 2007d). Preliminary oped by repeated substitution backcross as described evidence suggested that phenotypic and metabolic vari- by Kihara (1951). First, amphiploids H. chilense £ wheat ations in wheat are associated with different nuclear– are developed as described by Martin and Chapman cytoplasmic combinations (Atienza et al., 2007c; Atienza (1977). This step is essential since the hybrids between et al., 2008), including phenotypic traits such as height H. chilense and wheat are sterile while the amphiploids or quality traits like endosperm carotenoid content. are fertile. Backcrossing to the nucleus donor is repeated In other cases, the use of either wheat or H. chilense until H. chilense chromosomes are fully eliminated. After cytoplasm did not result in any phenotypic variation somatic chromosome counting, the cytoplasm origin has in Tritordeum (Atienza et al., 2007e). The genetic to be checked, since paternal inheritance of cytoplasm effects of the plasmon have been studied in several has also been reported (Soliman et al., 1987; Laser et al., species affecting different traits including yield (Loessl 1997; Aksyonova et al., 2005; Badaeva et al., 2006). et al., 2000), disease or pest resistance (Voluevich and Indeed, we have observed this phenomenon with both Buloichik, 1992; Matsui et al., 2002) and tolerance to H. chilense (Atienza et al., 2007d) and H. vulgare cyto- abiotic stresses (Uprety and Tomar, 1993; Shonnard plasms (Martin et al., 2008a) using the chloroplastic and Gepts, 1994; Zhang et al., 2003). Nevertheless, marker ccSSR4 (Chung and Staub, 2003). the most detailed studies have been performed in the Alloplasmic lines are very useful for elucidating plant Triticum–Aegilops complex (Tsunewaki et al., 1996, phylogeny and determining the genetic effect of different 2002; Tsunewaki, 2009) and in teosinte–maize combi- plasmons. Furthermore, since the discovery of cyto- nations (Allen, 2005). plasmic male sterility (CMS) in wheat (Kihara, 1951), Recently, parallel transcriptomic and metabolomic ana- breeders have been very interested in CMS systems, look- lyses have been carried out on three alloplasmic lines ing for a viable procedure for hybrid wheat production to investigate the effect of H. chilense, Ae. uniaristata (for a review, see Martin, 2009). and Ae. squarrosa cytoplasms on nuclear–cytoplasm Nuclear and cytoplasm variability in H. chilense 315 interaction with common wheat (Crosatti et al., 2010). Atienza SG, Satovic Z, Martin A and Martin LM (2005b) Genetic The gas chromatography-mass spectrometer metabolic diversity in Hordeum chilense Roem. et Schult. germplasm collection as determined by endosperm storage proteins. profiling of leaves revealed significant differences Genetic Resources and Crop Evolution 52: 127–135. between the alloplasmic lines and their euplasmic con- Atienza SG, Avila CM and Martin A (2007a) The development trol. Transcriptomic analyses using the Affimetrix 61k of a PCR-based marker for PSY1 from Hordeum chilense, wheat gene chip showed that more than 500 genes modi- a candidate gene for carotenoid content accumulation fied their behaviour in the H. chilense alloplasmic line in tritordeum seeds. Australian Journal of Agricultural compared with the euplasmic control (Crosatti et al., Research 58: 767–773. Atienza SG, Ballesteros J, Martin A and Hornero-Mendez D 2010). Most of them encoded for chloroplast/mitochon- (2007b) Genetic variability of carotenoid concentration and drion localized proteins. The simultaneous consideration degree of esterification among tritordeum ( £ Tritordeum of transcriptomic and metabolomic data underlined that Ascherson et Graebner) and durum wheat accessions. Journal the amino-acid biosynthetic pathways are highly depen- of Agricultural and Food Chemistry 55: 4244–4251. dent on the nuclear–cytoplasm interaction. Atienza SG, Martin AC and Martin A (2007c) Introgression of In conclusion, H. chilense is an interesting source wheat chromosome 2D or 5D into tritordeum leads to of variability for wheat breeding and the study of the free-threshing habit. Genome 50: 994–1000. Atienza SG, Martin AC, Ramirez MC, Martin A and Ballesteros J alloplasmic lines allows us to increase our understanding (2007d) Effects of Hordeum chilense cytoplasm on agro- of how nuclear and cytoplasmic genomes interact. Thus, nomic traits in common wheat. Plant Breeding 126: 5–8. this may open up new opportunities for plant improve- Atienza SG, Ramirez MC, Martin A and Ballesteros J (2007e) ment through cytoplasm modification. Effects of reciprocal crosses on agronomic performance of tritordeum. Russian Journal of Genetics 43: 865–868. Atienza SG, Martıń A, Pecchioni N, Platani C and Cattivelli L (2008) The nuclear–cytoplasmic interaction controls caro- Acknowledgements tenoid content in wheat. Euphytica 159: 325–331. Badaeva ED, Pershina LA and Bildanova LL (2006) Cytogenetic Our work in this area is supported by grants (to S. G. A.) analysis of alloplasmic recombinant lines (H. vulgare)–T. AGL2008-03720, and P09-AGR-4817 from Spanish Minis- aestivum unstable in fertility and viability. Russian Journal of Genetics 42: 140–149. try of Science and Innovation, Junta de Andalucıá and Bolot S, Abrouk M, Masood-Quraishi U, Stein N, Messing J, FEDER. C. R.-S. acknowledges financial support from Feuillet C and Salse J (2009) The inner circle of the cereal CSIC (JAE-Doc program). genomes. Current Opinion in Plant Biology 12: 119–125. Castillo A, Budak H, Martin AC, Dorado G, Borner A, Roder M and Hernandez P (2010) Interspecies and intergenus transferability of barley and wheat D-genome microsatellite References markers. Annals of Applied Biology 156: 347–356. Cherif-Mouaki S, Said M, Alvarez JB and Cabrera A (2011) Aksyonova E, Sinyavskaya M, Danilenko N, Pershina L, Sub-arm location of prolamin and EST-SSR loci on chromo- Nakamura C and Davydenko O (2005) Heteroplasmy and some 1Hch from Hordeum chilense. Euphytica. doi 10.1007/ paternally oriented shift of the organellar DNA composition s10681-010-0268-y. in barley–wheat hybrids during backcrosses with wheat Chung S-M and Staub JE (2003) The development and evalu- parents. Genome 48: 761–769. ation of consensus chloroplast primer pairs that possess Allen JO (2005) Effect of teosinte cytoplasmic genomes on highly variable sequence regions in a diverse array of maize phenotype. Genetics 169: 863–880. plant taxa. Theoretical and Applied Genetics 107: 757–767. Alvarez JB, Martin LM and Martin A (1999) Genetic variation for Crosatti C, Quansah L, Atienza SG, Mare C, Fait A and Cattivelli L carotenoid pigment content in the amphiploid Hordeum (2010) Exploitation of diversity in nuclear–cytoplasm chilense £ Triticum turgidum conv. durum. Plant Breeding interaction using alloplasmic wheat lines. 2nd International 118: 187–189. Symposium on Genomics of Plant Genetic Resources, Atienza SG, Gimenez MJ, Martin A and Martin LM (2000) Variability in monomeric prolamins in Hordeum chilense. Bologna, Italy, p. 86. Theoretical and Applied Genetics 101: 970–976. Hagras AAA, Kishii M, Sato K, Tanaka H and Tsujimoto H Atienza SG, Alvarez JB, Villegas AM, Gimenez MJ, Ramirez MC, (2005a) Extended application of barley EST markers for Martin A and Martin LM (2002) Variation for the low- the analysis of alien chromosomes added to wheat genetic molecular-weight glutenin subunits in a collection of background. Breeding Science 55: 335–341. Hordeum chilense. Euphytica 128: 269–277. Hagras AAA, Kishii M, Tanaka H, Sato K and Tsujimoto H Atienza SG, Ramirez CM, Hernandez P and Martin A (2004) (2005b) Genomic differentiation of Hordeum chilense Chromosomal location of genes for carotenoid pigments from H. vulgare as revealed by repetitive and EST in Hordeum chilense. Plant Breeding 123: 303–304. sequences. Genes and Genetic Systems 80: 147–159. Atienza SG, Avila CM, Ramirez MC and Martin A (2005a) Appli- Inostroza L, del Pozo A, Matus I, Castillo D, Hayes P, Machado S cation of near infrared reflectance spectroscopy to the and Corey A (2009) Association mapping of plant height, determination of carotenoid content in tritordeum for yield, and yield stability in recombinant chromosome breeding purposes. Australian Journal of Agricultural substitution lines (RCSLs) using Hordeum vulgare subsp. Research 56: 85–89. spontaneum as a source of donor alleles in a Hordeum 316 C. Rodrı´guez-Sua´rez et al. vulgare subsp. vulgare background. Molecular Breeding Nasuda S, Kikkawa Y, Ashida T, Rafiqul Islam AKM, Sato K and 23: 365–376. Endo TR (2005) Chromosomal assignment and deletion Kihara H (1951) Substitution of nucleus and its effects on mapping of barley EST markers. Genes and Genetic genome manifestations. Cytologia 16: 177–193. Systems 80: 357–366. Laser B, Mohr S, Odenbach W, Oettler G and Ku¨ck U (1997) Rodriguez-Suarez C, Gimenez MJ and Atienza SG (2010) Parental and novel copies of the mitochondrial orf25 Progress and perspectives for carotenoid accumulation gene in the hybrid crop-plant triticale: predominant tran- in selected Triticeae species. Crop Pasture Science 61: scriptional expression of the maternal gene copy. Current 743–751. Genetics 32: 337–347. Said M and Cabrera A (2009) A physical map of chromosome Loessl A, Goetz M, Braun A and Wenzel G (2000) Molecular 4Hch from H. chilense containing SSR, STS and EST-SSR markers for cytoplasm in potato: male sterility and contri- molecular markers. Euphytica 167: 253–259. bution of different plastid–mitochondrial configurations Shonnard GC and Gepts P (1994) Genetics of heat tolerance to starch production. Euphytica 116: 221–230. during reproductive development of common bean. Crop Martin AC (2009) Desarrollo de un nuevo sistema de andro- Science 34: 1168–1175. esterilidad en trigo (Triticum aestivum L.) utilizando la Soliman K, Fedak G and Allard RW (1987) Inheritance of orga- especie Hordeum chilense Roem. et Schultz. PhD Thesis, nelle DNA in barley and Hordeum £ Secale intergeneric University of Co´rdoba, Spain. hybrids. Genome 29: 867–872. Martin A and Chapman V (1977) A hybrid between Hordeum Tanksley SD and McCouch SR (1997) Seed banks and molecular chilense and Triticum aestivum. Cereal Research Com- maps: unlocking genetic potential from the wild. Science munications 5: 365–368. 277: 1063–1066. Martin A, Martıń LM, Cabrera A, Ramı´rez MC, Gimeńez MJ, Tsunewaki K (2009) Plasmon analysis in the Triticum–Aegilops Rubiales P, Hernańdez P and Ballesteros J (1998) The complex. Breeding Science 59: 455–470. potential of Hordeum chilense in breeding Triticeae Tsunewaki K, Wang G-Z and Matsuoka Y (1996) Plasmon species. In: Jaradat AA (ed.) Triticeae III. Enfield, CT: analysis of Triticum (wheat) and Aegilops. 1. Production Science Publications, pp. 377–386. of alloplasmic common whets and their fertilities. Genes Martin A, Alvarez JB, Martin LM, Barro F and Ballesteros J (1999) and Genetic Systems 71: 293–311. The development of tritordeum: a novel cereal for food Tsunewaki K, Wang G-Z and Matsuoka Y (2002) Plasmon anal- processing. Journal of Cereal Science 30: 85–95. ysis of Triticum (wheat) and Aegilops. 2. Characterization Martin AC, Atienza SG and Barro F (2008a) Use of ccSSR markers and classification of 47 plasmons based on their effects for the determination of the purity of alloplasmic wheat in on common wheat phenotype. Genes and Genetic Systems different Hordeum cytoplasms. Plant Breeding 127: 77: 409–427. 470–475. Uprety DC and Tomar VK (1993) Photosynthesis and drought Martin AC, Atienza SG, Ramirez MC, Barro F and Martin A resistance of Brassica carinata and its parent species. (2008b) Male fertility restoration of wheat in Hordeum Photosynthetica 29: 321–327. chilense cytoplasm is associated with 6H(ch)S chromosome van Ginkel M and Ogbonnaya F (2007) Novel genetic diversity addition. Australian Journal of Agricultural Research 59: from synthetic wheats in breeding cultivars for changing 206–213. production conditions. Field Crops Research 104: 86–94. Martin AC, Atienza SG, Ramirez MC, Barro F and Martin A (2009) Vaz Patto MC, Aardse A, Buntjer J, Rubiales D, Martin A Chromosome engineering in wheat to restore male fertility and Niks RE (2001) Morphology and AFLP markers in the msH1 CMS system. Molecular Breeding 24: 397–408. suggest three Hordeum chilense ecotypes that differ in Martin AC, Atienza SG, Ramı´rez M, Barro F and Martin A (2010) avoidance to rust fungi. Canadian Journal of Botany 79: Molecular and cytological characterization of an extra acro- 204–213. centric chromosome that restores male fertility of wheat in Voluevich EA and Buloichik AA (1992) Nuclear-cytoplasmic the msH1 CMS system. Theoretical and Applied Genetics interactions in wheat resistance to fungi pathogens: 121: 237–240. V. Quantitative resistance of alloplasmic line seedlings of Matsui K, Yoshida M, Ban T, Komatsuda T and Kawada N (2002) Penjamo 672 variety to powdery mildew. Genetika 28: Role of male-sterile cytoplasm in resistance to barley 82–88. yellow mosaic virus and Fusarium head blight in barley. Warburton M, Crossa J, Franco J, Kazi M, Trethowan R, Rajaram Plant Breeding 121: 237–240. S, Pfeiffer W, Zhang P, Dreisigacker S and Ginkel M (2006) Matus I, Corey A, Filichkin T, Hayes PM, Vales MI, Kling J, Bringing wild relatives back into the family: recovering Riera-Lizarazu O, Sato K, Powell W and Waugh R (2003) genetic diversity in CIMMYT improved wheat germplasm. Development and characterization of recombinant Euphytica 149: 289–301. chromosome substitution lines (RCSLs) using Hordeum Zhang A, Yu F and Zhang F (2003) Alien cytoplasm effects on vulgare subsp. spontaneum as a source of donor alleles in a phytosiderophore release in two spring wheats (Triticum Hordeum vulgare subsp. vulgare background. Genome 46: aestivum L.). Genetic Resources and Crop Evolution 50: 1010–1023. 767–772. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 317–320 ISSN 1479-2621 doi:10.1017/S1479262111000633

Improvement of crop protection against greenbug using the worldwide sorghum germplasm collection and genomics-based approaches

Yinghua Huang1,2* 1USDA-ARS Plant Science Research Laboratory, 1301 N. Western Road, Stillwater, OK 74075, USA and 2Department of Botany, Oklahoma State University, Stillwater, OK 74078, USA

Abstract Successful development of new sorghum cultivars and hybrids to ensure sustainable production depends largely on the availability of genetic resources with desirable traits such as pest resistance. Our recent research has focused on improvement of crop protection against greenbugs using the worldwide germplasm collection and genomics-based approaches. First, we conducted the systematic evaluation of a worldwide germplasm collection in order to identify new sources of greenbug resistance. Twenty-one resistant lines were identiﬁed, which offered new sources of resistance to sorghum breeding. Molecular markers used to assess the genetic diversity among those resistant lines suggested relatively diverse resistant sources in the sorghum germplasm collection. More recently, a mapping project was executed to associate the resistance genes with sorghum chromosomes. The mapping data indicated one major and a minor quantitative trait loci reside on chromosome 9 and are responsible for resistance to greenbug. In addition, cDNA microarrays were used to monitor greenbug-induced gene expression in sorghum plants. This study has developed a transcriptional proﬁle for sorghum in response to greenbug attack, which provides us with useful molecular information for discovery of greenbug resistance genes and a better understanding of the genetic mechanisms controlling host defences in sorghum.

Keywords: DNA microarray; genomics; greenbug aphid; molecular marker; sorghum

Introduction for bioethanol production in the USA, Australia and other developed countries. As with most crops, sorghum Sorghum (Sorghum bicolor) is a leading crop grown in is the host to many pest insects. Greenbug (Schizaphis more than 100 countries (Frederiksen and Odvody, graminum) is one of the major pests of sorghum as 2000; U.S. Grains Council, 2006). Grain sorghum is an well as wheat and barley worldwide (Hackerott et al., important staple food crop in Africa, South Asia and 1969; Andrews et al., 1993). For example, damage by Central America. Sorghum is grown as animal feed and greenbug to sorghum is estimated to cost US sorghum producers $248 million annually (INTSORMIL, 2006). One of the most effective and environmentally sound insect pest management approaches is the use of genetically resistant cultivars and hybrids as the core * Corresponding author. E-mail: [email protected] component of an integrated pest management pro- Mention of a trademark or proprietary product does not constitute a gramme. Several examples of successful deployment of guarantee or warranty of a product by the U.S. Department of Agriculture and does not imply its approval to the exclusion of other resistant cultivars can be cited, but these examples are products that may also be suitable. oftentimes short-lived due to the tremendous diversity 318 Y. Huang of biotypes in the target pest populations and evolution represents the great wealth of sorghum genetics of their virulence. New sources of resistance to these resources and is invaluable to both sorghum research key pests must be found continuously and incorporated and various crop improvement programmes. However, into high-performance breeding lines for cultivar/hybrid those genetic materials are of little value unless they are development. Host resistance mechanisms must be evaluated, documented and utilized. Thus, one of our characterized, and resistance genes/quantitative trait loci current research emphases has been the systematic evalu- (QTLs) and molecular markers linked to these traits ation of the entire US collection of sorghum germplasm need to be identified in order to assure movement into to identify new sources of greenbug resistance, and to commercial cultivars and hybrids through various breed- subsequently characterize the newly identified sources ing methods and to enable more efficient crop breeding. to facilitate breeding sorghum with insect resistance. Furthermore, plant genomics has proven to be a promis- In this project, these sorghum germplasm accessions ing new means for crop improvement. This paper takes from locations around the world were evaluated for sorghum–greenbug as the example to demonstrate the their response to greenbug feeding in the greenhouse current success and future potential in improvement of at the USDA-ARS Plant Science Research Laboratory, Still- crop protection against greenbugs using worldwide water, Oklahoma. As a result, 21 germplasm accessions germplasm collection and genomics-based approaches. were identified that possessed resistance to the greenbug, as shown in Table 1. When compared with susceptible checks, the newly identified accessions had relatively Evaluation of sorghum germplasm for new sources low greenbug damage scores, indicating their inherent of resistance to greenbugs resistance to greenbug. Among these resistant sources, four accessions showed high levels of resistance ranging Continuous improvement in crop defence is dependent from 1.0 to 2.8 on a 1 to 6 scale (1 ¼ no damage; on the availability of diverse genetic resources and 6 ¼ 100% damage). The other 17 accessions demon- judicious use of effective sources of resistance. At strated moderate resistance with resistance ratings from present, over 40,000 sorghum germplasm accessions 2.9 to 3.8. These sorghum accessions are resistant to have been captured and conserved in the USDA National greenbug biotype I and offer new resistance sources to Plant Germplasm System (NPGS, 2010). This collection sorghum breeding programmes.

Table 1. Greenbug damage scores of the new sources of sorghum germplasm lines

No. Seed ID PI number Origin Damage ratinga 1 1072 221719 South Africa 3.5 ^ 0.5bcdb 2 2-969 452752 Ethiopia 3.0 ^ 1.1de 3 3-416 455203 Ethiopia 3.2 ^ 1.2bcde 4 3-722 455512 Ethiopia 3.3 ^ 1.0cd 5 3-1022 455812 Ethiopia 3.8 ^ 1.0bcd 6 3-1698 456490 Ethiopia 3.5 ^ 1.7bcd 7 3-1711 456504 Ethiopia 3.4 ^ 1.0bcd 8 3-2419 457212 Ethiopia 3.7 ^ 0.8bcd 9 3-2513 457314 Ethiopia 3.1 ^ 1.1cde 10 4-2683 482903 Zimbabwe 3.5 ^ 1.3bcd 11 5-462 500963 Zambia 3.5 ^ 0.6bcd 12 6-384 515888 Togo 3.6 ^ 1.4bcd 13 7-855 535779 USA 2.0 ^ 0.7efg 14 7-943 536594 Honduras 3.3 ^ 1.0cd 15 7-1859 545501 Sudan 2.8 ^ 0.5de 16 7-2535 560387 South Africa 3.8 ^ 0.9bcd 17 8-17 562891 India 3.5 ^ 1.1bcd 18 9-99 585393 Nigeria 3.3 ^ 0.7bcde 19 9-2983 591008 USA 3.8 ^ 1.6bcd 20 10-562 596542 USA 2.6 ^ 1.8def 21 10-1231 607900 USA 1.1 ^ 0.1g 22 Resistant 550607 China 2.0 ^ 0.4efg 23 Susceptible Westland A line USA 5.8 ^ 0.2a a Based on greenbug damage to the sorghum seedlings, scores were recorded on a 1–6 scale, where 1 ¼ # 20%, 2 ¼ 20–40%, 3 ¼ 40–60%, 4 ¼ 60–80%, 5 $ 80% damage and 6 ¼ plant death. b Means with the same letter are not significantly different at probability level of 0.05. Crop protection against greenbug 319 Gene discovery through gene expression profiling coordinately modulated by versatile molecular regulators such as salicylic acid, jasmonic acid, abscisic acid and Identifying what genes are expressed in a particular phytohormones (Park et al., 2006). tissue or at a specific time is often very useful to deter- Although the application of microarray technology to mine their function. Recently, genomics techniques the analysis of plant defence responses is still a very such as DNA microarrays, large-scale gene expression recent approach (Zhu-Salzman et al., 2004; Park et al., profiling (transcriptome) and associated bioinformatics 2006), the initial microarray experiments on studies of analysis make it possible to monitor expression of all plant responses to attack by greenbugs have shown the genes simultaneously in any given tissue or following a promise for functional characterization of important pro- specific treatment (Alba et al., 2004). cesses such as plant defence against aphids. Furthermore, Using cDNA microarrays, we have recently examined these expression profiling studies completed to date have the transcriptional changes in sorghum seedlings and already identified an amazing number of genes that had compared the results from parallel systems, greenbug- not previously been implicated in plant defences against resistant and -susceptible genotypes, leading to the insects. Large sets of informative data were generated that detection of differentially expressed transcripts during led to the identification of many potential defence-related infestation by greenbug biotype I, corresponding to 157 transcripts. The ability to identify changes in gene sorghum genes. The experiments showed comprehen- expression, particularly in response to disease, insect sive gene activities resulted from up-regulating or activat- pest or abiotic stresses is significant since exploration of ing existing defence pathways in sorghum seedlings in gene activation in plant defence on a genome-wide response to greenbug feeding. Of the aphid-induced scale could be important in discovering novel defence genes identified in this study, 38 genes exhibited three- genes or strategies for crop improvement. fold or higher abundance in their expression and 26 genes were significantly reduced. For further analysis, the genes that showed differential expression were Development and application of molecular markers cloned and sequenced. The resulting sequences were in study of plant defence then annotated by comparison with GenBank databases using the BLASTX search program. Sequence similarity The development of DNA markers has facilitated the con- searches allowed putative functions to be assigned to struction of genetic maps for an economically important 16 cDNA clones/genes (Table 2) that were directly or plant species. A genetic map depicts the linear arrange- indirectly involved in host defence against greenbug ment of DNA markers along each chromosome and the attack. Our detailed studies also suggested that the genetic distances between adjacent markers. Once the defence responses against greenbug in sorghum are genetic maps are constructed, they can facilitate practical

Table 2. Identiﬁcation of the defence or defence-related genes that differentially expressed in sorghum seedlings in response to greenbug attack

cDNA IDa Gene product – putative biological functionb Fold changesc DR831455 Sulphur-rich/thionin-like protein 13.251 DR831570 b-Glucosidase 3.899 DR831456 Glucan endo-1,3-b-glucanase 2.218 DR831457 S-like RNase 1.801 DR831459 Cysteine proteinase inhibitor 3.324 DR831458 Cysteine proteinase 2.652 DR831460 Polyphenol oxidase 3.573 DR831462 Legumain-like protease 2.105 DR831463 Endo-1,4-b-glucanase 2.621 DR831464 Wound inductive gene 2.621 DR831465 Multiple stress-responsive zinc-finger protein 2.428 DR831466 Oxysterol-binding protein 2.135 DR831467 Cytochrome P450-like protein 2.757 DR831468 Cytochrome P450 monooxygenase 22.253 DR831470 Xa1-like protein 2.39 DR831471 OTU-like cystein domain-containing protein 21.066 a GenBank accession number. b BLASTX research was used to determine homologous genes and putative functions. c Values of signal intensity ratios showing up- or down-regulation. 320 Y. Huang applications in plant breeding such as the identification Dr. Sung-Jin Park and Dr. Yanqi Wu for their contri- of important agronomic traits (in many cases QTLs) and butions to the early phase of the research projects. marker-assisted selection. Genetic maps are also an important resource for plant gene isolation through map-based cloning and potential for modification without affecting other important traits. References Recently, simple sequence repeats (SSRs) or microsatellites have become the most important DNA marker tech- Alba R, Fei Z, Payton P, Liu Y, Moore SL, Debbie P, Cohn J, nology as they proved to be a more dependable, rapid D’Ascenzo M, Gordon JS, Rose JKC, Martin G, Tanksley and inexpensive tool for plant genotyping (Yang et al., SD, Bouzayen M, Jahn MM and Giovannoni J (2004) 1996). We have recently constructed a detailed SSR-based ESTs, cDNA microarrays, and gene expression profiling: tools for dissecting plant physiology and development. genetic map for sorghum (Wu and Huang, 2007). Using Plant J. 39: 697–714. the genetic map and SSR markers, we were able to locate Andrews DJ, Bramel-Cox PJ and Wilde GE (1993) New sources one major and a minor QTL on chromosome 9 that are of resistance to greenbug, biotype I in sorghum. Crop responsible for resistance to greenbug (Wu and Huang, Science 33: 198–199. 2008). The major resistance QTL on chromosome 9, desig- Frederiksen RA and Odvody GN (2000) Compendium of nated as QSsgr-09-01, was found to reside in the interval of Sorghum Diseases. APS Press: St. Paul, MN, pp. 1–4. Hackerott HL, Harvey TL and Ross WM (1969) Greenbug resist- 7.3 cM between Xtxp289 and Xtxp358, on the basis of ance in sorghums. Crop Science 9: 656–658. linkage and QTL analyses. QSsgr-09-01 accounted for Huang Y (2008) Development of EST-SSR markers for sorghum 54.5–80.3% of the phenotypic variation in genetic resist- and their transferability among cereal species In: Proceed- ance to greenbug biotype I. The second QTL identified ings of International Plant and Animal Genome Confer- on SBI-09 was flanked by Xtxp67 and Xtxp230, and desig- ence. January 12–16, 2008, San Diego, CA, p. 148 INTSORMIL (2006) Pest and disease resistant hybrids for U.S. nated as QSsgr-09-02. QSsgr-09-02 explained 1.3–5.9% producers INTSORMIL Report no. 4, August 1, 2006. of the resistance variation. With the rapidly increasing NPGS (2010) http://www.ars-grin.gov/npgs/searchgrin.html availability of cDNA clones and expressed sequence tags Park SJ, Huang Y and Ayoubi P (2006) Identification of (ESTs), we took the in silico mining approach to develop expression profiles of sorghum genes in response to green- EST-SSRs. From the available 25,456 ESTs or cDNA bug phloem-feeding using cDNA subtraction and microar- sequences, we have developed 2680 EST-SSRs (Huang, ray analysis. Planta 223: 932–947. U.S. Grains Council (2006) Building global markets for Ameri- 2008). These newly developed sorghum EST-SSR markers ca’s grains http://www.grains.org/index.ww. Accessed represent an additional resource for genetic mapping, com- August 22, 2006 parative genomics and evaluation of co-location between Wu Y and Huang Y (2007) An SSR genetic map of Sorghum QTLs and functionally associated markers in sorghum. bicolor (L.) Moench and its comparison to a published Among these newly identified markers, 200 selected genetic map. Genome 50: 84–89. EST-SSR markers were examined for transferability to related Wu Y and Huang Y (2008) Molecular mapping of QTLs for resistance to the greenbug Schizaphis graminum (Ron- cereal crops and the results showed their potential as mol- dani) in Sorghum bicolor (Moench). Theoretical and ecular markers in maize, sugarcane, rice, wheat and barley. Applied Genetics 117: 117–124. Yang W, de Oliveira AC, Godwin I, Schertz K and Bennetzen JL (1996) Comparison of DNA marker technologies in characterizing plant genome diversity: variability in Chinese Acknowledgements sorghums. Crop Science 36: 1669–1676. Zhu-Salzman K, Salzman RA, Ahn J-E and Koiwa H (2004) Tran- scriptional regulation of sorghum defence determinants The author gratefully acknowledges Angie Phillips for against a phloem-feeding aphid. Plant Physiology 134: her excellent technical assistance. The author also thanks 420–431. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 321–323 ISSN 1479-2621 doi:10.1017/S1479262111000219

An overlooked cause of seed degradation and its implications in the efﬁcient exploitation of plant genetic resources

Dionysia A. Fasoula* Agricultural Research Institute, PO Box 22016, 1516 Nicosia, Cyprus

Abstract The importance of plant genetic resources for the future of agricultural production and for achieving food security necessitates the study of the factors affecting their most efﬁcient exploitation, particularly in breeding programmes. The established negative correlation between competitive and yielding ability is emerging as an important, yet overlooked cause of seed and variety degradation. Because of this negative correlation, the low yielding, strong competing plants within the variety or the germplasm under study acquire a survival advantage over the high yielding, weak competing plants when propagated under dense stands, leading to a gradual cultivar degeneration and identity loss. Moreover, this gradual degeneration prevents selecting for the positive and novel adaptive variation that is endlessly released by the genome in response to biotic (e.g. mutating pathogens) and abiotic stresses. The application of nonstop selection on individual plants grown in the absence of competition using a novel selection equation demonstrates an effective means to counteract the negative effects and accelerate progress through selection.

Keywords: cereals; competition; moving complete blocks

Introduction of high yielding, weak competing ones (Yc), when seeds are propagated under dense stands. The efﬁcient exploitation of plant genetic resources is of utmost concern for agricultural production and for societies in general, as they represent the raw material of future elite cultivars and serve as an indicator of Materials and methods agricultural sustainability (Gepts, 2006). Seeds are the primary means of delivering the potential of plant genetic The starting material for selection was breeder’s seed of resources; thus, maintaining seed quality is of utmost a well-known local barley variety, cv. Athenaida, long importance. Seed quality is commonly referred to as maintained as a pure line and registered in the Cyprus the avoidance of the physiological causes affecting seed national variety catalogue. In 2007–08, a nonreplicated vigour and germination. This paper highlights the honeycomb trial (Fasoulas and Fasoula, 1995) of 600 existence of and the means to deal with an overlooked plants was established at the Athalassa experimental genetic cause of seed degradation due to the hidden station, with 1 m plant-to-plant distance, excluding inter- negative correlation between yielding and competitive plant competition. Standard agronomic practices were ability, which favours the gradual proliferation of low followed during the growing season. In 2008–09, seeds yielding, strong competing plants (yC) at the expense from six superior plants, representing six entries (families), along with seed from the control (original seed of cv. Athenaida) as the seventh, were grown in a replicated R7 honeycomb trial (Fasoulas and Fasoula, * Corresponding author. E-mail: [email protected] 1995) with 50 replications/family, summing up to a total 322 D. A. Fasoula of 350 plants. Seed from the ten best plants from each L3 of the two superior families, selected on the basis of 30 selection equation A (Fasoula, 2008), were advanced as 20 lines in 2009–10 and tested in a randomized complete H1 block (RCB) trial. 10 H1 ¼ð Þ2 £ ð Þ2 Equation A x=xr x=s ; 0

to Siete Cerros H1 where x is the yield in grams/plant and xr the average –10 yield of a representative sample of surrounding L3 Percent yield gain compared Percent plants (Supplementary Fig. S1, available online only –20 L3 at http://journals.cambridge.org). These plants, forming circular moving complete replicates, serve as a Absence of Mono Mixed competition culture culture common denominator that erases the masking effect of soil heterogeneity on single plant yields, while x and Fig. 1. Percent yield of the derived lines H1 and L3 s are the mean and standard deviation of the entry to compared to the original Siete Cerros in three conditions: (1) absence of competition, (2) monoculture and (3) mixed which each plant belongs. The stability parameter culture with Siete Cerros. Line H1, selected as a Yc, outper- 2 ðx=sÞ , termed coefficient of homeostasis (Fasoula, forms both Siete Cerros and L3 under conditions 1 and 2. 2008), converts the plant yield potential into crop yield Conversely, line L3, selected as a yC, outperforms Siete potential (Supplementary Fig. S2, available online only at Cerros and H1 under condition 3 only (based on data from http://journals.cambridge.org). The plants in the 2009–10 Fasoula, 1990). RCB trials involving the two higher yielding of the six lines on the basis of equation A, along with the control (original inability to efficiently select for yield under dense stand seed of cv. Athenaida), were grown in two locations and the gradual cultivar degeneration. (Athalassa and Zygi), in a four-row plots of 4 m long, with With competitive ability defined as the ability to gain a four replications. The two central rows of each plot growth advantage over other plants by interfering with were harvested. the equal sharing of growth resources, the negative correlation between yielding and competitive ability represents both an invisible cause of cultivar degradation Results and discussion and the principal reason hindering the response to selection for yield on a single plant basis under competition The negative correlation between yielding and (Fasoula and Fasoula, 1997). This helps to distinguish competitive ability between the undesirable competitive ability and the desirable attribute of the sometimes called ‘competitive In an important earlier piece of work, Kyriakou and ability against weeds’, also called ‘weed tolerance’ or Fasoulas (1985), selecting in a rye population in the ‘weed suppression ability’. As a hidden cause of seed presence (15 cm) and absence (90 cm) of interplant com- and variety degeneration and a detriment to the efficient petition, demonstrated that selection was effective in the response to selection, this negative correlation may absence of competition and ineffective in its presence. further explain the eloquently described by Zeven They hypothesized a strong negative correlation between (1999) practice of ‘inexplicable’ seed replacement yielding and competitive ability. In a systematic study of among traditional farmers. the correlation between yielding and competitive ability in wheat, Fasoula (1990) measured this correlation and found it to be high and negative r ¼ 20.94. As shown Table 1. Yield of the two central rows of the selected lines in Fig. 1, honeycomb selection within the soft wheat in % of the control (cv. Athenaida) in RCB trials in two locationsa cultivar Siete Cerros for high (H) and low (L) yield in the absence of competition, resulted in the isolation of Location Athalassa Location H-lines outyielding the L-lines in monoculture, but Entry (% yield) Zygi (% yield) lagging behind the L-lines in mixed stand with Siete Control 100b 100b Cerros. The occurrence of Yc and yC plants within Line 2 306a 219a Siete Cerros demonstrated that the traditional propa- Line 1 166b 123b gation under dense stand led to a gradual degeneration a The 100% mean yield of the control corresponds to 1213 g of the cultivar, due to the preferential proliferation of in location Athalassa and to 1688 g in location Zygi. Means the yC types at the expense of the Yc ones; hence the with different letters differ significantly at the 5% level. Seed degradation and plant genetic resources 323 Necessity of and means for nonstop selection – a General conclusions novel selection equation The invisible cause of seed and variety degradation due to The development of the selection equation (Fasoula, the negative correlation between yielding and competitive 2008) used in this study is a new addition in the bree- ability is counteracted by nonstop selection at ultra-wide ders’ toolbox. The results of the 2009–10 RCB trials plant spacings on the basis of equation A. Furthermore, (Table 1) show line 2 to outyield the original barley cul- the process of nonstop selection effectively exploits the tivar 2.5 times on the average, across both sites. Using capacity of plant genetic resources to steadily release adap- the equation A, all plants in the trial are assigned a unit- tive variation through the endogenous mechanisms of the less equation value, ranked on the basis of their crop genome in response to environmental stimuli, ensuring yield potential. All plants have equal opportunities to continuous progress through selection. be selected, effectively eliminating the masking effects of soil fertility that interfere with selection efficiency (Supplementary Figs. S1 and S2, available online only at http://journals.cambridge.org). This permits the appli- References cation of very high selection pressures (1–0.5%), leading to the reliable isolation of top yielding plants and Borlaug NE (2008) Stem rust never sleeps. New York Times,26April. Fasoula DA (1990) Correlations between auto-, allo- and nil- to the corresponding increase of progress through competition and their implications in plant breeding. selection. Euphytica 50: 57–62. In order to effectively exploit the capacity of crops to Fasoula DA and Fasoula VA (1997) Competitive ability and plant gradually build up resistance to abiotic and biotic stresses breeding. Plant Breeding Reviews 14: 89–138. and keep ahead of mutating pathogens, constant ‘non- Fasoula VA and Fasoula DA (2000) Honeycomb breeding: principles and applications. Plant Breeding Reviews 18: 177–250. stop selection’ (Fasoula and Fasoula, 2000) for high Fasoula VA (2008) Two novel whole-plant field phenotyping and stable crop yield in the absence of competition is equations maximize selection efficiency. In: Prohens J an important and necessary step. The process of nonstop and Badenes ML (eds) Modern Variety Breeding for Present selection, greatly enhanced by the use of equation A, is and Future Needs. Valencia: Editorial Universidad Politec- leading to the development of density-neutral cultivars, nica de Valencia, pp. 361–365. Fasoulas AC (2000) Building up resistance to Verticillium wilt in i.e. to cultivars that yield optimally over a wide range cotton through honeycomb breeding. In: Gillham FM (ed.) of plant densities, possessing also the ability to reduce New Frontiers in Cotton Research. Proceedings of 2nd the number of competitive weeds. Relevant data con- World Cotton Research Conference, 6–12 September firming the efficiency of nonstop selection in terms of 1998. Athens, Greece, pp. 120–124. developing resistance to soil pathogens are reported by Fasoulas AC and Fasoula VA (1995) The honeycomb selection designs. Plant Breeding Reviews 13: 87–139. Fasoulas (2000) in cotton. As to the perennial signifi- Gepts P (2006) Plant genetic resources conservation and cance of the efforts to keep pace with the mutating utilization: the accomplishments and future of a societal pathogens, it is of interest to consider the recent call insurance policy. Crop Science 46: 2278–2292. by Borlaug (2008) for a pulling together of the world’s Kyriakou DT and Fasoulas AC (1985) Effects of competition scientists to develop a new generation of resistant and selection pressure on yield response in winter rye (Secale cereale L.). Euphytica 34: 883–895. wheat varieties in order to avoid catastrophes from Zeven AC (1999) The traditional inexplicable replacement of epidemics, aptly remarking that diseases and pests seed and seed ware of landraces and cultivars: a review. never sleep. Euphytica 110: 181–191. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 324–326 ISSN 1479-2621 doi:10.1017/S1479262111000074

Cultivated and wild Solanum species as potential sources for health-promoting quality traits

Christina B. Wegener* and Gisela Jansen Julius Kuehn-Institute, Institute for Resistance Research and Stress Tolerance, Experimental Station for Potato Research, Sanitz, Germany

Abstract In this study, several genotypes of cultivated (Solanum tuberosum subsp. andigena, S. phureja) and wild, tuber-bearing Solanum species (S. chacoense, S. pinnatisectum) were examined for concentrations of antioxidants, soluble phenols and proteins in their tuber tissue. The potato genotypes differed considerably in all these traits. Amounts of antioxidants ranged from 0.06 to 4.22 mg/mg fresh weight (fw) when the ascorbic acid equivalent was measured and from 0.08 to 3.98 mg/mg fw for the trolox equivalent. The wild species S. pinnatisectum exhibited on average higher levels of both types of antioxidants than the other Solanum species, and it also had the highest quantities of soluble phenols and proteins in its tuber tissue. Among the species, S. phureja ranked on the lowest level of antioxidant potential.

Keywords: antioxidants; plant phenols; potatoes; soluble proteins

Introduction antioxidant potential assessed as ascorbic acid (ACE) and trolox equivalent (TXE) as well as their contents of Improvement of health-related quality traits like antioxi- soluble phenols and soluble proteins. dants, plant phenols, vitamins and anti-cancer compounds is a major challenge for potato breeding in the future (Van Gijssel, 2005). Regarding their consumption, Materials and methods potatoes are considered as a signiﬁcant antioxidant source in human nutrition (Lachman and Hamouz, Seed tubers of cultivated and wild Solanum species, 2005). Besides ascorbic acid, a-tocopherol and b-caro- each represented by two accessions and several geno- tene, also plant phenols like caffeic acid and chlorogenic types (Table 1), were introduced from the Leibniz Insti- acid are efﬁcient antioxidants (Byers and Perry, 1992). tute of Plant Genetics and Crop Plant Research, Potato When consumed in the diet, phenolic compounds have Genebank, Groß Lu¨sewitz, Germany. The Solanum the ability to protect human cells against oxidative species were chosen according to (1) information damage (Blomhoff, 2005). In view of health-related about resistance properties to major potato diseases quality traits, wild potatoes are increasingly seen as an such as soft rot, blackleg, late blight etc. (Hawkes, interesting gene pool in breeding. 1994), and (2) their ability to produce enough tubers In this study, two cultivated (Solanum tuberosum with an acceptable size ($10 mm in diameter) under subsp. andigena, adg; S. phureja, phu) and two wild, conditions used in this work. The following species tuber-bearing Solanum species (S. chacoense, chc; and accessions were involved in the test series: (adg) S. pinnatisectum, pnt), each represented by several 31881, 34155; (chc) 30161, 30180; (phu) 31455, 31467; accessions and genotypes, were examined for their (pnt) 31598, 31606. Ten plants per genotype were grown in 130 mm diameter pots under a shelter from April to October 2008. After harvest, tubers were stored in a controlled * Corresponding author. E-mail: [email protected] environment at 58C. The analyses described below were Cultivated and wild Solanum species 325

Table 1. Concentrations of water (ACE) and lipid-soluble antioxidants (TXE) in tuber tissue of cultivated and wild Solanum species

Antioxidant activity ACE (mg/mg fw) TXE (mg/mg fw) Number of Species genotypes Average Range Average Range adg 5 0.15 0.08–0.24 0.27 0.21–0.30 chc 6 0.54 0.19–0.85 0.59 0.32–0.79 phu 6 0.12 0.06–0.21 0.18 0.08–0.26 pnt 6 2.46 0.89–4.22 2.84 1.60–3.98 performed in duplicate with standard deviation #5%, well as in contents of soluble phenols and proteins and 20 tubers were randomly taken as an average (Table 2). Of the four Solanum species tested in this sample for each genotype. study, S. pinnatisectum exhibited on average the highest Assay of the antioxidant activity, i.e. ascorbic acid and antioxidant potential, including water (ACE) and lipid- trolox equivalent (TXE), was performed on a photochem (TXE) soluble antioxidants (Table 1). Pnt also exceeded instrument (Analytik Jena AG, Jena, Germany) as the other three Solanum species in quantities of soluble described by Wegener et al. (2009). The amounts of sol- phenols and soluble proteins present in its tuber tissue uble phenols in extract samples prepared from tuber (Table 2). Especially, pnt 31598-2 was outstanding in its tissue as detailed by Wegener et al. (2009) were measured antioxidant capacity with values of 4.22 mg/mg fresh using Folin–Ciocalteu reagent (Sigma-Aldrich, Tauf- weight (fw) for ACE and 3.98 mg/mg fw for TXE, and it kirchen, Germany) according to Cahill and McComb also displayed the highest level of soluble phenols (1992). The quantities of soluble proteins were deter- (2.84 g/kg fw) in its tuber tissue. This fact concurred mined in tuber tissue extracts by means of a Bradford with results of the years before, i.e. 2006/07 (Wegener assay using RotiR-Quant reagent (Roth, Germany), and Jansen, 2009). With it, pnt 31598-2 exhibited con- according to the manufacturer recommendations. siderably higher amounts of phenols as current potato cultivars, e.g. Adretta (0.45 g/kg fw), Romance (0.68 g/kg fw) and De´sireé (0.98 g/kg fw) analyzed in another work Statistic analyses (Wegener et al., 2009). The differences between pnt and the other Solanum The difference in phenol, protein and antioxidant species in ACE (adg and phu, P , 0.01; chc, P , 0.05), contents between pnt and the other Solanum species TXE (adg and phu, P , 0.001; chc, P , 0.01), soluble was compared by means of t-test, and P , 0.05 was phenols (all three, P , 0.01) and soluble proteins (adg considered significant. Correlation coefficients were and phu, P , 0.01) were statistically significant, except calculated between ACE and TXE, soluble phenols and protein contents of the wild species chc, which did antioxidants, and between proteins and the latter. not differ significantly from those of pnt (Table 2). In addition, chc revealed higher protein values than the Results and discussion cultivated species adg and phu (both, P , 0.05), and it also differed significantly in its antioxidant potential The cultivated and wild Solanum species and genotypes including ACE (adg and phu, P , 0.05) and TXE (both, varied notably in their antioxidant activities (Table 1) as P , 0.01) from the latter (Table 1). Among the species

Table 2. Concentrations of soluble phenols and proteins in tuber tissue of cultivated and wild Solanum species

Soluble phenols (g/kg fw) Soluble proteins (mg/ml) Number of Species genotypes Average Range Average Range adg 5 0.40 0.34–0.48 5.99 2.80–11.48 chc 6 0.51 0.25–0.66 11.13 7.08–12.68 phu 6 0.49 0.31–0.67 7.32 4.56–11.48 pnt 6 1.82 1.06–2.84 12.88 11.40–14.44 326 C. B. Wegener and G. Jansen tested in this work, S. phureja ranked on the lowest level Byers T and Perry G (1992) Dietary carotenes, vitamin C and of antioxidant potential (Table 1), and S. tuberosum vitamin E as protective antioxidants in human cancers. subsp. andigena displayed the lowest amounts of phe- Annual Review of Nutrition 12: 139–159. Cahill DM and McComb JA (1992) A comparison of changes in nols and proteins (Table 2) in its tuber tissue. phenylalanine ammonia-lyase activity, lignin and phenolic The ACE was correlated with the TXE (r ¼ 0.94, synthesis in the roots of Eucalyptus calophylla (field resist- P , 0.01, n ¼ 23). Moreover, a significant correlation ant) and E. marginata (susceptible) when infected with (P , 0.05, all) was observed between ACE and phenols Phytophthora cinnamomi. Physiological and Molecular (r ¼ 0.93), ACE and proteins (r ¼ 0.52), TXE and phenols Plant Pathology 40: 315–332. (r ¼ 0.96), and TXE and proteins (r ¼ 0.57). It seems Hawkes JG (1994) Origins in cultivated potatoes and species therefore that soluble phenols and soluble proteins accu- relationships. In: Bradshaw JE and Mackay GR (eds) mulated in tuber tissue substantially contributed to the Potato Genetics. Wallingford: CAB International, pp. 3–42. antioxidant activity of potatoes. Lachman J and Hamouz K (2005) Red and purple coloured pota- The results revealed that wild, tuber-bearing Solanum toes as a significant source in human nutrition – a review. species, i.e. pnt and chc examined in this work, are Plant Soil and Environment 51: 477–482. Van Gijssel J (2005) The potential of potatoes for attractive con- indeed an interesting source for health-related quality venience food: focus on product quality and nutritional traits like antioxidants, plant phenols and valuable value. In: Haverkort AJ and Struik PC (eds) Potato in proteins. Above all, an inclusion of S. pinnatisectum Progress. Wageningen: Academic Publishers, pp. 27–32. in breeding could be profitable for the enhancement of Wegener CB and Jansen G (2009) Antioxidants in wild and the antioxidant potential in new potato cultivars. cultivated potato species. In: Feldmann F, Alford DV and Furk C (eds) Crop Plant Resistance to Biotic and Abiotic Factors: Current Potential and Future Demands. Braunsch- weig: DPG Publishing, pp. 91–95. References Wegener CB, Jansen G, Ju¨rgens HU and Schu¨tze W (2009) Special quality traits of coloured breeding clones: antho- Blomhoff R (2005) Dietary antioxidants and cardiovascular cyanins, soluble phenols and antioxidant capacity. Journal disease. Current Opinion in Lipidology 16: 47–54. of the Science of Food and Agriculture 89: 206–215. q NIAB [2011]. This is a work of the U.S. Government Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 327–329 and is not subject to copyright protection in the United States. doi:10.1017/S1479262111000116 ISSN 1479-2621

Iron biofortiﬁcation of maize grain

Owen A. Hoekenga1*, Mercy G. Lung’aho2, Elad Tako2, Leon V. Kochian1 and Raymond P. Glahn1 1Robert W. Holley Center for Agriculture and Health, USDA-ARS, Ithaca, NY 14853, USA and 2Department of Food Science, Cornell University, Ithaca, NY 14853, USA

Abstract Mineral nutrient deficiencies are a worldwide problem that is directly correlated with poverty and food insecurity. The most common of these is iron deficiency; more than one-third of the world’s population suffer from iron deficiency-induced anaemia, 80% of which are in developing countries. The consequences of iron deficiency include increased mortality and morbidity rates, diminished cognitive abilities in children and reduced labour productivity, which in turn stagnates national development. The developed world has made tremendous success in alleviating nutrient deficiencies through dietary diversification, food product fortification, improved public health care and supplementation. In developing countries, these strategies are often expensive and difficult to sustain, especially in rural areas. The rural poor typically consume what they grow and are dependent upon a small number of staple crops for the vast majority of their nutrition. Therefore, genetic improvement of staple crops (biofortification) is the most cost-effective and sustainable solution to this global health problem. In this study, we describe a strategy to enhance iron nutritional quality in maize using a human cell culture (Caco-2)-based bioassay as a phenotyping tool to guide genetic analysis of the trait. We also report validation of this approach using an animal feeding study.

Keywords: biofortiﬁcation; Caco-2 bioassay; iron nutrition; maize; quantitative trait loci

Introduction their nutrition (Bouis, 2000; Welch and Graham, 2000). This limits the feasibility of processed food fortification Iron (Fe) deficiency is a worldwide problem that is as a micronutrient deficiency-alleviating tool for this directly correlated with poverty and food insecurity. group and emphasizes the importance of plant-based Approximately, one-third of the world’s population suffer solutions for human nutrition problems. from Fe deficiency-induced anaemia, 80% of which are Fe is less available for absorption into the human body in developing countries (Boccio and Iyengar, 2003). from vegetarian as opposed to non-vegetarian diets The consequences of Fe deficiency include increased (Fig. 1, adapted from Hunt (2003)). The influence of mortality and morbidity rates, diminished cognitive abil- biochemical factors on Fe availability depends on the ities of children and reduced labour productivity that form of Fe. Fe in plants exists primarily as non-haeme in turn stagnates national development (Caballero, Fe. One factor influencing non-haeme Fe bioavailability 2002). About 75% of the world’s poor households live is solubility; hence an increase in Fe concentration in rural areas, and the majority are small-scale farmers alone is not a sufficient answer to dietary Fe deficiency (Pinstrup-Andersen, 2002). The resource-poor typically problems (Lucca et al., 2001). Thus, evaluating the bioa- consume what they grow and are dependent upon a vailability of Fe is a necessity in order to improve the Fe small number of staple crops for the vast majority of nutritional quality in staple food crops. Given the high cost of quantifying Fe bioavailability via human and animal studies, in vitro screening of food samples represents the most feasible system for phenotyping large * Corresponding author. E-mail: [email protected] number of samples to identify factors and interactions 328 O. A. Hoekenga et al.

Non-vegetarian diet P<0.001 4 15 Vegetarian diet Fe levels were estimated by measuring the Fe storage protein ferritin by a commercial immunoradiometric 3 10 assay; ferritin is produced in linear response to Fe 2 uptake by the Caco-2 cells. Grain Fe concentration and bioavailable Fe values were used to map QTLs using 5 P<0.001 1 QTL Cartographer (Wang et al., 2006). BC2S2 families P<0.05 were derived from select IBM RI in both parental back- Amount absorbed (mg/d)

Amount consumed (mg/d) 0 0 Fe ZnCu Fe Zn Cu grounds and used to isolate QTL, using a combination of SSR markers and phenotypic selection. BC2S3 sister Fig. 1. Bioavailability is a key component of nutritional lines with contrasting Fe nutritional values were used to quality. While vegetarian and non-vegetarian diets may have similar quantities of nutrients such as Fe, the differ- derive new related inbreds and to create related hybrids, ence in solubility, absorption and ultimate utility (‘bioavail- to facilitate grain production for animal feeding studies. ability’) is significantly different and should be considered Hybrids with predicted high and low Fe nutritional in discussions of nutritional quality and breeding goals values were grown in 2008 and fed to newly hatched (Adapted from Hunt (2003)). Zn, zinc; Cu, copper. chickens to validate the results of the Caco-2 assay (Tako et al., 2010). that affect Fe bioavailability (Wienk et al., 1999). The current state of the art for in vitro screening is a two-stage assay of a simulated gastric and intestinal Results and discussion digestion of food together with feeding human intestinal epithelial cells, specifically the Caco-2 cell line (Glahn Given that the majority of Fe from plant-based foods is et al., 1998). We utilized the Caco-2 bioassay to evaluate inaccessible to those who consume them, enhancing Fe Fe bioavailability in maize with the intermated bioavailability has greater potential to improve nutritional B73 £ Mo17 recombinant inbred (IBM RI) population quality than merely increasing the amount of Fe present. (Lee et al., 2002). The data collected from the bioassay To that end, the Caco-2 bioassay was used as a phenotyp- allowed the identification of quantitative trait loci ing tool to evaluate Fe nutritional quality in a maize RI (QTLs) and have also been useful to guide selection for population. Once QTLs were identified, backcross enhanced (and diminished) Fe nutritional quality in derivatives from RI lines were isolated to combine the novel germplasm. superior (or inferior) alleles for the three largest QTLs. As backcross derivatives were constructed in both parental backgrounds, using both the superior and the inferior QTL configurations, hybrids were easily con- Materials and methods structed to be heterozygous essentially everywhere except the three QTL regions under selection. Hand- The IBM RI set was received from the Maize Genetics pollinated hybrids produced similar amounts of grain Cooperation Stock Center (Urbana, IL, USA) and grown (not significant, P , 0.05), while the potentially superior at the Musgrave Research Farm of Cornell University hybrids had more Fe in the grain (21.2 ^ 0.2 vs. (Poplar Ridge, NY, USA). The IBM RI set was planted in 18.8 ^ 0.4 mg Fe/g dry weight, mean ^ standard error 2001, 2003 and 2005. Grain samples from each year (SE)) and had more bioavailable Fe according to the were analyzed by an inductively coupled argon plasma Caco-2 bioassay (44.4 ^ 4.2 vs. 26.5 ^ 1.6 mg ferritin/mg atomic emission spectrophotometer to measure Fe con- total protein, mean ^ SE). The improvement in grain Fe centration. Fe bioavailability was measured in the IBM concentration (12%) was far smaller than the improve- RI from only the 2003 season using the Caco-2 in vitro ment in grain Fe bioavailability (67%), indicating that digestion method as described by Glahn et al. (1998). the majority of improvement in nutritional quality must Briefly, 50 g lots of clean grain were cooked, freeze have arisen from changes in grain chemistry that enhance dried and ground to fine powder. A subsample of 1 g Fe bioavailability. was then resuspended in water, treated with pepsin While the results of the Caco-2 bioassay have been (pH 2.0), neutralized, treated with pancreatin/bile salts consistent over several seasons, the impact of these results (pH 7.0) and then layered over a dialysis membrane is tempered by the artificial conditions of the bioassay. (15 kDa cut-off). Soluble Fe from the cooked and To validate the use of the Caco-2 bioassay as selection digested food matrix passed through the dialysis mem- tool, grain from the contrasting hybrids was fed to brane and was available to Caco-2 cells for uptake newly hatched chickens for 28 d. Bird health and Fe bio- (2 h feeding, 22 h recovery). Cells were harvested and availability were estimated from weekly blood tests washed, and total protein was then isolated. Bioavailable that measured haemoglobin (Fig. 2). In addition to the Iron biofortification of maize grain 329 14 genomic-based studies into the bases of Fe nutritional 12 quality in maize. We should be able to identify the underlying genes and compounds that improve Fe bio- 10 availability and transfer this knowledge to improve maize germplasm adapted to particular environments. This 8 knowledge can also be used to enhance other staple 6 crops using conventional breeding or biotechnology.

Group H, high bioavailable Fe maize diet (25 ppm)

Hemogloblin (g/dl) Hemogloblin 4 Group L, low bioavailable Fe maize diet (24 ppm) 2 Group C, positive control (commercial diet, 116 ppm) References

0 Boccio JR and Iyengar V (2003) Iron deficiency: causes, conse- Baseline Day 8 Day 15 Day 22 Day 28 quences, and strategies to overcome this nutritional Fig. 2. Poultry experiment validates predictions from problem. Biological Trace Element Research 94: 1–32. bioassay-based selection. Marker and phenotypic selection Bouis HE (2000) Enrichment of food staples through plant were used to develop hybrids with potentially superior and breeding: a new strategy for fighting micronutrient mal- inferior Fe nutritional quality. These efforts were evaluated nutrition. Nutrition 16: 701–704. with a 28 d feeding study in newly hatched chickens, where Caballero B (2002) Global patterns of child health: the role of haemoglobin levels were used to estimate Fe levels and nutrition. Annals of Nutrition and Metabolism 46: 3–7. overall animal health (n ¼ 6 birds, haemoglobin levels are Glahn RP, Lee OA, Yeung A, Goldman MI and Miller DD (1998) mean ^ SE). Caco-2 cell ferritin formation predicts nonradiolabeled food iron availability in an in vitro digestion/Caco-2 cell culture model. Journal of Nutrition 128: 1555–1561. experimental diets where the contrasting hybrids Hunt JR (2003) Bioavailability of iron, zinc, and other trace provided nearly all of the dietary Fe, a control diet was minerals from vegetarian diets. American Journal of Clinical Nutrition 78: 633S–639S. fed to a third group of birds and contained supplemen- Lee M, Sharopova N, Beavis WD, Grant D, Katt M, Blair D and tary Fe according to veterinary protocols. While the Hallauer A (2002) Expanding the genetic map of maize control and superior hybrid diets differed more than with the intermated B73 £ Mo17 (IBM) population. Plant fourfold for Fe concentration, birds fed both diets pro- Molecular Biology 48: 453–461. duced identical amounts of haemoglobin in response Lucca P, Hurrell R and Potrykus I (2001) Genetic engineering approaches to improve the bioavailability and the level of and were Fe replete. In contrast, the superior and inferior iron in rice grains. Theoretical and Applied Genetics 102: hybrid diets contained similar Fe concentrations but 392–397. produced highly significant differences in the birds for Pinstrup-Andersen P (2002) Food and agricultural policy for haemoglobin production and Fe repletion (Fig. 2). a globalizing world: preparing for the future. American These results indicate that the Caco-2 bioassay was a Journal of Agricultural Economics 84: 1201–1214. Tako E, Rutzke MA and Glahn RP (2010) Using the domestic useful phenotyping tool for evaluating Fe bioavailability, chicken (Gallus gallus)asanin vivo model for iron which could be useful for the improvement of any staple bioavailability. Poultry Science 89: 514–521. crop. In addition, we have produced novel maize Wang S, Basten C and Zeng Z (2006) Windows QTL Carto- varieties with altered Fe nutritional quality, which may grapher v2.5. Raleigh, NC: North Carolina State University. serve as donors for breeding programmes for biofortifica- Welch RM and Graham RD (2000) Breeding crops for enhanced micronutrient content. Plant and Soil 245: 205–214. tion of maize. As several BC2S2 families were used to Wienk K, Marx J and Beynen A (1999) The concept of iron create superior and inferior derivatives, these stocks bioavailability and its assessment. European Journal of will also be highly useful for genetic, agronomic and Nutrition 38: 51–75. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 330–333 ISSN 1479-2621 doi:10.1017/S1479262111000050

Polymorphism of waxy proteins in Spanish hulled wheats

C. Guzmań1*, L. Caballero2, M. V. Gutierrez1 and J. B. Alvarez1 1Departamento de Gene´tica, Escuela Tećnica Superior de Ingenieros Agrońomos y de Montes, Edificio Gregor Mendel, Campus de Rabanales, Universidad de Co´rdoba, ES-14071 Co´rdoba, Spain and 2Departamento de Mejora Gene´tica Vegetal, Instituto de Agricultura Sostenible, Consejo Superior de Investigaciones Cientı´ficas, Apdo. 4084, ES-14080 Co´rdoba, Spain

Abstract Hulled wheats are neglected crops that have potential in plant breeding programmes of modern durum and common wheat. Among these wheats, three species were widely cultivated in Spain until the mid 20th century: Triticum monococcum ssp. monococcum (einkorn), Triticum turgidum ssp. dicoccum (emmer) and Triticum aestivum ssp. spelta (spelt). One important aspect of wheat grain quality is starch composition, which is related to the action of waxy proteins. A collection of 536 accessions of Spanish hulled wheats was analyzed for waxy protein composition using sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE). Polymorphism was found for the Wx-A1, Wx-B1 and Wx-D1 proteins, including new and null alleles in the three species. An allelic variant with an electrophoretic mobility not previously described was found in einkorn wheat. In emmer and spelt, some alleles with different mobility were also found. A Wx-B1 null allele was detected in emmer wheat, and null alleles for Wx-A1, Wx-B1 and Wx-D1 were found in spelt wheat. The variations found could be used to enlarge the gene pool available to breeders, and to design new cultivars with different levels of amylose content.

Keywords: amylose content; genetic polymorphism; hulled wheats; waxy proteins

Introduction existed is conserved in germplasm banks. Due to the growing interest in natural food, some of these ancient Hulled wheats are the wild or cultivated species of crops are undergoing a revival in European agriculture. the Triticum genus, which have glumes that tightly enclose In the last decade, our research group has investigated the grains, even after normal threshing. In Spain, three three important collections of Spanish hulled wheats, species of hulled wheats were widely cultivated until mainly for genetic variation in seed storage proteins the late 1960 s: Triticum monococcum L. ssp. monococcum (Caballero et al., 2001, 2004a, b; Pflu¨ger et al., 2001; (2n ¼ 2x ¼ 14, AA), T. turgidum ssp. dicoccum Schrank Alvarez et al., 2006). The high variability detected in em. Thell. (2n ¼ 4x ¼ 28, AABB) and T. aestivum ssp. these species suggests that these collections are useful spelta L. em. Thell. (2n ¼ 6x ¼ 42, AABBDD). Nowadays, gene reservoirs that can be used in breeding programmes. only emmer and spelt wheat are still cultivated, and The starch composition of wheat grain has a primary solely in marginal farming areas of Asturias (North of influence on flour quality. Wheat starch consists of Spain). Fortunately, some of the biodiversity that once two types of glucose polymers, the essentially linear amylose and the highly branched amylopectin, in a ratio of 22–35:68–75%. Synthesis of amylose in the seed endosperm is carried out by the waxy proteins or gran- * Corresponding author. E-mail: [email protected] ule-bound starch synthases that are encoded by genes Polymorphism of waxy proteins 331 located in the Wx locus on the homoeologous group 7 30 mA/gel at 188C, continuing for 4 h after the tracking chromosomes (Ainsworth et al., 1993). In common dye migrated off the gel. Protein bands were visualized wheat, the Wx-A1 locus is located on chromosome 7AS, by silver staining. the Wx-B1 locus on 4AL (a segment of chromosome 7BS For two-dimensional PAGE (2D-PAGE), 8.0 mg of that has been translocated) and the Wx-D1 locus on starch was soaked at room temperature in 300 ml of lysis chromosome 7DS. Since starch properties such as gela- buffer (8 M urea, 2% ampholine pH 3.5–10 (Pharmacia) tinization, pasting and gelation depend on the amylose: and 5% DTT). After centrifugation, the supernatant amylopectin ratio (Zeng et al., 1997), these proteins are containing the solubilized proteins was subjected to very important in terms of flour quality. 2D-PAGE using isoelectric focusing (IEF) for the first The aim of this study was to assess waxy dimension and SDS-PAGE for the second. IEF gels con- protein polymorphism in a broad collection of Spanish tained 2.5% (v/v) ampholines (pH 3–10/5–8 and 8–10, hulled wheats. 1:1). Focusing commenced from the acidic end (0.01 M

H3PO4) at 200 V for 30 min, and continued at 400 V for 17 h, and then at 800 V for 1 h at room temperature. Material and methods After IEF, SDS-PAGE was conducted as described above.

Plant material Results and discussion In the current study, 536 Spanish hulled wheat lines (29 lines of einkorn, 87 lines of emmer and 420 lines The studies carried out on waxy proteins polymorphism of spelt) were analyzed. These lines were derived by have shown that variability is relatively low compared single seed selection from an equal number of acces- with other cereal grain proteins, such as seed storage sions obtained from the National Small Grain Collection proteins. However, Yamamori et al. (1994, 1995) found (Aberdeen, MD, USA), the Center for Genetic Resources five alleles for the Wx-A1 gene in common wheat, and (The Netherlands) and the Centro de Recursos Fitogene´- another six alleles have been described for the Wx-B1 ticos-INIA (Alcala´ de Henares, Spain). The plants were gene in common and durum wheats (Rodrı´guez-Quijano grown under field conditions during 2008, and several et al., 1998). More recently, Caballero et al. (2008) spikes per plant were protected to prevent random reported high variability for these proteins in some crossings. Durum wheat cultivars (cvs). Langdon and ancient wheats and related species. Mexicali and common wheat cultivar (cv). Chinese Ten waxy alleles (two for Wx-A m1 locus, two for Spring were used as standards. Wx-A1, three for Wx-B1 and three for Wx-D1) were detected in the lines herein evaluated (Table 1). In all cases, one of these alleles was clearly predominant, Starch extraction and electrophoretic analysis while the others were rare or very rare. For the Wx-A m1 locus, the allele Wx-A m1a0 that exhibited lower mobility Twenty milligrams of flour were mixed with 1 ml of than the Wx-A m1a was only found in a single accession distilled water and incubated at 48C for 24 h. The homo- (BGE-014269). Both alleles showed higher mobility than genate was filtered through Miracloth and centrifuged at that of the Wx-A1a allele detected in polyploid wheat 14,000 g for 1.5 min. The pellet was washed with 1 ml of buffer A (55 mM Tris–HCl, pH 6.8, 2.3% (w/v) SDS, 2% Table 1. Variability and allelic frequencies (in brackets) (w/v) dithiothreitol (DTT) and 10% (v/v) glycerol), detected for waxy genes in the three collections evaluated according to the method of Echt and Schwartz (1981). Then, 1 ml of buffer A was added to the pellet and left Einkorn Emmer Spelt for 30 min at room temperature. The pellet was washed Locus Alleles (n ¼ 29) (n ¼ 87) (n ¼ 420) three times with distilled water, once with acetone and Wx-A m1 a 28 (96.6) – – then air-dried. The residue was mixed with 80 mlof a0 1 (3.4) – – buffer A, heated in a boiling water bath for 2 min, Wx-A1 a – 87 (100) 384 (91.4) cooled on ice and centrifuged. b – 0 36 (8.6) Wx-B1 a – 85 (97.8) 317 (75.5) Aliquots of supernatant (20 ml) were loaded in vertical b – 1 (1.1) 49 (11.7) SDS-PAGE slabs in a discontinuous Tris–HCl–SDS buffer c0 – 1 (1.1) 54 (12.8) system (pH: 6.8/8.8) at a polyacrylamide concentration Wx-D1 a – – 418 (99.6) of 12% (w/v, cross-linker (C): 0.44%). The Tris–HCl/ b – – 1 (0.2) glycine buffer system of Laemmli (1970) was used. g – – 1 (0.2) H 0.067 0.023 0.189 Electrophoresis was performed at a constant current of e 332 C. Guzmań et al. 12345678 59 59 kDa kDa

Wx-A1 Wx-B1

Wx-A1 Wx-B1 Wx-A1

Fig. 1. Electrophoresis separation of waxy proteins in tetraploid wheats. One-dimensional SDS-PAGE (up) and two-dimensional IEF £ SDS-PAGE (down). Lanes are as follows: 1, KU 4213D (Wx-A1a and Wx-B1d); 2 and 4, Emmer-71 (Wx-A1a and Wx-B1a); 3 and 7, cv. Langdon (Wx-A1a and Wx-B1a); 5, cv. Mexicali (Wx-A1a and Wx-B1c0); 6, Emmer-49 (Wx-A1a and Wx-B1c0); 8, Emmer-52 (Wx-A1a and Wx-B1b).

(emmer and spelt). This is in agreement with studies that In conclusion, the current study has shown that suggest that the A genome of einkorn is substantially Spanish hulled wheats exhibit significant waxy protein different from the A genome present in emmer and variation, with some novel alleles at risk of erosion by spelt (Dvorak et al., 1993). genetic drift. Consequently, the safeguarding of these Emmer did not show polymorphism for the Wx-A1 and other hulled wheat accessions stored in germplasm locus, whereas in spelt 36, accessions that presented banks is fundamental for the maintenance of genetic the Wx-A1b allele (null type) were detected. The diversity. This diversity may be of use in breeding Wx-A1a alleles detected in both species were similar to programmes focussed on starch quality, for both the allele detected in the durum and common wheat modern wheats and these ancient crops that are cultivars used as standards (Langdon and Mexicali, and undergoing such a revival. Chinese Spring, respectively). The variability for the Wx-B1 locus was higher in both polyploid species. However, most of the emmer Acknowledgements accessions present the Wx-B1a allele, similar to the alleles found in cvs. Langdon and Chinese Spring. This research was supported by grant AGL2007-65685- The other two alleles were very rare. Because SDS- C02-02 from the Spanish Ministry of Science and PAGE can occasionally generate overlapping bands Innovation and the European Regional Development that yield misleading results, the accessions that Fund (FEDER) from the European Union. The first presented the null allele were analyzed by two- author thanks to the Spanish Ministry of Education dimensional electrophoresis (IEF £ SDS-PAGE). The and Science (FPU programme) for a predoctoral data revealed that the Wx-B1 gene is not expressed fellowship. in this accession (Fig. 1). The results were very similar for the Wx-D1 locus, with all accessions showing the Wx-D1a allele with References exception of one accession that presented the allele and another a novel allele named Wx-D1b Ainsworth CC, Clark J and Balsdon J (1993) Expression, Wx-D1 g, which exhibited a slightly lower electrophor- organization and structure of the genes encoding the etic mobility than the Wx-D1a allele. waxy protein (granule-bound starch synthase) in wheat. The mean expected heterocigosity (He) values were Plant Molecular Biology 22: 67–82. very low for all species (Table 1), representing Alvarez JB, Moral A and Martıń LM (2006) Polymorphism and genetic diversity for the seed storage proteins in Spanish approximately 13.4, 5.5 and 48.6% of the genetic diver- cultivated einkorn wheat (Triticum monococcum L. ssp. sity for einkorn, emmer and spelt, respectively, if the monococcum). Genetic Resources and Crop Evolution 53: allelic variants of each locus were distributed randomly. 1061–1067. Polymorphism of waxy proteins 333 Caballero L, Martıń LM and Alvarez JB (2001) Allelic variation Laemmli UK (1970) Cleavage of structural proteins during the of the HMW glutenin subunits in Spain accessions of assembly of the head of bacteriophage T4. Nature 227: spelt wheat. Theoretical and Applied Genetics 103: 680–685. 124–128. Pflu¨ger LA, Martıń LM and Alvarez JB (2001) Variation in Caballero L, Martıń LM and Alvarez JB (2004a) Genetic the HMW and LMW glutenin subunits from Spanish variability of the low-molecular-weight glutenin subunits accessions of emmer wheat (Triticum turgidum ssp. dicoc- in spelt wheat (Triticum aestivum ssp. spelta L. em cum Schrank). Theoretical and Applied Genetics 102: Thell.). Theoretical and Applied Genetics 108: 914–919. 767–772. Caballero L, Martıń LM and Alvarez JB (2004b) Variation Rodrı´guez-Quijano M, Nieto-Taladriz MT and Carrillo JM (1998) and genetic diversity for gliadins in Spanish spelt wheat Polymorphism of waxy proteins in Iberian hexaploid accessions. Genetic Resources and Crop Evolution 51: wheats. Plant Breeding 117: 341–344. 679–686. Yamamori M, Nakamura T, Endo R and Nagamine T (1994) Caballero L, Bancel E, Debiton C and Branlard G (2008) Waxy protein deficiency and chromosomal location of Granule-bound starch synthase (GBSS) diversity of coding genes in common wheat. Theoretical and Applied ancient wheat and related species. Plant Breeding 127: Genetics 89: 179–184. 548–553. Yamamori M, Nakamura T and Nagamine T (1995) Dvorak J, Terlizzi P, Zhang HB and Resta P (1993) The evolution Polymorphism of two waxy proteins in the emmer group of polyploidy wheats: identification of the A genome of tetraploid wheat, Triticum dicoccoides, T. dicoccum, species. Genome 36: 21–31. and T. durum. Plant Breeding 114: 215–218. Echt CS and Schwartz D (1981) Evidence for the inclusion of Zeng M, Morris CF and Batey Wrigley CW II (1997) Sources controlling elements within structural gene at the waxy of variation for starch gelatinization, pasting, and gelation locus in maize. Genetics 99: 275–284. properties in wheat. Cereal Chemistry 74: 63–71. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 334–337 ISSN 1479-2621 doi:10.1017/S1479262111000220

Molecular characterization of the Glu-Ay gene from Triticum urartu for its potential use in quality wheat breeding

M. V. Gutie´rrez, C. Guzmań, L. M. Martıń and J. B. Alvarez* Departamento de Gene´tica, Escuela Tećnica Superior de Ingenierıá Agrono´mica y de Montes, Edificio Gregor Mendel, Campus de Rabanales, Universidad de Co´rdoba, ES-14071 Co´rdoba, Spain

Abstract Triticum urartu Thum. ex Gandil. is a wild species identiﬁed as A-genome donor for polyploid wheats, which could be used as gene source for wheat breeding. The high-molecular weight glutenin subunits are endosperm storage proteins that are associated with bread-making quality. In T. urartu, these proteins are encoded by the Ax and Ay genes at the Glu-A u1 locus. The Ay gene of 17 Glu-A u1 allelic variants previously detected in this species has been analysed using PCR ampliﬁcation and digestion of the PCR products with two endonucleases (Dde I and Pst I). The combination of two restriction patterns has revealed variations between the active and inactive alleles, and within each type. This variation, especially that detected among the active alleles, could enlarge the high-quality genetic pool of modern wheat and be used for bread-making quality improvement in durum and common wheat.

Keywords: electrophoresis; genetic resources; glutenin; quality breeding; wild wheat

Introduction been identiﬁed as the A-genome donor of polyploid wheat (Dvorak et al., 1993). Nowadays, global climate change is one of the major The presence and variability of the endosperm storage problems facing humanity. For crops, this will require proteins are associated with the bread-making quality of the release of new cultivars able to adapt to a changing wheat. These proteins are divided into two main groups environment, without reducing quality standards or affect- (gliadins and glutenins) according to their molecular ing industrial food production according to the demands of characteristics (Payne, 1987). Glutenins are also divided a population highly sensitized to food quality (Godfray into high molecular weight (HMWGs) and low mole- et al., 2010). Several different studies have suggested that cular weight (B-LMWGs and C-LMWGs) subunits. The relatives and wild progenitors of wheat species could be HMWGs, encoded by genes at the Glu-1 loci located on interesting candidates for enlarging the gene pool of culti- the long arm of group-1 homoeologous chromosomes vated wheats (Hajjar and Hodgkin, 2007). At the diploid being the best studied (Payne, 1987), have been associated level, the main species of wild diploid wheat are Triticum with bread-making quality in common wheat (Cornish monococcum L. ssp. aegilopoides Link em. Thell. (syn. et al., 2006). Each Glu-1 locus contains two tightly linked T. boeoticum;2n ¼ 2x ¼ 14; AmAm) and T. urartu Thum. genes that encode for two types of HMWGs, called x- and ex Gandil (2n ¼ 2x ¼ 14; AuAu), the latter species having y-type (Harberd et al., 1986), although some of these genes are not expressed in cultivated wheats. In particular, the Ay subunit is absent in all durum and common wheats, while it is expressed in wild diploid and tetraploid wheats * Corresponding author. E-mail: [email protected] (Waines and Payne, 1987; Ciafﬁ et al., 1993), its presence Molecular characterization of the Glu-Ay gene 335 M 123456789C M1011121314151617CA

1353 bp 1353 bp

(a) (b)

Fig. 1. PAGE separation of PCR products from the Ay genes. (a) active Ay alleles; (b) inactive Ay alleles. M, X174 DNA-Hae III digest; C, cv. Cheyenne; A, cv. Alaga. being associated with an increase in bread-making quality The aim of the present study was the molecular charac- in wheat (Ciafﬁ et al., 1995). terization of the allelic variants for the Ay gene detected Alvarez et al. (2009) showed that the introgression of by Caballero et al. (2008) to obtain additional data prior T. urartu genome in durum wheat affects the gluten to their potential introgression in wheat. strength. However, these materials were developed from a single line, whereas Caballero et al. (2008) found as many as 17 allelic variants for the Glu-A u1 Materials and methods locus in a large collection of T. urartu. The Ax gene was found to be active in all these alleles, while the Ay Seventeen accessions of T. urartu that have the allelic subunit was detected in nine of them. variants found by Caballero et al. (2008) were analysed.

M 123456789CA 10 11 12 13 14 15 16 17 C A M

872 bp

310 bp

118 bp

72 bp

(a) (b) M123456789CA 10 11 12 13 14 15 16 17 C A M

872 bp

310 bp

118 bp

72 bp

Fig. 2. PAGE separation of PCR products from the Ay genes digested with Dde I and PstI (up and down, respectively). (a and c) digestion patterns of active Ay alleles; (b and d) digestion patterns of inactive Ay alleles. M, X174 DNA-Hae III digest; C, cv. Cheyenne; A, cv. Alaga. 336 M. V. Gutie´rrez et al.

DNA isolation was carried out from young leaf tissue showed similarities with the Ay alleles detected in cv. using the cetyl trimethyl ammonium bromide method Alaga (Fig. 2). (Stacey and Isaac, 1994). In some cases, only the combined use of the restriction The primers reported by D’Ovidio et al. (1995) were patterns from both endonucleases evidenced differences used to amplify the complete coding sequence of the between alleles. The actives alleles shown in lanes 1, 3, 4, Ay gene. PCR reactions mixtures were carried out in a 6, 7 and 9 were similar in total size (Fig. 1) and Dde I final volume of 20 ml composed of 1 £ Taq PCR buffer digestion (Fig. 2); however, the Pst I digestion separated (Promega), 125 ng of template DNA, 0.6 mM of each these alleles into two groups. The first group (lanes 1, forward and reverse primer, 1.5 mM MgCl2, 0.2 mM of 3, 4 and 7) presented five bands or fragments, whereas each deoxyribonucleotide and 1 U of Taq DNA poly- the second group (lanes 6 and 9) showed six. The differ- merase (Promega). DNA was subjected to an initial ence was that fragment 2 of group 1 was digested into denaturation step at 958C for 5 min, and the amplification two fragments for group 2: one fragment comigrated conditions were 35 cycles at 958C for 1 min, 608C for with the band 3 and the other produced a new band of 1 min and 728C for 2 min, followed by a final incubation approximately 120 bp that also appeared in the rest step at 728C for 8 min. of the restriction patterns. The same occurred for Dde I The amplicons (PCR products) were separated in digestion in the inactive alleles, where the alleles polyacrylamide gel electrophoresis (PAGE) gel with a showed in lanes 11 and 13 presented similar restriction discontinuous Tris–HCl buffer system (pH: 6.8/8.8) at a pattern, but the PstI digestion indicated that the two polyacrylamide concentration of 8% (w/v, crosslinker (C): alleles are different. 1.68%). These amplicons were digested using Dde I and Although further research needs to be carried out in Pst I endonucleases and separated in PAGE gel with a the future, such as the sequencing of these Ay alleles, discontinuous Tris–HCl buffer system (pH: 6.8/8.8) at a the results of this study demonstrate that the Ay alleles polyacrylamide concentration of 10% (w/v, C: 1.68%). presented in T. urartu, both active and inactive ones, are different from the alleles found in cultivated wheat. For breeding purposes, the variation detected for the active alleles would permit to expand the high-quality Results and discussion gene pool of the cultivated wheats. Consequently, the evaluation and characterization of the genetic resources The amplification of the complete coding sequence in the of this wild species are very important for its conservation accessions with and without active Ay subunits revealed a and potential use in wheat breeding programmes. single band of around 1500 bp, although with some small differences in size among them, in the accessions with Ay active subunits as well as in others with inactive ones (Fig. 1). This is in agreement with the findings of Cabal- Acknowledgements lero et al. (2008), who detected four Ay active subunits with differences in their mobility. This research was supported by grant AGL2007-65685- Some studies have suggested that digestion with endo- C02-02 from the Spanish Ministry of Science and Inno- nucleases could be a useful tool to evaluate the internal vation and the European Regional Development Fund differences between these alleles (Lafiandra et al., 1997; (FEDER) from the European Union. C. G. thanks the Alvarez et al., 1998). The amplicon digestion with Dde I Spanish Ministry of Education and Science (FPU (Fig. 2(a) and (b)) showed three restriction patterns programme) for a predoctoral fellowship. between the active Ay alleles, while in the inactive ones, five restriction patterns were identified. In the case of the digestion with Pst I (Fig. 2(c) and 2(d)) four patterns were revealed for the active Ay alleles and six References for the inactive ones. Although, in general, the restriction Alvarez JB, D’Ovidio R and Lafiandra D (1998) Analysis of patterns of the active alleles were different from those of x- and y-type genes presente at the Glu-A1 encoding high the inactive alleles, one of the patterns of the active molecular weight glutenin subunits in wild and cultivated alleles (lanes 2 and 5), together with the pattern of lane wheats. Proceedings of 9th Wheat Genetics Symposium, 10 (of an inactive allele), showed the same restriction pat- vol. 4. Saskatoon: University Extension Press, University of tern with both digestions (Fig. 2, lanes 2, 5 and 10). Saskatchewan, pp. 127–129. Alvarez JB, Caballero L, Nadal S, Ramı´rez MC and MartıńA On the other hand, all the restriction patterns found in (2009) Development and gluten strength evaluation of these T. urartu lines differed from those of the inactive introgression lines of Triticum urartu in durum wheat. Ay allele present in cv. Cheyenne, while some of them Cereal Research Communications 37: 243–248. Molecular characterization of the Glu-Ay gene 337 Caballero L, Martıń MA and Alvarez JB (2008) Allelic variation Godfray HCJ, Beddington JR, Crute IR, Haddad L, Lawrence D, for the high-and-low-molecular-weight glutenin subunits Muir JF, Pretty J, Robinson S, Thomas SM and Toulmin C in wild diploid wheat (Triticum urartu) and its comparison (2010) Food security: the challenge of feeding 9 billion with durum wheats. Australian Journal of Agricultural people. Science 327: 812–818. Research 59: 906–910. Hajjar R and Hodgkin T (2007) The use of wild relative in crop Ciaffi M, Lafiandra D, Porceddu E and Benedettelli S (1993) improvement: a survey of developments over the last Storage-protein variation in wild emmer wheat (Triticum 20 years. Euphytica 156: 1–13. turgidum ssp. dicoccoides) form Jordan and Turkey. Harberd NP, Bartels D and Thompson RD (1986) DNA restriction I. Electrophoretic characterization of genotypes. Theoreti- fragment variation in the gene family encoding high- cal and Applied Genetics 86: 474–480. molecular-weight (HMW) glutenin subunits of wheat. Bio- Ciaffi M, Lafiandra D, Turchetta T, Ravaglia S, Bariana H, Gupta chemical Genetics 24: 579–596. RB and MacRitchie F (1995) Bread-baking potential of Lafiandra D, Tucci GF, Pavoni A, Turchetta T and Margiotta B durum wheat lines expressing both x-and y-type subunits (1997) PCR analysis of x- and y-type genes present at at the Glu-A1 locus. Cereal Chemistry 72: 465–469. the complex Glu-A1 locus in durum and bread wheat. Cornish GB, Be´ke´s F, Eagles HA and Payne PI (2006) Prediction Theoretical and Applied Genetics 94: 235–240. of dough properties for bread wheats. In: Wrigley C, Payne PI (1987) Genetics of wheat storage proteins and the Be´ke´s F and Bushuk W (eds) Gliadin and Glutenin: The effects of allelic variation on bread-making quality. Unique Balance of Wheat Quality. St. Paul, MN: AACC Annual Review of Plant Physiology 38: 141–153. International Press, pp. 243–280. Stacey J and Isaac P (1994) Isolation of DNA from plants. D’Ovidio R, Masci S and Porceddu E (1995) Development of In: Isaac PG (ed.) Methods in Molecular Biology: Protocols a set of oligonucleotide primers specific for genes at for Nucleic Acid Analysis by Non-Radiactive Probes. the Glu-1 complex loci of wheat. Theoretical and Applied Totawa: Humana Press, pp. 9–15. Genetics 91: 189–194. Waines JG and Payne PI (1987) Electrophoretic analysis of Dvorak J, Terlizzi P, Zhang HB and Resta P (1993) The evolution the high-molecular-weight glutenin subunits of Triticum of polyploidy wheats: identification of the A genome monococcum, T. urartu, and the A genome of bread wheat species. Genome 36: 21–31. (T. aestivum). Theoretical and Applied Genetics 74: 71–76. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 338–341 ISSN 1479-2621 doi:10.1017/S147926211100013X

Protein disulphide isomerase promoter sequence analysis of Triticum urartu, Aegilops speltoides and Aegilops tauschii

Arun Prabhu Dhanapal1, Mario Ciafﬁ2, Enrico Porceddu2* and Elisa d’Aloisio1 1International Doctoral Programme on Agrobiodiversity, Scuola Superiore Sant’Anna – Plant Genetic Resources, Centro Ricerche ENEA Cassacia, Rome 00123, Italy and 2Dipartimento di Agrobiologia e Agrochimica, Universita` della Tuscia, Viterbo, Italy

Abstract Protein disulphide isomerase (PDI) catalyses the formation, reduction and isomerization of disulphide bonds in the newly synthesized secretory proteins. Plant PDIs have been shown to be involved in the folding and deposition of seed storage proteins, which makes this enzyme particularly interesting in wheat, as flour quality is strongly affected by composition and structure of seed storage proteins. In hexaploid wheat cultivar (AABBDD) Chinese Spring (CS), the genomic, complementary DNA and promoter sequences of the three homoeologous gene encoding PDI had been isolated and characterized in a previous study revealing high levels of sequence conservation. In this study, we report the isolation and sequencing of a ,700 bp region, comprising ,600 bp of the putative promoter region and 88 bp of the first exon of the typical PDI gene, in five accessions each from Triticum urartu (AA), Aegilops speltoides (BB) and Aegilops tauschii (DD). Sequence analysis indicated large variation among sequences belonging to the different genomes, while close similarity was found within each species and with the corresponding homoeologous PDI sequences of Triticum aestivum cv. CS (AABBDD) resulting in an overall high conservation of the regulatory motifs conferring endosperm-specific expression.

Keywords: promoter; protein disulphide isomerase; regulatory elements; variability and wild species

Introduction protein bodies of the endosperm. Wheat storage proteins are synthesized in the developing endosperm cells and Protein disulphide isomerase (PDI) catalyses the for- targeted to their endoplasmic reticulum lumen, wherein mation and isomerization of disulphide bonds in nascent they are folded and connected by intermolecular disul- proteins targeted to the endoplasmic reticulum. Different phide bonds to form large aggregates (Shewry and studies on its expression and intracellular localization in Tatham, 1997), whose molecular weight is directly related wheat and maize (Shimoni et al., 1995; Li and Larkins, to higher dough elasticity and better technological 1996), as well as the characterization of the rice esp2 quality. Even though wheat storage proteins have been mutant that lacks PDI expression (Takemoto et al., the object of a wide range of studies both at chemical 2002), suggest that PDI is involved in regulating the fold- and at genetic levels (reviewed by Shewry et al. (2003, ing of seed storage proteins and their deposition into the 2009)), knowledge of factors affecting their folding and deposition is still extremely limited. The genomic and complementary DNA sequences of three PDI homoeologous genes located in the chromosomes 4A, 4B and 4D of * Corresponding author. E-mail: [email protected] bread wheat cv. Chinese Spring (CS) and their putative PDI promoter analysis in wild species of wheat 339 promoters had been cloned and sequenced previously weight matrix with default parameters. Haplotype diver- (Ciaffi et al., 2006). The three genes showed a very sity, polymorphic sites and evolutionary relationship high sequence conservation in the coding region and were calculated using DnaSP v5 software (Librado and their exon/intron structures, which consisted of ten Rozas, 2009). Sequences were then searched for regulat- exons. The putative promoter sequences of the three ory elements in PlantCARE (http://sphinx.rug.ac.be:8080/ genes shared some regulatory motifs involved in PlantCARE/), a database of plant promoters, and PLACE endosperm-specific expression; consistently with this databases (http://www.dna.affrc.go.jp/htdocs/PLACE/; Higo observation, expression analysis of the three homoeo- et al., 1999). logous PDI genes showed the constitutive presence of their mRNAs in several wheat tissues, but the expression level was much higher in developing caryopses. This Results and discussion study reports the isolation and characterization of a ,700 bp region, comprising ,600 bp of the putative Analysis of the 120 sequences revealed that there was promoter sequence and 88 bp of the first exon of the no variability among plants within accessions, thus a PDI gene from wild diploid relatives of wheat, Triticum total of 15 sequences one from each accession were urartu (AA), Aegilops speltoides (BB) and Aegilops further analysed and compared with the three homoeo- tauschii (DD). logous sequences from CS (GPDI-4A, AJ868102; GPDIA- 4B, AJ868103; GPDI-4D, AJ868104; Ciaffi et al., 2006). The alignment of the five sequences derived from the Materials and methods accessions of T. urartu revealed the presence of two haplotypes: one common to the accessions from Leba- A total of 15 accessions, five each from T. urartu (AA), non, Turkey, Jordan and Syria and the other character- A. speltoides (BB) and A. tauschii (DD), were used in istic of the accession from Iran; in total, two variable this study. A total of 120 plants (8 plants/accession) sites were identified (Fig. 1). Both cases involved trans- were grown and analysed. Flag leaves were collected versions (G–T and C–G). Haplotype diversity (Hd) was at heading stage from plants grown in green house 0.400. This high level of conservation of PDI gene pro- (January–June 2008), immediately frozen in liquid nitro- moter in the proximal region of T. urartu may reflect gen and kept at 2808C until use. About 200 mg of leaf its important function or the effect of a bottleneck. tissue were ground in liquid nitrogen, and genomic The alignment of the five sequences from A. speltoides DNA was extracted using Sigma Gen Elute Plant Genomic revealed a higher number of polymorphic sites and the DNA Kit (G2N-350; Sigma Aldrich, St. Louis, MO, USA). presence of five haplotypes, each characteristic of one F2 (50-ACTCCAAATTTGGAACGGG-30) and R1 (50-GGTG- of the five analysed accessions. Out of the 16 identified AGCACTGCCTCGGG-30) primers were chosen to amplify single nucleotide polymorphisms (SNPs), whose the genomic DNA, and amplification products were positions are reported in Fig. 1, ten were transitions directly sequenced. Sequences were analysed in Chromas (A–G and C–T) and six were transversions (A–C, A–T, version 2.3 (http://technelysium.com.au/chromas.html) G–C and G–T); moreover, one Indel at position 27bp to identify any unresolved bases and subjected to visual from the ATG was also identified. Hd was 1.000. inspection. Multiple sequence alignment was performed The number of polymorphic sites found in A. speltoides with Clustal X software version 2.011 (Larkin et al., was thus higher than in T. urartu, consistently with 2007), International Union of Biochemistry as DNA its partial outcrossing nature. Finally, the alignment of

Fig. 1. SNP positions identified in the ,700 bp amplicon comprising the partial PDI promoter sequences and partial first exon from T. urartu (AA), A. speltoides (BB) and A. tauschii (DD). Dots denote consensus sequences. 340 A. P. Dhanapal et al. the sequences from the five analysed accessions of 269, 589

2 2 A. tauschii (DD) revealed the presence of only two haplotypes, one common to three out of the four

242, 547, analysed accessions from Iran and one shared by 2 2 accession number AE541-4 from Iran and the accession from Syria (Fig. 1). In total, only two variable sites were 223, 324, 480 492 442 426 79 2 2 2 2 2 2 2 identiﬁed (Fig. 1), both of which were transversions chii (DD) (G-T and C-G). Hd was 0.600. The presence of fewer polymorphic sites is probably partially due to

A. taus self-fertilizing nature of the species (Dvorak et al., 1998) and partially to the lower level of overall diversity present. The 15 sequences were further compared with their counterparts described previously in bread wheat cv. CS (Ciafﬁ et al., 2006), and the percentage AACA CGTGTCA of nucleotide identity is reported (Supplementary ,

b Table S1, available online only at http://journals. cambridge.org) Furthermore, the evolutionary relation- 296 , 300, b 2 ship between the 18 sequences, ﬁve sequences each , 2 b from T. urartu, A. speltoides, A. tauschii and the three 568 572 2 227270 481 CGAAAAGT358, 493 443427 NA ATGAC GTACGTG CACGTG homoeologous sequences from the bread wheat cv. CS, 2 274, , 79 TATTAAA b 2 2 2 2 2 2 2

2 was investigated (data not shown). The sequences 2 , , , , , , , b b b b b b b , b 552, 548 from A. speltoides (BB) closely clustered together with 2 2 485 83 362 447 431 243 497 247, 231 that located on chromosome 4B of CS, those from des (BB) 2 2 2 2 2 2 2 2 2 A. tauschii (DD) with that on 4D and those from T. urartu (AA) with that on 4A. All sequences were further searched for regulatory A. speltoi elements in two databases, PlantCARE and PLACE. Several regulatory elements, among which many involved in endosperm-specific expression, were identified. The TATA box, a cis-regulatory element found in the promoter region of many eukaryotic genes and considered as the core promoter sequence, was identified at position 279 from the ATG (Table 1). The combination of

588 AACA four motifs, namely GCN-4 motif, prolamine box, AACA The position of motifs for IG48766-1. 2

b and ACGT box, seems to be a minimal requirement for endosperm-specific expression (Wu et al., 1998, 2000). 492 CGAGTCA, CATGTCA 547, a 2 2 All four motifs were identified in the putative PDI promoter sequences analysed. Furthermore, the Skn-1-like (DD) 700 bp amplicon comprising the partial protein disulphide isomerase promoter sequences and partial first exon from 480 ACGAC 79 TATTAAA 356, 442426 GTACGTG CACGTG 323, 226 CGAAAAGT motif, an additional motif that together with ACGT and , 2 2 2 2 2 2 2 the prolamine box modulates the level of endosperm-

artu (AA) speciﬁc expression, was also identiﬁed. The presence A. tauschii

T. ur of these regulatory motifs is consistent with the higher PDI expression found in developing caryopses with respect to other tissues (Ciafﬁ et al., 2001). The regulatory

(BB) and elements are identified, and their consensus sequence and position are reported (Table 1). The present study revealed a high level of conservation of regulatory elements conferring endosperm-specific expression between sequences from CS and those isolated in this A. speltoides study. Further studies should involve large number of Regulatory motifs identified in the

(AA), accessions as well the characterization of the proximal and distal ends of the typical PDI promoter in diploid and tetraploid progenitor from diverse geographic Their consensus sequence and position are also reported. Skn-1-like ACGAC T. urartu Table 1. a Regulatory motifs ConsensusTATA box sequence Position from ATG Consensus TATTAAA sequence Position from ATG Consensus sequence Position from ATG GCN-4 4-likeACGT- box CGACTCA,CACGTG CATGTCA NA, not applicable. GTACGTG CACGTG AACA AACA PROLAMINE box TGAAAAGT origin of the world. PDI promoter analysis in wild species of wheat 341 References Shewry PR, Halford NG, Tatham AS, Popineau Y, Lafiandra D and Belton PS (2003) The high molecular weight subunits of Ciaffi M, Paolacci AR, Dominici L, Tanzarella OA and Porceddu E wheat glutenin and their role in determining wheat processing (2001) Molecular characterization of gene sequences properties. Advances in Food and Nutrition Research 45: coding for protein disulfide isomerase (PDI) in durum 219–302. wheat (Triticum turgidum ssp. durum). Gene 265: 147–156. Shewry PR, Underwood C, Wan Y, Lovegrove A, Bhandari D, Ciaffi M, Paolacci AR, d’Aloisio E, Tanzarella OA and Porceddu E Toole G, Mills ENC, Denyer K and Mitchell RAC (2006) Cloning and characterization of wheat PDI (protein (2009) Storage product synthesis and accumulation in disulfide isomerase) homoeologous genes and promoter developing grains of wheat. Journal of Cereal Science 50: sequences. Gene 366: 209–218. 106–112. Dvorak J, Luo MC, Yang ZL and Zhang HB (1998) The structure Shimoni Y, Zhu X, Levanoy H, Segal G and Galili G (1995) Puri- of the A. tauschii genepool and the evolution of hexaploid fication, characterization, and intracellular localization of wheat. Theoretical and Applied Genetics 97: 657–670. glycosylated protein disulfide isomerase from wheat Higo K, Ugawa Y, Iwamoto M and Higo H (1999) PLACE: a data- grains. Plant Physiology 108: 327–335. base of plant cis-acting regulatory DNA elements. Nucleic Takemoto Y, Coughlan SJ, Okita TW, Hikaru S, Masahiro O and Acids Research 26: 358–359. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan Tohihiro K (2002) The rice mutant esp2 greatly accumulates PA, McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez the glutenin precursor and deletes the protein disulfide R, Thompson JD, Gibson TJ and Higgins DG (2007) Clustal isomerase. Plant Physiology 128: 1212–1222. W and Clustal X version 2.0. Bioinformatics 23: 2947–2948. Wu CY, Suzuki A, Washida H and Takaiwa F (1998) The GCN4 Li CP and Larkins BA (1996) Expression of protein disulfide iso- motif in a rice glutelin gene is essential for endosperm- merase is elevated in the endosperm of the maize floury-2 specific gene expression and is activated by Opaque-2 in mutant. Plant and Molecular Biology 30: 873–882. transgenic rice plants. Plant Journal 14: 673–683. Librado P and Rozas J (2009) DnaSP v5: a software for Wu CY, Washida H, Onodera Y, Harada K and Takaiwa F comprehensive analysis of DNA polymorphism data. (2000) Quantitative nature of the Prolamin-box, ACGT Bioinformatics 25: 1451–1452. and AACA motifs in a rice glutelin gene promoter: minimal Shewry PR and Tatham AS (1997) Disulphide bonds in wheat cis-element requirements for endosperm-specific gene gluten proteins. Journal of Cereal Science 25: 207–227. expression. Plant Journal 23: 415–421. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 342–346 ISSN 1479-2621 doi:10.1017/S1479262111000232

Protein disulphide isomerase family in bread wheat (Triticum aestivum L.): genomic structure, synteny conservation and phylogenetic analysis

E. d’Aloisio1, A. R. Paolacci2, A. P. Dhanapal1, O. A. Tanzarella2, E. Porceddu2 and M. Ciafﬁ2* 1Scuola Superiore Sant’Anna, Pisa, Italy and 2Dipartimento di Agrobiologia e Agrochimica, Universita` della Tuscia, Viterbo, Italy

Abstract Eight genes encoding protein disulphide isomerase (PDI)-like proteins in bread wheat were cloned and characterized and their genomic structure was compared with that of homoeologous genes isolated from other plant species. Fourteen wheat cDNA sequences of PDI-like genes were amplified and cloned; eight of them were relative to distinct PDI-like genes, whereas six corresponded to homoeologous sequences. Also, the genomic sequences of the eight non-homoeologous genes were amplified and cloned. Phylogenetic analysis, which included eight genes encoding PDI-like proteins and the gene encoding the typical PDI, assigned at least one of them to each of the eight major clades identified in the phylogenetic tree of the PDI gene family of plants. The close chromosome synteny between wheat and rice was confirmed by the location of the homoeologous genes of the PDI family in syntenic regions of the two species. Within the same phylogenetic group, a high level of conservation, in terms of sequence homology, genomic structure and domain organization, was detected between wheat and the other plant species. The high level of conservation of sequence and genomic organization within the PDI gene family, even between distant plant species, might be ascribed to the key metabolic roles of their protein products.

Keywords: chromosome synteny; genomic structure; phylogenetic analysis; protein disulphide isomerase; Triticum aestivum

Introduction Cysteine, Glycine, Histidine, Cysteine) and for presence/ absence of specific domains and of the Lysine, Aspartic The protein disulphide isomerase (PDI) gene family acid, Glutamic acid, Leucine (KDEL) signal of retention includes the ‘typical PDI’, which catalyzes formation, in the ER. They are involved in folding and deposition of reduction and isomerization of disulphide bonds in seed storage proteins in several species (Li and Larkins, secretory proteins within the lumen of the endoplasmic 1996; Takemoto et al., 2002). Since wheat flour quality is reticulum (ER). The proteins encoded by the genes of the strongly affected by composition and structure of seed sto- plant PDI gene family cluster into eight phylogenetic rage proteins, the potential involvement of proteins of the classes (Houston et al., 2005) and differ for number and wheat PDI family in their folding and in the formation of position of the active thioredoxin-like site (CGHC: intra- and inter-molecular disulphide bonds makes their study particularly interesting. Genomic, cDNA and promoter sequences of the three homoeologous gene encoding the ‘typical’ PDI have been cloned and characterized * Corresponding author. E-mail: ciaffi@unitus.it (Ciaffi et al., 2006). Here, we report the isolation and Organization and phylogenesis of wheat PDI genes 343 characterization of eight new non-homoeologous genes at least one wheat gene had been cloned for each coding for PDI-like proteins, their assignment to group (Fig. 1). On the basis of the modular structure of the eight phylogenetic groups of the plant PDI family, the proteins, six of the eight subfamilies were clustered their chromosome location and the organization of their into two major clades, whereas the proteins of the sub- genomic sequences. families VI and VIII being highly diversified were considered as outgroups (Fig. 1). Since genes from P. patens, monocots and dicots formed three distinct Material and methods sub-clusters within each of the eight PDI phylogenetic groups (Fig. 1), the eight subfamilies would have Single plants of bread wheat cv. Chinese Spring (CS) and its emerged before the divergence of bryophytes and angio- nulli-tetrasomic (NT) lines were used for DNA and RNA sperms. Only three of the five PDI-like encoding genes extractions. The Dana-Faber Cancer Institute (DFCI) from C. reinhardtii were included in plant phylogenetic wheat gene index database (TaGI, version 11) was groups (CrPDI-4 in group V, CrPDI-5 in group VIII and BLAST-searched with the available sequences of PDI-like CrPDI-3 in group VI), indicating that only three PDI-like genes of rice (Houston et al., 2005) and the non-redundant genes would be common to both chlorophytes and strep- wheat PDI-like tentative consensus (TC) sequences ident- tophytes, which diverged over 1 billion years ago. The ified were used as template for 50 and 30 Rapid Amplifica- presence of multiple genes of the same species within tion of cDNA Ends (RACE) extension. Gene-specific single phylogenetic groups can be explained by dupli- primer pairs, designed on the 50 and 30 Un Translated cation events occurred either after the separation of the Region (UTR) sequences, were used for cDNA and geno- angiosperms from the briophytes or later, after the diver- mic amplifications, followed by cloning and sequencing sification of monocots and dicots. It is noteworthy that of the resulting amplicons. The evolutionary relationships only group VII hosted two paralogous genes of wheat between PDI and PDI-like genes of wheat and other plants not related to the allopolyploid origin of its genome. were studied by phylogeny reconstruction based on the The eight non-homoeologous genomic sequences alignment of 92 deduced amino-acid sequences of the fol- encoding the novel PDI-like proteins were amplified by lowing genes: 9 of wheat, 13 of Arabidopsis, 12 of poplar, the same primer combinations previously used to clone 10 of grapevine, 5 of soybean, 12 each of maize and rice, the cDNA sequences; their lengths were: (1) TaPDIL2-1: 14 of Physcomitrella patens and 5 of Chlamydomonas 5213 bp; (2) TaPDIL3-1: 4294 bp; (3) TaPDIL4-1: 3820 bp, reinhardtii. The chromosome location of the newly ident- (4) TaPDIL5-1: 5326 bp; (5) TaPDIL6-1: 2162 bp; (6) ified PDI-like gene sequences was determined through TaPDIL7-1: 2887 bp; (7) TaPDIL7-2: 2475 bp; (8) Southern analyses of CS and its NT lines using digoxi- TaPDIL8-1: 7034 bp. The exon–intron structure was genin-labelled probes. determined by their alignment with the corresponding cDNA sequences, which showed an almost perfect nucleotide match between cDNAs and exonic regions of the Results and discussion genomic sequences. The eight genes showed a complex genomic organization with following exon numbers: 12 The BLAST search using 12 PDI-like gene sequences of in TaPDIL2-1 and TaPDIL3-1;11inTaPDIL4-1;9in rice in the DFCI wheat gene index database fetched TaPDIL5-1;4inTaPDIL6-1;5inTaPDIL7-1 and nine TC sequences, one of them encoding the typical TaPDIL7-2;15inTaPDIL8-1 (Supplementary Fig. S1, PDI, whose three homoeologous genes had previously available online only at http://journals.cambridge.org). been cloned and characterized (Ciaffi et al., 2006). The Consistent with previous studies (Kersanach et al., 1994; eight TC sequences encoding PDI-like genes (Table 1) Petersen et al., 2006), the genes of wheat, Arabidopsis, were used as RACE template to isolate their 50 and 30 rice and P. patens clustering into the same phylogenetic extensions, subsequently validated by sequence analysis group revealed a high level of conservation of their struc- and the corresponding full-length cDNAs were cloned tural features (exon–intron pattern and number, size and by RT-PCR of RNA from various wheat tissues using position of the protein active sites; Supplementary Fig. S1, specific primer pairs designed in the 50 and 30 UTRs. Four- available online only at http://journals.cambridge.org), teen wheat cDNA sequences encoding PDI-like proteins whereas the intron–exon structure of the genes of the were identified by sequence analysis; eight of them alga C. reinhardtii was very different (data not shown). derived from distinct genes, whereas six corresponded The genes encoding the typical PDI had been to homoeologous sequences (Table 1). located in chromosome arms 4AL, 4BS and 4DS of Phylogenetic analysis included the nine non- bread wheat (Ciaffi et al., 2006). The chromosome homoeologous sequences of the wheat PDI family into locations of the eight wheat genes encoding PDI-like the eight phylogenetic groups identified in plants, then proteins, which were determined by Southern analysis 344

Table 1. Characteristics of the full-length cDNA sequences coding for wheat PDI-like proteins cloned in this study and their chromosome location

T. aestivum cv. CS O. sativa Full-length cDNA Orthologous rice gene TC sequence UTR50 UTR30 Open Reading (DFCI wheat Chromosome Previous Acc. number Chromosome Clonea (nt) (nt) Frame (ORF) (nt) gene index) locationb namec This study of cDNA Protein identity locationb TaPDIL2-1a 63 109 1767 TC301880 W6 OsPDIL1-4 OsPDIL2-1 AK071514 408/561 (72.73%) R2 TaPDIL3-1a 141 110 1626 TC353685 W7 OsPDIL1-5 OsPDIL3-1 AK073970 437/529 (82.61%) R6 TaPDIL4-1a 74 139 1104 TC300461 W1 OsPDIL2-1 OsPDIL4-1 AK103944 316/366 (86.34%) R5 TaPDIL4-1b 74 138 1104 TC300461 W1 OsPDIL2-1 OsPDIL4-1 AK103944 317/366 (86.61%) R5 TaPDIL5-1a 55 150 1323 TC317379 W5 OsPDIL2-3 OsPDIL5-1 AK062254 391/439 (89.07%) R9 TaPDIL5-1b 55 151 1323 TC317379 W5 OsPDIL2-3 OsPDIL5-1 AK062254 397/439 (90.43%) R9 TaPDIL6-1a 13 240 456 TC294820 W4 OsPDIL5-1 OsPDIL6-1 AK063663 121/146 (82.88%) R3 TaPDIL6-1b 13 255 450 TC294820 W4 OsPDIL5-1 OsPDIL6-1 AK063663 119/146 (81.51%) R3 TaPDIL7-1a 37 289 1242 TC287269 W2 OsPDIL5-2 OsPDIL7-1 AK069367 355/411 (86.37%) R4 TaPDIL7-1b 37 289 1254 TC287269 W2 OsPDIL5-2 OsPDIL7-1 AK069367 359/416 (86.30%) R4 TaPDIL7-1c 37 289 1242 TC287269 W2 OsPDIL5-2 OsPDIL7-1 AK069367 356/411 (86.62%) R4 TaPDIL7-2a 53 186 1257 TC287749 W6 OsPDIL5-3 OsPDIL7-2 ND 311/416 (74.76%) R2 TaPDIL7-2b 53 186 1257 TC287749 W6 OsPDIL5-3 OsPDIL7-2 ND 311/416 (74.76%) R2 TaPDIL8-1a 122 226 1458 TC301351 W2 OsPDIL5-4 OsPDIL8-1 AK099660 450/485 (92.78%) R7 ND, no cDNA sequence available, tBLASTn searches of Gramene identified predicted transcript GRMT00000163510 (OsPDIL5-3/OsPDIL7-2) and the amino acid sequence was predicted from the available genomic sequence Os02g34530 (Houston et al., 2005). a A code of two letters (Ta ¼ T. aestivum) followed by the suffix PDIL and by an Arabic number indicating the corresponding phylogenetic group was assigned to each sequence. Multiple sequences clustering into the same subfamily were designed by an additional number (1 and 2) and putative homoeologous gene sequences were distinguished with an additional letter (a–c). Corresponding TC sequences, identified in DFCI wheat gene index database, orthologous rice genes and their chromosomal location have also been reported. b The seven wheat homoeologous groups (w1–w7) are syntenic to the 12 rice chromosomes (r1–r12), in particular w1 ¼ r5 þ r10, w2 ¼ r7 þ r4, w3 ¼ r1, w4 ¼ r3 þ r11, w5 ¼ r12 þ r9 þ r3, w6 ¼ r2 and w7 ¼ r6 þ r8 (La Rota and Sorrells, 2004). c Nomenclature used by Houston et al. (2005). .d’Aloisio E. tal. et Organization and phylogenesis of wheat PDI genes 345

CrPDI-2 PpPDIL4-1 53 VvPDIL4-1 71 AtPDIL4-1 48 GmPDIL4-2 100 GmPDIL4-1 PtPDIL4-1 IV PtPDIL4-2 86 ZmPDIL4-2 OsPDIL4-1 TaPDIL4-1 OsPDIL4-2

ZmPDIL4-1 Clade II 100 CrPDI-4 97 PpPDIL5-1 VvPDIL5-1 100 GmPDIL5-1 PtPDIL5-1 AtPDIL5-1 V 75 AtPDIL5-2 TaPDIL5-1 73 OsPDIL5-1 ZmPDIL5-1 PpPDIL7-1 98 PtPDIL7-1 100 98 PtPDIL7-2 VvPDIL7-1 AtPDIL7-1 100 OsPDIL7-1 VII 100 TaPDIL7-1 ZmPDIL7-1 97 ZmPDIL7-2 98 OsPDIL7-2 TaPDIL7-2 CrRB60 AtPDIL2-2 97 GmPDIL2-1 PtPDIL2-1 89 VvPDIL2-1 93 VvPDIL2-2 AtPDIL2-1 100 100 TaPDIL2-1 99 OsPDIL2-1 II ZmPDIL2-1 ZmPDIL2-2 PpPDIL2-4 PpPDIL2-1 98 PpPDIL2-2 PpPDIL2-6 86 59 PpPDIL2-3 PpPDIL2-5 AtPDIL3-1 Clade I 100 AtPDIL3-2 PtPDIL3-1 100 VvPDIL3-1 TaPDIL3-1 III 100 OsPDIL3-1 ZmPDIL3-1 100 PpPDIL3-1 PpPDIL3-2 PpPDIL1-1 TaPDIL1-1 62 100 99 OsPDIL1-1 ZmPDIL1-1 ZmPDIL1-2 100 OsPDIL1-2 100 OsPDIL1-3 AtPDIL1-1 I AtPDIL1-2 VvPDIL1-2 PtPDIL1-1 84 PtPDIL1-2 VvPDIL1-1 GmPDIL1-1 CrPDI-5 100 PpPDIL8-1 65 AtPDIL8-1 98 AtPDIL8-2 VvPDIL8-1 PtPDIL8-1 VIII PtPDIL8-2 78 100 OsPDIL8-1 TaPDIL8-1 ZmPDIL8-1 CrPDI-3 67 PpPDIL6-1 91 AtPDIL6-1 PtPDIL6-1 100 TaPDIL6-1 VII ZmPDIL6-1 94 OsPDIL6-1 VvPDIL6-1 0.1 Fig. 1. Phylogenetic tree based on deduced amino-acid sequences of 92 PDI and PDI-like genes: nine of wheat (Ta), 13 of Arabidopsis (At), 12 of poplar (Pt), 10 of grapevine (Vv), 5 of soybean (Gm), 12 each of maize (Zm) and rice (Os), 14 of P. patens (Pp) and 5 of C. reinhardtii (Cr). Multiple alignment was performed by ClustalX 1.83 software using the Gonnet series as protein weight matrix and parameters set to three gap open penalty, 1.6 gap extension penalty, negative matrix on and divergent sequence delay set at 36% followed by manual adjustment. The phylogenetic tree was constructed by NEIGHBOR (PHYLIP 3.6). Distance matrices were estimated by PRODIST using the Percent Accepted Mutation (PAM) model of amino acid transition. To evaluate statistical significance, 1000 bootstrap replicates were generated by SEQBOOT. Numbers on main branches indicate bootstrap percentages for 1000 replicates. The most divergent PDI-like sequences of groups VI and VIII were set as outgroups. The major clades (I and II) and the eight phylogenetic groups (I–VIII) are enclosed by curly and square brackets, respectively. 346 E. d’Aloisio et al. of CS and its aneuploid NT lines, were compatible References with those of their rice orthologs, as expected on the basis of the main chromosome synteny between rice Ciaffi M, Paolacci AR, D’Aloisio E, Tanzarella OA and Porceddu E and wheat (Table 1). In fact, the seven wheat homoeo- (2006) Cloning and characterization of wheat PDI (protein logous groups (w1–w7) are syntenic to the 12 rice disulfide isomerase) homoeologous genes and promoter chromosomes (r1–r12): w1 ¼ r5 þ r10, w2 ¼ r7 þ r4, sequences. Gene 366: 209–218. Houston NL, Fan C, Xiang QY, Schulze JM, Jung R and Boston ¼ ¼ þ ¼ þ þ ¼ w3 r1, w4 r3 r11, w5 r12 r9 r3, w6 r2 RS (2005) Phylogenetic analyses identify 10 classes of the and w7 ¼ r6 þ r8 (La Rota and Sorrells, 2004), even protein disulfide isomerase family in plants, including though their collinearity has often been disrupted by single-domain protein disulfide isomerase-related proteins. extensive duplications and rearrangements of chromo- Plant Physiology 137: 762–778. some segments. The outputs of the sequence-based Kersanach R, Brinkmann H, Liaud MF, Zhang DX, Martin W and macro-collinearity between the Michigan State University Cerff R (1994) Five identical intron positions in ancient dupli- (MSU) rice pseudomolecules (Release 6.1) and the wheat cated genes of eubacterial origin. Nature 367: 387–389. La Rota M and Sorrells ME (2004) Comparative DNA sequence Expressed Aequence Tags (ESTs), whose locations on the analysis of mapped wheat ESTs reveals the complexity of wheat Bin map is reported in GrainGenes, were genome relationships between rice and wheat. Functional exploited to assess the syntenic relationships between and Integrative Genomics 4: 34–46. the regions of the wheat and rice chromosomes flanking Li CP and Larkins BA (1996) Expression of protein disulfide iso- the PDI-like encoding genes (Supplementary Table S1, merase is elevated in the endosperm of the maize floury-2 available online only at http://journals.cambridge.org). mutant. Plant and Molecular Biology 30: 873–882. Besides confirming the chromosome locations, the Petersen J, Teich R, Brinkmann H and Cerff R (2006) A “green” analysis based on Wheat Bin Mapped Markers assigned phosphoribulokinase in complex algae with red plastids: evidence for a single secondary endosymbiosis leading to the PDI-like genes to specific wheat chromosome arms. haptophytes, cryptophytes, heterokonts and dinoflagel- The close syntenic relationships between wheat and rice lates. Journal of Molecular Evolution 62: 143–157. confirmed by this study might be helpful for high-density Takemoto Y, Coughlan SJ, Okita TW, Satoh H, Ogawa M mapping of specific wheat chromosome regions, which and Kumamaru T (2002) The rice mutant esp2 greatly would be useful for chromosome walking and gene accumulates the glutenin precursor and deletes the protein positional cloning. disulfide isomerase. Plant Physiology 128: 1212–1222. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 347–351 ISSN 1479-2621 doi:10.1017/S1479262111000244

Protein disulphide isomerase family in bread wheat (Triticum aestivum L.): protein structure and expression analysis

A. R. Paolacci1, M. Ciafﬁ1*, A. P. Dhanapal2, O. A. Tanzarella1, E. Porceddu1 and E. d’Aloisio2 1Dipartimento di Agrobiologia e Agrochimica, Universita` della Tuscia, Viterbo, Italy and 2Scuola Superiore Sant’Anna, Pisa, Italy

Abstract The deduced amino-acid sequences of 17 protein disulphide isomerase (PDI) and PDI-like cDNAs of wheat assigned to nine homoeologous groups were searched for conserved motives by comparison with characterized sequences in different protein databases. The wheat protein sequences encoded by genes of different homoelogous groups showed a high level of structural similarity with the corresponding protein sequences of other species clustering into the same phylogenetic group. The proteins of ﬁve groups (I–V) share two thioredoxin-like active domains and show structural similarities with the corresponding proteins of higher eukaryotes, whereas those of the remaining three groups (VI–VIII) contain a single thioredoxin-like active domain. The expression analysis of the nine non-homoeologous wheat genes, which was carried out by quantitative RT-PCR in developing caryopses and in seedlings subjected to temperature stresses, showed their constitutive although highly variable transcription rate. The comprehensive structural and transcriptional characterization of the PDI and PDI-like genes of wheat performed in this study represents a basis for future functional characterization of the PDI gene family in the hexaploid context of bread wheat.

Keywords: expression analysis; protein disulphide isomerase; protein structure; temperature stress; Triticum aestivum

Introduction thioredoxin and contain the catalytic site (CXXC), whereas the b-type domains lack the active tetrapeptide, The protein disulphide isomerase (PDI) family includes but have a secondary structure similar to that of several genes whose products are responsible for thioredoxin. diversified metabolic functions, including secretory Phylogenetic analysis resolved the PDI family into protein folding, chaperone activity and redox signalling. eight groups (Houston et al., 2005), five with two PDI and PDI-like proteins differ for number and position TRX-like active domains (I–V) and three with only one of their thioredoxin-like active (type a) and inactive (VI–VIII). The purpose of this research was to character- (type b) domains, for presence/absence of other domains ize the genes of the PDI family in wheat and to compare and of the Lysine. Aspartic acid, Glutamic acid, Leucine the structure of their deduced amino-acid sequences (KDEL) retention signal in the endoplasmic reticulum and their expression with those of homoeologous (ER). The a-type domains are homoeologous to that of genes of other plant species. Former studies in wheat had been limited to the genes encoding the typical PDI (Ciaffi et al., 2006), which is of special interest for its potential involvement in determining the technological * Corresponding author. E-mail: ciaffi@unitus.it properties of flour. 348 A. R. Paolacci et al. Material and methods and Bicknell, 2003). TaPIDL8-1 has two transmembrane segments; the C-terminal segment is part of the DUF1692 Developing caryopses from 5 to 38 d after anthesis at domain, found in several proteins. ERGIC-32 is one of 5–6 d intervals (seven samplings) were collected from them (Breuza et al., 2004), which is localised in the 20 bread wheat plants (Triticum aestivum cv. Chinese ER–Golgi intermediate compartment (ERGIC), the first Spring). For temperature stress experiments, 20-d-old anterograde/retrograde sorting station in the mammalian CS seedlings were exposed to either 33 or 48C for 48 h; secretory pathway. Protein sorting seems only one of the each treatment included two biological replications of ERGIC functions. Other presumed functions are less 20 plants. Shoots were harvested just before (control), clear; it appears involved in protein quality control and after 24 h and after 48 h of temperature stress, frozen in in the retrotranslocation to the cytosol of permanently mis- liquid nitrogen and stored at 2808C. folded proteins. Further studies will be necessary to under- The deduced protein sequences were analysed by stand whether or not in plants the PDI-like proteins searching for conserved motifs in CDD, InterPro and belonging to the VIII phylogenetic group are localized in SMART databases; their subcellular locations were pre- a compartment similar to the ERGIC of mammalian and dicted by Target P1.1 and ChloroP 1.1, the presence of to investigate their function. the signal peptide was confirmed by Signal P3.0 and The functional features of the proteins encoded by the the transmembrane regions were determined by PDI gene family of wheat can be predicted on the basis TMHMM version 2.0; protein identity was estimated of their domain structure and of the presence/absence of using DNAMAN software. Quantitative RT-PCR (qRT- three major determinants whose role has been identified PCR) analyses and data normalization were performed in humans (Ellgaard and Ruddock, 2005). Besides the according to Paolacci et al. (2009). Two biological repli- active site tetrapeptide -CXHC-, typically -CGHC-, there cates, resulting from two different RNA extractions and are two additional prominent determinants for the qRT-PCR reactions, were used in quantification analysis; activity of the PDI-family proteins. A conserved arginine moreover, three technical replicates were analysed for modulates the pKa of the active-site cysteines and is each biological replicate. involved in the timing, allowing a single catalyst to act as isomerase and oxidase and facilitating the release of non-productive folding substrates. Moreover, the cataly- Results and discussion tic cycle for oxidation or reduction requires numerous proton transfer reactions; in the thioredoxins, a charged Structural characteristics and domain organization of the glutamic acid–lysine pair, located under the active site, deduced amino-acid sequences of three homoeologous is the proton acceptor. Table 1 reports the active-site tet- cDNAs coding for the typical PDI (Ciaffi et al., 2006) and rapeptide and the positions of conserved charge pair of 14 PDI-like cDNAs, of nine different homoeologous sequence and arginine residues determined on the groups are described in Table 1 and Supplementary Fig. basis of multiple alignments of the a-type domains S1 (available online only at http://journals.cambridge. of wheat PDI-like proteins and human typical PDI org). Except TaPDIL8-1, all proteins contained a putative (Supplementary Fig. S2, available online only at http:// signal peptide for translocation into the ER. TaPDIL2-1 journals.cambridge.org). and TaPDIL3-1 showed a multidomain organization similar The transcription level of the nine genes of the wheat to the typical PDI, with the addition of an N-terminal PDI family was investigated in a series of seed develop- domain (c) rich of acidic residues and a putative calcium- mental stages (Fig. 1(a–f)) from cellularisation to physio- binding site. TaPDIL4-1 had a D domain consisting of a logical maturity, showing a considerable variation C-terminal a-helical of about 100 amino acids with between the genes. The transcripts of some PDI genes unknown function and a potential ER-translocation containing two active thioredoxin domains, such as signal. Like all proteins of the fourth group, it lacked the TaPDIL1-1, TaPDIL2-1, TaPDIL4-1 and TaPDIL5-1, ER-retention signal; consequently, it might be targeted to increased dramatically during the early stages of seed a different subcellular location or be retained as part of development (5–10 d post anthesis; DPA), then declined an heteromeric complex; however, in D. discoideum, the steadily during the major part of the grain filling period C-terminal part of the D domain is responsible for the ER (15–27 DPA). Consequently, their temporal expression retention of the PDI-D protein (Monnat et al., 2000). The preceded that of seed storage protein genes and reached proteins of the seventh group have a transmembrane seg- a maximum before the more intense synthesis of gluten ment which, even without the KDEL signal, could retain proteins. For TaPDIL1-1, TaPDIL4-1 and TaPDIL5-1,itis the protein in the ER by anchoring it to the membrane; in noteworthy that a second peak of transcription was man, Drosophila and C. elegans, all proteins with similar detected by qRT-PCR between 27 and 32 DPA, when the structures are involved in developmental control (Clissold deposition of seed storage reserves decreased and seeds rti tutr n xrsino ha D genes PDI wheat of expression and structure Protein

Table 1. Characteristics of the wheat PDI and PDI-like proteins

N-glycosilation Domain Active Conserved charge Conserved Name Length Mw pI sites (putative) compositiona site sequence pair sequenceb Argb TaPDIL1-1a 515 56.59 4.99 1:N283 a-b-b0-a0 CGHC, CGHC E62-K96, E406-K439 R136, R475 TaPDIL1-1b 512 56.44 5.03 1:N283 a-b-b0-a0 CGHC, CGHC E62-K96, E406-K439 R136, R475 TaPDIL1-1c 515 56.63 4.96 1:N283 a-b-b0-a0 CGHC, CGHC E62-K96, E406-K439 R136, R475 TaPDIL2-1 588 63.80 4.61 2:N109,N212 c-a-b-b0-a0 CGHC, CGHC E123-K157, E464-K497 R193, R535 TaPDIL3-1 541 59.61 4.95 1:N150 c-a-b-b0-a0 CERS, CVDC L90- K124, E429-R462 R160, P492 TaPDIL4-1a 367 40.27 6.17 0 a8-a-D CGHC, CGHC E54-K87, E173-K211 R125, R244 TaPDIL4-1b 367 40.26 6.17 0 a8-a-D CGHC, CGHC E54-K87, E173-K211 R125, R244 TaPDIL5-1a 440 47.15 5.12 1:N170 a8-a-b CGHC, CGHC E51-K89, E188-K226 R119, R257 TaPDIL5-1b 440 47.22 5.36 1:N170 a8-a-b CGHC, CGHC E51-K89, E188-K226 R119, R257 TaPDIL6-1a 151 16.99 4.96 0 a CKHC Q56-S95 R126 TaPDIL6-1b 149 16.65 5.30 0 a CKHC Q54-S93 R124 TaPDIL7-1a 413 46.30 4.91 1:N275 a-b-b0-t CGHC D56-K90 R126 TaPDIL7-1b 417 46.62 4.91 2:N176,N279 a-b-b0-t CGHC D60-K94 R130 TaPDIL7-1c 413 46.32 4.87 2:N172,N275 a-b-b0-t CGHC D56-K90 R126 TaPDIL7-2a 418 46.40 5.12 0 a-b-b0-t CGHC D64-K98 R134 TaPDIL7-2b 418 46.34 5.03 1:N384 a-b-b0-t CGHC D64-K98 R134 TaPDIL8-1 485 54.41 6.90 ND t-a-t CYWS N164-K203 R249 a a, Active site containing the thioredoxin-like domain; b, inactive thioredoxin-like domain (superscript prime and degree symbols are included to distinguish multiple a and b domains on the basis of their position and not of sequence homology); c, acidic segment; D, Erp29c domain; t, transmembrane domain. b Positions of conserved charge pair sequence and arginine residues important for the catalytic activity of different proteins of the human PDI family were determined on the basis of multiple alignments of the a-type domains of wheat PDI-like proteins and human classical PDI (Accession number P07237) (Supplementary Fig. S2, available online only at http://journals.cambridge.org). ND, not determined because TaPDIL8-1 lacks a putative N-terminal signal peptide. Proteins without signal peptides are unlikely to be exposed to N-glycosilation machinery and thus may not be glycosylated in vivo even though they contain potential motifs. 349 350 A. R. Paolacci et al. started desiccating. Seemingly, the proteins encoded by TaPDIL4-1and TaPDIL8-1 was not affected by tempera- these genes may play an important role in later stages of ture treatments, whereas that of TaPDIL2-1 was signifi- seed development, when there is a dramatic increase of cantly down-regulated by both cold and heat stresses large gluten polymers stabilized by the formation and/or and that of TaPDIL7-2 was about twice lower in shoots the rearrangement of inter-chain disulphide bonds. exposed to low temperatures (Fig. 1(g)). A significant Analyses by qRT-PCR of six seedling samples induction was observed for at least one of the two consisting of two temperature treatments (4 and 338C) temperature treatments for the remaining five PDI-like each for 24 and 48 h and their controls (Fig. 1(g) and genes (Fig. 1(h)). The most significant variation in tran- (h)) indicated that the nine genes were differentially scription rate induced by temperature treatments was regulated by cold and heat stresses. The expression of observed for TaPDIL1-1 (typical PDI) and TaPDIL5-1.

(a)30 (b) 1.0×106

25 8.0×105 20 6.0×105 15 4.0×105 10 cDNA copies

Relative quantity Relative 5 PDIL1-1 5 2.0×10 PDIL2-1 0 0.0 (c)5 (d) 4.0×104 3.5×104 4 3.0×104 4 3 2.5×10 2.0×104 2 4 1.5×10 PDIL3-1

cDNA copies 4

Relative quantity Relative 1.0×10 1 PDIL7-1 5.0×103 PDIL7-2 0 0.0 (e)6 (f) 1.4×105 5 5 1.2×10 5 4 1.0×10 8.0×104 3 PDIL4-1 6.0×104 2 4 PDIL5-1 cDNA copies 4.0×10 Relative quantity Relative PDIL6-1 1 2.0×104 PDIL8-1 0 0.0 5 101521273238 5 101521273238 DPA DPA (g)2.5 (h) 7 PDIL2-1 6 PDIL4-1 2 PDIL7-2 5 PDIL8-1 1.5 4 3 PDIL1-1 1 PDIL3-1 2

Relative quantity Relative PDIL5-1 Relative quantity Relative 0.5 1 PDIL6-1 0.0 0 PDIL7-1 Contr 4°C/24 h4°C/48h Contr 33°C/24 h 33°C/48h Contr 4°C/24 h4°C/48h Contr 33°C/24 h 33°C/48h LT HT LT HT Fig. 1. Expression analysis of PDI and PDI-like genes in developing wheat caryopses and in wheat seedlings exposed to high- (HT) and low-temperature (LT) treatments. Relative (a, c and e) and absolute (b, d and f) quantification of the expression level in developing caryopses collected between 5 and 38 DPA of nine PDI and PDI-like genes. Relative (g and h) quantification of the expression level in six samples consisting of seedlings exposed to two temperature treatments (4 and 338C) for 24 and 48 h and their controls of nine PDI and PDI-like genes. The 14 cDNA pools (two biological replicates, seven developing caryopsis samples; a–f) and the 12 cDNA pools (two biological replicates, six seedling samples; g and h) were tested in triplicate and normalized using the geometric average of the relative expression of the two reference genes encoding cell division control protein and ADP-ribosylation factor. Relative expression levels of the nine genes were referred to that of a calibrator set to the value 1, which was represented by the developmental stage (a, c and e) or by the temperature treatments (g and h) with the lowest expression. Absolute expression levels of the nine genes were expressed as number of cDNA copies/(g of reverse transcribed total RNA (b, d and f). Normalized values of relative and absolute expressions of the nine genes are given as average ^ SD. Contr, control samples grown at 188C. Protein structure and expression of wheat PDI genes 351 References functional properties. European Molecular Biology Organ- ization Reports 6: 28–32. Breuza L, Halbeisen R, Jeno P, Otte S, Barlowe C, Hong W and Houston NL, Fan C, Xiang QY, Schulze JM, Jung R and Boston RS Hauri HP (2004) Proteomics of endoplasmic reticulum– (2005) Phylogenetic analyses identify 10 classes of the Golgi intermediate compartment (ERGIC) membranes protein disulfide isomerase family in plants, including from brefeldin A-treated HepG2 cells identifies ERGIC-32, single-domain protein disulfide isomerase-related proteins. a new cycling protein that interacts with human Erv46. Plant Physiology 137: 762–778. The Journal of Biological Chemistry 279: 47242–47253. Monnat J, Neuhaus EM, Pop MS, Ferrari DM, Kramer B and Ciaffi M, Paolacci AR, D’Aloisio E, Tanzarella OA and Porceddu E Soldati T (2000) Identification of a novel saturable endo- (2006) Cloning and characterization of wheat PDI (protein plasmic reticulum localization mechanism mediated by the disulfide isomerase) homoeologous genes and promoter sequences. Gene 366: 209–218. C-terminus of a Dictyostelium protein disulfide isomerase. Clissold PM and Bicknell R (2003) The thioredoxin-like fold: Molecular Biology of the Cell 11: 3469–3484. hidden domains in protein disulfide isomerases and other Paolacci AR, Tanzarella OA, Porceddu E and Ciaffi M (2009) chaperone proteins. Bioessays 25: 603–611. Identification and validation of reference genes for quanti- Ellgaard L and Ruddock LW (2005) The human protein tative RT-PCR normalization in wheat. BMC Molecular disulphide isomerase family: substrate interactions and Biology 10: 11. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 352–356 ISSN 1479-2621 doi:10.1017/S1479262111000141

Deployment of either a whole or dissected wild nuclear genome into the wheat gene pool meets the breeding challenges posed by the sustainable farming systems

Ciro De Pace1*, Marina Pasquini2, Patrizia Vaccino3, Marco Bizzarri1, Francesca Nocente2, Maria Corbellini3, Maria Eugenia Caceres4, Pier Giorgio Cionini4, Doriano Vittori1 and Gyula Vida5 1Dipartimento di Agrobiologia e Agrochimica, Universita` degli Studi della Tuscia, Viterbo, Italy, 2Unita` di ricerca per la valorizzazione qualitativa dei cereali (CRA-QCE), Roma, Italy, 3Unita` di ricerca per la selezione dei cereali e la valorizzazione delle varieta` vegetali (CRA-SCV), S. Angelo Lodigiano, Lodi, Italy, 4Dipartimento di Biologia Cellulare e Ambientale, Sez. di Biologia Cellulare e Molecolare, Universita` degli Studi di Perugia, Perugia, Italy and 5Agricultural Research Institute of the Hungarian Academy of Sciences, Martonva`sa`r, Hungary

Abstract Deploying whole and dissected nuclear genome of wild Triticeae species in the homoeologous wheat genetic background through inter-specific hybridization and introgression is a lower cost and effective option to prepare wheat germplasm with unexploited genes for disease resistance and enhanced grain yield and quality traits. The whole nuclear genomes of Dasypyrum villosum (Dv) and T. turgidum var durum have been combined, and an homoploid derivative of the original amphiploid displayed typical ‘farro’ spike morphology, tough rachis and the adaptive traits of Dv such as high resistance to diseases (caused by Tilletia tritici, Blumeria graminis f. sp. tritici, Puccinia triticina and P. graminis f. sp. tritici), heading earliness and fortified caryopses (high protein and micronutrient contents). The dissection of the Dv genome by either ‘Triticum aestivum cv Chinese Spring (CS) £ hexaploid amphiploid’ or ‘(CS £ Dv) £ CS’ hybridization and backcrossing provided wheat introgression breeding lines (IBLs) expressing one or more of the Dv adaptive traits. Molecular analyses revealed that either cryptic or Genomic In-situ Hybridization (GISH) detectable Dv chromatin introgression occurred in those IBLs. The IBLs, after 2 years of low-input field tests and genetic analyses in Italy and Hungary, showed simple inheritance, dominance and stability of the adaptive and disease resistance traits.

Keywords: Dasypyrum gene pool; gene transfer; introgression; wheat germplasm; wild plant species

Introduction and molecular cereal breeding programmes and proper wheat genetic resources are necessary for selecting the In order to meet the predicted global cereal grain suitable breeding lines (Baenziger et al., 2008). demand for the next decades, efficient conventional Many dominant genes for adaptation and trait enhancement have been lost during cereal crop domestication, but they have been retained in the genome of the wild components of the Triticeae gene pools (Dwivedi * Corresponding author. E-mail: [email protected] et al., 2008). In natural habitat, wild Triticeae species Wheat germplasm to meet breeding challenges 353 such as Dasypyrum villosum (Dv), whose genome was Dissection of the Dv genome by [T. aestivum cv exposed to million of years of climatic and environ- ‘Chinese Spring’ (CS) 3 Dv] 3 CS hybridization mental changes, are now expressing increased heading and backcrossing earliness, density stands and plant biomass. Deploying whole and dissected Dv nuclear genome in the homo- We selected six Dv-introgressed wheat breeding lines eologous wheat genetic background through inter- (IBLs), sharing a common CS (2n ¼ 6x ¼ 42; A0A0B0B0 specific hybridization and introgression could be a lower D0D0) genetic background, from a population of 150 cost and effective option to help wheat breeders to aneuploid lines developed at the University of Tuscia, merge and select the proper adapted gene pools to Viterbo, Italy, from the backcross (CS £ Dv) £ CS sustain the needed yearly grain yield increase. In this (Supplementary Fig. S2, available online only at study, we show that combining the Dv genome with http://journals.cambridge.org). The IBLs CS £ V63 and the T. turgidum var durum genomic background and CS £ V32 contained a disomic addition of 6V and a deploying dissected Dv genome in Triticum aestivum disomic substitution 6V(6B), respectively, and CS_1B-1V provided wheat genetic resources with new trait line contained a pair of 1BL-1VS chromosomes from a enhancements. spontaneous exchange between chromosome 1B of wheat and 1V of Dv. The IBLs CS £ V58, CS £ V59 and CS £ V60 were morphologically different from CS, Materials and methods although they did not contain apparent GISH detectable V chromatin (Minelli et al., 2005; Caceres et al., 2008). Combining the whole nuclear genomes of Dv and The IBLs were evaluated for resistance to Blumeria T. turgidum var. durum graminis f. sp. tritici (Bgt), Puccinia triticina (Pt) and P. graminis f. sp. tritici (Pgt) isolates in Italy and A hulled and brittle rachis Dv ecotype (2n ¼ 2x ¼ 14; VV) Hungary, and for end-use grain quality traits. collected near Bari (Puglia, Italy) was used as pollen parent in hybridization with the free-threshing and tough rachis T. turgidum var. durum cv ‘Modoc’ Dissection of the Dv genome by hybridization of (2n ¼ 4x ¼ 28; AABB) (Jan et al., 1986). The resulting T. aestivum cv CS and M 3 V-nb1 hexaploid hexaploid amphiploid (2n ¼ 42; AABBVV), labelled amphiploid M £ V-b1, was fertile and showed brittle rachis (Sup- plementary Fig. S1, available online only at http:// An F2-like breeding population with broad genetic journals.cambridge.org). A non-brittle rachis mutated diversity was obtained from selfing the F1 plants obtained amphiploid plant (M £ V-nb1; AABBVV; 2n ¼ 6x ¼ 42) after crossing the hexaploid amphiploid M £ V-nb1 to was discovered in the M £ V-b1 plot grown in 1987 (De CS (Supplementary Fig. S3, available online only at Pace et al., 2003). In the following years, other genetically http://journals.cambridge.org). After two generations of stable variants (‘Mut 7-04’, ‘Mut 12-04’ and ‘Mut 16-04’) selfing, three F4 lines, named ‘8-1’, ‘41-3’ and ‘Mut 3-04’, were found among the M £ V-nb1 plants. were tested for two consecutive growing seasons

Table 1. Mean values for heading time, plant height, number of spikelets/spike, 1000 kernel weight and protein content in M £ V-nb1 hexaploid amphiploid and derived variants by spontaneous mutation in comparison with ‘farro’ species used as controlsa

Heading time Plant height Number of 1000 Kernel Protein content Entry (days from 1st April) (cm) spikelets/spike weight (g) (% dry weight) Hexaploid amphiploid M £ V-nb1 114 101 14.7 27.1 21.1 Derived variants from spontaneous mutation in the hexaploid amphiploid Mut 7-04 116 105 14.6 23.5 21.5 Mut 12-04 124 104 16.3 42.6 16.5 Mut 16-04 124 106 15.9 42.6 16.9 Control ‘farro’ species T. monococcum 144 109 17.7 20.0 20.0 T. dicoccoides 132 125 15.5 35.5 22.0 T. spelta 135 124 17.0 39.8 18.8 a Data were taken from plots at S. Angelo Lodigiano (Lodi, Italy) in 2007 and 2008. 354

Table 2. Average values for test weight (TW), protein content (PC), SDS sedimentation volume (SSV), specific SDS sedimentation volume (SSSV ¼ SSV/PC), gluten index, rheological flour dough properties (Brabender farinograph and Chopin alveograph) and bread quality, determined for three inbred breeding lines obtained from a hybridization of T. aestivum cv CS and M £ V-nb1 hexaploid amphiploid Brabender farinograph Chopin alveograph Bread test TW PC SSV Gluten Stab. Degree of Water abs P L W Vol Height Line Year (kg/hl) (%dm) (ml) SSSV index (min) softness (BU) (%) (mm) (mm) P/L ( £ 1024 J) (cm3) (mm) 8-1 2008 74.9 12.6 91.0 7.2 99 18.2 23 57.8 104 128 0.81 400 710 100 2009 79.6 13.8 88.5 6.4 84 17.4 25 63.0 80 130 0.62 286 735 103 41-03 2008 76.4 12.3 89.0 7.2 98 14.6 25 57.7 92 119 0.77 334 685 101 2009 79.8 14.5 87.5 6.1 80 17.8 21 62.6 72 119 0.61 234 770 105 Mut 3-04 2008 77.0 12.6 91.0 7.2 99 13.1 33 58.1 93 136 0.68 375 720 105 2009 80.0 13.8 88.5 6.4 76 17.8 12 62.9 81 119 0.68 266 790 110 Control CS 2008 64.7 12.2 48.0 3.9 32 3.9 93 56.4 54 61 0.89 90 560 84 Bologna 2009 80.4 13.9 86.5 6.2 97 18.5 22 58.5 78 140 0.57 396 720 105 PR22R58 2009 76.4 12.1 87.5 7.2 97 19.0 21 54.5 56 115 0.49 219 695 98 Best check 2009 785 109 Stab., stability; BU, Brabender Units; abs, absorption; P, pression, measures the tenacity of the dough; L, length, measures the extensibility of the dough; W, work, measures the strength of the dough. a The values are mean of plots raised at two locations (S. Angelo Lodigiano and Viterbo) in Italy for 2 years (2008 and 2009). .D Pace D. C. tal. et Wheat germplasm to meet breeding challenges 355 (2007/2008 and 2008/2009), in the field at two sites: available online only at http://journals.cambridge.org) S. Angelo Lodigiano, SAL, near Lodi in northern Italy, expressed simple inheritance, due to Dv-derived Mende- and Tolentino, TOL, near Macerata in central Italy. The lian dominant genes governing resistance response to plots were managed using low-input criteria. The hexa- each of the Bgt, Pt and Pgt pathogen. The high SDS ploid wheat cultivars ‘Bologna’ and ‘PR22R58’ were sedimentation value of CS_1BL-1VS was traced to the used as checks. Heading time (days from 1st April), effect of one high-molecular weight glutenin subunit yield components and rheological properties of flour (1v in Supplementary Fig. S5, available online only at dough were evaluated at both sites. http://journals.cambridge.org) encoded at the Glu-V1 locus in 1VS of CS_1BL-1VS (Vaccino et al., 2010).

Results and discussion Dissection and deployment of the Dv genome by Combining the whole nuclear genomes of Dv and hybridization of T. aestivum cv CS and M 3 V-nb1 T. turgidum var durum hexaploid amphiploid

The hexaploid amphiploid M £ V-nb1 and the derived ‘New’ chromosome assortments were achieved using homoploid variants displayed ‘farro’ traits (tenacious the AABBVV hexaploid amphiploid as a bridge to com- glumes but tough rachis) and the typical adaptive traits bine the A and B genomes from durum wheat; the of Dv, such as high resistance to diseases (caused by A0,B0 and D0 genomes of bread wheat, and the V Tilletia tritici, Bgt, Pt and Pgt), fortified caryopses (þ17 genome of Dv. Tetraploid AA0BB0 lines, hexaploid to þ21% protein contents and .þ25% of Fe and Zn AA0BB0D0D0 lines, and aneuploid AA0BB0 lines with the content, compared with CS) and heading earliness. When addition of one or more V and D0 chromosomes were compared with the conventional ‘farro’ (T. monococcum, obtained (Supplementary Figs. S3 and S6, available T. dicoccoides, and T. spelta), new ‘farro’ types expressed online only at http://journals.cambridge.org). Promising earlier heading and shorter culms (Table 1), and also outcomes of this breeding scheme were the high fre- caryopses (Mut 12-04 and Mut 16-04) with larger size. quency of euploid segregants (Supplementary Fig. S3, available online only at http://journals.cambridge.org), the phenotypic uniformity observed in the progenies Dissection and deployment of the Dv genome by after only two generations of selfing that followed the ‘(CS 3 Dv) 3 CS’ hybridization and backcrossing CS £ (M £ V-nb1) triparental hybridization and the good agronomic values of those lines. The three inbred breed- This type of hybridization favoured the haploidization ing lines, ‘8-1’, ‘41-3’ and ‘Mut 3-04’, had chromosome 0 0 0 of the V genome in the heterohaploid A B D VF1, counting of 2n ¼ 42 and were as good as the best the random assortment (dissection) of V chromosomes check for yield components, grain quality and rheological in 7A0 þ 7B0 þ 7D0 þ 1V gametes and the formation of flour dough properties (Table 2 and Supplementary wheat IBLs with the disomic addition of one of the V Table S1, available online only at http://journals. chromosome or V chromosome arm after backcrossing cambridge.org). to CS. The IBLs CS £ V32 and CS £ V63 contained the In conclusion, it is suggested that a sustainable breed- chromosome 6V#4 (De Pace et al., 2011), conferring ing response to mitigate the severe threat to the world’s multiple disease resistance to virulent strains of Bgt, wheat supply is necessary (Hovmøller et al., 2010). Pt and Pgt. The IBL CS_1BL-1VS expressed superior We evidenced that deploying whole and dissected Dv Sodium Dodecyl Sulfate (SDS) sedimentation and rheolo- nuclear genome in the homoeologous wheat genetic gical properties of flour dough than CS. Other IBLs background, it is possible to prepare wheat germplasm (CS £ V58, CS £ V59 and CS £ V60) had cryptic Dv chro- with unexploited genes for rust disease resistance and matin introgression exhibiting a 27to29 d earliness in enhanced grain yield and quality traits. heading time compared with CS in the field. Using specific primers for several DNA targets in the genome of these lines revealed Dv alleles, but not CS alleles, at two loci (Vrn-A1 and Vrn-B3) involved in the vernaliza- Acknowledgements tion response pathway (Caceres et al., 2008).

F3 progenies derived from the hybridization of two This research was funded by the ‘Ministero delle disomic addition lines for chromosome 6V, CS £ V63 Politiche Agricole Alimentari e Forestali’, Italy, in the (þ6V#4) and CS þ 6V#1 (produced by Sears, 1982, framework of the research project FRUMIGEN (D.M. susceptible to Bgt, Pt and Pgt; Supplementary Fig. S4, 292/7303/05 issued oct. 12th, 2005). 356 C. D. Pace et al. References De Pace C, Vaccino P, Cionini PG, Pasquini M, Bizzarri M and Qualset CO (2011) Dasypyrum. In: Kole C (ed.) Wild Baenziger PS, Graybosch RA, Dweikat I, Wegulo SN, Hein GL Crop Relatives: Genomic and Breeding Resources, vol 1: and Eskridge KM (2008) Outstanding in their field: the Wild Relatives of Cereals. Heidelberg/New York: Springer. phenotype of the 21st century plant breeder. In: Appels Dwivedi SL, Upadhyaya HD, Stalker HT, Blair MW, Bertioli DJ, R, Eastwood R, Lagudah E, Mackay M, McIntyre L and Nielen S and Ortiz R (2008) Enhancing crop gene pools Sharp P (eds) Proceedings of the 11th International with beneficial traits using wild relatives. In: Janick J (ed.) Wheat Genetics Symposium. Available at http://hdl. Plant Breeding Review, vol. 30. New York: John Wiley & handle.net/2123/3325. Sydney: Sydney University Press. Sons, pp. 179–230. Caceres ME, Vaccino P, Corbellini M, Cionini PG, Sarri V, Polizzi Hovmøller MS, Walter S and Justesen AF (2010) Escalating threat E, Vittori D and De Pace C (2008) Flowering earliness in wheat inbred breeding lines derived from T. aestivum of wheat rusts. Science, 329 (no. 5990): 369. Chinese Spring x D. villosum hybridization is not related Jan CC, De Pace C, McGuire PE and Qualset CO (1986) Hybrids to allelic variation at the vernalization loci VRN-A1, and amphiploids of Triticum aestivum (L). and T. turgidum VRN-B1, and VRN-D1. In: Prohens J and Badenes ML (L) with Dasypyrum villosum (L.) Candargy. Zeitschrift fu¨r (eds) Modern Variety Breeding for Present and Future Pflanzenzu¨chtg 96: 97–106. Needs. Proceedings of the 18th EUCARPIA General Minelli S, Ceccarelli M, Mariani M, De Pace C and Cionini PG Congress 9–12 September 2008, Valencia, Spain: Editorial (2005) Cytogenetics of Triticum £ Dasypyrum hybrids Universidad Politećnica de Valencia, pp. 329–334. and derived lines. Cytogenetic and Genome Research 109: De Pace C, Jan CC, Caputi G and Scarascia Mugnozza GT (2003) 385–392. Genetical events occurring during and after Triticum turgi- Sears ER (1982) Activity report. Annual Wheat Newsletter 28: dum var. durum £ D. villosum hybridization recapitulate the population size and time span required for the tran- 121. sition from tetraploid to hexaploid wheat domestication. Vaccino P, Banfi R, Corbellini M and De Pace C (2010) Improv- Proceedings of 10th International Wheat Genetics ing the wheat genetic diversity for end-use grain quality Symposium, vol. 2. Rome, Italy: Istituto Sperimentale per la by chromatin introgression from the wheat wild relative Cerealicoltura, pp. 472–474. Dasypyrum villosum. Crop Science 50: 528–540. q NIAB 2011 Plant Genetic Resources: Characterization and Utilization (2011) 9(2); 357–360 ISSN 1479-2621 doi:10.1017/S1479262111000529

Identiﬁcation of root morphology mutants in barley

Riccardo Bovina1, Valentina Talame`1, Matteo Ferri1, Roberto Tuberosa1, Beata Chmielewska2, Iwona Szarejko2 and Maria Corinna Sanguineti1* 1Department of Agroenvironmental Science and Technology (DiSTA), University of Bologna, Italy and 2Department of Genetics, University of Silesia, Poland

Abstract In this study, a forward-genetics analysis was performed on a portion of TILLMore, a chemically mutagenized population of barley cv. ‘Morex’ (http://www.distagenomics.unibo.it/TILLMore/), to identify root morphology alterations by comparison with ‘Morex’ wild-type. For this purpose, a simple paper-roll approach was performed to identify phenotypic variants at the seedling stage.

The analysis of c. 1000 M4 families allowed us to identify c. 70 lines with altered root morphology. A more accurate phenotypic characterization of a portion of the mutant lines has been performed using stereomicroscopy and a scanning electron microscopy approach.

Keywords: forward-genetics; Hordeum vulgare; root hairs; root morphology; TILLING

Introduction wide spectrum of stimuli such as distribution of nutrients, obstacles, heat, light and oxygen availability (Eapen et al., In agriculture, water is usually the predominant factor 2003; Massa and Gilroy, 2003). limiting plant growth. If plants do not receive adequate At DiSTA, a mutagenized population of barley rainfall or irrigation, the resulting drought stress can (H. vulgare L.) cv. ‘Morex’ has been produced following reduce the growth more than all other environmental chemical treatment with sodium-azide (Talame` et al.,2008, stresses combined. In this context, crops characterized 2009). This Targeting Induced Local Lesions IN Genomes by a more effective use of water will be at an advantage. (TILLING; McCallum et al., 2000) resource, named TILLMore Among the morpho-physiological traits that have been (http://www.distagenomics.unibo.it/TILLMore/), is avail- shown to influence drought resistance, root morphology able for both forward- and reverse-genetics applications in (e.g. depth, angle, size, number of root-hairs, etc.) plays barley. We report the results of a forward-genetics analysis a crucial role (Richards, 2008; Shao et al., 2008; Hodge to identify mutants for root architecture. et al., 2009). In monocots, a systematic genetic analysis of root formation has been carried out mainly in maize and rice (de Dorlodot et al., 2007; Taramino et al., 2007; Hochholdinger and Tuberosa, 2009; Coudert et al., Materials and methods 2010). This notwithstanding, the information on the The phenotypic analysis for root-related traits was carried genetic control of root morphology and functions in out on a subset of c. 1000 M families from the TILLMore barley (Hordeum vulgare L.) is rather poor, with only 4 collection. The analysis was performed using a paper-roll one notable exception (Kwasniewski and Szarejko, approach as described by Woll et al. (2005). The method 2006). Additional evidence shows the complexity of is based on seedlings grown in rollers made of a wound root growth and how it is continually challenged by a sandwich of filter papers and kept in a growth chamber under controlled conditions (16/8 h photoperiod and 24/228C day/night temperature). The analysis was performed on c. 12 seeds/family per paper-roll. At the * Corresponding author. E-mail: [email protected] stage of 8-d-old seedling, the morphology of the seminal 358 R. Bovina et al. roots was visually compared to that of the wild type Table 1. List of barley mutant categories with altered root a cv. ‘Morex’. The phenotypic evaluation was repeated phenotype (three reps of 12 seeds/paper-roll) only for the putative No. of root-mutant mutants identified during the preliminary screening. families/total For those mutants that presented an altered root-hair Root-mutants no. of root-mutant phenotype, further observations were made in aeroponic Mutant categoryb (n) families (%) conditions. The germinated seeds were placed in sterile Short rootc 25 34.7 glass tubes covered with cotton bungs, one seed/tube. Very short rootd 12 16.7 Each tube with seed was then connected with an Short and thick root 16 22.2 empty tube and both parts were held together with Short and curly root 8 11.1 parafilm. The plants were grown for 5 d under controlled Curly root 3 4.2 Short root-hair 3 4.2 conditions. Microscopic observations and image analyses Highly geotropic 2 2.8 of 5- to 7-d-old seedlings were performed with the use Root-to-leaf 2 2.8 of Stemi 2000-C (Zeiss) stereomicroscope and AxioVision Hairless 1 1.4 LE (Carl-Zeiss) program. On the same samples, the term- Total 72 100.0 inal root segments (c. 1 cm portion) were excised and a Pictures of these mutants are available at http://www.dista. viewed with a Tesla BS-340 scanning electron microscopy unibo.it/TILLMore/check.php. The images can be viewed (SEM) microscope. by typing the mutant code number (e.g. highly-geotropic: 194, 3580; hairless: 5545; short root-hair: 5387, 2588, 1568; root-to-leaves: 2956, 3109) or search for the mutant category. b For each mutant category, we report the total Results and discussion number of mutants identified and their percentage among the root-mutant families. c Roots c. 50–60% shorter com- d The screening of 1000 TILLMore M4 lines showed altered pared with wild-type ‘Morex’. Roots c. 70–90% shorter root morphology for 72 families. One group of 26 compared with wild-type ‘Morex’. mutants showed fairly stable phenotype, while 47 lines showed unstable, indistinct or segregating phenotypes. root-hair (srh). A more detailed phenotyping of the The mutations identified affected different aspects of root-hair mutants was carried out growing the mutant root development and were classified as belonging lines in aeroponics and root observation was performed to nine main categories (Table 1). Some of the unstable using a stereomicroscope and a SEM microscope. The phenotypes may reflect genetic segregation or responses images showed a noticeable difference between the to environmental conditions like gases (e.g. O2, ethylene hairless mutants, with a perfectly smooth root surface, and CO2), light and physical impediments (Engvild and the short root-hair mutants, where the elongation and Rasmussen, 2004). Furthermore, shortened roots of the root hairs is blocked in the early stages of could sometimes be due to a pleiotropic effect of other development (Fig. 1). Recently, a similar survey was mutations such as dwarf or chlorophyll mutants. Almost conducted at the University of Silesia in Poland with all the detected mutants showed reduced length of semi- a population of barley mutants screened in aeroponic nal roots ranging from slightly short to extremely short. In conditions (Szarejko et al., 2005). This analysis allowed various cases, short-root mutants developed other altera- for the identification of 17 mutants with changes in tions such as curled morphology and thick appearance. root-hair development and the cloning of an expansin In barley, an experiment similar to what is herein gene related to root-hair formation in barley (Kwasniewski reported has been performed for the identification of and Szarejko, 2006). Except for this outcome, until now, short-root mutants (Nawrot et al., 2005). In this case, very little information is available on the molecular nine mutants showing significant shortening of seminal control of root-hair growth in barley and, in general, roots were identified in a collection of dwarf and semi- in almost all the monocots. dwarf lines obtained after mutagenic treatment of Remarkably, two TILLMore root mutant categories spring barley with N-nitroso-N-methylurea and sodium- presented alterations that, to our best knowledge, so far azide (Nawrot et al., 2005). The analysis on our popu- have not yet been described in barley. The most interes- lation was carried out on lines randomly chosen without ting is a mutant category where the root does not appear considering the phenotype of the visible portion of the and is replaced by a leaf (root-to-leaf). Further studies plant. Thus, the majority of mutants (c. 70%) showed on these materials may provide important information alterations in the root but not in other organs. for a better understanding of the signalling networks Four lines showed an altered root-hair formation. that lead to root/leaf differentiation. The second category Such mutations affected two stages of root-hair develop- of mutants presented a highly geotropic growth. Similar ment and were classified as root hairless (rhl) or short mutants of Arabidopsis, presenting an increased response Images of root-hair mutants 359 (a) For some of the TILLMore mutants with clear changes in root architecture, an appropriate series of backcrosses and outcrosses are in progress prior to their mapping in order to fix the mutations and then map and clone new genes involved in root formation and growth. From a physiological standpoint, these mutants may also help to better comprehend the role of roots in water and nutrient uptake (Gahoonia et al., 2001), association with symbiotic organisms (Chen et al., 2005) and root– 1 mm soil interaction, hence providing valuable information Morex for agronomic applications.

(b) Acknowledgements

The research was partially funded by International Atomic Energy Agency. We thank the Department of Agronomy and Plant Genetics, University of Minnesota, St. Paul, USA, for providing ‘Morex’ seed.

1 mm 2588 References

(c) Chen B, Roos P, Borggaard OK, Zhu YG and Jakobsen I (2005) Mycorrhiza and root hairs in barley enhance acquisition of phosphorus and uranium from phosphate rock but mycorrhiza decreases root to shoot uranium transfer. Plant Phytology 165: 591–598. Coudert Y, Pe´rin C, Courtois B, Khong NG and Gantet P (2010) Genetic control of root development in rice, the model cereal. Trends in Plant Science 15: 219–226. de Dorlodot S, Forster B, Page`s L, Price A, Tuberosa R and Draye X (2007) Root system architecture: opportunities and constraints for genetic improvement of crops. Trends 1 mm in Plant Science 12: 474–481. 5545 Eapen D, Barroso ML, Campos ME, Ponce G, Corkidi G, Dubrovsky JG and Cassab GI (2003) A no hydrotropic Fig. 1. Two root-hair mutants compared with Morex wild- response root mutant that responds positively to gravitrop- type (a) used as reference. The picture (b) shows a short ism in Arabidopsis. Plant Physiology 131: 536–546. root-hair mutant (cod. 2588), where the hairs are strongly Engvild KC and Rasmussen SK (2004) Root hair mutants of shortened. The image (c) shows an hairless mutant with a barley. Barley Genetics Newsletter 34: 13–15. smooth root (cod. 5545). All the pictures were obtained Gahoonia TS, Nielsen NE, Joshi PA and Jahoor A (2001) A root with a Stemi 2000-C (Zeiss) stereomicroscope at 50£ of hairless barley mutant for elucidating genetic of root hairs magnification. and phosphorus uptake. Plant and Soil 235: 211–219. Hochholdinger F and Tuberosa R (2009) Genetic and genomic dissection of maize root development and architecture. to gravitropism, lack the capacity to perceive a moisture Current Opinion in Plant Biology 12: 172–177. gradient that is thought to be important for controlling Hodge A, Berta G, Doussan C, Merchan F and Crespi M (2009) root orientation (Kobayashi et al., 2007). Consequently, Plant root growth, architecture and function. Plant and Soil the two barley highly geotropic mutants may actually 321: 153–187. Kobayashi A, Takahashi A, Kakimoto Y, Miyazawa Y, Fujii N, be hydrotropic mutants, unable to perceive water gradi- Higashitani A and Takahashi H (2007) A gene essential ent in the paper-roll system. Further studies are required for hydrotropism in roots. Proceedings of the National to validate this hypothesis. Academy of Science USA 104: 4724–4729. Further details on root mutants and an informative Kwasniewski M and Szarejko I (2006) Molecular cloning and set of pictures are available in an on-line database at characterization of beta-expansin gene related to root hair formation in barley. Plant Physiology 141: 1149–1158. http://www.distagenomics.unibo.it/TILLMore/. This Massa GD and Gilroy S (2003) Touch modulates gravity sensing database is publicly accessible and seed for mutants to regulate the growth of primary roots of Arabidopsis of interest is available on request. thaliana. Plant Journal 33: 435–445. 360 R. Bovina et al. McCallum CM, Comai L, Greene EA and Henikoff S (2000) of chemically induced mutants in barley. Plant Biotechnol- Targeting Induced Local Lesion IN Genomes (TILLING) ogy Journal 6: 477–485. for plant functional genomics. Plant Physiology 123: Talame` V, Bovina R, Salvi S, Sanguineti MC, Piffanelli P, 439–442. Lundquist U and Tuberosa R (2009) TILLING with TILLMore. Nawrot M, Szarejko I and Maluszynski M (2005) Barley mutants In: Shu QY (ed.) Induced Plant Mutations in the Genomics with short roots. Barley Genetics Newsletter 35: 3–8. Era. Joint FAO/IAEA Programme. Rome: Food and Agriculture Richards RA (2008) Genetic opportunities to improve cereal root Organization of the United Nations, pp. 240–242. system for dryland agriculture. Plant Production Science Taramino G, Sauer M, Stauffer JL Jr, Multani D, Niu X, Sakai H 11: 12–16. and Hochholdinger F (2007) The maize (Zea mays L.) RTCS Shao HB, Chu LY, Jaleel CA and Zhao CX (2008) Water-deficit gene encodes a LOB domain protein that is a key regulator stress-induced anatomical changes in higher plants. of embryonic seminal and post-embryonic shoot-borne Comptes Rendus Biologies 331: 215–225. root initiation. Plant Journal 50: 649–659. Szarejko I, Janiak A, Chmielewska B and Nawrot M (2005) Woll K, Borsuk LA, Stransky H, Nettleton D, Schnable PS and Genetic analysis of several root hair mutants of barley. Hochholdinger F (2005) Isolation, characterization, and Barley Genetics Newsletter 35: 36–38. pericycle-specific transcriptome analyses of the novel Talame` V, Bovina R, Sanguineti MC, Tuberosa R, Lundqvist U maize lateral and seminal root initiation mutant rum1. and Salvi S (2008) TILLMore, a resource for the discovery Plant Physiology 139: 1255–1267.