Somatic hybrids of Solanumtuberosum and species of the Solanum nigrum-complex and their backcross progeny

Somatische hybridenva n Solanum tuberosume n soorten uit het Solanumn/grum-comple x en hunterugkruisingsnakomelinge n PROMOTOR: prof.dr.ir. E. Jacobsen hoogleraar ind e plantenveredeling in het bijzonder ind e genetische variatie en reproductie

SAMENSTELLINGPROMOTIECOMMISSIE : prof.dr.ir. M.Koornnee f Wageningen Universiteit

prof.dr. J.L. van Went WageningenUniversiteit

dr. J.H.S.G.M. de Jong WageningenUniversiteit

prof.dr. R.G.F. Visser WageningenUniversiteit

dr. K.S. Ramulu PlantResearch International n ,!uO^>^'

KARIN HORSMAN

Somatic hybrids of Solariumtuberosum and species ofth e Solanum nigrum-complex and their backcross progeny

PROEFSCHRIFT ter verkrijging van degraa d van doctor op gezag van de rector magnificus van Wageningen Universiteit prof.dr.ir. L. Speelman, in hetopenbaa r te verdedigen op maandag 14me i 2001 des namiddagst evie r uur ind e Aula CIP-DATA KONINKLIJKE BIBLIOTHEEK, DEN HAAG

Horsman, K.

Somatic hybrids of Solatium tuberosuman dspecie s ofth e Solarium n/grum-complex and their backcross progeny / Karin Horsman - [S.l.:s.n.]

Thesis Wageningen University - with references - with summaries in English and Dutch.Laborator y of Plant Breeding, P.O. Box 386,6700 AJ, Wageningen, Nl.

ISBN 90-5808-392-6

Keywords Solanumtuberosum, potato , Solanumnigrum, protoplast fusion,somati c hybrids, backcross experiments, ovule rescue, genomic in situ hybridisation, meiosis,chromosom e pairing,AFL P STELLINGEN

1. Eengoed e groei van fusieprodukten in vitrozeg t niets over hungroeikrach t in vivo. (dit proefschrift)

2. De verhouding van het aantal genomen dat de betrokken soorten bijdragen aan een fusieprodukt is van invloed op de resultaten van terugkruisingsexperimenten met dit fusieprodukt. (dit proefschrift)

3. Het is onterecht dat fusieprodukten tot de genetisch gemodificeerde organismen moeten worden gerekend.

4. Evenals somatische verdubbeling biedt somatische hybridisatie van ver-verwante soorten in veel gevallen slechts een kleine kans op introgressie. (Lim etal., Chromosome Research 8:119-125)

5. Het plan van het College van Bestuur van de Universiteit van Amsterdam om het promotietraject van een vierjarige in een driejarige opleiding te veranderen door de begeleiding te intensiveren is vergelijkbaar met de verwachting dat een sporter een wereldrecord zal realiseren wanneer zijn coach daartoe gemaandword t door de sportbond.

6. De biologische landbouw zou een extra impuls kunnen krijgen wanneer zij genetische modificatie zou accepteren.

7. Het barbecue-gen ligt op het y-chromosoom.

8. De term "kenniseenheid" isverzonne n door een optimist.

9. De wetenschappen bedrijven en niet van mensen houden dat is als een fakkel ontsteken en de ogen sluiten. (Chinees gezegde)

Stellingen behorende bij het proefschrift 'Somatic hybrids of Solanum tuberosum and species of the Solanum ragrum-complex and their backcross progeny" door Karin Horsman, in het openbaar te verdedigen op 14 mei 2001,teWageningen. ABSTRACT

This thesis describes the production of somatic hybrids of potato and species of the S.n/grum-complex , wild relatives of the cultivated potato which are potentially interesting sources of genetic variation. The resulting fusion products could be used in backcross experiments with potato which, however, had a very low success rate and stopped after the production of BC2-progeny. To improve the crossability of the backcross derivatives, the genomic compositions of a BC1 and a BC2-genotype were successfully altered by adding two haploid potato genomes by somatic hybridisation. Unfortunately crossability did not improve, although the resulting BC1-somatic hybrids could be backcrossed again with potato. Through the application of genomic in situ hybridisation (GISH) itwa s shown that genomic constitutions of somatic hybrids and backcross derivatives were close to or exactly the expected constitutions and that homoeologouspairin goccurre di nth e BC1a swel la s inth e BC2generation . CONTENTS

CHAPTER 1 General introduction 9

CHAPTER2 Somatic hybridisation between Solanumtuberosum an d species ofth e 19 S. nigrumcomplex : selection ofvigorousl y growing andflowerin g plants

CHAPTER3 Successful first and second backcrosses of S. nigrum (+) S. tuberosum 33 somatic hybridswit h both Solanum parents

CHAPTER4 Alteration of the genomic composition of Solanumnigrum (+) potato 47 backcross derivatives bysomati c hybridisation: selection offusio n hybrids by DNA measurements and GISH

CHAPTER5 Prospects for the introgression of chromosomes from non-tuberous Solanum 59 nigrum into S. tuberosum:a qualitativ e analysis of the meiosis through GISH andAFLP-analysi s of backcross derivatives

CHAPTER6 General discussion 71

SUMMARY 79

SAMENVATTING 83

REFERENCES 87

NAWOORD 97

CURRICULUMVITA E 101

LIST OF PUBLICATIONS 103 1

General introduction Chapter 1

TRADITIONAL POTATO BREEDING SOMATIC HYBRIDISATION

In plant breeding wild relatives of a cultivated The discovery of the fusion of protoplasts species are interesting sources of genetic opened new prospects for further broadening of variation. In many instances these wild species the genetic base of potato because this have important characteristics that can not be technique offered the possibility to use distantly found within the gene pool of the cultivated related species (Shepard ef a/., 1983). Within species. The traditional method of introducing a the family of the Solanaceae somatic specific trait from a related species is sexual hybridisation has been applied in many hybridisation followed by recurrent instances, mostly for the introduction of a single backcrossing. In potato breeding several interesting trait. However, the consequence of species have been used for the transfer of the application of somatic hybridisation is that desirable traits into the potato gene pool. Many the complete genome of the donor species is of the potato cultivars bred in Europe and North introduced into the recipient species, or a large America carry genes from S. demissum, S. part in case of an asymmetric fusion. In either andigena or S. vernii (Hermsen, 1994), mainly case intensive backcrossing is inevitably but coding for resistance to late blight, cyst surprisingly, only a few successful somatic nematodes and viruses. hybridisation experiments followed by recurrent Often one or more reproductive barriers backcrossing were described in literature. have to be overcome in the process of Within the Solanaceae the combinations S. introgression of a specific trait. Depending on bulbocastanum (+) potato, S. brevidens (+) the type of barrier, application of, for instance, potato and potato (+) tomato have been studied rescue pollination or the use of bridging species most extensively. could be asolutio n to the problem.A successful example of the latter was the introgression of late blight resistance from diploid S. Somatichybrids ofS. bulbocastanum (+)potato bulbocastanum (B) into tetraploid potato (T), by using two late-blight-susceptible bridging S. bulbocastanum is a source for resistance to species, tetraploid S. acaule (A) and 2x S. late blight (Niederhausen and Mills, 1953) and phureja (P). The introgression scheme, which Meliodogyne chitwoodi races 1an d 2 (Brown ef also involved doubling of triploid hybrids, a/., 1989) and to M. hapla (Janssen ef a/., resulted in the ABPT clones among which 1996). S. bulbocastanum is extremely difficult some highly resistant segregants were to cross with potato but, as described discovered which have been used as breeding previously, by using bridging species it could be parents (Ramanna and Hermsen, 1971; made accessible for potato breeding (Hermsen Hermsen and Ramanna, 1973). and Ramanna, 1973). Others choose to bypass Despite the possibilities that were created the sexual incompatibility of these species by by these classical approaches, many non- applying somatic hybridisation like Helgeson ef tuberous Solarium species still cannot be al. (1998) who produced somatic hybrids sexually hybridised with potato and used for between diploid S. bulbocastanum and introgression breeding. tetraploid potato with the aim to transfer late

10 Chapter1 blight resistance. First and second generation and morphological analysis of the somatic backcrosses could easily be obtained after hybrids (Pehu ef al., 1989), analysis of their crossing with tetraploid potato. Late blight resistance against PVY, PVX and PLRV resistance was transferred to both generations. (Gibson ef al., 1988; Pehu at al., 1990; Brown et al. (1995) analysed S. Valkonen ef al.,1994) , determination of nuclear bulbocastanum (+) potato somatic hybrids DNA content, chromosome number and (Austin ef a/., 1993) and progeny for their flowering capacity (Valkonen et al., 1994) or resistance to Meliodogyne chitwoodi races 1 characterisation of morphological variation and and 2 and M.hapla. They reported male sterility cold resistance (Preiszner ef al., 1991). of the somatic hybrids and BC1 genotypes but However, in some instances the somatic crossing experiments with tetraploid potato hybrids were successfully used in backcross resulted in seeds. The BC2 population of 62 experiments. Ehlenfeldt and Helgeson (1987) plants was used for mapping the resistance to as well as Jacobsen ef al. (1993) produced M. chitwoodiiwit h RFLP's (Brown etal., 1996). progeny from S. brevidens (+) S. tuberosum somatic hybrids after crossing with the latter species. They both used tetraploid as well as Somatichybrids of S. brevidens (+)potato hexaploid hybrids with respectively two and four genomes of potato. Crossability of the tetraploid S. brevidens is an example of a non-tuberous somatic hybrids was poor with both 2x and 4x Solanaceae species which has been of interest potato genotypes, whereas the hexaploid for potato breeding because of its resistance to somatic hybrids crossed very well with potato leaf roll virus (PLRV), potato virus Y tetraploid potato. Rokka ef al. (1994) were also (PVY), potato virus X (Ehlenfeldt and Helgeson, successful in producing backcross progeny 1987; Gibson ef a/., 1988; Pehu ef al., 1990), from S. brevidens (+) potato somatic hybrids. In and several bacterial strains causing Erwinia their experiments the hexaploid somatic soft rot (Austin ef al.,1988 ) but itwa s not easily hybrids, although being aneuploids, could be accessible. Numerous attempts have been used as male parents, in contrast to Ehlenfeldt undertaken to sexually hybridise S. tuberosum and Helgeson (1987) and Jacobsen ef al. and S. brevidens (Ramanna and Hermsen, (1993) who were only successful when the 1981) but no hybrids resulted from these somatic hybrids were used as the female experiments. The alternative approach of parent. somatic fusion seemed to be much more Rokka ef al. (1994) and Jacobsen ef al. successful, as was shown by a number of (1993) only described the production of the first research groups (Fish ef al., 1988, Preiszner ef generation backcrosses whereas Williams ef al. al., 1991; Jacobsen ef al., 1993, Valkonen ef (1993) and McGrath ef al. (1994) analysed first al., 1994). Although in most instances the and second generation BC-genotypes with somatic hybrids were found to be at least molecular markers. Their results suggested the partially fertile, backcross hybrids of the S. occurrence of recombination by crossing-over brevidens (+) S. tuberosum fusion products are between potato and S. brevidens which is a not frequently reported. Some research was prerequisite for the introgression of useful traits mainly focused on the molecular, cytogenetic of S.brevidens into the potato gene pool.

11 Chapter1

In 1995 Watanabe ef al.describe d the first immature seeds resulting in six BC1 plants. The successful attempt of a direct cross between S. first step inth e backcross programme appeared brevidens and S. tuberosum by using a rescue to be by far the most laborious and difficult one pollinator, IVP35. Besides parental since it took only a few pollinations for the characteristics the F1 hybrids also expressed production of second backcross progeny S. phureja traits. A major advantage of the (Jacobsen ef al., 1994). A combined total of 97 production of these sexual hybrids is the fact BC2 plants from three populations were that, as a result of this, the somatic hybrids of screened for the presence of potato genotypes this combination of species no longer are with alien tomato chromosome additions. Out of considered to be genetically modified the 12 possible monosomic addition types, organisms. seven different ones (1,2 , 4, 6, 8, 10 and 12) were detected by Garriga-Caldere (1998). At present all 12 monosomic addition lines are Somatichybrids of potato (+)tomato available (Haider Ali ef al., in press).

One of the best studied combinations in somatic hybridisation is that of potato and SOMATIC HYBRIDS WITHINTH E BRASSICACEAE tomato (Lycopersicon esculentum). Melchers ef al. (1978) were the first to describe the Another plant family from which several species production of potato-tomato somatic hybrids but have been used in somatic fusion experiments progeny was never reported by them. Wolters are the Brassicaceae. The most important crop et al. (1991) analysed both symmetric and species within the Brassicacea are the species asymmetric somatic hybrids and of the genus Brassica: B. campestris (e.g. Schoenmakers ef al. (1993) tried Chinese cabbage, genome AA), 8. nigra unsuccessfully to obtain progeny from triploid (genome BB), 6. oleraceae (e.g. cabbage, fusion hybrids with two genomes of tomato and genome CC), 8. juncea (genome AABB), 8. one of potato. Fifteen years after Melchers ef napus (rapeseed, genome AACC), 8. carinata al. (1978) reported the first potato-tomato (genome BBCC). Most of the previously somatic hybrids, Jacobsen ef al. (1994) were mentioned species have been used in the first to succeed in making backcrosses of a interspecific somatic hybridisation experiments hexaploid somatic hybrid, containing two whereas a majority of the intergeneric and genomes of tomato and four of potato, with intertribal fusion experiments involved 8. napus tetraploid potato. Over 3400 immature seeds as one of the parents. were cultivated to obtain the first BC1 genotype which was male sterile but demonstrated female fertility. The production of another six Interspecificsomatic hybrids BC1 plants was described by Garriga-Caldere ef al. (1997) who used hexaploid somatic Interspecific somatic hybridisation within the hybrids of the same series as Jacobsen ef al. Brassicacea has been performed between (1994) asfemal e parents.Th e effort they had to several combinations of diploid and put in was much lower with only 406 cultured allopolyploid species. For increasing the

12 Chapter 1 genetic diversity of rapeseed, Brassica napus combinations. One of the goals was to (AACC) was resynthesised by a fusion of 8. introduce resistance to beet cyst nematod campestris(syn . 6. rapa, AA) and 8. oleraceae (BCN) from Raphanus sativa into rapeseed but (CC) (Sundberg ef al., 1987). Somatic the hybrids showed reduced fertility and could hybridisation inthi s combination of species was not be backcrossed to 8. napus (Lelivelt and not applied in order to circumvent crossing Krens, 1992). Experiments with Sinapisalba (+) barriers since sexual hybridisation in B. napus hybrids also failed in producing combination with chromosome doubling is progeny because the somatic hybrids were another possibility to resynthesise 6. napus mitotically unstable and sterile (Lelivelt, 1993). (Ozminkowski Jr. and Jourdan, 1994a and Somatic hybrids of S. alba (+) 6. y'uncea 1994b). however, were fertile and BC1 progeny was Important breeding goals for rapeseed are obtained after crossing with S. alba, just like resistances to blackleg (Phoma lingam, somatic hybrids between 8. napus and Eruca anamorph Leptosphaeria maculans) and to sativa which enabled backcrossing to rapeseed clubroot (Plasmodiophora brassicae) (Sjodin and introduction into breeding programmes and Glimelius, 1989a, b). Resistance to Phoma (Fahleson et al, 1988). The correlation between lingam was found in 6. nigra , 8.juncea and 8. chromosome elimination in somatic hybrids and carinata and after production of symmetric as genetic distance between the fusion parents well as asymmetric somatic hybrids between was investigated by Sundberg and Glimelius these gene-donors and rapeseed, resistant (1991) by comparing interspecific and hybrids were obtained. Stable inheritance and intergeneric somatic hybrids. They concluded possible introgression of the gene(s) for that a larger genetic distance resulted in a resistance to Phoma lingam have been higher degree of chromosome elimination. recorded in lines derived from a backcross programme of the original hybrid with rapeseed (Waara and Glimelius, 1995). Intertribal somatichybrids

Intertribal somatic hybrids within the Intergeneric somatichybrids Brassicacea can only be produced via somatic hybridisation. The first combination was Several intergeneric hybrids have also been Arabidopsis thaliana (+) 8. campestris (Gleba produced via protoplast fusions between and Hoffman, 1979, 1980), followed by A species from the genera Brassica, Eruca, thaliana (+) 6. napus (Forsberg ef al., 1994); Sinapis, Raphanus, Moricandia, Diplotaxis and Thlasi perfoliatum (Fahleson ef al., 1994b) or Trachystoma. In general all intergeneric Lesquerella fendleri (+) 8. napus combinations resulted in hybrid plants but, (Skarzhinskaya ef al., 1996); Barbarea vulgaris compared to the interspecific somatic hybrids, (+) 8. napus (Fahleson ef al., 1994b) or 8. they showed a larger variation in chromosome campestris. More spontaneous asymmetric number, with more aneuploid hybrids as well as hybrids were found among the intertribal polyploid hybrids. Attempts to produce hybrids than among the intrageneric and backcross progeny were not successful in all intergeneric hybrids. Interestingly, though,

13 Chapter1 when comparing the intertribal hybrids with fragment length polymorphisms (RFLP) and those obtained from hybridisation experiments amplified fragment length polymorphisms between more closely related species, no (AFLP) have opened immense possibilities to differences inth e frequency ofhybri d fusionso r monitor the progress of introgression. These hybrid shoot regeneration were recorded. techniques enable the determination of: 1) However, the intertribal combinations were, in whether alien genomes or chromosomes are general, more difficult to culture to mature intact; 2) intergenomic recombination; 3) plants in the greenhouse. This was especially transmission of the alien chromosomes toth e pronounced inth efusio n Barbarea vulgaris(+ ) progeny. 6. napus which never resulted in plants that GISH andth e RFLP-technique were both could grow under greenhouse conditions applied to the progeny of thefirs t and second (Fahlesonefa/., 1994a). backcross progeny of the potato-tomato Nevertheless, intertribal somatic hybrids somatic hybrids in order to improve the have been produced that differentiated and efficiency of introgression of genetic material developed into normal hybrid plants that could into cultivated potato. (Jacobsen ef a/,1995 ; even be backcrossed, like some ofth e Arabido- Garriga-Caldere ef a/., 1997; Garriga-Caldere Brassica hybrids (Forsberg ef a/., 1994)o r fi. ef a/., 1998). The combination of both napus (+) L fendleri asymmetric hybrids techniques resulted in the elucidation of the (Skarzhinskayaefa/, 1996). genomic constitution of all hybrids involved. With theapplicatio n of GISH it was possible to establish the number of alien tomato APPLICATION OF GISH AND GENETIC MARKERS chromosomes whereas RFLPs were used for FOR THE MONITORING OF THE PROCESS OF the identification of individual chromosomes. RECURRENTBACKCROSSIN G The first BC1 genotype was shown byusin g the GISH-technique to contain nine tomato There are quite a few bottlenecks for chromosomes besides the 24 potato introgression ofalie n chromosomes, like chromosomes. The RFLP-analysis indicated • Preferential elimination of alien the presence ofonl y sixdifferen t chromosomes chromosomes from the fusion hybridsan d which led to the conclusion that three backcross progenies; chromosomes were in disomic condition. So • extreme degree of sterility inth e backcross with these techniques it was clearly progenies; demonstrated that chromosomes from alien • lack of genetic recombination between the genera within the Solanaceae can be species involved; successfully transferred and that regularly • non-transmission of alien chromosomes plants areobtaine d with alien chromosomes in through the gametes. disomic condition. This phenomenon is the Besides these drawbacks the monitoring and result oflac k ofseparatio n ofth e chromatidso r evaluation of backcross progeny is not simple homologs during meiosis. and can be highly laborious. Fortunately, a A very powerful demonstration of the number of recent molecular techniques like application of GISH and FISH (fluorescence in genomic in situ hybridisation (GISH), restriction situ hybridisation) was performed by Kamstra

14 Chapter 1 during his analysis of Alstroemeria aurea and resistant to many biotic and abiotic factors and A. inodora hybrids. Chromosomes of both probably possess immense physiological and species could be distinguished with GISH genetic diversity. The greatest diversity and whereas FISH with repetitive probes (two concentration is found in the New World species specific and two ribosomal rDNA Tropics, particularly in South America. The repeats) was used to identify all individual generic type S. nigrum, commonly known as chromosomes (Kamstra etal., 1997).Th e same the black, garden or common nightshade is techniques were used to study pairing of predominantly an Eurasian species which does chromosomes and establish whether or not not occur naturally inSout h America (Edmonds, recombination had occurred between the two 1979). The S. nigrum complex comprises of species involved. It was shown that species with several levels of ploidy, ranging recombination had occurred frequently in the from diploid (S. americanum Miller; S. original hybrid since chromosomes with 1-4 chenopodioides Lam.) to tetraploid (S. villosum recombination points per chromosome were Miller) and hexaploid (S. nigrum L.) species. discovered in all BC1 genotypes (Kamstra ef Some of the previously mentioned species a/., 1999). In the recurrent backcross have been used in experiments with the aim to generations, on the other hand, it was recorded explore their genetic diversity. A few sexual that recombination was almost absent. Based hybrids of S. nigrum or S. villosum with on these results Kamstra concluded that respectively S. tuberosum and S. demissum, breeders that are aiming at introgression should were produced by Eijlander and Stiekema emphasise more on the production of sufficient (1994) who circumvented the strong barriers to BC1genotypes . sexual hybridisation through embryo rescue. These flowering sexual hybrids were found to be highly sterile and until now the production of S. nigrum ANDSPECIE S OF THES . nigrum backcross progeny has not been reported. COMPLEX One of the interesting traits was the resistance of the S. nigrum species to In view of the previously mentioned possibilities Phytophthora infestans, which is based on a it appears attractive to utilise more distantly strong hypersensitive reaction pattern (Colon ef related species in somatic hybridisation a/., 1993;Vleeshouwer s et al., 2000). S. nigrum experiments, for example the species of the S. as well as the sexual hybrids of S. nigrum nigrum complex which belong to the section related species and potato (Eijlander and Solarium, more usually referred to as the Stiekema, 1994) were shown to be highly section Morella (Seithe, 1962). Species of this resistant to P.infestans,. section are distributed in all major continents In the past, Binding ef al. (1982) have and are being used as leafy herbs and performed protoplast fusions between S. vegetables, as a source of fruits and for nigrum and S. tuberosum in order to transfer medicinal purposes in Africa and Southeast the atrazine resistance from black nightshade Asia (Edmonds and Chweya, 1997). However, to potato (Gressel ef a/., 1984), but successful in many cases they are considered to be weeds backcross experiments have not been reported which means they are highly adaptable, so far. In this thesis somatic hybridisation

15 Chapter 1 experiments with four species of the S. nigrum described. The results of the sixteen fusion complex aredescribed : combinations showed that somatic combining • S. americanum Miller: 2n = 2x = 24. Most abilities were influenced by the ploidy level as widespread, highly variable species, one of well as the genotype of the parental species. the ancestors of S. nigrum. It is found The main aim was to select flowering somatic throughout theworld . hybrids that could be used for backcross • S. chenopodioidesLam. :2n = 2x =24 . One experiments. ofth e most easy identifiable species, native to the eastern parts of South America, Chapter 3 describes successful backcrosses of might be an ancestor of S.villosum. Solanum nigrum (+) 2x potato somatic hybrids • S. villosum Miller: 2n =4 x =48 . It is thought with both S. nigrum and potato. A considerable to be native to Eurasia and has been difference in success rate was detected sparingly introduced to Australia, New between both backcross programmes. While Zealand and North America. It is one of the first and second backcross progeny with S. ancestors of S. nigrum, but its own nigrum could be obtained easily, it required ancestry is unknown. thousands of pollinations, resulting in 505 - S. nigrum L. ssp. schultesii: 2n = 6x= 72. It cultured ovules, to obtain two potato-BC1 is considered to be native to Eurasia, has genotypes. The production of BC2 progeny was not been found yet in South or Central equally laborious and so far no BC3 plants America. The derivation of this species could be obtained. The vigour and resistance to from S. villosuman d S. americanum is now Phytophthora infestans of twelve BC2 very well established (Edmonds, 1979). genotypes has been described.

Since no progeny was obtained from the BC2 AIMSAN DOUTLIN E OF THISTHESI S genotypes, alternative approaches were sought to overcome the sexual crossing barrier. In The principle aims of the work presented in this chapter 4 is described how by somatic thesis were the following: 1) To hybridise the hybridisation it was attempted to add two non-tuberous S. n/grum-species with cultivated additional potato genomes to the BC1-6738 potato; 2) to backcross the fusion products to and BC2-9017 to improve crossability of the potato; 3) to evaluate the BC progenies with backcross derivatives. An approach for the GISH and AFLPs and for resistance to P. selection of vigorous somatic hybrids, that can infestans; 4) to gain an insight into the aspects be used for making recurrent backcrosses, of introgression of alien chromosomes into involving flow cytometry to estimate the DNA potato. content of the regenerants, crossing experiments and genomic in situ hybridisation In chapter 2 the fusion experiments between results are discussed. Plants with a DNA diploid or tetraploid potato genotypes and four content higher than that of the BC1 or BC2 species belonging to the S. nigrum complex, genotypes were considered to be potential namely Solanum nigrum, S. villosum, S. somatic hybrids. The application of genomic in chenopodioides or S. americanum are situ hybridisation (GISH) made it possible to

16 Chapter 1 distinguish clearly between S. nigrum and parent and the somatic hybrid was missing in potato chromosomes in mitotic and meiotic BC1-6738, which is an indication that no major chromosome spreads which enabled us to chromosome elimination had taken place. A determine the genomic composition of hybrids. second backcross with 4x potato resulted in eleven BC2 genotypes with a near-pentaploid In chapter 5 the chromosome constitution and genomic constitution. Molecular cytological meiotic behaviour of a somatic hybrid of investigations on nuclei of BC2 tetrad cells Solariumnigrum (2n=6x= 72) and diploid potato showed that transmission of alien S. nigrum (2n=2x=24) and its BC1 and BC2 progeny was chromosomes to BC3 progeny is likely. Meiotic described. The somatic hybrid F21-26 was analysis of metaphase I in BC1-6738 and in found to be an octaploid with six genomes of S. BC2-9019 indicated clearly that allosyndetic nigrum and two of potato. A cross between pairing occurs in these genotypes. Association F21-26 and a 4x potato genotype produced of homoeologous chromosomes leading to both BC1-6738, a hexaploid with 36 chromosomes bivalent and trivalent formation was observed each of both species. With GISH it was but recombination based on crossingover was impossible to determine whether the 36 notdetected . chromosomes of S. nigrum represented three complete genomes. However, AFLP-data In chapter 6 the results of the earlier chapters showed that none of the AFLP-specific markers are discussed including the perspectives of this that was amplified in the S. nigrum fusion approach and material for future research.

17 Somatichybridisatio nbetwee nSolanum tuberosum an d species of the S. nigrum complex: selection of vigorously growing and flowering plants

Karin Horsman, Marjan Bergervoet and Evert Jacobsen.

The Graduate School of Experimental Plant Sciences, Wageningen University, Laboratory of Plant Breeding, P.O. Box 386,6700 AJ Wageningen,Th e Netherlands.

This chapterwas publishedin Euphytica 96:345-352(1997) with modifications

ABSTRACT

Fusionexperiment swer eperforme dbetwee ndiploi d(2n=2x=24 )o rtetraploi d(2n=4x=48 )potat ogenotype s andfou rspecie so fth eSolanum nigrum complex ,namel yS .nigrum (2n=6x=72) ,S .villosum (2n=4x=48) , S. chenopodioides (2n=2x=24) or S. americanum(2n=2x=2 4 and 2n=6x=72).Al lfiv e accessions ofth eS . n/grum-specieswer esuccessfull y hybridisedwit ha tleas ton eo fth epotat ogenotypes .Somati c combining abilitieswer e influenced byth e ploidy level aswel la s byth egenotyp e ofth e parentalspecies .Th e useo f kanamycin or hygromycin resistance as cell-selectable markersystem had no influence on somatic combining ability, butsuc h markers canb e useful to improve efficient selection of somatic hybrids in sufficient numbers. Atleas t 20%o fth e hybrids ofeac h successful combination performed well invitro. However, only 60 genotypes out of 761 somatic hybrids were vigorous as well as flowering inth e greenhouse.Analysi s ofth e DNA content ofsomati c hybrids could be used asa criterio nfo r the indirect selectionin vitro o fhybrid stha twer evigorou s inth egreenhouse . Flowering somatic hybridso fS .nigrum (+) 2xpotat o and S.americanum (+) 4x potato were selected with the aimo fintrogressio n ofresistanc e traitsafte r recurrent backcrossingwit hcultivate d potato.

19 Chapter 2

INTRODUCTION seventies (Melchers efa/. , 1978;Jacobse ne fa/. , 1992; Schoenmakers ef a/., 1993). Backcross In potato breeding several related species have progeny from the potato-tomato hybrids were been usedo na wid e scalefo rth e introduction of less readily obtained than from S. brevidens- useful traits into the gene pool of the cultivated potato somatic hybrids. Jacobsen ef al. (1994) potato. Although most tuber-bearing wild were the first to succeed in backcrossing potato Solariumspecie s can be crossed easily with the (+) tomato hybrids with potato. At present the cultivated potato,som e valuable source species transfer of a number of valuable traits, partly are recalcitrant inthi s respect (Hermsen, 1994). related to non-host resistance, is being Several approaches have been used to investigated in several BC-populations overcomethi sbarrier .Vi abridgin gspecie sgene s (unpublished data). from distantly related species of the series Black nightshade (S. nigrum) is another Bulbocastana,Pinnatisect a and Etuberosa have interesting non-tuberous species with useful beensuccessfull y introgressed intoth ecultivate d traits. Colon ef al, (1993) described a high level potato (Hermsen, 1966; 1983; Hermsen and of late blight resistance inS .nigrum (2n=6x ) and Ramanna, 1973; Hermsen and Taylor, 1979). the closely relatedspecie s S.villosum (2n=4x) .A The application of embryo rescue has been few sexual hybrids of S. nigrum or S. villosum another successfulapproac h inproducin g sexual with S. tuberosumo r S. demissum respectively, hybrids of potato and recalcitrant species, such were produced by Eijlander and Stiekema as three species of the series Etuberosa (1994). These sexual hybrids were found to be (Watanabe efal., 1995 ) or Solanumnigrum and highly sterile and until now, the production of S. villosum(Eijlande r and Stiekema, 1994). backcross progeny has notbee n reported.I nth e A third method, to circumvent strong past, Binding ef al. (1982) have performed crossability barriers, is somatic hybridisation. protoplast fusions between S. nigrum and S. However, when somatic hybrids are obtained, tuberosum in order to transfer the atrazine newbarrier s maybloc k subsequent backcrosses resistance from black nightshade to potato that are essential for introgression of desired (Gressel ef al., 1984), but successful backcross traits into the potato genome andfo r elimination experiments have not been reported so far. of the undesired traits. S. brevidens is an This paper describes the production and example of a species that has been used selection of vigorously growing and flowering successfully as a fusion parent in combination somatic hybridso fS .n/g/um-relate dspecie s and with potato, resulting in fertile somatic hybrids potato that are suitable for backcross thatcoul db ecrosse dwit hpotat o(Ehlenfeld t and experiments. For increasing the success of Helgeson, 1987; Jacobsen ef al., 1993). In this somatic hybridisation two factors were tested. way backcross progenies with resistances to Firstly, selective growth markers were applied Erwiniasof t rot (Austin ef a/., 1988), potato leaf with theobjectiv e to isolate somatic hybrids ona rollvirus ,potat oviru sX an d potato virusY (Pehu large scale.Secondly , the influence ofth e ploidy ef a/., 1990) were obtained. The fusion levels of the fusion parents on the production of combination ofpotat oan dtomat o {Lycopersicon somatic hybrids was evaluated. esculentum Mill.) is more complex. It has been studiedelaboratel y fora lon gtime ,startin g inth e

20 Chapter 2

MATERIALS AND METHODS Protoplast fusion procedure

Plant material Theprotoplas tfusio nprocedur ean dmedi a used were according to Jacobsen efal. (1992) . Solid- Fourspecie so fth eS .nigrum comple x andthre e callus growth medium (MS11), shoot induction genotypes ofS .tuberosum ssp .tuberosum wer e medium (MS12) and shoot elongation medium used in our fusion experiments (Table 1). This (MS13) were prepared according to Mattheij ef table includes the accession number, the ploidy al. (1992). When resistance genes in parental level, the abbreviation(s) of the species and the clones allowed the use of selective agents, serial number(s) of the plants that were kanamycinwa suse da tconcentration s of5 0mg/ l propagated in vitro and used in the fusion in TMD, 75 mg/l in MS11, 100 mg/l in MS12, experiments. The abbreviations and serial MS13, MS20 and MS30. Hygromycin was used numbers are used throughout this paper. at concentrations of 10mg/ l inTM D and 20 mg/l The S. nigrum ssp. schultesii accession in MS11, MS12, MS13, MS20 and MS30. (ngr, SN14-0, 2n=6x=72) was provided by the Gradina Botanica in Rumania. The accession of the diploid species S. chenopodioides (che, Identification and analysis of somatic hybrids SN13-1, 2n=2x=24) was collected at Miseglia, Apuane in Italy. In one fusion combination, a The first screening of the putative hybrids was hygromycin resistant genotype of S. based on the type and density of hairs on the chenopodioides (cheH, SN13-1H) was involved, stem of the regenerated shoots. The potato which was obtained by Agrobacterium genotypes had few long, non-glandular hairs on transformation (strain LBA 4404 with plasmid their stems,wherea s the stems ofspecie s ofth e PVU 1011). The accession of the species S. S. nigrum complex were covered with short, villosum(vll, SNF1-1,2n=4x=48 )wa scollecte da t glandular hairs. Hybrid plants had both types of Montpellier (France). Both S. americanum hairs and the density was intermediate. accessions (ame2*, Sam90-3, 2n=2x=24 and Hybridity of the plants was confirmed with amesx, Sam88-1, 2n=6x=72) were supplied by the isozyme marker 6-phosphogluconate the Botanical Garden of the University of dehydrogenase (6-PGDH) withth ephas t system Nijmegen (the Netherlands). (Jacobs ef a/., 1995). The S. tuberosum fusion parents were an Flowcytometri c analysiso fth e DNAconten t amylose-free (amt) mutant (tbr^, 1029/31, ofth eparenta l plantsan d hybridswa scarrie dou t 2n=2x=24) described byJacobse n etal. (1989) , according to De Laat ef al.(1987) . a kanamycin resistant transformant of 1029/31 (tbr2x, AM10K) and the Dutch potato cultivar Desiree (2n=4x=48). Propagation and culture ofhybrids

Invitro plant swer e propagatedo nMS3 0i nglas s tubeso r lowglas sjars .Vigou ran dmorpholog y of the hybrids were assessed in a glasshouse that complied with Government biosafety rules.

21 Chapter2

Table 1. Indicationan ddetail so fth eplan tmateria luse di nth efusio nexperiment so ffou rspecie so fth eS . nigrum complexwit hon etetraploi dan dtw odiploi dgenotype so fS . tuberosum.

Species Accession/variety Ploidy Abbreviation

S.nigrum 88BG 278 02 6x ngr ssp. schultesii S.chenopodioides 88BG186 09 2x chp 2x chp" S.americanum Sam 904750023 2x ame2x Sam 884750061 6x ame6x S.villosum - 4x vll

S. tuberosum 87.1029/31 2x 1029/31 2x AM10K Desiree 4x -

H =Hygromyci nresistance , K =kanamyci nresistanc e

RESULTS Regeneration experiments with mixtures ofS . nigruman dpotat o protoplasts resulted inshoot s Regeneration of fusion parents of bothspecie s but,a sexpected ,th efirs t shoots were always ofS . nigrumorigin . The regeneration capacity of all parental Theseobservation s indicatedtha tth eus eo f genotypes, S.n/grum-relate dspecie s aswel l as a cell-selectable marker, such as kanamycin potato genotypes, was evaluated in several resistance, could improve the efficiency of control experiments with the aimt o use this regeneration of somatic hybrids considerably. characteristica sa selectio ncriterio n inou rfusio n Applicationo ftw ocell-selectabl e markers,on ei n experiments. Based on these results, itwa s each fusion parent, could theoretically increase possiblet oasses swhethe rselectio nmarker slik e the efficiency to 100% of somatic hybrids.Al l kanamycin andhygromyci n resistancewoul d be three approaches using zero, oneo rtw o cell- necessary or mainly be useful in our fusion selectable markers have been applied in our experiments. Regenerants of the S. nigrum- experiments. relatedspecie swer eobserve dwithi non emonth , whereas the potato genotypes required on average two months tostar t regeneration. The Identification of somatichybrids regenerationo fth epotat ogenotyp e 1029/31 was studied inmor edetail .Afte rthre emonth s47 %o f Somatic hybrids were identified both thecall iha dproduce dregenerants , increasingt o morphologically (typean ddensit yo fth ehair so n 62% after four months and67 % after seven stems) and biochemically (isozyme patterns). K months. Desiree and AM10 behaved similarly. The reliability ofth e morphological identification

22 Chapter 2

ofsomati c hybridswa sevaluate do nth ebasi so f parent (AM10K) the kanamycin and the pooled results of two fusion experiments, S. chenopodioides the hygromycin resistance involving the combinations W/(+) 1029/31 and gene. The experiment yielded 1025 calli that ngr(+) 1029/31. All 147 regenerants obtained could grow on media with both antibiotics. from these two experiments were tentatively Regenerants wereobtained fro monl y4 8calli ,o f classified either as "somatic hybrid phenotype", which 42 (88%) were hybrids. Only 26% of the "parental phenotype" or "inconclusive" on the hybrid plants grew vigorously in vitro. basis of characteristics of their hairs. Unfortunately none of these hybrids, including Subsequently, this morphological classification the ones that grew well in vitro,wer e capable of was compared with results from the isozyme growing inth e greenhouse. analysis. Outo f 112somati c hybrids, 110 (98%) had been successfully identified on the basis of b. Fusion experiments withone cell-selectable their hairs only two being inconclusive. In the marker. In five fusion combinations only the category of 32 parental phenotypes only 20 potatofusio n parent (AM10K)carrie da cell-selec ­ (63%)wer eidentifie da ssuc hb ythei rhair san d7 table marker, the kanamycin resistance gene, (22%)genotype s haderroneousl y beenclassifie d whereas thefou r S.n/o/um-relate dspecie s were assomati c hybrids;fiv e(16% )wer e classified as wild type genotypes. The number of kanamycin "inconclusive". Not a single somatic hybrid that resistantcall irange dfro m25 0t o 1450acros sth e was detected byth e isozyme test,wa s classified five experiments. Regeneration efficiency of the as "parental phenotype" by stem hair type. calli varied from less then 1% with diploid S. From these results we concluded that the americanumt o 56%wit h the hexaploid S. ameri- structure and density of stem hairs were useful canum accession as fusion parent. The single morphological markers for initial screening of somatic hybridtha tresulte dfro mth e combination somatic hybrids amongth e regenerants fromth e ameix(+ ) AM10K grew vigorously in vitrobu t did two fusion combinations investigated. not grow well in the greenhouse. The fraction of Nevertheless, isozyme tests were necessary for hybrids among the other four combinations was the confirmation of hybrid nature. high, ranging from 83% to 97%. Such high frequencies were expected from fusion combinations in which the faster regenerating Analysis ofthe fusion experiments parent lacked the antibiotic resistance. The proportion ofplant swit hgoo d growthcapacit yin Sixteen different fusion combinations were vitrorange dfro m20 %t o69% ,dependin g onth e performed in which one of the indicated S. combination. Inth egreenhous e only 15(16% )o f n/grum-relatedspecie swa scombine dwit hon eo f thengr-tbfy hybridsan d2 (2% )o fth e ame^rtbr^ the three genotypes of cultivated potato. All hybrids grew vigorously and flowered well, combinations are listed inTabl e 2alon g withth e whereas the remaining 365 hybrids performed results obtained. poorly or died within one month. In the combination S. chenopodioides a. Fusion experiment with two cell-selectable (+) AM10K no essential differences were found markers. In this experiment both fusion parents between the results of the experiments with one contained a cell-selectable marker; the potato or two cell-selectable markers.

23 Chapter 2 c. Fusion experiments without cell-selectable S. nigrum,possibl ydu et oth ehig hploid yleve lo f markers. Ten fusion combinations were the expected hybrids. Nevertheless, out of the performed with genotypes lacking a cell- 355 regenerants from ame^ (+) Desiree, eight selectable marker. The experiments involved all hybridswit ha nexpecte dploid yleve lo f 10xwer e the S. n/grum-related species on the one hand obtained, though none of these hybrids was and the potato genotypes 1029/31 (2x) and cv. viablei nth egreenhouse .Th efusio nexperiment s Desiree (4x) on the other. From the five with the two diploid species S. chenopodioides experimentswit hth ediploi dpotat ofusio nparen t and S. americanum yielded both 21 hybrids thecombinatio nwit hS .amehcanum (2x )wa sno t which is respectively 8 and 20% of the successful. Only two S. americanum(2x ) plants regenerants that were obtained.Th e majority of andno ta singl ehybri dregenerate dfro mth e15 0 these hybrids performed rather well in vitro,bu t calli obtained. In the remaining four only plants of the combination of combinations, calli (450-1800) and regenerants ame2x(+ ) Desireeshowe dvigorou sgrowt hi nth e (29-86%) could easily be obtained. As a greenhouse andflowere d profusely. consequence of the absence of cell-selectable The absence of a cell-selectable marker markers, the percentages of hybrids among influenced the fraction of regenerants that was those regenerants were low. The highest shown to be somatic hybrids. Among the ten frequencies, 43 and 30%, resulted from the combinations investigated this fraction varied combinations with the two hexaploid species,S . between 0 and 0.43, which is much lower than americanum (6x) and S. nigrum respectively, thefraction s inexperiment s withon eo rtw o cell- whereas only 2% of the 1549 regenerants of S. selectable markers (0.82-1.00). villosum(+ ) 1029/31wer e somatic hybrids. An example of the time frame of The performance in vitro and in the regeneration of somatic hybrids and parental greenhouse of the hybrids of these four fusion plants is shown in Table 3. It indicates the combinations was to a large extent comparable number of regenerants within a period of 12 toth eperformanc eo fsomati chybrid sfro mfusio n months that resulted from the four fusion experiments with AM10K. Although between 36 combinations with S. chenopodioides. In both and 69%o f the hybrids of all four combinations combinations in which a selection marker was performedwel lin vitro, onlyth ecombinatio nwit h involved,somati c hybridswer efoun damon gth e S. nigrum resulted in a reasonable number of first regenerants. When the experiments were hybrids that was vigorous in the greenhouse. conducted without a cell-selectable marker, the Onlyon eo fth e chp-tbfy hybridsan dfou r outo f regeneration of somatic hybrids appeared to be 138 ame6x-tbr2x hybrids grew vigorously and delayed. Fromth ecombinatio nchp (+ ) 1029/31, flowered. not a single potato regenerant was obtained, The results of fusion experiments with the whereas in the combination of chp(+ ) Desiree tetraploid cv. Desiree deviated from those potato regenerants appeared, although obtainedwit hth ediploi dpotat ofusio n parentsi n regeneration was delayed. Time frames of several respects. Regenerants were obtained regeneration inth efusio ncombination s withth e from all five combinations, even with S. other S.raorum-species wer ecomparable . americanum (2x),bu t no hybrids resultedfro m the fusion combinations with S. villosum and

24 o CD

T- CO

TO' ^CO_ C S> = «> _ a> -2 o -o p CO O O CD CO CM •* O CO CO § Q- - > .c S CM s

a> O CO a> CD co •* co -^- 05 l~- LO Q O

CO a> c O •o CO CD o -Q a>- -Q c CO CD CO •<- CM CO CM O CO CD CO CM o _cr a> CO O CO O CM CD i i CD E to en a> co CO -=*• e a> o T— o O CD O o O O O CD CD CD

s en en co -c

i— CO l~- CO CO LO CO CD CD CM

• to s

"5 tn o •*- en T- en co CM S N OI CO Ifl -O en o o co CO ^t CM CM o CD r^ co in T- in co co T- CM f- -^ co E 2 >• a> ES» »§ .9.32! c <= o e8n, oo E « 1^- t^- O) CO CD St CD LO a: <5

CD LO CD CD O -I- CD CD LO CD CD CD CD LO CD CD CD — .a CM LO CD LO CT> LO LO CD O LO LO CD CM LO CD LO •*»•f — CM T- TT S E s CD in Tt •* o CM T- CO CO (D •*

CD .a ^ as jr o Ji CD C

CO o •* -v v v v ^^^^^ co a> a> co co H o °* g* ?* 2 ?*2 ?*2 3C5O 3 C5O 3 5CO 3 5CO 3 5C O .t-a .)t -c .to= -O .iJ= -O.t =J -a) a> ^ ^ ^ ^ ^r CM CM CM CM CM CO CO (O en CO 2 2 2 S S CD O O CD CD -CO -CO -CO -CO -CO t£ '+*'+"+"+"+' '+"+*'+"+"+' 4^""""+^ F OJ :s a> cz o3 CO 3 eg CM

OINCOIOIDSCOT- ' CN O) t~- _3>

o o o •<— cooto ' om

5 000^-0)1-11)0) 'coco f ^ .—. CN LO

E cu

CL E •p ° O CO

CD <£. c ^ lo CD E -S o^-oo 'on 'ID >n LO O CD i- CN " E o ^ .'"2» i -ss oooo'oo'O'o o eg

>; = O O O O O • O O ' O O •« o O -1= £. 5 ™.c & o CM CM TJ- co co ' on) ' in o> £= -Q T- T- csi T- h-

&•§• >• o =8 = CD CO o to co 5; °S

CD g. CJ5 •* O) CN ' ' 00 ' ' CM i-S co ,_- CB a) O "C ^ -*: CD O O O ' ' O ' ' O CD a> E .5 .? o -s o jS CD O E -3? I— CO co =S CD CO

Nn^tiniDMioo)"^ CN .o Chapter2

The results shown inTable s2 an d3 clearl y The somatic combining abilities of the indicate that inth ecombination s tested a cell- individual S. n/grum-related species with the selectable marker is not essential for the three different genotypes of potato are selectiono fsomati c hybridsi fa quic kan dreliabl e summarised in Table 4. It is apparent fromth e morphological method todetec t putative hybrids data that each of the wild species wasabl et o isavailable .However , acell-selectabl e markeri s form somatic hybrids with thecultivate d potato. useful in those combinations from which large Furthermore, the fraction of somatic hybrids numberso fregenerant s havet ob escreened , for among the regenerants was highly variable and instance,whe nth efraction o fhybrid samon gth e probably influenced by the genotype as wella s regenerants is small, or when only a small the ploidy level ofth efusio n parents. proportion of the somatic hybrids grows vigorously in the greenhouse. In the combinations from which nota single vigorous Ploidylevel of somatic hybrids hybrid was obtained,i tmigh t be effective to use different genotypes of the species involved. Theploid yleve lo fal lfusio n parents involvedwas Indications ofsuc h genotypic effects were found known. A total of 245 somatic hybrids were after comparison of potato cv. Aminca withcv . subjected to flow cytometric analysis to Desiree infusio n experiments (data not shown). determine their DNA content, which gave an Infiv efusio n combinations with cv.Aminca ,onl y indication ofth e ploidy level ofth e hybrids. The one somatic hybrid was obtained. DNAconten to fth edesire d hybrids should equal the sum ofth e parental DNA contents. Deviating DNA contents mayoriginat e from asymmetric fusion, loss of chromosomes after fusion, or Table 4.Somati c combining abilityo fth ethre e potato fusion parents with each of the five species of the S. multiple fusion events. Assignment of ploidy nigrum- complex resulting in somatic hybrids. levelst o plantswit ha deviatin g DNAconten twas complicated because of different C-values between the S. n/grum-related species and ame chp vll ngr ame ploidy 2x 2x 4x 6x 6x potato. Noselectio n ofsomati c hybridsin vitroo rin 1029/31 2x - + + ++ vivo was applied before the flow cytometric

K analysis,excep tamon gplant so fth ecombinatio n AM10 2x +- +* + ++ ++ S. nigrum(+) A M10 K, from which most ofth e Desiree 4x + + poorly performing hybrids were removed.Withi n eachfusio ncombinatio n somatic hybridswit hth e *wit h andwithou t hygromycin resistancei n expected or deviating ploidy levels were S. chenopodioides. distinguished (Table 5). Clearly different results - =n osomati c hybrids were obtained from comparable combinations +- = <1%o fth ecall i resulted ina somati c hybrid K + => 1% but<10 %o fth e calliresulte di na somati chybri d like chp (+) 1029/31 andchp (+) AM10 , which ++ =>10 %o fth e calli resultedi na somati c hybrid only differ in the presence or absence ofth e kanamycin resistance gene. The fusions with AM10K resulted ina higher proportion of plants

27 Chapter2 with the expected DNA content than the In case adeviatin g DNA content was observed, comparable combination with 1029/31. the hybridswit ha nexpecte d ploidy levelo f4 x or Exceptions were the two fusion combinations 6x mostly contained a surplus of DNA, buta with ame&x, which showed equally high deficit when 8x was theexpecte d ploidy. In the proportions of hybrids with the expectedDN A combination ngr{+) 2x potato itwa s shown that content, indicating that the use of a cell- vigorous plants had,a sa rule ,th eexpecte d DNA selectable marker does not result in a higher content,wherea s non-vigorous plantsha dmostl y fraction ofexpecte d hybrids inthi s combination. a deviating DNA content.

Table 5. Analysis of the DNA content of somatic hybrids from different fusion combinations

Fusion combination Expected Observed numbero fsomati chybrid s ploidy with with total expected deviating DNAcontent 1 DNAconten t chp (+)1029/3 1 4x 1 16 17 chp (+)AM10K 4x 6 13 19 chp" (+)AM10K 4x 6 6 12 vll (+)1029/3 1 6x 4 17 21 vll (+)AM10K 6x 23 20 43 ametx (+) 1029/31 8x 25 11 36 amesx (+)AM10K 8x 13 7 20 ngr (+)1029/3 1 8x 12 34 46 8x 15 3 18 ngfl (+)AM10K 6x 3 0 3 ameix (+) Desiree 6x 6 4 10 chp (+)Desire e

1 Hybridswit hth esu mo fth e parentalDN Acontent s±10% . 2 Mosto fth e poorlygrowin ggenotype sfro mthi scombinatio nha dbee n removed.

Descriptionof somatic hybrids morphologically verysimilar .The ydiffere d mainly in thesiz e ofthei r leaves. The somatic hybrids All successful fusion combinations resulted in with ngra s parental genotype hadth e largest hybrids with in vitro growth levels ranging from leaves in vitro (approximately 2 cm long), vigorous to poor (Table 2).Somati c hybrids of whereas the smallest leaves (0.5 cm) were different fusion combinations were observed on in vitro c/jp-hybrids. Hybrids with

28 Chapter 2 artier as one of the parents showed more dying. Somatic hybrids with ame^ as a parent anthocyanin pigmentation than hybrids from showed both in vitro as in the greenhouse other fusion combinations. intense anthocyanin pigmentation. The in vivogrowt h capacity of 682 somatic hybridswa sanalyse d inth egreenhouse . Ninety- seven somatic hybrid plants died in vitro before DISCUSSION they could be transferred to the greenhouse whichwa s partly duet oinfection .Onl ysixt yi fth e Production of somatichybrids 682 plants grew vigorously and flowered profusely. Fifty-three of those hybrids resulted The four parental species of the S. nigrum from three fusion combinations, i.e. complex that were used in our fusion ngr{+) 1029/31, no/(+) AM10 K and experiments are closely related. S. villosum ame2x (+) Desiree. Plants of the first two (2n=4x) and S. americanum (2n=2x) are combinations resembled each other. They were assumed to be progenitors of the hexaploid S. very vigorous and could reach a height of nigrum (Edmonds, 1979). Our results show that approximately two meters. Leaves were simple, allfiv e accessions ofth e four species were able dentate and of light green colour and in some tofor m somatic hybrids with potato. cases leaves were slightly irregular. All plants The sixteen investigated fusion showed anearl y aging of leaves.Th e sizeo f the combinations showed many differences in flowers was intermediate between those of somatic combining ability.Th e ploidy level ofth e potato and S. nigrum. Petals were white and species isa factor that may have influenced the folded back for three quarters of the length. results. Some combinations between parental Some tuber-like structures were obtained from plants with a relatively high ploidy level did not three somatic hybrids. Plants of these fusions yield viable hybrids. Although potato performs resembled ngr more than potato. poorlya thighe r ploidylevel s(Pijnacke ran dSre e The ame2x(+ ) Desiree hybrids had a Ramulu, 1990), high ploidy levels need not be completely different morphology. Plants reached troublesome since S. nigrum is still fertile as a height of one meter at most. Leaves were dodecaploid genotype (Singh and Roy, 1985). compound,irregular , and dark green.Th e plants Our results imply that, besides ploidy level, also produced many flower buds, which rarely the speciesa s such isa determinin g factor, asi s opened. Petals hada ting eo fpurpl ean d opened shown by the differences in somatic combining half, if they did at all. Some hybrids produced ability of the diploid species S. chenopodioides tuber-like structures. In contrast to the and S. americanum2xwit h S. tuberosum. ngr{+) 1029/31an dngr (+)AM10Khybrid sthes e The different results of fusion experiments plants resembled potato. with cv. Aminca and with cv. Desiree clearly Among the plants of three other fusion indicated a genotypic effect on somatic combinations, ameex (+) 1029/31, ame6x(+ ) combining ability. In interspecific somatic AM10K and chp (+) 1029/31, some well growing hybridisations,th eeffec t ofth egenotyp ewithi na and flowering hybrids were found. Plants of all specieso nsuccessfu l hybridisation hasno tbee n other combinations died or grew very poorly in described in detail in literature, but interspecific vivo. Some genotypes flowered shortly before sexual hybridisation has clearly shown the

29 Chapter 2

importance of this factor in many combinations. occurred among the regenerants as well. Either Hermsen (1966), for instance, presented the these regenerants were asymmetric hybrids in differences in berry set among 538 different which the tor-allele encoding for 6-PGDH was crosscombination s between2 0accession s ofS . lost, or the selection pressure was not strong acaule and 28 of S. bulbocastanum. Genotypes enough. of both species differed in their intercrossability from (very) good to bad. Recently Watanabe ef al.(1995 ) found indications for genotypic effects Vigourof somatic hybrids invitr o and in vivo in the results of their crosses between diploid potato and non-tuberbearing Solarium species. Ourresult sclearl ysho wtha tsomati c hybridstha t Since sexual and somatic interspecific are vigorousin vitroare easil y obtained,sinc e at hybridisation both result in the combination of least 20% of the hybrids of each successful genetic material of two different species in one combination performedwel lin vitro. Nevertheless plant, it is assumed that also in somatic it was observed that only from the combinations hybridisationth ecombinatio no fdifferen t parental ngr{+) 1029/31, ngr(+) AM10K and genotypes iso f importance. ameix(+ ) Desiree a substantial part of the hybrids was also vigorous in the greenhouse. Three other combinations resulted in a low Selection and detection of somatic hybrids percentage of relatively well growing hybrids:

chp(+) 1029/31, ame6x(+) 1029/31 and Comparison of the fractions of hybrids resulting ame6x(+)AM10K. The latter two combinations from experiments with zero, one or two cell- are comparable in ploidy to ngr (+)tbr2x- Among selectable markers ledt oth econclusio n that the the hybrids of seven other combinations, not a selection for these markers did not influence single plantwa sabl et ogro w inth e greenhouse. somatic combining ability. However, in some The DNA measurements showed that the combinations, a cell-selectable marker is useful percentage of hybrid plants with the expected to minimise the effort necessary to obtain and DNA content varied between the different identify a sufficient number of somatic hybrids. combinations but that most well-performing This is especially true when no easy detection hybrids inth egreenhous e hadth eexpecte d DNA method is available like the type and density of content. Thus, DNAmeasurement s can be used hairs onth estem .Th e useo ftw o cell-selectable as an additional means of indirect selection ofa markers did not havea nadvantag e overth e use higherfractio n ofhybrid stha t arevigorou s inth e of only one such marker. Contrary to our greenhouse. expectation,th efusio ncombinatio n inwhic h both The striking contrast betweenin vitroan din fusion parentscontaine da cell-selectabl e marker vivovigou rwa sobserve dearlie r inpotato-tomat o did not result in the isolation of 100% somatic hybrids.Severa lgenotypi ccombination s resulted hybrids.Th esam ecombinatio no f cell-selectable insomati c hybridstha t performedver ywel li nth e markers,use d infusio nexperiment s involving S. greenhouse, whereas plants of other genotypic tuberosum and S. brevidens, resulted in 100% combinations alldie din vivo(Garriga-Calder e ef somatic hybrids (Jacobsen ef al., 1993). In our al., 1997).Poo rperformanc e ofpotat o (+) tomato experiment, some S.chenopodioides genotypes somatic hybrids in vivo was also reported by

30 Chapter 2

Schoenmakers et at. (1993), but in their case incongruity isstronge rwhe nth especie sar e less also a low frequency of vigorous plants was related.However ,th efacto ro frelatednes so fth e obtained. species does notexplai n the striking differences A possible cause of poor vigour of somatic invigou r between hybrids of the different fusion hybrids in the greenhouse might be organelle- combinations we performed since the nucleus incongruity. However, in our hybrids phylogenetic distances between eacho fth efou r incongruity between chloroplasts and nucleus is S. n/arum-related species and potato are unlikely, since Gressel etat. (1984) showed that expected to be similar. the ngr (+) tbr hybrids, that resulted from their Itca nb econclude dtha tthre eou to f sixteen fusion experiments, grew well in vivo whether fusion combinations provided flowering hybrids they contained S. nigrum chloroplasts or potato that can be used in crossing experiments with chloroplasts. Wolters (1996) stated that somatic potato as backcross parent.

31 3

Successful first and second backcrosses ofS . nigrum (+) S. tuberosum somatic hybrids with both Solanum parents

KarinHorsman ,Richar d Fratini,Dirk-Ja n Huigenan d EvertJacobsen .

The Graduate School of Experimental Plant Sciences, Wageningen University, Laboratory of Plant Breeding, P.O. Box 386,6700A J Wageningen,Th e Netherlands.

This chapterwas publishedin Sex. Plant. Reprod. 12:144-151(1999) with modifications

ABSTRACT

Somatic hybrids of Solanum nigrum (+)2x potato were successfully crossed with S. nigrum and with potato. First and second generation backcross progeny with S.nigrum coul d easily be obtained.On eo f the BC1 genotypes wasalread y self-fertile. Backcrosses with potato hada much lower success rate. Only pollinations with tetraploid potato resulted in seed containing berries. TwoBC 1 genotypes were obtained after 5000 pollinations from which 505 ovules were cultured. Thefirs t BC1 genotypewa s vigorously growing in vitro and inth egreenhous e and flowered abundantly. Thesecon d BC1 showed many abnormalities and dropped its flowers before anthesis. Thefirs t BC1 was again crossed with tetraploid potato and also inthi s generation the success ratewa s low. Over 5000 pollinations resultedi n 1750 berries from which over 3000 ovuleswer e obtained.Twelv e plants germinatedfro m theseovules , which were nota svigorou s in vitro and in vivo asth eBC 1 parent. Some ofth e BC2 genotypes were usedfo r further backcrosses but sofa rn o BC3 plants could be obtained. BC1 andBC 2 genotypes that resulted from the backcross programme with potato were tested for their resistance to Phytophthora infestans.Th e BC1 genotype was asresistan t as the S.nigrum fusio n parent but among the eight BC2 genotypes scored six were resistant whereas twogenotype s showing lesions were concluded tob e susceptible.

33 Chapter 3

INTRODUCTION 3000 ovules that were cultured in vitro. The majority of 114 BC1 genotypes that was Somatic hybridisation followed by recurrent obtained, germinated from 865 ovules from backcrossing is a method for the introduction of crosses between the hexaploid somatic hybrids useful traits from recalcitrant species into a and tetraploid potato (Jacobsen ef al., 1993). recipient genome. In potato breeding this Ehlenfeldt and Helgeson (1987) reported similar approach has been used for several species results from their crossing experiments with like S. brevidens (Austin etal., 1988; Ehlenfeldt fusion hybrids of S. brevidens (+) potato. & Helgeson 1987; Jacobsen ef a/., 1993), S. The production of somatic hybrids between bulbocastanum (Austin ef a/., 1993) S. species of the S. nigrum complex and potato circaeifolium (Mattheij ef a/., 1992) and tomato has been described by Horsman ef al. (1997). (Lycopersicon esculentum) (Jacobsen ef a/., Diploid, tetraploid and hexaploid S. nigrum 1994). Successful backcrossing of somatic related species were hybridised with 2x andAx hybrids depends on various factors, like the potato genotypes in order to produce somatic phylogenetic distance of the parental species hybrids with a range of ploidy levels from 4x to and the genotypes of the fusion parents. The 10x. Only the fusion combinations 6x S. nigrum ratio of the parental genomes was shown to be (+) 2x potato and 2x S.americanum (+ ) Desiree a major factor in the backcrossing programmes resulted in plants that were vigorous in the of S. brevidens (+) potato and of potato (+) greenhouse. S. nigrum (+) 2x potato somatic tomato fusion hybrids. hybrids have also been described by Binding et Jacobsen et al. (1994) reported the first al. (1982) but no backcrosses were reported by successful backcross of a potato (+) tomato them. In the present report the successful first somatic hybrid with potato, 16 years after the and second generation backcrosses of the S. first somatic hybrids of this intergeneric nigrum (+) 2x potato fusion products with both combination were described (Melchers et al., potato and S. nigrum have been described. The 1978). Garriga-Caldere ef al. (1997) described influence is shown of the ratio of the genomes the production of more BC1 plants and of in these somatic hybrids on the success rate of several BC2 populations. All BC1 genotypes backcrosses with potato compared to had resulted from crosses of hexaploid somatic backcrosses with S. nigrum. hybrids containing four genomes of potato and The aim of the backcrossing programme two genomes of tomato with tetraploid potato with potato was the transfer of the resistance of genotypes as the male parents. Attempts to S. nigrum to Phytophthora infestans into produce backcross progeny from tetraploid cultivated potato. S. nigrum as well as the fusion products had not been successful. sexual hybrids of S. nigrum related species and Backcrosses of the hexaploid S. brevidens potato (Eijlander and Stiekema 1994) were (+) potato somatic hybrids with tetraploid potato shown to be resistant to P. infestans, based on were also more successful than backcrosses of a strong hypersensitive reaction (Colon et al., the tetraploid fusion products. Backcrosses with 1993). The genotypes that resulted from the diploid potato gave poor results on both backcrosses of S. nigrum (+) 2x potato fusion hexaploid and tetraploid somatic hybrids.A total products with potato were analysed for their of over 140 plants was obtained from more than resistance to P.infestans.

34 Chapter 3

MATERIALSAN D METHODS were grown in a greenhouse from February to October. The somatic hybrids and the Plantmaterial and culture backcross genotypes were grown according to government safety rules for genetically modified Somatic hybrids of three fusion combinations organisms. were used as female parents in crossing experiments: 1) 2x S. chenopodioides(+ ) 2x S. tuberosum clone 87.1029/31; 2) 2x S. Ovuleculture americanum (+) S. tuberosum cv. Desiree; 3) 6x S. nigrum (+) 2x S. tuberosum clone AM10. Berries of the somatic hybrids were harvested S. chenopodioides (2n = 2x = 24), S. between 10an d 15day s after pollination (DAP) americanum(2n =2 x =24 ) and S.nigrum (2n= for ovule rescue. Berries of the BC1 DJ93-6738 6x = 72) all belong to the S. nigrum complex of were harvested 16-52 DAP,bu t most frequently the section Solanum. The diploid S. tuberosum between 19 and 25 DAP. Ovule culture was genotype 87.1029/31 was an amylose-free performed using HLH-medium according to mutant described by Jacobsen ef a/. (1989). Neal and Topoleski (1983). Ovules were AM1 0 was a transformant of clone 87.1029/31, transferred tofres h medium every fortnight. containing the GUS-gene and the kanamycin resistance gene (Horsman era/., 1997). The S. chenopodioides (+) 1029/31 and STS-treatment the S. americanum (+) Desiree somatic hybrids were pollinated with the cultivars Desiree, BC2 genotype KH94-9017 was treated with Frieslander, Gloria and Katahdin and with the silverthiosulphate (STS) to prevent early flower tetraploid breeding clone AM66-42. The potato drop. STS was prepared by adding a 0.4 mM backcross parents for the S. nigrum (+) AM1 0 silvernitrate solution to anequa lvolum e ofa 3.2 somatic hybrids were: 1)th e tetraploid breeding mM sodiumthiosulphate solution containing clones AM66-42, HB93-7133-1, HB93-7133-2, 0.1% Tween 20. The entire plant in the HB93-7133-3, HB93-71334 and HB93-7133-5; greenhouse was sprayed once aweek . 2) the cultivars Desiree, Escort, Hertha, Monalisa, Mansour and Vital. The S. nigrum backcross parent was the same genotype as GUS assay used in the initial fusion experiments between diploid potatoan d S.nigrum. The method used for screening of the GUS- Backcrosses of the BC1 and BC2 activity was based on Jefferson ef al. (1987). genotypes were performed with the cultivars Young leaves of in vitro plants were incubated Adora, Desiree, Frieslander, Fresco, Gloria, in 50 mM sodium phosphate buffer pH 7.2 for 1 Katahdin, Mansour, Sebago, Turbo and the h. The leaves were transferred to GAB-buffer breeding cloneAM66-42 . containing 50 mM sodium phosphate, 10 mM Somatic hybrids and backcross genotypes EDTA (ethylene diamine tetra acetic acid), were maintained in vitro. Plants were grown on 0.1% (v/v) Triton X-100, 0.1% R-mercapto- basal MS mediumwit h 30g/ lsucrose .Al l plants ethanol and 1 mM X-Gluc (5-bromo 4-chloro 3-

35 Chapter 3 indol (i-glucuronide). Incubation took place Inoculations were performed in the afternoon. overnight at 37°C. Leaves were screened for After inoculation the lights were switched off the occurrence of blue coloured regions. and the moisturiser was switched on overnight. The occurrence of lesions was observed 3,4,5, 6 and 7day s after inoculation. Determination ofkanamycin resistance

Backcross genotypes were tested for the RESULTS presence of the NPTII-gene. Shoots of all genotypes were put on MS-medium containing Crossesof somatic hybrids of2x species of the 30 g/l sucrose and 100 mg/l kanamycin. After S. nigrum complex (+) 4x or 2x S. tuberosum one, two and four weeks the formation of roots withpotato varieties was determined. Absence of roots indicated susceptibility to kanamycin. From six hexaploid hybrids of the fusion combination 2x S. americanum (+) Desiree two plants of each hybrid were put in the Determination ofresistance to P. infestans greenhouse at three time points during the crossing season. As expected these hybrids In vitro plants were first transferred to the morphologically resembled potato more than S. greenhouse and four weeks later to a americanum since they contained two genomes screencage in order to adapt the plants to the of the latter species and four of potato. Leaves outdoor environment. At an age of eight to ten showed some abnormalities but flowers had a weeks the plants were transferred to a climate normal morphology. Early in the season all six chamber with a temperature of 15°C and a hybrids dropped most of their flower buds daylength of 16h. Three days after their transfer before anthesis but later in the season it was the plants were put in a plastic tent within the possible to carry out pollinations. Vigour and climate chamber with a moisturiser. Before flowering capacity varied between the different inoculation the moisturiser was switched on for S. americanum (+) Desiree hybrids. Two of four hours. them, F54-3 and F54-16, reached a height no P. infestans isolate 90128, race 1,3,4,6,7, more than 0.25 m, and produced only two and 8,10,11, which was provided byth e Department five flowers, respectively, that could be of Phytopathology of Wageningen University, pollinated. F54-1, F54-9 and F54-11 were was used to make the inoculum. Ten ml of moderately vigorous, had a height of sterilised water was added to two rye-agar approximately 0.4 m. and flowered better. A plates with mycelium of P. infestans. The plates total of 24, 16 and 33 pollinations respectively were put at a temperature of 4°C for two hours could be made. The most vigorous genotype to induce the release of zoospores. Five leaves was F54-10 of which 94 flowers could be starting with the first full-grown leaf were pollinated. In all these cases no berries were inoculated on the lower epidermis with a 10 ul obtained. drop of a zoospore suspension with a The fusion combination S. chenopodioides 4 concentration of 4.10 zoospores per ml. (+) 2x potato resulted in one flowering,

36 Chapter 3 tetraploid somatic hybrid. Most flowers dropped berry set were the genotypes with the lowest before anthesis. The remaining flowers of six production of ovules per berry. The six plants of this hybrid were pollinated with pollen genotypes with the lowest berry set had the of tetraploid potato varieties. A total of 31 highest production of ovules per berry. A total pollinations did not result in any berry set. number of 505 ovules was cultured from 1750 berries. Two of the cultured ovules developed into Crossing experiments with somatic hybrids of BC1 plants. The first BC1 genotype, 6xS . nigrum (+)2x potato DJ93-6738, resulted from the cross F21-26 x AM66-42. Rescue was applied at 15 DAP. After Backcrosses with S. tuberosum: BC1, BC2 and 14 weeks the first sign of germination was BC3: The most vigorous hybrids resulted from noticed. The seed coat cracked and two weeks fusion experiments between a hexaploid S. later agree n callus had emerged.Afte r another nigrum accession and a diploid potato two weeks during which shoot meristems had genotype. Eighteen somatic hybrids were developed the callus was transferred to shoot pollinated with diploid and tetraploid potato induction medium for twelve days. genotypes. Pollinations with diploid potato Subsequently the callus was cultured on MS13 mainly resulted in empty berries (data not + 30g/ lsucrose. Twelve shootswer e harvested shown). The results of pollinations with which were subcultured as individual tetraploid potato genotypes are summarised in genotypes. Table1 . No morphological differences were found Between 24 and 469 pollinations per between the twelve BC1-shoots and therefore genotype were made with twelve tetraploid they were considered to be one and the same potato genotypes. HB93-7133-2 was the only genotype. Isozyme analysis using SKDH potato genotype that did not induce berry demonstrated the hybrid nature of the BC1, formation.Al l other genotypes produced berries being a descendant of F21-26 and AM66-42. and ovules. The somatic hybrids F21-5, -7, -16 BC1 DJ93-6738 was a vigorously growing and -45 were removed from the crossing genotype that was able to reach a height of programme because of their stunted growth, several meters. As expected from the genomic parthenocarpic berry set and low number of constitution, leaf and flower morphology were ovules per berry. Among the remaining 14 intermediate between S. nigrum and S. genotypes berry set varied from 11t o64% . tuberosum. Some pollen could be collected Ovule rescue was performed between 9 from the flowers but their stainability with and 19 days after pollination (DAP). Most lactophenol acid fuchsin was just 10-15%. berries were harvested before 15 DAP because Formation of tuber-like structures was berry abscission increased exponentially after observed, whereas the somatic hybrids showed this point in time. Among the 14 selected only some stolons with afe w thickened parts. somatic hybrids the number of ovules per berry The second BC1, KH93-9100, germinated varied from 0.11 to 0.90. Ten somatic hybrids directly from an ovule that originated from the produced between 0.33 and 0.57 ovules per cross F21-59 x Desiree and was rescued at 13 berry. The eight genotypes with the highest DAP. Under in vitro conditions this genotype

37 Chapter3

Table 1.Result so fbackcrosse so fS .nigrum (+)2 xpotat osomati chybrid swit htetraploi dpotat ogenotypes .

Somatic Numbero f Numbero f Berry Number Ovules/ Numbero f hybrid pollinations berries set of berry BC1 plants ovules

F21-4 366 117 32% 39 0.33 0 F21-5 110 66 60% 5 0.08 0 F21-6 395 238 60% 33 0.14 0 F21-7 24 5 21% 0 0 - F21-8 469 227 48% 85 0.37 0 F21-9 465 189 41% 73 0.39 0 F21-15 275 106 39% 18 0.17 0 F21-16 29 1 3% 0 0 . F21-24 66 14 21% 6 0.43 0 F21-26 133 14 11% 8 0.57 1 F21-30 457 274 60% 29 0.11 0 F21-35 296 190 64% 67 0.35 0 F21-37 273 42 15% 38 0.90 0 F21-44 164 24 15% 10 0.42 0 F21-45 45 42 93% 4 0.10 0 F21-49 301 45 15% 25 0.56 0 F21-58 299 96 32% 36 0.38 0 F21-59 194 60 31% 29 0.48 1

Total 4362 1750 36% 505 0.40 2

was nota s vigorous as the BC1 described in showed the highest number of berries per the previous paragraph. In the greenhouse it pollination inal lthre e seasons with anaverag e reached a height of approximately 1 meter, of 39%, whereas cv. Gloria hada berry set of showed deformation of leaves anddroppe d its only 14%.Beside s the pollinator, also plant age flowers beforeopening . and temperature influenced the crossing results During three successive seasons crosses were considerably. For instance, berry set varied made with the vigorously growing BC1 from 15%i n 1995t o 28%i n 1996.Th e low genotype DJ93-6738. Itwa s mainly used as the percentage in 1995 was largely due toth e high female parent. Diploid and tetraploid potato temperatures in the greenhouse in July and genotypes, S. demissum and S. nigrum were August. used as pollinators. Pollinations with S. A total of 5474 pollinations with tetraploid demissum, S. nigrum and diploid potato potato genotypes resulted in 1251 berries genotypes did notresul t inberr y set. Only after (Table 2). Ovules were rescued from these pollination with pollen of tetraploid potato berries between 9an d 52DA P with an average genotypes seed containing berries were of 22.7 DAP. Seventy-six percent was rescued obtained (Table 2). Berry setdepende d among between 17an d2 4DAP .A tota lo f306 5 ovules other factors onth e pollinator. Cv. Frieslander was rescued from which less then 1%(2 7

38 Chapter 3 ovules) germinated. Fifteen embryos died the BC1 parent, but still showed abnormalities shortly after germination or showed severe because of the presence of S. nigrum abnormalities, whereas twelve embryos chromosomes. developed into complete plants. Ten of these KH94-9017 and KH96-9320 were less BC2genotype s showeda good growth capacity vigorous than the previously described in vitro (Table 3) which was comparable to the genotype but could still reach a height of vigour of the BC1 parent. Two genotypes were approximately 1 meter. KH94-9017 showed obviously lessvigorous . early senescence of leaves and dropped its All BC2 genotypes were analysed for their flower buds prematurely. STS treatment kanamycin resistance and GUS-activity, since prevented this early flower drop but also the potato fusion parent contained T-DNA with induced the formation of some spontaneous both genes (Table 3). Both BC1 genotypes and berries which were all empty. Most of the BC2 ten BC2 genotypes were still able to form roots genotypes produced spontaneous berries on kanamycin containing medium and showed during a particular part of the growing season, GUS-activity. Only KH96-9017 and KH96-9304 mostly at the end. KH96-9320 showed were kanamycin susceptible and had no GUS- spontaneous berry set at the beginning of the activity. Since the results of the tests for season. Later on 20 pollinations could be kanamycin resistance and GUS-activity were in carried out which resulted in five berries (Table agreement, the latter two genotypes were 2). Four ovules were rescued and cultured on considered not transgenic. HLH medium. After four weeks one of those Reciprocal crosses were also performed ovules germinated but the resulting seedling but pistil preparations showed that BC1 pollen stoppedin vitrogrowt ha t a heighto f 2cm . was not able to produce pollentubes . Some pollinations could be made on the BC2's All twelve BC2 genotypes were KH94-9072, KH94-9084, KH96-9379 and propagated and transferred to the greenhouse. KH96-9380 (Table 2). These genotypes were Compared to the BC1 parent all BC2's were stunted, showed abnormalities and produced less vigorous (Table 3). KH94-9019 was the spontaneous berries. Some of these most vigorously growing genotype that reached pollinations resulted in a few empty berries. a height of approximately 1.5 meter. Because KH94-9026 and KH96-9304 did not produce of the second backcross to potato, leaf any flowers. Grafts onto tomato or S. nigrum morphology was as expected to be more were made of all genotypes but no potato-like than leaf morphology of the BC1 improvement of flowering and/or vigour was DJ93-6738. KH94-9019 flowered abundantly observed. Three BC2 genotypes, KH94-9053, with normal looking white flowers. Stylar KH96-9345 and KH96-9365, were not able to preparations showed that pollen of several grow in vivoa t all. All plants of these genotypes tetraploid potato genotypes, for instance Gloria died a few weeks after they were transferred to and Desiree, were able to form pollen tubes, thegreenhouse. but 1155 pollinations did not result in any berry set (Table 2).Th e tuberstha twer e produced by Backcrosses withS . nigrum: BC1and BC2 this BC2 genotype and also by the other BC2 The S. nigrum (+) 2x potato hybrids were genotypes were more potato-like than tubers of backcrossedwit h S.nigrum a swell .

39 1-22 § 3 Z >™ E„

i_ CO ® 3 o Eos => cn ~ •z. > .£

E E-«

m c\i 3 a.o

in i-i oCD CO

a. CO 2f (o -^_ CN

O CO CD ion ^t CM O CO co in CD i— CM CM E^ -•3- T— T— in ^

03 Q. X 2 c: CM CM CM CM CM CM CM o 0) O OOO o O CD o00 o00 o0 0 o0 0 m m QQ jO "5 CO a> co 1^ <3) CM '^• 0 0)0 rV CO •>- f- t^ CO CM Is- CO IX. a> r^ co co co csi Q. CO oooO) O) C7> Oo) O) C7> O) CD o CO •4- •>*• -4 -4- CO CD CO c o> C7> C7> G) G) C7> O) C7) en a> X X X X XXX 1— O Q i£ *L XL ^ i£ :*: :*: a.

2S& CO w to >£ « .S o CD Di co o: d CE Oz c

.&• O 8 + + • + +

F CO CO

CO CD

I ICD .O) c

s * CD .£

CD .CD 1:1 o 5 + + + + + + + + a? §- O S

$co Q. < T— CO T— CO- *C O (DCOCC-OO •&o *i S£ CM CM -*d- CM CMC\ l T-C OC MC MC MC M c CO 05 CD

CD . „ CD CD CD CD £Z CZ CZ TZ" •a 1 CZ CZ ZJ CD CD jz _c mr? co CO CO O -CD -CD CM CO CO o CO SL l_ 2 I - "co co co -2 co -a> -2 .32 "l_ ZJ -CD co •£ LL. CO -CD CO "O LL 1— Sr CO 2 Q o S -03 X X X X X X X X X X X X < Q CO 0O CO CO CO oo CO OO CO CO CO CO "1 CO CO CO CO CO CO CO CO CO CO CO CO CO ". X X r^. 1^- l~- 1^- 1^- 1^. t^- r~- N- r~- N- 1^- CD CD CO CO CO CO CO CO co co CO co co co CM LO CO CO CO CO CO CO CO CO CO CO CO CO IK x— T— CJ> CD CD CD CD CD CD CD CD CD CD CD M- C U_ Q Q Q Q Q Q Q Q Q Q Q Q 8'o oI il co o h-C DC OC OC M• * ^-o m LOcj )o CO CO T— T— CML Ol> ~C O O CM•*! - CO h-C O co o o o OO o o COC OC OC OC OC O o >. COC D CDC DC DC DC DC D CDC DC D CDC DC D -*i -3 "i•" *i ^l "-= i *- *i i COi C OiC O Ci OC Oi C Oi i . CO <" 52 CD CDC DC DC DC D CDC DC D CDC DC D o CDC D O -5 X xx x x x x X X X X X X .2! Q. Q ^ •9 < m m x: ^ •*: m m x: m m ^ m <0 Q Chapter 3

Pollinations were made during the winter, which Resistance to P. infestans is not the optimal season for performing crossing experiments. The S. nigrum fusion The resistance to P. infestans of the somatic parent that was maintained in vitro, was used hybrid F21-26, the BC1 genotype DJ93-6738 as pollinator. A total of 31 pollinations on and eight BC2 genotypes was determined and somatic hybrid F21-6 resulted in 18 berries that compared to the resistance of S. nigrum. During contained 31 ovules. Although the number of two successive seasons drop inoculation ovules per berry resulting from backcrosses experiments were conducted with all previously with S. nigrum was higher than from mentioned genotypes, except for the BC2 backcrosses with potato, it was still significantly genotypes KH96-9320, KH96-9379 and lower than the number of ovules normally KH96-9380. Plants to test these genotypes observed in berries of S.nigrum o r potato. were naturally infected in the screencage in Thirteen BC1 plants were derived from 31 1997. The occurrence of lesions on these cultured ovules. Four of these BC1 plants were plants was determined inth e screenhouse. transferred to the greenhouse for further Results of the drop inoculation analysis and crossing experiments. One experiments and the spontaneous infection are genotype, DJ93-6732-1 was stunted and did summarised in Table 3. The somatic hybrid not flower. The remaining three genotypes, F21-26 and BC1 genotype DJ93-6738 were as DJ93-6732-2, -3 and -5, were vigorous plants resistant as the S. nigrum fusion parent. Some and did flower. Pollen stainability of BC1 plants necrotic spots could be observed on the 6732-2 and -3 was low, 1 and 10% inoculation site as they could also be found on respectively, but genotype 6732-5 was self- the resistant S. nigrumcontro l plants. fertile with 75%polle n stainability. The BC2 genotypes KH94-9017, KH94-9019, The three selected BC1 genotypes were KH94-9026, KH94-9072 and KH94-9084 were backcrossed with S. nigrum again. Berry set also tested with the drop inoculation method. was high, for instance 75% for 6732-3. Ovules From these genotypes KH94-9019 developed could be rescued from all berries, which easily lesions and was concluded to be susceptible. resulted in three BC2 populations of 12, 27 and The remaining four BC2 genotypes formed 11 plants respectively. Pollinations with pollen necrotic spots like S. nigrum or in some of 6732-5 on each of the two other BC1 instances slightly larger necrotic regions and genotypes, 6732-2 and 6732-3, also resulted in were considered resistant. two populations of 13 and 11 plants. Fifteen After the natural infection with P. infestans BC2 genotypes were obtained from selfing of in the screenhouse lesions were observed on 6732-5. leaves of KH96-9320. This genotype was Compared to the backcross programme therefore concluded to be susceptible to P. with potato, the success rate of backcrosses of infestans. The BC2 genotypes KH96-9379 and the S. nigrum (+) 2x potato somatic hybrids with KH96-9380 did not show any lesions. Most S. nigrum was much higher, and more likely these two BC2 genotypes were resistant. comparable to results from backcrosses of S. So from eight BC2 genotypes six were resistant Arevidens (+) potato somatic hybrids with and two were susceptible to potato late blight. potato.

42 Chapter 3

DISCUSSION ovules per berry was much higher. On average 35 ovules per berry were cultured in the potato Crosseswith somatic hybrids (+) tomato crossing experiments, whereas from the S.nigrum (+) potato crosses only 0.4 ovules This report described the first backcrosses that per berry could be rescued. Nevertheless only were made with S. nigrum (+) potato somatic one BC1 plant germinated from the almost hybrids. In 1982 Binding ef al. reported about 3500 ovules that resulted from the potato (+) the first fusion hybrids of this combination, but tomato backcrosses with potato. Crosses with crossing experiments were not mentioned. the4 x potato (+) tomato hybridsdi d not result in Eijlander and Stiekema (1994) succeeded in any progeny. obtaining sexual hybrids from the cross S. Differences in crossability were also nigrum x Desiree and S. nigrum x a diploid observed between S. brevidens (+) potato breeding clone. Until now successful somatic hybrids with different genomic backcrosses of these sexual hybrids have not constitutions. Ehlenfeldt and Helgeson (1987) beenreported . were the first to produce backcross progeny Backcrosses of the S. nigrum (+) potato from this combination. They used tetraploid somatic hybrids with both S. nigrum and with hybrids with two genomes of both species and potato were described in this report. hexaploid hybrids that resulted from fusion Considerable differences between both experiments between S. brevidens and a crossing programmes were observed. tetraploid potato genotype. Tetraploid somatic Backcrosses with S.nigrum had a high success hybrids crossed poorly with both 2x and 4x rate. Only a few pollinations were necessary to potato genotypes, whereas the hexaploid S. obtain twelve BC1 plants from which one was brevidens (+) potato somatic hybrids crossed already self-fertile. BC2 progeny was easily very well with tetraploid potato. Jacobsen ef al. obtained. (1993) showed similar results after Backcrosses of the S. nigrum (+) potato backcrossing S. brevidens (+) potato somatic fusion hybrids with potato had a much lower hybrids. In their experiments the hexaploid success rate. Over 4300 pollinations resulted in somatic hybrids with four genomes of potato 1750 berries from which 505 ovules were crossed better than the tetraploid fusion obtained. Only two BC1 plants resulted from in hybrids. vitro culture of these ovules. Jacobsen ef al. Rokka efal. (1994 ) were also successful in (1994) and Garriga-Caldere (1997) showed an producing backcross progenyfro m S.brevidens equally low success rate in their crossing (+) potato somatic hybrids. In their experiments with potato (+) tomato somatic experimentsthe hexaploid somatic hybrids, hybrids. Both tetraploid somatic hybrids (2x although being aneuploids, could be used as potato (+) 2x tomato) and hexaploid ones (4x male parents, in contrast to Ehlenfeldt and potato (+) 2x tomato were crossed with 4x Helgeson (1987) and Jacobsen ef al. (1993) potato genotypes. Crosses with the hexaploid who were only successful when the somatic somatic hybrids resulted in a berry set of 25%, hybridswer e useda sth efemal e parent. which was comparable to the berry set in our According to Ehlenfeld and Helgeson experiments. On the other hand, the number of (1987) their results with S. brevidens were in

43 Chapter 3 accordance with the EBN hypothesis. Genetic So in the crossing experiments of the studies of EBN have shown that a slight EBN potato (+) tomato somatic hybrids with potato excess on the female side can be overcome. the first backcross was the main obstruction. This might explain the success of the 6x Crosses of the S. nigrum (+) potato somatic (5EBN) S. brevidens (+) S. tuberosum hybrids hybrids with potato faced a new obstruction in in comparison to the 4x (3EBN) hybrids in every generation. This might be caused by the crossing experiments with potato. The EBN genomic constitution of the initial somatic hypothesis could not be applied to our crossing hybrid, which contained two genomes of potato experiments since no sufficient crossing data of and six of S. nigrum. The BC1 that resulted S. nigrum data with standard species are from a cross with 4x potato was expected to be available. Another aspect was that EBNs are a hexaploid genotype, that contained three based on normal seed set in ripe berries. Our genomes of both species. The plants that results related to crossing experiments in which resulted from a second backcross with 4x ovule culture was applied 2-3 weeks after potato were most likely to be pentaploid, but pollination. aneuploids for both species. During the backcross programme of the S. nigrum (+) potato hybrids the genomic ratio had to be Development ofBC2 and BC3progeny reversed. In the potato (+) tomato backcross The main difference between the S. nigrum (+) programme the somatic hybrids as well as the potato and the potato (+) tomato backcross BC1 and BC2 genotypes contained four experiments with potato, was the production of genomes of potato and an additional number of BC2 and BC3 progeny. The S. nigrum (+) tomato chromosomes that depended on the potato BC1 was again crossed with tetraploid generation. All hybrids had a surplus of potato potato genotypes, resulting in an average berry genomes. In this report we described crossing set of 23% and 2.5 ovules per berry, the latter experiments with somatic hybrids of a diploid S. being an improvement compared to the number mgrum-related species and cv. Desiree which of ovules per berry from the somatic hybrids. had a genomic ratio that was comparable to the Germination rate of the ovules was again low, hexaploid potato (+) tomato somatic hybrids. less than 1%. Only 12 plants with good in vitro Flowers of the S. americanum (+) Desiree vigour germinated from over 3000 cultured fusion hybrids were pollinated with pollen of ovules. In the greenhouse all BC2 genotypes several potato genotypes, but no berry set was were less vigorous than the BC1 parent. obtained although pollen tube growth was Further backcrosses with potato have not observed. Additional fusion experiments of S. resulted in BC3 progeny. americanum with potato cultivars Kanjer, In the potato (+) tomato backcrossing Elkana and Desiree resulted in new somatic programme BC2 progeny from different BC1 hybrids. The S. americanum (+) Kanjer hybrids genotypes (Garriga-Caldere et al., 1997) could showed good in vitrovigou r butwer e notabl e to easily be obtained. Ovule culture was still found survive in the greenhouse. Among the S. to be necessary, but mainly to improve the low americanum (+) Elkana and S. americanum (+) germination rate ofth e ovules. Desiree hybrids some vigorously growing and

44 Chapter 3 flowering genotypes were selected, but again Resistance to P. infestans pollinations did not yield berries. The fusion experiments should be repeated with other S. The somatic hybrid S. nigrum (+) potato, the americanum genotypes from different first backcross and eight second backcross accessions in order to select for combinations genotypes to potato that resulted from the with a higher cross compatibility. crossing experiments, were tested for their Another possibility might be to change the resistance to P. infestans. The somatic hybrid genomic ratio of the BC genotypes that have and the BC1 showed an equally high level of resulted from the S. nigrum (+) potato resistance as the initial S. nigrum fusion parent backcross experiments by somatic with a strong hypersensitive reaction. Colon ef hybridisation. Fusion experiments between the al. (1993) also observed this hypersensitive BC1 and diploid potato genotypes are expected reaction in S. nigrum and the sexual hybrid of to result in hybrids with three genomes of S. S. nigrumx diploid potato. nigrum and five of potato. The results of these Among the eight BC2 genotypes two kind of experiments will be described showed the formation of lesions and were elsewhere. concluded to be susceptible. The results of the Another option might be to use the P. infestans tests might indicate that one or microprotoplast technique (Ramulu er a/., more chromosomes were missing in the 1995), but the possibility to use this technique susceptible BC2 genotypes. It might also be depends on the genetics of the trait that has to possible that not all genomes of S. nigrum be transferred. Our aim was to transfer the contribute to the resistance since S. nigrum is resistance to P. infestans. So far nothing is expected to be an allohexaploid (Edmonds, known about the genetics of this resistance. 1979). Should it turn out to be a polygenic trait with loci on different chromosomes, the microprotoplast technique might not beth e optimal approach.

45 4 Alteration ofth e genomic composition of Solanum nigrum (+) potato backcross derivatives by somatic hybridisation: selection of fusion hybrids by DNA measurements and GISH

KarinHorsman 1,Tatyan a Gavrilenko2,Marja n Bergervoet1, Dirk-Jan Huigen1,Andr o TjinWon gJoe 1an d EvertJacobsen 1.

1 The Graduate Schoolo fExperimenta l Plant Sciences, Wageningen University, Laboratory ofPlan t Breeding, P.O. Box 386,6700 AJ Wageningen, The Netherlands. 2N.I .Vavilo v Institute ofPlan t Industry, B.Morskay a 42,190000 St.-Petersburg, Russia.

Thischapter was acceptedfor publication in Plant Breeding with modifications

ABSTRACT Fusion experiments were performed with a first (BC1-6738) anda second (BC2-9017) generation backcross hybrid which resulted from a cross between 6x S. nigrum (+)2 x potato somatic hybrids and potato. Because no progeny wasobtaine d from the BC2genotypes , alternative approaches were sought to overcome this sexual crossing barrier. Five potato genotypes, ofwhic h one contained the hygromycin resistance gene,were used inth e fusion experiments. All vigorous regenerants were used forth e estimation of nuclear DNA content through flow cytometry. Plants with a DNA content higher than that of the fusion parent BC1-6738 or BC2-9017 were considered potential somatic hybrids. Forty-nine potential somatic hybrids resulted from fusion experiments with BC1-6738, from which 20gre w vigorously inth egreenhous e andflowered , after pollination with several 4x potato cultivars eight genotypes produced seeded berries andfiv e genotypes gave seedless berries. In addition, eleven ofthes e thirteen somatic hybrids were selected forGISH-analysi st odetermin e their genomic composition. Nine of them hadexactl y or approximately the expected number of 36S . nigrum and 60 potato chromosomes. Inon e genotype only 22 instead of3 6 S. nigrum chromosomes were found andon epotat o chromosome waspossibl y missing aswell . Only five potential somatic hybrids were detected among the 79flow-cytometricall y analysed regenerants from BC2-9017 (+)2 x potato fusion experiments. Two ofthes e hybrids were rather vigorous and didflower , but pollinations with potato have not as yet, setan y berries.

47 Chapter4

INTRODUCTION fertility improved rapidly. Somatic hybrids of S. n/g/um-related species [S. americanum (2x), S. Several research groups have attempted to use chenopodioides (2x) and S. villosum (4/)] and Solanum nigrum as a donor species in potato potato with a more favourable genomic breeding, for instance for resistance to composition were produced as well (Horsman Phytophthora infestans (Colon ef a/., 1993). era/., 1997).Mos tfusio n combinations resulted Binding et al. (1982) produced somatic hybrids in plants that were not sufficiently vigorous in of S. nigrum and potato but did not report the greenhouse. Some vigorous plants of the successful backcross experiments. Eijlander fusion of 2x S. americanum (+) 4x potato were and Stiekema (1994) produced a few sexual selected but pollinations on these hybrids did hybrids but so far further crossing experiments not result in berries containing seed. haveno tbee ndescribed . In this contribution the alteration of the Successful backcross programmes of genomic composition of a first (BC1-6738) and interspecific somatic hybrids of several a second (BC2-9017) generation backcross combinations of Solanaceae species have hybridtha t have resultedfro m the S.nigrum (+ ) shown the influence of the genomic potato backcross programme with potato is composition of the fusion hybrid on the results described. The BC1 genotype is a vigorous of backcross experiments.Somati c hybrids with plant with three genomes each of S. nigrum an excess of genomes of the species that was and potato. In comparison to the BC1, the used as the backcross parent frequently gave selected BC2genotyp e was not as vigorous but better results in backcross experiments. For was still one of the best-performing genotypes example, in the potato (+) tomato backcross amongth e BC2's.Tw o genomes of potato were programme offspring could be obtained from added to these backcross hybrids by somatic crosses between hexaploid somatic hybrids hybridisation. The successful approach for the with four potato genomes and two tomato selection of the fusion hybrids is discussed. genomes andtetraploi d potato (Jacobsen etal., Putative somatic hybrids were selected after 1994; Garriga-Caldere et al., 1997). First measurements of the DNA content of the generation backcross hybrids were not regenerants. The hybrid nature of the selected obtained easily but the production of BC2 and fusion hybrids was confirmed by chromosome BC3 genotypes was less laborious. On the counting after genomic in situ hybridisation. other hand, backcrosses with tetraploid somatic Finally, the entire selection pathway of somatic hybrids possessing two potato and two tomato hybrids that grew vigorously in the greenhouse genomes were not successful. Somatic hybrids andflowere d isdescribed . of S. brevidens (+) potato showed similar results (Ehlenfeldt and Helgeson, 1987;

Jacobsen et al., 1993; Rokka ef al., 1994). The MATERIALS AND METHODS hexaploid genotypes with four genomes of potato were more fertile than the tetraploid Plantmaterials genotypes. After pollination with tetraploid potato, BC1 plants were obtained in low The two backcross genotypes thatwer e used in frequencies but in each following generation thefusio nexperiments ,a BC1an da BC2 plant,

48 Chapteri

were both hybrid derivatives of Solarium vitro. Plants were grown on basal MS medium nigrum and S. tuberosum (Horsman ef a/., with 30g/ lsucrose. 1999). The BC1-6738 was the result of successful backcrossing of a somatic hybrid of 6x S. nigrum and a 2x potato genotype, AM10, Somatichybridisation with a tetraploid potato genotype, AM66-42. Over 4000 pollinations and ovule rescue were The protoplast fusion procedure and media needed to obtain two hexaploid BC1 genotypes used were according to Jacobsen ef at. (1992). (Horsman ef a/., 1999) of which one genotype, Solid-callus growth medium (MS11), shoot BC1-6738, was vigorous, flowered profusely induction medium (MS12) and shoot elongation and could in its turn be crossed with tetraploid medium (MS13) were according to Mattheij ef potato. Over 5000 pollinations with several a/. (1992). When resistance genes in parental potato cultivars resulted in 12 BC2 genotypes. clones allowed the use of selective agents, The BC2 genotype BC2-9017 resulted from a kanamycin was used at concentrations of 50 cross of the BC1-6738 and the cv. Katahdin. mg/l in TMD, 75 mg/l in MS11, 100 mg/l in None ofth e BC2 genotypes was as vigorous as MS12, MS13, MS20 and MS30. Hygromycin the BC1 parent and most of the BC2's suffered was used at concentrations of 10 mg/l in TMD from severe growth abnormalities. Despite and 20 mg/l in MS11,MS12 , MS13, MS20 and many attempts, no BC3 progeny was obtained. MS30. The BC1-6738 contained the kanamycin resistance gene from the potato fusion parent of the initial hybrid, AM10. The kanamycin Selection ofpotential somatichybrids resistance gene appeared to be absent in BC2-9017. Regenerants from experiments in which one or Five 2x potato genotypes were used as both fusion parents contained a cell-selectable fusion parents in our experiments. PVU1 was a marker were put on medium containing hygromycin-resistant genotype which was kanamycin and/or hygromycin during callus obtained after Agrobacterium transformation growth, regeneration and plant growth. (strain LBA 4404 with plasmid PVU1011) of the BC1-6738 still contained the kanamycin amylose-free (amf) mutant 87.1031/29 resistance gene, which allowed the use of (Jacobsen ef a/., 1989). RH90-052-1 was a kanamycin for selection. One fusion diploid BC1 clone of a S. tuberosum x S. combination in which BC1-6738 was involved, microdontum hybrid, provided by Dr. Ronald BC1-6738 (+) 052-1, was carried out without Hutten of the Laboratory of Plant Breeding of kanamycin selection. Double selection with Wageningen University. The remaining three both kanamycin and hygromycin was applied in genotypes 6486-4, 6486-19 and 6487-9 all the fusion experiments of BC1-6738 (+) PVU1 resulted from crosses between 1024-2, an because the potato fusion parent PVU1 was Amfamf genotype, and dihaploids of the cv. hygromycin resistant. 'Gineke'. They were selected because of their All regenerants that were able to grow in favourable cell culture traits. Somatic hybrids vitro,wer e analysed flow cytometrically for their and backcross genotypes were maintained in DNA content. Regenerants with a DNA

49 •;= •.= := "> CO §> c co S CM t- 00 CO T- O 1- o o o >. T- CM I ° CE= a. co .c oa o •s E .J2 .ii CO a -2 EE • .2> o o >» , 2 73 > o. co .cz -

.22 o Ol CM -"St" CO CO Oi CO "^3 CM CO _>. TO o» E .0 S .O) o O >. r- to E > Q. CO SZ -S 5 o <. CO

CO CM m CO CO CO f T- in -g E. CO •* 9a . co

CD E CO O O O CO CD O CO CO 16 cu i2 CM CM •K D5 CZ P, P

co •<- -*• 00 T- CM LO ^ CD CO CO CM SP. E ^ CMIDS CC m fern £ £ 5

_co • i~ = O = "O CD t cooii- 0 l~- in •* 05 CO •* CD |v •* O r~- i^- "§-.§ g.| CO a> .w "s; cu E 5_ oS 5 CD O r C * I? C CO £

CO T- N CO N o co r»- o> •**• co CD CM •<- 1- CM CO

o o 00000 0 0 0 O =- £ inomom in 0 0 CD co E CDCOSOi- •* OO

2 OJ O O f CD .C>Z, t3 «2 CO CO CO ca" 2 -CD •£ cz cz cz cz cu co CO 3•>- 1 c-o^ co J*. jc^ o CO E • .CZ cu E CZ cz 0 0 >> cp a> > E V CD CO E 2 CM CO 00 CO 5 co 5 <2 co in •* •*r •>3- > -<3- > 0 co co co a. CD 0- <-» O + + + + + + + *O -*—f• 00 CO 00 CO CO tv t~- CO CO CO CO CO T 1•S «In N- h- t~- r^ [V CD ^O— <" ik CU co CO co co co Op cp E 1 1 • 1 S 2 o *— ^— • S3 CM CM SS OOOOO 0 O O 0 "~ !!. co ca co CD co 1— CO ca r— co 2

m Chapter 4 content higher than the BC1 or BC2 fusion After hybridisation the chromosomes were parent, were considered as potential somatic counterstained with a mixture of DAPI (4'6- hybrids. Flow cytometric analysis was carried diaminidine-2-phenylindole) (2 M9/ml) and out according to De Laat era/. (1987). propidium iodide (1 mg/ml) for 10 min and afterwards mounted in one drop of antifade solution. Crossingexperiments

All potential somatic hybrids were grown in a RESULTS greenhouse from February to October following the safety regulations stipulated by the Dutch Fusion experiments between BC1-6738 and 2x Government for growing genetically modified potato genotypes. organisms. Pollinations were performed with the tetraploid potato cultivars 'Bimonda', The number of calli that resulted from the five 'Desiree', 'Katahdin' and 'Mansour' and the fusion combinations of BC1-6738 (+) 2x potato tetraploid breeding clone AM66-42. varied between 600 and 1150 (Table 1, A). At Berries of the potential somatic hybrids least 24%o f these calli regenerated into plants. were harvested between three and five weeks Between 65 and 88% of the plants of the five after pollination for ovule rescue. Ovule culture different fusion combinations were not analysed was performed using HLH-mediumaccordin g to for hybrid nature because of poor in vitro Neal and Topoleski (1983). Ovules were growth. The remaining 307 regenerants were transferred to fresh medium every fortnight. analysed by DNA content measurement with a flow cytometer. All genotypes with a higher DNA content than the BC1-6738 fusion parent Genomici nsit u hybridisation (GISH) were considered to be potential fusion hybrids. Most of the analysed regenerants Mitotic chromosome spreads of root tips on appeared to be the BC1 fusion parent (Table 1, grease-free slides were made as described by A), as was expected since the BC1 genotype Pijnacker and Ferwerda (1984). Root tips were contained the kanamycin selection marker. The collected from in vitro grown plants, ten days fusion combination that was carried out without after propagation in vitro. GISH was performed kanamycin selection, also yielded 23 potato according to Schwarzacher and Heslop- regenerants. Despite the use of kanamycin and Harrison (1994). Modifications were mainly hygromycin a few potato regenerants were also described by Jacobsen et al. (1995). Probe detected among the regenerants of the DNA was sonicated into fragments of 1-5 kb. combination BC1-6738 (+) PVU1. Forty-nine Sonicated total genomic DNA was directly regenerants were concluded to be potential labelled with fluorescein-11-dUTP following the somatic hybrids because they had a higher nick translation protocol. DNA used for blocking relative DNA content than the BC1 fusion was autoclaved for 5 minutes into fragments of parent. In four fusion combinations the approximately 100 bp. For each preparation 50 percentage of somatic hybrids among the ng probe and 3-5 mg blocking DNA was used. regenerants analysed was rather low, between

51 Chapter4

4 and 7%. As expected 90% from the greenhouse and flowered aswell . At leaston e regenerants of the combination BC1-673 8 (+) flowering genotype per fusion combination PVU1 appeared tob e somatic hybrids, because could be selected but the majority of the both fusion parents contained a selection selected genotypes, 15 hybrids, had resulted marker. from the combination BC1-6738 (+) PVU1. All somatic hybrids were propagated in Since there were three fusion hybrids of this vitro and transferred toth e greenhouse. Teno f combination that were moderately vigorous and the 49somati c hybrids grew poorly in vitro and also able to flower, the total number of were also not able to grow well in vivo. Among flowering genotypes from this combinationwa s the remaining 39 somatic hybrids with agoo din 18 (Table 1, A). vitro vigour, 20genotype s were vigorous in the

Table 2. Results ofcrossin g experiments ofth e somatic hybrids BC1-6738 (+)2x potat o with tetraploid potato genotypes.

Genotype Fusion combination Relative Numbero f Number Number DNA- pollinations ofberrie s ofculture d content (berryset ) ovules

F101-2 BC1-6738 (+) 6486-4 1.18 44 6 (14%) 0 F102-1 BC1-6738(+) 6486-4 1.36 72 11 (15%) 1 F103-1 BC1-6738(+) 052-1 1.30 4 0 (0%) - F104-2 BC1-6738(+) 6487-9 1.49 16 5 (31%) 4

1 F108-1 BC1-6738(+)PVU1 1.12 11 0 (0%) - F108-2 " 1.24 82 38 (46%) 14 F108-3 " 1.25 10 0 (0%) - F108-4 " 1.22 13 0 (0%) - u F108-5 1.74 2 0 (0%) - F108-6 " 1.29 18 2 (11%) 0 F108-7 " 1.29 81 1 (1%) 1

F109-1 BC1-6738(+)PVU1i 1.30 45 1 (2%) 0 F109-2 ° 1.26 42 17 (40%) 0 F109-3 * 1.28 90 3 (3%) 1 F109-4 * 1.64 7 0 (0%) - F109-6 " 1.28 33 2 (6%) 2 F109-7 " 1.14 33 1 (3%) 0 F109-12 " 1.04 48 11 (23%) 26 F109-15 " 1.19 9 0 (0%) - F109-16 1.27 81 9 (11%) 5

1Serie s F108an dF10 9resulte dfro mdifferen t experimentswit hth esam ecombinatio no fgenotype s

52 Chapter 4

None of the somatic hybrids was as compared to the BC2fusio n parent,thes e three vigorous as the BC1 fusion parent itself which somatic hybrids showed improved vigour. The could reach a height of two to three meters in BC2 could grow to a height of about 1 m and the greenhouse. A majority of the twenty showed early senescence of leaves, whereas selected somatic hybrids was able to grow to a the new somatic hybrids could reach a height of height of approximately 1.5 m. Some of the approximately 1.5 m without early senescence hybrids showed morphological abnormalities of of their leaves. Only one of the selected leaves and flowers or dropped their flowers somatic hybrids, which had resulted from the beforeanthesis . combination BC2-9017 (+) 6487-9, was able to produce flowers. Eighteen pollinations were made but no berries were obtained. The Fusion experimentsbetween BC2-9017 and 2x remaining two genotypes dropped their flowers potatogenotypes before anthesis, liketh e BC2fusio n parent.

Fusion experiments with BC2-9017 were less successful. No calli could be harvested from Crossing experiments with somatic hybrids of two of the five fusion combinations, namely BC1-6738(+)2x potato. BC2-9017 (+) RH90-052-1 and BC2-9017 (+) 6486-19. The fusion experiment with potato Twenty somatic hybrids that resulted from the genotype 64864 resulted in 150 calli that did fusion experiments between BC1-6738 (+) 2x not regenerate into shoots (data not shown). potato were used in backcross experiments Only from the combinations BC2-9017 (+) (Table 2). From each genotype two plants were 6487-9 and BC2-9017 (+) PVU1 was a put in the crossing greenhouse. Tetraploid reasonable number of calli obtained, potato genotypes such as 'Katahdin', 'Mansour' respectively 800 and 600 (Table 1,B) . The first and 'Desiree' were used as male parents. combination, BC2-9017 (+) 6487-9, resulted in Because some genotypes such as F103-1, 138 regenerants from which 75 plants could be F108-5 and F109-6 dropped most of their analysed with the flow cytometer. Since no flower buds before they opened, only a few selection marker was used in this combination, flowers could be pollinated and these did not regenerants of both fusion parents were result in the formation of berries. On thirteen expected as well. Four somatic hybrids were somatic hybrids a total of fifteen pollinations or identified, whereas the remaining regenerants more was made. Genotypes F108-7, F109-1, were BC2 plants. In the second combination, F109-3, F109-6 and F109-7 had a very low BC2 (+) PVU1,th e potato fusion parent PVU1 berry set, between 1 and 6%. Some ovules contained the hygromycin selection marker. could be rescued from the berries of three of Only nine regenerants were obtained among these five genotypes. Genotypes F101-2, which one somatic hybrid was identified. Three F102-1, F108-6, and F109-16 had a berry set plantswer e potatoregenerants . between 11 and 15%. From the nine berries of All five somatic hybrids showed good in F109-16 five ovules could be cultured,wherea s vitro vigour (Table 1, B). In the greenhouse only one ovule from the eleven berries of three of these genotypes grew vigorously and, F102-1 and not a single ovule from the six

53 Chapter4

Table3 .Result so fth eGISH-analysi so fS .nigrum (+ )potat ohybrids .

Genotype Generation Relative Numbero f Numbero f DNA- S. nigrum S. tuberosum content chromo­ chromo­ somes somes

AM10 initial potato fusion parent 0.28 - 24 SN14-0 initial S.nigrum fusio n parent 1.00 72 -

F21-26 SN14-0(+)AM10 1.27 72 24

BC1-6738 F21-26x4/potato (AM6642) 1.00 36 36 6486-4, PVU1 2x potato fusion parents 0.27 - 24

F101-2 BC1-6738(+) 6486-41 1.18 4245 41/42 F102-1 BC1-6738(+) 648641 1.36 36 60 F104-2 BC1-6738 (+) 6487-9 1.49 36-42 60 F108-2 BC1-6738(+)PVU12 1.24 36 60 F108-6 ii 1.29 36 60 F108-7 II 1.29 36 58/59 F109-2 BC1-6738(+)PVU12 1.26 36 60 F109-3 II 1.28 36 60 F109-6 II 1.28 36 58-60 II F109-12 1.04 22 58-60 F109-16 II 1.27 36 60

1Serie sF10 1 and F102resulte dfro mdifferen texperiment swit h thesam e combinationo fgenotype s 2Serie sF10 8an d F109resulte dfro mdifferen texperiment swit hth esam ecombinatio no fgenotype s

Figure 1. Genomicconstitutio no fth einitia l somatic hybridF21-26 ,BC1-673 8an dthre epotentia lsomati chybrid s of BC1-6738 (+) 2x potato. In figure 1A, F21-26, potato chromosomes fluoresce yellow whereas S.nigrum chromosomes fluoresce red. In the pictures 1B to IE the S. nigrum chromosomes are yellow and the potato chromosomes are red. The number of S. nigrum chromosomes (ngf) and the number of S.tuberosum chromosomes (tbr) aregive n inparentheses .A . Initialsomati c hybrid F21-26(7 2ngr+ 24 tbr).B . BC1-6738(3 6 ngr+ 36tbr). C.F109-1 2(2 2ngr+ 58-60tbr). D .F101- 2(42-4 5ngr + 41/4 2tbr). E .F102- 1(3 6ngr+ 60 tbr).

54 Chapter 4

55 Chapter 4 berries of F101-2 and the two berries of F108-6 expected to contain 36 chromosomes of both was obtained. species. Counting of the differentially-stained The four remaining somatic hybrids, chromosomes after GISH confirmed these F104-2, F108-2, F109-2 and F109-12, had a chromosome numbers (Table 3, Figure 1B). berry set between 23 and 46%. Genotype Based on the results of the crossing F109-2 had a relatively high berry set of 40% experiments, eleven hybrid genotypes were but all berries were seedless. Sixteen selected to be analysed with GISH. The pollinations on F104-2 did result in five berries selection consisted of all eight genotypes from that contained four ovules. F108-2 produced 38 which at least one ovule could be rescued and berries (46%) from which 14 ovules could be three genotypes that had produced more than rescued. F109-12 had a berry set of 23% but one seedless berry (Table 2). Theoretically the the number of ovules that could be rescued somatic hybrids were expected to contain 96 from the eleven berries was high: 26 in total. All chromosomes: 36 of S. nigrum and 60 of S. genotypes on which at least 15 pollinations tuberosum. Six of the selected genotypes had were possible and genotypes that produced this expected genomic constitution (Table 3, berries from which ovules could be rescued,wil l Figure 1E) and five genotypes deviated from be used infutur e crossing experiments. this expectation. F108-7 and F109-6wer e most likely missing one or two potato chromosome but both contained 36 chromosomes of S. GISH-analysis nigrum. F104-2 had 60 chromosomes of potato origin but a higher number than 36 of S. nigrum The hexaploid S. nigrum genotype SN14-0 and chromosomes was identified. Between one and the diploid potato genotype AM10, the initial eight additional chromosomes or chromosome fusion parents, were used in a control fragments of S. nigrum per cell were found in experiment. Mitotic chromosome spreads from this genotype. F101-2 had a higher number of root tips were hybridised with total genomic S. nigrum chromosomes than expected but the DNA of one of the species as a probe and DNA number of potato chromosomes was only of the other species as blocking DNA. With slightly higher (41-42) than in the BC1-fusion GISH it was possible to differentiate clearly parent (Table 3, figure 1D). Therefore it is between S. nigrum and potato chromosomes. doubtful that this genotype is a somatic hybrid. The original somatic hybrid, F21-26 (Horsman The fifth genotype with a deviating ef a/., 1997), that had resulted from the somatic chromosome number was F109-12,th e somatic fusion of SN14-0 and AM1 0 was shown to hybrid with the highest number of ovules per contain the expected number of chromosomes berry. This genotype had only 22 S. nigrum of both species: 24 of potato and 72 of S. chromosomes left but did contain 58 to 60 nigrum, respectively (Table 3, Figure 1A). The chromosomes of potato (Table 3, Figure 1C). BC1 genotype BC1-6738, that was used as a The relative DNA content of somatic hybrid fusion parent in the experiments described in F109-12 already indicated that several this report, had resulted from a cross of the chromosomes might be missing since it was somatic hybrid F21-26 and the tetraploid potato lower than the sum of the DNA-contents of the genotype AM66-42. This BC1 genotype was fusion parents. Most somatic hybrids with a

56 Chapter4 chromosome number that was nearly or exactly cytometry. DNA content measurements have the expected chromosome number had a also been used fruitfully by several authors relative DNA content which equalled the sum of working with Solanaceae (Daunay ef a/., 1993; the parental DNA contents. DNA Pijnacker etal, 1989), although in S. brevidens measurements of genotypes F102-1 and (+) S. tuberosum somatic hybrids a low F104-2 were somewhat higher than expected. correlation between nuclear DNA content and GISH-analysis proved ten of the eleven number of chromosomes was found (Valkonen potential somatic hybrids that were selected ef a/., 1994). In our case DNA, content after DNA content measurement and backcross measurements could be used successfully for experiments, to be real somatic hybrids. In the detection of somatic hybrids among these fusion experiments the combination of regenerants, because of the differences in DNA-measurements and GISH has been ploidy level of the fusion parents. The shown to be a successful approach for the BC1-6738 and the BC2-9017 were 6x and selection offusio n hybrids. approximately 5x respectively. This made it possible to distinguish between BC1-6738 or BC2-9017 regenerants, potato self-fusions and

DISCUSSION the desired fusion hybrids. The additional application of GISH indicated the reliability of The idea of using somatic fusion for adjustment flow cytometric observations. Ten of the eleven of the genomic composition in order to solve selected vigorous potential somatic hybrids backcross problems is new. In this contribution appeared to be realfusio n products. the first steps have been made to investigate GISH-analysis of the most important basic the different factors of this approach, the most plant material showed the presence of the important one being the recognition of the expected number of chromosomes in the initial desired somatic hybrids. Detection markers like fusion hybrid F21-26: 24 of S. tuberosum and isozymes were not useful in our case since they 72 of S. nigrum. Observations of fusion hybrids were not able to discriminate between the of other combinations involving species from backcross hybrid and the potato fusion parent. different genera such as Solarium (+) A successful way to prescreen a large number Lycopersicon showed that in many instances of regenerants appeared to be flow cytometry. the expected number of chromosomes is not The combination of flow cytometry and GISH found in all hybrids. In 79% of the potato- analysis was shown to be an effective method tomato somatic hybrids analysed by Wolters for the selection of the fusion hybrids. and coworkers (1994) one or more In several publications, flow cytometry chromosomes were missing. Garriga-Caldere has been used for the analysis of DNA content ef al. (1997) analysed three potato-tomato of somatic hybrids. A high correlation between somatic hybrids with GISH and established the the 2C value and the number of chromosomes absence of one or two tomato chromosomes in was found in fusion products of several two of these hybrids. Besides this, somatic Brassicaceae (Fahleson ef a/., 1988). In the recombination was also observed in potato- same plant family Sundberg and Glimelius tomato somatic hybrids. (Wolters ef al., 1994; (1991) established ploidy levels using flow Garriga-Caldere etal., 1997). The phenomenon

57 Chapter4 of missing chromosomes in somatic hybrids frequencies as in the potato-tomato material was also observed in other plant families. earlier described. F109-12 was the most Allium ampeloprasum (Ax) (+) A. cepa (2x) interesting somatic hybrid from the crossability somatic hybrids were analysed with GISH by point of view. Inthi s hybrid a significant number Buitenveld et al. (1998). All fusion hybrids of S. nigrum chromosomes was missing. It has lacked 3 to 7 chromosomes of the tetraploid A. to be checked whether or not this hybrid is still ampeloprasum and one genotype had a lower resistant to P. infestans like the BC1 fusion number ofA. cepachromosome s as well. parent. The BC1 fusion parent BC1-6738, that During each successive step of the was used successfully in backcross selection scheme a substantial part of the experiments, also possessed the expected material had to be removed. Over 1100 number of potato and S. nigrum chromosomes: regenerants were obtained from the BC1 (+) 2x 36 of both species. Potato-tomato BC1 potato fusion combinations, but only 26% of the genotypes that had resulted from backcrosses regenerants appeared to be vigorous in vitro. with potato mostly did not contain the expected Among the 307 analysed plants 49 somatic number of 12 tomato chromosomes. GISH- hybrids were identified. Most of these hybrids analysis of six BC1 plants revealed the were vigorous in vitro but only 20 hybrids grew presence of 10 to 12 tomato chromosomes. vigorously in the greenhouse as well and were Identification of the individual tomato flowering. The selection of useful BC2-9017 (+) chromosomes with RFLP-markers and GISH 2x potato somatic hybrids showed a similar revealed the presence of some tomato pattern. Two vigorous and flowering genotypes chromosomes in disomic condition. Therefore were selected from 147 regenerants. Similar none of the six BC1 plants contained all twelve observations were made during the initial fusion tomato chromosomes (Garriga-Caldere ef al., experiments between S. nigrum related species 1997). In the S. n/grum-potato BC1 this and several potato genotypes (Horsman et al., phenomenon of disomy could not be checked 1997). Results described in this contribution because of the absence of S. nigrum- again showed the importance of producing a chromosome-specific markers. large quantity of starting material for the GISH-analysis of BC1-6738 (+) 2x potato selection of somatic hybrids that were useful for showed that most of the selected potential crossing experiments. The next step will be fusion hybrids contained the sum of the number more extensive backcrossing and embryo of parental chromosomes. Deviations were rescue experiments with the selected fusion found in both genomes but not in such high products.

58 Prospects for the introgression of chromosomes from non- tuberous Solanum nigrum into S. tuberosum: a qualitative analysis of the meiosis through GISH and AFLP-analysis of backcross derivatives

Karin Horsman, Evert Jacobsen and M.S. Ramanna

The Graduate School of Experimental Plant Sciences,Wageninge n University, Laboratory of Plant Breeding, P.O. Box 386,6700 AJ Wageningen,Th e Netherlands.

ABSTRACT

The chromosome constitution and meiotic behaviour of a somatic hybrid of Solanum nigrum (2n=6x= 72) and diploid potato (2n=2x=24) and its BC1 and BC2 progeny were studied. The application of Genomic in situ Hybridisation (GISH) made it possible to distinguish clearly between S. nigrum and potato chromosomes in mitotic and meiotic chromosome spreads. The somatic hybrid F21-26 was found to be an octaploid with six genomes of S. nigruman d two of potato.A cross between F21-26 and a 4x potato genotype produced BC1-6738, a hexaploid with 36 chromosomes each of both species. With GISH itwa s impossible to determine whether the 36 chromosomes of S. nigrumrepresente d three complete genomes. However, AFLP-data showed that none of the AFLP-specific markers that was amplified in the S. nigrum fusion parent and the somatic hybrid was missing in BC1-6738, which is an indication that no major chromosome elimination has taken place. A second backcross with 4x potato resulted in eleven BC2 genotypes with a near-pentaploid genomic constitution. Chromosome counts in nuclei of BC2 tetrad cells showed that transmission of alien S. nigrum chromosomes to BC3 progeny is likely since on average 9.1 S. nigrum chromosomes per BC2-microspore were detected. Meiotic analysis of metaphase I in BC1-6738 and in BC2-9019 indicated clearly that allosyndetic pairing occurs inthes e genotypes. Homoeologouspairin g in both bivalent andtrivalen t formation wasobserved .

59 Chapter5

INTRODUCTION exceptions of hybrids that were investigated cytologically are the interspecific F1 hybrids of The genus Solatium of the family of tuberous and non-tuberous Solanum species Solanaceae is a large one with more than 2000 (Ramanna and Hermsen, 1979; Hermsen efal., species. It includes tuber-bearing (tuberous) 1981) and the intergeneric hybrids of Solanum and non-tuberbearing (non-tuberous) forms. and Lycopersicon (Jacobsen ef al., 1994; With the aim to use the genetic variation that is Garriga-Caldere et al., 1997). From these present in this genus, extensive interspecific studies it was evident that a certain amount of hybridisation has been carried out in the past. homoeologous recombination occurs between Generally, it is easy to sexually hybridise and the chromosomes of tuberous and non- subsequently make backcrosses within the tuberous Solanum species (Hermsen ef al., group of tuberous Solarium species. 1981) whereas only a highly restricted amount Nevertheless, it is extremely difficult to cross of intergeneric recombination occurs in the tuberous and non-tuberous species. The first case of potato and tomato (Garriga-Caldere ef successful sexual hybrids of this kind were al., 1998). between S. pinnatisectum (tuberous) and With a view to introgress (a part of) species of the series Etuberosa, S. chromosomes of S. nigrum into potato, somatic etuberosum, S. brevidens and S. hybrids involving those species were produced fernandizianum (Hermsen and Taylor, 1979; and subsequently backcrossed to potato Ramanna and Hermsen, 1981). Later on (Horsman efal., 199 7 and 1999). BC1 and BC2 Watanabe et al. (1995) described the progeny was obtained and in order to utilise production of sexual hybrids between the same these genotypes in further crossing Etuberosa species and diploid S. tuberosum. experiments a cytological assessment was However, in the cases mentioned above all made applying the genomic in situ hybridisation species belong to the same section, namely (GISH) technique and a molecular analysis was Petota (Hawkes, 1978). Sexual hybrids carried out involving the AFLP (amplified between species of section Petota and species fragment length polymorphism) technique. of the S. nigrum complex were made by Special focus was on the possibilities of Eijlander and Stiekema (1994). S. nigrum and occurrence of recombination and the S. villosum were crossed with S. tuberosum transmission of alien chromosomes to the and S. demissum which resulted in only a progenies. In the light of these results the limited number of hybrid genotypes. prospects for introgression arediscussed . Besides sexual hybridisation somatic hybridisation has been performed using more distantly related species (Wolters, 1994). Some MATERIALS AND METHODS drawbacks of those experiments are that 1) few attempts have been made to produce extensive Plantmaterial backcrosses of the somatic hybrids; 2) hardly any cytological assessment has been All genotypes that were analysed were hybrids undertaken with somatic hybrids and their of S. nigrum and S. tuberosum. Somatic backcross derivatives, when available. A few hybridisation experiments between S. nigrum

60 Chapter 5

(2n=6x=72) genotype SN14-0 and diploid probe and S. tuberosum DNA as blocking or S. potato genotype AM1 0 gave rise to the tuberosum DNA as probe and S. nigrum DNA octaploid (2n=8x=96) somatic hybrid F21-26 as blocking following the procedure of (Horsman et al, 1997). The diploid S. Schwarzacher and Heslop-Harrison (1994), tuberosum genotype 87.1029/31, was an modifications described by Jacobsen ef al. amylose-free (amf) mutant described by (1995) and details in Horsman efal. (1999) . Per Jacobsen et al. (1989). AM10 was a preparation 50 ng probe and 3-5 mg blocking transformant of clone 87.1029/31, containing DNA was used. After hybridisation the the GUS-gene and the kanamycin resistance chromosomes were counterstained with a gene. mixture of DAPI (4'6-diaminidine-2- The BC1 genotype DJ93-6738, which will phenylindole) (2 ng/ml) and propidium iodide (1 be referred to as BC1-6738, resulted from a mg/ml) for 10 min and afterwards mounted in cross between F21-26 and the tetraploid one drop of antifade solution. (2n=4x=48) breeding clone AM66-42. Meiotic preparations were made of young BC1-6738 was again crossed with several anthers of flower buds that were collected tetraploid cultivars which resulted in eleven between ten and twelve o'clock in the morning. BC2 genotypes that were able to grow in vitro. One anther per flower bud was squashed in An overview of all genotypes and the parents of aceto-carmine and checked under a light- all BC2 genotypes can befoun d inTabl e 1. microscope in order to establish the meiotic Somatic hybrids and backcross genotypes stage. The remaining anthers of flowers with were maintained in vitro. Plants were grown on suitable meiotic stages were fixated in ethanol- basal MS mediumwit h 30 g/lsucrose . All plants acetic acid 3:1 for 30 minutes at room were grown in a greenhouse from February to temperature. Enzyme treatment of the anthers October. The somatic hybrids and the was carried out with 0.5% pectolyase Y23, backcross genotypes were grown according to 0.5% cytohelicase and 0.5% cellulase RS in 10 government safety rules for genetically modified mMcitrat e buffer pH 4.5 for two houres. organisms. Preparations were made following the protocol of Pijnacker and Ferwerda (1984). GISH with S. nigrum DNA or S. tuberosum Genomicin situ hybridisation(GISH) DNA as a probe was performed as described previously. For the analysis of the chromosome constitution mitotic chromosome spreads were made of root tips that were harvested from in AFLP-analysis vitro plants 7-10 days after propagation. Root tips were treated with 2 mM 8-hydroxyquinoline The AFLP-analysis was performed according to solution for 2.5 hours at 18°C after which they Van Eck ef al. (1995) which was based on the were fixed in ethanol-acetic acid 3:1. protocol as described by Zabeau and Vos Preparations were made on grease-free slides (1993) and Vos ef al. (1995). Three primer according to Pijnacker and Ferwerda (1984). combinations, involving primers with three GISH was carried out with S. nigrum DNA as selective bases,were used:

61 Chapter 5

1) E35/M48: Primer+ACA/Primer+CAC; to be a hexaploid. Thirty-six chromosomes of 2) E35/M50: Primer+ACA/Primer+CAT; each of both S. nigrum and potato were 3) E35/M58: Primer+ACA/Primer+CGT. counted in this genotype. The most logical explanation was that a normal gamete from the 8x somatic hybrid consisting of three genomes RESULTS of S. nigrum and one genome of potato was involved in the fertilisation process. The fusion An important pre-requisite for cytological of such a gamete with two genomes of the 4x analysis of the somatic hybrid F21-26 and its S. tuberosum backcross parent obviously backcross derivatives was that a distinction resulted in a hexaploid BC1. However, since could be made between the chromosomes of S. the individual chromosomes could not be nigrum and S. tuberosum. Genomic in situ identified in metaphase condition it was not hybridisation (GISH) made it possible to possible to determine whether these 72 distinguish the genomes from each other chromosomes represent six complete unambiguously in both mitotic and meiotic genomes. Theoretically it is possible that some preparations (Fig.1) . chromosomes are overrepresented whereas others, in that case, have to be under- represented. Somatic chromosome constitution of the fusion hybrid and BC1-genotype Pairing and transmission of chromosomes in Since the initial somatic hybrid was the result of BC1-6738 a fusion between diploid S. tuberosum (2n=2x=24) and hexaploid S. nigrum Being distant interspecific hybrids with high (2n=6x=72) the chromosome number of the chromosome numbers it was not possible to somatic hybrid was expected to be an octaploid obtain the chromosomes well spread out in (2/?=8x=96). As indicated in Table 1 this was cytological preparations of pollen mother cells. shown to be the case. A cross between the Therefore it was not possible to make a somatic hybrid and a tetraploid potato breeding quantitative analysis of the extent of clone gave rise to BC1-6738 which was shown chromosome pairing during metaphase I

Figure 1. Meiotic behaviour in a first generation (BC1-6738, A-C) and a second generation (BC2-9019, D-F) backcross derivative of S. nigrum(+ ) potato somatic hybrids with tetraploid potato as revealed by GISH. In all pictures except C the S. nigrumchromosome s fluoresce yellow and the potato chromosomes red. In picture C coloration is reversed. "N"= S. nigrum, "T"= S. tuberosum, the colour of the letter is similar to the fluorescence signal of the chromosome. White arrows point at allosyndatic paired chromosomes, light blue arrows at autosyndatic pairing. White arrowheads indicate univalents. A-C: Metaphase I stage of BC1-6738 showing N- univalents, T-univalents, autosyndatic pairing of both N- and T-chromosomes, homoeologous bivalents and homoeologous trivalents in all four possible configurations: NNT, NTN,TT N and TNT. D: Metaphase Istag e of BC2-9019 with a TTN-trivalent, several homologous bivalents and univalents. E: Late methaphase I, early anaphase Istag e of BC2-9019wit h laggingchromosome s of both potato and S. nigrum. F:Telophas e IIstag eo f BC2-9019.

62 Chapter5

63 Chapter 5 stages. However, it was possible to analyse in individual nuclei was estimated. This number cells for meiotic chromosome configuration at ranged from 13 to 19, the average being 16.7 metaphase I and to count chromosomes during per nucleus (Fig. 2) which is nearly half of the anaphase I and telophase II in both 6C1 and somatic number of S. nigrum chromosomes in BC2 plants (Fig. 1). Meiotic preparations of the this genotype (Fig. 2). somatic hybrid F21-26 could not be made since Because of the abnormal meiosis, BC1 no anthers with pollen mother cells undergoing plants showed a low fertility with a pollen meiotic division could be identified, despite the stainability of 10-15%. BC1 pollen was not able analysiso f a largequantit y offlora l buds. to penetrate pistils of4 x potatogenotypes . Meiotic analysis ofth e BC1-6738 genotype confirmed the genomic constitution established Chromosome constitution, pairing and from the somatic cell analysis, i.e., three transmissionin BC2 genomes each of both S. nigrum and potato were present. Because of the presence of this The chromosome constitution of eleven BC2 odd number of genomes (3x + 3x) in the plants was estimated from counting their BC1-6738, the chromosome pairing was number in root tip cells. The total number of complicated (Fig. 1A, B, C). Trivalents, chromosomes ranged from 47-49 to 59-61 bivalents and univalents were observed at (Table 1) with a number of S. nigrum metaphase I stages.Association s of more than chromosomes that variedfro m 14t o 20. Except three chromosomes could not be confirmed for KH94-9026 the chromosome numbers ofth e with certainty but trivalents were observed BC2 plants were concurrent with the pentaploid unequivocally. Because of the application of number of chromosomes that was expected GISH two types of trivalents could be from a6x-4 xcross . distinguished: those that involved only Meiosis was abnormal in BC2 plants homologous chromosomes (autosyndetic because of the pentaploid constitution (Fig. 1D, pairing) and a second type that involved E). As in the BC1 genotype multivalents, homoeologous chromosomes (allosyndetic bivalents and univalents were observed at pairing) (Fig. 1A, B, C). The latter types metaphase I stages of BC2-9019 (Fig. 1D). consisted of either one or two S. nigrum Both bivalents and trivalents displayed chromosomes with respectively two or one S. autosyndetic as well as allosyndetic tuberosum chromosomes. Within the group of chromosome associations as in the BC1. bivalents two types were distinguished as well: Despite the unbalanced distribution of homologous and homoeologous pairs (Figure chromosomes during anaphase I stages (Fig. 1A, B). A considerable amount of univalents of 1E), the second division seemed to occur both parentswa sals oobserve d (Fig.1A) . normally giving rise to spore tetrads of In view of the odd number of genomes of microspores. A notable feature was that the both parents in the BC1 plant, meiotic pairing chromosomes were well spread out in the and chromosome distribution during the first nuclei of each of the pollen mother cells (Fig. meiotic division was abnormal as expected. By 1F) so that it was possible to estimate the analysing a limited number of telophase II number of S. nigrum chromosomes which rang­ stages, the number of S. nigrum chromosomes ed from 7 to 12 with an average of 9.1 (Fig. 3)

64 CD 2 8 CD ° CD CD •* (D O) CO O T- _ JZl E m m m m• coi cio & £ 2 s •* in in •* CD Ol 3 ^a- m m m .° •*= in in in I— c o

£ Z CO o CM CO n*^t co co -a- -

CO CD oE 2 e to fc 5 o -2.e n £ E £ g ZJ .c CM CD TT •>«- in in ©CD SS OO CD O z CO o h- CO

cu E •«- CM CNI CM CNI CM CNI CNI CNI CNI CNI CNI o o o o o o o o o o O CD CO CO CO CO CO CO CO CO o o o CO CO CO

o so C1O Q. ,_ -o e 1_ c CD CD CD CNI CD •a "a CZ -Q CD ZJ cz CZ CD ^ CD £ O CD CD -CD ,2 CD -CD 0 CD SZ CD 5 CO co CO CO CO '55 CZ CD -Q CD CO JS CO .o> X CD 3 S c CNI e -CD Z3 b "i_ •CD CD -CD < O 2 LL 1— CD LL. O x: 0 ^ X X X X X X X X X X X X OO OO OO CO 00 OO 00 00 OO CO CO E CO CO CO CO CO CO CO CO CO CO CO CD CD 1^- I-- 1^- l~- r^- r-- 1^- r>- I-- r~- r~- -Q CD CD CO CD CO CO .Og) CNI 1 co co co CD 1 1 cp 1 E c CNI O 0 O 0 O O O O O 0 0 CO LL. CO CO CO CO CO CO CO CO CO CO CO o c o

& CM in -51- in CD r~ O CD CD CD <-> •* Q CO l~- CO c-> -s- CNI CNI CO CO l-~ CO CO CO co O CO <->0 CO C) CO CD 1^- CD CD O} CD a> CD en CD CD CD CD CM •* co co CD •sj- CO •^- •f CD -

13 14 15 16 17 18 19 Number of alien S. nigrum chromosomes

Figure2 .Chromosom e countsi nth enucle io ftetra dcell so fBC1-6738 .Fro ma tota lo f2 7tetra dcell sth enumbe r ofS .nigrum chromosomeswa sdetermined .O naverag e 16.7S .nigrum chromosome swer efoun dpe rnucleus .

7 8 9 10 11 12 Number of alien S. nigrum chromosomes

Figure3 .Chromosom e counts inth e nucleio ftetra dcell so fBC2-9019 .Fro ma tota lo f3 8tetra dcell sth enumbe r of S. nigrum chromosomes was determined. On average 9.1 S. nigrum chromosomes were found per nucleus.

66 Chapter5

AFLP-analysis Table 2. S. nigrum specific AFLP-bands in the descendants of S. nigrum (+) potato somatic hybrids. For the identification of the degree of homozygosity in S. nigrumAFL P fragments ofa Genotype Numbero f Percentageo f selfing population with4 8 descendants ofth e S. ngr- ngr-specific nigrum fusion parent were analysed. The initial chromosomen AFLP-bands S. nigrum fusion parent was found to be highly present homozygous since only 4% of all amplified F21-26 72 100% fragments appeared to be polymorphic. After a comparison of the AFLP-patterns of S. nigrum, BC1-6738 36 100% the potato fusion parent and the potato backcross parents a total of 66 markers was BC2-9365 14 50% BC2-9304 15 65% selected from the remaining fragments for the BC2-9345 15 48% analysis of the fusion product, the BC1 and BC2-9026 16 52% BC2 derivatives. BC2-9320 16 68% All of the selected fragments were BC2-9017 17 58% BC2-9084 17 66% amplified in the somatic hybrid F21-26 as well BC2-9380 18 60% as in the BC1-6738 (Table 2). Since the BC2-9019 19 67% distribution of these AFLPs on the BC2-9379 20 71% chromosomes of the S. nigrum genomes is unknown, this can not lead to any final conclusion about the presence of the complete the BC2s andth e number of S. nigrum markers S. nigrum genome in these genotypes. could not be detected. Relatively large However, together with the observation of 72 S. differences in the number of S. nigrum specific nigrum chromosomes in F21-26 and 36 in bands were found between BC2 genotypes with BC1-6738,ther e was a strong indication that no an equal number of S. nigrum chromosomes. major chromosome elimination had occurred For example, a difference of 17% inth e number during somatic hybridisation and the first of S. nigrum specific fragments was shown backcross. between BC2-9304 and BC2-9345, both of In the AFLP patterns of the ten BC2 which contain 15 chromosomes of S. nigrum. genotypes that were analysed the percentage This could indicate a difference in the number of S. nigrum specific markers that was detected of single chromosomes and induced diplo- varied from 48 to 71, indicating that, as chromosomes betweenthos etw o genotypes. expected, chromosome loss occurred in this generation (Table 2). BC2-9365, which was possessing the smallest number of S. nigrum DISCUSSION chromosomes of all BC2's,showe d only 50%o f the S. nigrum bands whereas BC2-9379 with Several attempts have been undertaken to still 20 S. nigrum chromosomes had retained hybridise distantly related species within the 71%. However, a linear relationship between Solanaceae (Wijbrandi, 1989; Wolters, 1994). the number ofalie n S.nigrum chromosome s in

67 Chapter 5

However, there have been relatively few 1979; Hermsen and Ramanna, 1981). In these studies to investigate backcrosses and monitor biparental and trispecific triploid hybrids the backcross progeny extensively. The results trivalents were observed in low frequencies reported in this article are an example in which (respectively 0-3 and 1-4 per cell) which the backcross progeny of a fusion hybrid has indicated the occurrence of recombination been analysed.A s far as is known, it isth e only between the species involved. Taxonomically successful attempt to backcross a S. nigrum (+) these species are more closely related S. tuberosum hybrid (Horsman ef a/., 1999). A (Hawkes, 1978) than the species used in our similar study was carried out on potato (+) investigation, S. tuberosum and S. nigrum, tomato somatic hybrids and their backcross which belong to different sections of the progeny (Jacobsen ef a/., 1994, Garriga- Solanaceae. This means that the considerable Caldere ef a/., 1997). In that case the species amount of allosyndetic pairing observed in belonged to different genera, viz. Solarium and BC1-6738 and BC2-9019 indicates that Lycopersicon. Application of the GISH allosyndetic pairing is much more widespread. technique made it possible to distinguish It may also be noted that allosyndetic pairing in between potato and tomato chromosomes intergeneric hybrid derivatives, like BC1 and (Jacobsen ef a/., 1995). Fortunately, in our BC2 of potato (+) tomato somatic hybrids experiments it was also possible to discriminate (Garriga-Caldere ef al., 1999), is much more between the chromosomes of S. nigrum and restricted than in our hybrids. A quantitative potato with GISH,althoug h both species belong analysis might be very useful for the to the same genus,Solanum. establishment of phylogenetic trends. When Two important aspects emerged from our backcross derivatives with lower ploidy levels cytological analysis of BC1-6738 and will become available it might be possible to BC2-9019: 1) clear evidence was found for quantify the extent of pairing and show allosyndetic pairing, of importance for recombination. So far in none of the backcross homoeologous recombination; 2) the S. nigrum derivatives recombination could be shown with chromosomes are transmitted in reasonable GISH. This is probably because the numbers from the somatic hybrid F21-26 to euchromatic parts of chromosomes in BC1-6738 and from BC1-6738 to the BC2 solanaceous crops such as tomato (Ramanna progeny. Looking at the number of S. nigrum and Prakken, 1967) and potato (Ramanna and chromosomes in telophase II nuclei of the BC2 Wagenvoort, 1976) are highly condensed. plants, it seems to be likely that a relatively high Because of this, crossover segments that are frequency of the alien chromosomes will be generally distal, cannot be made visible easily transmitted to BC3 progeny as well. unless large segments of chromosomes are Regarding the pairing between involved in crossingover. chromosomes of a tuberous and a non- No preferential elimination of S. nigrum tuberous species the only cytologically chromosomes seemed to have occurred in the analysed Solanum hybrids were between S. somatic hybrid and the BC1-6738 since the pinnatisectum and the Etuberosa species S. latter genotype contained exactly 36 S. nigrum etuberosum, S. fernandizianum and the hybrid chromosomes, half the number of S. nigrum S. brevidens x S. etuberosum (Ramanna etal., chromosomes of the somatic hybrid. It is

68 Chapter 5

possible that these 36 chromosomes do not varied from 1.7 till 3.4 tomato chromosomes, represent three genomes exactly. One or more depending on the chromosomal constitution of chromosomes could be overrepresented the BC1 genotype. The BC1 that contained 11 whereas others, in that case, have to be single tomato chromosomes transmitted, on underrepresented. Besides, the AFLP data average, 1.7 chromosomes to its BC2 showed clearly that all S. nigrum specific descendants. Another BC1 mother plant with 8 markers that were present in the initial S. single and 2 disomic tomato chromosomes nigrum fusion parent could also be detected in (Garriga-Caldere ef a/., 1998) transferred on the BC1-6738, which indicates that all parts of average 3.4 alien chromosomes to the BC2 the chromosomes that were covered by the progeny. In our experiments we established a AFLPs are retained after the first backcross. transmission of 14 to 20 S. nigrum Since it is unknown at present whether the chromosomes from BC1 to the BC2 genotypes. markers used by us are dispersed over all The S. n/grum-complex is a highly chromosomes of the S. nigrum genome, no polymorphic taxon with world wide distribution conclusions can be drawn about the and the species are most successful as weeds. completeness of this genome in BC1-6738. It is attractive to transfer the chromosomes of However, AFLP-analysis of the first potato- these species into potato in order to widen the tomato BC1, which contained 50%o fth e twelve genetic basis for some of the desirable tomato chromosomes, showed that from the characters like resistances against pests and tomato-specific AFLP-markers also 50% could diseases. In this context it is worthwhile to still be detected in this genotype (unpublished stably integrate alien S. nigrum chromosomes data). into potato, even if the traditional method of In potato-tomato hybrids preferential introgression has to be practiced. Fortunately it elimination of tomato chromosomes was is possible to apply the GISH and AFLP already observed in the BC1 plants (Garriga- techniques to the S. n/grum-potato hybrids Caldere ef a/., 1997). None of the five BC1 which facilitates the identification of genotypes plants that were the result of a cross between a with desirable alien chromosomes among the potato-tomato somatic hybrid and 4x potato backcross derivatives and thus enhances the contained all twelve chromosomes of tomato. chance of success. The transmission from BC1 to BC2 plants

69 General discussion Chapter6

The useo f taxonomically related species in well in the greenhouse, whereas plants of other plant breeding as asourc e of genetic variation genotypic combinations all died in vivo is very common in plant breeding. Several (Garriga-Caldere ef a/., 1997). Poor methods are being used inorde r tomak e the performance of potato (+) tomato somatic genes of these species available, from which hybrids in vivo was also reported by sexual hybridisation followed by recurrent Schoenmakers ef al. (1993), but inthei r case backcrossing is most common. Inthi s thesis the also a low frequency of vigorous plants was application of one of the alternative approaches obtained. was described, namely somatic hybridisation Difficulties with rooting and culturing of followed by recurrent backcrossing. The somatic hybrids were also encountered in successive steps of our approach will be Brassicaceae, mainly within intertribal discussed in this chapter and some alternatives combinations. Brassica napus (+) Thlaspi will be compared. perfoliatum fusion products were difficult to transfer to the greenhouse (Fahleson et al, 1994b) but difficulties were even more PRODUCTION OFSOMATI C HYBRIDS pronounced in fusions between B. napus(+ ) Barbarea vulgaris, which never resulted in In order to produce large populations of starting plants that could grow under greenhouse material for backcross experiments, four conditions (Fahleson ef al., 1994a). The different species of the S. nigrum complex with difference with our somatic hybrids was, that different ploidy levels were somatically the B. napus (+) Barbarea vulgaris were also hybridised tothre e potato genotypes. Mosto f difficult to culture in vitro, since their in vitro the fusion combinations resulted in somatic growth was rather slow whereas oursomati c hybrids that were vigorous in vitrobu t appeared hybrids grew vigorously in vitro. Another to be inviable after transferring to the troublesome combination wasB. napus (+) greenhouse. Only from the three combinations Lesquerella fendlerifro m which only 5 of the 53 ngr(+) 1029/31, ngr(+)MW and symmetric somatic hybrids could be established ame2x(+ ) Desiree a substantial part ofth e in the greenhouse (Skarzinskaya et al., 1996). hybrids was also vigorous in the greenhouse. Partial elimination of DNA ofth e wild species Three other combinations resulted in a low using irradiation appeared to improve the in percentage of relatively well growing hybrids: vivo performance of B. napus (+)L fendleri chp (+) 1029/31, ames* (+) 1029/31 and fusion products since 32 % ofth e asymmetric ameex(+)A M10 K. Among the hybrids of the hybrids rooted and grew in vivo. In fusion remaining combinations, not a single plant was combinations of Solanaceae species, for able to grow inth e greenhouse. instance L. esculentum (+) L peruvianum, The striking contrast between in vitro and irradiation had a negative effect on vigour and in vivo vigour has not been described indetai l fertility of the hybrids (Wijbrandi efal., 1990 ) in literature, however, it was observed earlier in The problem of establishing somatic some combinations within the Solanaceae. hybrids that are able togro w vigorously inth e Several potato-tomato genotypic combinations greenhouse seems to occur mainly among resulted in somatic hybrids that performed very combinations of more distantly related species.

72 Chapter6

However, the factor of relatedness can not The main difference between the S. explain the differences that we observed nigrum (+) potato and the potato (+) tomato between the hybrids of the different fusion backcross experiments with potato, was the combinations since all of the species of the S. ability to produce of BC2 and BC3 progeny. ro'grum-complex are expected to have the same The S. nigrum (+) potato BC1 was again phylogenetic distance to potato. Our results crossedwit h tetraploid potato genotypes, which indicate that there might be a major genotypic resulted in 12 plants out of 3000 cultured influence on the results of somatic hybridisation ovules. In the greenhouse all BC2 genotypes experiments. AFLP results showed that were less vigorous than the BC1 parent and no apparently the hexaploid S. americanum that BC3 progeny could be obtained. In the potato was used in our experiments, is an accession (+) tomato backcrossing programme BC2 and of S. nigrum. Nevertheless its performance in BC3 progeny from different BC1 genotypes somatic hybridisation experiments was quit could easily be obtained (Garriga-Caldere efal. different from the performance of the official S. 1997). nigrumaccessio n that we used. Differences in crossability were also observed between S. brevidens (+) potato somatic hybrids with different genomic BACKCROSS EXPERIMENTS constitutions. Ehlenfeldt and Helgeson (1987) were the first to produce backcross progeny Octaploid S. nigrum (+) potato somatic hybrids from this combination. They used tetraploid were backcrossed to both S. nigrum and hybrids with two genomes of both species and potato. Considerable differences were hexaploid hybrids that resulted from fusion observed between both crossing programmes experiments between S. brevidens and a since backcrosses with S. nigrum were rather tetraploid potato genotype. Tetraploid somatic easy to accomplish whereas backcrosses with hybrids crossed poorly with both 2x and 4x potato had a very low success rate. Only 31 potato genotypes, whereas the hexaploid S. pollinations with S. nigrum pollen were brevidens (+) potato somatic hybrids crossed necessary in order to obtain twelve BC1 plants very well with tetraploid potato. Jacobsen ef al. from which one was already self-fertile. From (1993) showed similar results after these BC1 descendants BC2 progeny could backcrossing S. brevidens (+) potato somatic easily be obtained. On the contrary over 4300 hybrids. In their experiments the hexaploid pollinations were necessary in order to obtain somatic hybrids with four genomes of potato two BC1 plants with potato. crossed much better than the tetraploid fusion Potato (+) tomato somatic hybrids showed hybrids. an equally low success rate in their crossing So the production of BC1 progeny was the experiments with potato (Jacobsen ef a/., 1994; main obstruction in the crossing experiments Garriga-Caldere ef a/., 1997). Both tetraploid with potato genotypes of the potato (+) tomato somatic hybrids [2x potato (+) 2x tomato] and as well as the S. brevidens (+) potato somatic hexaploid ones [4x potato (+) 2x tomato] were hybrids, from which only the hexaploid somatic crossed with 4x potato genotypes from which hybrids with an excess of potato genomes were only the latter resulted in backcross progeny. successful. Crosses of the S. nigrum (+) potato

73 Chapter 6

somatic hybrids with potato faced a new of potato and three of S. nigrum. Subsequently obstruction in every generation. This might be the plants were used in backcross experiments caused by the fact that during the backcross inwhic h they were pollinated with potato pollen. programme of the S. nigrum (+) potato hybrids The first preliminary crossing results did not the genomic ratios had to be reversed. The show a major improvement ofth e crossability of initial somatic hybrid contained a surplus of S. the somatic hybrids as compared to the BC1- nigrum genomes and only two genomes of fusion parent. Two genotypes had a higher potato but the aim of the backcross program berry set than the BC1-6738 (23%) but the was to obtain potato material with only the P. number of ovules per berry of these two infestans resistance of S. nigrum.Th e BC1 that genotypes was lower. In fact, none of theBC1 - resulted from a cross with 4x potato was somatic hybrids had a higher number of ovules expected to be a hexaploid genotype that per berry than the BC1-6738. Crossing contained three genomes of both species. The experiments were continued with a selection of plants that resulted from a second backcross the BC1-somatic hybrids which have resulted in with 4x potato were most likely to be four backcross plants (unpublished results). pentaploid, but aneuploid imbalanced genomes Two ofthes e plants were even obtained without for potato as well as S. nigrum were obtained, ovule culture, directly from mature seeds. which might be an explanation for their low Unfortunately the backcross plants are less vigour. vigorous than the BC1-somatic hybrids from In the potato (+) tomato backcross which they were obtained and berryset seems programme the somatic hybrids as well as the to be low. Until now one seed was obtained. BC1 and BC2 genotypes contained four genomes of potato and an additional number of tomato chromosomes that depended on the ASSESSMENT OF THE POSSIBILITIES OF generation. All hybrids in every generation had INTROGRESSION a surplus of potato genomes containing sufficient balanced genetic information for the Our approach was aimed at introgression of development of normally growing plants. In our interesting traits from S. nigrum into the potato crossing programme the S. americanum (+) genome. A prerequisite for success is the Desiree hybrids had a genomic ratio that was occurrence of recombination between comparable to the hexaploid potato (+) tomato chromosomes of the two species. Application of somatic hybrids but, unfortunately, did not GISH showed homoeologous association in the produce any berries after pollination. BC1-6738 and BC2-9019. Despite a clear In an attempt to improve the backcross differentiation between the chromosomes of the ability of the BC1-genotype with potato, the two parental species it was not possible to genomic composition was modified by somatic establish whether or not there were fusion. Two potato genomes were added by recombinant chromosomes in the progenies. somatic hybridisation in order to create a This failure to detect homoeologous surplus of potato genomes in this genotype. recombination might be due to two reasons: Twenty somatic hybrids could be selected • There might be considerable recombination which were expected to contain five genomes but it isdifficul t to detect.

74 Chapter 6

• There is no (or very little) crossing-over homoeologous chromosome pairing and between the chromosomes of potato and S. crossing-over, direct evidence of recombinant nigrum; segments could not be demonstrated cytologically either in Solanum or Lycopersicon In species with large chromosomes like Lily, species hybrids. However, one exception is a with one of the largest genomes in the plant report written by Parokonny ef a/. (1997) who kingdom with a 2C value of 72 pg, intergenomic detected recombinant segments in L recombination in the backcross derivatives can esculentumx L peruvianum hybrids. easily be monitored through GISH (Karlov ef The question is raised why cytological a/., 1999). In species with small genomes, like detection of recombinant segments has not the ones that belong to the Solanaceae (2C been possible in our S. nigrum (+) potato fusion value is approximately 2.0 pg), it is much more hybrids. One problem in general seems to be difficult to show introgression. the distribution of repetitive DNA on small In backcross derivatives of potato (+) chromosomes like those of Solanum and tomato somatic hybrids, in which the Lycopersicon. More than 77% of the repetitive homoeologous chromosomes were equally well DNA is situated in the so-called centromeric differentiated as in our study, no genetic heterochromatin, occupying the proximal parts recombination was detected between potato of the chromosomes (Peterson ef a/., 1996). and tomato chromosomes. This might be The distal euchromatic parts, where chiasmata expected because of the phylogenetic distance and crossing-over have been shown to occur, between potato and tomato since they belong have very little repetitive DNA so that it is to different genera, Solarium and Lycopersicon difficult to differentiate these parts with GISH. respectively. Potato and S. nigrum however, This might be the cause for the detection of the both belong to the genus Solanum, although alleged recombinant fragments in the L being tuberous and non-tuberous species, esculentum x L peruvianum hybrids respectively. In view of the taxonomic position (Parokonny ef a/., 1997). Moreover, the relatively more homoeologous pairing and euchromatic parts in both potato and tomato crossing-over is expected in potato (+) S. are extremely contracted during metaphase nigrum hybrids than in potato (+) tomato stages of mitotic chromosomes (Ramanna and hybrids. This was indeed shown to be the case Prakken, 1967; Ramanna and Wagenvoort, in our experiments. Other combinations of 1976). tuberous and non-tuberous species already Thus, in the absence of convenient showed a considerable amount of cytological methods for the detection of homoeologous chromosome pairing, in sexual recombination, the use of molecular mapping hybrids of S. etuberosum and S. pinnatisectum methods such as the RFLP and AFLP- (Ramanna and Hermsen, 1979; Ramanna and technique appear to be the best alternatives. Hermsen, 1982) as well as in somatic hybrids involving S. etuberosum or S. brevidens for instance (Novy and Helgeson, 1994; McGrath ef a/., 1994). In these cases, however, despite the occurrence of indirect evidence for

75 Chapter 6

ALTERNATIVES TO THE SOMATIC HYBRIDISATION yield (McKendry etal., 1996),dependin g on the APPROACH genotype. Substitution 1R was developed by If recombination between chromosomes of S. extensive selection for several resistances in a nigrum and potato does not occur it might be wheat-rye backcross programme, combined necessary to turn to an alternative approach with cytogenetic monitoring which first resulted like: in monosomic addition lines for the 1R- • Production of addition lines, combined with chromosome and later on in substitution lines. inducedtranslocation ; The translocation lines occurred either • Sexual hybridisation, combined with the spontaneous or were induced by irradiation. useo f 2n-gamet.es. The spontaneous translocation developed from The above options will be discussed in the breakages of wheat and rye univalent following paragraphs. chromosomes at the centromeres, followed by exchange and fusion of the respective chromosome arms (Mettin ef a/., 1978). Productionof addition or substitution lines Induced translocation was applied to backcross progeny of a cross between triticale andwheat , One of the most successful examples of resulting in a translocation in the cultivar substitution of a chromosome isth e substitution 'Amigo' (Sebesta etal., 1994). of the 1B chromosome of wheat by 1R of rye. Within the Solanaceae some examples of Nowadays many commercial wheat varieties the creation of addition lines can be found as posses this 1R substitution or one of the well. A set of monosomic alien addition lines of 1AL/1RS, 1BL/1RS or 1D/1RS translocations S. lycopersicoides chromosomes in a tomato which were derived from it. A survey in 1998 of background was assembled with sexual 454 European wheat varieties found 1RS in methods (Chetelat ef a/., 1998). The second 17% of the genotypes (Kazman ef a/., 1998). example is the selection of a complete series of Approximately 330genotype swit h 1RSfro m 35 monosomic alien tomato addition lines in the different countries were identified by background of the cultivated potato (Garriga- Rabinovich (1998). The short arm of 1R Caldere etal., 1997; Haider Ali etal., inpress) . provides for resistance to yellow rust (Puccinia Although both combinations involve a Solanum stniformis), leaf rust (P. recondita), stem rust and a Lycopersicon species, it should be (P. graminis) (Bartos ef a/., 1973), powdery emphasised that S. lycopersicoides and L mildew (Erysiphe graminis) (Heun and Friebe, esculentum are much more closely related to 1990) and insects (Martin ef a/., 1976). each other than S. tubersoum and L Depending on the wheat genotype into which it esculentum (Rick, 1979) and therefore is introduced, 1RS may also have a positive homoeologous pairing is occurring at a lower effect on yield (Villareal ef a/., 1991; Schlegel frequency in the latter combination than in and Meinel, 1994). Unfortunately 1RS can also hybrids of S. lycopersicoides and L have a negative effect on the bread making esculentum. As a result of this, diploid quality (Martin and Stewart, 1990) or on flour recombinants were identified in progenies of

76 Chapter 6 the S. lycopersicoides addition lines, which Direct hybridisation and the useof 2n-gametes demonstrated the possibility of transferring monogenic characters from S. lycopersicoides A consequence of somatic hybridisation of two to tomato (Chetelat ef a/., 2000). The tomato distantly related species is that the resulting addition lines were established despite hybrids will be allopolyploids. Experiences in abnormal behaviour of the tomato the past have shown that the prospects for chromosomes like preferential chromosome intergenomic recombination between alien elimination, precocious disjunction and division. chromosomes in allopolyploids is minimal Despite these drawbacks transmission of the because of preferential pairing during meiosis tomato chromosomes is fairly high, which in the fusion hybrid. Sexual hybridisation indicates that the addition lines can be used for between diploid potato (2n=2x=2A) and S. further introgression studies. So far nigrum, or one of the other species of the S. introgression of tomato traits into a potato mgrum-complex, might therefore be an background was not observed, but in one case attractive alternative because the resulting a reciprocal translocation was detected hybrids have the advantage of having only a (Garriga-Caldere era/., 1999). homoeologue as a pairing partner for each Pairing between chromosomes of potato chromosome so preferential pairing cannot and chromosomes of tomato seems to occur occur. Such distant allodiploid hybrids can be less frequently than pairing of potato and S. especially useful when they produce 2n- nigrum chromosomes and no preferential gametes, as has been recently shown in Lilium chromosome elimination was observed in (Lim ef a/., 2000). An LA-hybrid of Lilium potato-S. nigrum hybrids. Therefore the longiflorum (L) and an Asiatic Lily hybrid (A) prospects for spontaneous introgression of S. that produced 2n-gametes, was used as a nigrum traits into the potato background from pollinator to produce backcross progeny to the addition lines might be more positive. If, Asiatic Lily hybrid (A), resulting in ALA-BC1 however, introgression is not detected, an progeny. Three to four homoeologous alternative might be found in induced crossovers were observed in some translocation. With this method the addition line chromosomes besides the assortment of alien of interest is irradiated with the aim of chromosomes during the process of sexual translocating the piece of alien chromosome polyploidisation. Another progeny of somatically containing the desirable character to another, polyploidised L/7/um-hybrids was analysed as preferably homoeologous chromosome. This well, showing no evidence of intergenomic technique, however, can only be used for recombination at all (Karlov ef a/., 1999). simple, screenable characters and cannot be Comparable results were found in Alstroemeria controlled since the translocating fragment can (Kamstra ef a/., 1999). Alstroemeria aurea x A. transfer to any other chromosome. Another inadora hybridswer e backcrossed toA. inadora disadvantage is that it is not unlikely that which resulted in six BC1 plants: two undesirable traits might be introduced aswell . aneuploids (2n=2x + 1) and four triploids (2n=3x=24). Recombinant chromosomes were found in all BC1 genotypes, indicating homoeologous recombination during meiosis in

77 Chapter 6 the species hybrid. Analysis of the meiosis in several examples in which successful the triploid BC1-genotypes showed preferential hybridisation was achieved through long years pairing ofA. inodorahybrids . of trials. Crosses between tuberous and non- tuberous species of the section Petota for Finally, while considering the alternatives example were achieved unexpectedly for using the species of the S. nigrum complex (Hermsen and Taylor, 1979). Sexual hybrids of for introgression of desirable traits into potato, it S. lycopersicon and S. lycopersicoides were should be pointed out that the present study successfully backcrossed to Lycopersicon after has demonstrated that homoeologous pairing 40 years, first to L.pennelli (Rick, 1951; Rick ef between S. nigrum and potato chromosomes a/., 1986, Rick ef a/., 1988), later on to L. does occur. Because the alien chromosomes esculentum (Chetelat ef a/., 1997). In the case are transmitted through the gametes, of potato-tomato fusion hybrids, although introgression might be possible. An example of produced by Melchers ef al.i n 1978, they were intergeneric hybrids in which homoeologous backcrossed to potato 15 years later (Jacobsen pairing occurred just as in our BC1-6738 are ef al., 1994). Sexual hybridisation between S. those of Lolium and Festuca (King ef a/., 1998; nigrum x potato was shown to be possible Zwierzykowski ef a/., 1998). In the allopolyploid (Eijlander and Stiekema, 1994) but was hybrids of these species homoeologous pairing extremely laborious. These examples illustrate has led to introgression of characters between that the discovery and use of appropriate these genera. It also illustrates the relevance of genotypes can enable us to achieve success improving the crossability of the complex even inth e case of crosses that were supposed hybrids to the recurrent parent and of improving to be extremely difficult or impossible. So the efficiency of monitoring introgression and whatever approach will be chosen for the selection. introgression of traits from S. nigrum into Despite the extreme difficulty of achieving potato, it will most likely be a matter of long success through sexual hybridisation, the breath. possibility cannot be ruled out. There are

78 SUMMARY Summary

Thespecie so fth eSolanum nigrum-comp\ex ar e Only the somatic hybrids of S. nigrum (+) 2x wild relatives of the cultivated potato and potato were successfully backcrossed to both potentially interesting sources of genetic potato and S. nigrum. First and second variation.Th etraditiona l methodo fintrogressin g generation backcross progeny with S. nigrum a specific trait from a related species is sexual couldeasil yb eobtained .Self-fertilit ywa salread y hybridisation followed by recurrent backcrossing restored in one of the BC1 genotypes. butofte n oneo rmor ereproductiv e barriers have Backcrosses with potato had a much lower to beovercom e inthi s process.Th ediscover yo f success rate. Only pollinations with tetraploid the fusion of protoplasts opened new prospects potato resulted in seed containing berries. Two for further broadening of the genetic base of BC1 genotypes were obtained after 5000 potato because this technique offered the pollinationsfro mwhic h50 5ovule swer ecultured . possibility to usedistantl y related species. Thefirs t genotype, BC1-6738,wa s a vigorously Fusion experiments were performed growing genotype, both in vitro and in the between diploid (2n=2x=24) or tetraploid greenhouse, and flowered abundantly. The (2n=4x=48) potato genotypes and four species second genotype, BC1-9001, showed many belonging to the S. nigrum complex, namely abnormalities and dropped its flowers before Solanum nigrum (2n=6x=72), S. villosum anthesis. BC1-6738 was again crossed with (2n=4x=48),S .chenopodioides (2n=2x=24 )o rS . tetraploid potato and also in this generation the americanum(2n=2x=2 4 and 2n=6x=72).Al lfiv e success rate was low. Over 5000 pollinations accessions ofth e four species ofth e S.nigrum- resulted in 1750 berries from which over 3000 complexwer eabl et ofor mfusio n hybridswit ha t ovuleswer eobtained .Twelve plantsgerminate d least one of the potato genotypes but some fromthes eovules ,whic hwer eno ta svigorou sin combinationswer emor esuccessfu ltha nothers . vitroan din vivoa sth e BC1 parent.Som eo f the Itwa s shown that the ploidy level aswel l as the BC2 genotypes were used for further genotype were factors that influenced the backcrosses but so far no BC3 plants could be somatic combining abilities. In some obtained. combinations the cell-selectable markers The application of genomic in situ kanamycin or hygromycin resistance were used hybridisation (GISH) made it possible to butthe ydi d not influenceth esomati c combining distinguishclearl ybetwee n S.nigrum an dpotat o abilities considerably. However, such markers chromosomesi nmitoti can dmeioti cchromosom e can be useful to improve efficient selection of spreads. It was used to determine the genomic somatichybrid si nsufficien tnumbers .Almos thal f constitution ofth e somatic hybrid F21-26 and its (373)o fth e76 1somati chybri dplant s performed backross progeny. F21-26 was found to be an well in vitro, which was in striking contrast with octaploidwit hsi x genomes ofS .nigrum an dtw o their performancein vivo. Onl y6 0genotype sou t of potato. BC1-6738, the result of a cross of 761 somatic hybrids were vigorous in the betweenF2 1-2 6an d4 xpotato ,appeare dt ob ea greenhouse andwer e able to flower. hexaploidwit h 36 chromosomes ofeac ho f both Vigorous somatic hybrids of the fusion species. Unfortunately it was impossible to combinations 2xS. americanum (+) Desiree, S. determine with GISH whether the 36 chenopodioides(+ ) 2x potato and S. nigrum (+) chromosomes of S. nigrum represented three 2x potato were used in backross experiments. complete genomes. However, AFLP-data

80 Summary

showed that none of the AFLP-specific markers experiments, of which one contained the thatwer e amplified inth e S.nigrum fusio nparen t hygromycin resistance gene. All vigorous and in the somatic hybrid was missing in regenerants that resulted from the fusion BC1-6738, which is an indication that no major experiments were used for the estimation of chromosome elimination hastake nplace .Mos to f nuclear DNA content through flow cytometry. the eleven BC2 genotypes that were analysed Plantswit ha DN Aconten t higher thantha t ofth e had a near-pentaploid genomic constitution with BC1o rBC 2genotype swer econsidere d potential 14-20 chromosomes of S. nigrum and between somatic hybrids. A total of forty-nine potential 33 and 43 chromosomes of potato. somatic hybridsresulte dfro mfusio n experiments The meiotic behaviour of BC1-6738 and with BC1-6738 ,fro mwhic h2 0gre wvigorousl y in BC2-9019 was also studied with GISH. the greenhouse anddi dflower . Eight genotypes Chromosome counts innucle io fBC 2tetra d cells produced seeded berries and five genotypes showed that transmission of alien S. nigrum gave seedless berries after pollination with chromosomes to BC3 progeny is likely. An several potato cultivars. Only five potential average of 9.1 S. nigrum chromosomes per somatic hybrids were detected among the 79 microspore in BC2-9019wa s observed. Meiotic flow-cytometrically analysed regenerants from analysis of metaphase I in BC1-6738 and in BC2-9017(+ )2x potat ofusio nexperiments .Tw o BC2-9019 indicated clearly that allosyndetic of these hybrids were rather vigorous and did pairing in both bivalent and trivalent formation flower, but pollinations with potato did not yet occurs in these genotypes, which is of give any berry set. importancefo rhomoeologou srecombination .S o Eleven potential somatic hybrids of far in none of the backcross derivatives BC1-6738 (+) 2x potato were selected for GISH- recombination could be shown with GISH. analysis todetermin ethei rgenomi c composition. BC1-6738 and eight BC2 genotypes that Theoretically the somatic hybridswer e expected resultedfro mth e backcross programwit h potato tocontai n9 6chromosomes : 36o f S.nigrum an d were tested for their resistance to Phytophthora 60 of S. tuberosum. Six of the selected infestans.Th eBC 1genotyp ewa sa sresistan t as genotypes had this expected genomic the S. nigrumfusio n parent but among the eight constitution andfiv egenotype s deviatedfro mthi s BC2 genotypes that were scored, six were expectation. Two of these five, F108-7 and resistantwherea stw ogenotype s showed lesions F109-6, were most likely missing one or two on the inoculated leaves, indicating they were potato chromosomes but contained 36 susceptible. chromosomes of S. nigrum. One genotype, Since no progeny was obtained from the F104-2,ha d6 0chromosome s ofpotat oorigi nbu t BC2 genotypes, alternative approaches were between one andeigh tadditiona l chromosomes sought to overcome this sexual crossing barrier. or chromosome fragments of S. nigrum per cell With the aim to improve the crossability of the were found.Th efourt h genotype, F101-2, hada backcross derivatives, it was attempted to alter higher number of S. nigrum chromosomes than the genomic composition of BC1-673 8 and expected (42-45) but the number of potato BC2-9017b yaddin gtw ohaploi dpotat ogenome s chromosomes was only slightly higher (41-42) by somatic hybridisation. Five diploid potato than in the BC1-fusion parent which makes it genotypes were used in these fusion doubtful that this genotype is a true somatic

81 Summary hybrid. The fifth genotype with a deviating It can be concluded that traits from S. chromosome number was F109-12 which had nigrumhav e becomeavailabl e forth e cultivated only 22 S. nigrum chromosomes left but did potatowit hth eai do fprotoplas tfusion . However, contain 58t o 60 chromosomes of potato. introgression of these traits by repetitive Crossing experiments were continued with backcrossing with potato is much more selected BC1-somatic hybrids,whic h resulted in complicated then initially expected although it four backcross progeny plants. Two of these seemst ob epossible .Larg enumber so fstartin g plantswer eeve nobtaine dwithou t ovuleculture , material and a lot of work and time seem to be directly from mature seeds. Unfortunately these necessary to accomplish the introgression of S. backcross plants were less vigorous than the n/grum-traits into the cultivated potato. somatic hybrids from which they were derived and berry set seemed to be low.

82 SAMENVATTING Samenvatting

De soorten van het Solanum n/grom-complex fusieprodukten groeide bijna de helft goed in zijn wilde verwanten van onze cultuuraardappel vitro (373)maa r in de kas bleken de meeste en zijn in potentie interessante bronnen van planten niet in staat omt e overleven. Slechts genetische variatie. Detraditionel e manier om 60 hybriden waren in staat om in de kast e een bepaalde eigenschap van een verwante groeien en om bloement e vormen. soort te introduceren inee n cultuurgewas is het Groeikrachtige fusieprodukten van de kruisen vand etwe e soorten gevolgd door het combinaties 2x S. americanum (+) Desiree,S . maken van herhaalde terugkruisingen. In veel chenopodioides (+)2x aardappe l enS .nigrum gevallen wordt deze aanpak bemoeilijkt door (+) 2x aardappel zijn gebruikt om het voorkomen van verschillende kruisings- terugkruisingen te maken. Alleen deS . nigrum barrieres De ontdekking van de mogelijkheid (+) aardappel hybriden leverden nakomelingen van hetfusere n vanprotoplaste n bood nieuwe op, zowel na kruising met aardappel als na perspectieven voor het verbreden van de kruising met S. nigrum. Vooral de terug­ genetische basis van aardappel omdat deze kruisingen metS . nigrum bleken eenvoudig te techniek het mogelijk maakte om ook minder zijn want erware n slechts enkele bestuivingen verwante soorten tegebruiken . nodig voor deeerst e BC1 en BC2-planten.Bi j Vier soorten uit het S. nigrum complex, een vand e BC1-planten wasd e mogelijkheid namelijk Solanum nigrum (2n=6x=72), S. tot zelfbevruchting alweer hersteld. Terug­ villosum (2n=4x=48), S. chenopodioides kruisingen met aardappel waren veel minder (2n=2x=24) of S. americanum (2n=2x=24 and succesvol. Alleen bestuiving met tetraploide 2n=6x=72) zijn gebruikt in fusie-experimenten aardappelgenotypen leidde tot bessen met met diplo'ide (2n=2x=24) en tetraploide zaad. In totaal waren er meer dan 5000 (2n=4x=48) aardappelgenotypen. Alle vijf bestuivingen nodig waaruit 505 zaadknoppen accessies van de vier eerder genoemde werden uitgeprepareerd voor het verkrijgenva n soorten waren in staat om fusieprodukten te twee BC1-nakomelingen. Heteerst e genotype, vormen met minstens een van de aardappel­ BC1-6738, wasee ngroeikrachtig e plant zowel genotypen maar sommige combinaties bleken in vitro alsi n de kase n bloeide uitbundig.D e succesvoller te zijn dan andere. Het tweede BC1, BC1-9001, had veel afwijkingen ploidieniveau en het genotype waren factoren en liet vroegtijdig zijn bloemen vallen. die van invloed waren opd e resultaten van de Ook BC1-6738 werd met tetraploide fusie-experimenten, aangeduid als de aardappel teruggekruist en wederom was het somatische combinatie geschiktheid. In slagingspercentage laag. Meer dan 5000 sommige fusie-experimenten is gebruik bestuivingen resulteerden in 1750 bessen gemaakt van kanamycine en/of hygromycine waaruit 3000 zaadknoppen uitgeprepareerd resistentiegenen waarmee fusieprodukten op konden worden. Hieruit zijn twaalf BC2-planten celniveau geselecteerd konden worden. Deze voortgekomen die geen van alien zo selectiemerkers hadden geen effect op de groeikrachtig waren alsd eBC1-oude r zelf. Met somatische combinatie geschiktheid maar een aantal vand e BC2-genotypen is getracht bleken wel nuttigt e zijn voor het verbeteren van een volgende generatie terugkruisings- de efficientie van de selectie van grote nakomelingent emake n maar tot opdi tmomen t aantallen fusieprodukten. Van de 761 isda t niet gelukt.

84 Samenvatting

Met behulp van genomische in situ trivalenten werden waargenomen waarbij zowel hybridisatie (GISH) was het mogelijk om trivalenten van twee aardappelchromosomen onderscheid te maken tussen S. nigrum- en en een S. n/grum-chromosoom werden aardappelchromosomen in mitotische en gevonden als trivalenten van twee S. nigrum- meiotische chromosoompeparaten. Daardoor chromosomen met een aardappelchromosoom. kon de genomische samenstelling van Allosyndetische paring is een belangrijke fusieprodukt F21-26 en zijn nakomelingen voorwaarde voor homeologe recombinatie. In worden bepaald. F21-26 bleek een octaploidt e geen van de terugkruisingsnakomelingen is zijn metze sgenome n vanS . nigrum en twee echter een recombinant chromosoom m.b.v. genomen van aardappel. BC1-6738, een GISH waargenomen. nakomeling van F21-26 en een tetraploid BC1-6738 en acht BC2-nakomelingen uit aardappel genotype, bleek een hexaploidt ezij n het terugkruisingsprogramma met aardappel met 36 chromosomen van elk van beide zijn getoetst op hun resistentie tegen soorten. Helaas was het met GISH niet Phytophthora infestans. BC1-6738 was even mogelijk om te bepalen ofd e 36 chromosomen resistent alsd e S. n/grum-fusieouder maar de van S. nigrum drie volledige genomen resistentie splitste uiti nd eBC2-generatie .Va n vertegenwoordigden. AFLP-gegevens hebben de acht BC2-genotypen bleken erze s resistent echter aangetoond dat alle AFLP-specifieke te zijn maar opd egei'noculeerd e bladerenva n merkers die in de S. n/grum-fusieouder de overige twee genotypen werden lesies voorkwamen, ook in het fusieprodukt en waargenomen waaruit geconcludeerd werdda t BC1-6738 werden geamplificeerd, wat aangeeft zevatbaarwaren. dat er waarschijnlijk geen grate chromosoom- Omdat tot op heden geen nakomelingen eliminaties hebben plaatsgevonden. Het zijn verkregen uitterugkruisinge n van deBC2 - grootste deel van de elf geanalyseerde BC2- genotypen met aardappel, isnaa r een methode genotypen had een chromosoomaantal dat gezocht om deze kruisingsbarriere te ongeveer overeenkomt met dat van een doorbreken. In een poging om de kruisbaarheid pentaploid met 14-20 S. n/grum-chromosomen te verbeteren, is getracht de genomische en tussen de 33 en 43 chromosomen van samenstelling van BC1-6738 en BC2-9017 te aardappel. veranderden door via protoplastenfusie twee Van BC1-6738 en BC2-9019 is ook de extra haploide aardappelgenomen toe te meiose bestudeerd met behulp van GISH. voegen. In totaal zijn vijf diploide Chromosoomtelling in nuclei vanBC 2 tetraden aardappelgenotypen, waarvan er een het lieten zien dat transmissie van S. nigrum hygromycineresistentiegen bezat, gebruikt in chromosomen van de BC2 naar BC3- deze fusie-experimenten. Van alle nakomelingen zeer waarschijnlijk is. Gemiddeld groeikrachtige regeneranten die daaruit werden 9.1 S.nigrum chromosomen geteldpe r resulteerden, is het DNA-gehalte bepaald met microspore van het genotype BC2-9019. behulp van flow cytometry. Alle planten met Meiotische analyse van de metafase I van een hoger DNA-gehalte dan BC1-6738 of zowel BC1-6738 als BC2-9019 toonde aanda t BC2-9019, werden beschouwd als potentiele allosyndetische paring regelmatig voorkomt in fusieprodukten. Uit de fusie-experimenten met deze genotypen. Zowel bivalenten als BC1-6738 zijn in totaal 49 potentiele

85 Samenvatting fusieprodukten voortgekomen, waarvan er S. n/gn/m-chromosomen (42-45) maar bevatte twintig goed groeiden in de kas en bloemen slechts een beperkt aantal extra aardappel­ vormden. Na bestuiving met verschillende chromosomen dan de BC1-fusieouder aardappelgenotypen werden van acht geno- waardoor het twijfelachtig is dat dit typen bessen metzaad geoogst en gaven vijf daadwerkelijk een fusieprodukt is. Het vijfde genotypen zaadloze bessen. Onder de 79 genotype met een van het verwachte aantal regeneranten van de BC2-9017 (+) 2x afwijkend aantal chromosomen hadslecht s 22 aardappel fusie-experimenten werden slechts S. n/grum-chromosomen en 58-60 aardappel­ vijf potentiele fusieprodukten gevonden. Twee chromosomen. van deze hybriden waren redelijk groeikrachtig De terugkruisingsexperimenten zijn voort- en bloeiden maar bestuivingen met gezet met een selectie van de BC1- aardappelpollen resulteerden niet in bessen. fusieprodukten. Tot nu toe hebben die Van elf potentiele fusieprodukten met geresulteerd in vijf nakomelingen waarvan er BC1-6738 als fusieouder is de genoom- twee zelfs zijn verkregen zonder samenstelling bepaald met behulp van GISH. zaadknopcultuur direct uit afgerijpt zaad. Theoretisch werd verwacht dat zij 96 Helaas waren ook deze terugkruisings- chromosomen zouden hebben: 36 van S. nakomelingen minder groeikrachtig dan de nigrum en 60va n S. tuberosum.Ze s genotypen somatische hybriden waaruit ze zijn verkregen bleken exact deze verwachtte en isd e besvorming slecht. genoomsamenstelling te hebben maar de Geconcludeerd kan worden dat m.b.v. andere vijf weken daar vanaf. Twee vandez e protoplastenfusie de eigenschappen van S. vijf genotypen, F108-7 en F109-6, misten nigrum nuvoo r deaardappe l beschikbaar zijn. hoogstwaarschijnlijk een of twee aardappel- Introgressie van deze eigenschappen d.m.v. chromosomen maar hadden wel 36 herhaald terugkruisen van fusieprodukten lijkt chromosomen van S. nigrum. Een genotype, mogelijk maar is lastiger dan oorspronkelijk F104-2, had het verwachtte aantal van 60 verwacht. Veel uitgangsmateriaal, inspanning aardappelchromosomen maar er werden en geduld lijken nodig om de S. nigrum- tussen de 1-8 extra S. n/gmm-chromosomen eigenschappen daadwerkelijk over te brengen en/of chromosoomfragmenten gevonden. Het naar de cultuuraardappel. vierde genotype, F101-2, had een hoger aantal

86 REFERENCES References

AUSTIN S, LOJKOWSKA E, EHLENFELDT MK, KELMAN A, HELGESON JP (1988) Fertile interspecific somatic hybrids ofSolanum: A nove l source ofresistanc e toErwinia sof t rot. Phytopathology 78:1216-1220

AUSTIN S, POHLMAN JD, BROWN CR, MOJTAHEDI H, SANTO GS, DOUCHES D, HELGESON JP (1993) Interspecific somatic hybridization between Solanum tuberosumL and S.bulbocastanum DUN. as a means oftransfearin g nematode resistance.A m Potato J 70:485-495

BARTOS P, VALKOUN J, KOSNER J, SLOVNCIKOVA V (1973) Rust resistance of some European wheat cultivars derived from rye.4t h IntWhea t Genet Symp, Columbia, MO, USA. Pp145-14 6

BINDING H, JAIN SM, FINGER J, MORDHORST G, NEHLS R, GRESSEL J (1982) Somatic hybridization of an atrazine resistant biotype of Solanum nigrum with Solanum tuberosum. Part 1: Clonal variationi n morphology and inatrazin e sensitivity. Theor Appl Genet 63:273-277

BROWN CR, MOJTAHEDI H, SANTO GS (1989) Comparison of reproductive efficiency of Meloidogyne chitwoodi on Solanumbulbocastanum insoi l andin vitro tests . Plant Dis 73:957-959

BROWN CR, MOJTAHEDI H, SANTO GS (1995) Introgression ofresistanc e toColumbi a and Northern root- knot nematodes from Solanumbulbocastanum into cultivated potato. Euphytica 83:71-78

BROWN CR, YANG C-P, MOJTAHEDI H, SANTO GS, MASUELLI R (1996) RFLP analysis of resistance to Columbia root-knot nematode derived from Solanum bulbocastanum ina BC2-population. Theor Appl Genet 92: 572-576

BUITENVELD J, Suo Y, VAN LOOKEREN CAMPAGNE MM, CREEMERS-MOLENAAR J (1998) Production and characterization ofsomati c hybrid plants between leek (Alliumampeloprasum L. )an d onion (Allium cepa L). Theor Appl Genet 96:765-775

CHETELAT RT, CISNEROS P, STAMOVA I, RICK CM (1997) A male-fertile Lycopersicon esculentum x Solanum lycopersicoides hybrid enables direct backcrossing to tomato at the diploid level. Euphytica 95:99-108

CHETELAT RT, RICK CM, CISNEROS P, ALPERT KB, DEVERNA JW (1998) Identification, transmission,an d cytological behavior of Solanum lycopersicoides Dun. Monosomic alien addition lines intomat o (Lycopersicon esculentum Mill.). Genome 41:40-50

CHETELAT RT, MEGLIC V (2000) Molecular mapping of chromosome segments introgressed from Solanum lycopersicoides into cultivated tomato (Lycopersicon esculentum). Theor Appl Genet 100: 232-241

COLON LT, EIJLANDER R, BUDDING DJ, VAN IJZENDOORN MT, PIETERS MMJ, HOOGENDOORN J (1993) Resistance to potato late blight (Phytophthora infestans (Mont.) de Bary) inSolanum nigrum, S. villosuman d their sexual hybridswit h S.tuberosum an d S. demissum.Euphytic a 66:55-64

DAUNAY MC, CHAPUT MH, SIHACHAKR D, ALLOT M, VEDEL F, DUCREUX G (1993) Production and characterization of fertile somatic hybrids of eggplant (Solanum melongena L) with Solanum aethiopicumL . Theor Appl Genet 85:841-850 References

DE JONG JH, WOLTERS AMA, KOK JM, VERHMR H, VAN EDEN J (1993) Chromosomes pairing and potential for intergeneric recombination in some hypotetraploid somatic hybrids of Lycopersicon esculentum(+ ) Solanumtuberosum. Genome 36:1032-1041

DE LAAT AMM, GOHDE W, VOGELZANG MJDC (1987) Determination ofploid y ofsingl e plants and plant populations byflo w cytometry. Plant Breeding 99:303-307

EDMONDSJ M (1979) Biosystematicso fSolanum sectio n Solanum(Maurella) . In: Hawkes JG,Leste r RN, Skelding AD (ed), The Biology andtaxonom y of Solanaceae. Academic Press, London, pp 529-551

EDMONDS JM, CHWEYAJ A (1997) Black nightshades. Solanumnigrum L. and related species. Promoting the conservation and use ofunderutilize d and neglected crops. 15. Institute ofPlan t Geneticsan d Crop Plant Research, Gatersleben/International Plant Genetic Resources Institute, Rome, Italy

EHLENFELDT MK, HELGESON JP (1987) Fertility ofsomati c hybrids from protoplast fusions ofSolanum brevidens and S. tuberosum.Theo r Appl Genet 73:395-402

EIJLANDER R, STIEKEMA WJ (1994) Biological containment ofpotat o (Solanum tuberosum): outcrossing to the related wild species black nightshade (Solanum nigrum) and bittersweet (Solanum dulcamara). Sex Plant Reprod 7:29-40

FAHLESON J, RAHLEN L, GLIMELIUS K (1988) Analysis of plants regenerated from protoplast fusions between Brassica napusan d Erucasativa. Theo r Appl Genet 76:507-512

FAHLESON J, DIXELIUS J, SUNDBERG E, GLIMELIUS K (1988) Correlation between flow cytometric determination of nuclear DNAconten t and chromosome number in somatic hybrids within Brassicaceae. Plant Cell Reports 7:74-77

FAHLESON J, ERIKSSON I, GLIMELIUS K(1994a ) Intertribal somatic hybrids between Brassica napusan d Barbarea vulgaris- production ofi nvitr o plantlets. Plant Cell Rep 13:411-416

FAHLESON J, ERIKSSON I, LANDGREN M, STYMNE S, GLIMELIUS K (1994b) Intertribal somatic hybrids between Brassica napus and Thlaspi perfoliatum with high content of T.perfb/zafum-specifi c nervonic acid. Theor Appl Genet 87:795-804

FISH N, KARP A, JONES MGK (1988) Production ofsomati c hybrids byelectrofusio n inSolanum. Theor Applied Genet 76:113-117

FORSBERG J, LANDGREN M, GLIMELIUS K (1994) Fertile somatic hybrids between Brassica napusan d Arabidopsis thaliana. Plant Sci Limerick 95:213-223

GARRIGA-CALDERE F, HUIGEN DJ, FILOTICO F, JACOBSEN E, RAMANNA MS (1997) Identification of alien chromosomes through GISH and RFLP analysis and the potential for establishing potato lines with monosomicaddition s of tomato chromosomes. Genome 40:666-673

GARRIGA-CALDERE F, HUIGEN DJ, ANGRISANOA , JACOBSENE , RAMANNA MS (1998) Transmission of alien tomato chromosomes from BC1 toBC 2 progenies derived from backcrossing potato (+) tomato fusion hybrids topotato : the selection of single additions forseve n different tomato chromosomes. Theor Appl Genet 96:155-163

89 References

GARRIGA-CALDERE F, HUIGEN DJ, JACOBSEN E, RAMANNA MS (1999) Prospects for introgressing tomato chromosomes into thepotat o genome: anassessmen t through GISH analysis. Genome 42: 282- 288

GIBSON RW, JONES MGK, FISH N(1988 ) Resistance topotat o leaf roll virus and potato virus Yi n somatic hybrids between dihaploid Solariumtuberosum an d S.brevidens. Theor ApplGene t 76:113-117

GLEBA YY, HOFFMAN F (1979) "Arabidobrassica": plant-genome engineering by protoplast fusion. Naturwissenschaften 66:547-554

GLEBA YY, HOFFMAN F(1980 ) "Arabidobrassica": anove l plant obtained byprotoplas t fusion. Planta 149: 112-117

GRESSEL J, COHEN N, BINDING H (1984) Somatic hybridization of an atrazine resistant biotype of Solatium nigrum with Solanum tuberosum. 2. Segregation of plastomes. Theor Appl Genet67 : 131-134

HAIDER ALI SN, RAMANNA SM, JACOBSEN E, VISSER R. Establishment of a complete series of a monosomic tomato chromosome addition lines in the cultivated potato using RFLP and GISH analyses. Theor App. Genet: inpres s

HAWKES JG (1978) Biosystematics ofth e potato. In: Harris PM (ed) The potato crop; the scientific basis for improvement. Chapman and Hall, London, UK,p p15-6 9

HELGESON JP, POHLMAN JD, AUSTIN S, HABERLACH GT, WIELGUS SM, RONIS D, ZAMBOUM L, TOOLEY P, MCGRATH JM, JAMES RV, STEVENSON WR (1998) Somatic hybrids between Solanum bulbocastanum andpotato : a newsourc e of resistance to late blight. Theor Applied Genet96 : 738-742

HERMSEN JGTH (1966) Crossability, fertility and cytogenetic studies in Solanum acaule x Solanum bulbocastanum. Euphytica 15:149-155

HERMSEN JGTH, RAMANNA MS (1973) Double-bridge hybrids ofSolanum bulbocastanum and cultivarsof Solanum tuberosum. Euphytica 22:457-466

HERMSEN JGTH, TAYLOR LM (1979) Successful hybridization ofnon-tuberou s Solanum etuberosum Lind. and tuber-bearing S.pinnatisectum Dun. Euphytica 28:1-7

HERMSEN JGTH, RAMANNA MS, SAWOR Z (1981) The effect ofchromosom e doubling onfertility , meiotic behaviour and crossability ofSolanum etuberosumx S.pinnatisectum. Euphytica 30:33-39

HERMSEN JGTH (1983) Utilization ofwid e crosses in potato breeding. Report ofa planning conference on present andfutur e strategies forpotat o breeding and improvement. International Potato Center, Lima, Peru, pp 115-132

HERMSEN JGTH (1994) Introgression of genes from wild species, including molecular and cellular approaches. In: Bradshaw J, MacKay G(eds ) Potato Genetics. Cab International, Wallingford, UK, pp 515-538

90 References

HEUN M, FRIEBEA E (1990) Introgression ofpowder y mildew resistance from ryeint o wheat. Phytopath 80:242-245

HORSMAN K, BERGERVOET JEM, JACOBSEN E(1997 ) Somatic hybridization between Solanumtuberosum and species of the S. nigrum complex: Selection of vigorously growing and flowering plants. Euphytica 96: 345-352

HORSMAN K, FRATINIR , HUIGEN DJ, JACOBSENE (1999) Successfulfirs t and second backcrosseso f S. nigrum (+)S . tuberosum somatic hybrids with both Solanum parents. SexPlan t Reprod 12: 144- 151

JACOBS JME, VANEC KHJ , ARENS P, VERKERK-BAKKER B, TE LINTEL HEKKERT B, BASTIAANSSEN HJM, EL- KHARBOTLY A, PEREIRA A, JACOBSEN E, STIEKEMA WJ (1995) A genetic map of potato (Solanum tuberosum) integrating molecular markers, including transposons, and classical markers. Theor Appl Genet 91:289-300

JACOBSEN E, HOVENKAMP-HERMELINK JHM, KRIJGSHELD HT, NIJDAM H, PIJNACKER LP, WITHOLD B, FEENSTRAWJ (1989) Phenotypic and genotypic characterization ofa n amylose-free starch mutant of potato. Euphytica 44:43-48

JACOBSEN E, REINHOUT P, BERGERVOET JEM, DE LOOFF J, ABIDIN PE, HUIGEN DJ, RAMANNA MS (1992) Isolation and characterization of potato-tomato somatic hybrids using an amylose-free potato mutant as parental genotype. Theor Appl Genet 85:159-164

JACOBSEN E, MALVAR R, HUIGEN DJ, BERGERVOET JEM, RAMANNA MS (1993) Isolation and characterization ofsomati c hybrids ofdiploi d Solanum tuberosum and Solanum brevidensan d the use of amylose-free starch mutationfo rdetectio no fintrogression . Euphytica 69:191-201

JACOBSEN E, DANIEL MK, BERGERVOET-VAN DEELEN JEM, HUIGEN DJ, RAMANNA MS (1994) The first and second backcross progeny of the intergeneric fusion hybrids ofpotat o andtomat o after crossing with potato.Theo r Appl Genet 88:181-186

JACOBSEN E, DE JONG JH, KAMSTRA SA, VAN DEN BERG PMMM, RAMANNA MS (1995) Genomic in situ hybridization (GISH) andRFL P analysis for the identification of alien chromosomes in the backcross progeny ofpotat o (+) tomato fusion hybrids. Heredity 74:250-257

JANSSEN GJW, VAN NOREL A, VERKERK-BAKKER B, JANSSEN R (1996) Resistance to Meloidogyne chitwoodi,M. fallaxan d M. hapla inwil d tuber-bearing Solanum spp. Euphytica 92:287-294.

JEFFERSON RA, KAVANAGH TA, BEVAN M (1987) GUS-fusions: ^-glucuronidase as a sensitivean d versatile gene fusion marker in higher plants. EMBOJ 6:3901-390 7

KAMSTRA SA, KUIPERS AGJ, DE JEU MJ, RAMANNA MS, JACOBSEN E (1997) Physical localization of repetitive DNAsequence s in Alstroemeria: karyotyping oftw ospecie s with species-specific and ribosomal DNA. Genome 40:652-658

KAMSTRA SA, KUIPERS AGJ, DE JEU MJ, RAMANNA MS, JACOBSEN E (1999) The extent and position of homoelogous recombination in a distant hybrid of Alstroemeria: a molecular cytogenetic assessment offirs t generation backcross progenies. Chromosoma 108: 52-63

91 References

KARLOV Gl, KHRUSTALEVA LI, LIM KB, VAN TUYL JM (1999) Homoeologous recombination in 2n-gametes producing interspecific hybrids ofLilium (Liliaceae) studied bygenomi c insitu hybridization (GISH). Genome 42: 681-686

KAZMAN ME, LEIN V, ROBBELEN G (1998) The 1B.1RS translocation in recently developed European wheats. In: Lelley T (ed), Current topics in plant cytogenetics related to plant improvement. WUV- Univ-Verl,Austria , pp 334-341

KING IP, MORGAN WG, ARMSTAED IP, HARPER JA, HAYWARD MD, BOLLARD A, NASH JV, FORSTER JW, THOMA HM (1998) Introgression mapping in the grasses. I. Introgression of Festuca pratensis chromosomes and chromosome segments intoLolium perenne. Heredity 81:462-467

LELIVELT CLC, KRENS FA(1992 ) Transfer ofresistanc e toth e beet cyst nematode (Heteroderaschachtii Schm.) into Brassica napus L.gen e pool through intergeneric somatic hybridization with Raphanus sativus L. TheorApp lGene t 83:887-894

LELIVELT C, LEUNISSEN E, FREDERIKS, HJ, HELSPER J, KRENS FA (1993) Transfer of resistance to the beet cyst nematode (Heterodera schachtii Schm) from Synapsis alba L.(whit e mustard) toth e Brassica napus L. gene pool bymean s ofsexua l and somatic hybridization.Theo r Appl Genet 85:688-696

LIM K-B, CHUNGJ-D, VAN KRONENBURG BCE, RAMANNA MS (2000) Introgression ofLilium rubellum Baker chromosomes into L longiflorumThunb. : agenom e painting study ofth e F1 hybrid, BC1 and BC2 progenies.Chromosom e Research 8:119-125

MARTIN DJ, STEWARTB G (1990) Dough stickiness inrye-derive d wheat cultivars. Euphytica 51:77-86

MATTHEIJ WM, EIJLANDER R, DE KONING JRA, LOUWES KM (1992) Interspecific hybridization between the cultivated potato Solanum tuberosum subspecies tuberosum L. and the wild species S. circaeifolium subsp. circaeifolium Bitter exhibiting resistance to Phytophthora infestans (Mont.)d e Bary and Globoderapallida (Stone) Behrens. 1. Somatic hybrids.Theo r Appl Genet 83:459-466

MCKENDRY AL, TAGUE DN, FINNEY PL, MISKIN KE (1996) Effect of 1BL.1R So n milling and baking quality of soft redwinte r wheat. Crop Sci 36:848-851

MCGRATH JM, WIELGUS SM, UCHTYIL TF, KIM-LEE H, HABERLACH GT, WILLIAMS CE, HELGESON JP (1994) Recombination of Solanum brevidens chromosomes in the second backcross generation from a somatic hybrid with S. tuberosum.Theor . Appl. Genet 88:917-924

MELCHERS G, SACRISTAN MD, HOLDER AA (1978) Somatic hybrid plants of potato and tomato regenerated from fused protoplasts. Carlsberg Res Commun 43:203-218

METTIN D, BLUTHNER WD, WEINRICH M (1978). Studies on the nature and possible origin of the spontaneously translocated 1B-1R chromosome inwheat . Wheat InfSer v 47,48:12-16

NEAL CE, TOPOLESKI LD(1983 ) Effects ofth e basal medium ongrowt h ofimmatur e embryo's invitro .J AmerHort Sci 108:434-438

NIEDERHAUSEN JS, MILLS WR (1953) Resistance of Solanum species to Phytophthora infestans in Mexico. Phytopathology 43:456-457

92 References

Now RG, HELGESON JP (1994) Somatic hybrids between Solanum etuberosum and diploid, tuber bearing Solanumclones .Theo r Appl Genet 89:775-782

OZMINKOWSKI JRRH , JOURDAN P (1994a) Comparing the resynthesis of Brassica napus L.b y interspecific somatic andsexua l hybridization. I. Producing and Identifying Hybrids. J AmerSo c HortScil19 : 808-815

OZMINKOWSKI JRRH , JOURDAN P (1994b) Comparing the resynthesis of Brassica napus L.b y interspecific somatic and sexual hybridization. II. Hybrid Morphology and Identifying Organelle Genomes. JAme r Soc Hort Sci119: 816-823

PAROKONNYAS , MARSHALL JA, BENNET MD, COCKING EC, DAVEY MR, POWER JB (1997) Homoeologous pairing and recombination in backcross derivatives of tomato somatic hybrids [Lycopersicon esculentum(+ ) L peruvianum].Theo r Appl Genet 94:713-723

PEHU E, KARP A, MOORE K, STEELE S, DUNCKLEY R, JONES MGK (1989) Molecular, cytogenetic and morphological characterization ofsomati c hybrids ofdihaploi d Solanum tuberosum and diploidS . brevidens. Theor.Appl . Genet. 78:696-704.

PEHU E, GIBSON RW, JONES MGH, KARP A (1990) Studies onth egeneti c basis ofresistanc e topotat o leaf roll virus, potato virus Yan d potato virus Xi nSolanum brevidens using somatic hybridso f Solanumbrevidens and Solanum tuberosum. Plant Science 69:95-101

PETERSON DG, PRICE HJ, JOHNSTON JS, STACK SM (1996) DNA content of heterochromatin and euchromatin intomat o (Lycopersicon esculentum) pachytene chromosomes. Genome 39:77-82

PIJNACKER LP, FERWERDA MA (1984) Giemsa C-banding ofpotat o chromosomes. CanJ Gene t Cytol 26: 415-419

PIJNACKER LP, FERWERDA MA, PUITE KJ, SCHAART JG (1989) Chromosome elimination and mutation in tetraploid somatic hybrids ofSolanum tuberosum and Solanum phureja. Plant Cell Reports 8: 82- 85

PIJNACKER LP, SREE RAMULU K(1990 ) Somaclonal variation inpotato :a karyotypi c evaluation. ActaBo t Neerl 39:163-169

PREISZNER J, FEHER A, VEISZ O, SUTKA J, DUDITS D (1991) Characterization of morphological variation and cold resistance ininterspecifi c somatic hybrids between potato {Solanumtuberosum L. )an d S. brevidens Phil. Euphytica 57:37-49

PUITE KJ, TEN BROEKE W, SCHAART JG (1988) Inhibition ofcel l wal synthesis improves flow cytometric sorting ofpotat o heterofusions resulting inhybri d plants. Plant Sci 56:61-68

RABINOVICH SV (1998) Importance ofwheat-ry e translocations forbreedin g modern cultivars of Triticum aestivumL . Euphytica 100:323-340

RAMANNA MS, PRAKKEN R (1967) Structure and homology between pachytene and somatic metaphase chromosomes oftomato . Genetica 38:115-133

93 References

RAMANNA MS, HERMSEN JGTH (1971) Somatic chromosome elimination and meiotic chromosome pairing in the triple hybrid 6x-{Solanum acaule x S .bulbocastanum) x 2x - S.phureja. Euphytica 20:470- 481

RAMANNA MS, WAGENVOORT M (1976) Identification ofth e trisomic series indiploi d Solanum tuberosum L, group tuberosum I.Chromosome s identification. Euphytica 30:15-31

RAMANNA MS, HERMSEN JGTH (1979) Unique meiotic behaviour in Fi plants from acros s between anon - tuberous and atuberou s Solanumspecie s insectio n petota. Euphytica 28:9-15

RAMANNA MS, HERMSEN JGTH (1981) Structural hybridity inth eserie s Etuberosa ofth e genus Solanum andit sbearin g on crossability. Euphytica 30:15-31

RAMANNA MS, HERMSEN JGTH (1982) Gene transfer from non-tuberous to tuberous Solanum species: Evidence from meiosis inhybrids . Euphytica 31:565-572

RAMULU KS, DIJKHUIS P, RUTGERS E, BLAAS J, VERBEEK WHJ, VERHOEVEN HA, COLIJN-HOOYMANS CM (1995) Microprotoplast fusion technique: ane w tool forgen e transfer between sexually-incongruent plant species. Euphytica 85:255-268

RICK CM (1951) Hybrids between Lycopersicon esculentum Mill. AndSolanum lycopersicoidesDun . Genetics 37:741-744

RICK CM (1979) Biosystematic studies in Lycopersicon and closely related species of Solanum. In: Hawkes JG, Lester RN, Skelding AD(eds ) The Biology and Taxonomy of the Solanaceae. Academic Press, London, pp 667-678

RICK CM, DEVERNA JW, CHETELAT RT, STEVENS MA (1986) Hybrids between Lycopersicon esculentum and Solanumlycopersicoides Dun. Proc NatlAca d Sci USA 83:3580-3583

RICK CM, CHETELAT RT, DEVERNA JW (1988) Recombination insesquidiploi d hybrids of Lycopersicon esculentumx Solanumlycopersicoides and derivatives. Theor Appl Genet 76:647-655

ROKKAV-M, XUY-S , KANKILA J, KUUSELAA, PULLI S, PEHU E (1994) Identification ofsomati c hybrids of dihaploid Solanumtuberosum lines and S.brevidens byspecifi c RAPD patterns and assessmento f disease resistance ofth e hybrids. Euphytica 80:207-21

SCHLEGEL R, MEINEL A (1994) Aquantitativ e trait locus (QTL) onchromosom e arm1R S of rye and its effect on yield performance of hexaploid wheat. Cereal Res Commun 22:7-13

SCHOENMAKERS HCH, WOLTERS AMA, NOBEL EM, DE KLEIN CMJ, KOORNNEEF M (1993) Allotriploid somatic hybrids ofdiploi d tomato (Lycopersicon esculentum Mill.) and monoploid potato (Solanum tuberosum L). Theor Appl Genet 87:328-336

SCHWARZACHER T, HESLOP-HARRISON JS (1994) Direct fluorochrome labeled DNA probes for direct fluorescent in situ hybridization to chromosomes. In: Isaac PG (ed) Methods inMolecula r Biology, vol. 28: Protocols for Nucleic Acid Analysis by Non-radioactive Probes. Humana Press Inc Totowa NJ,pp8-17

94 References

SEBESTA EE, WOOD JR EA, PORTER DR, WEBSTER JA, SMITH EL (1994) Registration of Amigo wheat germplasmresistan tt o greenbug.Cro p Sci 34:293

SEITHEA (1962 ) Die Haarartende r Gattung Solatium L. und ihretaxonomisch eVerwertun g BotJah r 81: 261-336

SHEPARD JF, BIDNEY D, BARSBY T, KEMBLE R(1983 ) Genetic transfer in plants through interspecific protoplastfusion . Science 219:683-688

SINGH RN, ROY SK(1985 ) Cytomorphological studies of dodecaploid (12x) Solanum nigrum Linn. Cytologia 50:59-68

SJODINC , GUMELIUS K (1989a) Brassicanaponigra, a somati c hybrid resistant to Phomalingam. Theor. Appl. Genet. 77:651-656

SJODIN C, GUMELIUS K(1989b ) Transfer of resistance against Phoma lingam to Brassica napus by somatic hybridization combinedwit htoxi n selection.Theor .Appl . Genet. 78:513-520

SKARZHINSKAYA M, LANDGREN M, GUMELIUS K (1996) Production ofintertriba l somatic hybrids between Brassicanapus L . and Lesquerellafendleri (Gray ) WatsTheo rApp l Genet 93:1242-1250

SUNDBERG E, LANDGREN M, GLIMELIUS K(1987 ) Fertility andchromosom e stability inBrassica napus resynthesised byprotoplas tfusion . TheorApp l Genet 75:96-104

SUNDBERG E, GLIMELIUS K (1991) Effects of parental ploidy level and genetic divergence on chromosome elimination and chloroplast segregation in somatic hybrids within Brassicaceae. Theor Appl Genet 83:81-88

VALKONEN JPT, XUY-S , ROKKA V-M, PULLI S, PEHU E (1994) Transfer of resistance to potato leafroll virus, potato virus Y and potato virus X from Solanum brevidens to S. tuberosum through symmetric anddesigne dasymmetri c somatic hybridisation.An nApp l Biol 124:351-362

VALKONEN JPT, WATANABE KN, PEHU E(1994 ) Analysis ofcorrelatio n between nuclear DNA content, chromosome number, and flowering capacity ofasymmetri c somatic hybrids ofdiploi d S. brevidens and(di)haploi d S. tuberosum.Jp nJ Gene t 69:525-536

VAN ECKHJ , ROUPPE VAN DER VOORT J, DRAAISTRA J, VAN ZANDVOORT P, VANENCKEVOR T E, SEGERS B, PELEMAN J, JACOBSEN E, HELDER J, BAKKER J (1995) The inheritance and chromosomal localization ofAFL P markers ina non-inbre dpotat ooffspring . Molecular Breeding 1:397-410

VILLAREAL RL, BANUELOS O, MUJEEB-KAZI A (1991) The effect of chromosome 1B/1R translocation on the yield potential ofcertai n springwheat s (Triticumaestivum L). Plant Breed 106:77-81

VLEESHOUWERS VGAA, VAN DOOIJEWEERT W, GOVERS F, KAMOUN S, COLON LT (2000) The hypersensitive response isassociate dwit h hostan d nonhost resistancet oPhytophthora infestans. Planta 210:853-864

95 References

Vos P, HOGERS R, BLEEKER M, REIJANS M, VANDE R LEE T, HORNES M, FRIJTERS A, POT J, PELEMAN J, KUIPER M, ZABEAU M (1995) AFLP:a ne w technique forDN A fingerprinting. NuclAci d Res 23 (21): 4407-4414

WAARA S, GLIMELIUS K(1995 ) Thepotentia l ofsomati c hybridization incro p breeding. Euphytica85 : 217-233

WATANABE KN, ORRILLO M, VEGA S, VALKONEN JPT, PEHU E, HURTADO A, TANKSLEY SD (1995) Overcoming crossing barriers between nontuber-bearing and tuber-bearing Solatium species: towards potato germplasm enhancement with a broad spectrum of solanaceaous genetic resources. Genome 38:27-355

WIJBRANDI J(1989 ) Isolation and characterisation ofsomati c hybrids between Lycopersiconesculentum andL peruvianum. PhDThesis , Wageningen Agricultural University, The Netherlands, pp. 104

WIJBRANDI J, POSTHUMA A, KOK JM, RIJKEN R, VOSJGM , KOORNNEEF M (1990) Asymmetric somatic hybrids between Lycopersicon esculentum and irradiated Lycopersicon peruvianum. Theor Appl Genet 80:305-312

WILLIAMS CE, WIELGUS SM, HABERLACH GT, GUENTHER C, KIM-LEE H, HELGESON JP (1993) RFLP analysis ofchromosoma l segregation inprogen y from an interspecific hexaploid somatic hybrid between Solanum brevidensan d Solanumtuberosum. Genetics 135:1167-1173

WOLTERS AMA, SCHOENMAKERS HCH, VAN DER MEULEN-MUISERS JJM, VAN DER KNAAP E, DERKS FHM, KOORNNEEF M, ZELCER A (1991) Limited DNA eliminationfrom the irradiated potato parent infusio n products ofalbino Lycopersicon esculentum and Solanum tuberosum. Theor Appl Genet 83: 225- 232

WOLTERS AMA, SCHOENMAKERS HCH, KAMSTRA S, EDEN J, KOORNNEEF M, DE JONG JH (1994) Mitotic and meiotic irregularities insomati c hybrids ofLycopersicon esculentum and Solanum tuberosum. Genome 37: 726-735

WOLTERSAM A (1996) Analysis ofnuclea r and organellar DNA in somatic hybrids between solanaceous species. PhD thesis,Wageninge n Agricultural University, The Netherlands, pp 1-152

ZABEAU M, VOS P (1993) Selective restriction fragment amplification: a general method for DNA fingerprinting. European PatentApplicatio n 92402629.7; Publication number EP 0534858 A1

ZWIERZYKOWSKI, Z, TAYYAR R, BRUNELL M, LUKASZEWSKI AJ (1998) Genome recombination in intergeneric hybrids between tetraploid Festuca pratensis and Lolium multiflorum. TheJourna lo f Heredity 89:324-328

96 NAWOORD Nawoord

Eindelijk klaar! ledereen die mij een beetje kent weet dat dat een hele opluchting is. Mijn contract als Aio was tenslotte al een aantal jaren geleden afgelopen maar door een aantal "vertragende factoren" heeft het verschijnen van dit proefschrift nogalo p zich latenwachten . Eerst even terug naar hetbegin .

September 1993 ben ik als Aio begonnen bij Plantenveredeling nadat ik een jaar bij het CPRO had gewerkt als NOP-per (NOP staat voor Na-doctoraal Onderzoeksproject) en tot de conclusie was gekomenda t iktoc hech tverde r wilde in hetonderzoek . Die kans kreeg ik dus envo l enthousiasme ben ik aan de slag gegaan. Het project werd gefinancierd door de "Stichting Bevordering Veredelingsonderzoek" waarin de aardappelkweekbedrijven Kama, Stet Holland en Hettema participeerden en waarvan prof. Hermsen adviseur was. Twee keer perjaa r werd er vergaderd met de kwekers wat altijd heel gezellig was en waar ik altijd kon rekenen op belangstelling voor mijn resultaten. Hetwa s iedere keer echter ook duidelijk dat mijn materiaal nog ver van de praktijk afstond.

MijnAio-onderzoe k maakte deel uit van een grater project waar ook een postdoc in zat, Ronald Hutten. Ik had nog nooit iemand zo enthousiast over aardappels en de aardappelveredeling horen praten als hij en al snel was ik ingewijd in de wereld van de pollenpreparaten, de veldproeven, de 2x-4x-kruisingen, de knollenlijsten, het chips bakken, de 1e, 2s*etc . jaars en ga zo maar door. Ondertussen was ik bezig met plantjes in glazen potjes en labproefjes maar het was duidelijk dat mijn werk een hoger doel had: de perfecte pieper!

Die plantjes in die glazen potten dat waren fusieprodukten (als alles goed was gegaan), stuk voor stuk gemaakt door Marjan Bergervoet. Marjan, eigenlijk zou je voor de gein eens uit moeten rekenen hoeveel glazen potten en buizen je door je handen hebt laten gaan en hoeveel liter medium daarmee gemoeid was ina l die projecten diej e van plantmateriaal hebtvoorzien . Hetzo u me niet verbazen als je daar een zwembad mee kunt vullen en dan bedoel ik niet zo'n pierebadje maar eentje van wedstrijdformaat.

In het lab brak de FISH en GISH-periode in voile hevigheid uit op het moment dat ik aan mijn project begon. De eerste beginselen hadden we opgepikt bij Erfelijkheidsleer bij Hans de Jong maar als snel konden we bij Plantenveredeling ook aan de slag. In het begin was het nog een beetje pionieren met o.a. een waterbad dat op zolder onder het stof vandaan was gehaald maar Anja, Silvan en Fransesc wisten daar in rap tempo verandering in te brengen en zo ontstond er een heus FISH-lab dat garant stond voor zeer kleurrijke resultaten.Tanya , thank you for helping met o paint my material.

En toen werd het zomer. "Kom", zei Dirk-Jan Huigen, "we gaan de kas in, kruisingen maken." En zo stond ik ineens labeltjes te schrijven, bloemetjes te bestuiven en alles netjes bij te houden in mijn kruisingsboekje. Dat klinkt allemaal romantisch maar oh wee als het echt zomer werd, dan was het nauwelijks meer uit te houden en ging de lol er snel vanaf. De suggestie van Dirk-Jan was om dan wat eerder op de ochtend te beginnen. Een klein probleempje, voor 9 uur 's ochtends bevind ik mij nog in mijn "sub-optimale" periode, om maar eens een understatement te gebruiken. Er zat dus weinig anders op dan om even flink af te zien. Gelukkig heb ik daar ook wel wat hulp bij gehad: Richard Fratini en

98 Nawoord

Andro Tjin Wong Joe hebben beiden tijdens hun afstudeervakken ook menige druppel gezweet in de kas.

Na een dag op het lab of in de kas was het lekker om in de aardappelkamer even na te praten en een potje "bommen te vegen". Met Heleen, Herman en de twee Ronalden heb ik menig nuttig en nutteloos gesprek gevoerd. Herman altijd even kritisch, Ronald E. ge'i'nteresseerd in alles, Ronald H. "nu even niet want het is half een en dus tijd om te klaverjassen" en Heleen altijd spontaan en gezellig. We hebben toch heelwa t afgelachen en dat was zeer heilzaamtege n stress en neerslachtigebuien .

En dan moest er natuurlijk ook met enige regelmaat serieus wetenschappelijk gediscussieerd en gepland worden. Evert, jij was altijd in staat om mij in korte tijd weer bij de les te krijgen. Wat ik erg bewonderd heb isda t hoe drukj e het ook had,j e altijd binnen een paar dagen reageerde op plannen en concept-artikelen. Dr. Ramanna,than k youfo r your help,especiall y during the final stages ofwritin g this thesis. You were always an inspiration. Professor Hermsen, bedankt voor alle adviezen en aanmoedigingen bij het schrijven van de verschillende hoofdstukken.

En toen was mijn contract afgelopen en liep iktege n de eerste vertragende factor aan,d egeboort e van onze dochter. Lieve Robin, van alle vertragende factoren was jij absoluut de leukste. Najou w geboorte werd het een stuk moeilijker om aan schrijven toe te komen want spelen met jouw was veel leuker. Daarnaast had ikal s nieuwbakken ouder geen idee "what hit me"e nware n zaken als gebroken nachten niet echt bevorderlijk voor de schrijverij. Maar opjou w manier hebj e me ook geholpen. Op de moeilijke momenten was jij altijd in de buurt om me even af te leiden en niets werkt zo therapeutisch als twee kleine armpjes om mijn nek, een natte zoen op m'n wang en een lach van oor tot oor die je me geeft zonder dat ik er iets voor hoeft te doen

De tweede vertragende factor diende zich aan in de vorm van een baan toen Robin bijna een jaar oud was. En nuwer k ik op het Bestuurscentrum en ben ik een van die ambtenaren in "het witte gesticht". Ik moet jullie vertellen, ik heb het er ontzettend naar mijn zin. Leuke collega's die ook allemaal erg mee hebben geleefd bij de laatstefas e van mijn promotie, leukwerk , watwi lj e nog meer! Integenstellin g tot wat veel mensen denken, is het een bewuste keuze van mij geweest en ben ik er niet per ongeluk terechtgekomen. ledereen die niet begrijpt "waar die lui in het Bestuurscentrum de hele dag mee bezig zijn" wil ik uitnodigen om een keer bij mij langs te komen, ik leg het graag een keer uit. Daarmee zeg ik niet dat iedereen moet weten wat er centraal allemaal gebeurt want wat wij doen daar gaat het nietom . Onderzoek en onderwijs, daar gaat het binnen Wageningen UR om en ook centraal wordt dat heus wel begrepen.

Maar het schrijven moest doorgaan en het moge duidelijk zijn dat dit boekje er niet had gelegen zonder de steun van een aantal mensen.

Lieve pap en mam, bedankt voor al jullie steun en hulp. Mam, bedankt voor het oppassen, mede namens Robin. Ik weet dat ze het altijd beregezellig vindt als het weer oma-dag is. Even voor alle

99 Nawoord duidelijkheid, je hoeft menie t meer achter m'n vodden aant e zitten ensame n te spannen metEver t want het is KLAAR! Pap, iedereen is jaloers op zo'n handige vader als jij. Nog bedankt voor die prachtige badkamer. Heb je de bouwtekening voor de erker, de uitbouw achter end edakkapelle n al klaar? Niet stiekem beginnen alswi je rnie t zijn!

Lieve Willem, ikwee t datj e hier op zit te wachten dus een speciale alinea voor jou: alspartne r van een Aio zitj e inee nlasti g pakket. Eigenlijk wil je er alles aandoe n om het proces te bespoedigen maar aanmoedigingen in de trend van "moet je eigenlijk niet aan je proefschrift gaan werken?" worden je meestal niet indan k afgenomen. Maar alsj e daarentegen alles aan het eigen gezonde verstand van de promovenda overlaat en wijselijk je mond houdt, loop je het risico datj e "gebrek aan belangstelling" wordt verweten. Kortom,j edoe t het niet snel goed! Op de een ofander e manier hebben wij het ertoc h zonder kleerscheuren vanaf gebracht en dat isvoora l aan jou tedanken . Geen onvertogen woord is er overj e lippen gekomen. "Behind every strongwoma n there hast ob ea stron gman. "

Lieve familieleden, vrienden envriendinnen , bedankt voor aljulli e support. Ik zal jullie nog hard nodig hebben want na 14me i hebi k zoveel tijd over dati k memisschie n wel gavervelen . Dusko m gerust langs,d ekoffi e staat klaar.

Nu heb ikno g een stap voor de boeg:d e verdediging van mijn proefschrift op 14 mei. Gelukkig wordi k bijgestaan door twee paranimfen die zich nietge k laten makene n die ikbi jdez e de taak geef om ervoor te zorgen dati k mijn hoofd ookkoe l houdt. Richard, als broer enzu shebbe n we heel wat robbertjes geknokt maar we hebben ookhee l watlo l gehad enda thebbe n weno g steeds (the famous Horsman- humor he). Marjan,zo' n ervaren paranimfal sji j sleept mije rvas t wel doorheen. Op naar de eindstreep!

O/Vtv^

100 CURRICULUM VITAE

Karin Horsman werd op 14 november 1966 in Arnhem geboren als dochter van Jan Theunis Horsman en Kitty Aalders. Vanaf 1979 bezocht zij het VWO in Zevenaar alwaar zij in 1985 haar diploma haalde. Inda t zelfde jaar startte zij haar studie Plantenveredeling aan de Landbouwhogeschool te Wageningen. Haar afstudeerrichting waren somatische celgenetica en virologie. Haar stage in de tomatenveredeling bracht ze door in Israel bij het veredelingsbedrijf Hazera Seeds in Kiriat Gat. In 1991 studeerde ze af aan de Landbouwuniversiteit Wageningen waarna ze parttime schermdocente werd bij dezelfde instelling. In 1992 werkte ze bij de afdeling Moleculaire Biologie van het CPRO-DLO aan de mapping van nematodenresistentiegenen in aardappel. In 1993 begon zij bij de vakgroep Plantenveredeling van de Landbouwuniversiteit Wageningen aan het promotieproject waarvan het resultaat hier voor u ligt in het proefschrift getiteld "Somatic hybrids of Solanum tuberosum and species of the Solarium nigrum- complex and their backcross progeny". Sinds 1999 is zij als beleidsmedewerker in dienst van Wageningen Universiteit en Research Centrum, in eerste instantie bij de afdeling Onderzoek- en Onderwijsbeleid maar sindsjanuar i 2000 bij de stafafdeling Onderzoekstrategie.

101 LIST OF PUBLICATIONS

KOORNNEEF M, BADE J, HANHART C, HORSMAN K, SCHEL J, SOPPE W, VERKERK R, ZABEL P (1993) Characterization and mapping ofa gene controlling shoot regeneration intomato . The Plant J.3 (1): 131-141.

JACOBS JME, H.J. VAN ECK HJ, HORSMAN K, ARENS PFP, VERKERK-BAKKERB , JACOBSENE , PEREIRAA , STIEKEMA WJ (1994) Mapping ofresistanc e toth e potato cyst nematode Globodera rostochiensis from thewil d potato species Solanumvemei. Molecula r Breeding 2:51-60.

HORSMAN K, JACOBSEN E(1994 ) Comparison ofgenome s ofa tuberou s and a non-tuberous Solanum species through GISH of somatic hybrids and their backcrosses. Kew Chromosome ConferenceI V Abstract book, p.77

BASTIAANSSEN HJM, HORSMAN K, JACOBSEN E, RAMANNA MS (1995) Chromosome differentiation in Solanum species. In: P.E. Brandham and M.D. Bennett (editors). Kew Chromosome Conference IV, pp.281-290 .

HORSMAN K, JACOBSEN E (1996) Analysis of. S. n/grum-potato somatic hybrids and their backcross derivatives with Genomic insitu Hybridization. In: Abstracts of Conference Papers, Posters and Demonstrations ofth e 13th Triennial Conference ofth e European Association for Potato Research, pp. 128-129

HORSMAN K, BERGERVOET M, JACOBSEN E (1997) Somatic hybridization between S. tuberosum and species of the S. n/grum-complex: selection of vigorously growing and flowering plants. Euphytica 96:345-352

Gavrilenko T, HORSMAN K, BERGERVOETM , JACOBSENE , PETER K (1998) Genomic in situ hybridization identifies chromosome composition in the somatic hybrid Solanum tuberosum (+) Solanum nigrum and it's sexual andfusio n progenies. In: Breeding research on potatoes. Proceedings of an international symposium, Gross Lusewitz, Germany, pp 49-50

HORSMAN K, FRATINI R, HUIGEN DJ, JACOBSENE (1999) Successful first and second backcrosses of S. nigrum (+) S. tuberosum somatic hybrids with both Solanum parents. Sex Plant Reprod 12: 144- 151

HORSMAN K, GAVRILENKOT , BERGERVOETM , HUIGEN DJ, JOEATW , JACOBSENE (2001) Alteration of the genomic composition ofSolanum nigrum (+) potato backcross derivatives by somatic hybridisation: selection offusio n hybrids by DNA measurements and GISH. In press

HORSMAN K, JACOBSEN E, RAMANNA MS. Prospects for the introgression of chromosomes from non- tuberous Solanum nigrum into S. tuberosum: a qualitative analysis ofth e meiosis through GISH andAFLP-analysi s of backcross derivatives. Submitted

103 The investigations described in this thesis were performed at the Laboratory of Plant Breeding, Wageningen University, and were part of the programme of the graduate school 'Experimental Plant Sciences'. The research was supported by the 'Stichting Bevordering Veredelingsonderzoek' and the Ministry of Economic Affairs (Senter).

Drukwerk: Grafisch bedrijf Ponsen & Looijen, Wageningen

104