UNIVERSIDAD DE CASTILLA LA MANCHA FACULTAD DE CIENCIAS AMBIENTALES Y BIOQUÍMICA DEPARTAMENTO DE CIENCIAS AMBIENTALES

Gene expression reprogramming during the formation of feeding sites induced by plant endoparasitic nematodes

TESIS DOCTORAL JAVIER CABRERA CHAVES (TOLEDO, 2016)

UNIVERSIDAD DE CASTILLA LA MANCHA FACULTAD DE CIENCIAS AMBIENTALES Y BIOQUÍMICA DEPARTAMENTO DE CIENCIAS AMBIENTALES ÁREA DE FISIOLOGÍA VEGETAL

Gene expression reprogramming during the formation of feeding sites induced by plant endoparasitic nematodes

Javier Cabrera Chaves

2016 UNIVERSIDAD DE CASTILLA LA MANCHA FACULTAD DE CIENCIAS AMBIENTALES Y BIOQUÍMICA DEPARTAMENTO DE CIENCIAS AMBIENTALES ÁREA DE FISIOLOGÍA VEGETAL

Gene expression reprogramming during the formation of feeding sites induced by plant endoparasitic nematodes

Memoria presentada por el licenciado Javier Cabrera Chaves para optar al grado de Doctor por la Universidad de Castilla- La Mancha.

Trabajo dirigido por la Dra. Carolina Escobar Lucas de la Universidad de Castilla- La Mancha

Toledo, 2016

Vº Bº del Director de Tesis El Doctorando

Dra. Carolina Escobar Lucas Javier Cabrera Chaves

INDEX INDEX

CONCEPTUAL GRAPH SUMMARY

CHAPTER 1: Introduction

Purpose of the chapter

Overview of Root-Knot Nematodes and Giant Cells

Developmental Pathways Mediated by Hormones in Nematode Feeding Site

The Power of Omics to Identify Plant Susceptibility Factors and to Study Resistance to Root-knot Nematodes

AIMS AND OBJECTIVES

CHAPTER 2: Holistic analyses on plant- nematode interactions

Purpose of the chapter

Distinct and conserved transcriptomic changes during nematode- induced giant cell development in tomato compared with Arabidopsis: a functional role for gene repression

Differentially expressed small RNAs in Arabidopsis galls formed by Meloidogyne javanica: a functional role for miR390 and its TAS3-derived tasiRNAs

NEMATIC: a simple and versatile tool for the in silico analysis of plant–nematode interactions

CHAPTER 3: Molecular parallelisms between lateral roots and giant cell and gall formation

Purpose of the chapter

A role for LATERAL ORGAN BOUNDARIES-DOMAIN 16 during the interaction Arabidopsis–Meloidogyne spp. provides a molecular link between lateral root and root-knot nematode feeding site development Genes co-regulated with LBD16 in nematode feeding sites inferred from in silico analysis show similarities to regulatory circuits mediated by the auxin/cytokinin balance in Arabidopsis

CHAPTER 4: A new method for the phenotyping of the giant cells

Purpose of the chapter

Phenotyping nematode feeding sites: three-dimensional reconstruction and volumetric measurements of giant cells induced by root-knot nematodes in Arabidopsis

INTENGRATED DISCUSSION

CONCLUSIONS

CONCEPTUAL GRAPH SUMMARY Chapter 1: Introduction

Chapter 2: Holistic Studies A general overview of the plant nematode interaction Chapter 3: LBD16 & Lateral Roots and the research Chapter 4: 3D Reconstruction performed in this field

Developmental pathways Holistic studies performed mediated by hormones in the susceptible and and altered during the resistant interactions with plant- nematode root- knot nematodes interactions

A method for the 3D phenotping of the giant cells. Insights into their volumes and morphological features

Transcriptomic and The transcriptomes of molecular mechanisms are early developing giant cells common between lateral from tomato and root and giant cells and arabidopsis showed a gall formation. LBD16 is a common massive and key gene in both functional gene repression processes

A tool to facilitate the selection of genes frm the data generated on trasncriptomes from plant- nematode interactions and other fields

miRNA and tasiRNA Regulation of LBD16 co- pathways for gene regulated genes differs silencing are functional in between giant cells and galls and the rasiRNAs are syncytia. Differences in highly induced. Molecular the balance parallelisms with lateral Auxins/Cytokinins between roots both feeding cells Chapter 1: Introduction

Chapter 2: Holistic Studies A general overview of the plant nematode interaction Chapter 3: LBD16 & Lateral Roots and the research Chapter 4: 3D Reconstruction performed in this field

Developmental pathways Holistic studies performed mediated by hormones in the susceptible and and altered during the resistant interactions with plant- nematode root- knot nematodes interactions

A method for the 3D phenotping of the giant cells. Insights into their volumes and morphological features

Transcriptomic and The transcriptomes of molecular mechanisms are early developing giant cells common between lateral from tomato and root and giant cells and arabidopsis showed a gall formation. LBD16 is a common massive and key gene in both functional gene repression processes

A tool to facilitate the selection of genes frm the data generated on trasncriptomes from plant- nematode interactions and other fields

miRNA and tasiRNA Regulation of LBD16 co- pathways for gene regulated genes differs silencing are functional in between giant cells and galls and the rasiRNAs are syncytia. Differences in highly induced. Molecular the balance parallelisms with lateral Auxins/Cytokinins between roots both feeding cells CHAPTER 1: Introduction Chapter 1: Introduction

Chapter 2: Holistic Studies A general overview of the plant nematode interaction Chapter 3: LBD16 & Lateral Roots and the research Chapter 4: 3D Reconstruction performed in this field

Developmental pathways Holistic studies performed mediated by hormones in the susceptible and and altered during the resistant interactions with plant- nematode root- knot nematodes interactions

A method for the 3D phenotping of the giant cells. Insights into their volumes and morphological features

Transcriptomic and The transcriptomes of molecular mechanisms are early developing giant cells common between lateral from tomato and root and giant cells and arabidopsis showed a gall formation. LBD16 is a common massive and key gene in both functional gene repression processes

A tool to facilitate the selection of genes frm the data generated on trasncriptomes from plant- nematode interactions and other fields

miRNA and tasiRNA Regulation of LBD16 co- pathways for gene regulated genes differs silencing are functional in between giant cells and galls and the rasiRNAs are syncytia. Differences in highly induced. Molecular the balance parallelisms with lateral Auxins/Cytokinins between roots both feeding cells Aim of the chapter.

This introductory chapter is composed by three different review articles with the aim of offering an insight into the current state of the art of the plant- nematode interaction research. It is mainly centred in the root-knot nematodes and particularly in molecular aspects of the interaction.

The first review article entitled “Overview of Root-Knot Nematodes and Giant Cells” discusses general aspects of the morphology, the life cycle including the reproduction of the nematodes of the genus Meloidogyne spp., as well as the impact of this plague in the agriculture worldwide. Moreover, an overview of the present knowledge on the morphological and molecular changes occurring during the development of the root- knot nematode feeding cells, the giant cells, is also shown. The article gives a hint of the research on giant cells and galls biology and development based largely on microscopy and molecular biology techniques.

This overview is followed by a more detailed description of those plant developmental pathways mediated by hormones that could be contributing to the formation of the giant cells induced by Meloidogyne spp. or to the syncytia induced by cyst nematodes. This review entitled “Developmental Pathways Mediated by Hormones in Nematode Feeding Sites” discusses the parallelisms found at the gene regulation level between the processes of the development of different plant tissues or organs such as the root apical meristem or the lateral roots and the formation of the nematode feeding sites, and how nematode secretions should interfere with hormone- regulated developmental pathways in the roots to establish their feeding sites.

Under the title of “The Power of Omics to Identify Plant Susceptibility Factors and to Study Resistance to Root-knot Nematodes (RKNs)”, the last section of this introductory chapter reviews the approaches based on omic techniques for the study of the RKNs interaction. The vast data generated in this holistic studies helped to a better understanding of the giant cells and galls development, and to the identification of putative plant susceptibility factors; some of them helpful to the development of biotechnological tools for nematode control.

Moreover, this introduction makes reference to some of the results obtained during the experimental work of this thesis, highlighting some of the main conclusions obtained. In this way, the results obtained during my PhD have been incorporated to the state of the art of the research on this field, helping to placing them in context and to understand their significance. CHAPTER ONE

Overview of Root-Knot Nematodes and Giant Cells

Carolina Escobar1,a, Marta Barcalaa, Javier Cabrera, Carmen Fenoll Laboratory of Plant Physiology, Department of Environmental Sciences, Universidad de Castilla-La Mancha, Toledo, Spain 1Corresponding author: E-mail: [email protected]

Contents 1. Introduction to Plant Parasitic Nematodes 2 2. General Aspects of Root-Knot Nematodes (RKNs) 5 3. The Morphology and Reproduction of RKNs 8 4. The Life Cycle of RKNs 12 5. Giant Cells (GCs): From Vascular Cells to Nourishing Cells 15 6. Holistic Approaches to Tackle GCs Specific Gene Expression 22 7. Conclusions 23 Acknowledgements 24 References 24

Abstract Root-knot nematodes (RKNs) are ubiquitous parasites with an amazing capacity to interact with a very large variety of plant species. They are sedentary endoparasitic nematodes that depend on the induction of a permanent feeding site in living roots to complete their life cycle. RKNs interfere with the genetic programmes of their hosts to transform root vascular cells into giant cells (GCs) through the injection of nematode effectors from their oesophageal glands. Dramatic rearrangements in GCs cytoskeleton, alteration of cell cycle mechanisms, such as mitosis and endoreduplication, readjust- ment of enzymes involved in carbohydrate synthesis and degradation are among those processes modified in GCs. GCs act as sinks to provide nutrients for life cycle comple- tion from J2 larvae to adult females. The female produces an egg offspring protected by a gelatinous matrix and the free-living stage, J2, hatch from these eggs, completing the nematode life cycle. The model species Arabidopsis thaliana allowed easy in vivo obser- vations of the interaction by video-enhanced contrast light microscopy on infected roots, and the wide range of existing genetic and molecular tools of this plant model has extended its use. Holistic approaches to tackle gene expression combined with cell biology techniques, as isolation of GCs by laser capture microdissection, allowed GC-specific transcriptomic analysis, generating vast lists of differentially expressed

aBoth authors have contributed equally to this work.

Advances in Botanical Research, Volume 73 ISSN 0065-2296 © 2015 Elsevier Ltd. http://dx.doi.org/10.1016/bs.abr.2015.01.001 All rights reserved. 1 j 2 Carolina Escobar et al.

genes. However, the design of consistent functional hypothesis about these genes and their products will require the development of in silico analysis tools for comparisons among the transcriptomes of plantenematode compatible interactions. The under- standing of the processes subjacent to GC differentiation and maintenance, as well as a deeper knowledge of RKN mode of parasitism, will provide tools for new control methods of these devastating agricultural pests.

1. INTRODUCTION TO PLANT PARASITIC NEMATODES

Nematodes are pluricellular organisms, classified within the large phylum Nematoda that encompasses unsegmented roundworms. Nematodes are widespread in almost all ecosystems and habitats throughout the planet, including different soils, marine and fresh waters. These ubiquitous organisms have proved an amazing adaptability to diverse and extreme environments from deserts to the arctic pole. They also show varied lifestyles (with representatives from free-living to parasitic species) and food resources (plants, bacteria, animals and fungi) (Perry & Moens, 2011). There are nematodes detrimental to agriculture, parasites of animal and humans, but also beneficial species, such as the entomopathogenic nematodes used in crop protection as insect control agents (Lacey & Georgis, 2012; Ravichandra, 2008), as well as free-living nematodes involved in soil nutrient turnover. So far, more than 25.000 spp. have been included in the phylum (Zhang, 2013) but this number is constantly increasing as new species are discovered or redescribed (Elling, 2013). Classic taxonomy proposed two classes, based on morphological and anatomical characters (Chromadorea and Adeno- phorea), which diverged over 550 million years ago. Recently, a more comprehensive phylogenetic classification based mainly on molecular anal- ysis of small subunit of ribosomal DNA (ssUrDNA) was proposed: Chroma- dorea and Enoplea (De Ley & Blaxter, 2002; De Ley & Blaxter, 2004; van Megen et al., 2009)(Table 1). Nematode species included within the Chro- madorea class in the suborder Tylenchina (Table 1) have an especial rele- vance due to their great economic impact on agriculture. Plant parasitic nematodes affect frequently the root system, where they produce extensive damage such as galling and necrosis. As an indirect consequence of infection, aboveground plant parts are altered, showing a reduced growth, leaf chlo- rosis, poor yield and wilting. Crop losses, are sometimes underestimated because plant symptoms after the infection are unspecific and can be errone- ously identified as resulting from nutritional deficiencies or abiotic stress. vriwo otKo eaoe n in Cells Giant and Nematodes Root-Knot of Overview

Table 1 Phylogenetic Classification of Plant Parasitic Nematodes According to De Ley and Blaxter (2002) Class Subclass Order Suborder Superfamily Family Subfamily Genus Rotylenchus Hoplolaiminae Hoplolaimidae Hoplolaimus Rotylenchulinae Rotylenchulus Cactodera Globodera Tylenchoidea Heteroderinae Chromadorea Chromadoria Rhabditida Tylenchina Heteroderidae Heterodera Punctodera Meloidogyninae Meloidogyne Pratylenchinae Pratylenchus Pratylenchidae Nacobbinae Nacobbus Criconematoidea Tylenchulidae Tylenchulinae Tylenchulus Longidorus Longidorinae Dorylaimina Dorylaimoidea Longidoridae Paralongidorus Dorylaimia Dorylaimida Enoplea Xiphineminae Xiphinema

Enoplia Enoplida 3 4 Carolina Escobar et al.

Nematodes also represent an important economic issue in some leisure busi- ness such as golf courses, turfs along the world and in ornamental crops (Crow, 2005, 2007; Crow & Luc, 2014; Rahman Khan, Khan, & Mohide, 2005). Most plant parasitic nematodes suffer four moults throughout their development from the juvenile stage (stages 1e4, J1eJ4) until reaching the adult stage. Transition from J1 to J2 usually takes place within the egg, and after this first moult the egg hatches releasing the J2, which repre- sents for the majority of the species the infective stage (Perry & Moens, 2011). J2 larvae are mostly microscopic (from 250 mm to 12 mm in length) and live in soils without feeding until they find a suitable host. Then, J2 invade and feed on living plants through a protrusible oral stylet that they use to puncture cells and to feed from them. Throughout their develop- mental stages, nematodes usually maintain a vermiform, worm-like shape. However, in several nematode species, such as Meloidogyne spp., Heterodera spp., Rotylenchus spp. and Tylenchulus spp., adult females adopt a swollen, pear-like or kidney-like shape (Decraemer & Hunt, 2013). Plant parasitic nematodes are classified according to their lifestyle and feeding habits. Those that penetrate the host root to feed from different inner cell types are classified as endoparasites, whereas the nematodes that feed externally by inserting their mouth stylets into root cells from the root surface are called ectoparasites. They are further subclassified into sedentary, when they have a sessile stage, or migratory (Decraemer & Hunt, 2013). Examples of genera included in all these categories are found among the major agriculturally relevant nematode species. For instance, the sedentary ectoparasite Tylenchulus semipenetrans (citrus nematode) is responsible for losses in citrus and olive trees and, to a lesser extent, grape- vines. The lance and the needle nematodes (Hoplolaimus spp. and Longidorus spp. respectively) are migratory ectoparasites that cause considerable losses in turf grasses and lawns, corn crops and grape vineyards. Migratory ectoparasitic nematodes are particularly relevant, as some act as virus vectors (e.g. Xiphinema spp., a grapevine pathogen). Among the endoparasitic nematodes, there are migratory species (e.g. Pratylenchus spp., a major prob- lem in fruit trees) and sedentary ones, which constitute a most relevant group in agriculture. Sedentary endoparasitic nematodes show the most sophisticated parasitism behaviour; they develop an intimate relationship within their hosts, inducing highly specialized ‘pseudo-organs’ to provide them with a continuous source of food. This group is represented by the root-knot Overview of Root-Knot Nematodes and Giant Cells 5 nematode (RKN; Meloidogyne spp.) and the cyst nematodes (e.g. Heterodera and Globodera spp.), receiving their names from the characteristic structures formed in the roots after their infection: the galls or knots and the syncytia. Recently, phylogeny methods based on ssUrDNA (van Megen et al., 2009) support the idea that the similar parasitism behaviour of root-knot and cyst nematodes has been acquired by convergent evolution between both groups rather than the existence of a common ancestor (Castagnone-Sereno, Danchin, Perfus-Barbeoch, & Abad, 2013; Castagnone-Sereno, Skantar, & Robertson, 2011; Perry & Moens, 2011). Plant damage caused by plant parasitic nematodes is mostly due to the reduced availability of water and nutrients because of nematode feeding and disturbance of root anatomy. Nematode-produced wounding also pre- dispose the plant to other soil pathogens attack, what is sometimes favoured by pathogenic bacteria or fungi carried by the nematode itself (Back, Haydock, & Jenkinson, 2002; Jones & Goto, 2011; Stanton & Stirling, 1997). For example, wilt fungus Fusarium oxysporum can interact with RKNs in complex diseases, affecting tomato, cabbage or watermelon (Bergeson, Van Gundy, & Thomason, 1970; Fassuliotis & Rau, 1969; Jenkins & Coursen, 1957; Sumner & Johnson, 1972) and Ralstonia solanacea- rum bacteria can increase tomato wilt when RKNs are present (Valdez, 1978). For cyst nematodes, complex diseases are found mainly in potato and soybean crops (Back et al., 2002). RKN species are polyphagous and can feed on almost all vascular plants tested (Jones & Goto, 2011), whereas cyst nematodes often show a more specific host preference and usually can parasitize a single plant species (e.g. Globodera spp. only infect potato). This Chapter will focus on the general biology of the RKN, while cyst nematodes will be reviewed in Chapter 2.

2. GENERAL ASPECTS OF ROOT-KNOT NEMATODES (RKNs)

Meloidogyne is a genus formed by more than 90 species (Jones et al., 2013), some of them including several races (Eisenback & Triantaphyllou, 1991; Ravichandra, 2008). Only a few species are referred as major agricultural pests, as they were considered the most abundant and damaging: Meloidogyne incognita, Meloidogyne javanica, Meloidogyne arenaria from Mediter- ranean and tropical areas, and the temperate species Meloidogyne hapla. Addi- tionally, species previously considered minor agricultural pests as Meloidogyne 6 Carolina Escobar et al. enterolobii, Meloidogyne paranaensis or Meloidogyne exigua (from tropical and subtropical regions), and Meloidogyne fallax, Meloidogyne minor or Meloidogyne chitwoodi (from temperate regions) are emergent parasites that receive increasing attention (Elling, 2013; Moens, Perry, & Starr, 2009) as they are raising as important agriculture threats. Some of them, such as M. chitwoodi, M. enterolobii or M. fallax, have been included in the 2013 quarantine pest list from the European and Mediterranean Plant Protection Organization. As previously indicated, RKNs are extremely polyphagous parasites. Meloidogyne spp. such as M. incognita, M. javanica, M. hapla, M. arenaria, M. enterolobii, M. fallax or M. chitwoodi show a broad host range, being able to parasitize vegetable crops, fruit trees and ornamental plants, whereas other species show a more restricted host range, as M. minor (grasses, potato and tomato) or Meloidogyne hispanica (peach, sugar beet, tomato). In accordance to this, species with narrower host ranges show more restricted geographical localizations, but as their host range widens, they show a global distribution (Triantaphyllou, 1985). Control strategies in agriculture cover the use of chemicals (nematicides and fumigants), biological control with nematode antagonists, physical methods, such as solarization and fallowing, cultural methods as crop rota- tion, as well as the use of resistant plants. The use of chemicals is gradually vanishing due to their toxicity and environmental contamination potential. The frequently used methyl bromide, a broad spectrum and economically viable pesticide, has been banned in the European Union since 2010 (Kearn, Ludlow, Dillon, O’Connor, & Holden-Dye, 2014) and other countries are reducing its use. Organophosphate- and carbamate-based nematicides are also restricted. Those belonging to the fluoroalkenyl thioether group are effective against RKN showing a lower impact on the environment as compared to organophosphate- and carbamate-based nematicides and new nematicides derived from biologically active compounds such as those found in garlic are being developed (Kearn et al., 2014). However, effective chemical pesticides against these complex eukaryotes will mostly be poten- tially harmful for other organisms. Biological control has resulted in a low effective strategy unless applied in combination with other techniques (Viaene, Coyne, & Kerry, 2006). The use of nematode antagonists that can be predators, parasites or pathogens such as the fungi Verticillium spp. and Fusarium spp., or the bacteria Pasteuria penetrans, is at its initial days. Despite being an ecofriendly strategy, few com- mercial products containing viable organism for biological control are avail- able (Stanton & Stirling, 1997; Timper, 2011). Overview of Root-Knot Nematodes and Giant Cells 7

Crop rotation with nonhost species or resistant cultivars has provided good results for RKN control. Despite few poor or nonhost plant species are available, cover crops as marigolds (Tagetes spp.) or perennial grasses (such as bahiagrass (Paspalum notatum) and bermudagrass (Cynodon dactylon L. Pers.)) have been effective to control populations of M. arenaria, M. hapla, M. incognita and M. javanica (Hooks, Wang, Ploeg, & McSorley, 2010; Netcher & Taylor, 1979). With regard to resistant cultivars, several genes from tomato (Mi genes; Ammiraju, Veremis, Huang, Roberts, & Kaloshian, 2003; Rossi et al., 1998; Veremis, van. Heusden, & Roberts, 1999; Yaghoobi, Kaloshian, Wen, & Williamson, 1995), prunus (Ma and RMia genes; Claverie et al., 2004; Lu, Sossey-Alaoui, Reighard, Baird, & Abbott, 1999), carrot (Mj genes; Alietal.,2014) and pepper (Me genes; Djian-Caporalino et al., 2007) have been described to confer resistance to many Meloidogyne spp. However so far only the tomato Mi-1 gene has been cloned and successfully transferred to commercial cultivars (Devran &Sog€ ut,€ 2010). Mi-1 confers resistance to three Meloidogyne spp. (M. jav- anica, M. incognita and M. arenaria), but this resistance is easily overcome when soil temperature increases (reviewed by Williamson (1998)). In addi- tion, the isolation of virulent Meloidogyne spp. populations in tomato cultivars carrying the Mi-1 gene questioned the durability of the Mi-resis- tance (Jacquet et al., 2005) and prompted the suggestion of a relationship between resistance breakdown and Mi gene dosage (Jacquet et al., 2005). Moreover, the durability of the Me gene seems to be influenced not only by allelic dosage but also by the genetic background, since other genes or quantitative trait loci may be contributing to resistance (Djian- Caporalino et al., 2014). All these strategies should be combined in an integrated pest manage- ment (IPM) plan for effective control of RKN population in the field. A detailed evaluation of the cropping systems and accurate diagnosis of RKN species must be performed for an IPM successful implementation. Differences regarding host preferences that exhibit races of a determined species (e.g. for M. incognita all 4 races described can infect tomato cv. Rutgers, whereas only races 3 and 4 can parasite cotton cv. Deltapine (Hartman & Sasser, 1985; Mahdy, 2002)) should be considered. Therefore, designing an IPM is very laborious and overall it needs to be locally designed. Consequently, there is still a clear need to deeply understand the molecular basis of the RKNeplant interaction, including the develop- ment and maintenance of the specific feeding structures induced in the plant host, galls and giant cells (GCs). This knowledge together with that of the 8 Carolina Escobar et al. nematode biology could establish an emerging creative ground to develop new tools for RKN control.

3. THE MORPHOLOGY AND REPRODUCTION OF RKNs

RKNs display a conserved basic body plan throughout their life stages, with morphological features used for species identification. Briefly, J2 outer- most body structure consists of a body wall encompassing three layers: the cuticle, the hypodermis (also known as epidermis) and the somatic muscles. The cuticle is a flexible, semipermeable exoskeleton with a noncellular, multilayer structure that is newly synthesized and secreted by the epidermis in each moult. Cuticle layers (cortical, medial and basal layer) can vary in thickness throughout the nematode life stages or can even be absent as is the case of the medial layer in adult females (Decraemer & Hunt, 2013; Eisenback, 1985). The cuticle is a collagenous matrix covered by an outer coat (epicuticle) mainly made of glycoproteins and other surface-associated proteins. This coat is probably involved in host immunity response (Decraemer & Hunt, 2013; Eisenback, 1985). The cuticle allows solute diffusion and water and gas exchange with the medium to compensate the lack of either respiratory or circulatory system. In females, cuticular morphological features of the perineum (the region surrounding the vulva and anus) are used for the perineal pattern analysis, i.e. a characteristic pattern of ridges and annulations stablish differences among RKN species. Beyond the musculature, digestive, reproductive and nervous systems are found within the RKN body. The digestive system consists of a mouth with a retractile stylet (Figure 1(A)e(C)) connected to an oesophagus (or pharynx) which ends in an intestine and a rectum. Within the oesophagus there is a median bulb or metacorpus containing a metacorporal valve (Figure 1(A)e(C)) responsible for the suction force necessary for nutrient uptake and for pumping out gland secretions coming from the dorsal and subventral glands. These glands play a main role during parasitism, including invasion, establishment and feeding site development. During the prepara- sitic stage, the predominant glands are the two subventral glands, involved in releasing cell wall-degrading enzymes such as cellulases or pectinases (Davis, Hussey, & Baum, 2004; Jaubert, Laffaire, Abad, & Rosso, 2002). However, during the parasitic stage, once the nematode establishes, the dorsal gland become more active. Morphological changes of these glands reflect their predominance during each stage, and thus the subventral glands Overview of Root-Knot Nematodes and Giant Cells 9 reach their maximum size before invasion and begin to shrink as a nematode settles. On the contrary, the dorsal gland maximum size is described for the adult female stage (Hussey & Mims, 1990). The oesophageal gland secretions (dorsal and subventral) are released in spherical granules that vary in size, composition and morphology not only depending on nematode develop- mental stage, but also depending on nematode species (Hussey & Davis, 2004). The intestine serves as storage organ where many lipid granules can be easily observed under light microscopy (Figure 1(A) and (B)). The diges- tive system ends in the rectum, with an anus at the end in females whereas in males is joined to the reproductive system to form the cloaca. In females, rectal glands opening to the anus are responsible for the secretion of the gelatinous matrix where eggs are embedded as they are deposited. In addition, the adult female body is almost filled by the gonads, a pair of tubular organs that converge in a vagina that opens to the outside by a vulva. In preparasitic J2, the reproductive system consists on a genital primordium that will develop into either ovaries or testis as soon as the J2 starts to feed. In J2, the nervous system mainly controls the somatic musculature and sensory perception through the chemoreceptor organs (amphids and sensilla at the head, and phasmids at the posterior end). A distinctive feature of the nervous system is the nerve ring, that encircles the oesophagus behind the medium bulb (Eisenback, 1985), and is the coordinating centre for the nervous system. Meloidogyne spp. usually reproduce by mitotic parthenogenesis (e.g. M. incognita, M. javanica or M. arenaria) although some species, as M. hapla (race A) or M. chitwoodi, multiplies by facultative meiotic parthenogenesis (Berg, Fester, & Taylor, 2008; Eisenback & Triantaphyllou, 1991). The female-to-male ratio is variable, though in general few males are produced and only under suboptimal conditions (e.g. insufficient nutrients, crowding or low temperature Davide & Triantaphyllou, 1967; Decker, 1989; Snyder, Opperman, & Bird, 2006). This decision is taken during the J2 parasitic stage (Triantaphyllou, 1973), but so far signals that promote this change have not been unravelled. Contrary to the adult female, the males are motile and vermiform, range from 1.100 to 2.000 mm in length and leave the host root right after the final moult (Eisenback & Triantaphyllou, 1991). Males can grow up to four times that of the J2 length (Figure 1(I)). They also display a distinctive visible feature of the reproductive system, the spicules, hook-like structures (Stanton & Stirling, 1997) to duct sperm during mating. 10

(A) (D) (E) (F)

(B) (G) (H)

(C) (I) (J) aoiaEcbre al. et Escobar Carolina Figure 1 Root-knot nematode life cycle. A schematic diagram with pictures illustrating some of the key stages during the interaction. (A) Cells Giant and Nematodes Root-Knot of Overview Photograph of a developing Meloidogyne javanica J2 inside the egg. (B) Recently hatched M. javanica J2. (C) Close-up of an M. javanica J2 anterior body. For A, B and C stylet is indicated by a black arrow, median bulb by a white arrow and lipid globules by a black arrowhead. (D) Schematic diagram of an Root-knot nematode (RKN) life cycle as a time course of the progression of the infection represented in the same root. RKNs are black-coloured for easy location. Starting at the bottom of the diagram, a J2 penetrates the root at the elongation zone, mi- grates towards the tip and it turns 180 to enter the vascular cylinder, where it induces several giant cells (GCs). By 3 days post infection (dpi), an incipient gall has formed around the nematode including the GCs. The nematode gradually grows and develops into a female while GCs and galls enlarge, and eventually the pear-shaped mature female lays an egg mass that protrudes from the root surface. (E) Mature gall of Arabidopsis thaliana with adult female posterior region exposed outside the root and laying eggs within the gelatinous matrix. (F) Enlarged adult female of M. javanica showing the typical pear-like shape. (G) Incipient gall in A. thaliana plant at 3 dpi. (H) Overview of tomato roots infected with M. javanica showing profuse galling. (I) Two infective juveniles (J2, black arrow) an adult male (black arrowhead) show M. javanica motile stages nearby an A. thaliana root tip. (J) Initial stages of an M. javanica J2 migration in an Arabidopsis root, turning at the root tip to enter the vascular cylinder. Scale bars in A, B, C represent 20 mm, 0.2 mm in E, F and I, and 0.1 mm G and J. (See colour plate) 11 12 Carolina Escobar et al.

4. THE LIFE CYCLE OF RKNs

An RKN life cycle can be completed within 20e40 days, but its length is influenced by environmental conditions such as the temperature, to a lesser extent, soil moisture, and by the host species (Ravichandra, 2008; Rohini, Ekanayaka, & Di Vito, 1986). A typical RKN life cycle (Figure 1(D)) begins with the hatched J2s (Figure 1(B)), that are attracted towards the host roots (Figure 1(I)) after sensing chemical gradients of root diffusates (Teillet et al., 2013) with their chemosensory sensilla, the amphids (Perry & Moens, 2011). So far, only CO2 has been identified as a prime long distance attractant for plant parasitic nematodes, including M. incognita (Robinson, 2002). Additional attractants are amino acids, sugars and other metabolites (Bird, 1959; Perry, 2001; Prot, 1980; Robinson, 2002). When a suitable root tip of a host is located in the soil, nematodes penetrate preferably behind the elongation zone, and migrate intercellularly down to the root tip. The reason why RKN move downwards towards the meristem in order to enter the vascular cylinder, is probably that the Caspar- ian strip at the endodermis represents an insuperable barrier to their stylet. In fact, nematodes that do not orientate correctly within this region of the root are unable to induce a feeding site and eventually leave the root (von Mende, 1997; Wyss, Grundler, & Munch,€ 1992). The precise signals J2 might sense to orientate themselves once inside the root and to move towards the root tip are unknown. So far only CO2 has been proved to be an attractant (Robinson, 2002), but it also has been suggested that pH gradients and even electric fields could guide them (Bird, 1959; von Mende, 1997). Molecular determinants in the plant cell surfaces (linked to cell walls or mobile in the apoplast) may as well be perceived by the migrating J2 to identify their pathway towards the root tip. Penetration and migration are accomplished by a combination of chemical and mechanical tools. Nematodes secrete a mixture of cell wall-degrading enzymes and use their stylet and head to push and separate the softened mid lamella that cements the root cells together (Perry & Moens, 2011). The secretion of cell wall- degrading enzymes to the apoplast for host invasion is a feature common to other plant endoparasitic nematodes during migration and also to pathogenic bacteria and fungi (Perry & Moens, 2011). RKNs secrete cellu- lases (endoglucanases), endoxylanases, pectatelyases and polygalacturonases produced by their subventral glands (Davis, Haegeman, & Kikuchi, 2011; Perry & Moens, 2011; Wieczorek et al., 2014). Phylogenetic studies suggest Overview of Root-Knot Nematodes and Giant Cells 13 that plant parasitic nematodes acquired this capacity by ancient horizontal gene transfer from bacteria (Perry & Moens, 2011). It has also been suggested that acquisition of prokaryotic genes from the glycosyl hydrolase family 5 by sedentary endoparasitic nematodes could have occurred by pass on by a relative ancestor rather than by new horizontal gene transfer (Kyndt, Haegeman, & Gheysen, 2008; Rybarczyk-Myd1owska et al., 2012). Once the J2 reach the root tip, they rotate 180 (Figure 1(D) and (J))to enter the vascular cylinder and move upwards until near the differentiation zone where they select several vascular cells to induce the formation of a feeding site (Bird, Opperman, & Williamson, 2009; Perry & Moens, 2011). Upon feeding site development, the J2 becomes sedentary. The selected cells (usually five to eight cells) start to undergo dramatic morphological and metabolic changes, to become nutrient sinks. The most obvious morpho- logical characteristic is their enlargement, and due to this feature, they were named by Treub (1886) GCs (Figures 1(D) and 2(A)). Additionally, cortex cells surrounding the GCs divide and become hypertrophied and the pericycle cells proliferate (Figure 2(A); Berg et al., 2008). The xylem in the vicinity is grossly disrupted and GCs are encaged by a newly devel- oped intricate xylem network (Bartlem, Jones, & Hammes, 2014; Christie, 1936). Around GCs also protophloem is formed and proliferates dramati- cally (Absmanner, Stadler, & Hammes, 2013). Thus, the result is the forma- tion of a unique pseudo-organ called gall containing the GCs (Figure 1(D), (G) and (H)). However, some Meloidogyne spp. have been described to cause small or no galling (e.g. Meloidogyne artiellia, M. chitwoodi, M. fallax, M. minor or M. paranaensis; Elling, 2013; Vovlas et al., 2005) in particular hosts, showing a limited hyperplasia and proliferation of surrounding tissues. Meloi- dogyne kikuyensis develops a different gall that resembles the nodules induced by rhizobium. This gall is located on one side of the root, and the GCs are encaged within the so-called feeding socket (Eisenback, Dodge, & Odge, 2012). A detailed record of the first stages of parasitism was reported by using video-enhanced contrast light microscopy (Wyss et al., 1992). The ability of Meloidogyne spp. to parasite Arabidopsis thaliana (Sijmons, Grundler, Von Mende, Burrows, & Wyss, 1991) allowed a real progress in the understand- ing of the hostepathogen interaction. This was not only because Arabidop- sis have thin, translucent roots that permit a direct observation of initials stages of parasitism inside the plant. In the last 10 years, scientists made important contributions to the knowledge of the molecular basis of the 14 Carolina Escobar et al.

(A) (B)

(C)

Figure 2 Morphology of giant cells induced by root-knot nematodes. Semi-thin sec- tions of Meloidogyne javanica induced galls stained with toluidine. (A) Cross-section of a 3 days post infection (dpi) Arabidopsis thaliana gall showing giant cells (GCs) (*). Scale bar represents 50 mm. (B) Nicotiana tabacum longitudinal cross-section of a 7 dpi gall showing a partial view of three multinucleate GCs with ameboid nuclei (pur- ple stain) and prominent nucleoli (black arrows) scale bar represents 20 mm. (C) 360 rotation views of a 3D reconstruction image of a fully developed GC system from an Arabidopsis gall 7 dpi with M. javanica. Note the irregular shape and ragged surface of the eight GCs that comprise this individual feeding site. (See colour plate) planteendoparasitic nematodes interaction by using other advantages provided by this simple model plant such as having a small genome, being easy to transform and with multiple genetic, functional, transcriptomic, pro- teomics tools already developed (reviewed in Gheysen and Fenoll (2011)). Very recently, a novel technique enabling nondestructive, long-term obser- vations of live nematodes in planta based on the nematode fluorescent label with the lipid analogue PKH26, allowed to observe their behaviour, development, and morphology for the full duration of the parasite’s life cycle by confocal microscopy (Dinh, Brown, & Elling, 2014). Inside the gall, the nematode becomes sessile by atrophy of their somatic musculature, except for the head, and will alternate periods of feeding from the different GCs, having three consecutive moults (to J3, J4 and adult fe- male). Only J2 will feed and after the last moult, the adult females resume Overview of Root-Knot Nematodes and Giant Cells 15 feeding (Lewis & Perez, 2004). Neither J3 nor J4 have a functional stylet and hence they do not feed (Manzanilla-Lopez & Bridge, 2004). Under favourable conditions and sufficient nutrients, J4 suffers the final moult to the female adult stage. By this time, the adult females have adopted the typical pear-like shape (Figure 1(D)e(F)), have enlarged over 500 times the J2 volume (Shepperson & Jordan, 1974), and begin to deposit hundreds of eggs containing the larvae in a gelatinous matrix of glycoproteinaceous nature that they secrete (Figure 1(D) and (E); Sharon & Spiegel, 1993). This matrix is a barrier to water loss (Wallace, 1968) and provides a protec- tion to developing larvae from external pathogenic agents like bacteria and fungi. In Meloidogyne spp., the egg mass is exposed outside the root, due to the enlargement of the female, whose posterior body portion can protrude outside the gall (Figure 1(D) and (E)), making eggs more accessible to the rhizosphere microorganisms. Antimicrobial activity has been described for the gelatinous matrix (Orion, Kritzman, Meyer, Erbe, & Chitwood, 2001; Sharon & Spiegel, 1993), but if this matrix represents anything else apart from a physical barrier still must be elucidated. Anyhow, the egg mass does not represent a resistant form as in the case of the cyst nematodes. J2 usually hatches from the egg after its complete development to start a new life cycle (Figure 1(B)). Transition from J1 to J2 occurs inside the egg within the egg mass (Figure 1(A)).

5. GIANT CELLS (GCs): FROM VASCULAR CELLS TO NOURISHING CELLS

RKNs were described as plant pathogens from late 1880s (reviewed in Berg et al. (2008)). Initial research described their morphology and it is not until mid-1900s when the first studies focused on the nematode-induced plant morphological changes (Christie, 1936; Ravichandra, 2008). More detailed morphological features of the feeding cells induced in the plant hosts were already described in the 1960s by light and electron microscopy analysis (Bird, 1961; Huang & Maggenti, 1969). Nowadays, it still results a challenge to elucidate those cell processes involved in the dramatic morpho- logical and physiological changes induced in the initial root cells transformed into a specialized structure for the nematode feeding, the GCs. During this process, the first evidence of a developing GC inside the root vascular cylinder is the appearance of binucleate cells near the nematode head (de Almeida Engler & Favery, 2011). Subsequently, new mitotic cycles with 16 Carolina Escobar et al. uncoupled cytokinesis will lead to the multinucleate status of the GCs (Figure 2(A) and (B)). According to the former histological description, mitosis promotion was confirmed by the specific expression of genes encod- ing mitotic cyclins and the corresponding cyclin-dependent kinases (CDKs) involved in transitions through S-G2-M phases. Some of them are AtCYCB1;1, AtCYCA1;2, AtCDKB1;1, AtCDKA;1 and several D-type cyclin-coding genes, (de Almeida Engler et al., 1999; Barcala et al., 2010; Niebel et al., 1996). A clear increase in DNA content has been also confirmed (de Almeida Engler & Gheysen, 2013), probably due to repeated endoreduplication cycles, although other unconventional ways of DNA amplification (e.g. defective mitoses or nuclear fusion) are also suggested. These processes might help GC expansion. Yet, one of the characteristic features of GCs is those repeated cycles of mitosis, as they do not take place within syncytia. However, endoreduplication occurs in both feeding structures, as it does mitosis in the adjacent cell layers (see Chapter 4 for details on cell cycle). Finally, the nuclei of GCs are large, with irregular lobed shape and with large conspicuous nucleoli (Figure 2(B), Berg et al., 2008; Christie, 1936). As previously indicated, some Meloidogyne spp. are capable of inducing feeding sites with little galling. These feeding sites have been studied in detail and revealed fewer but larger nuclei (Vovlas et al., 2005), what could support the idea that increased DNA content might be necessary for GC expansion. So far GC precursor cells have been described as vascular cells. From his- tological observations, parenchymatic cells within the stele that surround the nematode head are generally accepted as their initial cells after being trig- gered by oesophageal gland secretions (Berg et al., 2008). However, the pre- cise cell type chosen by the nematode and GC ontogeny is still unclear. Accordingly, global transcriptomic similarities were encountered between early developing GCs dpi and suspension cells differentiating into xylem el- ements (Barcala et al., 2010). More recently, Cabrera, Diaz-Manzano, et al. (2014) described the crucial role of a transcription factor from the Lateral Organ Boundaries Domain family, LBD16, during GC development and confirmed the importance of the pericycle, a root meristematic tissue, during gall ontogeny similarly to that of lateral root formation. LBD16 is a molec- ular transducer integrated in a signalling cascade mediated by auxins for lateral root and gall formation. These findings strongly suggest that nema- todes might alter pre-existing developmental pathways in the precursor cells of GCs, probably interfering with transduction cascades modulated by hor- mones, such as auxins or cytokinins (see further discussion in Chapter 7). Overview of Root-Knot Nematodes and Giant Cells 17

Kostoff and Kendall (1930) already suggested a putative role of nematode secretions during GC development. These secretions, that contain nematode effectors, are currently the focus of numerous studies and are addressed in several chapters in this volume. Transcriptomic and proteomic studies of iso- lated nematode glands confirmed the presence of putative secreted proteins and peptides with a possible function during invasion but also during the feeding site formation (reviewed in Quentin, Abad, & Favery (2013); Rosso & Grenier (2011)). The availability of whole genome sequence for M. incog- nita and M. hapla (Abad et al., 2008; Opperman et al., 2008, further details in Chapter 10) allowed in silico searches of putative effectors, but no clear pic- ture of how the effectors are synthesized and secreted is available as yet (Berg et al., 2008; Mitchum et al., 2013; Quentin et al., 2013). Some effectors have been localized inside the feeding cells (Mj-NULG1a (Lin et al., 2013), Mi-EFF1 (Jaouannet et al., 2012; Zhang, Davies, & Elling, 2015)) whereas others locate in the apoplast (Vieira et al., 2011). In addition, comparisons among different nematode groups have revealed that some effectors are com- mon to phytonematodes and others are lifestyle specific(Tucker & Yang, 2013), supporting the convergent parasitism style theory. Nematode subven- tral glands are more active during the preparasitic stage, and the effectors secreted in this stage are to assist during migration. In contrast, once the nem- atode becomes sedentary the dorsal gland is more active; and it has a main role in feeding site development and maintenance (Quentin et al., 2013). These effectors target host cellular processes such as cell cycle, transport or hormone signalling pathways, by mimicking or interfering with host regula- tors (Mitchum et al., 2013; Tucker & Yang, 2013). So far, few effectors have been functionally assessed by plant-mediated RNAi assays to attenuate nem- atode parasitism (reviewed in Dinh et al. (2014); Elling & Jones (2014) reviewed in chapters “Function of Root Knot Nematode Effectors and Their Targets in Plant Parasitism” and “Application of Biotechnology for Nema- tode Control in Crop Plants”). Identification of the plant targets of nematode effectors is also crucial to understand the plant regulatory networks that nem- atodes perturb for feeding site development and maintenance. Thus, effectors exhibit an enormous potential to develop biotechnological based strategies for the nematode control (see Chapter “Function of Root Knot Nematode Effectors and Their Targets in Plant Parasitism” in this book). The GC becomes a typical highly metabolically active cell with a dense cytoplasm containing abundant organelles (endoplasmic reticulum (ER), ribosomes, mitochondria or Golgi bodies) (Berg et al., 2008; Christie, 1936). The large central vacuole is also fragmented into smaller ones 18 Carolina Escobar et al.

(Figure 2(A)) and chloroplast-like structures with starch accumulation are observed (Ji et al., 2013). Chaperones that may assist protein folding in cells with high metabolic activity, such as small heat shock proteins, are induced (Barcala et al., 2008). GCs constitute a sink of nutrients for the developing nematode and therefore, the metabolism of carbohydrates and amino acids is highly activated in these cells (Baldacci-Cresp et al., 2012; Gautam & Poddar, 2014; Machado et al., 2012). Sensitive metabolomics techniques recently confirmed that galls induced by M. incognita in Medicago truncatula present an elevated content of starch, sucrose, glucose, malate, fumarate and diverse amino acids (Phe, Tyr, Val, Glu, Asp) (Baldacci-Cresp et al., 2012). Similar results were found for roots of coffee and bitter gourd infected with M. exigua and M. incognita, respectively (Gautam & Poddar, 2014; Machado et al., 2012). Accordingly, the regulation of sugar, amino acid, þ water and Ca2 membrane transporters is altered in the nematode feeding sites (NFS), as many transporters were differentially expressed in galls (Barcala et al., 2010; Hammes et al., 2005; Marella et al., 2013). Some genes coding amino acid transporters as AtCAT6, upregulated upon M. incognita infection, showed no evident functional role (Hammes, Nielsen, Honaas, Taylor, & Schachtman, 2006). In contrast, loss of function of other genes such as AtAAP3 and AtAAP6 impaired M. incognita infection (Marella et al., 2013). Sucrose is assumed to be the main source of carbohydrates for the nematode, and Arabidopsis mutants sus1/sus4, cinv1 and cinv1/ cinv2 for the two main enzymes that cleave the sucrose, invertases (INVs) and sucrose synthases (SUSs), showed an increased in gall formation by M. javanica (Cabello et al., 2014). In agreement with this, AtCINV2, AtSUS1, and AtSUS4 are upregulated in GCs and/or galls (Barcala et al., 2010, further information can be found in Chapter 5). As gall development progresses, the GCs keep on enlarging. Their vol- ume increase by 100 fold from 3 dpi to 40 dpi in Arabidopsis (Cabrera et al., 2015). This increase in volume probably corresponds to the stage when the adult female requires the highest nutrient supply for growth and egg produc- tion. Cells after 3D reconstruction appeared with almost no sphericity in accordance to the presence of abundant protuberances, crevices and lobules that provided an irregular shape (Figure 2(C); Cabrera et al., 2015). The irregular shape of GCs showed in Figure 2(C), as well as their enormous volume augment, should be accompanied of profound changes on the cyto- skeleton organization. RKN induce long-term changes in the organization of the cytoskeleton during GCs expansion, i.e. microtubule and actin cytoskeleton disruption and rearrangements occur. A large number of Overview of Root-Knot Nematodes and Giant Cells 19 unusual, randomly oriented actin bundles and cables were also observed (de Almeida Engler et al., 2004; Caillaud, Lecomte, et al., 2008; de Almeida Engler & Favery, 2011). Accordingly, upregulation of tubulin- and actin- coding genes was described by de Almeida Engler (2004). Other essential pro- teins involved in cytoskeleton dynamics such as formins may act as nucleating proteins stimulating the de novo polymerization of actin filaments control- ling the assembly of actin cables. Those actin cables are probably required to guide the vesicle trafficking needed for an increasing demand of plasma membrane and cell wall biogenesis (Caillaud, Abad, & Favery, 2008). Other associated proteins such as actin depolymerizing factor (ADF; AtADF2) or microtubule-associated proteins (MAPs; AtMAP65-3) are crucial for GC development and their corresponding genes were differentially regulated within GC (Caillaud, Lecomte, et al., 2008; Clement et al., 2009). Likewise, data from genome-wide transcriptomic analysis of galls and/or GCs showed upregulation of some formin-coding genes through gall growth and GC enlargement (de Almeida Engler et al., 2004; Barcala et al., 2010; Favery et al., 2004; Jammes et al., 2005), supporting the importance of the cytoskel- eton during GC formation. Cytoskeleton reorganization is also particularly relevant during the special mitotic cycles occurring in GCs. In this respect, the presence of abnormal spindles and phragmoplasts in developing GCs was reported (Caillaud, Abad, et al., 2008; Banora et al., 2011). Misaligned phragmoplasts may also interfere with cell wall formation, thus resulting in multinucleate cells and wall stub formation (de Almeida Engler et al., 2004; de Almeida Engler & Favery, 2011; Jones & Payne, 1978). The pres- ence of these wall stubs together with repeated mitotic cycles, are essential to distinguish GC differentiation from that of syncytia, also multinucleated but formed by fusion of adjacent cells, as suggested initially by Bird, 1961.Cyto- skeleton dynamics is a complex and crucial process that requires sensitive cell biology techniques combined to molecular biology for a direct observation inside the galls girth. In the future, stress should be put to unveil these early processes of cytoskeleton reshuffling in GCs. A network of microtubules positioned along to the plasma membrane lin- ing the cell wall ingrowths (CWI) formed in GCs has been described (Berg et al., 2008). The formation of these CWIs was already described by Jones & Northcote, 1972, and are believed to increase and facilitate solute transport from adjacent xylem and phloem cells (Sobczak, Fudali, & Wieczorek, 2011). CWIs are concentrated in areas facing different tissues and not only vascular elements, but are particularly abundant opposite to xylem (Berg et al., 2008; Jones & Gunning, 1976; Jones & Northcote, 1972), what is in 20 Carolina Escobar et al. accordance to the upregulation of genes encoding membrane transporters as mentioned before. The CWIs develop from previous small patches of cell wall thickenings that increase their length and thickness. However, this process is not uniform across all the GCs cell wall, resulting in an irregularly thickened cell wall. Associated to the CWI progress, stacks of Golgi and ER are found, typical of transfer cell (TC) development, where Golgi vesicles release cell wall components (Berg et al., 2008) External signals for initiating cell wall deposition and CWI development are still unknown. Syncytia and GCs differ in their ontogeny and global transcriptional signatures, but both develop CWIs to facilitate high rates of apoplastic/symplastic molecular exchange, showing TC characteristics. The presence of CWI can allow GCs to compensate the considerable decrease in the surface/volume ratio as these cells expand (Cabrera et al., 2015). Similarities between transcriptional changes observed during the early stages of nematode feeding cells (NFC) formation and those described during differentiation of TCs suggest that auxin and ethylene might be putative signals triggering TC-like morphology of NFCs (Cabrera, Barcala, Fenoll, & Escobar, 2014; Rodiuc, Vieira, Banora, & de Almeida Engler, 2014). Although still scarce, there are some other data linking TC regulatory signals to NFSs. For instance, ZmMRP-1 (Gomez, Royo, Guo, Thompson, & Hueros, 2002) encoding a primary sensor of the putative signals for TCs differentiation, is activated in galls as compared to the rest of the root (Barrero et al., 2009). Another plant cell structure specialized in intercellular transport of molecules, the plasmodesmata (PD), are relevant for cyst nematode feeding site formation, as PDs represent the starting point for cell wall dissolution and therefore cell fusion (Grundler, Sobczak, & Golinowski, 1998; Hoth & Schneidereit, 2005; Hoth, Stadler, Sauer, & Hammes, 2008; Jones & Payne, 1978). With regard to RKNs, recent research conducted by Hofmann, Youssef-Banora, de Almeida-Engler, and Grundler (2010) reported symplastic connection between GCs and phloem, whose function- ality may vary depending on GC developmental stage or even on host species (Grundler & Hofmann, 2011). In contrast, previous studies accepted that GCs were symplastically isolated (Jones & Dropkin, 1976) despite the existence of PD (Hoth et al., 2008). Further research is needed to clarify this important topic regarding GCs PD connections. Sedentary plant parasitic nematodes produce within the feeding cell cytoplasm one of the most striking and so far poorly characterized key structures for nutrient withdrawal from the GCs and successful para- sitism: the feeding tube (FT). FTs were first identified unequivocally by Overview of Root-Knot Nematodes and Giant Cells 21

(Rahman Razak & Evans, 1976), although they had been highlighted in the initial studies of Nemec (1911, 1932) as proteinaceous threads. In elec- tron microscopy analysis, FTs are cylinders with an electron translucent lumen, connected to the stylet orifice in one side and blind at the distal end. They are described in different groups of sedentary endoparasites (Berg et al., 2008; Hussey & Mims, 1990; Rebois, 1980; Rumpenhorst, 1984; Sobczak, Golinowski, & Grundler, 1999); and even in the migratory ectoparasite Trichodorus similis (Wyss, Jank-Ladwig, & Lehmann, 1979). However, their structure may be genus specific (reviewed in Berg et al. (2008)). Meloidogyne incognita FTs showed a crystalline structure (Hussey & Mims, 1990; Nemec, 1932) suggesting a proteinaceous nature, whereas FTs from Rotylenchulus reinformis and Heterodera schachtii do not show such a regular structure (Rebois, 1980; Sobczak et al., 1999). In addition, electron energy loss spectroscopy analysis provided further support of the FT puta- tive protein composition, revealing a high content of nitrogen and sulphur (reviewed in Berg et al. (2008)). It is accepted that FTs are formed by a rapid reaction of nematode secretions with unknown components of feeding site cytoplasm. However there are controversial hypothesis regarding whether FT composition is solely from nematode or plant material or a combination of both (Berg et al., 2008). FTs are formed each time the nematode pierces the cell wall for feeding and is then discarded; so, several FTs are encoun- tered within a particular GC (Jones & Goto, 2011). During FT formation, the cell membrane seems to remain nearly intact, although it is unclear whether a small choke is open up, and only a small callose deposition has been reported at the cell wall disruption point, similarly to the feeding plug described for cyst nematodes. Functional FTs are intimately related to the endomembrane system, especially to ER, and this association has been described for FT from other genera, suggesting a relevant role for either FTs formation or nematode feeding (Berg et al., 2008). Whenever the stylet is retracted, the FT is abandoned, and no more endomembrane system can be observed in its proximity, thus suggesting that the nematode might provide a signal for active FT to recruit ER and other endomem- brane complexes for their formation. FTs are thought to serve as molecular sieves during nutrient withdraw to avoid blockage of the stylet by large particles (proteins or even organelles) as the nematode pumps away the cell content. They may also serve to discriminate specific cell components, fine-tuning the composition of the nematode diet. To date, there is no indication that nematodes use FTs to inject secretions in the feeding cells. Several studies have been conducted to elucidate the maximum size of 22 Carolina Escobar et al. solutes that can be uptaken through cyst nematodes derived-FT pores by using fluorescent molecules (green fluorescent protein (GFP) (Goverse et al., 1998; Urwin, Moller, Lilley, McPherson, & Atkinson, 1997), mono- meric red fluorescent protein (mRFP) (Valentine et al., 2007) or dextrans (Bockenhoff€ & Grundler, 1994)), but still contentious data have not allowed proposing a clear explanation. This issue has been recently approached by Eves-van den Akker et al. (2014). They remarked the struc- tural differences between RKN and cyst nematodes FTs, as uniform discrete pores were formed in GCs and heteroporous in syncytia. To date, the specific composition, mechanisms of formation and action and detailed functions of FTs await for further insight.

6. HOLISTIC APPROACHES TO TACKLE GCs SPECIFIC GENE EXPRESSION

During gall and GC ontogeny a profound reprogrammation of gene expression takes place, as encountered in transcriptomic analysis such as microarray (Barcala et al., 2010; Jammes et al., 2005; Portillo et al., 2009) and massive sequencing (Ji et al., 2013; Cabrera et al., unpublished). Precise single cell isolation techniques as microaspiration or laser capture microdis- section combined to global transcriptomic analysis constituted a step forward to the understanding of the specific transcriptomic signatures of GCs (Barcala et al., 2010; Fosu-Nyarko, Jones, & Wang, 2009; Portillo et al., 2013; Ramsay, Wang, & Jones, 2004; Wang, Potter, & Jones, 2003; Ji et al., 2013). It allowed bypassing the complexity of the gall transcriptome that included all the different tissues present in this pseudo-organ, and to stablish differences between whole gall and GC-specific transcriptomes. Recently, RNA-sequencing approaches for miRNA differential expression analysis increased the complexity of this scenario (Hewezi, Howe, Maier, & Baum, 2008; Kyndt et al., 2012; Cabrera et al. unpublished), as miRNAs have come up as key signal molecules, controlling and regulating many cellular processes at transcriptional, post-transcriptional and translational level (Yang, Xue, & An, 2007). Those holistic approaches to gene expression generated vast lists of differentially expressed genes available in public databases and publications and valuable information of general tendencies for gene expression in the NFS. However, classifying detailed information of the regulation of particular genes or gene groups through cross-comparisons among complex Overview of Root-Knot Nematodes and Giant Cells 23 data sets, or obtaining customized gene selections through sequential comparative and filtering is not an easy task. This had limited the design of consistent functional hypothesis about genes and gene products of GCs based on holistic gene expression data. One of the first data-mining spread- sheet tool, specifically designed for comparisons among transcriptomes of plantenematode compatible interactions is NEMATIC (NEMatodee Arabidopsis Transcriptomic Interaction Compendium; Cabrera, Bustos, Favery, Fenoll, & Escobar, 2014 http://www.uclm.es/grupo/gbbmp/ english/nematic.asp). It combines available transcriptomic data for the inter- action between Arabidopsis and plant endoparasitic nematodes with data from different transcriptomic analyses regarding hormone and cell cycle regulation, development, different plant tissues, cell types and various biotic stresses, facilitating efficient in silico studies on plantenematode biology. However, there is an increasing need to develop additional user friendly in silico analysis tools that may include other plant species and biological processes.

7. CONCLUSIONS

RKNs depend on a specifically developed cell type from their initial root vascular cells to complete its life cycle. Those GCs are induced and probably maintained by nematode secretions delivered through their sty- lets. Many questions regarding GC ontogeny and functioning remain unanswered. To date, only a few players of the complex regulatory net- works taking place during GCs development have emerged, and the un- derstanding of how these organisms can interact with their hosts in such a subtle manner is fragmentary. Yet, integrative analysis of proteomics and transcriptomics together with genetics and molecular and cell biology tools are facilitating its comprehension. However, the complexity of an evolving interaction makes its analysis a challenge, i.e. feeding site cell sta- tus is continuously changing as it differentiates, controlled by nematode nutritional needs. Therefore, comparisons and inferred conclusions from the analysis of galls/GCs at selected infection points should be taken cautiously. Furthermore, valuable data were also obtained from the study of nematode putative effectors and their molecular interactions to their host targets, as well as the downstream responses, pointing out common and spe- cific regulatory pathways manipulated by RKN and/or cyst nematodes. 24 Carolina Escobar et al.

ACKNOWLEDGEMENTS This work was supported by the Spanish Government (AGL2010-17388 and AGL2013- 48787-R to C. Escobar, and CSD2007-057 and PCIN-2013-053 to C. Fenoll). J.Cabrera was supported by fellowships from the Ministry of Economy and Competitiveness, Spain.

REFERENCES Abad, P., Gouzy, J., Aury, J.-M., Castagnone-Sereno, P., Danchin, E. G. J., Deleury, E., et al. (2008). Genome sequence of the metazoan plant-parasitic nematode Meloidogyne incognita. Nature Biotechnology, 26, 909e915. Absmanner, B., Stadler, R., & Hammes, U. Z. (2013). Phloem development in nematode- induced feeding sites: the implications of auxin and cytokinin. Frontiers in Plant Science, 4, 241. Ali, A., Matthews, W. C., Cavagnaro, P. F., Iorizzo, M., Roberts, P. A., & Simon, P. W. (2014). Inheritance and mapping of Mj-2, a new source of root-knot nematode (Meloi- dogyne javanica) resistance in carrot. Journal of Heredity, 105, 288e291. de Almeida Engler, J., De Vleesschauwer, V., Burssens, S., Celenza, J. L., Jr., Inze, D., Van Montagu, M., et al. (1999). Molecular markers and cell cycle inhibitors show the impor- tance of cell cycle progression in nematode-induced galls and syncytia. The Plant Cell, 11, 793e808. de Almeida Engler, J., & Favery, B. (2011). The plant cytoskeleton remodelling in nematode induced feeding sites. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant-nematode interactions (pp. 369e393). Springer Netherlands. de Almeida Engler, J., & Gheysen, G. (2013). Nematode-induced endoreduplication in plant host cells: why and how? Molecular Plant-Microbe Interactions, 26,17e24. de Almeida Engler, J., Van Poucke, K., Karimi, M., De Groodt, R., Gheysen, G., Engler, G., et al. (2004). Dynamic cytoskeleton rearrangements in giant cells and syncytia of nema- tode-infected roots. Plant Journal, 38,12e26. Ammiraju, J. S., Veremis, J. C., Huang, X., Roberts, P. A., & Kaloshian, I. (2003). The heat-stable root-knot nematode resistance gene Mi-9 from Lycopersicon peruvianum is localized on the short arm of chromosome 6. Theoretical and Applied Genetics, 106, 478e484. Back, M. A., Haydock, P. P. J., & Jenkinson, P. (2002). Disease complexes involving plant parasitic nematodes and soilborne pathogens. Plant Pathology, 51, 683e697. Baldacci-Cresp, F., Chang, C., Maucourt, M. l. M., Deborde, C., Hopkins, J., Lecomte, P., et al. (2012). (Homo)glutathione deficiency impairs root-knot nematode development in Medicago truncatula. PLoS Pathogens, 8, e1002471. Banora, M. Y., Rodiuc, N., Baldacci-Cresp, F., Smertenko, A., Bleve-Zacheo, T., Mellilo, M. T., et al. (2011). Feeding cells induced by phytoparasitic nematodes require gamma-tubulin ring complex for microtubule reorganization. PLoS Pathogens, 7, e1002343. Barcala, M., García, A., Cabrera, J., Casson, S., Lindsey, K., Favery, B., et al. (2010). Early transcriptomic events in microdissected Arabidopsis nematode-induced giant cells. Plant Journal, 61, 698e712. Barcala, M., Garcia, A., Cubas, P., Almoguera, C., Jordano, J., Fenoll, C., et al. (2008). Distinct heat-shock element arrangements that mediate the heat shock, but not the late-embryogenesis induction of small heat-shock proteins, correlate with promoter activation in root-knot nematode feeding cells. Plant Molecular Biology, 66, 151e164. Barrero, C., Royo, J., Grijota-Martinez, C., Faye, C., Paul, W., Sanz, S., et al. (2009). The promoter of ZmMRP-1, a maize transfer cell-specific transcriptional activator, is induced at solute exchange surfaces and responds to transport demands. Planta, 229, 235e247. Overview of Root-Knot Nematodes and Giant Cells 25

Bartlem, D. G., Jones, M. G. K., & Hammes, U. Z. (2014). Vascularization and nutrient delivery at root-knot nematode feeding sites in host roots. Journal of Experimental Botany, 65, 1789e1798. Bergeson, G. B., Van Gundy, S. D., & Thomason, I. J. (1970). Effect of Meloidogyne javanica on rhizosphere microflora and Fusarium wilt of tomato. Phytopathology, 69, 1245e1249. Berg, R. H., Fester, T., & Taylor, C. G. (2008). Development of the root-knot nematode feeding cell. In R. H. Berg, & C. G. Taylor (Eds.), Cell biology of plant nematode parasitism (pp. 115e152). Berlin: Springer Berlin Heidelberg. Bird, A. F. (1959). The attractiveness of roots to the plant parasitic nematodes Meloidogyne javanica and M. hapla. Nematologica, 4(4), 322e335. Bird, A. F. (1961). The ultrastructure and histochemistry of a nematode-induced giant cell. Journal of Biophysical and Biochemical Cytology, 11, 701e715. Bird, D., Opperman, C., & Williamson, V. (2009). Plant infection by root-knot nematode. In R. H. Berg, & C. G. Taylor (Eds.), Cell biology of plant nematode parasitism (pp. 1e13). Springer Berlin Heidelberg. Bockenhoff,€ A., & Grundler, F. M. W. (1994). Studies on the nutrient uptake by the beet cyst nematode Heterodera schachtii by in situ microinjection of fluorescent probes into the feeding structures in Arabidopsis thaliana. Parasitology, 109, 249e255. Cabello, S., Lorenz, C., Crespo, S., Cabrera, J., Ludwig, R., Escobar, C., et al. (2014). Altered sucrose synthase and invertase expression affects the local and systemic sugar metabolism of nematode-infected Arabidopsis thaliana plants. Journal of Experimental Botany, 65, 201e212. Cabrera, J., Barcala, M., Fenoll, C., & Escobar, C. (2014). Transcriptomic signatures of trans- fer cells in early developing nematode feeding cells of Arabidopsis focused on auxin and ethylene signaling. Frontiers in Plant Science, 5, 107. Cabrera, J., Bustos, R., Favery, B., Fenoll, C., & Escobar, C. (2014). NEMATIC: a simple and versatile tool for the in silico analysis of plant-nematode interactions. Molecular Plant Pathology, 15, 627e636. Cabrera, J., Díaz-Manzano, F. E., Barcala, M., de Almeida Engler, J., Engler, G., Fenoll, C., et al. (2015). Phenotyping nematode feeding sites: three dimensional reconstruction and volumetric measurements of giant cells induced by root-knot nematodes in Arabidopsis. New Phytologist. http://dx.doi.org/10.1111/nph.13249. Cabrera, J., Diaz-Manzano, F. E., Sanchez, M., Rosso, M.-N., Melillo, T., Goh, T., et al. (2014). A role for lateral organ boundaries-domain 16 during the interaction Arabidopsis-Meloidogyne spp. provides a molecular link between lateral root and root- knot nematode feeding site development. New Phytologist, 203, 632e645. Caillaud, M. C., Abad, P., & Favery, B. (2008). Cytoskeleton reorganization, a key process in root-knot nematode-induced giant cell ontogenesis. Plant Signaling & Behavior, 3, 816e818. Caillaud, M. C., Lecomte, P., Jammes, F., Quentin, M. M., Pagnotta, S., Andrio, E., et al. (2008). MAP65e3 microtubule-associated protein is essential for nematode-induced gi- ant cell ontogenesis in Arabidopsis. The Plant Cell, 20, 423e437. Castagnone-Sereno, P., Danchin, E. G., Perfus-Barbeoch, L., & Abad, P. (2013). Diversity and evolution of root-knot nematodes, genus Meloidogyne: new insights from the genomic era. Annual Review of Phytopathology, 51, 203e220. Castagnone-Sereno, P., Skantar, A., & Robertson, L. (2011). Molecular tools for diagnostics. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant- nematode interactions (pp. 443e464). Springer Netherlands. Christie, J. R. (1936). The development of root-knot nematode galls. Phytopathology, 26, 1e22. Claverie, M., Bosselut, N., Lecouls, A. C., Voisin, R., Lafargue, B., Poizat, C., et al. (2004). Location of independent root-knot nematode resistance genes in plum and peach. Theoretical and Applied Genetics, 108, 765e773. 26 Carolina Escobar et al.

Clement, M., Ketelaar, T., Rodiuc, N., Banora, M. Y., Smertenko, A., Engler, G., et al. (2009). Actin-depolymerizing factor2-mediated actin dynamics are essential for root- knot nematode infection of Arabidopsis. The Plant cell, 21, 2963e2979. Crow, W. T. (2005). How bad are nematode problems on Florida’s golf courses? Florida Turf Digest, 22,10e12. Crow, W. T. (2007). Understanding and managing parasitic nematodes on turfgrasses. In Handbook of turfgrass management & physiology (pp. 351e374). Boca, Raton: CRC Press. Crow, W. T., & Luc, J. E. (2014). Field efficacy of furfural as a nematicide on turf. Journal of Nematology, 46,8e11. Davide, R. G., & Triantaphyllou, A. C. (1967). Influence of the environment on develop- ment and sex differentiation of root-knot nematodes. Nematologica, 13, 111e117. Davis, E. L., Haegeman, A., & Kikuchi, T. (2011). Degradation of the plant cell wall by nematodes. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant-nematode interactions (pp. 255e272). Springer Netherlands. Davis, E. L., Hussey, R. S., & Baum, T. J. (2004). Getting to the roots of parasitism by nematodes. Trends in Parasitology, 20, 134e141. De Ley, P., & Blaxter, M. (2002). Systematic position & phylogeny. In D. L. Lee (Ed.), The biology of nematodes (pp. 1e30). London: Taylor & Francis. De Ley, P., & Blaxter, M. (2004). A new system for Nematoda: combining morphological characters with molecular trees, and translating clades into ranks and taxa. In R. Cook, & D. J. Hunt (Eds.), Nematology monographs and perspectives (pp. 633e653). Leiden: E.J. Brill. Decker, H. (1989). Plant nematodes and their control (phytonematology). Brill. Decraemer, W., & Hunt, D. J. (2013). Structure and classification. In R. N. Perry, & M. Moens (Eds.), Plant Nematology (pp. 3e39). Wallingford, UK: CABI publ. Devran, Z., & Sog€ ut,€ M. A. (2010). Occurrence of virulent root-knot nematode populations on tomatoes bearing the Mi gene in protected vegetable-growing areas of Turkey. Phytoparasitica, 38, 245e251. Dinh, P. T., Brown, C. R., & Elling, A. A. (2014). RNA Interference of effector gene Mc16D10L confers resistance against Meloidogyne chitwoodi in arabidopsis and potato. Phytopathology, 104, 1098e1106. Djian-Caporalino, C., Fazari, A., Arguel, M. J., Vernie, T., VandeCasteele, C., Faure, I., et al. (2007). Root-knot nematode (Meloidogyne spp.) Me resistance genes in pepper (Capsicum annuum L.) are clustered on the P9 chromosome. Theoretical and Applied Genetics, 114, 473e486. Djian-Caporalino, C., Palloix, A., Fazari, A., Marteu, N., Barbary, A., Abad, P., et al. (2014). Pyramiding, alternating or mixing: comparative performances of deployment strategies of nematode resistance genes to promote plant resistance efficiency and durability. BMC Plant Biology, 14, 53. Eisenback, J. D. (1985). Detailed morphology and anatomy of second-stage juveniles, males, and females of the genus Meloidogyne (root-knot nematode). In J. N. Sasser, & C. C. Carter (Eds.), An advanced treatise on Meloidogyne,Vol.1(pp. 47e78). Raleigh: North Carolina State University Graphics. Eisenback, J. D. D., Dodge, D. J., & Odge, D. J. D. (2012). Description of a unique, complex feeding socket caused by the putative primitive root-knot nematode, Meloidogyne kikuyensis. Journal of Nematology, 44, 148e152. Eisenback, J. D., & Triantaphyllou, H. H. (1991). Root-knot nematodes: Meloidogyne species and races. In W. R. Nickle (Ed.), Manual of agricultural nematology (pp. 191e274). New York: Marcel Dekker. Elling, A. A. (2013). Major emerging problems with minor Meloidogyne species. Phytopa- thology, 103, 1092e1102. Elling, A. A., & Jones, J. T. (2014). Functional characterization of nematode effectors in plants. Methods in Molecular Biology, 1127, 113e124. Overview of Root-Knot Nematodes and Giant Cells 27

Eves-van den Akker, S., Lilley, C. J., Ault, J. R., Ashcroft, A. E., Jones, J. T., & Urwin, P. E. (2014). The feeding tube of cyst nematodes: characterisation of protein exclusion. PLoS One, 9, e87289. Fassuliotis, G., & Rau, G. J. (1969). The relationship of Meloidogyne incognita acrita to the inci- dence of cabbage yellows. Journal of Nematology, 1, 219e222. Favery, B., Chelysheva, L. A., Lebris, M., Jammes, F., Marmagne, A., de Almeida-Engler, J., et al. (2004). Arabidopsis formin AtFH6 is a plasma membraneeassociated protein upre- gulated in giant cells induced by parasitic nematodes. The Plant Cell, 16, 2529e2540. Fosu-Nyarko, J., Jones, M. G. K., & Wang, Z. (2009). Functional characterization of tran- scripts expressed in early-stage Meloidogyne javanica-induced giant cells isolated by laser microdissection. Molecular Plant Pathology, 10, 237e248. Gautam, S. K., & Poddar, A. N. (2014). Study on protein and sugar content in Meloidogyne incognita infested roots of bitter gourd. International Journal of Current Microbiology and Applied Sciences, 3, 470e478. Gheysen, G., & Fenoll, C. (2011). Arabidopsis as a tool for the study of plant-nematode interactions. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant-nematode interactions (pp. 139e156). Springer Netherlands. Gomez, E., Royo, J., Guo, Y., Thompson, R., & Hueros, G. (2002). Establishment of cereal endosperm expression domains: identification and properties of a maize transfer cell- specific transcription factor, ZmMRP-1. Plant Cell, 14, 599e610. Goverse, A., Biesheuvel, J., Wijers, G. J., Gommers, F. J., Bakker, J., Schots, A., et al. (1998). In planta monitoring of the activity of two constitutive promoters, CaMV 35S and TR20, in developing feeding cells induced by Globodera rostochiensis using green fluorescent protein in combination with confocal laser scanning microscopy. Physiological and Molec- ular Plant Pathology, 52, 275e284. Grundler, F. W., & Hofmann, J. (2011). Water and nutrient transport in nematode feeding sites. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant- nematode interactions (pp. 423e439). Springer Netherlands. Grundler, F. M. W., Sobczak, M., & Golinowski, W. (1998). Formation of wall openings in root cells of Arabidopsis thaliana following infection by the plant-parasitic nematode Heterodera schachtii. European Journal of Plant Pathology, 104, 545e551. Hammes, U. Z., Nielsen, E., Honaas, L. A., Taylor, C. G., & Schachtman, D. P. (2006). AtCAT6, a sink-tissue-localized transporter for essential amino acids in Arabidopsis. Plant Journal, 48, 414e426. Hammes, U. Z., Schachtman, D. P., Berg, R. H., Nielsen, E., Koch, W., McIntyre, L. M., et al. (2005). Nematode-induced changes of transporter gene expression in Arabidopsis roots. Molecular Plant-Microbe Interactions, 18, 1247e1257. Hartman, K. M., & Sasser, J. N. (1985). Identification of Meloidogyne species on the basis of differential host test and perineal pattern morphology. In K. R. Barker, C. C. Carter, & J. N. Sasser (Eds.), An advanced treatise on Meloidogyne, Vol. 2 (pp. 69e78). Raleigh: North Carolina State University Graphics. Hewezi, T., Howe, P., Maier, T. R., & Baum, T. J. (2008). Arabidopsis small RNAs and their targets during cyst nematode parasitism. Molecular Plant-Microbe Interactions, 21,1622e1634. Hofmann, J., Youssef-Banora, M., de Almeida-Engler, J., & Grundler, F. M. W. (2010). The role of callose deposition along plasmodesmata in nematode feeding sites. Molecular Plant- Microbe Interactions, 23, 549e557. Hooks, C. R. R., Wang, K.-H., Ploeg, A., & McSorley, R. (2010). Using marigold (Tagetes spp.) as a cover crop to protect crops from plant-parasitic nematodes. Applied Soil Ecology, 46, 307e320. Hoth, S., & Schneidereit, A. (2005). Nematode infection triggers the de novo formation of unloading phloem that allows macromolecular trafficking of green fluorescent protein into syncytia. Plant Physiology, 138, 383e392. 28 Carolina Escobar et al.

Hoth, S., Stadler, R., Sauer, N., & Hammes, U. Z. (2008). Differential vascularization of nematode-induced feeding sites. Proceedings of the National Academy of Sciences of the United States of America, 105, 12617e12622. Huang, C. S., & Maggenti, A. R. (1969). Wall modifications in developing giant cells of Vicia faba and Cucumis sativus induced by root knot nematode, Meloidogyne javanica. Phytopa- thology, 59, 931e937. Hussey, R. S., & Davis, E. L. (2004). Nematode esophageal glands and plant parasitism. In Z. X. Chen, S. Y. A. Chen, & D. W. Dickson (Eds.), Nematology advances and perspectives, volume I. Nematode morphology, physiology and ecology (pp. 258e293). Wallingford, UK: CABI publ. Hussey, R. S., & Mims, C. W. (1990). Ultrastructure of esophageal glands and their secretory granules in the root-knot nematode Meloidogyne incognita. Protoplasma, 156,9e18. Jacquet, M., Bongiovanni, M., Martinez, M., Verschave, P., Wajnberg, E., & Castagnone- Sereno, P. (2005). Variation in resistance to the root-knot nematode Meloidogyne incognita in tomato genotypes bearing the Mi gene. Plant Pathology, 54,93e99. Jammes, F., Lecomte, P., de Almeida-Engler, J., Bitton, F., Martin-Magniette, M. L., Renou, J. P., et al. (2005). Genome-wide expression profiling of the host response to root-knot nematode infection in Arabidopsis. Plant Journal, 44, 447e458. Jaouannet, M., Perfus-Barbeoch, L., Deleury, E., Magliano, M., Engler, G., Vieira, P., et al. (2012). A root-knot nematode-secreted protein is injected into giant cells and targeted to the nuclei. New Phytologist, 194, 924e931. Jaubert, S., Laffaire, J. B., Abad, P., & Rosso, M. N. (2002). A polygalacturonase of animal origin isolated from the root-knot nematode Meloidogyne incognita. FEBS Letters, 522, 109e112. Jenkins, W. R., & Coursen, B. W. (1957). The effect of root-knot nematodes, Meloidogyne incognita acrita and M. hapla, on Fusarium wilt of tomato. Plant Disease Report, 182e186. Ji, H., Gheysen, G., Denil, S., Lindsey, K., Topping, J. F., Nahar, K., et al. (2013). Transcrip- tional analysis through RNA sequencing of giant cells induced by Meloidogyne graminicola in rice roots. Journal of Experimental Botany, 64, 3885e3898. Jones, M. G. K., & Dropkin, V. H. (1976). Scanning electron microscopy of nematode- induced giant transfer cells. Cytobios, 15, 149e161. Jones, M. G. K., & Goto, D. B. (2011). Root-knot nematodes and giant cells. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant-nematode interactions (pp. 83e100). Springer Netherlands. Jones, M. G. K., & Gunning, B. E. S. (1976). Transfer cells and nematode induced giant cells in Helianthemum. Protoplasma, 87, 273e279. Jones, J. T., Haegeman, A., Danchin, E. G. J., Gaur, H. S., Helder, J., Jones, M. G. K., et al. (2013). Top 10 plant-parasitic nematodes in molecular plant pathology. Molecular Plant Pathology, 14, 946e961. Jones, M. G. K., & Northcote, D. H. (1972). Multinucleate transfer cells induced in coleus roots by the root-knot nematode, Meloidogyne arenaria. Protoplasma, 75, 381e395. Jones, M. G., & Payne, H. L. (1978). Early stages of nematode-induced giant-cell formation in roots of Impatiens balsamina. Journal of Nematology, 10,70e84. Kearn, J., Ludlow, E., Dillon, J., O’Connor, V., & Holden-Dye, L. (2014). Fluensulfone is a nematicide with a mode of action distinct from anticholinesterases and macrocyclic lactones. Pesticide Biochemistry and Physiology, 109,44e57. Kostoff, D., & Kendall, J. (1930). Cytology of nematode galls on Nicotiana roots. Zentralbl Bakteriol Parasitenk, 81,86e91. Kyndt, T., Denil, S., Haegeman, A., Trooskens, G., Bauters, L., Van Criekinge, W., et al. (2012). Transcriptional reprogramming by root knot and migratory nematode infection in rice. New Phytologist, 196, 887e900. Overview of Root-Knot Nematodes and Giant Cells 29

Kyndt, T., Haegeman, A., & Gheysen, G. (2008). Evolution of GHF5 endoglucanase gene structure in plant-parasitic nematodes: no evidence for an early domain shuffling event. BMC Evolutionary Biology, 8, 305. Lacey, L. A., & Georgis, R. (2012). Entomopathogenic nematodes for control of insect pests above and below ground with comments on commercial production. Journal of Nema- tology, 44. Lewis, E. E., & Perez, E. E. (2004). Aging and developmental behavior. In G. Randy, & A. L. Bilgrami (Eds.), Nematode behaviour (pp. 151e176). Lin, B., Zhuo, K., Wu, P., Cui, R., Zhang, L-h., & Liao, J. (2013). A novel effector protein, MJ-NULG1a, targeted to giant cell nuclei plays a role in Meloidogyne javanica parasitism. Molecular Plant-Microbe Interactions, 26,55e66. Lu, Z. X., Sossey-Alaoui, K., Reighard, G. L., Baird, W. V., & Abbott, A. G. (1999). Devel- opment and characterization of a codominant marker linked to root-knot nematode resistance, and its application to peach rootstock breeding. Theoretical and Applied Genetics, 99, 115e122. Machado, A., Campos, V., da Silva, W. R., Campos, V., Zeri, A., & Oliveira, D. (2012). Metabolic profiling in the roots of coffee plants exposed to the coffee root-knot nema- tode, Meloidogyne exigua. European Journal of Plant Pathology, 134, 431e441. Mahdy, M. (2002). Biological control of plant parasitic nematodes with antagonistic bacteria on different host plants. In Hohen Landwirtschaftlichen Fakult€at. Bonn: Rheinischen Friedrich-Wilhelms-Universit€at. Manzanilla-Lopez, R. H. E. K., & Bridge, J. (2004). In Z. X. Chen, S. Y. Chen, & D. W. Dickson (Eds.), Plant diseases caused by nematodes (pp. 636e716). Wallingford, UK: CABI Publishing. Marella, H. H., Nielsen, E., Schachtman, D. P., Taylor, C. G., Danforth, D., Science, P., et al. (2013). The amino acid permeases AAP3 and AAP6 are involved in root-knot nem- atode parasitism of Arabidopsis. Molecular Plant-Microbe Interactions, 26,44e54. van Megen, H., van den Elsen, S., Holterman, M., Karssen, G., Mooyman, P., Bongers, T., et al. (2009). A phylogenetic tree of nematodes based on about 1200 full-length small subunit ribosomal DNA sequences. Nematology, 11, 927e950. von Mende, N. (1997). Invasion and migration behaviour of sedentary nematodes. In C. Fenoll, F. M. W. Grundler, & S. A. Ohl (Eds.), Cellular and molecular aspects of plant-nematode interactions (pp. 51e64). Springer Netherlands. Mitchum, M. G., Hussey, R. S., Baum, T. J., Wang, X., Elling, A. A., Wubben, M., et al. (2013). Nematode effector proteins: an emerging paradigm of parasitism. New Phytologist, 199, 879e894. Moens, M., Perry, R., & Starr, J. (2009). Meloidogyne speciesea diverse group of novel and important plant parasites. In R. N. Perry, M. Moens, & J. L. Starr (Eds.), root-knot nematodes (p. 483). Wallingford, UK: CABI publ. Nemec, B. (1911). Uber€ die Nematodenkrankheiten der Zuckerrube.€ Zeitschrift f€ur P Anzenkrankheiten, 21,1e10. Nemec, B. (1932). Uber€ die Galen von Heterodera schachtii auf der Zuckerrube.€ Studies from the Plant Physiological Laboratory of Charles University Prague, 4,1e14. Netcher, C., & Taylor, D. P. (1979). Physiological variation with the genus Meloidogyne and its implications on integrated control. In Root-knot nematodes (Meloidogyne species), systematics, biology and control (pp. 269e294). London: Acad. Press. London. Niebel,A.,DeAlmeidaEngler,J.,Hemerly,A.,Ferreira,P.,Inzé,D.,VanMontagu,M., et al. (1996). Induction of cdc2a and cyc1At expression in Arabidopsis thaliana during early phases of nematode-induced feeding cell formation. Plant Journal, 10, 1037e1043. Opperman, C. H., Bird, D. M., Williamson, V. M., Rokhsar, D. S., Burke, M., Cohn, J., et al. (2008). Sequence and genetic map of Meloidogyne hapla: a compact nematode 30 Carolina Escobar et al.

genome for plant parasitism. Proceedings of the National Academy of Sciences of the United States of America, 105, 14802e14807. Orion, D., Kritzman, G., Meyer, S. l. F., Erbe, E. F., & Chitwood, D. J. (2001). A role of the gelatinous matrix in the resistance of root-knot nematode (Meloidogyne spp.) eggs to microorganisms. Journal of Nematology, 33, 203e207. Perry, R., & Moens, M. (2011). Introduction to plant-parasitic nematodes; modes of parasitism. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant-nematode interactions (pp. 3e20). Springer Netherlands. Perry, R. N. (2001). An evaluation of types of attractants enabling plant-parasitic nematodes to locate plant roots. Russian Journal of Nematology, 13,83e88. Portillo, M., Cabrera, J., Lindsey, K., Topping, J., Andres, M. F., Emiliozzi, M., et al. (2013). Distinct and conserved transcriptomic changes during nematode-induced giant cell development in tomato compared with Arabidopsis: a functional role for gene repression. New Phytologist, 197, 1276e1290. Portillo, M., Lindsey, K., Casson, S., García-Casado, G., Solano, R., Fenoll, C., et al. (2009). Isolation of RNA from laser-capture-microdissected giant cells at early differentiation stages suitable for differential transcriptome analysis. Molecular Plant Pathology, 10, 523e535. Prot, J. C. (1980). Migration of plant-parasitic nematodes towards roots. Revue de Nématologie, 3, 305e318. Quentin, M., Abad, P., & Favery, B. (2013). Plant parasitic nematode effectors target host defence and nuclear functions to establish feeding cells. Frontiers in Plant Science, 4. Rahman Khan, M., Khan, S. M., & Mohide, F. (2005). Root-knot nematode problem of some winter ornamental plants and its biomanagement. Journal of Nematology, 37,198e206. Rahman Razak, A., & Evans, A. A. F. (1976). An intracellular tube associated with feeding by Rotylenchulus reniformis on cowpea root. Nematologica, 22, 182e189. Ramsay, K., Wang, Z., & Jones, M. G. K. (2004). Using laser capture microdissection to study gene expression in early stages of giant cells induced by root-knot nematodes. Molecular Plant Pathology, 5, 587e592. Ravichandra, N. G. (2008). Plant nematology. New Delhi, India: I. K. International Pvt Ltd. Rebois, R. V. (1980). Ultrastructure of a feeding peg and tube associated with Rotylenchulus reniformis in cotton. Nematologica, 26, 396e405. Robinson, A. (2002). Host finding by plant-parasitic nematodes. In E. E. Lewis, J. Campbell, & M. Sukhdeo (Eds.), The behavioral ecology of parasites (pp. 89e110). Wallingford, UK: CABI publ. Rodiuc, N., Vieira, P., Banora, M. Y., & de Almeida Engler, J. (2014). On the track of transfer cell formation by specialized plant-parasitic nematodes. Frontiers in Plant Science, 5,160. Rohini, K., Ekanayaka, H. M., & Di Vito, M. (1986). Life cycle and multiplication of Meloi- dogyne incognita on tomato and eggplant seedlings. Tropical Agriculturist, 142. Rossi, M., Goggin, F. L., Milligan, S. B., Kaloshian, I., Ullman, D. E., & Williamson, V. M. (1998). The nematode resistance gene Mi of tomato confers resistance against the potato aphid. Proceedings of the National Academy of Sciences of the United States of America, 95, 9750e9754. Rosso, M. N., & Grenier, E. (2011). Other nematode effectors and evolutionary constraints. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant- nematode interactions (pp. 287e307). Springer Netherlands. Rumpenhorst, H. J. (1984). Intracellular feeding tubes associated with sedentary plant parasitic nematodes. Nematologica, 30,77e85. Rybarczyk-Myd1owska, K., Maboreke, H. R., van Megen, H., van den Elsen, S., Mooyman, P., Smant, G., et al. (2012). Rather than by direct acquisition via lateral gene transfer, GHF5 cellulases were passed on from early Pratylenchidae to root-knot and cyst nematodes. BMC Evolutionary Biology, 12, 221. Overview of Root-Knot Nematodes and Giant Cells 31

Sharon, E., & Spiegel, Y. (1993). Glycoprotein characterization of the gelatinous matrix in the rooteknot nematode Meloidogyne javanica. Journal of Nematology, 25, 585e589. Shepperson, J. R., & Jordan, W. C. (1974). Observations on in vitro survival and develop- ment of Meloidogyne. Proceedings of the Helminthological Society of Washington, 41, 254. Sijmons, P. C., Grundler, F. M. W., Von Mende, N., Burrows, P. R., & Wyss, U. (1991). Arabidopsis thaliana as a new model host for plant-parasitic nematodes. Plant Journal, 1, 245e254. Snyder, D. W., Opperman, C. H., & Bird, D. M. (2006). A method for generating Meloido- gyne incognita males. Journal of Nematology, 38, 192e194. Sobczak, M., Fudali, S., & Wieczorek, K. (2011). Cell wall modifications induced by nematodes. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant-nematode interactions (pp. 395e422). Springer Netherlands. Sobczak, M., Golinowski, W., & Grundler, F. (1999). Ultrastructure of feeding plugs and feeding tubes formed by Heterodera schachtii. Nematology, 363e374. Stanton, J. M., & Stirling, G. (1997). Nematodes as plant parasites. In J. F. Browm, & H. J. Ogle (Eds.), Plant pathogens and plant diseases (pp. 127e141). Armidale, Australia: University of New England. Sumner, D. R., & Johnson, A. W. (1972). The effect of nematodes and crop sequence Fusarium wilt of watermelon. Phytopathology, 62, 791. Teillet, A., Dybal, K., Kerry, B. R., Miller, A. J., Curtis, R. H. C., & Hedden, P. (2013). Transcriptional changes of the root-knot nematode Meloidogyne incognita in response to Arabidopsis thaliana root signals. PLoS One, 8, e61259. Timper, P. (2011). Utilization of biological control for managing plant-parasitic nematodes. In K. Davies, & Y. Spiegel (Eds.), Progress in biological control, Biological control of plant-parasitic nematodes: Building Coherence between microbial ecology and molecular mech- anisms (pp. 259e289). Springer Netherlands. Treub, M. (1886). Quelques mots sure les effets du parasitisme de l’Heterodera javanica dans les racines de la canne a sucre. Annales du Jardin botanique de Buitenzorg, 6,93e96. Triantaphyllou, A. C. (1973). Environmental sex differentiation of nematodes in relation to pest management. Annual Review of Phytopathology, 11, 441e462. Triantaphyllou, A. C. (1985). Cytogenetics,cytotaxonomy and phylogeny of root-knot nematodes. In J. N. Sasser, & C. C. Carter (Eds.), An advanced teatrise on meloidogyne. Vol. I, biology and control (pp. 113e126). Raleigh, USA: North Carolina State University Graphics. Tucker, M. L., & Yang, R. (2013). A gene encoding a peptide with similarity to the plant IDA signaling peptide (AtIDA) is expressed most abundantly in the root-knot nematode (Meloidogyne incognita) soon after root infection. Experimental Parasitology, 134,165e170. Urwin, P. E., Moller, S. G., Lilley, C. J., McPherson, M. J., & Atkinson, H. J. (1997). Continual green-fluorescent protein monitoring of cauliflower mosaic virus 35S promoter activity in nematode-induced feeding cells in Arabidopsis thaliana. Molecular Plant-Microbe Interactions, 10, 394e400. Valdez, R. B. (1978). Nematodes attacking tomato and their control. In First international symposium on tropical tomato (pp. 136e152). AVRDC publ. Valentine, T. A., Randall, E., Wypijewski, K., Chapman, S., Jones, J., & Oparka, K. J. (2007). Delivery of macromolecules to plant parasitic nematodes using a tobacco rattle virus vector. Plant Biotechnology Journal, 5, 827e834. Veremis, J. C., van Heusden, A. W., & Roberts, P. A. (1999). Mapping a novel heat-stable resistance to Meloidogyne in Lycopersicon peruvianum. Theoretical and Applied Genetics, 98, 274e280. Viaene, N., Coyne, D. L., & Kerry, B. R. (2006). Biological and cultural management. In R. Perry, & M. Moens (Eds.), Plant nematology. Wallingford, UK: CABI. 32 Carolina Escobar et al.

Vieira, P., Danchin, E. G. J., Neveu, C., Crozat, C., Jaubert, S., Hussey, R. S., et al. (2011). The plant apoplasm is an important recipient compartment for nematode secreted proteins. Journal of Experimental Botany, 62, 1241e1253. Vovlas, N., Rapoport, H. F., Díaz, R. M. J., Castillo, P., Nematologia, B., Nazionale, C., et al. (2005). Differences in feeding sites induced by root-knot nematodes, Meloidogyne spp., in Chickpea. Phytopathology, 95, 368e375. Wallace, H. R. (1968). The influence of soil moisture on survival and hatch of Meloidogyne javanica. Nematologica, 14, 231e242. Wang, Z., Potter, R. H., & Jones, M. G. K. (2003). Differential display analysis of gene expression in the cytoplasm of giant cells induced in tomato roots by Meloidogyne javanica. Molecular Plant Pathology, 4, 361e371. Wieczorek, K., Elashry, A., Quentin, M., Grundler, F. M. W., Favery, B., Seifert, G. J., et al. (2014). A distinct role of pectate lyases in the formation of feeding structures induced by cyst and root-knot nematodes. Molecular Plant-Microbe Interactions, 27, 901e912. Williamson, V. M. (1998). Root-Knot nematode resistance genes in tomato and their poten- tial for future use. Annual Review of Phytopathology, 36, 277e293. Wyss, U., Grundler, F. M. W., & Munch,€ A. (1992). The parasitic behavior of second- stage juveniles in Meloidogyne incognita in roots of Arabidopsis thaliana. Nematologica, 38, 98e111. Wyss, U., Jank-Ladwig, R., & Lehmann, H. (1979). On the formation and ultrastructure of feeding tubes produced by Trichodorid nematodes. Nematologica, 25, 385e390. Yaghoobi, J., Kaloshian, I., Wen, Y., & Williamson, V. M. (1995). Mapping a new nematode resistance locus in Lycopersicon peruvianum. Theoretical and Applied Genetics, 91, 457e464. Yang, T., Xue, L., & An, L. (2007). Functional diversity of miRNA in plants. Plant Science, 172, 423e432. Zhang, Z. Q. (2013). Animal biodiversity: an update of classification and diversity in 2013. Zootaxa, 3703,5e11. Zhang, L., Davies, L. J., & Elling, A. A. (2015). A Meloidogyne incognita effector is imported into the nucleus and exhibits transcriptional activation activity in planta. Molecular Plant Pathology, 16(1), 48e60. CHAPTER SEVEN

Developmental Pathways Mediated by Hormones in Nematode Feeding Sites

Javier Cabrera, Fernando E. Díaz-Manzano, Carmen Fenoll, Carolina Escobar1 Laboratory of Plant Physiology, Department of Environmental Sciences, Universidad de Castilla-La Mancha, Toledo, Spain 1Corresponding author: E-mail: [email protected]

Contents 1. Introduction 168 2. Nematode Peptide Hormones as Interceptors of Plant Development to Form 171 Feeding Sites 3. Auxins, Lateral Root Formation and Feeding Sites 175 4. Giant Cell Morphogenesis and Transfer Cell Nature 179 Acknowledgements 181 References 182

Abstract Sedentary plant endoparasitic (root-knot and cyst) nematodes induce the formation of their feeding sites by directing the transdifferentiation of normal plant root cells into nem- atode feeding cells, namely giant cells (GCs) and syncytia. In the past years, transcriptomic analyses combined with molecular cell biology have revealed dramatic and specific changes in gene expression in syncytia and GCs. Among the genes whose expression is modified to establish feeding sites are those involved in hormone-regulated develop- mental pathways in the roots, particularly those related to auxins and cytokinins. The high concentrations of auxins and cytokinins in galls and syncytia have been described in detail by the use of reporter genes driven by specific promoters as ‘sensors’ of both phytohor- mones, such as DR5, ARR5 or TCS. Moreover, several molecular evidences link the forma- tion of nematode feeding sites (NFSs) to developmental processes such as maintenance of the root apical meristem, lateral root initiation or vascular tissue development, in which the two hormones are involved. The mechanisms that nematodes use to interfere with plant developmental pathways are unclear, but some seem to involve nematode secreted molecules, such as the CLE-like and the CEP peptides. Only in a few cases, plant hormone transduction and developmental circuits hijacked by nematodes to induce and maintain feeding sites have been studied in detail. Analysis combining hormone genetic sensors, mutants and comparative transcriptomics lead to the identification of

Advances in Botanical Research, Volume 73 ISSN 0065-2296 © 2015 Elsevier Ltd. http://dx.doi.org/10.1016/bs.abr.2014.12.005 All rights reserved. 167 j 168 Javier Cabrera et al.

relevant plant regulators that are exploited for NFS differentiation. We present the cur- rent knowledge connecting the hormonal-controlled developmental processes of the root with the development of the NFS, which seem to be different for GCs and syncytia. For instance, LBD16 and WRKY23, two key transcription factors in the signal transduction leading to lateral root formation mediated by auxins, play distinctive roles during gall/GC and syncytia formation, respectively. However, the expression of either gene in the feeding site is not strictly plant auxin-dependent, indicating that their regulation by nematodes differs in some aspects from the endogenous pathways operating in normal root development. We also highlight the evidences linking gall and GC ontogeny to the pericycle and discuss the transfer cell-like identity of feeding cells.

1. INTRODUCTION

Plant endoparasitic nematodes induce the formation of sophisticated feeding structures inside the root that operate as physiological sinks to supply nutrients to the nematode (Perry & Moens, 2011). Among the most damaging groups of plant parasitic nematodes are the root-knot nematodes (RKNs; Meloidogyne spp.) and the cyst nematodes (CNs; Heterodera spp. and Globodera spp.), representing major threats to agriculture (Bird et al., 2009; Moens, Perry, & Starr, 2009). One of the most remarkable changes regarding cell morphogenesis directly induced by nematode effectors (see Chapters 11, 12 and 13) is the formation of giant cells (GCs) by RKNs and syncytia by CNs. They are cells specifically induced by nematodes to sustain their feeding and their obligated development inside the plant. Together, both nematode groups are able to infect almost all species of agri- cultural crops, as RKNs show a polyphagous behaviour (Moens et al., 2009), suggesting that the nematodes interfere with biological processes shared by most plant species in order to develop their feeding sites. One possibility is that they may ‘hijack’, at least partially, fundamental mechanisms of plant development, necessary for the survival and appropriate plant performance, as described for molecular transducers common to lateral root (LR) and gall formation (Cabrera, Díaz-Manzano, et al., 2014). Understanding nematode feeding site (NFS) formation based on a deep knowledge of the develop- mental processes occurring in a noninfected root was proposed before the blast of the omics (Scheres et al., 1997). In the age of trancriptomics, when hundreds of genes have been identified as differentially expressed during the process of plant–nematode interaction, it becomes a prerequisite to con- nect those molecular evidences to the signalling cascades mediating develop- mental processes in a noninfected root. Several evidences show that plant Developmental Pathways Altered in Nematode Feeding Sites 169 parasitic nematodes develop their feeding sites through modulation or inter- loping of those developmental mechanisms present in the plant. This chapter summarizes the advances in this topic. The molecular mechanisms used by microorganisms to interfere with plant processes are surprisingly subtle but can effectively modify predefined plant developmental patterns. A recent example is the phytoplasma virulence protein SAP54 that promotes the degradation of flowering regulatory proteins, generating a short circuit in a developmental process that transforms flowers into leaves, helping attrac- tiveness to leaf-hopper vectors for phytoplasma reproduction and propaga- tion (MacLean et al., 2014). Similarly, effector molecules secreted by the nematodes seem to interfere with developmental pathways, although still there is a lack of clear evidence about the particular transduction cascades modified or perturbed by nematode effectors (this chapter and Chapters 11 and 12; Lee et al., 2011). In order to modify the molecular pathways used by the plants to develop their basic functional structures, it is reasonable to presume that the nema- tode interferes with the upstream hormonal control of these particular trans- duction cascades. In this respect, early experiments already pointed to the importance of auxins and cytokinins in the development of NFSs induced within the root, as increased concentrations of both phytohormones or their precursors were detected in galls induced by RKNs (Balasubramanian & Rangaswami, 1962; Krupasagar & Barker, 1969). More recently, mass spec- trometric analysis confirmed the presence of auxins and cytokinins in the secretions of Heterodera schachtii and Meloidogyne incognita (De Meutter et al., 2003, 2005), which suggests a role for both hormones during NFS development. Moreover, the use of ‘hormone-sensor systems’ based on re- porter genes evidences the activation of both auxin and cytokinin signalling pathways in the formation/maintenance of GCs, galls and syncytia. Early ex- periments showing the activation of the auxin-responsive promoter GH3 in the galls formed by RKNs in white clover (Hutangura, Mathesius, Jones, & Rolfe, 1999) have been further confirmed with the use of the synthetic auxin responsive promoter DR5. It showed a clear activation in the GCs, galls and syncytia induced by plant parasitic nematodes in Arabidopsis either with b-glucuronidase or Green Fluorescent Protein at early (Grunewald, Cannoot, Friml, & Gheysen, 2009; Karczmarek, Overmars, Helder, & Gov- erse, 2004) and late NFS developmental stages (Absmanner, Stadler, & Hammes, 2013; Cabrera, Díaz-Manzano, et al., 2014). The activation of the DR5-based sensor occurs very early during GC formation, as brief incu- bation with the GUS substrate in conditions where diffusion is minimized 170 Javier Cabrera et al. shows a strong signal specifically within the GCs (Figure 1(A)). However, the signal was also present in the surrounding vascular cells after longer in- cubation times, suggesting that, not surprisingly, auxins are also present in the vascular cell layers surrounding the GCs (Figure 1(B) and (C); Cabrera, Díaz-Manzano, et al., 2014). Similarly, the activation of cytokinin-regulated genes has been demonstrated by using the responsive promoter ARR5, induced during early stages of RKN establishment (Lohar et al., 2004), and the synthetic cytokinin-responsive promoter TCS, that is induced in syncytia formed by H. schachtii in Arabidopsis but not in galls (Absmanner et al., 2013). Thereby, the overexpression in Lotus japonica of CKX, the enzyme that catalyzes the degradation of cytokinins, resulted in a reduction in the infection by RKNs (Lohar et al., 2004). The auxin–cytokinin crosstalk is considered as the main hormonal con- trol system regulating the developmental processes occurring in the roots, such as, the root apical meristem (RAM) maintenance, LR emergence or vascular tissue development (Bielach, Duclercq, Marhavý, & Benkova, 2012; Bishopp, Help, et al., 2011; Dello Ioio et al., 2007). There are several molecular evidences showing that both hormones act antagonistically,

(A)(B) (C)

Figure 1 Transgenic Arabidopsis line DR5::GUS showing GUS activity in 4 dpi galls induced by Meloidogyne javanica at different incubation times in the GUS staining so- lution. (A) After 2 h incubation, GUS activity is centred in giant cells. (B) After 4 h incu- bation, staining spreads to the adjacent vascular tissue. (C) With overnight incubation, the signal covers most of the gall vascular tissues. N, nematode; *, giant cells. Zoom in images of (A) and (B) is shown in the panels below. Bars: 100 mm. (See colour plate) Developmental Pathways Altered in Nematode Feeding Sites 171 i.e. auxin induces cell division in the meristems, while cytokinin stimulates the differentiation of these cells (reviewed in Bielach et al., 2012; Bishopp, Benkova, & Helariutta, 2011; Moubayidin, Di Mambro, & Sabatini, 2009; Saini, Sharma, Kaur, & Pati, 2013). Hormone signalling function in the plant–nematode interaction is not restricted to the interference with developmental pathways to establish the feeding site as they also play a key role in plant defences (review in Chapter 6). We present in this chapter the state of the art regarding our understanding of how nematodes interfere with hormone-regulated developmental path- ways in the roots to establish their feeding sites, particularly those related to auxins and cytokinins. New data regarding morphometric parameters after GC reconstruction are also discussed in relation to their acquisition of transfer cell (TC)-like nature.

2. NEMATODE PEPTIDE HORMONES AS INTERCEPTORS OF PLANT DEVELOPMENT TO FORM FEEDING SITES

One of the major evidences demonstrating that nematodes actively interfere with the programmed development of the roots came from the dis- covery of a protein, Hg-SYV46, secreted by the CN Heterodera glycines that contains a structural motif of the CLAVATA3/ESR-related (CLE) family in Arabidopsis (Wang et al., 2001, 2005). CLV3-like peptide hormones have been shown to have multiple functions in many aspects of plant develop- ment and morphogenesis (Leasure & He, 2012). The similarities between CLV3 and the nematode peptide Hg-SYV46 are not merely structural. The overexpression of Hg-SYV46 in clv3 mutant plants rescues their pheno- type, while Hg-SYV46 overexpression in wild-type plants results in strong downregulation of WUSCHEL (WUS), similar to that occurring in the plants overexpressing CLV3 (Wang et al., 2005). In Arabidopsis, membrane receptor kinases and several of the WUSCHEL-RELATED HOMEOBOX (WOX) transcription factors participate in CLE signalling (Leasure & He, 2012). While CLV3 acts in the shoot apical meristem repressing the expres- sion of WUS and regulating the shoot apical meristem stem cell number through a negative feedback loop (Schoof et al., 2000), CLE40, another CLV3-like peptide, acts in the RAM regulating the expression of WOX5 (Stahl, Wink, Ingram, & Simon, 2009) through the receptor-like kinase CRINKLY4 (ACR4). Overexpression of CLE40 alters the expression of WOX5 and promotes the differentiation of distal columella stem cells to 172 Javier Cabrera et al. columella cells (Stahl et al., 2009). Since the description of Hg-SYV46, more CLE-like genes have been described in H. glycines (Wang et al., 2010), H. schachtii (Patel et al., 2008; Wang et al., 2011) and Globodera rostochiensis (Guo, Ni, Denver, Wang, & Clark, 2011; Lu et al., 2009; see Chapters 11 and 12 for detailed explanation). Functional analysis via targeting these nem- atode CLE-like genes in plants by RNAi or by the infection of CLE knock- down plants showed a decrease in the infection rate and in the size of syncytia either in Arabidopsis or in soybean (Bakhetia, Urwin, & Atkinson, 2007; Patel et al., 2008; Guo et al., 2015). These constitute experimental evidences that confirm the role of CLE-like peptides in syncytia development (Replo- gle et al., 2011, 2013). All these data together suggest a role of the nematode peptide hormones in the development of the syncytia in Arabidopsis, possibly interfering with the aforementioned plant developmental pathways. The 16D10 gene from M. incognita encoding a secretory peptide with a CLE-like sequence also showed functional characteristics of a component of a CLE-related pathway. In vivo expression of 16D10 dsRNA in Arabidopsis resulted in an increase in the resistance against RKN (Huang, Allen, Davis, Baum, & Hussey, 2006). In addition, the overexpression of 16D10, that directly interacts with SCARECROW-like transcription factors (SCR; Huang et al., 2006), does not rescue the clv3 phenotype but stimulates root growth in Arabidopsis and tobacco, i.e. calli were formed in tips cut for subculturing (Huang et al., 2006). These results demonstrate 16D10 be- ing an effector that substantially alters plant development. SCR, expressed specifically in the endodermis and cortex/endodermis initial cells of the root (Di Laurenzio et al., 1996), is a key regulator of radial patterning in the Arabidopsis root (Levesque et al., 2006) and is also directly activated by SHORT-ROOT (SHR; Levesque et al., 2006), therefore regulating root meristem identity and root development. Strikingly, transcripts from both SCR and SHR were downregulated in isolated GCs at 3 dpi (Barcala et al., 2010); however, transcriptomes specific of GCs are not available in earlier time points when developmental switches concerning cell develop- ment are probably crucial. Thus, the 16D10 gene constitutes another example of how nematodes could interfere with hormonal controlled devel- opmental pathways of the root to generate their feeding sites. The putative role of CLE-like nematode peptides might not be merely related to the meristems development. Yet, another group of CLE peptides such as tracheary element differentiation inhibitory factor (TDIF), a peptide hormone derived from CLE41/44 (Ito et al., 2006) that induces the expres- sion of WOX4 in cambium cells (Hirakawa, Kondo, & Fukuda, 2010), Developmental Pathways Altered in Nematode Feeding Sites 173 participates in vascular development by promoting the division of cambium cells preventing their differentiation into xylem. Changes in auxin levels, such as those happening in nematode feeding cells, are necessary as the trigger signal for vascular development mediated by the TDIF/WOX4 pathway (Donner, Sherr, & Scarpella, 2009; Scarpella, Marcos, Friml, & Ber- leth, 2006; Wenzel, Schuetz, Yu, & Mattsson, 2007). This agrees with the high degree of similarity found between the transcriptomes of 3 dpi GCs and suspension cells treated with brassinolide/boric acid that are differenti- ating into tracheids (Barcala et al., 2010; Kubo et al., 2005). These findings strongly point to provascular cells as putative precursors of the GCs, and are in agreement to initial data based on histological observations that proposed metaxylem, protoxylem or xylem parenchyma cells as the initial cells that develop into GCs (Bird, 1961; Bird & Koltai, 2000; Christie, 1936; Dropkin & Nelson, 1960; Niebel et al., 1993; Williamson & Hussey, 1996). In addi- tion, proliferating tracheids and phloem elements have been described around GCs and syncytia (Absmanner et al., 2013; Bartlem, Jones, & Hammes, 2013; Hoth, Stadler, Sauer, & Hammes, 2008). The development of these vascular elements is governed by the balance between auxins and cytokinins, as those vascular cells from the galls that differentiate into phloem elements respond to auxins, but not to cytokinins, before differentiation (Absmanner et al., 2013). In contrast, the phloem around syncytia responded to both auxins and cytokinins (Absmanner et al., 2013). An example of the interference of the RKNs with the process of the vascularization in the galls is that overexpression of the enzyme chorismate mutase, secreted by Meloi- dogyne javanica, inhibits the final differentiation of root vascular cells (Doyle & Lambert, 2003), a phenotype that can be rescued by adding indole-3-acetic acid (IAA). This suggests that nematode-secreted chorismate mutase acts by reducing IAA levels. Interestingly, both CNs and RKNs encode in their genome proteins homologues to plant chorismate mutases (Bekal, Niblack, & Lambert, 2003; Chronis, Chen, Skantar, Zasada, & Wang, 2014; Huang et al., 2005; Jones et al., 2003; Vanholme, Haegeman, Jacob, Cannoot, & Gheysen, 2009). Therefore, the secretion of chorismate mutases by Plant Parasitic Nematodes (PPNs) might be a way to interfere with the auxin/ cytokinin balance within the vascular cylinder in order to redirect the differ- entiation of vascular elements to NFCs. Auxin gradients, generated by the PIN proteins, have been shown to be a common signal for the formation of different new organs in the plant (Benkova et al., 2003; Vanneste & Friml, 2009). In this context, tomato plants treated with the polar auxin transport inhibitor NPA showed a 174 Javier Cabrera et al. reduction in the establishment of CNs and abnormal syncytia development (Goverse et al., 2000). Grunewald et al. (2009) studied the differential expression of PIN coding genes during early syncytia development in Ara- bidopsis confirming opposite regulation for different members of the family. A model was proposed in which PIN1 mediates the influx of auxin to the initial syncytia cells and PIN3 and PIN4 distribute the accumulated auxin laterally, allowing the expansion of the NFS (Grunewald et al., 2009). Thereby, CN infection rates and syncytia development are affected in pin3 mutant plants (Grunewald et al., 2009). Together with PIN1, another auxin influx carrier, LAX3, that is expressed in the syncytium and in cells to be incorporated into the syncytium together with LAX1 (Lee et al., 2011), allow the syncytia growth by increasing the auxin levels in the neighbouring cells. Strikingly, an effector protein from H. schachtii (Hs19C07) can interact with LAX3 (Lee et al., 2011). In addition, AUX1, a closely related AUX/ LAX family member, was upregulated in syncytia developed by H. schachtii and in galls induced by M. incognita in Arabidopsis (Mazarei, Lennon, Puthoff, Rodermel, & Baum, 2003). Both the aux1/lax3 double mutant and the aux1/lax1/lax2/lax3 quadruple mutant showed significant decrease in the number of female nematodes at both 14 and 30 dpi (Lee et al., 2011), sug- gesting that the LAX3–Hs19C07 interaction could alter auxin levels to promote syncytia establishment. The analysis of Arabidopsis mutant lines pin1/ttg-1 and pin2 after H. schachtii infection, resulting in the reduction of nematode and syncytia development, confirmed the role of auxin trans- porters during the plant–nematode interaction (Goverse et al., 2000). How- ever, most of the existing data are based on CNs and further analysis should be made in galls induced by RKNs to know the putative role of these pro- teins in this process. Novel classes of peptide hormones as C-terminally encoded peptide (CEP) are emerging as regulators of the developmental process leading to gall formation. CEP genes have been identified in M. incognita and Meloido- gyne hapla genomes (Bobay et al., 2013; Goverse & Bird, 2011) but not in the false RKN Nacobbus aberrans (Eves-van den Akker et al., 2014), nor in CNs (H. glycines and G. rostochiensis), migratory nematodes (Radopholus similis and Pratylenchus coffeae) or the free-living nematode (Caenorhabditis elegans; Bobay et al., 2013). In particular, MhCEP11 from M. hapla shows a significant sequence homology (Bobay et al., 2013) with the Arabidopsis-encoded CEP1 that is expressed in LR primordia. Overexpression of AtCEP1 with a constitutive promoter resulted in a reduced number of cells in the RAM (Ohyama, Ogawa, & Matsubayashi, 2008). In Medicago truncatula, Developmental Pathways Altered in Nematode Feeding Sites 175 the constitutive overexpression of MtCEP1 altered root development in several ways, by inhibition of LR formation, enhancement of nodulation and cortical, epidermal and pericycle cell divisions (Imin, Mohd- Radzman, Ogilvie, & Djordjevic, 2013). Future functional studies involving these molecules will elucidate their role as putative regulators of the gall and/or GC formation.

3. AUXINS, LATERAL ROOT FORMATION AND FEEDING SITES

CEPs are one of the most recently identified molecules that relate a root developmental process, LR formation, to gall development (Imin et al., 2013), but it is not the only connecting link. The aforementioned local increase in auxin levels favoured by PIN proteins is needed as well during the formation of LRs (Benkova et al., 2003). There are several pieces of ev- idence that point to similarities and molecular connections between the pro- cesses of LR and NFS development. Among them, it was shown that the auxin-insensitive tomato mutant diaegotropica, dgt (Richardson & Price, 1982), which lacks LRs, was resistant to M. incognita and developed smaller syncytia upon CN infection (Goverse et al., 2000). In tomato and M. trun- catula, two transcription factors, KNOX and PHAN, are induced in both GCs and LR meristems (Bird & Koltai, 2000; Koltai, Dhandaydham, Opperman, Thomas, & Bird, 2001). In Arabidopsis the downregulation of the Knotted1-like homeobox (KNOX) transcription factor KNAT6 yields an increment in the number of LRs (Dean, Casson, & Lindsey, 2004), in agreement to the suggested antagonistic action between auxin and KNOX transcription factors in organogenesis (reviewed in Scofield & Mur- ray, 2006). Moreover, Barthels et al. (1997) used a promoter-tagging strat- egy to identify specific regulatory regions differentially activated in NFS as compared to uninfected roots. Surprisingly, among the 103 promoter tag lines that displayed a distinct activation response to nematode infection, 39 also exhibited induction at LR initiation sites. This has been further confirmed by in silico analysis of transcriptomes from galls and GCs in Ara- bidopsis that showed an enrichment of characteristic genes from LR initial cells in the transcriptome of 3 dpi GCs and galls (Cabrera, Díaz-Manzano, et al., 2014). LRs originate from divisions in the xylem pole pericycle (XPP) cells following and auxin-mediated signalling pathway. Two XPP marker lines, J0121 and J0192, showed strong and distinct GFP expression in the galls formed by M. javanica (Cabrera, Díaz-Manzano, et al., 2014; 176 Javier Cabrera et al.

(A) (C)

(B)

Figure 2 Transgenic lines of lateral root and XPP pericycle markers and LBD16 express- ing either GFP or GUS in galls formed by Meloidogyne javanica. (A) J0121 >> GFP at 4 dpi. (B) J0192 >> GFP at 4 dpi. (C) pLBD16:GUS at 7 dpi (Cabrera, Díaz-Manzano, et al., 2014). Bars: 100 mm. (See colour plate)

Figure 2(A) and (B)). Strikingly, in both lines GFP expression was mostly observed at both sides of the vascular cylinder and progressed inwards during gall development, which partially differs with the expression pattern found during LR formation (Cabrera, Díaz-Manzano, et al., 2014; Laplaze et al., 2005). In addition, the regular anticlinal and periclinal divisions observed during LR formation (Lavenus et al., 2013) were substituted by abnormal division planes in the cells proliferating inside the galls (Cabrera, Díaz-Man- zano, et al., 2014). The expression pattern of LBD16, the gene whose pro- moter drives GFP expression in the J0192 enhancer trap line, mimics that of the J0192 line (Figure 2(C)). LBD16 expression is detected in galls from M. javanica and Meloidogyne arenaria from 1 dpi up to 11–15 dpi and was regu- lated by auxins, as shown by its inhibition by a-(phenyl ethyl-2-one)- indole-3-acetic acid (PEO-IAA), an antagonist of IAA, similar to what happens in LR primordia (Cabrera, Díaz-Manzano, et al., 2014; Lee, Kim, Lee, & Kim, 2009; Okushima, Fukaki, Onoda, Theologis, & Tasaka, 2007). Although its expression at early stages appeared to correlate with the presence of auxins in the same cell types, at later stages the mere presence of Developmental Pathways Altered in Nematode Feeding Sites 177 auxins in the gall was not sufficient to activate LBD16 expression (Cabrera, Díaz-Manzano, et al., 2014). These results may indicate the necessity for a threshold level of auxins in the gall to allow LBD16 expression, which would mimic the scenario that occurs during the first divisions of LR devel- opment. However, the absence of signal in the syncytia formed in the LBD16::GUS line, where the ‘auxin sensor’ DR5 was also activated (Karcz- marek et al., 2004), suggests that other signals apart from auxins could be contributing to the early activation of LBD16 in galls and GCs. In this respect, secretions from M. incognita juveniles were also able to induce LBD16 expression in Arabidopsis leaf protoplasts, suggesting an activation of the LBD16 promoter by nematode secretions, in an autonomous manner. Furthermore, a reduction in gall formation of at least 20% was observed in LBD16 loss of function lines as compared to wild type controls (Cabrera, Díaz-Manzano, et al., 2014). Interestingly, the expression pattern of the marker line ProCycB1;1:CycB1;1(NT)-GUS, active only during the G2/M transition, mimicked that of J0192 in XPP cells during early nematode establishment, and in most cells inside the vascular cylinder of the gall as the infection progressed. These results suggest that the founder cells con- tained in the XPP that divide to form a new LR (Péret, Larrieu, & Bennett, 2009) also divide during early gall formation (Cabrera, Díaz-Manzano, et al., 2014). LBD16 loss of function lines showed abnormal GC development, pointing to a role of the pericycle during this process. The importance of XPP-specific genes during infection in Arabidopsis was further demonstrated by genetic ablation using a J0121 >> DTA line that showed a dramatic reduction in the infection and in the size of the GCs as compared to a con- trol J0121 >> GFP line (Cabrera, Díaz-Manzano, et al., 2014). The induction of LBD16 during gall formation not only connects this process with LR formation but with the generation of calli (Cabrera, Díaz-Manzano, et al., 2014; Demeulenaere & Beeckman, 2014). LBD tran- scription factors and pericycle cells have been shown to be essential for the generation of calli from different organs through ectopic activation of an LR developmental program (Sugimoto, Jiao, & Meyerowitz, 2010). The ectopic expression of LBDs triggers spontaneous callus formation but their suppression inhibits the process (Fan, Xu, Xu, & Hu, 2012); whether or not gall development is somehow related to this process of callus formation through the activation of an LR initiation-like program remains to be eluci- dated. In silico data comparison supported this hypothesis, as those genes co-regulated with LBD16 in different transcriptomes were integrated in sig- nalling cascades mediated by auxins during LR and callus formation, as a 178 Javier Cabrera et al. particular feature of early developing RKN feeding sites (3 dpi) distinct to CNs. In contrast, cytokinin-induced genes were enriched in syncytia, whose transcriptomes hold a high similarity with the transcriptome of shoot regen- eration from callus, modulated by cytokinins (Cabrera, Bustos, Favery, Fenoll, & Escobar, 2014; Cabrera, Fenoll, & Escobar, in press). In agreement with these analyses, subtle changes in the balance between cytokinin and auxin levels could be mediating the appearance of chloroplast-like structures inside 7 dpi GCs induced by Meloidogyne graminicola in rice, as showed by confocal microscopy (Ji et al., 2013; Kyndt, Vieira, Gheysen, & de Almeida-Engler, 2013) and in syncytia (Szakasits et al., 2009). Transcrip- tomic studies performed in isolated GCs and syncytia reflected the predom- inance of genes regulated by phytohormones among the differentially expressed genes as compared to noninfected tissues (reviewed in Cabrera, Bustos, et al., 2014; Cabrera, Fenoll, et al., in press; Escobar, Horowitz, & Mitchum, 2011). In this way, a direct in silico comparison of the transcrip- tomes of isolated GCs and syncytia at early developmental stages in Arabidop- sis and the transcriptomes from seedlings treated with exogenous phytohormones contributed to increase the vast number of genes regulated by hormones that are also induced or repressed in NFCs (Cabrera, Bustos, et al., 2014; Cabrera, Fenoll, et al., in press). Clear differences between the hormone-related transcriptional balances of the two NFC types were found. While the percentage of auxin-induced genes stands out (26%) in GCs, in syncytia there are 21% of cytokinin-induced genes. On the con- trary, in GCs there are more cytokinin-repressed genes as compared to those upregulated; in syncytia the number of cytokinin upregulated genes is higher than the number of downregulated ones. Interestingly, the number of genes repressed by auxins or cytokinins in GCs or syncytia was high, suggesting that gene repression driven by these hormones may be also crucial for the development of the NFS (Cabrera, Bustos, et al., 2014; Cabrera, Fenoll, et al., in press). WRKY23 is another transcription factor regulated by auxins and induced by both RKN and CN. Loss of function lines showed an increased resistance to CNs (Grunewald et al., 2009). WRKY23 acts downstream of the signalling cascade (SLR/IAA14)-(auxin responsive factors) (ARF7/ ARF19) during NFs and LR formation (Grunewald et al., 2008, 2012) and is needed for LR development through the stimulation of local flavonol biosynthesis (Grunewald et al., 2012). However, WRKY23 expression is activated in NFS by an independent auxin pathway, suggesting the existence of other nematode-dependent signals in regulating WRKY23 expression. Developmental Pathways Altered in Nematode Feeding Sites 179

This could also be the case for LBD16 expression that is activated by nem- atode secretions as previously mentioned (Cabrera, Díaz-Manzano, et al., 2014). Other members of the LBD transcription factor family, such as LBD41, are activated in both M. incognita and H. schachtii feeding sites (Fuller, Lilley, Atkinson, & Urwin, 2007), in agreement with transcriptomic analysis of GCs in Arabidopsis (Barcala et al., 2010). The expression patterns of crucial molecular components of the auxin signalling pathway, encoded by genes of the ARFs family, have been addressed during syncytium development in Arabidopsis (Hewezi, Piya, Richard, & Rice, 2014). At early infection stages, 2–3 days after H. schachtii infection, ARF3, 6, 10–12, 14, 15 and 20–22 were expressed inside the developing syncytium, while ARF1, 2, 4, 5, 9, 18 and 19 were active in both syncytial and neighbouring cells. ARF7 and 17 were mainly expressed at the edges of the syncytial and neighbouring cells and ARF8 and 16 showed a weak response to H. schachtii infection (Hewezi et al., 2014). At 9–10 days after infection, ARF1–3, 7, 17 and 20–22 were expressed in fully developed syncytium, whereas the expression of the other ARFs was restricted to the syncytial cells around the nematode head (Hewezi et al., 2014). Although still under investigation, the differential expression patterns of the ARF genes seem to be essential for the correct development of the CNs feeding cells. All these data suggest a subtle and complex regulation of auxin-mediated pathways based on a tight temporal and spatial control of molecular components such as ARFs in CNs feeding sites. A lack of knowledge of the regulation of these genes after RKN infection is still faced.

4. GIANT CELL MORPHOGENESIS AND TRANSFER CELL NATURE

In the previous sections, we described parallelisms between feeding site formation and developmental programs during the plant life cycle, such as LR formation. Although syncytia and GCs differ in their ontogeny and global transcriptional signatures, both develop cell wall ingrowths (CIs) to facilitate high rates of apoplastic/symplastic solute exchange. Both feeding site types also show similarities to the TCs that appear in different plant or- gans during plant development (reviewed in Offler, McCurdy, Patrick, & Talbot, 2003). In GCs the amplification of the plasma membrane surface area could be up to 20-fold (reviewed in Jones & Goto, 2011). Syncytia induced by H. schachtii are symplastically isolated at 10–15 dpi (Hofmann, Wieczorek, 180 Javier Cabrera et al.

Blochl,€ & Grundler, 2007) and the CIs are smaller in male- than in female- developed syncytia, suggesting that CI size control is based on the nutrient demand of the nematode (reviewed in Sobczak & Golinowski, 2008). Recently, three-dimensional reconstruction and volume measurements of GCs in Arabidopsis (Cabrera, Díaz-Manzano, et al., in press) brought some interesting findings that might explain, at least partially, their TC character- istics. The abnormally large size of the GCs implies a reduction in their sur- face area to volume ratio (S/V ratio; Cabrera, Díaz-Manzano, et al., in press), a factor that may compromise its functioning during nematode nourishing. Thus, the extensive formation of wall ingrowths lined with plasma mem- brane at certain developmental stages that defines them as TCs (Jones & Dropkin, 1976; Siddique, Sobczak, Tenhaken, Grundler, & Bohlmann, 2012) could be a response for a functional requirement to compensate the decrease of the S/V ratio as the GCs expand (reviewed in Cabrera, Díaz- Manzano, et al., in press; Rodiuc, Vieira, Banora, & de Almeida-Engler, 2014). Moreover, size regulation affects cell function in multiple ways, e.g. not only can nutrient and water movement be changed by altered sur- face/volume ratio, but also intercellular signalling might be influenced by changes in cell size and geometry. Recent studies in yeast revealed that tran- scription also changes specifically in response to cell dimensions (Sablowski & Dornelas, 2014; Wu, Rolfe, Gifford, & Fink, 2010). Cell wall anatomy and composition in NFCs are quite similar to other TCs that develop in plants, and it is mainly composed of polysaccharides such as cellulose, hemicelluloses and pectin (reviewed in Rodiuc et al., 2014). Although regulatory signals for TC differentiation are not well known, the transcription factor ZmMRP-1, that has been described as a key component in the pathway leading to the formation of the TCs (Gomez, Royo, Guo, Thompson, & Hueros, 2002), is also induced in Arabidopsis galls (Barrero et al., 2009). Auxins and ethylene are the major phytohormones described as regulators of TC differentiation from several cell types (Dibley et al., 2009; Thiel et al., 2008; Thiel, Hollmann, et al., 2012; Thiel, Riewe, et al., 2012; Xiong, Li, Kang, & Chourey, 2011; Zhou et al., 2010). Com- mon genes related to the auxin signalling cascades and transport from the IAA/ARF/PIN families play a role and/or show differential regulation in both TCs and NFCs (reviewed in Cabrera, Barcala, Fenoll, & Escobar, 2014; this chapter). Genes related with ethylene synthesis and signalling as those encoding membrane receptors, aminocyclopropane-1-carboxylic (ACC) acid oxidases or ACC acid synthase, are also to be induced in NFS (Cabrera, Barcala, et al., 2014; Tucker, Xue, & Yang, 2010). Increased Developmental Pathways Altered in Nematode Feeding Sites 181 concentration of ethylene in tomato galls was described long ago (Glazer, Orion, & Apelbaum, 1983). Moreover, functional analysis performed with ethylene-related mutants reinforce the importance of this hormone during plant–nematode interaction as ethylene-overproducing mutants eto2 and eto3 were more susceptible to CNs (Goverse et al., 2000; Wubben, Su, Rodermel, & Baum, 2001). Interestingly, ethylene overproduction in eto2 mutants stimulated the formation of CIs or protuberances in syncytia along the vascular tissue, at late infection stages (Goverse et al., 2000), providing direct evidence for a putative role of ethylene in the stimulation of syncytia TC identity. On the other hand, mutants compromised in the ethylene sig- nalling cascade (etr1-1, ein2-1, ein3-1, eir1-1 and axr2) showed a lower sus- ceptibility to infection by CNs (Goverse et al., 2000; Wubben et al., 2001). Interestingly, the most clarifying study of a functional implication in TCs characteristic of NFCs, such as the CIs formation, comes from the analysis of uridine diphosphate (UDP)-glucose dehydrogenase (UGD) coding genes. UGDs act through oxidation of UDP-glucose producing several cell wall polysaccharides. UGD2 and UGD3 are necessary for the production of CIs in syncytia and loss of function in double mutants severely affected nem- atode development (Siddique et al., 2012). In conclusion, there are several evidences that nematodes may interfere or partially ‘hijack’ signal transduction pathways used by the plant to initiate and/ or maintain developmental processes where auxins and cytokinins play a cen- tral role, such as SAM and RAM maintenance and LR and vascular tissue for- mation. In addition, nematodes may also interfere with transduction pathways leading to the differentiation of specialized plant cell types such as TCs. The understanding of the plant molecular components necessary to reprogramme normal plant cells into NFCs are an outstanding topic to be deciphered, what would contribute to the basic understanding of the plant–nematode interac- tion. This will also help to identify the interactions of nematode effectors with plant components, and ultimately, their mode of action while interfering with the plant developmental programs. Furthermore, this knowledge will consti- tute a powerful tool for engineering nematode resistance in plants and to direct the search for specific nematode control strategies.

ACKNOWLEDGEMENTS This work was supported by the Spanish Government (AGL2010-17388 and AGL2013- 48787-R to C. Escobar; and CSD2007-057 and PCIN-2013-053 to C. Fenoll). J. Cabrera and F. E. Díaz-Manzano were supported by predoctoral fellowships from the Spanish Government. 182 Javier Cabrera et al.

REFERENCES Absmanner, B., Stadler, R., & Hammes, U. Z. (2013). Phloem development in nematode- induced feeding sites: the implications of auxin and cytokinin. Frontiers in Plant Science, 4. Bakhetia, M., Urwin, P. E., & Atkinson, H. J. (2007). QPCR analysis and RNAi define pharyngeal gland cell-expressed genes of Heterodera glycines required for initial interac- tions with the host. Molecular Plant–Microbe Interactions, 20(3), 306–312. Balasubramanian, M., & Rangaswami, G. (1962). Presence of indole compounds in nema- tode galls. Nature, 194, 774–775. Barcala, M., García, A., Cabrera, J., Casson, S., Lindsey, K., Favery, B., et al. (2010). Early transcriptomic events in microdissected Arabidopsis nematode-induced giant cells. The Plant Journal, 61(4), 698–712. Barrero, C., Royo, J., Grijota-Martinez, C., Faye, C., Paul, W., Sanz, S., et al. (2009). The promoter of ZmMRP-1, a maize transfer cell-specific transcriptional activator, is induced at solute exchange surfaces and responds to transport demands. Planta, 229(2), 235–247. Barthels, N., van der Lee, F. M., Klap, J., Goddijn, O. J., Karimi, M., Puzio, P., et al. (1997). Regulatory sequences of Arabidopsis drive reporter gene expression in nematode feeding structures. The Plant Cell, 9(12), 2119–2134. Bartlem, D. G., Jones, M. G., & Hammes, U. Z. (2013). Vascularization and nutrient delivery at root-knot nematode feeding sites in host roots. Journal of Experimental Botany, 65(7), 1789–1798. Bekal, S., Niblack, T. L., & Lambert, K. N. (2003). A chorismate mutase from the soybean cyst nematode Heterodera glycines shows polymorphisms that correlate with virulence. Mo- lecular Plant–Microbe Interactions, 16(5), 439–446. Benkova, E., Michniewicz, M., Sauer, M., Teichmann, T., Seifertova, D., Jurgens,€ G., et al. (2003). Local, efflux-dependent auxin gradients as a common module for plant organ formation. Cell, 115(5), 591–602. Bielach, A., Duclercq, J., Marhavý, P., & Benkova, E. (2012). Genetic approach towards the identification of auxin–cytokinin crosstalk components involved in root development. Philosophical Transactions of the Royal Society B: Biological Sciences, 367(1595), 1469–1478. Bird, A. F. (1961). The ultrastructure and histochemistry of a nematode-induced giant cell. The Journal of Biophysical and Biochemical Cytology, 11(3), 701–715. Bird, D. M., & Koltai, H. (2000). Plant parasitic nematodes: habitats, hormones, and horizon- tally acquired genes. Journal of Plant Growth Regulation, 19(2), 183–194. Bird, D. M., Williamson, V. M., Abad, P., McCarter, J., Danchin, E. G., Castagnone- Sereno, P., et al. (2009). The genomes of root-knot nematodes. Annual Review of Phyto- pathology, 47, 333–351. Bishopp, A., Benkova, E., & Helariutta, Y. (2011). Sending mixed messages: auxin-cytokinin crosstalk in roots. Current Opinion in Plant Biology, 14(1), 10–16. Bishopp, A., Help, H., El-Showk, S., Weijers, D., Scheres, B., Friml, J., et al. (2011). A mutually inhibitory interaction between auxin and cytokinin specifies vascular pattern in roots. Current Biology, 21(11), 917–926. Bobay, B. G., DiGennaro, P., Scholl, E., Imin, N., Djordjevic, M. A., & Mck Bird, D. (2013). Solution NMR studies of the plant peptide hormone CEP inform function. FEBS Letters, 587(24), 3979–3985. Cabrera, J., Barcala, M., Fenoll, C., & Escobar, C. (2014). Transcriptomic signatures of trans- fer cells in early developing nematode feeding cells of Arabidopsis focused on auxin and ethylene signalling. Frontiers in Plant Science, 5. Cabrera, J., Bustos, R., Favery, B., Fenoll, C., & Escobar, C. (2014). NEMATIC: a simple and versatile tool for the in silico analysis of plant–nematode interactions. Molecular Plant Pathology, 15(6), 627–636. Developmental Pathways Altered in Nematode Feeding Sites 183

Cabrera, J., Díaz-Manzano, F. E., Barcala, M., de Almeida-Engler, J., Engler, G., Fenoll, C., et al. Phenotyping nematode feeding sites: three dimensional reconstruction and volu- metric measurements of giant cells induced by root-knot nematodes in Arabidopsis. New Phytologist, in press. Cabrera, J., Díaz-Manzano, F. E., Sanchez, M., Rosso, M.-N., Melillo, T., Goh, T., et al. (2014). A role for LATERAL ORGAN BOUNDARIES-DOMAIN 16 during the interaction Arabidopsis – Meloidogyne spp. provides a molecular link between lateral root and root-knot nematode feeding site development. New Phytologist, 203, 632–645. Cabrera, J., Fenoll, C., & Escobar, C. Genes co-regulated with LBD16 in nematode feeding sites inferred from in silico analysis show similarities to regulatory circuits mediated by the auxin/cytokinin balance in Arabidopsis. Plant Signaling and Behaviour, in press. Christie, J. R. (1936). The development of root-knot nematode galls. Phytopathology, 26, 1–22. Chronis, D., Chen, S., Skantar, A. M., Zasada, I. A., & Wang, X. (2014). A new chorismate mutase gene identified from Globodera ellingtonae and its utility as a molecular diagnostic marker. European Journal of Plant Pathology, 139(2), 239–246. De Meutter, J., Robertson, L. E. E., Parcy, F., Mena, M., Fenoll, C., & Gheysen, G. (2005). Differential activation of ABI3 and LEA genes upon plant parasitic nematode infection. Molecular Plant Pathology, 6(3), 321–325. De Meutter, J., Tytgat, T., Witters, E., Gheysen, G., Van Onckelen, H., & Gheysen, G. (2003). Identification of cytokinins produced by the plant parasitic nematodes Heterodera schachtii and Meloidogyne incognita. Molecular Plant Pathology, 4(4), 271–277. Dean, G., Casson, S., & Lindsey, K. (2004). KNAT6 gene of Arabidopsis is expressed in roots and is required for correct lateral root formation. Plant Molecular Biology, 54(1), 71–84. Dello Ioio, R., Linhares, F. S., Scacchi, E., Casamitjana-Martinez, E., Heidstra, R., Costantino, P., et al. (2007). Cytokinins determine Arabidopsis root-meristem size by controlling cell differentiation. Current Biology, 17(8), 678–682. Demeulenaere, M. J., & Beeckman, T. (2014). The interplay between auxin and the cell cy- cle during plant development. In E. Zazímalova, J. Petrasek, & E. Benkova (Eds.), Auxin and its role in plant development (pp. 119–141). Vienna: Springer. Di Laurenzio, L., Wysocka-Diller, J., Malamy, J. E., Pysh, L., Helariutta, Y., Freshour, G., et al. (1996). The SCARECROW gene regulates an asymmetric cell division that is essential for generating the radial organization of the Arabidopsis root. Cell, 86(3), 423–433. Dibley, S. J., Zhou, Y., Andriunas, F. A., Talbot, M. J., Offler, C. E., Patrick, J. W., et al. (2009). Early gene expression programs accompanying trans-differentiation of epidermal cells of Vicia faba cotyledons into transfer cells. New Phytologist, 182(4), 863–877. Donner, T. J., Sherr, I., & Scarpella, E. (2009). Regulation of pre-procambial cell state acqui- sition by auxin signaling in Arabidopsis leaves. Development, 136(19), 3235–3246. Doyle, E. A., & Lambert, K. N. (2003). Meloidogyne javanica chorismate mutase 1 alters plant cell development. Molecular Plant–Microbe Interactions, 16(2), 123–131. Dropkin, V. H., & Nelson, P. E. (1960). The histopathology of root-knot nematode infec- tions in soybeans. Phytopathology, 50(6), 442–447. Escobar, C., Horowitz, S. B., & Mitchum, M. G. (2011). Transcriptomic and proteomic analysis of the plant response to nematode infection. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant–nematode interactions (pp. 157– 173). Dordrecht, Heidelberg, London & New York: Springer. Eves-van den Akker, S., Lilley, C. J., Danchin, E. G., Rancurel, C., Cock, P. J., Urwin, P. E., et al. (2014). The transcriptome of Nacobbus aberrans reveals insights into the evolution of sedentary endoparasitism in plant-parasitic nematodes. Genome Biology and Evolution, 6(9), 2181–2194. 184 Javier Cabrera et al.

Fan, M., Xu, C., Xu, K., & Hu, Y. (2012). LATERAL ORGAN BOUNDARIES DOMAIN transcription factors direct callus formation in Arabidopsis regeneration. Cell Research, 22(7), 1169–1180. Fuller, V. L., Lilley, C. J., Atkinson, H. J., & Urwin, P. E. (2007). Differential gene expression in Arabidopsis following infection by plant-parasitic nematodes Meloidogyne incognita and Heterodera schachtii. Molecular Plant Pathology, 8(5), 595–609. Glazer, I., Orion, D., & Apelbaum, A. (1983). Interrelationships between ethylene produc- tion, gall formation, and root-knot nematode development in tomato plants infected with Meloidogyne javanica. Journal of Nematology, 15(4), 539. Gomez, E., Royo, J., Guo, Y., Thompson, R., & Hueros, G. (2002). Establishment of cereal endosperm expression domains identification and properties of a maize transfer cell–spe- cific transcription factor, ZmMRP-1. The Plant Cell, 14(3), 599–610. Goverse, A., & Bird, D. (2011). The role of plant hormones in nematode feeding cell formation. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant–nematode interactions (pp. 325–347). Dordrecht, Heidelberg, London & New York: Springer. Goverse, A., Overmars, H., Engelbertink, J., Schots, A., Bakker, J., & Helder, J. (2000). Both induction and morphogenesis of cyst nematode feeding cells are mediated by auxin. Mo- lecular Plant–Microbe Interactions, 13(10), 1121–1129. Grunewald, W., Cannoot, B., Friml, J., & Gheysen, G. (2009). Parasitic nematodes modulate PIN-mediated auxin transport to facilitate infection. PLoS Pathogens, 5(1), e1000266. Grunewald, W., De Smet, I., Lewis, D. R., Lofke,€ C., Jansen, L., Goeminne, G., et al. (2012). Transcription factor WRKY23 assists auxin distribution patterns during Arabidop- sis root development through local control on flavonol biosynthesis. Proceedings of the National Academy of Sciences of the United States of America, 109(5), 1554–1559. Grunewald, W., Karimi, M., Wieczorek, K., Van de Cappelle, E., Wischnitzki, E., Grundler, F. M., et al. (2008). A role for AtWRKY23 in feeding site establishment of plant-parasitic nematodes. Plant Physiology, 148(1), 358–368. Guo, Y., Ni, J., Denver, R., Wang, X., & Clark, S. E. (2011). Mechanisms of molecular mimicry of plant CLE peptide ligands by the parasitic nematode Globodera rostochiensis. Plant Physiology, 157(1), 476–484. Guo, X., Chronis, D., De La Torre, C. M., Smed, J., Wang, X., & Mitchum, M. G. (2015). Enhanced resistance to soybean cyst nematode Heterodera glycines in transgenic soybean by silencing putative CLE receptors. Plant Biotechnol J. http://dx.doi.org/10.1111/ pbi.12313. Hewezi, T., Piya, S., Richard, G., & Rice, J. H. (2014). Spatial and temporal expression patterns of auxin response transcription factors in the syncytium induced by the beet cyst nematode Heterodera schachtii in Arabidopsis. Molecular Plant Pathology, 15(7), 730–736. Hirakawa, Y., Kondo, Y., & Fukuda, H. (2010). TDIF peptide signaling regulates vascular stem cell proliferation via the WOX4 homeobox gene in Arabidopsis. The Plant Cell, 22(8), 2618–2629. Hofmann, J., Wieczorek, K., Blochl,€ A., & Grundler, F. M. (2007). Sucrose supply to nematode-induced syncytia depends on the apoplasmic and symplasmic pathways. Jour- nal of Experimental Botany, 58(7), 1591–1601. Hoth, S., Stadler, R., Sauer, N., & Hammes, U. Z. (2008). Differential vascularization of nematode-induced feeding sites. Proceedings of the National Academy of Sciences of the United States of America, 105(34), 12617–12622. Huang, G., Allen, R., Davis, E. L., Baum, T. J., & Hussey, R. S. (2006). Engineering broad root-knot resistance in transgenic plants by RNAi silencing of a conserved and essential root-knot nematode parasitism gene. Proceedings of the National Academy of Sciences of the United States of America, 103(39), 14302–14306. Developmental Pathways Altered in Nematode Feeding Sites 185

Huang, X., Tian, B., Niu, Q., Yang, J., Zhang, L., & Zhang, K. (2005). An extracellular pro- tease from Brevibacillus laterosporus G4 without parasporal crystals can serve as a pathogenic factor in infection of nematodes. Research in Microbiology, 156(5), 719–727. Hutangura, P., Mathesius, U., Jones, M. G., & Rolfe, B. G. (1999). Auxin induction is a trigger for root gall formation caused by root-knot nematodes in white clover and is associated with the activation of the flavonoid pathway. Functional Plant Biology, 26(3), 221–231. Imin, N., Mohd-Radzman, N. A., Ogilvie, H. A., & Djordjevic, M. A. (2013). The peptide- encoding CEP1 gene modulates lateral root and nodule numbers in Medicago truncatula. Journal of Experimental Botany, 64(17), 5395–5409. Ito, Y., Nakanomyo, I., Motose, H., Iwamoto, K., Sawa, S., Dohmae, N., et al. (2006). Dodeca-CLE peptides as suppressors of plant stem cell differentiation. Science, 313(5788), 842–845. Ji, H., Gheysen, G., Denil, S., Lindsey, K., Topping, J. F., Nahar, K., et al. (2013). Transcrip- tional analysis through RNA sequencing of giant cells induced by Meloidogyne graminicola in rice roots. Journal of Experimental Botany, 64(12), 3885–3898. Jones, M. G. K., & Dropkin, V. H. (1976). Scanning electron microscopy of nematode- induced giant transfer cells. Cytobios, 15(58–59), 149. Jones, J. T., Furlanetto, C., Bakker, E., Banks, B., Blok, V., Chen, Q., et al. (2003). Char- acterization of a chorismate mutase from the potato cyst nematode Globodera pallida. Mo- lecular Plant Pathology, 4(1), 43–50. Jones, M. G., & Goto, D. B. (2011). Root-knot nematodes and giant cells. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant–nematode interac- tions (pp. 83–100). Dordrecht, Heidelberg, London & New York: Springer. Karczmarek, A., Overmars, H., Helder, J., & Goverse, A. (2004). Feeding cell development by cyst and root-knot nematodes involves a similar early, local and transient activation of a specific auxin-inducible promoter element. Molecular Plant Pathology, 5(4), 343–346. Koltai, H., Dhandaydham, M., Opperman, C., Thomas, J., & Bird, D. (2001). Overlapping plant signal transduction pathways induced by a parasitic nematode and a rhizobial endosymbiont. Molecular Plant–Microbe Interactions, 14(10), 1168–1177. Krupasagar, V., & Barker, K. R. (1969). Increased cytokinin concentrations in tobacco infected with the root-knot nematode Meloidogyne incognita. Phytopathology, 56, 885. Kubo, M., Udagawa, M., Nishikubo, N., Horiguchi, G., Yamaguchi, M., Ito, J., et al. (2005). Transcription switches for protoxylem and metaxylem vessel formation. Genes and Development, 19(16), 1855–1860. Kyndt, T., Vieira, P., Gheysen, G., & de Almeida-Engler, J. (2013). Nematode feeding sites: unique organs in plant roots. Planta, 238(5), 807–818. Laplaze, L., Parizot, B., Baker, A., Ricaud, L., Martiniere, A., Auguy, F., et al. (2005). GAL4- GFP enhancer trap lines for genetic manipulation of lateral root development in Arabi- dopsis thaliana. Journal of Experimental Botany, 56(419), 2433–2442. Lavenus, J., Goh, T., Roberts, I., Guyomarc’h, S., Lucas, M., De Smet, I., et al. (2013). Lateral root development in Arabidopsis: fifty shades of auxin. Trends in Plant Science, 18(8), 450–458. Leasure, C. D., & He, Z. H. (2012). CLE and RGF family peptide hormone signaling in plant development. Molecular Plant, 5(6), 1173–1175. Lee, C., Chronis, D., Kenning, C., Peret, B., Hewezi, T., Davis, E. L., et al. (2011). The novel cyst nematode effector protein 19C07 interacts with the Arabidopsis auxin influx transporter LAX3 to control feeding site development. Plant Physiology, 155(2), 866–880. Lee, H. W., Kim, N. Y., Lee, D. J., & Kim, J. (2009). LBD18/ASL20 regulates lateral root formation in combination with LBD16/ASL18 downstream of ARF7 and ARF19 in Arabidopsis. Plant Physiology, 151(3), 1377–1389. 186 Javier Cabrera et al.

Levesque, M. P., Vernoux, T., Busch, W., Cui, H., Wang, J. Y., Blilou, I., et al. (2006). Whole-genome analysis of the SHORT-ROOT developmental pathway in Arabidopsis. PLoS Biology, 4(5), e143. Lohar, D. P., Schaff, J. E., Laskey, J. G., Kieber, J. J., Bilyeu, K. D., & Bird, D. M. (2004). Cytokinins play opposite roles in lateral root formation, and nematode and Rhizobial symbioses. The Plant Journal, 38(2), 203–214. Lu,S.W.,Chen,S.,Wang,J.,Yu,H.,Chronis,D.,Mitchum,M.G.,etal.(2009).Struc- tural and functional diversity of CLAVATA3/ESR (CLE)-like genes from the potato cyst nematode Globodera rostochiensis. Molecular Plant–Microbe Interactions, 22(9), 1128–1142. MacLean, A. M., Orlovskis, Z., Kowitwanich, K., Zdziarska, A. M., Angenent, G. C., Immink, R. G., et al. (2014). Phytoplasma effector SAP54 hijacks plant reproduction by degrading MADS-box proteins and promotes insect colonization in a RAD23- dependent manner. PLoS Biology, 12(4), e1001835. Mazarei, M., Lennon, K. A., Puthoff, D. P., Rodermel, S. R., & Baum, T. J. (2003). Expres- sion of an Arabidopsis phosphoglycerate mutase homologue is localized to apical meri- stems, regulated by hormones, and induced by sedentary plant-parasitic nematodes. Plant Molecular Biology, 53(4), 513–530. Moens, M., Perry, R. N., & Starr, J. L. (2009). Meloidogyne species – a diverse group of novel and important plant parasites. In R. N. Perry, M. Moens, & J. L. Starr (Eds.), Root-knot nematodes (pp. 1–13). United Kingdom: CAB International. Moubayidin, L., Di Mambro, R., & Sabatini, S. (2009). Cytokinin – auxin crosstalk. Trends in Plant Science, 14(10), 557–562. Niebel, A., de Almeida-Engler, J., Tire, C., Engler, G., Van Montagu, M., & Gheysen, G. (1993). Induction patterns of an extensin gene in tobacco upon nematode infection. The Plant Cell, 5(12), 1697–1710. Offler, C. E., McCurdy, D. W., Patrick, J. W., & Talbot, M. J. (2003). TRANSFER cells: cells specialized for a special purpose. Annual Review of Plant Biology, 54, 431–454. Ohyama, K., Ogawa, M., & Matsubayashi, Y. (2008). Identification of a biologically active, small, secreted peptide in Arabidopsis by in silico gene screening, followed by LC-MS- based structure analysis. The Plant Journal, 55(1), 152–160. Okushima, Y., Fukaki, H., Onoda, M., Theologis, A., & Tasaka, M. (2007). ARF7 and ARF19 regulate lateral root formation via direct activation of LBD/ASL genes in Arabidopsis. The Plant Cell, 19(1), 118–130. Patel, N., Hamamouch, N., Li, C., Hussey, R. S., Mitchum, M. G., Baum, T. J., et al. (2008). Similarity and functional analyses of expressed parasitism genes in Heterodera schachtii and Heterodera glycines. Journal of Nematology, 40(4), 299. Péret, B., Larrieu, A., & Bennett, M. J. (2009). Lateral root emergence: a difficult birth. Jour- nal of Experimental Botany, 60(13), 3637–3643. Perry, R. N., & Moens, M. (2011). Introduction to plant-parasitic nematodes: modes of parasitism. In J. Jones, G. Gheysen, & C. Fenoll (Eds.), Genomics and molecular genetics of plant–nematode interactions (pp. 3–20). Dordrecht, Heidelberg, London & New York: Springer. Replogle, A., Wang, J., Bleckmann, A., Hussey, R. S., Baum, T. J., Sawa, S., et al. (2011). Nematode CLE signaling in Arabidopsis requires CLAVATA2 and CORYNE. The Plant Journal, 65(3), 430–440. Replogle, A., Wang, J., Paolillo, V., Smeda, J., Kinoshita, A., Durbak, A., et al. (2013). Syn- ergistic interaction of CLAVATA1, CLAVATA2, and RECEPTOR-LIKE PROTEIN KINASE 2 in cyst nematode parasitism of Arabidopsis. Molecular Plant–Microbe Interactions, 26(1), 87–96. Richardson, L., & Price, N. S. (1982). Host–parasite relationships of Meloidogyne incognita and the dia-geotropica tomato mutant. Journal of Nematology, 14, 465–466. Developmental Pathways Altered in Nematode Feeding Sites 187

Rodiuc, N., Vieira, P., Banora, M. Y., & de Almeida-Engler, J. (2014). On the track of trans- fer cell formation by specialized plant-parasitic nematodes. Frontiers in Plant Science, 5. Sablowski, R., & Dornelas, M. C. (2014). Interplay between cell growth and cell cycle in plants. Journal of Experimental Botany, 65(10), 2703–2714. Saini, S., Sharma, I., Kaur, N., & Pati, P. K. (2013). Auxin: a master regulator in plant root development. Plant Cell Reports, 32(6), 741–757. Scarpella, E., Marcos, D., Friml, J., & Berleth, T. (2006). Control of leaf vascular patterning by polar auxin transport. Genes and Development, 20(8), 1015–1027. Scheres, P., Sijmons, P. C., van den Berg, C., McKhann, H., de Vrieze, G., Willemsen, V., et al. (1997). Root anatomy and development, the basis for nematode parasitism. In C. Fenoll, F. M. Grundler, & S. A. Ohl (Eds.), Cellular and molecular aspects of plant–nem- atode interactions (pp. 25–38). Netherlands: Kluwer Academic Publishers. Schoof, H., Lenhard, M., Haecker, A., Mayer, K. F., Jurgens,€ G., & Laux, T. (2000). The stem cell population of Arabidopsis shoot meristems is maintained by a regulatory loop between the CLAVATA and WUSCHEL genes. Cell, 100(6), 635–644. Scofield, S., & Murray, J. A. (2006). KNOX gene function in plant stem cell niches. Plant Molecular Biology, 60(6), 929–946. Siddique, S., Sobczak, M., Tenhaken, R., Grundler, F. M., & Bohlmann, H. (2012). Cell wall ingrowths in nematode induced syncytia require UGD2 and UGD3. PloS One, 7(7), e41515. Sobczak, M., & Golinowski, W. (2008). Structure of cyst nematode feeding sites. In R. H. Berg, & C. G. Taylor (Eds.), Cell biology of plant nematode parasitism (pp. 153– 187). Berlin: Springer. Stahl, Y., Wink, R. H., Ingram, G. C., & Simon, R. (2009). A signaling module controlling the stem cell niche in Arabidopsis root meristems. Current Biology, 19(11), 909–914. Sugimoto, K., Jiao, Y., & Meyerowitz, E. M. (2010). Arabidopsis regeneration from multiple tissues occurs via a root development pathway. Developmental Cell, 18(3), 463–471. Szakasits, D., Heinen, P., Wieczorek, K., Hofmann, J., Wagner, F., Kreil, D. P., et al. (2009). The transcriptome of syncytia induced by the cyst nematode Heterodera schachtii in Ara- bidopsis roots. The Plant Journal, 57(5), 771–784. Thiel, J., Hollmann, J., Rutten, T., Weber, H., Scholz, U., & Weschke, W. (2012). 454 Transcriptome sequencing suggests a role for two-component signalling in cellularization and differentiation of barley endosperm transfer cells. PloS One, 7(7), e41867. Thiel, J., Riewe, D., Rutten, T., Melzer, M., Friedel, S., Bollenbeck, F., et al. (2012). Dif- ferentiation of endosperm transfer cells of barley: a comprehensive analysis at the micro- scale. The Plant Journal, 71(4), 639–655. Thiel, J., Weier, D., Sreenivasulu, N., Strickert, M., Weichert, N., Melzer, M., et al. (2008). Different hormonal regulation of cellular differentiation and function in nucellar projec- tion and endosperm transfer cells: a microdissection-based transcriptome study of young barley grains. Plant Physiology, 148(3), 1436–1452. Tucker, M. L., Xue, P., & Yang, R. (2010). 1-Aminocyclopropane-1-carboxylic acid (ACC) concentration and ACC synthase expression in soybean roots, root tips, and soybean cyst nematode (Heterodera glycines)-infected roots. Journal of Experimental Botany, 61(2), 463–472. Vanholme, B., Haegeman, A., Jacob, J., Cannoot, B., & Gheysen, G. (2009). Arabinogalac- tan endo-1,4-b-galactosidase: a putative plant cell wall-degrading enzyme of plant- parasitic nematodes. Nematology, 11(5), 739–747. Vanneste, S., & Friml, J. (2009). Auxin: a trigger for change in plant development. Cell, 136(6), 1005–1016. Wang, X., Allen, R., Ding, X., Goellner, M., Maier, T., de Boer, J. M., et al. (2001). Signal peptide-selection of cDNA cloned directly from the esophageal gland cells of the soybean cyst nematode Heterodera glycines. Molecular Plant–Microbe Interactions, 14(4), 536–544. 188 Javier Cabrera et al.

Wang, J., Lee, C., Replogle, A., Joshi, S., Korkin, D., Hussey, R. S., et al. (2010). Dual roles for the variable domain in protein trafficking and host-specific recognition of Heterodera glycines CLE effector proteins. New Phytologist, 187(4), 1003–1017. Wang, X., Mitchum, M. G., Gao, B., Li, C., Diab, H., Baum, T. J., et al. (2005). A parasitism gene from a plant-parasitic nematode with function similar to CLAVATA3/ESR (CLE) of Arabidopsis thaliana. Molecular Plant Pathology, 6(2), 187–191. Wang, J., Replogle, A., Hussey, R. S., Baum, T. J., Wang, X., Davis, E. L., et al. (2011). Identification of potential host plant mimics of CLAVATA3/ESR (CLE)-like peptides from the plant-parasitic nematode Heterodera schachtii. Molecular Plant Pathology, 12(2), 177–186. Wenzel, C. L., Schuetz, M., Yu, Q., & Mattsson, J. (2007). Dynamics of MONOPTEROS and PIN-FORMED1 expression during leaf vein pattern formation in Arabidopsis thaliana. The Plant Journal, 49(3), 387–398. Williamson, V. M., & Hussey, R. S. (1996). Nematode pathogenesis and resistance in plants. The Plant Cell, 8(10), 1735. Wubben, M. J., Su, H., Rodermel, S. R., & Baum, T. J. (2001). Susceptibility to the sugar beet cyst nematode is modulated by ethylene signal transduction in Arabidopsis thaliana. Molecular Plant–Microbe Interactions, 14(10), 1206–1212. Wu, C. Y., Rolfe, P. A., Gifford, D. K., & Fink, G. R. (2010). Control of transcription by cell size. PLoS Biology, 8(11), e1000523. Xiong, Y., Li, Q. B., Kang, B. H., & Chourey, P. S. (2011). Discovery of genes expressed in basal endosperm transfer cells in maize using 454 transcriptome sequencing. Plant Molec- ular Biology Reporter, 29(4), 835–847. Zhou, W., Wei, L., Xu, J., Zhai, Q., Jiang, H., Chen, R., et al. (2010). Arabidopsis tyrosyl- protein sulfotransferase acts in the auxin/PLETHORA pathway in regulating postem- bryonic maintenance of the root stem cell niche. Plant Cell Online, 22(11), 3692–3709. 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al.

The Power of Omics to Identify Plant Susceptibility Factors and to Study Resistance to Root-knot Nematodes

Javier Cabrera, Marta Barcala, Carmen Fenoll and al., 2015). RKNs secretions interfere with the plant Carolina Escobar* molecular pathways to differentiate specialized cell types into their feeding sites (Truong et al., 2015), called giant Universidad de Castilla-La Mancha, Facultad de Ciencias cells (GCs) as they are far larger than the surrounding cells Ambientales y Bioquímica, Avda. Carlos III s/n, Toledo, in the vascular cylinder (Cabrera et al., 2015). During the Spain development of the GCs, surrounding vascular cells divide *Corresponding Author: [email protected] profusely and cortical cells hypertrophy which, together with GC growth, lead to the formation of a knot or gall in the root that is the typical structure that characterizes the Abstract infection by these nematodes (Escobar et al., 2015). GCs Technology has contributed to the advances on the inside the galls act as transfer cells that sink the nutrients genomic, transcriptomic, metabolomic and proteomic from the plant to help the nematode complete its life cycle analyses of the plant-root-knot nematode (RKN) (Rodiuc et al., 2014). The differentiation of vascular cells interaction. Holistic approaches to obtain expression into highly specialized GCs requires massive profiles, such as cDNA libraries, differential display, q-PCR, transcriptional reprogramming to allow changes in the cell microarray hybridization, massive sequencing, etc., have cycle machinery (de Almeida-Engler et al., 2015), lipid, increased our knowledge on the molecular aspects of the carbohydrate and amino acid metabolisms (Brown et al., interaction and have triggered the development of 2015; Siddique and Grundler, 2015) or cell wall biotechnological tools to control this plague. An important composition (Wieczorek, 2015) among others. limitation, however, has been the difficulty of cross- comparative analysis of these data. The construction of a In the last 10-15 years, technology has assisted the database, NEMATIC, compiling microarray data available advances in the genomic, transcriptomic, metabolomic and in Arabidopsis of the interaction with plant endoparasitic proteomic analyses which have contributed to the huge nematodes facilitated the in silico analysis, but is not increase in the research on the molecular aspects of the sufficient for the handling of "omic" information of different plant-RKNs interaction. Hence, studies based on different plant species. Omics combined with cell isolation omics have tried to decipher those molecular pathways techniques have shed some light on the heterogeneous leading to the successful establishment of the nematode in expression signatures of nematode induced gall tissues, the root and to GCs differentiation, as well as those i.e., plant defences are specifically inhibited in giant cells pathways governing the plant resistance that lead to within the gall aiding the nematode for a successful incompatible plant-RKN interactions. Expectations on the establishment. The natural resistance against RKNs varies development of biotechnological tools based on this from an early hypersensitive reaction before the molecular knowledge to control this plague were, at least establishment of the nematode, to the arrest of gall growth. partly, the fundament of the research. A quick online search The molecular bases of these mechanisms, not fully in the Pubmed library using the keyword "Meloidogyne" understood yet, could disclose powerful targets for the retrieves a higher number of publications with this word in development of biotechnology based tools for nematode the abstract within the last 15 years than in all the 20th control. century (1132 and 1041, respectively). Moreover, 246 of the publications since 2000 contain in the abstract the Introduction words "gene expression". These data reflect the intense Plant parasitic nematodes (PPNs) constitute a major pest research on this field in recent years which has resulted for crop plants worldwide, causing economic and yield from the development of the omics techniques that might losses estimated in $118 billion and 11%, respectively open a new horizon not only for the basic understanding of (McCarter et al., 2008). To these numbers, losses in non- the plant-nematode interaction, but also to design effective food crops including ornamentals, turf and forest trees and economically viable tools for its control. In this review should be added, increasing even further the global impact we describe the main studies based on different omics of these parasites on society and on the environment analyses focusing mainly in those with relevant data on (McCarter et al., 2008). Root- knot nematodes (RKNs; plant defence responses and resistance. Meloidogyne spp.) are sedentary endoparasites considered one of the most harmful group of PPNs as they can infect a Transcriptomic analyses of plant-RKN interactions wide range of plant species (Escobar et al., 2015). RKNs cDNA libraries and differential display penetrate into the roots through the elongation zone and Differential screening of cDNA libraries from infected root migrate intercellulary towards the root apex entering into tissue compared to the cDNA of uninfected roots has been the vascular cylinder where they establish and inducing a broadly used to uncover those genes responsive to the feeding site necessary for life cycle completion (Escobar et infection of different plant species when microarrays or

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 53 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al. massive sequencing techniques were not available (Figure Based also on a cDNA library from 5 weeks-old M. 1; Table S1). First attempts to achieve a global gene incognita tomato galls, Van der Eycken et al. (1996), found expression analysis were performed by Wilson et al., in a gene named LEMMI9, in addition to other genes such as 1994 by constructing a cDNA library from RNA of 1-2 those coding for two extensins. LEMMI9 encodes a late months-old hand dissected GCs induced by M. incognita in embryogenesis related protein whose induction was tomato roots (Figure 1; Table S1). 58 cDNA clones were confirmed in GCs by in situ hybridization. It shows a high more abundant in GCs compared to the uninfected roots homology to cotton LEA14, a hydrophilic protein belonging (Bird and Wilson, 1994; Wilson et al., 1994). Among them, to the late embryogenesis abundant (LEA) class of proteins some were not expressed in uninfected roots but in other that accumulate both during embryogenesis and in parts of the plant, suggesting that the pathways interfered vegetative tissues under water stress (Dure, 1993), and during GC maintenance were not necessarily specific for whose function has been proposed to be ion sequestering roots. Only 12 cDNA clones could be associated to a during water deficit. Hence, it was proposed that the putative homolog using a Blast algorithm, including a accumulation in GCs of LEMMI9 might be related to putative MYB transcription factor (Bird and Wilson, 1994). osmotic changes during their development and functioning More recently, several publications have confirmed the (Van der Eycken et al., 1996). LEMMI9 promoter was also involvement of the MYB family in the regulation of the plant active in potato, and a nematode responsive cis-element responses to different abiotic and biotic stresses (Ambawat was identified by deletions and EMSA analysis (Escobar et et al., 2013) including plant-RKN interaction, with several al., 1999). This gene was one of the firsts to be obtained MYB transcription factors resulting differentially regulated in from transcriptomic studies, and demonstrated to play a different transcriptomic analyses (e.g. Jammes et al., 2005; role in the plant-nematode interaction as evidenced by the Barcala et al., 2010; see next section).

Figure 1. Different omics techniques used to decipher the molecular mechanisms underlying the interaction of Meloidogyne spp. with plants. The use of giant cells, galls and/or infected root tissues from dicotyledonous and monocotyledonous plant species are indicated. We acknowledge Prof. Dr. Godelieve Gheysen and Ji Hongli and Kamrun Nahar from Ghent University for kindly providing us with the pictures of galls and giant cells from rice.

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 54 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al. fact that its silencing in tomato hairy roots produced a As a result of a proteomic study (see next section), a cotton decrease in nematode reproduction (Plovie et al., 2003). protein correlated with RKN resistance was identified. It was DE at 8 dpi when nematode development is arrested Most of the early attempts to disclose the gall/GC in the incompatible interaction (Callahan et al., 1997). transcriptomes were made in well-developed GCs at 1-2 Analysis with a specific antiserum recognized the 14 kDa months after infection, therefore retrieving no information protein and revealed a negative correlation between the on the establishment and early development of the GCs. presence of this protein in root tissues and root galling from The first cDNA library constructed from very early stages of a large group of RKN resistant breeding lines, and the plant-nematode interaction during the penetration- susceptible cultivars of cotton (Callahan et al., 2004). migration phase was performed using 12 hours post Additionally, it helped to identify a gene named MIC-3 inoculation (hpi) nematode-resistant tomato root tips (Meloidogyne Induced Cotton3) from a cDNA library of 10 (carrying Mi-1 gene) infected with M. javanica (Lambert et dpi M. incognita galls that was induced in a nematode- al., 1999). Eight genes from the cDNA library were resistant line of cotton (Gossypium hirsutum; Zhang et al., characterized, some of them associated with defence or 2002) as compared to the expression in a susceptible line stress responses, such as those encoding an ascorbate (Callahan et al., 1997; Zhang et al., 2002) corresponding to reductase or a peroxidase. However, these 8 genes were this protein. More recently, the authors demonstrated that expressed in both the compatible and incompatible the over-expression of the protein in a susceptible line interaction (Lambert et al., 1999). This is in accordance reduces egg production by 60-75% compared to the non- with the production of reactive oxygen species (ROS) at transgenic lines (Wubben et al., 2015). At least 15 MIC-like the early stages, not only in the incompatible but also in the cDNAs are expressed in the roots of susceptible and compatible interaction as confirmed experimentally (Melillo resistant cotton lines (Wubben et al., 2008), differing in the et al., 2011). However, the H2O2 levels were lower in the timing of induction. In resistant lines their maximum compatible than in the incompatible interaction, which expression occurs before the appearance of visible signs of suggests a role of ROS in nematode resistance. resistance, while susceptible lines showed a delayed Accordingly, a cDNA library of M. incognita galls at 3 days expression. This suggests that the MIC gene family is part post inoculation (dpi) in Medicago sativa roots revealed 6 of a root-specific defence response mechanism in cotton induced cDNAs. Two of them, an isoflavone-like protein that operates independently from other known pathways and a metallothionein-like protein have a potential role in such as gossypol biosynthesis, lipid peroxidation, and ROS scavenging (Potenza et al., 2001). In this respect, it PR10 expression (Wubben et al., 2008). MIC-3 constitutes was recently shown that the cyst nematode Heterodera a clear example of how proteomic and transcriptomic schachtii stimulates the formation of ROS by an NADPH- studies could complement each other to assist the oxidase to modulate the distribution of dead cells and to development of biotechnological tools based on allow the proper formation of their feeding site (Siddique et modifications of the plant defence systems against the al., 2014). Differential display was also used to identify RKNs. alterations of gene expression of A. thaliana root galls caused by M. incognita at early infection stages (2, 3, 4, 5 Although some of the previously mentioned papers used and 7 dpi; Vercauteren et al., 2001). The differential root material enriched in GCs for the construction of the expression of six genes was confirmed by DNA gel blot cDNA library, e.g. hand dissected galls, there was a serious hybridization. A general defence mechanism operating technical difficulty to isolate the GCs from the rest of the during this compatible interaction was suggested because gall where they were immersed. GCs are the nourishing of the enhanced expression of genes encoding proteinase cells and their differentiation and continuous maintenance inhibitors and peroxidases, both described during other is strictly necessary for the nematode to complete its life plant-pathogen interactions (e.g. Koiwa et al., 1997; Habib cycle (Hussey et al., 1989). It was not until 2003 when the and Fazili, 2007). Interestingly, most of the induced genes first transcriptomic study from isolated GCs was carried out were also localized by in situ hybridisation within the gall (Wang et al., 2003). In this work, 25 dpi GCs from tomato tissues, among them, two genes encoding a trypsin roots inoculated with M. javanica were isolated using a inhibitor and a peroxidase were expressed mainly in the pressure probe system (Wang et al., 2001; 2003). The endodermal cells surrounding the expanding GCs, which mRNA extracted was further used for differential display can be considered as part of a host wound reaction RT-PCR, finding that 73 genes were up-regulated and 8 (Vercauteren et al., 2001). Their expression pattern in the genes down-regulated (Wang et al., 2003). Many of those outer layers of the galls, but not within the GCs, is in genes were classified into functional categories that reflect accordance, essentially, with the data obtained in a highly the high metabolic rate of the GCs more than the defence sensitive analysis performed later by combining microarray response or process related to the GCs development. analysis with cell biology techniques. This analysis involved These results were not surprising due to the age of the GCs isolation from the rest of the gall tissues in GCs, 25 dpi, when they are fully developed nourishing the Arabidopsis, tomato and Medicago (Barcala et al., 2010; nematode. However, the authors demonstrated the Damiani et al., 2012; Portillo et al., 2013). In general, it was importance of GCs isolation preceding the transcriptomic observed that defence genes were repressed within the analysis, as the fold change of the up- regulated genes GCs, whereas defence responses were observed in the was much lower using GC-enriched root segments than rest of the gall. that from isolated GCs. This confirms the strong dilution effect of GC mRNAs when whole infected tissues are used (Wang et al., 2003). Later, 87 and 54 transcripts were

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 55 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al. identified from a cDNA library of laser captured isolated of GCs as transfer cells to feed the nematode (Cabrera et GCs from tomato galls at 4 and 7 dpi respectively (Fosu- al., 2014a; Rodiuc et al., 2014), an up- regulation of auxin, Nyarko et al., 2009). It was suggested that the presence amino acid and sucrose transporters was observed and initial movement of the nematode juveniles (J2s) (Hammes et al., 2005). The second microarray study between cells may activate defence and stress reported for RKNs was a full time-course along different transduction pathways in the host root, as transcripts infection stages, from establishment to maturation (7, 14 encoding pathogenesis-related proteins (e.g. TSI-1 and and 21 dpi) using CATMA microarrays representing 22089 stress related proteins such as superoxide dismutases) Arabidopsis genes (Jammes et al., 2005). A total of 3373 were induced at 4 and 7 dpi as compared to uninfected genes were differentially expressed (DE) at some of the cells surrounding the vascular cylinder (Fosu-Nyarko et al., selected infection times, increasing greatly the number of 2009). In contrast, most of the later transcriptomic analyses genes found DE in the galls as compared to those found in based on microarrays or massive sequencing combined the previous cDNA library screens. Such a number of DE with laser microdissection of GCs pointed to gene genes indicated a vast molecular reprogramming of the repression of these groups of genes within GCs, not only in cells forming the gall, including the GCs. The number of Arabidopsis but in tomato and Medicago as well (Barcala et genes up and down- regulated were similar (1606 and al., 2010; Daimiani et al., 2012; Portillo et al., 2013). 1742, respectively), highlighting the importance of gene repression in the compatible interaction (Jammes et al., Even though in the last decade researchers preferred 2005). Moreover, the number of repressed genes related to holistic transcriptomic techniques such as microarrays and plant defence was remarkable, suggesting a necessity for high throughput sequencing analyses, cDNA libraries have the down- regulation of defence pathways in the gall for a been used in other studies when no microarray chips were proper establishment during the compatible interaction available, i.e. in the interactions of M. arenaria with Arachis (Jammes et al., 2005). The gene repression in galls was hypogaea and M. javanica with Glycine max (Tirumalaraju further confirmed in later microarray analysis of galls from et al., 2011; de Sa et al., 2012). cDNA libraries from Arabidopsis infected with M. incognita (Fuller et al., 2007) resistant and susceptible peanut lines confirmed the higher where two-thirds of the 259 DE were down-regulated. number of genes related to stress categories in the Those repressed genes were mainly classified in the resistant line than in the susceptible one, although both of response to abiotic or biotic stresses (Fuller et al., 2007). them presented genes from this category. The resistant cultivar (NemaTAM)-specific transcripts included those The first transcriptome combining cell isolation techniques encoding putative PR proteins, patatin-like proteins and and DNA-microarrays was performed with RNA extracted universal stress proteins (USP). The expression of most from GCs at 3 dpi isolated by laser capture micro- PR proteins is contingent upon pathogen infection and is dissection from frozen sections of hand dissected mediated through salicylic acid (SA), jasmonic acid (JA) Arabidopsis-M. javanica galls (Barcala et al., 2010). It and ethylene signal transduction pathways (Tirumalaraju et detected 1161 DE genes, of which 851 were down- al., 2011). In contrast, the number of genes related to regulated and 310 were up-regulated. Consistently, most of transcriptional regulation was higher in the susceptible than the stress related genes were down-regulated in GCs. in the resistant lines, probably reflecting an intense Hence, a distinctive feature of the GCs transcriptome was reprogramming needed for GC establishment (Tirumalaraju the general down-regulation of genes classified in the et al., 2011). secondary metabolism and stress groups. In particular, most genes involved in the phenylpropanoid pathway, Microarrays typically part of a defence mechanism against pathogen The first studies reporting DNA microarrays for the study of attack, including PAL1, PAL2, 4CL3 and CAD (Dixon et al., the plant- Meloidogyne interaction constituted a prologue 2002) were down-regulated. Furthermore, many WRKY, for a veritable explosion of analyses based on this MYB, AP2/EREBP and bZIP transcription factor family technique (Figure 1; Table S1). The sequencing of the members, induced by biotic or abiotic stresses (Bakshi and Arabidopsis genome coupled to the valuable information Oelmüller, 2014; Mizoi et al., 2012; Liu, et al., 2015) were from other research fields, made this plant a suitable down-regulated in GCs, but not in the transcriptome of system to study the transcriptional profiles of galls and whole galls. However, defence and PRP genes directly or GCs. The first study was centred in the regulation of indirectly regulated by SA and JA were not DE in 3 dpi GCs membrane transporters in Arabidopsis roots infected with (Barcala et al., 2010). The study by Barcala et al. (2010) M. incognita at 1, 2 and 4 weeks post inoculation (wpi; allowed comparison of the GC transcriptome to that of the Hammes et al., 2005). Using the GeneChip® Arabidopsis whole gall at the same developmental stage (3 dpi), ATH1 Genome Array from Affymetrix, the authors focused revealing that their transcriptional profiles are different, with their study in the expression of the 805 membrane-proteins only 120 shared genes out of a total of 1161 DE in GCs. A represented in the chip (Hammes et al., 2005). 50 group of distinctive stress and defence related genes were transporters including members of 19 of the approximately clearly induced in galls, mainly related to biotic stress and 80 transporter gene families identified in Arabidopsis secondary metabolism. In contrast, most genes of those showed differential expression at 1, 2 and 4 wpi in whole categories were repressed in GCs. This study also infected roots compared to the uninfected control (Hammes confirmed the high dilution of GC mRNAs when the whole et al., 2005). Some of them were specifically expressed in gall is analysed, as the fold changes of the common DE the galls and not in the rest of the root, as demonstrated by genes were much higher in GCs than in galls. Therefore, it qPCR analysis and GUS assays. Consistent with the role served to characterize the transcriptional changes

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 56 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al. associated to GCs formation, which were proved to be (Table 1). This list of 88 stress and resistance related distinct to that of the rest of gall tissues (Barcala et al., genes is an example of the vast information that 2010). transcriptomics produced in the last decade and the usefulness of tools like NEMATIC, which allows easy cross Vast amount of transcriptomic data for the plant-nematode comparison through several databases and transcriptomes. interaction was generated in the last 10 years. However, Genes listed in Table 1 could be useful for basic studies that quantity of data was not directly translated into a gain and for biotechnology based strategies for the RKN control. of knowledge about gene functions or metabolic pathways within the feeding sites, neither to the development of Microarray based transcriptomic has been successfully genetic tools for biotechnological nematode control. One applied not only in Arabidopsis but also in the crop model reason for this lag was the difficulty of cross-comparative tomato (Solanum lycopersicum; Figure 1; Table S1). analysis of these data. A first attempt to overcome this Pathogenesis-related genes, as defensins, were DE in limitation was the construction of a database called galls induced by M. javanica in tomato roots (Bar-Or et al., NEMATIC (Nematode-Arabidopsis Transcriptomic 2005). The up-regulation of HIN1 (harpin-induced gene 1), Interactive Compendium) that includes most of the induced by bacterial harpins (a class of virulent/avirulent transcriptomic data generated in Arabidopsis related to proteins) in tomato and potato, was suggested as a feeding sites induced by endoparasitic nematodes putative virulence/avirulence mechanism in the RKN- (Cabrera et al., 2014b). This tool was designed for the tomato interaction (Bar-Or et al., 2005). Differences selection of putative target genes, for the development of between the compatible and the incompatible interaction biotechnological tools and to contribute to the basic was studied in more detail using microarray technology knowledge of the plant-nematode interaction. NEMATIC is with resistant and susceptible tomato plants infected with completed with transcriptomic data from several biological M. incognita (Mi-susceptible) and M. hapla (Mi- resistant; processes: plant biotic and abiotic interactions, hormonal Schaff et al., 2007). During a time course study at 12, 36 regulation, specific transcriptomes for different root cell and 72 hpi and 2 wpi, they found 58 genes that exhibited types, etc. (Cabrera et al., 2014b). It allows to compare differential regulation in resistant roots compared to lists of genes commonly or distinctly regulated throughout uninfected roots (Schaff et al., 2007). 31 genes were gall development or in syncytia (induced by cyst induced in the incompatible interaction, encoding proteins nematodes), and/or GCs (Table 1; Table S2). A direct such as a glycosyltransferase, a peroxidase, and an example is the use of NEMATIC for the comparison of all ethylene-responsive protein that were not DE in the stress- related genes contained in Mapman (Thimm et al., susceptible plants at 12 to 36 hpi. In absence of 2004) across all the GC and gall transcriptomes compiled nematodes, the same glycosyltransferase corresponded to in the tool. Applying this filter, a list of 200 genes related to the only gene which was induced in the transcriptome from stress/disease resistance that were DE at some of the resistant (Mi-1/Mi-1) Motelle as compared to the developmental stages of GCs or galls was retrieved (Table susceptible (Mi-) Moneymaker cultivar. Furthermore, the S2). From a biotechnological point of view, the most silencing of this gene restored the susceptibility to interesting defence-related genes are those down- nematodes in the resistant cultivar Motelle, suggesting that regulated at the early stages of GCs or gall development (3 this gene participates in the resistance mechanism. A and 7 dpi). By adding these criteria, the number of genes in similar approach, but focusing in the roles of JA and SA, the list was reduced to 88 (Table 1). Some of them (Table was used by Bhattarai et al., 2008. RNA was extracted 1) are strongly down-regulated, as KTI1, involved in cell from root tips infected at 24 hpi. in cultivar Motelle death modulation during plant-pathogen interactions (Li et (resistant) and cultivar Moneymaker (susceptible). After al., 2008), or PR1, a typical gene induced by different microarray hybridisation, resistant root tips presented twice pathogens and a marker for the systemic acquired as many DE genes than the susceptible line (1497 and resistance (Asai et al., 2014). Other highly down-regulated 750, respectively). Several genes involved in JA genes are PR4, a pathogenesis-related protein that biosynthesis and genes regulated by JA were expressed in participates in the resistance to Fusarium head blight both compatible and incompatible interactions, inferring disease in wheat (Zhuang et al., 2012), and SPX1, a Pi- that the JA signalling pathway had a role in basal defence dependent inhibitor of PHR1 and a master regulator of Pi but not in Mi-1-mediated resistance to RKNs (Bhattarai et starvation responses (Puga et al., 2014). Other genes al., 2008). Two of the transcription factors found, obtained in this selection that have been already related SlWRKY72a and SlWRKY72b, were up-regulated in with disease resistance are HSP90 (Takahashi et al., resistant but not in susceptible roots after RKN infection 2003), TGA1, a redox-controlled regulator of systemic (Bhattarai et al., 2010). The silencing of these genes in acquired resistance (Fobert and Després, 2005), or HS1, resistant plants allowed the infestation by RKN in contrast induced by exposure to SA and JA and with antimicrobial to non-transgenic resistant plants, showing that and antifungal activity (Park et al., 2007). Several disease SlWRKY72a and SlWRKY72b are required for nematode resistance proteins (AT1G58170, AT2G21100, AT4G19920, resistance in Motelle (Bhattarai et al., 2010). Furthermore, AT5G40910, AT5G51630), pathogenesis related proteins their silencing in susceptible Moneymaker plants increased (AT4G25780, AT4G33720, AT4G36010, AT5G40020) and nematode reproduction, indicating that they have a role in germin-like proteins (AT1G10460, AT1G18980, the basal defence response of the plant as well (Bhattarai AT3G05950, AT3G62020, AT4G14630, AT5G39100, et al., 2010). AT5G39150, AT5G39180) are also down regulated in GCs and galls at early and medium stages of development

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 57 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al.

Table 1. List of genes from the stress Mapman category that appear as down- regulated at early and/or medium stages (3 and 7 dpi) of giant cells (GCs) and gall development in the compatible interaction between Arabidopsis thaliana and Meloidogyne javanica or M. incognita. M values for each gene in each of the experiments (Jammes et al., 2005; Barcala et al., 2010) are indicated. "- -" indicates "Non-differentially expressed". M value GC Gall Gall Gene ID 3dpi 3dpi 7dpi Description At1g04430 -- -- -1,07 dehydration-responsive protein-related At1g05850 -1,48 -- -- POM1: Encodes an endo chitinase-like protein AtCTL1. At1g10460 -0,99 -- -- GLP7: germin-like protein (GLP7) At1g11960 -- -- -0,98 early-responsive to dehydration protein-related At1g13930 -1,79 -- -1,18 Involved in response to salt stress. At1g18980 -- -- -0,71 germin-like protein, putative At1g20440 -- -- -2,45 COR47: Belongs to the dehydrin protein family At1g21080 -- -- -0,91 DNAJ heat shock N-terminal domain-containing protein At1g23120 -2,63 -- -- major latex protein-related At1g30360 -- -- -1,61 ERD4 (early-responsive to dehydration 4) At1g31850 -1,67 -- -- dehydration-responsive protein, putative At1g35260 -2,63 -- -- MLP165 (MLP-LIKE PROTEIN 165) At1g52560 -1,54 -- -- 26.5 kDa class I small heat shock protein-like At1g58170 -1,55 -- -- disease resistance-responsive protein-related At1g64140 -- -- -1,11 loricrin-related At1g65390 -1,39 -- -- ATPP2-A5 (ARABIDOPSIS THALIANA PHLOEM PROTEIN 2 A5) At1g70850 -2,52 -- -0,79 MLP34 (MLP-LIKE PROTEIN 34) At1g70880 -- -2,62 -- Bet v I allergen family protein At1g70890 -1,79 -- -- MLP43 | MLP43 (MLP-LIKE PROTEIN 43) At1g72500 -- -- -0,71 inter-alpha-trypsin inhibitor heavy chain-related KTI1: Encodes a trypsin inhibitor involved in modulating programmed cell death in plant-pathogen At1g73260 -3,05 -- -- interactions. At1g73330 -- -- -1,17 DR4: encodes a plant-specific protease inhibitor-like protein At1g75030 -1,17 -- -- TLP-3: encodes a PR5-like protein ERD14: Encodes a dehydrin protein whose expression is induced early on in response to dehydration At1g76180 -- -- -1,85 stress. At1g77260 -2,15 -- -- dehydration-responsive protein-related At1g78040 -1,23 -- -- pollen Ole e 1 allergen and extensin family protein At2g01520 -2,09 -- -- MLP328 (MLP-LIKE PROTEIN 328); copper ion binding At2g01530 -1,42 -- -- MLP329 (MLP-LIKE PROTEIN 329); copper ion binding At2g14610 -4,07 -- -- PR1: PR1 gene expression is induced in response to a variety of pathogens. COR413-PM1: encodes an alpha form of a protein similar to the cold acclimation protein WCOR413 in At2g15970 -1,53 -- -1,59 wheat. At2g19990 -- -- -0,75 PR-1-like protein homolog differentially expressed in resistant cultivars by powdery mildew infection. At2g21100 -1,13 -- -- disease resistance-responsive protein-related At2g21620 -1,87 -- -- RD2: Encodes gene that is induced in response to dessication At2g24550 -- -- -0,83 unknown protein At2g33790 -- -2,26 -- AGP30: pollen Ole e 1 allergen protein At2g39200 -- -- -0,88 MLO12: homolog of the barley mildew resistance locus o (MLO) protein. At2g45130 -- -- -0,69 ATSPX3, SPX3 | SPX3 (SPX DOMAIN GENE 3) At2g47770 -- -- -0,89 TSPO: Encodes a membrane-bound protein designated AtTSPO At3g01420 -- -- -1,97 DOX1: Encodes an alpha-dioxygenase involved in protection against oxidative stress and cell death. At3g04720 -4,15 -- -- PR4: Encodes a protein similar to the antifungal chitin-binding protein hevein from rubber tree latex. At3g05880 -2,12 -- -1,32 RCI2A: Induced by low temperatures, dehydration and salt stress and ABA. At3g05950 -1,23 -- 1,22 germin-like protein, putative At3g10985 -1,82 -1,14 -1,59 SAG20: A senescence-associated gene induced in response to a fungal protein that causes necrosis. At3g17020 -0,88 -- -- universal stress protein (USP) family protein At3g17210 -1,21 -- -- HS1: Encodes a heat stable protein with antimicrobial and antifungal activity. At3g23300 -1,41 -- -- dehydration-responsive protein-related At3g28940 -1,92 -- -- avirulence induced gene (AIG) protein, putative At3g44670 -1,32 -- -- ATP binding / nucleoside-triphosphatase/ nucleotide binding / protein binding / transmembrane receptor At3g47540 -1,40 -1,02 0,69 chitinase, putative At3g48450 -2,29 -- 1,00 nitrate-responsive NOI protein, putative At3g50460 -0,97 -- -- HR2: Homolog of RPW8 At3g50970 -- -- -2,05 LTI30: Belongs to the dehydrin protein family At3g59930 -4,03 -- -- Encodes a defensin-like (DEFL) family protein. At3g62020 -- -1,10 -- GLP10: germin-like protein (GLP10)

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 58 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al.

At3g62190 -1,39 -- -- DNAJ heat shock N-terminal domain-containing protein At4g09940 -1,43 -- -- avirulence induced gene (AIG1) family protein At4g11650 -2,99 -- -- OSM34: osmotin-like protein At4g14630 -- -- -1,90 GLP9: germin-like protein At4g19920 -- -1,16 -- disease resistance protein (TIR class), putative At4g22214 -- -1,27 -- Encodes a defensin-like (DEFL) family protein. At4g23680 -1,41 -- -- major latex protein-related At4g24190 -0,77 -- -- SHD: an ortholog of GRP94, involved in regulation of meristem size and organization. At4g25780 -- -- -0,81 pathogenesis-related protein, putative At4g28240 -1,19 -- -0,68 wound-responsive protein-related At4g28480 -1,13 -- -- DNAJ heat shock family protein At4g30650 -2,85 -- -- low temperature and salt responsive protein, putative At4g33300 -- -- -0,74 ADR1-L1 (ADR1-like 1); ATP binding / protein binding At4g33720 -0,80 -- -- pathogenesis-related protein, putative At4g36010 -- -- -0,82 pathogenesis-related thaumatin family protein At4g36040 -- -- -0,68 J11: DNAJ heat shock N-terminal domain-containing protein (J11) At5g20150 -3,40 -- -- ATSPX1, SPX1 | SPX1 (SPX DOMAIN GENE 1) At5g33355 -- -1,05 -- Encodes a defensin-like (DEFL) family protein. At5g39100 -- -1,40 -- GLP6: germin-like protein (GLP6) At5g39150 -2,33 -1,51 -- germin-like protein, putative At5g39180 -1,77 -1,46 -- germin-like protein, putative At5g39730 -1,81 -- -- avirulence induced gene (AIG) protein-related At5g40020 -- -1,77 -- pathogenesis-related thaumatin family protein At5g40910 -1,18 -- -- disease resistance protein (TIR-NBS-LRR class), putative TTR1: Encodes a member of WRKY Transcription Factor that confers resistance to tobacco ringspot At5g45050 -1,59 -- -- nepovirus. At5g45490 -- -- -1,14 disease resistance protein-related At5g49910 -1,25 -- -- cpHsc70-2: Stromal heat shock protein involved in protein import into chloroplast. At5g51630 -1,25 -- -- disease resistance protein (TIR-NBS-LRR class), putative At5g52060 -- -- -1,05 BAG1: A member of Arabidopsis BAG proteins, plant homologs of mammalian regulators of apoptosis. At5g52640 -- -- -0,99 HSP90.1: Encodes a cytosolic heat shock protein AtHSP90.1. At5g54855 -1,19 -- -- pollen Ole e 1 allergen and extensin family protein At5g64890 -- -- -0,81 PROPEP2 (Elicitor peptide 2 precursor) At5g64930 -0,81 -- -- CPR5: Regulator of expression of pathogenesis-related (PR) genes. At5g65210 -- -- -0,79 TGA1: Encodes TGA1, a redox-controlled regulator of systemic acquired resistance.

The first microarray study performed with RNA extracted number of galls formed. In contrast, a remarkable induction from isolated GCs induced by M. javanica in a model crop, (above nine-fold) was observed for this peroxidase during tomato (S. lycopersicum) was carried out by laser capture the incompatible interaction with a resistant cultivar microdissection of 3 and 7 dpi GCs (Portillo et al., 2013). carrying the Mi-1 gene, S. lycopersicum cv Motelle (Mi-1/ The transcriptomes of the isolated GCs were compared Mi-1) as compared to the susceptible near-isogenic line S. with that of the complete galls at 1, 3, 7 and 14 dpi, lycopersicum cv Moneymaker (Portillo et al., 2013). showing a higher repression state in 3 dpi GCs (86% of the Transcriptomic studies during the plant-nematode DE genes) than at 7 dpi when the GCs are already well interaction have been also performed on a model plant for developed (56% of the DE genes). This repression trend in legume genetics, Medicago truncatula (Damiani et al., GCs at early developing stages was already observed in 2012) challenged with M. incognita. Laser microdissected Arabidopsis (Barcala et al., 2010). Moreover, genes down- GCs were compared with laser microdissected regulated in early developing GCs were robustly conserved neighbouring cells in 7 dpi galls from Medicago using between Arabidopsis and tomato (Portillo et al., 2014). The Affymetrix GeneChip®_Medicago genome (Damiani et al., phenylpropanoid pathway was significantly over- 2012). 36% of the down-regulated genes and 78% of those represented among the repressed genes in both tomato up-regulated in GCs were not expressed in neighbouring and Arabidopsis GCs, particularly genes involved in lignin cells, reinforcing the importance of the genomic studies biosynthesis, such as those coding for a group of which focus specifically in the feeding cells (Damiani et al., peroxidases, together with genes from a biotic stress 2012). Medicago GCs also showed down-regulation of subcategory encoding protease inhibitors (Portillo et al., secondary metabolism and defence related genes as 2013). Infection tests with a tomato line overexpressing the compared to the neighbouring cells, similar to that TPX-1 peroxidase, highly repressed in GCs from occurring in Arabidopsis and tomato (Barcala et al., 2010; Arabidopsis and tomato, showed a 35% of reduction in the Portillo et al., 2013). All these data demonstrate that the

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 59 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al. down-regulation of stress-defence related genes is a resistance genes (R genes) and only one of these induced hallmark of GCs during the compatible interaction. transcripts has a validated counterpart in eggplant, pointing to major sequence divergence and/or lack of counterparts Microarray technology has been successfully used as well in eggplants for this group of induced S. torvum candidate with other species for which genomic information is scarce. R genes (Bagnaresi et al., 2013). Some examples are Glycine max infected with M. incognita during a compatible interaction (Ibrahim et al., 2011). At 12 High throughput sequencing dpi, six of seven members of the lipoxygenase (LOX) gene In the last years, the development of high throughput family were up-regulated as compared to non-infected sequencing techniques opened a new field for the plants. LOX genes are essential for the biosynthesis of transcriptomic analyses of the RKN- plant interaction oxylipins, which have an important function in the plant (Figure 1). Interesting examples are those from monocots defence response against wounding and pathogen attack. as Oryza sativa (Kyndt et al., 2012a; 2014; Ji et al., 2013). Interestingly, some members have been recently reported Consistently with the findings described in previously to have significant roles during plant-nematode discussed microarray experiments, the galls induced by M. interactions. LOX4 has a key role controlling plant defence graminicola in rice roots at 3 dpi showed down-regulation in Arabidopsis when infected with M. javanica and H. of genes related to biotic stresses and to the schachtii, as proved by the fact that loss-of-function phenylpropanoids pathway, responsible for the mutants were more susceptible to nematodes than control biosynthesis of different metabolites, such as lignin plants, and showed increased levels of JA and transcripts precursors, flavonoids and hydroxycinnamic acid esters from the defence related genes encoding allene oxide and SA (Kyndt et al., 2012). However, genes related to synthase, allene oxide cyclase, ethylene-responsive metabolic processes such as protein and sucrose transcription factor 4 and phenylalanine ammonia lyase biosynthesis, were up-regulated (Kyndt et al., 2012a). The (AOS, AOC2, ERF4, PAL1) (Ozalvo et al., 2014). Genes sequencing of the mRNAs obtained from laser captured encoding PR proteins were induced at 12 dpi in soybean microdissected isolated GCs from M. graminicola galls in galls formed by M. incognita after microarray analysis with rice showed the same tendency, with the defence and Affymetrix Soybean GeneChip containing 37,500 probe secondary metabolism related genes being strongly down- sets (Ibrahim et al., 2011). Additionally, although its precise regulated (Ji et al., 2013). Moreover, the gene regulation role on the plant-nematode interaction is not yet known, it trend in GCs was almost the same at 7 and 14 dpi. has been reported that SA induced the expression of PR-1 However, genes related with epigenomic changes as those and SA treatment of tomato plants inoculated with M. coding for AGO or DCL proteins, showed up-regulation at 7 incognita enhanced the synthesis of PR-1, which resulted dpi but not at 14 dpi (Ji et al., 2013). in a significant increase in resistance to the nematode (Nandi et al., 2003). In order to achieve novel insights into incompatible interactions, comprehensive gene expression profiling was Interestingly, the Rk locus from cowpea (Vigna unguiculata) also obtained by pyrosequencing (Beneventi et al., 2013). confers resistance against several species of RKNs, but Transcriptomic studies in soybean lines resistant to specific the mechanism of resistance is substantially different to strains and races of RKNs identified DE transcripts for that of the characterized Mi-1 gene in tomato (Williamson, several members of the lipid transfer protein (LTPs) family. 1999; Mantelin et al., 2013). Rk confers a delayed but For instance, PR-14, an important player in the general strong resistance mechanism without a hypersensitive plant stress response, was induced only in the reaction-mediated cell death process, which allows incompatible interaction whereas PR-1, PR-2 and PR-5 nematode development but blocks reproduction. Gene protein families showed increased expression in the ontology based functional categorization of genes DE in compatible interaction. The data also suggest a key role of RK plants, as compared to a susceptible nearly isogenic glycosyltransferases, auxins and components of gibberellin line, revealed that the typical defence response was related genes in the resistance reaction. In addition, some partially suppressed in resistant roots, even at 9 dpi, overlaps between salt stress responses and this particular allowing nematode juvenile development. However, interaction were also observed, such as the induction of differences in ROS concentrations, induction of toxins and annexins and of negative regulators of JA signalling (JAZ/ other defence related genes seem to play a role in this TIFY). A complex model integrating putative crosstalk unique resistance mechanism (Das et al., 2010). Another mechanisms between plant hormones, mainly gibberellins example of transcriptome analysis on a non-model species and auxins, which can be crucial to modulate the levels of is that of Solanum torvum, used worldwide as rootstock for ROS in the resistance reaction to nematode invasion, has eggplant cultivation because of its vigour and resistance/ been proposed (Beneventi et al., 2013). tolerance to bacterial, fungal wilts and RKNs. The transcriptome was obtained by combining 454 Recently, high throughput sequencing has also been used pyrosequencing and microarray technology and it was to study the transcriptome of the interaction between M. compared to that of the phylogenetically-related nematode- javanica and tomato hairy roots (Iberkleid et al., 2015). This susceptible eggplant species S. melongena. The is one of the few experiments in which the transcriptome of expression profiling of S. torvum challenged with M. a wild type plant is compared with a transgenic plant incognita points to sesquiterpenoids and chitinases as overexpressing a nematode-secreted protein. Local major effectors of nematode resistance. In addition, 16 transcriptional changes were monitored in tomato hairy root transcripts of S. torvum were annotated as disease lines with constitutive expression of the M. javanica fatty

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 60 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al. acid- and retinol-binding protein (Mj-FAR-1), compared with 0001) with GUS expression was clearly activated in galls control roots without inoculation and after 2, 5 and 15 dpi and GCs (Barthels et al., 1997) was selected. Further (Iberkleid et al., 2015). Mj-FAR-1 is involved in nematode Southern analysis and inverse-PCR were used to identify development and reproduction in tomato roots and plants the promoter of the gene WRKY23, as the tagged locus over-expressing Mj-FAR-1 showed increased susceptibility and its induction by nematodes was confirmed by the same to the infection by M. javanica (Iberkleid et al., 2013). A expression pattern of a pWRKY23::GUS line under RKN total of 3970 genes were DE in over-expressing and control infection (Grunewald et al., 2008). Grunewald et al. (2008) lines (2069 up-regulated and 2205 down-regulated), further hypothesised that the induction of WRKY23 in the feeding confirming the down-regulation of the phenylpropanoids cells could led to a reduction in the plant basal defences. and secondary metabolism pathways (Iberkleid et al., The GUS expression in Att0728, another line showing 2015), which might facilitate nematode infection. distinct induction in galls, has been shown to be driven by the promoter of the transcription factor JAZ1/TIFY10A mRNA high throughput sequencing was also used to (Grunewald et al., 2009), a negative regulator of JA- unravel the mechanism of nematode resistance by responsive genes (Thines et al., 2007). sequencing the transcriptome of roots from M. sativa susceptible and resistant lines at 10 dpi after infection with Hormone mediated systemic defence against RKNs has M. incognita (Postnikova et al., 2015). Most of the genes been broadly studied in rice infested with M. graminicola DE were unique to either resistant or compatible cultivars, (Nahar et al., 2011; Kyndt et al., 2012b; Nahar et al., 2013). including a high number of R genes (36 and 121, Defence induced genes, as shown by qPCR, were found to respectively). These results suggest that the presence of R be down-regulated in infected roots at early stages of genes in the susceptible interaction may be part of a development, while they were mostly up-regulated in general or passive defence reaction while those DE in the infected roots under treatment with the systemic defence- incompatible interaction may be responsible for the inducing compounds JA and ethephon (ethylene), resistance (Postnikova et al., 2015). Moreover, they indicating the ability of the nematodes to supress the corroborate as well the significant repression of plant defence pathways (Nahar et al., 2011). This suppression defences in the susceptible line but not in the resistant one was further confirmed by qPCR at 3 dpi, as general (Postnikova et al., 2015). defence genes (OsPR10 and OsPR1a), as well as OsPAL1, OsPR1b, the JA carboxyl methyl transferase gene Small RNAs (sRNAs) DE in the systemic response of the (OsJMT1) and ethylene biosynthesis and signalling genes plants to RKN have been also analysed by high throughput (OsAC07 and OsEIN2a) were down-regulated in shoots of sequencing (Zhao et al., 2015). Sequencing of sRNAs was infected plants while the PR genes were up- regulated in performed on tomato phloem tissue after RKN inoculation roots (Kyndt et al., 2012b). Accordingly, in Arabidopsis in WT and spr2 (suppressor of prosystemin-mediated plants infected with M. incognita, the genes PR1, PR2, response 2, JA-deficient). Interestingly, a negative PR3 and PR5 were induced in roots while PR1 to PR5 correlation was observed between miR319 and its target were down-regulated in leaves (Hamamouch et al., 2011). TCP4 in leaf, stem, and root under RKN stress, implying Similar results were found in tomato, although PR genes that the miR319/TCP4 module is involved in the systemic were described as being down-regulated in roots in defensive response (Zhao et al., 2015). Reverse genetics addition to shoots, but a strong up-regulation of PR genes also demonstrated that the miR319/TCP4 module affected was found in the shoots of resistant tomato genotypes JA synthetic genes and the endogenous JA level in leaves, (Molinari et al., 2014). thereby mediating RKN resistance (Zhao et al., 2015). Genome-wide analysis of the HSP20 gene family in Other approaches and tools soybean after M. incognita infection revealed six Apart from the holistic techniques previously described, responsive GmHSP20 genes by qRT-PCR (Lopes-Caitar et other molecular tools have been used to study gene al., 2013). Some of them were down-regulated in a expression during gall formation. This is the case for real- susceptible line but up- regulated in the resistant genotype. time quantitative PCR (qPCR) to measure specific The proteins encoded by these genes are promising transcripts, GUS reporter lines or promoter trap lines to targets for developing crop varieties that are better adapted check promoter activity in feeding sites, or in situ RT-PCR, to biotic and abiotic stresses. QPCR analysis has also which have been successfully used and reported in many pointed to the down-regulation of thionins, peptides articles to show the regulation of genes or to validate those involved in plant defence, in galls induced by M. results obtained by holistic studies. In this section, due to graminicola at early and medium stages of development (Ji the vast amount of publications related to the plant- RKN et al., 2015). Addtionally, the expression of six PAL genes interaction using these techniques, we will only discuss related to the phenylpropanoids pathway, in three maize some of those which describe the regulation of defence- or genotypes that were good, moderate, and poor hosts for M. stress- related groups or families of genes. One of the incognita showed that ZmPAL4 was most strongly pioneer analysis using genomic tools to uncover the expressed in the most-resistant maize line (Starr et al., regulation of gene expression from a holistic point of view 2014), suggesting a role for this pathway in the defence was the large-scale screening of 1472 Arabidopsis against the RKNs as previously commented. promoter tag lines, harbouring a promoter-less GUS gene, challenged with cyst or with RKNs (Barthels et al., 1997). The above compilation represents only a brief selection of From this screen, the line Att0001 (Arabidopsis thaliana tag the successful use of techniques as qPCR and GUS

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 61 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al. reporter lines to study the defence mechanisms of the ferredoxin-NADP reductase isozyme 2, and glutathione S- plants against the root-knot nematodes. There are a transferase (Villeth et al., 2015). Some of them, like the multitude of studies, however, which are based on the use hydroxyacid oxidase and the gamma-type carbonic of these techniques to decipher the expression of individual anhydrase, probably participate in the hypersensitive genes that could be related with the plant- nematode response of the highly resistant cultivar. The authors also interaction. found resistance related proteins such as patatin or the resistance associated protein RPP13 (Villeth et al., 2015). Proteomic and Metabolomic analyses on plant-RKN At later stages of development (35-40 dpi), 2D interaction electrophoresis has been also used to unravel the Proteomic and metabolomic profiles of plants infected with proteome of the tripartite interaction M. artiellia-Fusarium RKN are very scarce as compared to the literature oxysporum, in fungus resistant cultivars of Cicer arietinum regarding transcriptomic studies. Moreover, traditional (chickpea) (Palomares-Rius et al., 2011). One of the model plants as tomato and Arabidopsis have not been cultivars lost the resistance to F. oxysporum after M. used for this purpose, which somehow limits the possibility artiellia interaction. Class I quitinases were significantly to combine the data with the genomic information available induced in the cultivar maintaining F. oxysporum (Foc-5) in different databases to gain further molecular knowledge resistance after M. artiellia infection. This differential of the interaction. One of the first proteomic studies was response could thus be interpreted as a key determinant of made in a Medicago sativa cultivar resistant to M. the maintenance of Foc-5-resistance after infection with incognita, Moapa 69, which lacks the typical hypersensitive both pathogens. Additionally, a class III peroxidase and a R response identified in tomato and tobacco (Potenza et al, protein of the NBS-LRR class increased 5.5-fold and 4 fold, 1996). Differences in J2 behaviour between the susceptible respectively in the cultivar that lost the F. oxysporum and resistant cultivars first appeared 48 to 72 hours after resistance, after Meloidogyne artiellia infection compared infection, coinciding with the entry of J2 into the vascular with the control, but did not change after infection by either cylinder of the susceptible line Lahontan. Bidimensional pathogen alone. The class III peroxidase and catalase level electrophoresis (2D electrophoresis) of translation products changes reinforce the idea that redox perturbation is of RNA from infected roots of both cultivars revealed no associated with the chickpea response to M. artiella specific transcript variation in the resistant cultivar as infection. compared to that of the susceptible one and to a general mechanical wounding response (Potenza et al., 1996). In In this respect, an antioxidant molecule involved in the contrast, another pioneer study was performed via one-and cellular redox homeostasis is glutathione and two dimensional polyacrylamide gel electrophoresis homoglutathione, (h)GSH. Its crucial role in plant signalling (PAGE) on a resistant cotton line that also showed under biotic stress is well known (Dubreuil-Maurizi and nematode development arrest soon after infection Poinssot, 2012). The (h)GSH metabolism is modified in (Callahan et al.,1997). From this study, a highly abundant nematode-induced root galls as the (h)GSH content is protein that temporally correlated with the resistance significantly higher in galls than in roots at late infection response was detected and later identified as MIC-3 stages. A detailed analysis of glutathione (GSH) and (Meloidogyne Induced Cotton3) whose over-expression homoglutathione in galls, (hGSH), demonstrated the reduces egg production by 60-75% (Wubben et al., 2015). importance of these compounds for the success of M. incognita infection in Medicago truncatula. Moreover, the Franco et al. (2010) studied the proteome of resistant impairment of nematode development and reproduction in cotton and coffee cultivars after M. paranaensis and M. (h)GSH-depleted galls was linked to a modified carbon incognita infection using 2D electrophoresis. They found 4 metabolism at the feeding site (Baldacci-Cresp et al., up and 5 down-regulated proteins in cotton infected roots 2012). Hence, a substantial modification of starch and γ- at 6 dpi, and 13 up and 2 down-regulated at 10 dpi together aminobutyrate metabolism and of malate and glucose with 7 induced / 3 repressed proteins at 6 dpi in coffee content in (h)GSH-depleted galls compared to control roots (Franco et al., 2010). Two of those proteins from coffee and was detected by untargeted proton Nuclear Magnetic one from cotton were identified as a chitinase, a Resonance. Eighteen of the 37 quantified metabolites were pathogenesis related protein and a quinone reductase 2, related to sugar, organic acid and amino acid metabolism. respectively, all related to disease resistance (Franco et al., GABA was specifically detected in (h)GSH-depleted galls 2010). In this respect, overexpression of quinone (Baldacci-Cresp et al., 2012) and it is known to accumulate reductases in Arabidopsis caused hypersensitive reaction in the plant in response to abiotic and biotic stresses to necrotrophic fungi (Heyno et al., 2013). Likewise, when (Kinnersley and Turano, 2000). The reproduction of M. comparing a highly resistant and a slightly resistant hapla in transgenic plants accumulating GABA is lower cowpea line at 3, 6 and 9 dpi after challenging both with M. than in control plants (McLean et al., 2003). Although incognita, significant differential expression profiles were details regarding how (h)GSH content can modulate observed. Interestingly, 13 unique proteins were identified metabolism are still lacking, there is evidence suggesting in the highly resistant cultivar compared to the slightly that the control of the plant cell redox status through the resistant line (Villeth et al., 2015). Several of the induced modification of the (h)GSH content may be a key regulator proteins are associated to the oxidative stress responses in of the GC potential to sustain nematode feeding. the highly resistant cultivar, including a multicatalytic endopeptidase complex (proteasome), hydroxyacid oxidase, gamma-type carbonic anhydrase family protein,

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 62 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al.

Concluding Remarks Spain. We acknowledge Prof. Dr. Godelieve Gheysen and Meloidogyne spp. is one of the most devastating Ji Hongli and Kamrun Nahar from Ghent University for pathogens for plants. Technology has contributed kindly provide us with the pictures of galls and giant cells massively to the advances on genomic, transcriptomic, from rice in Figure 1, Alberto de Marcos from University of metabolomic and proteomic analyses of plant-Meloidogyne Castilla- La Mancha for the plant designs in Figure 1 and interactions, increasing the knowledge on the molecular Celia Martin for the reading of the manuscript. aspects of the RKNs interaction in the last 10-15 years. Studies based on different omics tried to decipher the References molecular pathways leading to the successful establishment of the nematode in the root and to GC Ambawat, S., Sharma, P., Yadav, N.R., and Yadav, R.C. differentiation, as well as those governing the plant (2013). MYB transcription factor genes as regulators for resistance leading to incompatible plant-RKN interactions. plant responses: an overview. Physiol Mol Biol Plants 19, Different holistic approaches to obtain gene expression 307-321. profiles, such as cDNA libraries, differential display, q-PCR, Asai, S., Rallapalli, G., Piquerez, S.J.M., Caillaud, M.-C., hybridization of microarrays, massive sequencing etc., Furzer, O.J., Ishaque, N., Wirthmueller, L., Fabro, G., anticipated the development of biotechnological tools Shirasu, K., and Jones, J.D.G. (2014). Expression based on this molecular knowledge to control this plague. Profiling during Arabidopsis/Downy Mildew Interaction However, the vast amount of transcriptomic data for the Reveals a Highly-Expressed Effector That Attenuates plant-nematode interaction has not always correlated Responses to Salicylic Acid. PLoS Pathog 10, e1004443. directly with an increase in the identification of genes Bagnaresi, P., Sala, T., Irdani, T., Scotto, C., Lamontanara, functionality or metabolic pathways within the feeding sites, A., Beretta, M., Rotino, G.L., Sestili, S., Cattivelli, L., and neither with an enhancement of genetic tools for Sabatini, E. (2013). Solanum torvum responses to the biotechnological purposes leading to nematode control. root-knot nematode Meloidogyne incognita. BMC One of the limitations has typically been the difficulty of Genomics 14, 540. cross- comparative analysis of these data. A first attempt to Bakshi, M., and Oelmüller, R. (2014). WRKY transcription achieve this objective was the construction of a database, factors. Plant Signaling and Behavior 9, e27700. NEMATIC, compiling microarray data available in Baldacci-Cresp, F., Chang, C., Maucourt, M., Deborde, C., Arabidopsis for the interaction with plant endoparasitic Hopkins, J., Lecomte, P., Bernillon, S., Brouquisse, R., nematodes. It also integrates transcriptomic data from Moing, A., Abad, P., et al. (2012). (Homo)glutathione several biological processes, plant biotic and abiotic deficiency impairs root-knot nematode development in interactions, hormonal regulation, specific transcriptomes Medicago truncatula. PLoS Pathog 8, e1002471. for different root cell types, etc., but is not sufficient for the Bar-Or, C., Kapulnik, Y., and Koltai, H. (2005). A broad handling of "omic" information of different plant species, characterization of the transcriptional profile of the thus it should be expanded to other plant species. compatible tomato response to the plant parasitic root knot nematode Meloidogyne javanica. European Journal The main information obtained from the RKN interaction of Plant Pathology 111, 181-192. based on omics is that plant defence shut down at early Barcala, M., Garcia, A., Cabrera, J., Casson, S., Lindsey, infection stages within GCs should serve the nematode for K., Favery, B., Garcia-Casado, G., Solano, R., Fenoll, C., its successful establishment and development. Omics and Escobar, C. (2010). Early transcriptomic events in combined with cell isolation techniques have also shed microdissected Arabidopsis nematode-induced giant some light on the heterogeneous expression signatures of cells. Plant J 61, 698-712. nematode induced gall tissues, showing contrasted Barthels, N., van der Lee, F.M., Klap, J., Goddijn, O.J., expression profiles in the GCs compared to the rest of the Karimi, M., Puzio, P., Grundler, F.M., Ohl, S.A., Lindsey, tissues. The natural resistance gene Mi-1 and their K., Robertson, L., et al. (1997). Regulatory sequences of homologues, are based in the generation of a typical Arabidopsis drive reporter gene expression in nematode hypersensitive reaction, modulating plant defence feeding structures. Plant Cell 9, 2119-2134. pathways before the establishment of the nematode. Beneventi, M.A., da Silva, O.B., Jr., de Sa, M.E., Firmino, However, there are also other resistance modes in other A.A., de Amorim, R.M., Albuquerque, E.V., da Silva, crops that allow the arrest of gall formation at early stages M.C., da Silva, J.P., Campos Mde, A., Lopes, M.J., et al. of development without a typical hypersensitive reaction. (2013). Transcription profile of soybean-root-knot Omics are contributing to increase the knowledge on the nematode interaction reveals a key role of phythormones molecular bases of nematode resistance. This is an open in the resistance reaction. BMC Genomics 14, 322. field still to be fully uncovered that could help generating Bhattarai, K.K., Atamian, H.S., Kaloshian, I., and Eulgem, powerful targets to the development of biotechnology T. (2010). WRKY72-type transcription factors contribute based tools for nematode control. to basal immunity in tomato and Arabidopsis as well as gene-for-gene resistance mediated by the tomato R gene Acknowledgements Mi-1. Plant J 63, 229-240. This work was supported by the Spanish Government Bhattarai, K.K., Xie, Q.G., Mantelin, S., Bishnoi, U., Girke, (AGL2013-48787 to C. Escobar, and CSD2007-057 and T., Navarre, D.A., and Kaloshian, I. (2008). Tomato PCIN-2013-053 to C. Fenoll) and by the Castilla-la Mancha susceptibility to root-knot nematodes requires an intact Government (PEII-2014-020-P to CF). Javier Cabrera was jasmonic acid signaling pathway. Mol Plant Microbe supported by a fellowship from the Ministry of Education, Interact 21, 1205-1214.

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 63 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al.

Bird, D.M., and Wilson, M.A. (1994). DNA sequence and Escobar, C., Barcala, M., Cabrera, J., and Fenoll C. (2015). expression analysis of root-knot nematode-elicited giant Overview of Root-Knot Nematodes and Giant Cells. In cell transcripts. Mol Plant Microbe Interact 7, 419-424. Advances in Botanical Research Volume 73, C. Escobar Brown, S., Ionit, I., Buki, P., and Kolomiets, M. (2015). The and C. Fenoll, eds. (Oxford, UK: Elsevier Academic Role of Lipid Signalling in Regulating Plant-Nematode Press), pp. 1-32. Interactions. In Advances in Botanical Research Volume Escobar, C., De Meutter, J., Aristizábal, F.A., Sanz-Alférez, 73, C. Escobar and C. Fenoll, eds. (Oxford, UK: Elsevier S., del Campo, F.F., Barthels, N., Van der Eycken, W., Academic Press), pp. 139-166. Seurinck, J., van Montagu, M., Gheysen, G., et al. Cabrera, J., Barcala, M., Fenoll, C., and Escobar, C. (1999). Isolation of the LEMMI9 Gene and Promoter (2014a). Transcriptomic signatures of transfer cells in Analysis During a Compatible Plant-Nematode early developing nematode feeding cells of Arabidopsis Interaction. Molecular Plant-Microbe Interactions 12, focused on auxin and ethylene signaling. Front Plant Sci 440-449. 5, 107. Fobert, P.R., and Després, C. (2005). Redox control of Cabrera, J., Bustos, R., Favery, B., Fenoll, C., and systemic acquired resistance. Current Opinion in Plant Escobar, C. (2014b). NEMATIC: a simple and versatile Biology 8, 378-382. tool for the in ilico analysis of plant-nematode Fosu-Nyarko, J., Jones, M.G., and Wang, Z. (2009). interactions. Molecular Plant Pathology 15, 627-636. Functional characterization of transcripts expressed in Cabrera, J., Diaz-Manzano, F.E., Barcala, M., Arganda- early-stage Meloidogyne javanica-induced giant cells Carreras, I., de Almeida-Engler, J., Engler, G., Fenoll, C., isolated by laser microdissection. Mol Plant Pathol 10, and Escobar, C. (2015). Phenotyping nematode feeding 237-248. sites: three-dimensional reconstruction and volumetric Franco, O.L., Pereira, J.L., Costa, P.H., Rocha, T.L., measurements of giant cells induced by root-knot Albuquerque, E.V., Grossi-de-Sa, M.F., Carneiro, R.M., nematodes in Arabidopsis. New Phytol 206, 868-880. Carneiro, R.G., and Mehta, A. (2010). Methodological Callahan, F.E., Jenkins, J.N., Creech, R.G., and Lawrence, evaluation of 2-DE to study root proteomics during G.W. (1997). Changes in root cotton proteins correlated nematode infection in cotton and coffee plants. Prep with resistance to root knot nematode development. J Biochem Biotechnol 40, 152-163. Cot Sci 1, 38-47. Fuller, V.L., Lilley, C.J., Atkinson, H.J., and Urwin, P.E. Callahan, F.E., Zhang, X.D., Ma, D.P., Jenkins, J.N., (2007). Differential gene expression in Arabidopsis Hayes, R.W., and Tucker, M.L. (2004). Comparison of following infection by plant-parasitic nematodes MIC-3 protein accumulation in response to root-knot Meloidogyne incognita and Heterodera schachtii. Mol nematode infection in cotton lines displaying a range of Plant Pathol 8, 595-609. resistance levels. J Cot Sci 8, 186-190. Grunewald, W., Karimi, M., Wieczorek, K., Van de Damiani, I., Baldacci-Cresp, F., Hopkins, J., Andrio, E., Cappelle, E., Wischnitzki, E., Grundler, F., Inze, D., Balzergue, S., Lecomte, P., Puppo, A., Abad, P., Favery, Beeckman, T., and Gheysen, G. (2008). A role for B., and Herouart, D. (2012). Plant genes involved in AtWRKY23 in feeding site establishment of plant- harbouring symbiotic rhizobia or pathogenic nematodes. parasitic nematodes. Plant Physiol 148, 358-368. New Phytol 194, 511-522. Grunewald, W., Vanholme, B., Pauwels, L., Plovie, E., Inze, Das, S., Ehlers, J.D., Close, T.J., and Roberts, P.A. (2010). D., Gheysen, G., and Goossens, A. (2009). Expression Transcriptional profiling of root-knot nematode induced of the Arabidopsis jasmonate signalling repressor JAZ1/ feeding sites in cowpea (Vigna unguiculata L. Walp.) TIFY10A is stimulated by auxin. EMBO Rep 10, 923-928. using a soybean genome array. BMC Genomics 11, 480. Habib, H., and Fazili, K.M. (2007). Plant protease de Almeida-Engler, J., Vieira, P., Rodiuc, N., Grossi de Sa, inhibitors: a defense strategy in plants. Biotechnology M.F., and Engler, G. (2015). The Plant Cell Cycle and Molecular Biology Reviews. 2, 068-085. Machinery: Usurped and Modulated by Plant-Parasitic Hamamouch, N., Li, C., Seo, P.J., Park, C.M., and Davis, Nematodes. In Advances in Botanical Research Volume E.L. (2011). Expression of Arabidopsis pathogenesis- 73, C. Escobar and C. Fenoll, eds. (Oxford, UK: Elsevier related genes during nematode infection. Mol Plant Academic Press), pp. 91-118. Pathol 12, 355-364. de Sa, M.E., Conceicao Lopes, M.J., de Araujo Campos, Hammes, U.Z., Schachtman, D.P., Berg, R.H., Nielsen, E., M., Paiva, L.V., Dos Santos, R.M., Beneventi, M.A., Koch, W., McIntyre, L.M., and Taylor, C.G. (2005). Firmino, A.A., and de Sa, M.F. (2012). Transcriptome Nematode-induced changes of transporter gene analysis of resistant soybean roots infected by expression in Arabidopsis roots. Mol Plant Microbe Meloidogyne javanica. Genet Mol Biol 35, 272-282. Interact 18, 1247-1257. Dixon, R.A., Achnine, L., Kota, P., Liu, C.-J., Reddy, M.S.S., Heyno, E., Alkan, N., and Fluhr, R. (2013). A dual role for and Wang, L. (2002). The phenylpropanoid pathway and plant quinone reductases in host-fungus interaction. plant defence—a genomics perspective. Molecular Plant Physiol Plant 149, 340-353. Pathology 3, 371-390. Iberkleid, I., Sela, N., and Brown Miyara, S. (2015). Dubreuil-Maurizi, C., and Poinssot, B. (2012). Role of Meloidogyne javanica fatty acid- and retinol-binding glutathione in plant signaling under biotic stress. Plant protein (Mj-FAR-1) regulates expression of lipid-, cell Signal Behav 7, 210-212. wall-, stress- and phenylpropanoid-related genes during Dure, L. (1993). The LEA proteins of higher plants. In nematode infection of tomato. BMC Genomics 16, 272. Control of Plant Gene Expression. D. P. S. Verma, ed. Iberkleid, I., Vieira, P., de Almeida Engler, J., Firester, K., (Cyty, Country: CRC Press), pp. 325-335. Spiegel, Y., and Horowitz, S.B. (2013). Fatty acid-and

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 64 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al.

retinol-binding protein, Mj-FAR-1 induces tomato host soybean: comprehensive sequence, genomic susceptibility to root-knot nematodes. PLoS One 8, organization and expression profile analysis under e64586. abiotic and biotic stresses. BMC Genomics 14, 577. Ibrahim, H.M., Hosseini, P., Alkharouf, N.W., Hussein, E.H., Mantelin, S., Bhattarai, K.K., Jhaveri, T.Z., and Kaloshian, I. Gamal El-Din Ael, K., Aly, M.A., and Matthews, B.F. (2013). Mi-1-Mediated Resistance to Meloidogyne (2011). Analysis of gene expression in soybean (Glycine incognita in Tomato May Not Rely on Ethylene but max) roots in response to the root knot nematode Hormone Perception through ETR3 Participates in Meloidogyne incognita using microarrays and KEGG Limiting Nematode Infection in a Susceptible Host. PLoS pathways. BMC Genomics 12, 220. ONE 8, e63281. Jammes, F., Lecomte, P., de Almeida-Engler, J., Bitton, F., McCarter, J.P. (2008). Molecular approaches toward Martin-Magniette, M.L., Renou, J.P., Abad, P., and resistance to plant-parasitic nematodes. In Cell Biology Favery, B. (2005). Genome-wide expression profiling of of Plant Nematode Parasitism, R. H. Berg and C. G. the host response to root-knot nematode infection in Taylor, eds. (Berlin, Germany: Springer), pp. 239-267. Arabidopsis. Plant J 44, 447-458. McLean, M.D., Yevtushenko, D.P., Deschene, D., Van Ji, H., Gheysen, G., Denil, S., Lindsey, K., Topping, J.F., Cauwenberghe, O.R., Makhmoudova, A., Potter, J.W., Nahar, K., Haegeman, A., De Vos, W.H., Trooskens, G., Bown, A.W., and Shelp BJ (2003) Overexpression of Van Criekinge, W., et al. (2013). Transcriptional analysis glutamate decarboxylase in transgenic tobacco plants through RNA sequencing of giant cells induced by confers resistance to the northern root-knot nematode. Meloidogyne graminicola in rice roots. J Exp Bot 64, Mol. Breed. 11, 277-285 3885-3898. Melillo, M.T., Leonetti, P., Leone, A., Veronico, P., and Ji, H., Gheysen, G., Ullah, C., Verbeek, R., Shang, C., De Bleve-Zacheo, T. (2011) ROS and NO production in Vleesschauwer, D., Hofte, M., and Kyndt, T. (2015). The compatible and incompatible tomato-Meloidogyne role of thionins in rice defence against root pathogens. incognita interactions. Eur J Plant Pathol 130, 489-502. Mol Plant Pathol. Mizoi, J., Shinozaki, K., and Yamaguchi-Shinozaki, K. Kinnersley, A.M., and Turano, F.J. (2000). Gamma (2012). AP2/ERF family transcription factors in plant Aminobutyric Acid (GABA) and Plant Responses to abiotic stress responses. Biochimica et Biophysica Acta Stress. Critical Reviews in Plant Sciences 19, 479-509. (BBA) - Gene Regulatory Mechanisms 1819, 86-96. Koiwa, H., Bressan, R.A., and Hasegawa, P.M. (1997). Molinari, S., Fanelli, E., and Leonetti, P. (2014). Expression Regulation of protease inhibitors and plant defense. of tomato salicylic acid (SA)-responsive pathogenesis- Trends in Plant Science 2, 379-384. related genes in Mi-1-mediated and SA-induced Kyndt, T., Denil, S., Haegeman, A., Trooskens, G., Bauters, resistance to root-knot nematodes. Mol Plant Pathol 15, L., Van Criekinge, W., De Meyer, T., and Gheysen, G. 255-264. (2012a). Transcriptional reprogramming by root knot and Nahar, K., Kyndt, T., De Vleesschauwer, D., Hofte, M., and migratory nematode infection in rice. New Phytol 196, Gheysen, G. (2011). The jasmonate pathway is a key 887-900. player in systemically induced defense against root knot Kyndt, T., Fernandez, D., and Gheysen, G. (2014). Plant- nematodes in rice. Plant Physiol 157, 305-316. parasitic nematode infections in rice: molecular and Nahar, K., Kyndt, T., Hause, B., Hofte, M., and Gheysen, G. cellular insights. Annu Rev Phytopathol 52, 135-153. (2013). Brassinosteroids suppress rice defense against Kyndt, T., Nahar, K., Haegeman, A., De Vleesschauwer, D., root-knot nematodes through antagonism with the Hofte, M., and Gheysen, G. (2012b). Comparing jasmonate pathway. Mol Plant Microbe Interact 26, systemic defence-related gene expression changes upon 106-115. migratory and sedentary nematode attack in rice. Plant Nandi, B., Sukul, N.C., Banewee, N., Sengupta, S., Das, P., Biol (Stuttg) 14 Suppl 1, 73-82. and Sinha BSP (2003). Induction of pathogenesis-related Hussey, R.S. (1989). Disease-inducing secretions of plant- protein by salicylic acid and resistance to root-knot parasitic nematodes. Annual Review of Phytopathology. nematode in tomato. Indian J. Nematology. 33,111-116. 27,123-141. Ozalvo, R., Cabrera, J., Escobar, C., Christensen, S.A., Lambert, K.N., Ferrie, B.J., Nombela, G., Brenner, E.D., Borrego, E.J., Kolomiets, M.V., Castresana, C., Iberkleid, and Williamson, V.M. (1999). Identification of genes I., and Brown Horowitz, S. (2014). Two closely related whose transcripts accumulate rapidly in tomato after members of Arabidopsis 13-lipoxygenases (13-LOXs), root-knot nematode infection. Physiological and LOX3 and LOX4, reveal distinct functions in response to Molecular Plant Pathology 55, 341-348. plant-parasitic nematode infection. Mol Plant Pathol 15, Li, J., Brader, G., and Palva, E.T. (2008). Kunitz Trypsin 319-332. Inhibitor: An Antagonist of Cell Death Triggered by Palomares-Rius, J.E., Castillo, P., Navas-Cortes, J.A., Phytopathogens and Fumonisin B1 in Arabidopsis. Jimenez-Diaz, R.M., and Tena, M. (2011). A proteomic Molecular Plant 1, 482-495. study of in-root interactions between chickpea Liu, J., Osbourn, A., and Ma, P. (2015). MYB Transcription pathogens: the root-knot nematode Meloidogyne artiellia Factors as Regulators of Phenylpropanoid Metabolism in and the soil-borne fungus Fusarium oxysporum f. sp. Plants. Molecular Plant 8, 689-708. ciceris race 5. J Proteomics 74, 2034-2051. Lopes-Caitar, V.S., de Carvalho, M.C., Darben, L.M., Park, S.-C., Lee, J.R., Shin, S.-O., Park, Y., Lee, S.Y., and Kuwahara, M.K., Nepomuceno, A.L., Dias, W.P., Hahm, K.-S. (2007). Characterization of a heat-stable Abdelnoor, R.V., and Marcelino-Guimaraes, F.C. (2013). protein with antimicrobial activity from Arabidopsis Genome-wide analysis of the Hsp20 gene family in

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 65 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al.

thaliana. Biochemical and Biophysical Research Takahashi, A., Casais, C., Ichimura, K., and Shirasu, K. Communications 362, 562-567. (2003). HSP90 interacts with RAR1 and SGT1 and is Paulson, R.E., and Webster, J.M. (1972). Ultrastructure of essential for RPS2-mediated disease resistance in the hypersensitive reaction in roots of tomato, Arabidopsis. Proceedings of the National Academy of Lycopersicon esculentum L., to infection by the root-knot Sciences of the United States of America 100, nematode, Meloidogyne incognita. Physiological Plant 11777-11782. Pathology 2, 227-234. Thimm, O., Blasing, O., Gibon, Y., Nagel, A., Meyer, S., Plovie, E., De Buck, S., Goeleven, E., Tanghe, M., Kruger, P., Selbig, J., Muller, L.A., Rhee, S.Y., and Stitt, Vercauteren, I., and Gheysen, G. (2003). Hairy roots to M. (2004). MAPMAN: a user-driven tool to display test for transgenic nematode resistance: think twice. genomics data sets onto diagrams of metabolic Nematology 5, 831-841. pathways and other biological processes. Plant J 37, Portillo, M., Cabrera, J., Lindsey, K., Topping, J., Andres, 914-939. M.F., Emiliozzi, M., Oliveros, J.C., Garcia-Casado, G., Thines, B., Katsir, L., Melotto, M., Niu, Y., Mandaokar, A., Solano, R., Koltai, H., et al. (2013). Distinct and Liu, G., Nomura, K., He, S.Y., Howe, G.A., and Browse, conserved transcriptomic changes during nematode- J. (2007). JAZ repressor proteins are targets of the induced giant cell development in tomato compared with SCF(COI1) complex during jasmonate signalling. Nature Arabidopsis: a functional role for gene repression. New 448, 661-665. Phytol 197, 1276-1290. Tirumalaraju, S.V., Jain, M., and Gallo, M. (2011). Postnikova, O.A., Hult, M., Shao, J., Skantar, A., and Differential gene expression in roots of nematode- Nemchinov, L.G. (2015). Transcriptome analysis of resistant and -susceptible peanut (Arachis hypogaea) resistant and susceptible alfalfa cultivars infected with cultivars in response to early stages of peanut root-knot root-knot nematode Meloidogyne incognita. PLoS One nematode (Meloidogyne arenaria) parasitization. J Plant 10, e0118269. Physiol 168, 481-492. Potenza, C., Thomas, S.H., and Sengupta-Gopalan, C. Truong, N.M., Nguyen, C.N., Abad, P., Quentin, M., and (2001). Genes induced during early response to Favery, B. (2015). Function of Root-Knot Nematode Meloidogyne incognita in roots of resistant and Effectors and Their Targets in Plant Parasitism. In susceptible alfalfa cultivars. Plant Sci 161, 289-299. Advances in Botanical Research Volume 73, C. Escobar Potenza, C.L., Thomas, S.H., Higgins, E.A., and Sengupta- and C. Fenoll, eds. (Oxford, UK: Elsevier Academic Gopalan, C. (1996). Early Root Response to Press), pp. 293-324. Meloidogyne incognita in Resistant and Susceptible Van der Eycken, W., de Almeida Engler, J., Inze, D., Van Alfalfa Cultivars. J Nematol 28, 475-484. Montagu, M., and Gheysen, G. (1996). A molecular study Puga, M.I., Mateos, I., Charukesi, R., Wang, Z., Franco- of root-knot nematode-induced feeding sites. Plant J 9, Zorrilla, J.M., de Lorenzo, L., Irigoyen, M.L., Masiero, S., 45-54. Bustos, R., Rodríguez, J., et al. (2014). SPX1 is a Vercauteren, I., Van Der Schueren, E., Van Montagu, M., phosphate-dependent inhibitor of PHOSPHATE and Gheysen, G. (2001). Arabidopsis thaliana Genes STARVATION RESPONSE 1 in Arabidopsis. Expressed in the Early Compatible Interaction with Root- Proceedings of the National Academy of Sciences 111, Knot Nematodes. Molecular Plant-Microbe Interactions 14947-14952. 14, 288-299. Rodiuc, N., Vieira, P., Banora, M.Y., and de Almeida Engler, Villeth, G.R., Carmo, L.S., Silva, L.P., Fontes, W., J. (2014). On the track of transfer cell formation by Grynberg, P., Saraiva, M., Brasileiro, A.C., Carneiro, specialized plant-parasitic nematodes. Front Plant Sci 5, R.M., Oliveira, J.T., Grossi-de-Sa, M.F., et al. (2015). 160. Cowpea-Meloidogyne incognita interaction: Root Schaff, J.E., Nielsen, D.M., Smith, C.P., Scholl, E.H., and proteomic analysis during early stages of nematode Bird, D.M. (2007). Comprehensive transcriptome profiling infection. Proteomics 15, 1746-1759. in tomato reveals a role for glycosyltransferase in Mi- Wang, Z., Potter, R.H. and Jones, M.G.K. (2001). A novel mediated nematode resistance. Plant Physiol 144, approach to extract and analyse cytoplasmic contents 1079-1092. from individual giant cells in tomato roots induced by Siddique, S., and Grundler, F.M.W. (2015). Metabolism in Meloidogyne javanica. Int. J. Nematol. 11, 219-225. Nematode Feeding Sites. In Advances in Botanical Wang, Z., Potter, R.H., and Jones, M.G.K. (2003). Research Volume 73, C. Escobar and C. Fenoll, eds. Differential display analysis of gene expression in the (Oxford, UK: Elsevier Academic Press), pp. 119-138. cytoplasm of giant cells induced in tomato roots by Siddique, S., Matera, C., Radakovic, Z.S., Shamim Hasan, Meloidogyne javanica. Molecular Plant Pathology 4, M., Gutbrod, P., Rozanska, E., Sobczak, M., Torres, 361-371. M.A., and Grundler, F.M.W. (2014). Parasitic Worms Wieczorek, K. (2015). Cell Wall Alterations in Nematode- Stimulate Host NADPH Oxidases to Produce Reactive Infected Roots. In Advances in Botanical Research Oxygen Species That Limit Plant Cell Death and Volume 73, C. Escobar and C. Fenoll, eds. (Oxford, UK: Promote Infection. 7, ra33 Elsevier Academic Press), pp. 61-90. Starr, J.L., Yang, W., Yan, Y., Crutcher, F., and Kolomiets, Williamson, V.M. (1999). Plant nematode resistance genes. M. (2014). Expression of Phenylalanine Ammonia Lyase Curr Opinion Plant Biol 2,327-331. Genes in Maize Lines Differing in Susceptibility to Wilson, M.A., Bird, D.M., and van der Knaap, E. (1994). A Meloidogyne incognita. J Nematol 46, 360-364. Comprehensive Subtractive cDNA Cloning Approach to

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 66 7. Plant Susceptibility Factors and Root-knot Nematodes Cabrera et al.

Identify Nematode-Induced Transcripts in Tomato. Other Publications of Interest Phytopathology 84, 299-303. Xu, J.-P. (2014). Next-generation Sequencing: Current Wubben, M., Callahan, F., Hayes, R., and Jenkins, J. Technologies and Applications. Caister Academic Press, (2008). Molecular characterization and temporal U.K. ISBN: 978-1-908230-33-1 expression analyses indicate that the MIC (Meloidogyne He, Z. (2014). Microarrays: Current Technology, Induced Cotton) gene family represents a novel group of Innovations and Applications. Caister Academic Press, root-specific defense-related genes in upland cotton U.K. ISBN: 978-1-908230-49-2 (Gossypium hirsutum L.). Planta 228, 111-123. Poptsova, M.S. (2014). Genome Analysis: Current Wubben, M., Callahan, F., Velten, J., Burke, J., and Procedures and Applications. Caister Academic Press, Jenkins, J. (2015a). Overexpression of MIC-3 indicates a U.K. ISBN: 978-1-908230-29-4 direct role for the MIC gene family in mediating Upland Fuentes, M. and LaBaer, J. (2014). Proteomics: Targeted cotton (Gossypium hirsutum) resistance to root-knot Technology, Innovations and Applications. Caister nematode (Meloidogyne incognita). Theoretical and Academic Press, U.K. ISBN: 978-1-908230-46-1 Applied Genetics 128, 199-209. Murillo, J., Vinatzer, B.A., Jackson, R.W. and Arnold, D.L. Zhang, X.D., Callahan, F.E., Jenkins, J.N., Ma, D.P., (2015). Bacteria-Plant Interactions: Advanced Research Karaca, M., Saha, S., and Creech, R.G. (2002). A novel and Future Trends. Caister Academic Press, U.K. ISBN: root-specific gene, MIC-3, with increased expression in 978-1-908230-58-4 nematode-resistant cotton (Gossypium hirsutum L.) after Caranta, C., Aranda, M.A., Tepfer, M. and Lopez-Moya, root-knot nematode infection. Biochim Biophys Acta J.J. (2011). Recent Advances in Plant Virology. Caister 1576, 214-218. Academic Press, U.K. ISBN: 978-1-904455-75-2 Zhao, W., Li, Z., Fan, J., Hu, C., Yang, R., Qi, X., Chen, H., Nannipieri, P., Pietramellara, G. and Renella, G. (2014). Zhao, F., and Wang, S. (2015). Identification of jasmonic Omics in Soil Science. Caister Academic Press, U.K. acid-associated microRNAs and characterization of the ISBN: 978-1-908230-32-4 regulatory roles of the miR319/TCP4 module under root- Herold, K.E. and Rasooly, A. (2009). Lab-on-a-Chip knot nematode stress in tomato. J Exp Bot. Technology: Fabrication and Microfluidics. Caister Zhuang, Y., Gala, A., and Yen, Y. (2012). Identification of Academic Press, U.K. ISBN: 978-1-904455-46-2 Functional Genic Components of Major Fusarium Head Blight Resistance Quantitative Trait Loci in Wheat Cultivar Sumai 3. Molecular Plant-Microbe Interactions 26, 442-450.

Curr. Issues Mol. Biol. Vol. 19. Omics in Plant Disease Resistance. Vijai Bhadauria (Editor). 67 AIMS AND OBJECTIVES 1. To study the specific gene reprogramming occurring in the plant parasitic nematode feeding sites, including putative mechanisms of gene repression: Similarities and differences are established between two model plants Arabidopsis and tomato.

2. To study the relevance of molecular pathways mediated by phytohormones defined as key components of other root developmental programs during the feeding site formation. 3. To uncover functional molecular pathways in the correct establishment and development of the giant cells and galls during the nematode infection.

CHAPTER 2: Holistic analyses on plant- nematode interactions Chapter 1: Introduction

Chapter 2: Holistic Studies A general overview of the plant nematode interaction Chapter 3: LBD16 & Lateral Roots and the research Chapter 4: 3D Reconstruction performed in this field

Developmental pathways Holistic studies performed mediated by hormones in the susceptible and and altered during the resistant interactions with plant- nematode root- knot nematodes interactions

A method for the 3D phenotping of the giant cells. Insights into their volumes and morphological features

Transcriptomic and The transcriptomes of molecular mechanisms are early developing giant cells common between lateral from tomato and root and giant cells and arabidopsis showed a gall formation. LBD16 is a common massive and key gene in both functional gene repression processes

A tool to facilitate the selection of genes frm the data generated on trasncriptomes from plant- nematode interactions and other fields

miRNA and tasiRNA Regulation of LBD16 co- pathways for gene regulated genes differs silencing are functional in between giant cells and galls and the rasiRNAs are syncytia. Differences in highly induced. Molecular the balance parallelisms with lateral Auxins/Cytokinins between roots both feeding cells Aim of the chapter.

The last section of the introduction highlighted the importance of omic approaches for the study of the development of the galls and giant cells and how they have contributed to a better understanding of these processes. The differentiation of a vascular cells into highly specialized feeding cells, the giant cells, is accompanied by massive transcriptional changes in these cells. Barcala et al. 2010 analysed the transcriptomes of galls induced by Meloidogyne javanica at 3 days post infection and also of their laser captured microdissected giant cells in Arabidopsis. Later, the transcriptomes of galls induced by M. javanica at 1, 3, 7 and 14 days post infection and their laser microdissed giant cells at 3 and 7 days post infection in tomato were performed by Dra. Mary Portillo during her PhD. Here I present the article “Distinct and conserved transcriptomic changes during nematode-induced giant cell development in tomato compared with Arabidopsis: a functional role for gene repression” in which these transcriptomes from tomato galls and giant cells are presented and the comparative analysis with the transcriptomes obtained from Arabidopsis is also performed. My contribution to this article is focused in this comparative analysis. Our main conclusion is that the gene repression is a hallmark of early developing giant cells, differently to that occurring in galls, and that this gene repression is conserved in giant cells in both species, Arabidopsis and tomato. Moreover, we showed that this gene repression seems to have a role for the proper gall development, as the overexpression of a commonly down- regulated peroxidase resulted in a reduction in the number of infections.

With these results in mind, in a step forward we investigated if this massive gene down- regulation could be explained by the small RNA population present in the galls, as this is a common way to obtain such a massive repression in the cells. Therefore, we performed the sequencing of the small RNAs present in 3 days post infection galls from Arabidopsis in the article “Differentially expressed small RNAs in Arabidopsis galls formed by Meloidogyne javanica: a functional role for miR390 and its TAS3-derived tasiRNAs”. Here we show that the main difference between galls and control roots is that there is a higher proportion of 24 nucleotides small RNA instead of 21 nt in galls while in control roots there is a higher proportion of those small RNA sequences of 21 nt. Therefore, a high proportion of miRNAs (mostly 21 nt in Arabidopsis) were found as down- regulated in galls while a massive up- regulation of 24 nt sequences was found. These last sequences participate in the RNA-directed DNA methylation (RdDM) that could explain the gene repression found in the giant cells. Moreover, we demonstrated that the miRNA390a, one of the most induced in the galls and the tasiRNA3a are active in the galls and have a role during the gall formation. Therefore the activation of these two pathways of sRNAs mediated silencing are described for the first time in galls induced by M. javanica in Arabidopsis.

The vast amount of data generated in the holistic analyses of the plant- nematode interaction was scattered in different databases and dozens of publications reviewed in the Introduction. With the aim of facilitating the managing of this knowledge we designed a tool called NEMATIC, published in “NEMATIC: a simple and versatile tool for the in silico analysis of plant–nematode interactions”, that compiles most of the plant- nematode transcriptomes from Arabidopsis. Moreover, a collection of different transcriptomes from other plant research fields with interest for the analysis of the plant- nematode interaction were integrated. NEMATIC has proved to be a useful tool for the discovery of new candidate genes for the analysis of their role in the development of the giant cells and galls, as described in the next chapter. Research

Distinct and conserved transcriptomic changes during nematode-induced giant cell development in tomato compared with Arabidopsis: a functional role for gene repression

Mary Portillo1,*, Javier Cabrera1,*, Keith Lindsey2, Jen Topping2, Maria Fe Andres3, Mariana Emiliozzi3, Juan C. Oliveros4, Gloria Garcıa-Casado4, Roberto Solano4, Hinanit Koltai5, Nathalie Resnick5, Carmen Fenoll1 and Carolina Escobar1 1Facultad de Ciencias Ambientales y Bioquımica, Universidad de Castilla-La Mancha, Avenida de Carlos III s/n, 45071, Toledo, Spain; 2Integrative Cell Biology Laboratory, School of Biological and Biomedical Sciences, Durham University, Durham, DH1 3LE, UK; 3ICA CSIC, Proteccion Vegetal, Serrano 115 dpdo, 28006, Madrid, Spain; 4Centro Nacional de Biotecnologıa CSIC, Darwin3, Campus Universidad Autonoma de Madrid, 28049, Spain; 5Institute of Plant Sciences ARO, Volcani Center, 50250, Bet-Dagan, Israel

Summary Author for correspondence:  Root-knot nematodes (RKNs) induce giant cells (GCs) from root vascular cells inside the Carolina Escobar galls. Accompanying molecular changes as a function of infection time and across different Tel: +34 925 26 88 00 ext 5476 species, and their functional impact, are still poorly understood. Thus, the transcriptomes of Email: [email protected] tomato galls and laser capture microdissected (LCM) GCs over the course of parasitism were Received: 16 June 2012 compared with those of Arabidopsis, and functional analysis of a repressed gene was Accepted: 15 November 2012 performed.  Microarray hybridization with RNA from galls and LCM GCs, infection–reproduction tests New Phytologist (2013) 197: 1276–1290 and quantitative reverse transcription-polymerase chain reaction (qRT-PCR) transcriptional doi: 10.1111/nph.12121 profiles in susceptible and resistant (Mi-1) lines were performed in tomato.  Tomato GC-induced genes include some possibly contributing to the epigenetic control of Key words: Arabidopsis, gene repression, GC identity. GC-repressed genes are conserved between tomato and Arabidopsis, notably giant cell, laser capture microdissected/ those involved in lignin deposition. However, genes related to the regulation of gene expres- microdissection (LCM), Meloidogyne, perox- sion diverge, suggesting that diverse transcriptional regulators mediate common responses idase, tomato, transcriptome. leading to GC formation in different plant species. TPX1, a cell wall peroxidase specifically involved in lignification, was strongly repressed in GCs/galls, but induced in a nearly isogenic Mi-1 resistant line on nematode infection. TPX1 overexpression in susceptible plants hindered nematode reproduction and GC expansion.  Time-course and cross-species comparisons of gall and GC transcriptomes provide novel insights pointing to the relevance of gene repression during RKN establishment.

from RKN juveniles that interacts in vitro with a SCARE- Introduction CROW-like plant transcription factor (TF) (Huang et al., 2006). Root-knot nematodes (RKNs) establish an intimate interaction Drastic changes in plant gene expression occur in GCs and sur- with their hosts, transforming four to seven differentiated root rounding cells (reviewed in Gheysen & Fenoll, 2002; Escobar vascular cells into giant cells (GCs) that undergo mitoses without et al., 2011; Hewezi et al., 2012). Transcriptional profile changes complete cytokinesis, followed by endoreduplication (Caillaud in Arabidopsis (Hammes et al., 2005; Jammes et al., 2005) and et al., 2008a; De Almeida & Favery, 2011). Although the molec- tomato galls (Bar-Or et al., 2005; Schaff et al., 2007) reveal ular basis of this transformation is not yet well understood, differences as a function of the infection time. Transcriptomic nematode secretions injected into the plant cells may trigger sig- analyses during feeding cell development have only been nals for GC differentiation and maintenance, some mimicking performed in Arabidopsis and soybean for microaspirated and plant peptides that control proliferation/differentiation balances laser capture microdissected (LCM)-isolated syncytia in compati- of stem cells from the vascular bundles, for example, the CLE- ble interactions at early–medium infection points, revealing like peptides (reviewed in Rosso & Grenier, 2011) or 16D10 differences as a function of the infection stage in soybean, but not in Arabidopsis (Ithal et al., 2007; Klink et al., 2007; Szakasits et al., 2009). However, similar analyses are still lacking for GCs. *These authors contributed equally to this work. Likewise, comparisons of transcriptional profiles from excised

1276 New Phytologist (2013) 197: 1276–1290 Ó 2013 The Authors www.newphytologist.com New Phytologist Ó 2013 New Phytologist Trust New Phytologist Research 1277 infected roots with their correspondent isolated syncytia are lim- (20–30 second-stage juveniles (J2s); Portillo et al., 2009; Fig. 1). ited to Arabidopsis and soybean (Klink et al., 2005, 2007; Ithal Equivalent sampling and biological materials were used for all et al., 2007). Only one study has compared expression patterns of experiments, including gall hand-dissection and cryosectioning galls with LCM GCs at 3 d post-inoculation in Arabidopsis, before GC LCM (Portillo et al., 2009). identifying many differentially expressed genes (DEGs) in GCs, but not in galls; thus, both structures show distinct gene expres- sion profiles (Barcala et al., 2010). (a) (b) GC content isolation has been performed in mature tomato GCs (Bird & Wilson, 1994; Wang et al., 2001, 2003) and also, at very early infection points (48–72h), using LCM followed by microarray hybridization (Portillo et al., 2009) and followed by a cDNA library construction (Fosu Nyarko et al., 2009). However, cross-species comparisons of transcriptomic data from galls/GCs or syncytia have not been reported, in spite of their potential to identify unique or conserved plant responses during plant– (c) (d) nematode interactions (Escobar et al., 2011). Extensive down-regulation of gene expression has been described in Arabidopsis (GCs and galls) and in tomato galls after nematode establishment, particularly of stress-related genes (Schaff et al., 2007; Caillaud et al., 2008b; Barcala et al., 2010), suggesting the participation of putative nematode suppressors of plant defense (reviewed in Smant & Jones, 2011). For example, (e) (f) peroxidase-coding genes are repressed in the compatible interac- tion, but up-regulated in tomato Mi (Bar-Or et al., 2005; Schaff et al., 2007) and in soybean resistant plants (Klink et al., 2009, 2010). The repression of particular gene subsets, such as those involved in defense and secondary metabolism, occurs in patho- genic and symbiotic interactions (Maunoury et al., 2010; Schlink, 2010 Moreau et al., 2011; Damiani et al., 2012), but also during cell differentiation (Sawa et al., 2005; Ito & Sun, 2009). In this respect, GCs have been suggested to derive from developing tracheary elements (Bird, 1996; Barcala et al., 2010) through an as yet undefined nematode-triggered cell differentiation process. (g) To gain insight into GC and gall differentiation, we compared the differential transcriptomes of tomato LCM GCs and whole galls formed by Meloidogyne javanica over the course of parasit- ism, as well as the transcriptomes of tomato and Arabidopsis LCM GCs in the same conditions. Distinct pathways, such as Fig. 1 Collected material and monitoring of the infection with those leading to lignin deposition, were highly conserved between Meloidogyne javanica second-stage juveniles (J2s) for expression analysis. Arabidopsis and tomato during GC development. By contrast, At the top, the images on the right correspond to hand-dissected tomato genes involved in the regulation of gene expression were distinct. galls and, on the left, their corresponding transverse Araldite® resin The overexpression of a peroxidase involved in tomato lignifica- sections. (a–d) Tomato (Solanum lycopersicum) hand-dissected galls at 1, tion, TPX1, repressed in the compatible and induced in the 3, 7 and 14 d post-infection (dpi) were collected. (a) The typical initial gall swelling was evident at 1 dpi (24 h post-infection; hpi). Identical incompatible interaction (Mi-1), impaired nematode feeding site experimental conditions were used to obtain giant cells (GCs) during their development, suggesting the importance of gene repression initiation and differentiation process: (b) at 3 dpi, the first developmental during nematode establishment. stages of GCs are clearly identified; (c) at 7 dpi, GCs are expanding and show an increase in cytoplasmic density with multiple nuclei. Section bars, 100 lm; gall bars, 1 cm. (e) As nematode penetration is not synchronous, Materials and Methods the gall formation and position relative to the root tips were strictly monitored every 24 h. (f) Equivalent primary root tissues (in age and Plant material, nematode infections and gall sections position) of uninfected seedlings were used as controls for the gall transcriptomic analysis and laser capture microdissection (LCM) of control Seeds from tomato (Solanum lycopersicum Mill, cv Moneymaker) tissue; the positions of the root tips were labeled every 24 h. (g) Infected were germinated in vitro in Petri dishes (125 mm) on Gamborg root segments with necrotic areas (black arrows) from resistant À1 S. lycopersicum cv Motelle (Mi-1/Mi-1) were collected for quantitative B5 medium (Gamborg et al., 1968) supplemented with 20 g l reverse transcription-polymerase chain reaction (qRT-PCR) 2 d post- sucrose and 1.5% agar. Each primary tomato root was inoculated inoculation. Equivalent root segments were also collected as controls (not just behind the root tip with sterile freshly hatched M. javanica shown). Asterisks in sections indicate GCs.

Ó 2013 The Authors New Phytologist (2013) 197: 1276–1290 New Phytologist Ó 2013 New Phytologist Trust www.newphytologist.com New 1278 Research Phytologist

Seedlings of two isogenic tomato lines, susceptible Money- galls, four independent time points were examined (1, 3, 7 and maker (mi/mi) and resistant Motelle (Mi-1/Mi-1) genotypes, 14 dpi vs their respective noninfected root segments at an equiva- were grown in small clay pots with 10 ml of quartz sand at lent developmental stage to that of the infected plants; Fig. 1; 25 Æ 2°C, 18 h light. Twenty-one-day-old plants were inoculated samples and labeling in Supporting Information Table S12). with 100 freshly hatched M. javanica J2s per plant; noninoculat- Data were collected in the Cy3 and Cy5 channels. Background ed plants were kept as controls (see legend to Figs 1, 7). correction and normalization of the expression data were per- Transgenic Solanum lycopersicum cv Pera plants overexpressing formed using LIMMA, as described by Adie et al. (2007). First, the cell wall peroxidase TPX1 driven by the 35S promoter were the dataset was filtered based on spot quality. A strategy of adap- homozygous for a single T-DNA insertion (TP3 line; El Mansouri tive background correction was used that avoids exaggerated vari- et al., 1999; Figs 7, 8). Statistical analysis of infection and repro- ability of log ratios for low-intensity spots; for local background duction parameters after ANOVA was set to a significance level of correction, the ‘normexp’ method in LIMMA was used to adjust P  0.05 using the The SPSS statistics package, IBM (Armonk, the local median background. The resulting log ratios were print- NY, USA). Control wild-type plants were nontransformed plants tip loess normalized for each array to have a similar distribution from the same cultivar. across arrays and to achieve consistency among arrays, and log Galls at 3 and 7 dpi from 35S::TPX1 plants and controls were ratio values were scaled using the median absolute value as scale hand-dissected, fixed and embedded as in Barcala et al. (2010). estimator (Smyth & Speed, 2003). Linear model methods were Sections were stained with 1% toluidine blue in 1% borax solu- used to determine DEGs. Each probe was tested for changes tion (TAAB). The TrakEM2 plug-in from FIJI was used to mea- in expression over replicates by an empirical Bayes moderated sure the GC area (Cardona et al., 2012). Two representative galls t-statistic (Smyth, 2004). To overcome the problem of multiple from each genotype were entirely sectioned into 2-lm slices and testing, corrected P values (q values) were calculated following the areas were quantified (Fig. 8). Benjamini & Hochberg (1995). Gene expression differences (always between galls or GCs vs their corresponding controls) were considered to be significant with q < 0.05.When several immobi- Gall and GC RNA isolation, amplification and aRNA probe lized spots (IDs) corresponded to the same unigene (SGN-U), the preparation representative probe with the highest differential expression level, Independent RNA extractions from 100 mg of hand-dissected among those with q < 0.05, was used for further analysis. Data galls or equivalent, uninfected controls at 1, 3, 7 and 14 dpi from TOM1 microarray spots were downloaded from the Tomato (Fig. 1) were performed, and the quality was assessed as in Portillo Functional Genomics Database: http://ted.bti.cornell.edu/. Hier- et al. (2006). More than 12 (uninfected plants) or 18 (infected archical clustering (HCL) was calculated using MultiExperiment plants) plates, each containing eight individuals, were used for Viewer (MeV4.1; http://www.tm4.org/mev; Saeed et al., 2006) each independent experiment and time point. A total of six inde- with Pearson uncentered metric distance and a complete linkage. pendent experiments was performed (i.e. 3000 hand-dissected For clustering analyses, we selected those genes with q < 0.05 at any galls and 4400 uninfected root fragments). RNA from two inde- of the developmental stages assayed (1, 3, 7, 14 dpi). Functional pendent experiments was pooled in equimolar ratios to obtain one categories of DEG were obtained from the Solanaceous Mapman ‘independent biological replicate’ per infection time analyzed. In ontology for TOM1 chip. A gene was considered to be ‘distinctive’ total, three ‘independent biological replicates’ were processed. A of a particular developmental stage, or of galls or GCs, when it was single round of linear amplification using a MessageAmpe II only differentially expressed in one of the situations, but not in the aRNA Amplification Kit (Casson et al., 2005) was performed. other, considering a threshold of q < 0.05. RNA isolation and aRNA probe preparation from LCM GCs at 3 and 7 dpi and from control vascular cells from cryosections Real-time quantitative reverse transcription-polymerase were performed as described by Portillo et al. (2009). Gene anno- chain reaction (qRT-PCR) and in situ RT-PCR tation and putative functions for expressed sequence tags (ESTs) are described on the Solanaceae Genomics Network (SGN), qRT-PCR was carried out as in Portillo et al. (2009) (primers in http://www.sgn.cornell.edu/index.pl. Table S12). Statistical analysis after ANOVA was set to a signifi- cance level of P  0.05 using the the SPSS statistics package, IBM (Armonk, NY, USA) package. Microarray analysis In situ RT-PCR was performed in single 3- and 7-dpi galls Microarray hybridization, normalization and statistical analysis (primers in Table S12). Fixation and PCR-SYBR reaction were of the expression data were performed as described by Portillo as described in Gal et al. (2006). Two adjacent sections for each et al. (2009) and Barcala et al. (2010). Three or four slides were gall were used, one as a control PCR without primers, and hybridized independently with aRNA from three to four inde- repeated twice (Fig. 6). Tissues were immediately observed using pendent biological replicates, including one dye-swap for both a confocal microscope (Olympus IX81, Tokyo, Japan) to record analyses: galls and LCM GCs. Two independent time points the fluorescence signal (excitation and emission wavelengths of were assayed for GCs (3- and 7-dpi GCs vs their respective con- 578 and 603 nm, respectively). The BA505I filter and the argon trol cells from vascular cylinders of noninfected tissue at an 488-nm laser beam were modified for no fluorescence back- equivalent developmental stage to that of the infected plants); for ground signal in the controls.

New Phytologist (2013) 197: 1276–1290 Ó 2013 The Authors www.newphytologist.com New Phytologist Ó 2013 New Phytologist Trust New Phytologist Research 1279

(a) Identification and analysis of tomato–Arabidopsis homologs TBLASTN analysis to identify the tomato–Arabidopsis homologs was performed using, as query, AGI protein sequences of DEG from Arabidopsis 3-dpi GCs induced by M. javanica J2s, down- loaded with ‘Sequence Bulk Download and Analysis’ from TAIR10 (http://www.arabidopsis.org/tools/bulk/sequences/ index.jsp). The output sequences were confronted to the database TOM1 re-sequences from the ‘Tomato Functional Genomics Database’ (http://ted.bti.cornell.edu/cgi-bin/TFGD/array/blast. cgi; Fei et al., 2011). From zero to five TOM1 probe sequences producing statistically significant alignments (E-value < 0.01) were retrieved for each Arabidopsis protein. To confirm TBLASTN matches, a reciprocal BLASTX was performed, confronting trans- lated tomato probes with the Arabidopsis TAIR10 protein data- (b) base. When the same Arabidopsis protein as in the TBLASTN analysis was matched in the BLASTX with a statistically significant value (E-value < 0.01), the genes were considered to be homologs (Table S3). It is important to note that the great majority of the considered homologs (94.4%) showed E-values < 1E-05. Only 2.7% of the considered homologs were chosen with E-values between 1E-02 and 1E-03.The log ratios relative to their own 2 Fig. 2 Global changes in differential gene expression during tomato controls of tomato and Arabidopsis DEG were used to obtain co- (Solanum lycopersicum) giant cell (GC) differentiation. (a) Number of regulated and contrastingly regulated genes. The over-representa- differentially expressed genes (DEGs; q < 0.05) in laser capture tion of Mapman categories within the shared genes between both microdissected (LCM) GCs at 3 and 7 d post-infection (dpi) with respect to species was contrasted with the chi-squared test set at P < 0.05. LCM cells from the vascular cylinder of noninfected roots. (b) Venn diagrams showing temporal changes in DEGs. The intersections of the diagrams represent genes co-regulated in both infection stages. Genes Results that did not display a uniform expression pattern as a function of the infection time are not included in the Venn analysis, but are listed in Supporting Information Table S13. Data have been deposited at GEO, Unique gene expression profiles during tomato GC www.ncbi.nlm.nih.gov/geo (accession no. GSE30048). development Transcriptomic analysis of LCM GCs at 3 and 7 dpi vs control clearly repressed. In the ‘Stress’, ‘DNA’, ‘Cell wall’ and ‘Transport’ vascular cells from uninfected roots revealed that 12% of the categories, no induced genes were found. Similarly, at 3 and 7 dpi, 8500 genes represented in the TOM1 chip were DEGs (false dis- most genes from the ‘RNA’ category (WRKY, bHLH and NAC covery rate cut-off, q < 0.05). Most genes with q < 0.05 also TF families, among others) were repressed, as were ‘Stress’ and showed a fold change (FC) > 1.9. For simplicity, the terms ‘Miscellaneous’ (over-represented with P < 0.05; Fig. 3; Table S2). induced/up-regulated and repressed/down-regulated are used At 7 dpi, the number of DEGs increased considerably and the throughout the text to mean transcript levels higher or lower than ‘Protein’ (‘Protein synthesis’) and ‘DNA’ (‘DNA synthesis/Chroma- the corresponding controls, respectively. tin structure’ and ‘Histones’) categories showed more induced than At 3 dpi, 307 DEGs were identified, most of them down- repressed genes (over-represented, P < 0.05; Fig. 3). Interestingly, regulated (86%). By contrast, at 7 dpi, a larger number of DEGs genes encoding 40, 50 and 60S ribosomal proteins were up-regu- (1000) was obtained and the percentages of up- and down- lated exclusively at 7 dpi. Most induced distinctive DEGs at 7 dpi regulated genes were similar (43.5% and 56.5%, respectively; (11 of 12) encoded histones, such as H3 and H4, together with genes Fig. 2a). At 3 dpi, only 51 genes were DEGs (i.e. were distinctive encoding a histone deacetylase, a WD-40 repeat protein (MSI)puta- for this stage; see Materials and Methods). Of these, 42 (82.3%) tively involved in nucleosome assembly, and a DEAD/DEAH box were down-regulated. By contrast, 744 DEGs were distinctive at helicase (Dicer-like) (Table S2). Thus, the expression of genes 7 dpi, 53.6% were up-regulated and 46.4% were repressed related to chromatin remodeling, maintenance and protein synthesis (Fig. 2b; Table S1). is associated with particular GC developmental stages. Biologically functional information for the two GC develop- mental stages was obtained from Solanaceous Mapman ontology Conserved and nonconserved gene expression patterns in (Urbanczyk-Wochniak et al., 2006; Fig. 3). In both GC stages, a Arabidopsis and tomato GCs large proportion of DEGs (323) were classified in the ‘Not assigned function’ category (including genes with ‘No ontology’ and The observed trends in global expression, such as large numbers of ‘Unknown function’). At 3 dpi, most DEGs in all categories were repressed genes in early developing GCs, were similar in tomato

Ó 2013 The Authors New Phytologist (2013) 197: 1276–1290 New Phytologist Ó 2013 New Phytologist Trust www.newphytologist.com New 1280 Research Phytologist

coordinately down-regulated (76.5%), but only one of 21 was up- regulated (0.7%; Table 1a). Similar results were observed at 7 dpi for repressed genes, whereas half of the up-regulated genes showed opposite regulation (Table 1a). In addition, a reciprocal BLASTX to confront the 482 significant counterparts found between Ara- bidopsis and tomato validated 93.4% (450) as the percentage of translated nucleotide sequences from tomato matching the same Arabidopsis protein sequence (E-value < 0.01; Table 1b; Table S3). An interactive spread sheet is provided to simplify searches for corresponding homologous DEGs in GCs from tomato and Arabidopsis (Table S4). Although some caution in interpretation must be taken because of the lower unigene representation in the TOM1 microarray rela- tive to that of Arabidopsis, genes down-regulated in early develop- Fig. 3 Overview of differentially expressed genes in tomato (Solanum ing GCs are robustly conserved between the two species. The lycopersicum) giant cells classified into functional categories (Mapman). phenylpropanoid pathway within the ‘Secondary metabolism’ cat- Bars indicate the number of genes per category. Asterisks mark categories egory was significantly over-represented among repressed genes in over-represented with a statistical significance (P < 0.05) using the both tomato and Arabidopsis, particularly genes involved in lignin – Wilcoxon Rank Sum test with Benjamini Hochberg correction (Benjamini biosynthesis, a group of peroxidases from the category ‘Miscella- & Hochberg, 1995), indicative of significant differential expression profiles in comparison with the rest of the categories. Infection time points in nea’ and genes in a small biotic stress subcategory encoding prote- colors as indicated. ase inhibitors. These results were confirmed by an independent analysis based on different categorizations, both being highly coin- and Arabidopsis (Barcala et al., 2010; this work). Thus, we exam- cident for the phenylpropanoid pathway (Table 2, Mapman; ined their putatively conserved expression patterns. We retrieved Fig. 4a, GeneOntology (GO)). When the same group of genes protein sequences for the 1161 Arabidopsis DEGs in LCM GCs was placed in KEGG, all were included in the lignin biosynthesis (Barcala et al., 2010) from TAIR and confronted them against the and cross-linking routes (Fig. 4b). Thus, the down-regulation of six putative open reading frames (ORFs) of the tomato database genes encoding proteins in the lignin monomer synthesis and lig- TOM1 re-sequences from the ‘Tomato Functional Genomics nin cross-linking pathways was consistently conserved between Database’ (see the Materials and Methods section) by TBLASTN the two species. Other over-represented categories in the GO anal- (Table S3). Eight-hundred and sixty-six Arabidopsis genes ysis for commonly regulated genes were found, notably genes matched homologous sequences in tomato. Thus, 74.6% of the related to auxin signaling, stress responses and transport, which Arabidopsis GC DEGs had a putative homolog protein counter- were also mostly down-regulated (Fig. 4a). part in tomato (77% from the up-regulated and 75% from the However, a large number of genes (Table 1a) were distinctly repressed Arabidopsis DEGs). Of these, 43% were DEGs in both regulated and were not shared by tomato and Arabidopsis GCs, tomato (at 3 and 7 dpi) and Arabidopsis (at 3 dpi) GCs. Interest- particularly those related to the regulation of gene expression and ingly, from the 132 DEGs at 3 dpi in both species, 101 were to housekeeping metabolic processes (Fig. 4a). In agreement with

Table 1 Giant cell differentially expressed gene (DEG) homologs

(a) TBLASTN analysis from the Arabidopsis DEGs in laser capture microdissected (LCM) giant cells (GCs) confronted to the tomato database Homologs in tomato*

3 dpi 7 dpi 3 + 7 dpi

Up-regulated genes in Arabidopsis Same regulation in tomato 1 (0.7%) 49 (13.0%) 50 (12.7%) Different regulation in tomato 20 (15.1%) 55 (14.6%) 57 (14.5%) Down-regulated genes in Arabidopsis Same regulation in tomato 101 (76.5%) 206 (54.8%) 218 (55.6%) Different regulation in tomato 10 (7.6%) 66 (17.5%) 67 (17.1%)

(b) BLASTX reciprocal analysis from tomato to the Arabidopsis database † E value range

À À À À <1 9 e 5 1 9 e 2/1 9 e 5 > 1 9 e 2 No hit Total significant counterparts

TBLASTN 454 (94.2%) 28 (5.8%) ––482 BLASTX 436 (90.4%) 14 (2.9%) 10 (2.1%) 22 (4.6%) 450 (93.4% from those found in TBLASTN)

*Percentages and absolute data from the up- and down-regulated homologs are shown. †An E-value < 0.01 was considered for significant homology. dpi, days post-inoculation.

New Phytologist (2013) 197: 1276–1290 Ó 2013 The Authors www.newphytologist.com New Phytologist Ó 2013 New Phytologist Trust New Phytologist Research 1281

Table 2 Mapman functional categories of common repressed genes from did not show any gene commonly regulated between tomato and the tomato and Arabidopsis giant cell transcriptomes; over-represented Arabidopsis (yellow label in Table S5). In addition, all TF fami- < v2 categories and subcategories inferred from P 0.05 of the are shown in lies have either tomato or Arabidopsis homologs noncommonly bold regulated. However, all DEGs from lateral organ boundaries Mapman Bin Category v2 P value (LBDs), a transcription regulator acting as repressor of auxin- inducible gene expression (AUX/IAA) and CONSTANS like zinc 6 Gluconeogenesis 1.076 0.300 finger TFs (C2C2 (Zn) CO-like) were co-regulated (Table S5). /Glyoxylate cycle 8 TCA/Org. transformation 0.160 0.689 9 Mitochondrial electron 0.000 0.983 Comparative expression profiles in tomato GCs and galls at transport/ATP synthesis 10 Cell wall 1.105 0.293 different developmental stages 11 Lipid metabolism 0.002 0.962 A complete transcript profile using the TOM1 microarray in galls 13 Amino acid metabolism 1.067 0.302 from migration/establishment (1 dpi) to gall/GC differentiation 15 Metal handling 0.317 0.574 – 16 Secondary metabolism 3.136 0.077 (3 dpi) and maturity (7 14 dpi; Fig. 1) identified 2414 ESTs 16.1 Secondary metabolism. 1.068 0.301 with significant differences (q < 0.05) to uninfected controls at Isoprenoids one or more infection stages. These represent 1839 (22%) unig- 16.2 Secondary metabolism. 6.303 0.012 enes of the 8500 in the array. An interactive searching sheet is Phenylpropanoids 16.2.1 Secondary metabolism. 7.220 0.007 supplied that extracts gall and GC transcriptional profiles of any Phenylpropanoids. particular tomato ID, or from a list of IDs, at all infection stages Lignin biosynthesis (galls: 1, 3, 7, 14 dpi; GCs: 3, 7 dpi; Table S6). 16.5 Secondary metabolism. 0.057 0.811 Marked expression pattern differences as a function of the Sulfur-containing infection time were observed in galls (Fig. S1). At very early 16.8 Secondary metabolism. 2.816 0.093 Flavonoids stages, most genes were down-regulated, notably at 1 dpi (71%), 16.10 Secondary metabolism. 0.308 0.579 but also at 3 dpi (60%). As infection progressed (7 and 14 dpi), Simple phenols the proportion of up- and down-regulated genes was balanced 17 Hormone metabolism 3.687 0.055 (46% up, 54% down and 46% up, 54% down, respectively) and 20 Stress 0.310 0.578 the DEG number was considerably higher than at early stages. 21 Redox 0.291 0.590 24 Biodegradation of 0.614 0.433 This agrees with HCL, revealing that global transcriptional pro- xenobiotics files between 1 and 3 dpi were more similar than those at 7 and 26 Miscellanea 5.727 0.017 14 dpi, sorted in a separate cluster (Fig. 5a). At 1 dpi, 99 were 26.2 Miscellanea.UDP 2.563 0.109 exclusively DEGs, mostly down-regulated, whereas, at later glucosyl and stages, similar proportions of distinctive up- and down-regulated glucuronyl transferases – 26.7 Miscellanea.Oxidases 0.286 0.593 genes were observed (47 50%; Fig. S1; Tables S6, S7). A small 26.8 Miscellanea.Nitrilases 0.286 0.593 number were DEGs during all infection stages (Fig. S1). 26.9 Miscellanea.Glutathione 1.701 0.192 Significant differential expression profiles (P < 0.05) were S transferases encountered in the ‘Cell wall’ category, as the induced gene num- 26.12 Miscellanea. 6.218 0.013 ber increased over the infection course. By contrast, the percent- Peroxidases 26.13 Miscellanea.Acid and 2.331 0.127 age of repressed genes in the ‘Stress’ category (23%) was other phosphatases markedly high at 1 dpi, but lower at later time points (Fig. S2). 26.16 Miscellanea. 0.308 0.579 Hence, of the 99 distinctive genes at 1 dpi (Table S7), 66 were in Myrosinases-lectin- the ‘Stress’ category. DEGs with expression level variations at dif- jacalin ferent infection points are listed in Table S8. 26.21 Miscellanea.Protease 4.251 0.039 inhibitor/seed storage To study differences between gene expression changes in GCs /lipid transfer protein and galls, we compared their differential transcriptomes in sam- (LTP) family protein ples obtained with equivalent biological sampling, using the same 27 RNA 0.001 0.976 microarray platform, at the same developmental stages (3 and 29 Protein 0.058 0.810 7 dpi) and with the same data analysis tools. Each time point was 30 Signaling 2.966 0.085 31 Cell 0.155 0.694 compared independently, but cross-comparisons between analy- 33 Development 2.463 0.117 ses were also considered. Less than half of GC DEGs at 3 and 34 Transport 0.953 0.329 7 dpi were also detected as gall DEGs (148 of 307 and 469 of 1000, respectively). This indicates that many DEGs were GC dis- tinctive (Fig. 5; Table S9). Distinctive GC DEGs at 3 dpi were this, some TF groups, such as the b-zip, a TF family with a predominantly down-regulated, whereas, at 7 dpi, the proportion DNA-binding domain including a single C(2)-C(2) zinc finger of up- and down-regulated GC-distinctive genes was balanced (C2C2 (Zn) DOF) and another family of zinc finger TFs (Fig. 5). When all co-expressed GC and gall DEGs were com- containing the GATA domain (GATA), or homeobox families, pared, the log2 values of their expression values were generally

Ó 2013 The Authors New Phytologist (2013) 197: 1276–1290 New Phytologist Ó 2013 New Phytologist Trust www.newphytologist.com New 1282 Research Phytologist

(a)

(b)

Fig. 4 Comparison of tomato (Solanum lycopersicum) and Arabidopsis giant cell (GC) transcriptomes. (a) Categorization of tomato–Arabidopsis homologs following GeneOntology (GO) processes. Only over- represented categories (with P < 0.05 from the v2) of the commonly regulated and exclusive (not co-regulated in both plant species) genes are shown. (b) Adapted pathway from KEGG for lignin biosynthesis and cross-linking enzymes. Genes represented are the conserved homologs between tomato and Arabidopsis in GCs. Green squares, repressed genes.

lower in galls than in GCs (Table S9). A group of distinctive GC At 7 dpi, ‘Protein’, ‘Metabolism’, ‘RNA’ and ‘Miscellaneous’ genes (eight at 3 dpi and 276 at 7 dpi) was obtained when all categories showed the highest number of distinctive GC and DEGs from galls at 1, 3, 7 and 14 dpi were compared with those gall genes. Eleven members of the histone family were solely of 3- and 7-dpi LCM GCs (Table S10). induced in 7-dpi GCs, none being differentially expressed in Functional classification of distinctive gall and GC genes either 7-dpi galls or 3-dpi GCs. Other genes putatively involved expressed at the same developmental stage revealed differences in in chromatin structure maintenance or remodeling, such as the predominant biological processes altered. The largest number SNF2, seem to be distinctive of 3-dpi GCs (Fig. S4; Table S9). of distinctive genes was classified in the ‘Not assigned’ category The subcategory ‘Protein synthesis’, containing genes encoding (Fig. S3; Table S9). At 3 dpi, GC-distinctive genes were mostly ribosomal proteins, included a large number of DEGs, in both repressed (grouped in ‘Stress’, ‘RNA’, ‘Protein’, ‘Secondary metab- galls and GCs, but encoded different proteins (Table S9). One olism’ and ‘Transport’), whereas gall-distinctive genes were mostly of the major categories of distinctive genes in GCs and galls at induced (Fig. S3), essentially for the categories ‘RNA: regulation both developmental stages was ‘RNA’, containing mainly TFs of transcription’ and ‘Protein: protein degradation’. Interestingly, that could be key regulators of cell differentiation. A detailed the ‘Stress’ category contained a high proportion of induced genes analysis of the ‘Regulation of transcription’ subcategory showed in galls in contrast with GCs. ‘Miscellaneous’ was the group with all members of TFs containing the WRKY domain, the largest number of distinctive up-regulated genes in galls and WRKYGQK, (WRKY) and the basic helix-loop-helix (bHLH) down-regulated genes in GCs, encoding proteins such as glucos- families repressed in GCs, but not in galls. Most AP2 yltransferases and peroxidases, among others (Fig. S3). (APETALA2) and EREBPs (ethylene-responsive element

New Phytologist (2013) 197: 1276–1290 Ó 2013 The Authors www.newphytologist.com New Phytologist Ó 2013 New Phytologist Trust New Phytologist Research 1283

(a) and 7 dpi, but induced in galls, including those encoding patho- genesis-related proteins (e.g. in Tables S9, S11). These patterns indicate that GCs and the other gall tissues exhibit independent and differential transcriptional profiles.

Confirmation of microarray data qRT-PCR was performed for 15 genes, selected by criteria such as similar expression profiles over infection time, high FC or pref- erentially expressed in galls or GCs (primer sequences in Table S12). The reference gene for normalization encoded a hypothetical protein (SGN-U150992, 1-1-6.4.11.14) and showed steady expression as a function of the infection time in galls and GCs. qRT-PCR revealed transcriptional patterns similar to those found in microarray hybridization; only one gene coding an omega-6 fatty acid desaturase showed a different expression pattern at 7 dpi in galls with both methods (underlined in Table 3). Some were not DEGs in the microarray at certain infec- tion points (white boxes in Table 3), although a clear tendency to up- or down-regulation was observed, coincident with qRT-PCR (b) data. In situ RT-PCR hybridization for selected genes preferentially expressed in GCs (40S ribosomal protein (40S) and a putative serine/threonine kinase (STK)) revealed that both accumulated in 7-dpi GCs (Table 3; Fig. 6b,d), but not in their corresponding controls (Fig. 6a,c). In agreement with the GC induction of 40S and STK from the microarrays (Table 3), the signal observed by in situ PCR in GCs was absent or much lower in adjacent cells (Table 3; Fig. 6b,d), regardless of the differences in detection sen- sitivity. Transcripts for the same genes in control uninfected roots were barely detectable; an example is provided for STK in Fig. S5. Hence, the high validation rate of the microarray data with qRT- PCR, combined with the in situ RT-PCR, reflects a significant reliability of the transcriptomic data.

Fig. 5 Time course of transcript profiles of tomato (Solanum lycopersicum) Functional characterization of a tomato GC-repressed galls and comparison between differentially expressed genes (DEGs) of peroxidase giant cells (GCs) and galls. (a) Hierarchical cluster analysis of tomato gall DEGs at 1, 3, 7 and 14 d post-infection (dpi). The numbers within the tree To elucidate the biological relevance of early gene down-regula- correspond to the bootstrap values after 1000 iterations. (b) Total DEGs in tion in galls and GCs, particularly of genes involved in lignin GCs and those common to galls. The number of ‘GC-distinctive genes’ increased in the lower fold change (FC) ranges. deposition, during compatible interactions, we used tomato plants over-expressing the basic cell wall peroxidase TPX1. TPX1 binding proteins), AP2/EREBP, and all members of a TF was consistently down-regulated in tomato galls from early stages family containing the no apical meristem, NAM, domain (1 dpi) at the infection points analyzed and strongly down-regu- (NAC) were repressed in galls and GCs at both stages. Some lated in GCs (3 and 7 dpi; FCs of – 2.7 and –4.6, respectively; members of the homeobox domain TFs (HB), C2C2 (Zn) Table 3). In Arabidopsis GCs, 10 peroxidase-coding genes were DOF and GATA zinc finger families, among others, were up- repressed; eight had tomato homologs, all co-repressed in tomato regulated in galls, but not in GCs. Interestingly, repression lev- and Arabidopsis GCs. These included four putative Arabidopsis els were less pronounced in galls than in GCs, suggesting that homologs of TPX1 (ID: 1-1-3.2.20.9) (AT2G35380; GCs contribute strongly to the gene repression found in galls AT2G41480; AT4G11290; AT5G51890; Fig. 4). TPX1 has (Fig. S4). been formerly characterized as being directly involved in lignin Another category showing clear differences between GCs and biosynthesis in tomato (El Mansouri et al., 1999). galls was ‘Metabolism’. All genes associated with lipid, amino We first analyzed TPX1 expression in a homozygous resistant acid and secondary metabolism were repressed in GCs at 3 and line (Motelle, Mi-1/Mi-1) triggered by M. javanica 2 d after inoc- 7 dpi, but most were either induced or not DEGs in galls. Simi- ulation when compared with a nearly isogenic tomato susceptible larly, most biotic stress-related genes were repressed in GCs at 3 line (Moneymaker, mi-1/mi-1; Schaff et al., 2007). The results

Ó 2013 The Authors New Phytologist (2013) 197: 1276–1290 New Phytologist Ó 2013 New Phytologist Trust www.newphytologist.com New 1284 Research Phytologist

Table 3 Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) validation of gall microarray hybridization data

Relative expression levels (log2)

Galls 1 dpi Galls 3 dpi Galls 7 dpi Galls 14 dpi GCs GCs 3 dpi 7 dpi Description qPCR Array qPCR Array qPCR Array qPCR Array Array Array

Putative LRP (lateral root primordia) 1 2.03 0.90 2.40 1.67 2.40 1.96 2.49 2.48 1.71 1.85 Wound-induced protein Sn-1 1.33 0.08 2.82 2.74 3.83 4.08 3.54 4.47 1.95 À0.44 Late-embryogenesis protein lea5 À0.95 À1.83 À0.65 À2.00 À1.78 À1.58 À1.41 À1.71 À3.34 À4.31 Xyloglucan endotransglycosylase (XTR4) À1.63 À1.22 À1.47 À1.69 À2.31 À1.65 À1.98 À3.03 À2.49 À2.98 Pathogenesis-related protein PR-1 precursor À1.52 À2.05 À0.81 À1.09 À2.57 À1.45 À2.00 À1.99 À2.82 À4.63 Peroxidase (TPX1) À0.51 À1.13 À0.64 À0.97 À1.15 À2.29 À2.25 À2.66 À2.68 À4.58 Omega-6 fatty acid desaturase 0.14 0.27 1.59 0.71 À1.13 1.04 À2.14 1.02 1.44 2.29 Histone H3 [Arabidopsis thaliana] À0.28 À0.45 À0.09 À0.23 1.07 À0.11 0.99 1.45 0.20 3.01 HD-Zip transcription factor Athb-14 (HD) À0.28 À0.04 0.74 0.26 2.20 1.10 1.87 0.85 2.37 1.66 WRKY family transcription factor 1.35 0.85 1.81 1.34 1.54 1.51 1.75 1.66 1.05 À0.55 Homeotic protein VAHOX1 - tomato À0.89 À0.75 À0.26 À0.69 À0.60 À0.44 0.18 À0.83 À1.04 À1.37 Receptor protein kinase-related protein [Arabidopsis 0.68 0.71 1.85 1.07 0.14 1.15 1.27 1.07 0.96 À0.26 thaliana] Putative serine/threonine kinase similar to NAK (STK) * 0.40 0.09 0.42 0.33 0.003 0.58 1.38 0.55 1.20 1.45 40S ribosomal protein S17 (40S) * 0.73 0.35 À0.17 0.01 0.32 0.40 0.32 0.54 0.22 1.13 Expansin precursor 3.64 0.17 3.48 0.67 2.03 1.51 2.03 1.51 0.36 À0.22

The columns indicate the means of qRT-PCR data from three independent biological replicates (P < 0.05) and microarray log2 ratios of differentially expressed genes (DEGs) from galls and giant cells (GCs) at 1, 3, 7 and 14 dpi vs their corresponding uninfected primary root fragments and laser capture microdissected (LCM) vascular cells, CCs, respectively, taken as controls. Green squares, repressed genes (À) and red squares, induced genes (+), with q < 0.05. White boxes represent no DEGs (q > 0.05) from the microarray at the infection points indicated. Those data with opposite expression in qRT-PCR relative to the microarray are underlined. *Genes also tested by in situ PCR.

confirmed the repression observed in the microarray for the com- (a) (b) patible interaction (Fig. 7a). By contrast, a remarkable induction (above nine-fold) was observed during the incompatible interac- tion (Motelle; Fig. 7a); both infection-site RNAs were compared against their noninfected controls (see Materials and Methods). These data indicate a contrasting behavior of TPX1 during RKN compatible and incompatible interactions in tomato. Subsequently, infection tests were performed in a homozygous TPX1-overexpressing line (TP3) from a tomato susceptible culti- var, Solanum lycopersicum cv Pera, carrying a single T-DNA inser- (c) (d) tion, which showed apparently normal root growth in vitro, but exhibited peroxidase activity and lignin content higher than con- trols. It was selected among three independent lines because it presented the highest peroxidase activity (El Mansouri et al., 1999). In TP3, nematode infection was severely impaired, as shown by the significantly reduced number of galls formed per plant (c. 35% reduction compared with wild-type controls; P < 0.05; Fig. 7b). Nematode reproduction parameters, such as the number of eggs and number of egg masses/root dry weight, were strongly affected (35 and 25% reduction, respectively Fig. 6 In situ reverse transcription-polymerase chain reaction ( RT-PCR) hybridization in longitudinal sections of tomato (Solanum lycopersicum) Fig. 7d,e). The number of egg masses/gall was also reduced (35% giant cells (GCs) at 7 d post infection (dpi). 40S, ribosomal protein S1; STK, reduction, Fig. 7c). Consistently, galls in which the nematode putative serine/threonine kinase. Asterisks indicate GCs. (a, c) Sections with had successfully established showed less expanded GCs than con- control reactions, no primers added. (b, d) Specific amplification signals are trol galls at 3 and 7 dpi (Fig. 8a). GC areas in gall sections observed as SYBR fluorescence in the GC cytoplasm (long white arrows). showed significant differences (P < 0.05) between the nontrans- Fluorescence of the nuclei also corresponds to SYBR, as it nonspecifically binds double-stranded DNA (Gal et al., 2006). Hence, only cytoplasmic genic control and the TPX1-overexpressing line (TPX1-OE)at label corresponds to specific transcript signals. Bars, 50 lm. Fluorescence in both infection stages, with more than a four-fold variation at the cytoplasm of cells adjacent to GCs (short white arrows in b). 7 dpi (Fig. 8b).

New Phytologist (2013) 197: 1276–1290 Ó 2013 The Authors www.newphytologist.com New Phytologist Ó 2013 New Phytologist Trust New Phytologist Research 1285

(a) (a)

(b) (c) (d) (e)

(b) Fig. 7 Functional analysis of tomato (Solanum lycopersicum) peroxidase TPX1. (a) Fold change (FC) values after quantitative reverse transcription- polymerase chain reaction (qRT-PCR) of TPX1 transcripts from susceptible (Moneymaker; mi/mi) and resistant (Motelle, Mi-1/Mi-1) nearly isogenic lines infected with Meloidogyne javanica second-stage juveniles (J2s). Bars indicate standard errors from three independent experiments with at least 10 plants each. Infection sites were hand-dissected 2 d after inoculation together with equivalent root segments from uninfected plants. (b–e) Infection and reproduction parameters in TPX1- overexpressing plants (TPX1-OE) relative to wild-type controls, both Solanum lycopersicum cv Pera. The number of galls per plant was assessed at 7–15 d after inoculation and reproduction parameters (number of egg masses per galls, number of eggs and egg masses per root dry weight) at 60–70 d after inoculation. The root dry weight was used to Fig. 8 Functional analysis of tomato peroxidase TPX1 in tomato (Solanum normalize most of the parameters measured. Data are representative from lycopersicum) plants. (a) Ten-micrometer representative sections of fixed six independent in vitro experiments. Standard errors are represented. galls from wild-type controls and TPX1-overexpressing plants (TPX1-OE) Asterisks mark statistically significant differences (P < 0.05) between TPX1- in a susceptible cultivar, S. lycopersicum cv Pera. Asterisks mark giant cells l OE and wild-type controls. (GCs); bars, 100 m. (b) Average of the total area occupied by GCs from consecutive gall sections measured for two independent representative galls of TPX1-OE and wild-type controls at each infection point (dpi, days Discussion post-infection). Bars indicate standard errors from seven sections. Asterisks mark statistically significant differences (P < 0.05) between TPX1-OE and A comprehensive examination of the still poorly understood wild-type controls. transcriptional events associated with GC differentiation in crops, such as tomato, may reveal information relevant for nematode control. This is the first time that differentially expressed tran- the later stages of gall development (7–14 dpi, Figs S1, S2). The scripts of GCs from equivalent developmental stages have been percentage of repressed genes was close to 71% in early stages, compared between Arabidopsis and tomato, under the same but only c. 50% in later stages, in accordance with the observed experimental conditions and data processing. This approach changes in GCs (Fig. 2). This indicates that transcriptional identified a group of genes consistently down-regulated in Ara- changes during the first stages of nematode infection include the bidopsis and tomato feeding sites; among them is a peroxidase, selective down-regulation of transcription, perhaps to allow ini- TPX1, with a contrasting behavior in compatible and incompati- tial migration/establishment, appropriate GC induction and gall ble interactions. TPX1 overexpression in a susceptible cultivar formation (Schaff et al., 2007). severely interfered with RKN infection and reproduction. When tomato GC and gall differential transcriptomes at the same developmental stages (3 and 7 dpi) and from the same experiments were compared, many DEGs in GCs were not Comparative expression profiles of tomato GCs and galls at DEGs in galls (Figs 5, S3), in agreement with Arabidopsis galls/ different developmental stages GCs at 3 dpi (Barcala et al., 2010). This suggests that tomato GC A clear boundary was found in the molecular changes detected differentiation also requires specific transcriptional changes dif- between the initial stages of nematode infection (1–3 dpi) and ferent from the rest of the gall.

Ó 2013 The Authors New Phytologist (2013) 197: 1276–1290 New Phytologist Ó 2013 New Phytologist Trust www.newphytologist.com New 1286 Research Phytologist

One of the categories with contrasting expression between GCs Particularly relevant is a set of induced genes that might regu- and their corresponding galls is ‘Stress’, with a high proportion of late the GC differential expression pattern and cellular identity, genes repressed in GCs, but not DEGs or induced in galls (i.e. as their products are involved in chromatin remodeling and tran- pathogenesis-related proteins, PRPs; Fig. S3; Table S9), in accor- scriptional control. Among them are members of the NF-Y gene dance with previous observations in Arabidopsis GCs (Barcala family of TFs with putative roles in cellular differentiation (Sie- et al., 2010) and tomato galls at 1, 5 and 10 dpi (Bar-Or et al., fers et al., 2009), induced from 3 dpi and increasing at 7 dpi. 2005; Bhattarai et al., 2008). The vast majority of ‘Stress’ genes Likewise, the only member of the MYB family induced in GCs is either co-expressed in GCs and galls, or GC distinctive, were a homolog of PHANTASTICA, involved in leaf organogenesis repressed (Table S9), suggesting that nematodes trigger a defense and cell differentiation, as reported in tomato and Medicago response in galls different from their GCs. Other obligate bio- truncatula GCs (Koltai & Bird, 2000; Koltai et al., 2001). Other trophic pathogens, such as phytopathogenic bacteria (Truman genes encode a chromomethylase (CMT) involved in DNA et al., 2006; Aslam et al., 2008), biotrophic and hemibiotrophic methylation (Lindroth et al., 2001), proteins related to histone fungi (Cooper et al., 2008) and symbiotic microorganisms methylation and deacetylation (HDAC), matrix-attachment region (Maunoury et al., 2010; Damiani et al., 2012), can disable host (MAR)-binding proteins, MSI (multicopy suppressor of ira1) defenses. Although putative nematode suppressors of plant proteins important for chromatin organization, condensation or defense have been suggested (Hewezi et al., 2010; reviewed by modification (Hollender & Liu, 2008) and a DICER-LIKE Smant & Jones, 2011), knowledge on the mechanisms governing coding gene with functions in the miRNA-mediated silencing this suppression is still sparse. Consistent with all of these machinery, all highly expressed in GCs (Tables S1, S2, S6; observations, we found that the vast majority of genes encoding Fig. S4; Kohler et al., 2003). These proteins can influence pro- TFs and enzymes related to chromatin modification were GC dis- cesses correlated with gene silencing and are key to the mainte- tinctive and not similarly regulated in galls (Figs S3, S4; nance of epigenetic gene expression patterns (Smith et al., 2007; Table S9). Knizewski et al., 2008), and their induction might be related to the observed large-scale gene repression during feeding site devel- opment. Recent analyses have reported several nematode-secreted Reprogramming GCs in tomato proteins with putative nuclear localization, DNA binding or Within developing GCs, 1051 transcripts were differentially chromatin modification domains, one directly localized in the expressed vs vascular control LCM cells from noninfected roots. nucleus (Bellafiore et al., 2008; Jaouannet et al., 2012). In agree- The DEG number increased at 7 dpi, suggesting increased tran- ment with these observations are those obtained in the scription complexity between 3 and 7 dpi (Fig. 2). At 7 dpi, 64 of Arabidopsis GC transcriptome (Barcala et al., 2010). All this evi- 65 genes classified in ‘Protein synthesis’, encoding mainly ribo- dence indicates that the output of secreted nematode effectors somal proteins and translation elongation factors, were up-regu- might be a series of dynamic changes in chromatin structure. As lated, in agreement with other transcriptomic studies in tomato these processes are necessary for cell identity and differentiation, and Arabidopsis galls (Jammes et al., 2005; Schaff et al., 2007; they might participate in the triggering of GC fate in its initials, Bhattarai et al., 2008; Tables S1, S2). This could be related to the developing root vascular stem cells. In addition to these genes higher transcription rate and increased protein synthesis needed up-regulated in GCs, the data suggest that transcriptional repro- for GC differentiation and growth (Bird, 1961). gramming of GCs acts in concert with large-scale gene repression, One of the first detectable signs of GC initiation is nuclear particularly obvious at 3 dpi (Fig. 2). For example, genes encod- division with partial cytokinesis, the number of nuclei per GC ing TFs from the WRKY, bHLH and NAC families, involved in nearly doubling during each of the first 4 d after infection (Starr, stress responses or in secondary wall thickening (Toledo-Ortiz 1993; Caillaud et al., 2008a). Functional and expression analyses et al., 2003; Olsen et al., 2005; Eulgem & Somssich, 2007; of CCS52 (a cell cycle switch promoting endoreduplication) Zhong et al., 2008), were all repressed in 3- and 7-dpi GCs indicate that endoreduplication takes place in early-developing (Fig. S4; Table S5). GCs (reviewed in De Almeida & Favery, 2011). G1/S transition and S-phase genes, that is, those coding for histones, minichro- Down-regulation of gene expression in RNK feeding sites is mosome maintenance (MCM) proteins and DNA helicases, were conserved in Arabidopsis and functionally relevant in up-regulated in tomato GCs (Tables S1, S2). MCM proteins are tomato part of a complex that regulates DNA replication initiation (Shultz et al., 2007). In Arabidopsis, the MCM family member It is generally accepted that one of the first events in feeding site PROLIFERA (PRL) was expressed in GCs from the earliest formation is the reprogramming of developmental fate in vascu- developmental stages (Huang et al., 2003). Furthermore, the lar cells towards GC identity (Williamson & Hussey, 1996). expression of 11 histone-coding genes increased between 3 and Although some evidence was provided by meta-analysis of tran- 7 dpi, notably the core histones H3 and H4 (FCs of 8 and 11, scriptomic data (Barcala et al., 2010), no demonstration of the respectively; Table S2). H4 deposition and modification have molecular events during successful GC formation from vascular been related to cell proliferation, and this is a well-characterized initials has been provided to date. According to our GC-specific gene up-regulated during the G1/S transition (Sanchez et al., transcriptome at very early stages, part of this reprogramming 2008). could rely on large-scale repression of gene expression (Fig. 2;

New Phytologist (2013) 197: 1276–1290 Ó 2013 The Authors www.newphytologist.com New Phytologist Ó 2013 New Phytologist Trust New Phytologist Research 1287

Table 1). Gene down-regulation has been described in resistance phase in resistant soybean (Klink et al., 2009, 2010). Arabidopsis and tomato gall transcriptomes at mid–late infection By contrast, peroxidases were repressed in Arabidopsis syncytia at stages and from the expression analysis of several gene promoters, 5 and 15 dpi (Szakasits et al., 2009). Contrasting results from but evidence has been focused mainly on stress-related genes, as syncytia could indicate differential regulation depending on plant also observed in this study (Caillaud et al., 2008b; Fig. S3). For species or may require more detailed analysis. However, in example, all WRKYs with a putative tomato–Arabidopsis homo- tomato and Arabidopsis GCs, TPX1 and its homologs, as well as log were repressed, either co-regulated or not (Table S5). genes for lignin biosynthesis, were consistently repressed (Fig. 4, Although, none was identified as WRK72a or b, both involved in Table 3), and may be necessary for gall and/or GC differentia- basal resistance in tomato and also related to Mi-1-dependent tion. Further support for this interpretation is that TPX1 tran- immunity (Bhattarai et al., 2010), the data are in agreement with scripts were highly induced in infected Mi-1 resistant plants plant defense suppression in GCs. Recently, the repression of relative to the susceptible mi-1 line (Fig. 7), which agrees with defense-related genes has been suggested as a common target for the assumption that peroxidase repression is characteristic of rhizobia and nematodes at the cellular level (Damiani et al., tomato–nematode compatible interactions relative to incompati- 2012). We present the first direct molecular evidence that tran- ble interactions involving resistant Mi-1 genotypes (Schaff et al., scriptional repression patterns are highly conserved in early 2007). phases of GC development in tomato and Arabidopsis, as 76.5% The repression of particular signaling cascades occurs during of the tomato–Arabidopsis homologs found in TOM1 were cell differentiation (Sawa et al., 2005; Ito & Sun, 2009). RKN co-repressed in both species, whereas only 0.7% were co-induced secretions, such as the CLE-like 16D10 peptide, may play a in 3-dpi GCs (Table 1). role in GC formation from procambial cells by restricting their In silico comparison revealed that genes from the phenylpropa- differentiation into xylem elements (reviewed in Gheysen & noid pathway, coding enzymes involved in lignin biosynthesis Fenoll, 2011). Moreover, GCs might also arise from partially and deposition, were significantly over-represented, considering differentiated tracheary elements (Bird, 1996; Barcala et al., the genes with a corresponding counterpart in Arabidopsis and 2010). In this study, we provide strong evidence that the tomato GCs, consistently in three independent analyses (Table 2; repression of a gene encoding an enzyme involved in lignin Fig. 4). Phenylpropanoids are a widely diverse group of mole- deposition/cross-linking is important for successful plant–nem- cules, some involved directly in plant defenses, such as salicylic atode interaction, probably during GC establishment. This acid (SA) (Chen et al., 2009; Fraser & Chapple, 2011). However, might also apply to other enzymes, as the process seems to be all phenylpropanoid-related DEGs in tomato and Arabidopsis highly conserved between tomato and Arabidopsis. Given the were found in the same lignin precursor biosynthesis route, but striking amplitude and cross-species conservation of transcrip- none was involved in the synthesis of other secondary metabo- tional down-regulation in incipient GCs, together with the lites, such as SA or flavonoids, down-stream from cinnamic acid increased expression of many genes involved in chromatin and feruloyl-CoA (Chen et al., 2009; Fraser & Chapple, 2011; remodeling, it seems likely that cell fate manipulation through Fig. 4). Furthermore, many genes down-regulated in GCs reprogramming of these genes is at the heart of the cell differ- (3–7 dpi) and in early-developing galls (1–3 dpi, but not at entiation process triggered by nematodes. Their targeted over- 7–14 dpi) play a direct role in lignin content and composition, expression or the identification of gain-of-function variants demonstrated through mutant analysis in Arabidopsis (as genes would provide additional tools for integrated nematode man- encoding cinnamate-4-hydroxylase (C4H), 4-coumarate: CoA agement in crops. ligase (4CL), p-coumarate 3-hydroxylase (C3H), caffeoyl CoA 3- However, the transcriptional profiles of the genes putatively O-methyltransferase (CCoAOMT), cinnamoyl-CoA reductase involved in the regulation of gene expression were not conserved (CCR), ferulate 5-hydroxylase (F5H) and caffeic acid between the two species (Figs 4, S4, Table S5), suggesting that O-methyltransferase (COMT); Fraser & Chapple, 2011; Fig. 4). gene networks or signaling cascades leading to downstream Repression of this particular group of genes might be crucial for responses for GC differentiation and/or maintenance might appropriate gall and GC formation at the initial stages. Further- diverge. This, together with the suggested high redundancy in more, plant peroxidases are highly diverse and grouped into sev- plant targets for nematode effectors or in other plant proteins eral classes (PeroxiBase; Passardi et al., 2007); some have been needed for feeding site development (Gheysen & Fenoll, 2011), demonstrated to participate directly in lignin cross-linking in could be part of the strategy of polyphagous nematodes able to Arabidopsis (Almagro et al., 2009). In tomato, TPX1 mediates a infect most plant species. late step in lignin monomer cross-linking (Lucena et al., 2003), and was drastically and consistently repressed in GCs and galls as Acknowledgements a function of the infection time from the early stages (Table 3). Moreover, nematode infection and reproduction were severely We thank Maria Sanchez for technical assistance and Dr Ques- impaired in TPX1-overexpressing tomato susceptible plants, as ada for kindly providing the TP3 line. This work was supported was GC expansion (Figs 7, 8). However, genes involved in the by grants from the Spanish Government to C.E. (AGL2010- phenylpropanoid pathway and some peroxidases were up-regu- 17388) and C.F. (CSD2007-057), and the Castilla-La Mancha lated in susceptible soybean syncytia from early to medium infec- Government to C.F. and C.E. (PCI08-0074-0294); R.S. tion stages (Ithal et al., 2007; Klink et al., 2007) and during the acknowledges grants BIO2010-21739, EUI2008-03666 and

Ó 2013 The Authors New Phytologist (2013) 197: 1276–1290 New Phytologist Ó 2013 New Phytologist Trust www.newphytologist.com New 1288 Research Phytologist

CSD2007-00057-B; K.L. and J.T. were supported by the UK Cooper AJ, Latunde-Dada AO, Woods-Tor A, Lynn J, Lucas JA, Crute IR, Biotechnology and Biological Sciences Research Council. The Holub EB. 2008. Basic compatibility of Albugo candida in Arabidopsis thaliana data obtained from the gall and GC RNA hybridizations have and Brassica juncea causes broad-spectrum suppression of innate immunity. Molecular Plant–Microbe Interactions 21: 745–756. been submitted to GEO, www.ncbi.nlm.nih.gov/geo (accession Damiani I, Baldacci-Cresp F, Hopkins J, Andrio E, Balzergue S, Lecomte P, no. GSE30048). Puppo A, Abad P, Favery B, Herouart D. 2012. Plant genes involved in harbouring symbiotic rhizobia or pathogenic nematodes. New Phytologist 194: 511–522. References De Almeida J, Favery B. 2011. Unravelling the plant cell cycle in nematode induced feeding sites. In: Jones J, Gheysen G, Fenoll C, eds. Genomics and Adie BA, Perez-Perez MM, Godoy MP, Sanchez-Serrano JJ, Schmelz EA, molecular genetics of plant–nematode interactions. Dordrecht, the Netherlands: Solano R. 2007. ABA is an essential signal for plant resistance to pathogens Springer, 349–368. affecting JA biosynthesis and the activation of defenses in Arabidopsis. Plant El Mansouri I, Mercado JA, Santiago-Domenech N, Pliego-Alfaro F, Valpuesta Cell 19: 1665–1681. V, Quesada MA. 1999. Biochemical and phenotypical characterization of Almagro L, Gomez Ros LV, Belchi-Navarro S, Bru R, Ros Barcelo A, Pedreno~ transgenic tomato plants over-expressing a basic peroxidase. Physiologia MA. 2009. Class III peroxidases in plant defence reactions. Journal of Plantarum 106: 355–362. Experimental Botany 60: 377–390. Escobar C, Sigal B, Mitchum M. 2011. Transcriptomic and proteomic analysis Aslam SN, Newman MA, Erbs G, Morrissey KL, Chinchilla D, Boller T, Jensen of the plant response to nematode infection. In: Jones J, Gheysen G, Fenoll C, TT, De Castro C, Ierano T, Molinaro A et al. 2008. Bacterial polysaccharides eds. Genomics and molecular genetics of plant–nematode interactions. Dordrecht, suppress induced innate immunity by calcium chelation. Current Biology 18: the Netherlands: Springer, 157–176. 1078–1083. Eulgem T, Somssich IE. 2007. Networks of WRKY transcription factors in Barcala M, Garcia A, Cabrera J, Casson S, Lindsey K, Favery B, Garcia-Casado defense signaling. Current Opinion in Plant Biology 10: 366–371. G, Solano R, Fenoll C, Escobar C. 2010. Early transcriptomic events in Fei Z, Joung JG, Tang X, Zheng Y, Huang M, Lee JM, McQuinn R, Tieman microdissected Arabidopsis nematode-induced giant cells. The Plant Journal DM, Alba R, Klee HJ et al. 2011. Tomato functional genomics database: a 61: 698–712. comprehensive resource and analysis package for tomato functional genomics. Bar-Or C, Kapulnik Y, Koltai H. 2005. A broad characterization of the Nucleic Acids Research 39: D1156–D1163. transcriptional profile of the compatible tomato response to the plant parasitic Fosu-Nyarko J, Jones MG, Wang Z. 2009. Functional characterization of root knot nematode Meloidogyne javanica. European Journal of Plant Pathology transcripts expressed in early-stage Meloidogyne javanica-induced giant cells 111: 181–192. isolated by laser microdissection. Molecular Plant Pathology 10: 237–248. Bellafiore S, Shen Z, Rosso MN, Abad P, Shih P, Briggs SP. 2008. Direct Fraser CM, Chapple C. 2011. The phenylpropanoid pathway in Arabidopsis. identification of the Meloidogyne incognita secretome reveals proteins with host The Arabidopsis Book 9: e0152. cell reprogramming potential. PLoS Pathogens 4: e1000192. Gal TZ, Aussenberg ER, Burdman S, Kapulnik Y, Koltai H. 2006. Expression of Benjamini Y, Hochberg Y. 1995. Controlling the False Discovery Rate: a a plant expansin is involved in the establishment of root knot nematode practical and powerful approach to multiple testing. Journal of the Royal parasitism in tomato. Planta 224: 155–162. Statistical Society 57: 289–300. Gamborg OL, Miller RA, Ojima K. 1968. Nutrient requirements of suspension Bhattarai KK, Xie QG, Mantelin S, Bishnoi U, Girke T, Navarre DA, cultures of soybean root cells. Experimental Cell Research 50: 151–158. Kaloshian I. 2008. Tomato susceptibility to root-knot nematodes requires an Gheysen G, Fenoll C. 2002. Gene expression in nematode feeding sites. Annual intact jasmonic acid signaling pathway. Molecular Plant–Microbe Interactions Review of Phytopathology 40: 191–219. 21: 1205–1214. Gheysen G, Fenoll C. 2011. Arabidopsis as a tool for the study of plant– Bhattarai KK, Atamian HS, Kaloshian I, Eulgem T. 2010. WRKY72-type nematode interactions. In: Jones J, Gheysen G, Fenoll C, eds. Genomics and transcription factors contribute to basal immunity in tomato and Arabidopsis molecular genetics of plant–nematode interactions. Dordrecht, the Netherlands: as well as gene-for-gene resistance mediated by the tomato R gene Mi-1. The Springer, 139–156. Plant Journal 63: 229–240. Hammes UZ, Schachtman DP, Berg RH, Nielsen E, Koch W, McIntyre Bird AF. 1961. The ultrastructure and histochemistry of a nematode- LM, Taylor CG. 2005. Nematode-induced changes of transporter gene induced giant cell. Journal of Biophysical and Biochemical Cytology 11: expression in Arabidopsis roots. Molecular Plant–Microbe Interactions 18: 701–715. 1247–1257. Bird DM. 1996. Manipulation of host gene expression by root-knot nematodes. Hewezi T, Howe PJ, Maier TR, Hussey RS, Mitchum MG, Davis EL, Baum Journal of Parasitology 82: 881–888. TJ. 2010. Arabidopsis spermidine synthase is targeted by an effector protein of Bird DM, Wilson MA. 1994. DNA sequence and expression analysis of root- the cyst nematode Heterodera schachtii. Plant Physiology 152: 968–984. knot nematode-elicited giant cell transcripts. Molecular Plant–Microbe Hewezi T, Maier T, Nettleton D, Baum T. 2012. The Arabidopsis Interactions 7: 419–424. MicroRNA396-GRF1/GRF3 regulatory module acts as a developmental Caillaud MC, Dubreuil G, Quentin M, Perfus-Barbeoch L, Lecomte P, de regulator in the reprogramming of root cells during cyst nematode infection. Almeida Engler J, Abad P, Rosso MN, Favery B. 2008b. Root-knot Plant Physiology 159: 321–335. nematodes manipulate plant cell functions during a compatible interaction. Hollender C, Liu Z. 2008. Histone deacetylase genes in Arabidopsis development. Journal of Physiology 165: 104–113. Journal of Integrative Plant Biology 50: 875–885. Caillaud MC, Lecomte P, Jammes F, Quentin M, Pagnotta S, Andrio E, de Huang GR, Dong AR, Davis EL, Baum TJ, Hussey RS. 2006. A root-knot Almeida Engler J, Marfaing N, Gounon P, Abad P et al. 2008a. MAP65-3 nematode secretory peptide functions as a ligand for a plant transcription microtubule-associated protein is essential for nematode-induced giant cell factor. Molecular Plant–Microbe Interactions 19: 463–470. ontogenesis in Arabidopsis. The Plant Cell 20: 423–437. Huang X, Springer PS, Kaloshian I. 2003. Expression of the Arabidopsis MCM Cardona A, Saalfeld S, Schindelin J, Arganda-Carreras I, Preibisch S, Longair gene PROLIFERA during root-knot and cyst nematode infection. M, Tomancak P, Hartenstein V, Douglas R. 2012. TrakEM2 software for Phytopathology 93:35–41. neural circuit reconstruction. PLoS ONE 7:1–8. Ithal N, Recknor J, Nettleton D, Maier T, Baum TJ, Mitchum MG. 2007. Casson S, Spencer M, Walker K, Lindsey K. 2005. Laser capture microdissection Developmental transcript profiling of cyst nematode feeding cells in soybean for the analysis of gene expression during embryogenesis of Arabidopsis. The roots. Molecular Plant–Microbe Interactions 20: 510–525. Plant Journal 42: 111–123. Ito T, Sun B. 2009. Epigenetic regulation of developmental timing in floral stem Chen Z, Zheng Z, Huang J, Lai Z, Fan B. 2009. Biosynthesis of salicylic acid in cells. Epigenetics 4: 564–567. plants. Plant Signaling & Behavior 4: 493–496.

New Phytologist (2013) 197: 1276–1290 Ó 2013 The Authors www.newphytologist.com New Phytologist Ó 2013 New Phytologist Trust New Phytologist Research 1289

Jammes F, Lecomte P, de Almeida-Engler J, Bitton F, Martin-Magniette ML, Rosso M-N, Grenier E. 2011. Other nematode effectors and evolutionary Renou JP, Abad P, Favery B. 2005. Genome-wide expression profiling of the constraints. In: Jones J, Gheysen G, Fenoll C, eds. Genomics and molecular host response to root-knot nematode infection in Arabidopsis. The Plant Journal genetics of plant–nematode interactions. Dordrecht, the Netherlands: Springer, 44: 447–458. 287–307. Jaouannet M, Perfus-Barbeoch L, Deleury E, Magliano M, Engler G, Vieira P, Saeed AI, Bhagabati NK, Braisted JC, Liang W, Sharov V, Howe EA, Li J, Danchin EGJ, Rocha MD, Coquillard P, Abad P et al. 2012. A root-knot Thiagarajan M, White JA, Quackenbush J. 2006. TM4 microarray software nematode-secreted protein is injected into giant cells and targeted to the nuclei. suite. Methods in Enzymology 411: 134–193. New Phytologist 194: 924–931. Sanchez MdeL, Caro E, Desvoyes B, Ramirez-Parra E, Gutierrez C. 2008. Klink VP, Alkharouf N, MacDonald M, Matthews B. 2005. Laser capture Chromatin dynamics during the plant cell cycle. Seminars in Cell and microdissection (LCM) and expression analyses of Glycine max (soybean) Developmental Biology 19: 537–546. syncytium containing root regions formed by the plant pathogen Heterodera Sawa S, Demura T, Horiguchi G, Kubo M, Fukuda H. 2005. The ATE genes glycines (soybean cyst nematode). Plant Molecular Biology 59: 965–979. are responsible for repression of transdifferentiation into xylem cells in Klink VP, Hosseini P, Alkharouf N, Matthews B. 2009. A gene expression Arabidopsis. Plant Physiology 137: 141–148. analysis of syncytia laser microdissected from the roots of the Glycine max Schaff JE, Nielsen DM, Smith CP, Scholl EH, Bird DM. 2007. Comprehensive (soybean) genotype PI 548402 (Peking) undergoing a resistant reaction after transcriptome profiling in tomato reveals a role for glycosyltransferase in Mi- infection by Heterodera glycines (soybean cyst nematode). Plant Molecular mediated nematode resistance. Plant Physiology 144: 1079–1092. Biology 71: 525–567. Schlink K. 2010. Down-regulation of defense genes and resource allocation into Klink VP, Overall C, Alkharouf N, MacDonald M, Matthews B. 2007. Laser infected roots as factors for compatibility between Fagus sylvatica and capture microdissection (LCM) and comparative microarray expression analysis Phytophthora citricola. Functional & Integrative Genomics 10: 253–264. of syncytial cells isolated from incompatible and compatible soybean (Glycine Shultz RW, Tatineni VM, Hanley-Bowdoin L, Thompson WF. 2007. Genome- max) roots infected by the soybean cyst nematode (Heterodera glycines). Planta wide analysis of the core DNA replication machinery in the higher plants 226: 1389–1409. Arabidopsis and rice. Plant Physiology 144: 1697–1714. Klink VP, Overall C, Alkharouf N, MacDonald M, Matthews B. 2010. Siefers N, Dang KK, Kumimoto RW, Bynum WE, Tayrose G, Holt BF. 2009. Microarray detection call methodology as a means to identify and compare Tissue-specific expression patterns of Arabidopsis NF-Y transcription factors transcripts expressed within syncytial cells from soybean (Glycine max) roots suggest potential for extensive combinatorial complexity. Plant Physiology 149: undergoing resistant and susceptible reactions to the soybean cyst nematode 625–641. (Heterodera glycines). Journal of Biomedicine and Biotechnology 2010:1–30. Smant G, Jones J. 2011. Suppression of plant defences by nematodes. In: Jones J, Knizewski L, Ginalski K, Jerzmanowski A. 2008. Snf2 proteins in plants: gene Gheysen G, Fenoll C, eds. Genomics and molecular genetics of plant–nematode silencing and beyond. Trends in Plant Science 13: 557–565. interactions. Dordrecht, the Netherlands: Springer, 273–286. Kohler C, Hennig L, Bouveret R, Gheyselinck J, Grossniklaus U, Gruissem W. Smith LM, Pontes O, Searle I, Yelina N, Yousafzai FK, Herr AJ, Pikaard CS, 2003. Arabidopsis MSI1 is a component of the MEA/FIE Polycomb group Baulcombe DC. 2007. An SNF2 protein associated with nuclear RNA complex and required for seed development. The Embo Journal 22: 4804– silencing and the spread of a silencing signal between cells in Arabidopsis. The 4814. Plant Cell 19: 1507–1521. Koltai H, Bird DM. 2000. Epistatic repression of PHANTASTICA and class 1 Smyth GK. 2004. Linear models and empirical Bayes methods for assessing KNOTTED genes is uncoupled in tomato. The Plant Journal 22: 455–459. differential expression in microarray experiments. Statistical Applications in Koltai H, Dhandaydham M, Opperman C, Thomas J, Bird D. 2001. Genetics and Molecular Biology 3:1–26. Overlapping plant signal transduction pathways induced by a parasitic Smyth GK, Speed T. 2003. Normalization of cDNA microarray data. Methods nematode and a rhizobial endosymbiont. Molecular Plant–Microbe Interactions 31: 265–273. 14: 1168–1177. Starr JL. 1993. Dynamics of the nuclear complement of giant cells induced by Lindroth AM, Cao X, Jackson JP, Zilberman D, McCallum CM, Henikoff S, Meloidogyne incognita. Journal of Nematology 25: 416–421. Jacobsen SE. 2001. Requirement of CHROMOMETHYLASE3 for Szakasits D, Heinen P, Wieczorek K, Hofmann J, Wagner F, Kreil DP, Sykacek maintenance of CpXpG methylation. Science 292: 2077–2080. P, Grundler FM, Bohlmann H. 2009. The transcriptome of syncytia induced Lucena MA, Romero-Aranda R, Mercado JA, Cuarteros J, Valpuesta V, by the cyst nematode Heterodera schachtii in Arabidopsis roots. The Plant Quesada MA. 2003. Structural and physiological changes in the roots of Journal 57: 771–784. tomato plants over-expressing a basic peroxidase. Physiologia Plantarum 118: Toledo-Ortiz G, Huq E, Quail PH. 2003. The Arabidopsis basic/helix-loop- 422–429. helix transcription factor family. The Plant Cell 15: 1749–1770. Maunoury N, Redondo-Nieto M, Bourcy M, Van de Velde W, Alunni B, Truman W, de Zabala MT, Grant M. 2006. Type III effectors orchestrate Laporte P, Durand P, Agier N, Marisa L, Vaubert D et al. 2010. a complex interplay between transcriptional networks to modify basal Differentiation of symbiotic cells and endosymbionts in Medicago truncatula defence responses during pathogenesis and resistance. The Plant Journal nodulation are coupled to two transcriptome-switches. PLoS ONE 5: e9519. 46:14–33. Moreau S, Verdenaud M, Ott T, Letort SB, de Billy FO, Niebel A, Gouzy JM, Urbanczyk-Wochniak E, Usadel B, Thimm O, Nunes-Nesi A, Carrari F, Davy Carvalho-Niebel F, Gamas P. 2011. Transcription reprogramming during root M, Blasing O, Kowalczyk M, Weicht D, Polinceusz A et al. 2006. Conversion 712 nodule development in Medicago truncatula. PLoS ONE 6: e16463. of MapMan to allow the analysis of transcript data from Solanaceous species: Olsen AN, Ernst HA, Leggio LL, Skriver K. 2005. NAC transcription factors: effects of genetic and environmental alterations in energy metabolism in the structurally distinct, functionally diverse. Trends in Plant Science 10:79–87. leaf. Plant Molecular Biology 60: 773–792. Passardi F, Theiler G, Zamocky M, Cosio C, Rouhier N, Teixera F, Margis- Wang Z, Potter RH, Jones MGK. 2001. A novel approach to extract and analyse Pinheiro M, Ioannidis V, Penel C, Falquet L et al. 2007. PeroxiBase: the cytoplasmic contents from individual giant cells in tomato roots induced by peroxidase database. Phytochemistry 68: 1605–1611. Meloidogyne javanica. International Journal of Nematology 11: 219–225. Portillo M, Fenoll C, Escobar C. 2006. Evaluation of different RNA extraction Wang Z, Potter RH, Jones MGK. 2003. Differential display analysis of gene methods for small quantities of plant tissue: combined effects of reagent type expression in the cytoplasm of giant cells induced in tomato roots by and homogenization procedure on RNA quality-integrity and yield. Physiologia Meloidogyne javanica. Molecular Plant Pathology 4: 361–371. Plantarum 128:1–7. Williamson VM, Hussey RS. 1996. Nematode pathogenesis and resistance in Portillo M, Lindsey K, Casson S, Garcia-Casado G, Solano R, Fenoll C, Escobar plants. The Plant Cell 8: 1735–1745. C. 2009. Isolation of RNA from laser-capture-microdissected giant cells at Zhong R, Lee C, Zhou J, McCarthy RL, Ye ZH. 2008. A battery of transcription early differentiation stages suitable for differential transcriptome analysis. factors involved in the regulation of secondary cell wall biosynthesis in Molecular Plant Pathology 10: 523–535. Arabidopsis. The Plant Cell 20: 2763–2782.

Ó 2013 The Authors New Phytologist (2013) 197: 1276–1290 New Phytologist Ó 2013 New Phytologist Trust www.newphytologist.com New 1290 Research Phytologist

Table S5 List of Arabidopsis–tomato transcription factors homo- Supporting Information logs Additional supporting information may be found in the online version of this article. Table S6 ID Interactive tomato gene finder of giant cell (GC) and gall differentially expressed genes (DEGs) as a function of Fig. S1 Microarray-based temporal changes of gall differentially the infection stage expressed genes (DEGs). Table S7 List of gall-exclusive differentially expressed genes Fig. S2 Overview of gall differentially expressed genes (DEGs) (DEGs) at 1 d post-infection (dpi) classified in functional categories by Mapman. Table S8 Examples of gall differentially expressed genes (DEGs) Fig. S3 Summary of exclusive and co-expressed differentially from different categories (Mapman) expressed genes (DEGs) between giant cells (GCs) and galls clas- sified by Mapman. Table S9 Examples of exclusive and co-regulated genes in galls and giant cells (GCs) by Mapman Fig. S4 Mapman map fitting of differentially expressed genes (DEGs) related to ‘RNA-regulation of transcription’ in galls and Table S10 List of distinctive giant cell (GC) genes giant cells (GCs). Table S11 List of genes with opposite expression patterns in Fig. S5 In situ reverse transcription-polymerase chain reaction giant cells (GCs) and galls (RT-PCR) hybridization of uninfected root longitudinal sections as controls. Table S12 Primers used for quantitative reverse transcription- polymerase chain reaction (qRT-PCR), in situ PCR analysis and Table S1 List of giant cell (GC)-distinctive differentially microarray sample labeling expressed genes (DEGs) with a fold change (FC) > 1.9 and < À1.9 Table S13 List of differentially expressed genes (DEGs) with no uniform pattern during infection in galls and giant cells (GCs) Table S2. Examples of giant cell (GC) differentially expressed genes (DEGs) from different categories (Mapman) Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the Table S3 Detailed information of TBLASTN and TBLASTX authors. Any queries (other than missing material) should be reciprocal analysis directed to the corresponding author for the article.

Table S4 Interactive search of tomato–Arabidopsis homologous differentially expressed genes (DEGs) in giant cells (GCs)

New Phytologist is an electronic (online-only) journal owned by the New Phytologist Trust, a not-for-profit organization dedicated to the promotion of plant science, facilitating projects from symposia to free access for our Tansley reviews.

Regular papers, Letters, Research reviews, Rapid reports and both Modelling/Theory and Methods papers are encouraged. We are committed to rapid processing, from online submission through to publication ‘as ready’ via Early View – our average time to decision is <25 days. There are no page or colour charges and a PDF version will be provided for each article.

The journal is available online at Wiley Online Library. Visit www.newphytologist.com to search the articles and register for table of contents email alerts.

If you have any questions, do get in touch with Central Office ([email protected]) or, if it is more convenient, our USA Office ([email protected])

For submission instructions, subscription and all the latest information visit www.newphytologist.com

New Phytologist (2013) 197: 1276–1290 Ó 2013 The Authors www.newphytologist.com New Phytologist Ó 2013 New Phytologist Trust Research

Differentially expressed small RNAs in Arabidopsis galls formed by Meloidogyne javanica: a functional role for miR390 and its TAS3-derived tasiRNAs

Javier Cabrera1, Marta Barcala1, Alejandra Garcıa1, Ana Rio-Machın2,Clemence Medina3, Stephanie Jaubert- Possamai3, Bruno Favery3, Alexis Maizel4, Virginia Ruiz-Ferrer5, Carmen Fenoll1 and Carolina Escobar1 1Universidad de Castilla-La Mancha, Facultad de Ciencias Ambientales y Bioquımica, Avda. Carlos III, s/n 45071 Toledo, Spain; 2Molecular Cytogenetics Group, Human Cancer Genetics Programme, Centro Nacional Investigaciones Oncologicas (CNIO), C/Melchor Fernandez Almagro, 3, 28029 Madrid, Spain; 3INRA, Universite Nice Sophia Antipolis, CNRS, UMR 1355- 7254 Institut Sophia Agrobiotech, 06900 Sophia Antipolis, France; 4Centre for Organismal Studies University of Heidelberg, Im Neuenheimer Feld, 230-69120 Heidelberg, Germany; 5Centro de Investigaciones Biologicas, CSIC, Av. Ramiro de Maeztu 9, 28040 Madrid, Spain

Summary Author for correspondence: Root-knot nematodes (RKNs) induce inside the vascular cylinder the giant cells (GCs) Carolina Escobar embedded in the galls. The distinctive gene repression in early-developing GCs could be facili- Tel: +34925268800, ext. 5476 tated by small RNAs (sRNA) such as miRNAs, and/or epigenetic mechanisms mediated by Email: [email protected] 24nt-sRNAs, rasiRNAs and 21-22nt-sRNAs. Therefore, the sRNA-population together with Received: 10 August 2015 the role of the miR390/TAS3/ARFs module were studied during early gall/GC formation. Accepted: 25 September 2015 Three sRNA libraries from 3-d-post-inoculation (dpi) galls induced by Meloidogyne javanica in Arabidopsis and three from uninfected root segments were sequenced following New Phytologist (2015) Illumina-Solexa technology. pMIR390a::GUS and pTAS3::GUS lines were assayed for nema- doi: 10.1111/nph.13735 tode-dependent promoter activation. A sensor line indicative of TAS3-derived tasiRNAs bind- ing to the ARF3 sequence (pARF3:ARF3-GUS) together with a tasiRNA-resistant ARF3 line Key words: galls, giant cells, meloidogyne, (pARF3:ARF3m-GUS) were used for functional analysis. miRNAs, rasiRNAs, silencing, small RNAs, The sRNA population showed significant differences between galls and controls, with high tasiRNAs. validation rate and correspondence with their target expression: 21-nt sRNAs corresponding mainly to miRNAs were downregulated, whilst 24-nt-sRNAs from the rasiRNA family were mostly upregulated in galls. The promoters of MIR390a and TAS3, active in galls, and the pARF3:ARF3-GUS line, indicated a role of TAS3-derived-tasiRNAs in galls. The regulatory module miR390/TAS3 is necessary for proper gall formation possibly through auxin-responsive factors, and the abundance of 24-nt sRNAs (mostly rasiRNAs) con- stitutes a gall hallmark.

cytosine methylation in DNA, a landmark of RNA-directed Introduction DNA methylation (PolIV-RdDM). Recently, a genetic RdDM Regulatory small RNAs (sRNAs) are a group of short noncoding pathway was uncovered in Arabidopsis. It utilizes 21–22-nt RNAs (20–24 nucleotides (nt) long) with diverse roles in gene siRNAs as well as 24-nt siRNAs, and methylates ta-siRNAs (Wu silencing at the transcriptional and post-transcriptional levels. et al., 2012; Kanno et al., 2013) and active transposable elements Arabidopsis sRNAs are dominated by short-interfering RNAs (TEs) by the combined activities of RDR6, DCL2, DCL4 and (siRNAs) and by microRNAs (miRNAs; Axtell, 2013; Bologna AGO1 (Nuthikattu et al., 2013). These 21–22-nt siRNAs guide & Voinnet, 2014). Among the endogenous Arabidopsis siRNAs AGO6 to its chromatin targets to establish TE expression- there are different sRNAs groups such as trans-acting short- dependent DNA methylation (Wu et al., 2012). Hence, a core of interfering RNAs (ta-siRNAs) and repeat-associated small inter- different Dicer-like proteins (DCLs), RNA-dependent RNA fering RNAs (rasiRNAs). Ta-siRNAs are produced through polymerases (RDRs) and Argonaute proteins (AGOs) participate miRNA-guided cleavage of noncoding primary transcripts that in the biogenesis and action of miRNAs, tasiRNAs and rasiRNAs are then converted into dsRNA by RDR6, whereas rasiRNAs are (Axtell, 2013; Bologna & Voinnet, 2014). generated mostly from transposon loci and DNA repeats. In recent years, next-generation sequencing has made clear AGO1/7- ta-siRNA complexes mediate the cleavage of mRNA the importance of sRNA regulation in different abiotic and from coding genes, whereas AGO4/6- rasiRNA complexes guide biotic plant stresses (Sunkar et al., 2012; Balmer & Mauch-

Ó 2015 The Authors New Phytologist (2015) 1 New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 2 Research Phytologist

Mani, 2013; Weiberg et al., 2014; Harfouche et al., 2015). between resistant/susceptible plants. A functional role for a Pathogen attack triggers massive miRNA changes, exerting regu- miRNA has been demonstrated only for miR396 in Arabidopsis latory roles through alteration of hormone pathways, or manip- syncytia by Hewezi et al. (2012). It was shown that miR396 ulating silencing pathways to counteract miRNA-mediated downregulation following GRF1/GRF3 induction is necessary for defenses, thus regulating plant immunity (reviewed in Balmer & correct syncytia initiation. However, subsequent miR396 induc- Mauch-Mani, 2013). For example, miR393, implicated in bac- tion was necessary once the syncytium is established and it does terial resistance by repressing auxin signaling, was upregulated not incorporate new cells. So far, DE sRNAs and their functional in response to the bacterial effector flg22, and targets TIR1, roles in RKN feeding sites have not been described, except for AFB2 and AFB3 which allow the stabilization of auxin signaling systemic stress responses in phloem after RKN inoculation, where repressor Aux/IAA proteins (Navarro et al., 2006) and viruses the miR319/TCP4 module acts as a systemic signal responder use suppressor proteins that interfere with the silencing machin- modulating a systemic defensive response mediated by jasmonic ery (Jagga & Gupta, 2014).It can therefore be concluded that acid (Zhao et al., 2015). miRNAs are very likely to be fundamental players in the concert Here we analyze the DE sRNAs during early gall develop- of broad-spectrum disease resistance (reviewed in Balmer & ment, by using three sRNA libraries from independent biologi- Mauch-Mani, 2013). cal replicates of hand-dissected 3-dpi galls formed by Plant sedentary endoparasitic nematodes are among the most M. javanica in Arabidopsis, and uninfected root segments at damaging parasites, causing severe agricultural losses equivalent positions of the root to obtain three control libraries. (Mitkowski & Abawi, 2003). Two types of sedentary endopar- We found significant differences between galls and control roots asitic nematodes are described depending on the type of feed- in the sRNA population. In galls, 21-nt sequences correspond- ing cells that they induce in roots, namely giant cells (GCs) ing mostly to miRNAs were downregulated, whereas 24-nt inside galls or knots for root-knot nematodes (RKN; sequences from rasiRNAs were upregulated. We studied the Meloidogyne spp; Escobar et al., 2015) and syncytia for cyst expression pattern and functional role of miR390, one of the nematodes (Heterodera spp. and Globodera spp.; Bohlmann, few upregulated miRNAs in galls and GCs at early infection 2015). The differentiation of a root vascular cell into a GC or stages. Its promoter is active in GCs and gall vascular tissues a syncytium, both highly specialized transfer cells, requires dra- and it regulates TAS3-derived tasiRNAs formation in galls. matic changes in gene expression (reviewed in Escobar et al., TAS3 was necessary for proper gall development, possibly 2011, 2015). Generalized gene repression is characteristic of through the control of auxin-responsive factors. This is the first early-developing GCs and galls, and constitutes a signature of report of a functional role during plant–nematode interactions early-developing GCs (Schaff et al., 2007; Caillaud et al., 2008; of a highly conserved ta-siRNA from bryophytes to vascular Barcala et al., 2010; Portillo et al., 2013) which includes plant plants. defense-related genes (reviewed in Smant & Jones, 2011; Hewezi & Baum, 2015). For example, peroxidase-coding genes Materials and Methods are repressed in compatible interactions, but upregulated in soybean resistant plants (Klink et al., 2009, 2010) and in Biological materials, growth conditions and nematode tomato resistant cultivars homozygous for Mi-1 (Bar-Or et al., inoculation 2005; Schaff et al., 2007). This is consistent with the func- tional role of tomato TPX1 in resistance to Meloidogyne All Arabidopsis thaliana (L.) Heynh lines were in Col-0 back- javanica (Portillo et al., 2013). The mechanisms mediating this ground. Seeds were sterilized, and plants grown and inoculated as massive gene repression in early-developing GCs/galls are in Cabrera et al. (2014a). unknown, but could involve a general differential expression For functional analysis, four independent infection tests (DE) of sRNAs. were performed for Col-0 (number of plants (n) = 280) and Few massive sequencing experiments have been performed to mir390a-2 (n = 217), and three independent infection tests uncover the role of sRNAs in the plant–nematode interaction with Col-0 (n = 114) and TAS3a-1 (n = 114). Gall number and those which have were mainly focused in cyst nematodes per main root was determined under a stereo microscope at (Hewezi et al., 2008; Li et al., 2012; Xu et al., 2014; Zhao et al., 14 dpi. For diameter measurements, at least 21 galls at 14 2015). In Arabidopsis infected with Heterodera schachtii, Hewezi dpi from three independent experiments were hand dissected, et al. (2008) identified 16 miRNAs DE in syncytia at 4 and/or photographed and measured with IMAGEJ (US National Insti- 7 d post inoculation (dpi). Several of them targeted transposons tutes of Health, Bethesda, MD, USA). For significant differ- or retrotransposons of different types, suggesting a role for these ences on infection level and gall diameter a Student’s t-test, miRNAs in controlling TE movement (Hewezi et al., 2008; P < 0.05, was performed using SPSS (IBM, Armonk, NY, Hewezi & Baum, 2015). Arabidopsis rdr and dcl mutants altered USA). in essential genes for sRNA biogenesis showed reduced suscepti- Meloidogyne javanica Treub, 1885 was maintained and in vitro bility to H. schachtii (Hewezi et al., 2008). Two sequencing infection assays were performed as described in Barcala et al. experiments shed light to the sRNA population from resistant (2010). For M. incognita, Arabidopsis seeds were stratified in and susceptible soybean lines infected with H. glycines (Li et al., 0.59 MS/0.8% agar plates at 23°C and 16 h : 8 h, light : dark 2012; Xu et al., 2014), showing several DE miRNAs and siRNAs photoperiod. Twenty days after germination, roots were

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 3 inoculated with 200 J2 per plant previously sterilized with 0.01% maximal copy number of miRNAs on reference (20 nt), maxi- HgCl2 and 0.7% streptomycin. mal free energy allowed for a miRNA precursor (18 kcal mol 1), maximal space between miRNA and miRNA* (300 nt), minimal base pairs of miRNA and RNA extraction, sRNA library construction and statistical miRNA* (16 nt), maximal bulge of miRNA and miRNA* analysis (4 nt), and maximal asymmetry of miRNA/miRNA* duplex RNA extractions from three independent samples of c. 300 3-dpi (4 nt). galls and three of c. 300 control root segments were performed Target prediction was performed on the Plant Small RNA and quality-assessed as in Portillo et al. (2009). Each RNA sam- Target Analysis Server (psRNATarget; http://plantgrn.noble.org/ ple extracted from pooled galls or control roots from three or four psRNATarget/; Dai & Zhao, 2011). The following restrictions independent experiments is thus considered a biological replicate. were established: maximum expectation (3.0), length for comple- Three of these RNA samples were independently analyzed by mentarity scoring (20), allowed maximum energy to unpair the massive sequencing for gall or control roots. In each independent target site (UPE; 25); flanking length around target site for target experiment, 20 plates containing 10 plants each (n = 200) were accessibility analysis (17 bp in upstream/13 bp in downstream); also used to collect galls or control root segments from uninfected and range of central mismatch leading to translational inhibition plants. Samples were quick-frozen in liquid nitrogen and stored (9–11 nt). at 80°C. Illumina-Solexa Sequencing-By-Synthesis technology was used q-PCR validation of the expression of miRNAs and their to generate the small-RNA libraries at Beijing Genomics Institute targets, histological analysis of GUS expression and gall (Shenzhen, China). Clean sequences were obtained after discard- phenotyping ing low-quality reads, 50 primer contaminants, and those without 30 primer, without the insert tag, with poly A and/or shorter than Expression levels of selected miRNAs DE in the libraries were ® 18 nt. analyzed by qRT-PCR using TaqMan Small RNA assays. The A reads per million (RPM) value was obtained for each probes used were: aly-miR156 h (463816_mat), miR163 (343), sequence in each library (number of reads for a sequence/number miR167d (000350), miR390a (001409miR775 (008366_mat), of total reads for all sequences in that library 9 106). Each miR780.2 (464297_mat), miR839 (008535_mat) and aly- sequence was tested for expression changes by a two-tailed miR857 (006362_mat)). qRT-PCR was performed as described heteroscedastic t-test using statistic software R (R Development by Frenquelli et al. (2010)). The independent RNA samples used Core Team, 2008) between their three independent RPM values for RNAseq were used for cDNA synthesis. Total RNA (10 ng) in galls and their three independent RPM values in control roots. was reverse-transcribed with the MicroRNA Reverse Transcrip- Expression differences between galls and roots were considered tion Kit (Applied Biosystems, Foster City, CA, USA) in a final significant when P < 0.05. Fold change (FC) was used to measure reaction volume of 5 ll. Independent reverse transcription reac- the expression differences between galls and control roots as the tions were set for each miR-snoR pair. qRT-PCR was performed ratio between the average of the RPM values of the three samples using TaqMan Fast Universal PCR Master Mix on a 7500 Fast in each group. The full raw sequencing data were submitted to Real-Time PCR System following the manufacturer’s protocol. the GEO database (http://www.ncbi.nlm.nih.gov/geo/) with the Each independent cDNA template was run in triplicate (384-well accession number: GSE71563. plate). Data were normalized to the expression of two small â nucleolar RNAs (predesigned TaqMan small RNA control assays snoR101 and snoR66; Applied Biosystems) and relative sRNA sequences annotation expression was calculated using the comparative Ct method DΔC Clean reads were mapped to the Arabidopsis or M. incognita (2 t). Tendencies were similar with both normalizers. genomes (http://www.arabidopsis.org/ or http://www6.inra.fr/ Validation of the transcript level for predicted miRNA target meloidogyne_incognita) by SOAP (Short Oligonucleotide Anal- genes was performed by qRT-PCR. RNA samples used for ysis Package; http://soap.genomics.org.cn/) and sRNAs were cat- sequencing were reverse transcribed with the High-Capacity egorized into types: rRNA, tRNA and snRNA were analyzed in cDNA Reverse Kit (Applied Biosystems), using 500 ng of RNA GenBank and in RFAM; known miRNAs and their isomiRs per reaction and following manufacturer’s instructions. Normal- (miRBase19), repeats, exon and intron were analyzed by BGI ization was referred to GAPC2 according to the method of Bar- Genomics in-house database. cala et al. (2010). Relative expression was calculated using the DΔ Unannotated sequences were screened with Mireap comparative Ct method (2 Ct). Primer sequences are listed in (http://sourceforge.net/projects/mireap/) which predicts novel Supporting Information Table S1. miRNA by exploring secondary structure, DICER cleavage site For all q-PCR data a Student’s t-test was performed (P < 0.05) and minimum free energy (MFE) of the unannotated small to identify FC differences between galls and uninfected controls. RNA tags. The following restrictions were imposed: minimal For tissue localization of GUS activity, seedlings from miRNA sequence length (18 nt), maximal miRNA sequence pMIR390a::GUS, pTAS3::GUS, pARF3:ARF3-GUS and pARF3: length (25 nt), minimal miRNA reference sequence length ARF3 m-GUS lines were treated and phenotyped as described in (20 nt), maximal miRNA reference sequence length (23 nt), Cabrera et al. (2014a).

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 4 Research Phytologist

(Fig. 2a), most of them ranging from 20 to 24 nt, the typical Results length of small RNAs processed by DICER (Ghildiyal & Zamore, 2009). Read-length distribution in each library was sim- Overview of sRNAs expressed in galls and control roots ilar except for the 24-nt sequences, which were consistently more Illumina-Solexa Sequencing-By-Synthesis technology was used to abundant in gall than in control root libraries (Fig. 2a). By con- generate three independent sRNA libraries from c. 300 hand- trast, 20- and 21-nt reads (the typical length of miRNAs; collected 3-dpi galls each, and three libraries from their corre- Ghildiyal & Zamore, 2009) were consistently larger in root than sponding control samples (see the Materials and Methods sec- in gall libraries (Fig. 2a). tion; Fig. 1a). An average of 10.1 million raw reads (Solexa 50 nt Fifty-eight percent of the unique sRNA sequences in the six reads) were obtained for the three gall libraries (G1, G2 and G3), libraries were present exclusively in galls, whilst only 16.2% were and an average of 9.8 million for control roots (R1, R2, R3) exclusive from control roots and 25.8% were shared between (Fig. 1b). After filtering reads, the number of clean reads was both (Fig. 2b). The matched sequences were aligned against reduced to an average of 9.8 million for galls and 9.5 million for sequences in GenBank, Rfam and miRBase to find previously control roots (Fig. 1b). The similar overall numbers obtained for known sRNAs. Fig. 2(c) summarizes the number of sequences the three libraries in both sample types indicate a high uniformity found in each library corresponding to different sRNA types. among the three biological replicates used for the analysis. An Approximately half of the unique sequences (54% and 50% for average of 2121 427 clean reads corresponded to unique sRNA galls and control roots, respectively) were unannotated; that is, sequences in the three gall libraries, almost twice the 1241 364 either they could not be mapped to the Arabidopsis genome (and reads obtained for the control root libraries (Fig. 1b). In the six could be of nematode origin) or they hit Arabidopsis sequences libraries, 99.9% of the reads had a length between 18 and 29 nt not classified as sRNAs (Fig. 2c).

(a) (b)

Fig. 1 Schematic representation of sRNA library construction and clean read generation. (a) Approximately 900 at 3-d-post-infection (dpi)-galls inducedby Meloidogyne javanica in Arabidopsis thaliana were collected in three independent biological replicates (left panel) for construction of gall libraries (G1, G2 and G3). Equivalent root segments from uninfected plants grown in the same conditions were collected also in triplicate for uninfected root libraries (R1, R2 and R3; right panel). Dotted lines in the pictures indicate the hand-dissected segments collected. (b) Number of discarded or retained reads is indicated in each step, as well as the percentage that this number represents with respect to the total number of high-quality reads for each library. Clean reads are obtained after discarding reads without 30 primers, with 50 primers contaminants, without the insert tag, with poly A, or shorter than 18 nt, and low-quality sequences. Unique reads represent those distinct sequences read.

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 5

(a) (c)

(b)

Fig. 2 General sRNA analysis in galls induced by Meloidogyne javanica in Arabidopsis thaliana and control libraries. (a) Distribution of clean reads (in %) in the six libraries catalogued into length categories. Note the marked abundance of the 24-nt sequences in gall libraries, with an average of 3803 611 sequences as compared to the 2826 085 found in libraries from control roots. (b) Venn diagram indicating number and percentage of unique reads exclusive and/or or shared by galls and control roots libraries. (c) Schematic representation of the annotation process of the reads by alignment to the genome and adscription to different small RNA categories (miRNA, rRNA, tRNA, snRNA, snoRNA, repeat, exon, intron). Note that the abundance of intron and exon antisense sequences is higher in galls than in control libraries.

upregulated and repressed/downregulated are used throughout Differential expression of sRNAs and identification of their the text to mean mature sRNA levels higher or lower than in putative targets in the Arabidopsis genome their corresponding controls, respectively. Interestingly, the From the 338 known Arabidopsis miRNAs, 288 were detected average number of reads corresponding to known miRNAs in in at least one library, whereas 50 were not identified in any the three control root libraries (24 227) was 1.7-fold that of read through the six libraries (Table S2). The overall numbers the three gall libraries (14 617), suggesting a global downregu- of reads for each miRNA among the libraries from either galls lation of miRNA in galls (Table S2). Accordingly, after normal- or control roots presented similar trends (Table S2), reinforcing ization of the reads as ‘reads per million’ and analysis of DE the validity of the sequencing in the three biological replicates. miRNAs (see the Materials and Methods section), only 11 of From the 265 different miRNAs identified in each library, 242 the 288 miRNAs detected were upregulated (P < 0.05), whereas were common to both conditions, 23 were exclusive of galls 51 were downregulated in galls (P < 0.05; Table 1). From the and 23 were only present in control root libraries (Table S2). 62 DE miRNAs, large miRNA families as miR166 and For simplicity, the terms differentially expressed (DE), induced/ miR169 were consistently downregulated in galls (Table 1). By

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 6 Research Phytologist

Table 1 Average number of reads for known miRNAs in gall and control Table 1 (Continued) root libraries Average no. of reads Average no. of reads Gall libraries Root libraries FC P-value Gall libraries Root libraries FC P-value ath-miR2111b-5p 2 21 10.82 0.001 ath-miR156i 5 1 7.06 0.032 ath-miR5637 0 4 12.40 0.001 ath-miR839 19 5 3.74 0.028 ath-miR172e 2 35 17.67 0.000 ath-miR5657 64 18 3.47 0.017 ath-miR851-3p 10 3 2.94 0.016 The three libraries for galls induced by Meloidogyne javanica in ath-miR390a 53 602 19 921 2.70 0.007 Arabidopsis thaliana and roots are represented separately. Fold change ath-miR390b 53 586 19 917 2.70 0.007 (FC) and statistical significance (P-value) for the differentially expressed ath-miR391 42 16 2.62 0.002 miRNAs are indicated. Red, induced, and green, repressed miRNAs in galls ath-miR5655 113 45 2.46 0.042 tested by q-PCR in Fig. 3(a). ath-miR775 790 368 2.13 0.012 ath-miR156h 1478 720 2.02 0.020 ath-miR5643a 58 38 1.50 0.047 contrast, miR390a and miR390b were highly upregulated ath-miR159b 312 459 1.52 0.049 (FC = 2.70) in galls (Table 1) as were two members of the ath-miR166a 127 138 191 088 1.54 0.027 miR156 family (miR156 h-i; FC = 7.06; Table 1), although the ath-miR166g 121 322 182 951 1.55 0.025 majority of the members of this family were downregulated ath-miR166c 121 384 183 056 1.55 0.025 ath-miR166e 121 289 182 914 1.55 0.025 (miR156a-g, j). ath-miR166d 121 384 183 059 1.55 0.025 Fold change values for the DE miRNAs were validated by ® ath-miR166f 121 284 182 909 1.55 0.025 TaqMan assay-based real-time PCR (Fig. 3a). Seven miRNAs ath-miR166b 121 344 183 020 1.55 0.025 were selected for validation, three induced and four repressed in ath-miR5645f 113 176 1.59 0.044 galls, representative of different expression patterns in both ath-miR5645e 112 174 1.59 0.036 ath-miR5645a 112 176 1.60 0.037 libraries; that is, a high number of readings (e.g. miR390a), low ath-miR5645b 112 176 1.60 0.037 number of readings (e.g. miR857), and intermediate number of ath-miR165a 40 918 65 318 1.65 0.044 readings (e.g. miR775s). qPCR was done after cDNA synthesis ath-miR165b 39 533 63 672 1.67 0.044 with the same RNA samples used for the construction of the six ath-miR5635a 75 130 1.78 0.011 small RNA libraries, and resulted in a high validation rate of the ath-miR5635d 260 503 1.98 0.005 ath-miR156j 431 1009 2.40 0.010 sequencing results (6 out of 7 tested; Fig. 3a). Different sensitivi- ath-miR156b 191 210 461 888 2.47 0.013 ties of both techniques might explain the nonvalidated expression ath-miR156a 191 007 461 738 2.47 0.013 pattern of miR167d. ath-miR156c 191 007 461 738 2.47 0.013 The potential target genes of the known DE miRNAs were ath-miR156e 190 500 461 235 2.48 0.013 downloaded from the Plant MicroRNA Database (http://bioin ath-miR156f 190 497 461 238 2.48 0.013 ath-miR156d 192 256 470 427 2.50 0.014 formatics.cau.edu.cn/PMRD/; Zhang et al., 2010) and were ath-miR5653 448 1102 2.54 0.001 also predicted by using the psRNATarget web server ath-miR399e 5 15 3.02 0.028 (http://plantgrn.noble.org/psRNATarget/) accomplishing ath-miR156g 187 569 3.13 0.000 restrictive parameters (see the Materials and Methods section). ath-miR169f 41 129 3.30 0.004 The TAIR10 transcripts database was used as the reference ath-miR857 383.30 0.032 ath-miR169j 20 70 3.53 0.023 genome, identifying 108 putative target genes for the upregu- ath-miR169n 20 70 3.53 0.023 lated and 222 for the downregulated miRNAs in galls, respec- ath-miR169l 20 70 3.55 0.026 tively (Table S3). Some of the predicted targets had been ath-miR169g-5p 37 128 3.55 0.004 experimentally validated in different biological systems, for ath-miR163 56 200 3.66 0.015 example, those for miR390, miR156 or miR172 that preferen- ath-miR169i 20 72 3.67 0.021 ath-miR169h 6 21 3.79 0.023 tially target TAS3, trans-acting short-interfering RNA 3 ath-miR169k 6 21 3.79 0.023 (Montgomery et al., 2008), SQUAMOSA PROMOTER ath-miR167d 67 263 3.99 0.041 BINDING PROTEINS (SPLs; Wu & Poethig, 2006) and R- ath-miR169d 33 126 4.01 0.004 homologous Arabidopsis protein (RAP) (Aukerman & Sakai, ath-miR169e 33 126 4.01 0.004 2003), respectively. Classification into functional categories ath-miR169m 6 23 4.03 0.024 ath-miR5644 26 102 4.09 0.029 with Mapman (Thimm et al., 2004) determined that most tar- ath-miR399b 11 46 4.12 0.003 gets belong to the category of ‘Regulation of Transcription’, ath-miR172b-3p 108 482 4.53 0.029 particularly to the transcription factors subcategory, with 35 ath-miR172a 108 483 4.54 0.028 transcription factors among the putative targets of downregu- ath-miR319b 6 27 4.88 0.003 lated miRNAs and 8 among the targets of the upregulated ath-miR780.2 4215.45 0.003 ath-miR5648-3p 8 70 8.58 0.034 miRNAs in galls (Table S3). It is also important to point out ath-miR2111a-5p 2 21 10.82 0.001 the identification of 26 transposable elements (TEs) as putative targets of the DE miRNAs (Table S3).

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 7

(a)

Fig. 3 Quantitative-PCR (q-PCR) analysis of (b) differentially expressed miRNAs and their predicted targets in galls induced by Meloidogyne javanica in Arabidopsis thaliana. (a) miRNA expression values in galls relative to uninfected controls by RNAseq (gray) and q-PCR (black). All miRNAs tested showed significant expression differences between galls and controls (P < 0.05) by RNA-seq and q-PCR, except miR839 by q-PCR with P < 0.16. (b) Expression of six putative miRNA targets measured by q-PCR. All were repressed in galls vs control roots (P < 0.05) agreeing with microarray data (Jammes et al., 2005; Barcala et al., 2010). FC, fold change.

Fig. 5a). Moreover, galls formed in the mir390a-2 line were A role for miR390 during root-knot nematode infection smaller than those formed in Col-0 plants (P < 0.05; Fig. 5b). miR390a-b were highly expressed in galls, showing the highest This suggests that mir390a expression in galls is important for number of readings among those up-regulated miRNAs successful nematode infection. described (c. 53 000 each; Table 1). As MIR390a (At2g38325) is In roots, miR390 controls the biogenesis of TAS3 derived also predominantly expressed in roots (Marin et al., 2010), we trans-acting short-interfering RNAs (tasiRNAs), and this func- analyzed its promoter activation pattern during nematode infec- tion is crucial for lateral root growth (Marin et al., 2010). The tion. A GUS reporter line carrying a 2.6-kb promoter region pTAS3a::GUS line showed activation along the vascular tissue of from MIR390a (Fig. 4a–c; Marin et al., 2010) was specifically uninfected roots and in M. javanica-induced galls (Fig. S1c,d; activated within 4-dpi galls (Fig. 4a) induced by M. javanica in Fig. 4d–f). The activation observed at 4 dpi was maintained in 7- Arabidopsis roots, whereas in uninfected roots it was only and 15-dpi galls (Fig. 4d–f). In semi-thin sections of 4 dpi galls, detected in lateral root primordia and in the elongation zone the GUS signal was intense in GCs as well as in their neighboring (Fig. S1a,b). Hence, pMIR390a::GUS activation pattern was con- cells (Fig. 4h). Thus, the expression patterns of pTAS3a::GUS sistent with our sequencing and q-PCR results (Table 1; Fig. 3a). and pMIR390a::GUS overlap in galls, similarly to those described GUS activity in galls was maximal at 7 dpi (Fig. 4b) and for lateral roots (Marin et al., 2010). TAS3a role in galls was ana- decreased at medium-late infection stages (15 dpi; Fig. 4c). GUS lyzed by performing infection tests in the mutant line TAS3a-1 activity was localized in the GCs and surrounding cells within the (GABI 621G08), with only 40% of wild-type TAS3a transcript vascular cylinder in semi-thin 4-dpi gall sections (Fig. 4g). The levels; Marin et al., 2010). This mutant line presented a 40% expression pattern of pMIR390a::GUS was similar in 5- and 7- reduction in the percentage of galls per main root with respect to dpi galls induced by the related species M. incognita (Fig. S1e, f). the control (Fig. 5c). The galls formed in TAS3a-1 were not sig- We then investigated its functional role during gall develop- nificantly different in size to those of the control (Fig. 5d). It has ment. The homozygous line mir390a-2, carrying a T-DNA inser- been demonstrated that miR390a, TAS3-derived tasiRNAs and tion located 30 bp upstream of the MIR390a (At2g38325) the auxin responsive factors ARF2, ARF3 and ARF4, constitute transcription start site, showed a strong reduction in the accumu- an auxin-responsive regulatory module controlling lateral root lation of the mature miR390a in roots, compared to the wild- growth, in such a way that active TAS3-derived tasiRNAs regulate type line (Marin et al., 2010) and is therefore considered a loss- these ARFs directly in lateral roots by degrading their transcripts of-function line. When tested for nematode infection, this line (Marin et al., 2010). Accordingly, a GUS-based sensor line that is presented a significant decrease (P < 0.05) in infection levels not induced when the TAS3-derived tasiRNAs are actively bind- (23% less galls per main root than the wild-type Col-0 line; ing the ARF3 sequence, pARF3:ARF3-GUS, was not induced in

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 8 Research Phytologist

(a) (b) (c)

(d) (e) (f)

(g) (h)

Fig. 4 Overlapping expression patterns of pMIR390a::GUS and pTAS3::GUS in galls. pMIR390a::GUS is active in galls induced by Meloidogyne javanica in Arabidopsis at 4 d post infection (dpi) (a), reaching its maximum expression at 7 dpi (b). Expression disappears at 15 dpi (c). pTAS3::GUS is activated in the vascular cylinder and inside galls at 4 dpi (d). Its expression is maintained at 7 (e) and 15 (f) dpi. GUS signal was localized in giant cells (GCs) and surrounding cells in the vascular cylinder in semi-thin sections of 4-dpi galls from pMIR390a::GUS (g) and pTAS3::GUS (h). Asterisks indicate GCs. Black arrowhead indicates nematode. Bars: (a–f) 500 lm; (g, h) 100 lm.

galls (Fig. 6a). By contrast, a line harboring an ARF3 sequence analyzed using Mireap software (http://sourceforge.net/projects/ resistant to the cleavage by TAS3-derived tasiRNAs (pARF3: mireap/) to predict their secondary structures, DICER cleavage ARF3m-GUS; induced in the tissues where the promoter is sites and MFE (minimum free energy; 40/74 kcal mol 1; active) was activated in galls at 3 dpi (Fig. 6b). GUS staining in Bonnet et al., 2004). Six hundred and two sequences ranging this line showed the same pattern in GCs and surrounding cells from 19 to 25 nt and present in at least one of the six libraries as the MIR390a promoter and the TAS3a promoter lines presented structural hairpin characteristics. However, only 11 of (Figs 4g,h, 7c). These results confirmed that TAS3-derived these sequences ranging from 21 to 23 nt were DE (P < 0.05) in tasiRNAs were active in galls and GCs, as they could degrade their galls as compared to control roots, three of them being upregu- described mRNA target, ARF3. lated and eight downregulated (Table 2). From them, one upreg- ulated and two downregulated also presented MFE values similar to those described for Arabidopsis miRNAs (Table 2). Interest- Identification of putative nematode-responsive novel ingly, one was exclusively read in control root libraries, another miRNAs one only in galls, and another sequence was abundant in both, Those sRNA sequences that mapped in the Arabidopsis genome although upregulated in galls (Table 2). The potential targets for but were not classified as any previously described miRNA were these three putative sRNAs were predicted, finding among them

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 9

(a) (b)

Fig. 5 The role of miRNA390 and TAS3 during the infection of Arabidopsis thaliana by Meloidogyne javanica. Mutant lines infected with Meloidogyne javanica showed a significant decrease in the infection rate. A 23% decrease in the infection level was (c) (d) observed for the mutant mir390a-2 line (a), as well as a reduction in the size of the formed galls (b). TAS3a-1 mutant line presented a significant reduction of 40% in the percentage of galls per main root with respect to the Col-0 control (c), but galls were not smaller than those of the control (d). Statistical analysis was performed with three independent experiments per line using ANOVA; significant differences with Col-0 are indicated by asterisks, P < 0.05; values are means SE.

(a) (b) (c)

Fig. 6 TAS3-derived tasiRNAs are active in galls and giant cells (GCs) as they could degrade ARF3. At 3 d post infection (dpi), pARF3:ARF3-GUS is not induced in galls induced by Meloidogyne javanica in Arabidopsis thaliana (a), whereas pARF3: ARF3m-GUS, a mutated version resistant to TAS3, is activated in galls (b), and giant cells (c). Asterisks indicate giant cells. N, nematode. Bars: (a, b) 500 lm; (c) 100 lm. genes related to stress, hormone metabolism, development or cell A double validation of some of these putative targets by q- cycle (Table S4). PCR showed the same tendencies as in microarray analysis (Fig. 3b; Barcala et al., 2010; Jammes et al., 2005). The selected miRNAs and the mRNA abundance of their predicted targets Comparative expression analysis of miRNAs and their showed opposite behavior as expected. targets The expression of the predicted targets for the upregulated Overview and differential expression of 24-nt sRNAs and miRNAs in 3-dpi galls (Tables 3, S3) was checked among down- rasiRNAs regulated genes identified in previous studies, and vice versa. The transcriptomes used were from 3-dpi microdissected GCs and The length distribution overview of the sequences found in the galls (Barcala et al., 2010), and 7 dpi-galls (Jammes et al., 2005; six libraries showed a marked increase in the number of 24-nt Cabrera et al., 2014b). In these comparisons, six downregulated reads in galls compared to control roots (Fig. 2a). To investigate genes (P < 0.05) were predicted as putative targets of six upregu- the importance of this group of 24-nt sRNAs in galls, a DE anal- lated miRNAs described in this work (Table 3). We found 15 ysis was performed after normalization of reads as RPM. From upregulated transcripts that were putative targets of downregu- the 2909 193 unique 24-nt sequences found in at least one of the lated known and novel miRNAs in our analysis (Table 3). six libraries, 1.94% (corresponding to 56 455 sequences) were Among them is MYB33 (targeted by miR159b and miR319b), DE statistically (Fig. S2). From those, 35% (19 970) were the only MYB transcription factor upregulated in 3-dpi GCs expressed exclusively in galls, whereas 2557 (4.5%) were exclusive (Barcala et al., 2010). from control roots (Fig. S2). From the 33 928 sequences shared

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 10 Research Phytologist

Fig. 7 Distribution of rasiRNAs from Meloidogyne javanica galls and control roots in the Arabidopsis genome. Repeat-associated small interfering RNAs in gall and control libraries are shown across the five Arabidopsis chromosomes. x-Axis indicates chromosome position and y-axis the number of reads.

Table 2 Novel Arabidopsis miRNAs identified in galls induced by Meloidogyne javanica and control root libraries

Total number of Reads

Length Minimum Free Energy G1 G2 G3 C1 C2 C3 P value

GAAGGTAGAGTTGTAGAGAGGTT 23 61.3 16 11 10 0 0 0 0.020 AGAGAGAGAGAGAGAAGAGCAA 22 28.2 11 5 8 0 0 0 0.042 TTGAGGGGGGTGTATTAATAATA 23 24 9 14 15 0 0 0 0.022 GAAAGGAATTTGCGGTAGATATA 23 73.2 1515 1622 1813 1988 2356 2643 0.004 GGGGGCTTTTTGAGAATTGGCAC 23 31.6 28 0 40 72 62 83 0.035 TGATGAACTCGCAATTAGACGTA 23 36.9 0 0 0 10 13 12 0.006 TGGATAGAGATGAGTGATGGATA 23 22.3 0 0 0 970 1409 1331 0.009 GGAATGATGAGGAAAAATGTA 21 18 0 0 0 15 13 15 0.005 GTGGAAGAAAGAGAAGATGATA 22 47.2 0 0 0 17 19 24 0.001 GGTCATGATCGCGGTCACGGT 21 69.1 0 0 0 17 13 19 0.009 ATGGAAGAGTGATATGGATAA 21 44.4 0 0 0 40 70 64 0.020

The length, minimum free energy, total number of reads in the six libraries and statistical significance (P-value) for the differentially expressed novel miRNAs identified in galls and control roots are indicated. In bold those sequences with a minimum free energy suitable for miRNAs.

by galls and controls, 67.7% (22 974) were upregulated in galls, remarkable differences in their expression levels in galls as com- whereas only 10 954 were repressed (Fig. S2). Therefore, from pared to the control roots. Comparison of these results to those the 56 455 sequences of 24-nt DE, 42 944 (76%) were either relative to the sRNAs of 20–21 nt described above, mostly down- exclusive or upregulated in galls (Fig. S2). Strikingly, up to 1537 regulated in galls, indicate a contrasted regulation between 20– sequences of 24 nt presented a FC higher than + 20, showing 21-nt sRNAs and 24-nt sRNAs in galls.

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 11

Table 3 Predicted targets of differentially expressed (DE) miRNAs miRNA Target ID Description ath-miR156h AT3G05165 sugar transporter, putative ath-miR156i AT2G17220 Protein kinase superfamily protein ath-miR5643a AT1G45130 BGAL5 | beta-galactosidase 5 ath-miR5655 AT4G28270 ATRMA2, RMA2 | RING membrane-anchor 2 ath-miR5657 AT1G20950 Phosphofructokinase family protein NewSeq1 AT5G51690 ACS12 | 1-amino-cyclopropane-1-carboxylate synthase 12 ath-miR156a,b,c,d,e,f,g,j AT5G50670 squamosa promoter-binding protein, putative ath-miR156j AT5G08620 STRS2, ATRH25 STRS2 (STRESS RESPONSE SUPPRESSOR 2) ath-miR159b/ath-miR319b AT5G06100 MYB33, ATMYB33 MYB33 (MYB DOMAIN PROTEIN 33); DNA binding/transcription factor ath-miR163 AT3G44870 S-adenosyl-L-methionine:carboxyl methyltransferase family protein ath-miR163 AT5G37990 S-adenosyl-L-methionine-dependent methyltransferases superfamily protein ath-miR165a,b/ath-miR166a,b,c,d,e,f,g AT2G34710 PHB, ATHB14, ATHB-14, PHB-1D PHB (PHABULOSA); DNA binding/transcription factor ath-miR165a,b/ath-miR166a,b,c,d,e,f,g AT1G52150 ATHB-15, ATHB15, CNA, ICU4 ATHB-15; DNA binding/transcription factor ath-miR167d AT3G61310 DNA-binding family protein ath-miR169d,e,f,g,h,i,j,k,l,m,n AT5G06510 NF-YA10 NF-YA10 (NUCLEAR FACTOR Y, SUBUNIT A10); transcription factor ath-miR169d,e,f,g,h,i,j,k,l,m,n AT3G05690 NF-YA2 (NUCLEAR FACTOR Y, SUBUNIT A2) ath-miR169 h,i,j,k,l,m,n AT1G72830 HAP2C, ATHAP2C, NF-YA3 NF-YA3 (NUCLEAR FACTOR Y, SUBUNIT A3); transcription factor ath-miR172a,b-3p AT5G12900 unknown protein ath-miR172a,b-3p AT4G29430 rps15ae ribosomal protein S15A E ath-miR2111a-5p,b-5p AT2G19590 ACO1, ATACO1 ACC oxidase 1 ath-miR399b,e AT2G33770 UBC24, ATUBC24, PHO2 PHO2 (PHOSPHATE 2); ubiquitin-protein ligase

Predicted Arabidopsis genes differentially expressed (DE) in 3-d-post-inoculation (dpi) giant cells and 3–7-dpi galls induced by Meloidogyne javanica and M. incognita in Arabidopsis thaliana (Jammes et al., 2005; Barcala et al., 2010) as putative targets of the differentially expressed miRNAs described within the manuscript. Red, induced, and green, repressed in galls.

Those sRNAs associated with DNA, rasiRNAs (most of them (Allen et al., 2005) to higher plants (Axtell et al., 2006). Our 24-nt long; Axtell, 2013), from control and gall samples were study was performed with three independent biological replicates mapped in the five Arabidopsis chromosomes. Their distribution for galls and for control roots. Accordingly, we had a high valida- was concentrated in the same areas (mostly centromeric) for con- tion rate for the abundance tendencies of miRNAs and their trols and galls, although the total rasiRNAs number was much putative gene targets by independent techniques based on q-PCR higher in galls than in controls (Fig. 7), similarly to that occur- and microarrays (Fig. 3). Galls are pseudo-organs formed by the ring for the gall-exclusive and control root-exclusive rasiRNAs expansion of giant cells (GCs), the proliferation of the neighbor- (Fig. S3). From the 27 groups or families of repetitive sequences ing vascular tissues and the hypertrophy of cortical cells in identified in the six libraries, 16 were induced in galls with response to the nematode infection (Escobar et al., 2015). There- respect to the controls and only two were less abundant (Table 4; fore, sRNAs detected in this study by high-throughput sequenc- P < 0.05). All of these results suggest that the abundance or 24-nt ing could be originated from GCs and from any of these cell sRNAs (most probably also rasiRNAs), constitutes a gall hall- types. mark as well. We obtained sequences of 21 and 24-nt as the most abundant in the six libraries (Fig. 2). This is consistent with the scarce exist- ing data on massive sRNAs sequencing in cyst-nematode feeding Discussion sites (Hewezi et al., 2008, 2012; Li et al., 2012). We found differ- Under the generic name of small RNA (sRNA), a plethora of ences in sRNA length distribution between galls and control diverse RNA molecules of 20–30 nt that have emerged as major roots: 21-nt sRNAs (the length of most plant miRNAs; Axtell, regulators in plants still have little-known roles in plant–nema- 2013) were more abundant in uninfected root libraries than in tode interactions. The only data concerning root-knot nematode gall libraries, whereas 24-nt sRNAs, (the typical length of (RKN)–plant interactions were the role of miRNAs in systemic rasiRNAs; Axtell, 2013) were amply more abundant in galls. changes caused by the infection (Zhao et al., 2015). Here we Only 11 miRNAs were upregulated in galls, whereas 51 miRNAs describe for the first time the sRNA population differential were downregulated. These data strongly suggest a general down- expression (DE) in RKN-induced galls compared to uninfected regulation of miRNAs in early-developing galls, as reported for root segments. Our results indicate that the abundance of 24-nt 4-d-post-inoculation (dpi)-syncytia (Hewezi et al., 2008). Most sRNAs (most probably rasiRNAs) constitutes a gall hallmark, at downregulated miRNAs in galls are induced or have a confirmed least at early infection stages (Fig. 2). We also describe the first role in other stress conditions or nutritional statuses (Guleria functional role during a plant–nematode interaction for a et al., 2011), with the exception of two miRNAs from the same tasiRNA, TAS3a, regulated by miR390a in galls. This regulatory family (miR156i, h), probably related to hypoxia (Guleria et al., module is highly conserved from the moss Physcomitrella patens 2011), which were induced in galls (Table 1). Although hypoxia

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 12 Research Phytologist

Table 4 Repeat associated small interfering RNA (rasiRNAs) in Meloidogyne javanica galls and control root libraries

Unique Reads

Type G1 G2 G3 R1 R2 R3 FC P value

DNA/En-Spm Antisense 12 324 10 919 11 389 5193 5921 6432 1.9 0.001 Sense 16 168 14 660 15 019 5722 6334 7043 2.3 0.001 DNA/Harbinger Antisense 3068 2672 2842 1893 1964 2194 1.4 0.007 Sense 3942 3446 3639 2089 2194 2404 1.6 0.003 DNA/hAT-Ac Antisense 7114 6390 7022 4393 4373 4826 1.5 0.001 Sense 6072 5246 5578 3138 3284 3491 1.6 0.002 DNA/MuDR Antisense 62 474 57 081 58 722 39 919 42 600 46 227 1.3 0.001 Sense 65 456 60 306 62 294 40 907 43 668 47 507 1.4 0.000 DNA/TcMar-Mariner Antisense 278 227 261 278 301 317 1.2 0.038 LINE/L1 Antisense 25 452 23 074 24 568 15 505 16 393 17 736 1.4 0.001 Sense 40 596 36 065 39 289 22 187 23 629 25 230 1.6 0.001 LINE? Sense 1167 1027 1105 780 854 911 1.3 0.008 LTR/Copia Antisense 28 689 25 912 27 391 14 858 15 972 17 389 1.6 0.001 Sense 38 615 35 199 36 448 17 196 18 791 20 583 1.9 0.000 LTR/Gypsy Antisense 54 906 49 407 51 941 24 441 26 914 29 642 1.9 0.001 Sense 69 655 63 209 65 910 32 429 36 191 39 162 1.8 0.000 Other/Composite Antisense 712 699 676 554 545 676 1.1 0.011 RC/Helitron Antisense 22 238 20 305 20 770 16 397 16 985 18 957 1.2 0.007 rRNA Antisense 6329 5392 6131 4247 4018 4593 1.3 0.011 Satellite/centr Antisense 1459 1396 1328 478 554 645 2.4 0.000 Sense 1066 1037 967 439 536 576 1.9 0.000 Satellite Antisense 6815 6131 5993 2650 2811 3123 2.1 0.003 Sense 4410 4211 4066 2096 2112 2375 1.9 0.000 SINE/tRNA Antisense 3010 2581 3136 3269 3416 3697 1.2 0.032 SINE? Sense 210 202 237 139 132 142 1.5 0.007

Number of unique reads in the six libraries are listed. Fold change (FC) and statistical significance (P-value) for the differentially expressed groups of repeat associated small interfering RNA (rasiRNAs).

inside galls has not been confirmed, indirect evidences come from Other gall-repressed miRNAs are associated with developmental the activation in tobacco GCs induced by Meloidogyne javanica processes, for example, MIR166/165 genes, which regulates shoot of a hemoglobin promoter that responds to low oxygen tension apical meristem and floral development in parallel to the (Ehsanpour & Jones, 1996). Additionally, other hypoxia-related WUSCHEL-CLAVATA pathway (Jung & Park, 2007), or genes were also induced in 3-dpi galls (Barcala et al., 2010; miR156 (a, b, c, d, e, f, j) that regulates the juvenile-to-adult Cabello et al., 2014), for example, a gene encoding a class I non- transition (Wu et al., 2009). In this respect, cyst nematodes symbiotic leghemoglobin with high oxygen affinity (At2g16060; secrete peptides similar to CLAVATA-LIKE ELEMENTS (CLE- Hunt et al., 2002), an alcohol dehydrogenase-coding gene like) plant ligands, which have a role during syncytia formation (At1g77120), and SUS1/SUS4, upregulated in roots under (Replogle et al., 2011, 2013). Other CLE-like peptides, such as hypoxia (Bieniawska et al., 2007), 16D10 from RKNs, interact with transcription factors such as Examples of stress-induced miRNAs that were repressed in SCARECROW, a key regulator of radial patterning in the Ara- galls include members of the miR169 family induced under bidopsis root (Levesque et al., 2006). Thus, the repression of drought (Li et al., 2008; Zhao et al., 2011), low temperature these miRNAs in galls might be related to developmental pro- (Zhou et al., 2007; Lee et al., 2010), high soil salinity (Zhao cesses contributing to maintain a balance between cell prolifera- et al., 2009), N deficiency (Zhao et al., 2010) and UV-B radia- tion and differentiation in the complex gall structure (Table 1). tion (Zhou et al., 2007). Further examples are miR391 and Although the reasons and functional consequences of massive miR399, which are prominently induced upon phosphate starva- miRNA repression in galls is currently far from being under- tion and involved in inorganic phosphate homeostasis mainte- stood, downregulation of several miRNAs has been reported in nance (Lundmark et al., 2010; Guleria et al., 2011), and several plant species during interactions with different pathogens miR780, induced under N starvation (Liang et al., 2012). Their such as fungi, viruses or bacteria (Balmer & Mauch-Mani, repression might indicate that N and P status are maintained in 2013). In Arabidopsis infected with Heterodera schachtii, miR396 galls, probably being a basic requirement to sustain nematode downregulation and subsequent GRF1/GRF3 induction are nec- feeding. This is also in accordance with the global downregula- essary for correct syncytia initiation (Hewezi et al., 2012). In tion of stress-related genes in early-developing GCs and galls in mammals, general miRNA-mediated gene repression contributes Arabidopsis and tomato (Jammes et al., 2005; Schaff et al., 2007; to cancer by promoting transposable element (TE) expression Caillaud et al., 2008; Barcala et al., 2010; Portillo et al., 2013). that benefits the evolving tumor by leading to genomic instability

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 13

(Shalgi et al., 2010). Accordingly, we have identified several accordance, MET1 is upregulated in Arabidopsis GCs (Barcala miRNAs as being downregulated in early-developing galls that et al., 2010) and a chromomethylase involved in DNA methyla- may target TEs in the Arabidopsis genome (Tables S2, S3). tion (Lindroth et al., 2001) is induced in tomato GCs (Portillo The most abundant miRNA induced in 3-dpi galls is miR390 et al., 2013). Moreover, previous microarray results showed a (Table 1). miR390a and TAS3-derived tasiRNAs define a path- drastic repression of gene expression in GCs as compared to con- way that regulates leaf patterning establishing leaf polarity and trol roots. This tendency was conserved in early-developing GCs controls lateral root growth by repressing the ARF family mem- (3 dpi) and galls of Arabidopsis and tomato (Jammes et al., 2005; bers ARF2, ARF3 and ARF4 (Hunter et al., 2006; Nogueira et al., Barcala et al., 2010; Portillo et al., 2013). Thus, in this context, it 2007; Marin et al., 2010). Previously, a molecular link between is possible that epigenetic processes regulated by 24-nt sRNAs lateral root formation and galls was demonstrated by the crucial mediate gene repression during early stages of GC differentiation role during gall formation of LBD16, an essential gene for lateral from their vascular precursors. A similar abundance of 24-nt root development which is regulated by auxins (Cabrera et al., sRNAs was found in a Verticillium-sensitive cultivar of Gossypium 2014a), similarly to miR390a. Here we describe another regula- hirsutum, infected with Verticillium dahlia, a soil-borne fungal tory network shared by galls and lateral roots, based on the regu- pathogen (Yang et al., 2013). Suppression of the RdDM silencing latory action of miR390 on TAS3-derived tasiRNAs (Figs 4–6). pathway increased Arabidopsis resistance to Pseudomonas syringae The expression patterns of miR390a and TAS3a promoters over- (Dowen et al., 2012), suggesting that this pathway is important lapped in galls, particularly in the vascular tissue and GCs for bacterial infection. (Fig. 4). Loss-of-function lines showed increased resistance to In soybean, DNA hypermethylation has been observed in M. javanica and their galls were smaller than in control plants cyst nematode-resistant lines with multiple Rhg1 copies (Cook (Fig. 5). Remarkably, TAS3-derived tasiRNAs were active in et al., 2014). Strikingly, tasiRNA transcripts can also generate galls, as proven by the fact that an ARF3::GUS-based sensor line 24-nt sRNAs in plants (Allen et al., 2005; Khraiwesh et al., showed no signal in wild-type galls, whereas a line carrying an 2010). These induce cytosine methylation at tasiRNA- ARF3 mutation resistant to TAS3-derived tasiRNAs showed generating loci (Wu et al., 2012), similarly to 21-nt secondary GUS activity restricted to the gall centre, including GCs (Fig. 6). siRNAs generated by DCL4 that participate in post- These results concur with our former data from microarrays, transcriptional silencing of exogenous targets through AGO1 where ARF4, another demonstrated target of TAS3-derived (Cuperus et al., 2010). In this respect, a recent differential tasiRNAs in lateral roots (Marin et al., 2010), was repressed in 3- methylation study (methyloma) of Arabidopsis DNA extracted dpi GCs (Barcala et al., 2010). Thus, we identify for the first time from soybean roots infected with H. schachtii (Rambani et al., in the Arabidopsis–RKN interaction a complex regulatory mod- 2015) showed DNA methylation changes during the compati- ule that involves two sRNA types: an miRNA highly abundant in ble interaction that impact a large number of protein coding galls, miR390, and the TAS3-derived tasiRNAs, actively func- genes locally in the syncytium and systemically. A significant tioning in both galls and GCs. Although, Arabidopsis is not a portion of the differentially methylated genes was among genes natural host for RKNs, the transcriptomes of early GCs induced with differential expression in the soybean syncytium. These by M. javanica in Arabidopsis and tomato roots showed a high results point to a novel role of syncytia-induced DNA methy- similarity (Portillo et al., 2013). Furthermore, genes such as the lation in regulating gene expression changes during parasitism peroxidase TPX1, downregulated in Arabidopsis GCs at 3 dpi as (Rambani et al., 2015). well as its corresponding ortholog in tomato (Barcala et al., 2010; In conclusion, a plethora of sRNAs is present in early-developing Portillo et al., 2013) with a demonstrably crucial role for GCs/ galls. Among them, some miRNAs will contribute to the gene gall formation in tomato, provide another specific example. silencing observed in GCs via target mRNA degradation, These findings validate holistic approaches in the Arabidopsis translation inhibition, or by regulation of ta-siRNAs, such as genetic model for simplicity, and launch the potential to transfer miR390-TAS3. The scenario gets more complex, as one of the gall this knowledge to crop plants. molecular signatures is the high abundance of 24-nt sRNAs known Galls showed a higher abundance in 24-nt sRNA sequences to mediate epigenetic regulation of gene expression through chro- than control roots, most of them being gall-exclusive or upregu- matin remodeling. Future research will elucidate these sRNA- lated (Fig. S2). Most of these sequences match repetitive mediated mechanisms that probably orchestrate the developmental sequences in the genome, like TEs (Table 4; Fig. 7). More than processes participating in gall and GC differentiation. 16 out of 26 transposon families predicted to produce the rasiRNAs identified in uninfected roots and galls were over- Acknowledgements represented in galls (Table 4). Among them, hAT and PIF/ Harbinger families are highly enriched in euchromatic regions in This work was supported by the Spanish Government maize and have been related to heterosis (He et al., 2013). (AGL2013-48787 to C.E. and PCIN-2013-053 to C.F.) and by rasiRNAs play a pivotal role in gene silencing through the RNA- the Castilla-la Mancha Government (PEII-2014-020-P to C.F.). dependent DNA methylation (RdDM) pathway (Weiberg et al., J.C. was supported by a fellowship from the Ministry of Educa- 2014). In RdDM, the RNA polymerase IV/RDR2/DCL3/ tion, Spain. C.M. was supported by a PhD grant from INRA- AGO4/6/9 module induces DNA methylation through 24 nt SPE Dpt and the Conseil Regional Provence-Alpes-Cote^ d’Azur sRNAs, which is maintained by MET1 and CMT3. In (PACA). B.F. was supported by the French Government

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 14 Research Phytologist

(National Research Agency, ANR) through the ‘Investments for Cuperus JT, Montgomery TA, Fahlgren N, Burke RT, Townsend T, Sullivan the Future’ LabEx SIGNALIFE (ANR-11-LABX-0028-01). CM, Carrington JC. 2010. Identification of MIR390a precursor processing- defective mutants in Arabidopsis by direct genome sequencing. Proceedings of the National Academy of Sciences, USA 107: 466–471. Dai X, Zhao PX. 2011. psRNATarget: a plant small RNA target analysis server. Author contributions Nucleic Acids Research 39: W155–W159. J.C., C.F. and C.E. planned and designed the research. J.C., Dowen RH, Pelizzola M, Schmitz RJ, Lister R, Dowen JM, Nery JR, Dixon JE, Ecker JR. 2012. Widespread dynamic DNA methylation in response to biotic M.B., A.G., A.R-M., C.M. and A.M. performed experiments stress. Proceedings of the National Academy of Sciences, USA 109: E2183–E2191. and analysed data. J.C., V.R-F, S.J-P, B.F, C.F and C.E. wrote Ehsanpour AA, Jones MG. 1996. Glucuronidase expression in transgenic tobacco the manuscript. All authors made changes to the initial roots with a parasponia promoter on infection with Meloidogyne javanica. – manuscript and revised the final version. Journal of Nematology 28: 407 413. Escobar C, Barcala M, Cabrera J, Fenoll C. 2015. Overview of root-knot nematodes and giant cells. In: Escobar C, Fenoll C, eds. Advances in botanical References research: plant nematode interactions, vol. 73. Oxford, UK: Elsevier, 1–32. Escobar C, Sigal B, Mitchum M. 2011. Transcriptomic and proteomic analysis Allen E, Xie Z, Gustafson AM, Carrington JC. 2005. microRNA-directed of the plant response to nematode infection. In: Jones J, Gheysen G, Fenoll C, phasing during trans-acting siRNA biogenesis in plants. Cell 121: 207–221. eds. Genomics and molecular genetics of plant–nematode interactions. Dordrecht, Aukerman MJ, Sakai H. 2003. Regulation of flowering time and floral organ the Netherlands: Springer, 157–176. identity by a microRNA and its APETALA2-like target genes. Plant Cell 15: Frenquelli M, Muzio M, Scielzo C, Fazi C, Scarfo` L, Rossi C, Ferrari G, Ghia P, 2730–2741. Caligaris-Cappio F. 2010. MicroRNA and proliferation control in chronic Axtell MJ. 2013. Classification and comparison of small RNAs from plants. lymphocytic leukemia: functional relationship between miR-221/222 cluster Annual Review of Plant Biology 64: 137–159. and p27. Blood 115: 3949–3959. Axtell MJ, Jan C, Rajagopalan R, Bartel DP. 2006. A two-hit trigger for siRNA Ghildiyal M, Zamore PD. 2009. Small silencing RNAs: an expanding universe. biogenesis in plants. Cell 127: 565–577. Nature Reviews Genetics 10:94–108. Balmer D, Mauch-Mani B. 2013. Small yet mighty – microRNAs in plant- Guleria P, Mahajan M, Bhardwaj J, Yadav SK. 2011. Plant small RNAs: microbe interactions. Microrna 2:72–79. biogenesis, mode of action and their roles in abiotic stresses. Genomics Barcala M, Garcia A, Cabrera J, Casson S, Lindsey K, Favery B, Garcia-Casado Proteomics Bioinformatics 9: 183–199. G, Solano R, Fenoll C, Escobar C. 2010. Early transcriptomic events in Harfouche L, Haichar Fel Z, Achouak W. 2015. Small regulatory RNAs and the microdissected Arabidopsis nematode-induced giant cells. Plant Journal 61: fine-tuning of plant–bacteria interactions. New Phytologist 206:98–106. 698–712. He G, Chen B, Wang X, Li X, Li J, He H, Yang M, Lu L, Qi Y, Wang X et al. Bar-Or C, Kapulnik Y, Koltai H. 2005. A broad characterization of the 2013. Conservation and divergence of transcriptomic and epigenomic variation transcriptional profile of the compatible tomato response to the plant parasitic in maize hybrids. Genome Biology 14: R57. root knot nematode Meloidogyne javanica. European Journal of Plant Pathology Hewezi T, Baum TJ. 2015. Gene silencing in nematode feeding sites. In: Escobar 111: 181–192. C, Fenoll C, eds. Advances in botanical research: plant nematode onteractions, vol. Bieniawska Z, Paul Barratt DH, Garlick AP, Thole V, Kruger NJ, Martin C, 73. Oxford, UK: Elsevier, 221–239. Zrenner R, Smith AM. 2007. Analysis of the sucrose synthase gene family in Hewezi T, Howe P, Maier TR, Baum TJ. 2008. Arabidopsis small RNAs and Arabidopsis. Plant Journal 49: 810–828. their targets during cyst nematode parasitism. Molecular Plant-Microbe Bohlmann H. 2015. Introductory chapter on the basic biology of cyst nematodes. Interactions 21: 1622–1634. In: Escobar C, Fenoll C, eds. Advances in botanical research: plant nematode Hewezi T, Maier TR, Nettleton D, Baum TJ. 2012. The Arabidopsis interactions, vol. 73. Oxford, UK: Elsevier, 33–59. microRNA396-GRF1/GRF3 regulatory module acts as a developmental Bologna NG, Voinnet O. 2014. The diversity, biogenesis, and activities of regulator in the reprogramming of root cells during cyst nematode infection. endogenous silencing small RNAs in Arabidopsis. Annual Review of Plant Plant Physiology 159: 321–335. Biology 65: 473–503. Hunt PW, Klok EJ, Trevaskis B, Watts RA, Ellis MH, Peacock WJ, Dennis ES. Bonnet E, Wuyts J, Rouze P, Van de Peer Y. 2004. Evidence that microRNA 2002. Increased level of hemoglobin 1 enhances survival of hypoxic stress and precursors, unlike other non-coding RNAs, have lower folding free energies promotes early growth in Arabidopsis thaliana. Proceedings of the National than random sequences. Bioinformatics 20: 2911–2917. Academy of Sciences, USA 99: 17 197–17 202. Cabello S, Lorenz C, Crespo S, Cabrera J, Ludwig R, Escobar C, Hofmann J. Hunter C, Willmann MR, Wu G, Yoshikawa M, de la Luz Gutierrez-Nava M, 2014. Altered sucrose synthase and invertase expression affects the local and Poethig SR. 2006. Trans-acting siRNA-mediated repression of ETTIN and systemic sugar metabolism of nematode-infected Arabidopsis thaliana plants. ARF4 regulates heteroblasty in Arabidopsis. Development 133: 2973–2981. Journal of Experimental Botany 65: 201–212. Jagga Z, Gupta D. 2014. Supervised learning classification models for prediction Cabrera J, Bustos R, Favery B, Fenoll C, Escobar C. 2014a. NEMATIC: a of plant virus encoded RNA silencing suppressors. PLoS ONE 9: e97446. simple and versatile tool for the in silico analysis of plant–nematode Jammes F, Lecomte P, de Almeida-Engler J, Bitton F, Martin-Magniette ML, interactions. Molecular Plant Pathology 15: 627–636. Renou JP, Abad P, Favery B. 2005. Genome-wide expression profiling of the Cabrera J, Diaz-Manzano FE, Sanchez M, Rosso MN, Melillo T, Goh T, Fukaki host response to root-knot nematode infection in Arabidopsis. Plant Journal H, Cabello S, Hofmann J, Fenoll C et al. 2014b. A role for LATERAL 44: 447–458. ORGAN BOUNDARIES-DOMAIN 16 during the interaction Arabidopsis- Jung J-H, Park C-M. 2007. MIR166/165 genes exhibit dynamic expression Meloidogyne spp. provides a molecular link between lateral root and root-knot patterns in regulating shoot apical meristem and floral development in nematode feeding site development. New Phytologist 203: 632–645. Arabidopsis. Planta 225: 1327–1338. Caillaud MC, Dubreuil G, Quentin M, Perfus-Barbeoch L, Lecomte P, de Kanno T, Yoshikawa M, Habu Y. 2013. Locus-specific requirements of DDR Almeida Engler J, Abad P, Rosso MN, Favery B. 2008. Root-knot nematodes complexes for gene-body methylation of TAS genes in Arabidopsis thaliana. manipulate plant cell functions during a compatible interaction. Journal of Plant Molecular Biology Reporter 31: 1048–1052. Plant Physiology 165: 104–113. Khraiwesh B, Arif MA, el Seum GI, Ossowski S, Weigel D, Reski R, Frank W. Cook DE, Bayless AM, Wang K, Guo X, Song Q, Jiang J, Bent AF. 2014. 2010. Transcriptional control of gene expression by microRNAs. Cell 140: Distinct copy number, coding sequence, and locus methylation patterns 111–122. underlie Rhg1-mediated soybean resistance to soybean cyst nematode. Plant Klink VP, Hosseini P, Matsye P, Alkharouf NW, Matthews BF. 2009. A gene Physiology 165: 630–647. expression analysis of syncytia laser microdissected from the roots of the Glycine

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 15

max (soybean) genotype PI 548402 (Peking) undergoing a resistant reaction Replogle A, Wang J, Paolillo V, Smeda J, Kinoshita A, Durbak A, Tax FE, after infection by Heterodera glycines (soybean cyst nematode). Plant Molecular Wang X, Sawa S, Mitchum MG. 2013. Synergistic interaction of Biology 71: 525–567. CLAVATA1, CLAVATA2, and RECEPTOR-LIKE PROTEIN KINASE 2 in Klink VP, Hosseini P, Matsye PD, Alkharouf NW, Matthews BF. 2010. cyst nematode parasitism of Arabidopsis. Molecular Plant–Microbe Interactions Syncytium gene expression in Glycine max([PI 88788]) roots undergoing a 26:87–96. resistant reaction to the parasitic nematode Heterodera glycines. Plant Schaff JE, Nielsen DM, Smith CP, Scholl EH, Bird DM. 2007. Comprehensive Physiology and Biochemistry 48: 176–193. transcriptome profiling in tomato reveals a role for glycosyltransferase in Mi- Lee H, Yoo SJ, Lee JH, Kim W, Yoo SK, Fitzgerald H, Carrington JC, Ahn JH. mediated nematode resistance. Plant Physiology 144: 1079–1092. 2010. Genetic framework for flowering-time regulation by ambient temperature- Shalgi R, Pilpel Y, Oren M. 2010. Repression of transposable-elements – a responsive miRNAs in Arabidopsis. Nucleic Acids Research 38: 3081–3093. microRNA anti-cancer defense mechanism? Trends in Genetics 26: 253– LevesqueMP,VernouxT,BuschW,CuiH,WangJY,BlilouI,HassanH, 259. Nakajima K, Matsumoto N, Lohmann JU et al. 2006. Whole-genome analysis of Smant G, Jones J. 2011. Suppression of plant defences by nematodes. In: Jones J, the SHORT-ROOT developmental pathway in Arabidopsis. PLoS Biology 4: e143. Gheysen G, Fenoll C, eds. Genomics and molecular genetics of plant–nematode Li W-X, Oono Y, Zhu J, He X-J, Wu J-M, Iida K, Lu X-Y, Cui X, Jin H, Zhu J- interactions. Dordrecht, the Netherlands: Springer, 273–286. K. 2008. The Arabidopsis NFYA5 transcription factor is regulated Sunkar R, Li Y-F, Jagadeeswaran G. 2012. Functions of microRNAs in plant transcriptionally and posttranscriptionally to promote drought resistance. Plant stress responses. Trends in Plant Science 17: 196–203. Cell Online 20: 2238–2251. Thimm O, Blasing O, Gibon Y, Nagel A, Meyer S, Kruger P, Selbig J, Muller Li X, Wang X, Zhang S, Liu D, Duan Y, Dong W. 2012. Identification of LA, Rhee SY, Stitt M. 2004. MAPMAN: a user-driven tool to display soybean microRNAs involved in soybean cyst nematode infection by deep genomics data sets onto diagrams of metabolic pathways and other biological sequencing. PLoS ONE 7: e39650. processes. Plant Journal 37: 914–939. Liang G, He H, Yu D. 2012. Identification of nitrogen starvation-responsive Weiberg A, Wang M, Bellinger M, Jin H. 2014. Small RNAs: a new paradigm microRNAs in Arabidopsis thaliana. PLoS ONE 7: e48951. in plant-microbe interactions. Annual Review of Phytopathology 52: 495–516. Lindroth AM, Cao X, Jackson JP, Zilberman D, McCallum CM, Henikoff S, Wu G, Park MY, Conway SR, Wang JW, Weigel D, Poethig RS. 2009. The Jacobsen SE. 2001. Requirement of CHROMOMETHYLASE3 for sequential action of miR156 and miR172 regulates developmental timing in maintenance of CpXpG methylation. Science 292: 2077–2080. Arabidopsis. Cell 138: 750–759. Lundmark M, Korner CJ, Nielsen TH. 2010. Global analysis of microRNA in Wu G, Poethig RS. 2006. Temporal regulation of shoot development in Arabidopsis Arabidopsis in response to phosphate starvation as studied by locked nucleic thaliana by Mir156 and its target SPL3. Development 133:3539–3547. acid-based microarrays. Physiologia Plantarum 140:57–68. Wu L, Mao L, Qi Y. 2012. Roles of dicer-like and argonaute proteins in TAS- Marin E, Jouannet V, Herz A, Lokerse AS, Weijers D, Vaucheret H, Nussaume derived small interfering RNA-triggered DNA methylation. Plant Physiology L, Crespi MD, Maizel A. 2010. miR390, Arabidopsis TAS3 tasiRNAs, and 160: 990–999. their AUXIN RESPONSE FACTOR targets define an autoregulatory network Xu M, Li Y, Zhang Q, Xu T, Qiu L, Fan Y, Wang L. 2014. Novel miRNA and quantitatively regulating lateral root growth. Plant Cell 22: 1104–1117. phasiRNA biogenesis networks in soybean roots from two sister lines that are Mitkowski NA, Abawi GS. 2003. Root-knot nematodes. Plant Health Instructor resistant and susceptible to SCN race 4. PLoS ONE 9: e110051. 2011: doi: 10.1094/PHI-I-2003-0917-0. Yang X, Wang L, Yuan D, Lindsey K, Zhang X. 2013. Small RNA Montgomery TA, Howell MD, Cuperus JT, Li D, Hansen JE, Alexander AL, and degradome sequencing reveal complex miRNA regulation during cotton Chapman EJ, Fahlgren N, Allen E, Carrington JC. 2008. Specificity of somatic embryogenesis. Journal of Experimental Botany 64: 1521–1536. ARGONAUTE7-miR390 interaction and dual functionality in TAS3 trans- Zhang Z, Yu J, Li D, Zhang Z, Liu F, Zhou X, Wang T, Ling Y, Su Z. acting siRNA formation. Cell 133: 128–141. 2010. PMRD: plant microRNA database. Nucleic Acids Research 38: Navarro L, Dunoyer P, Jay F, Arnold B, Dharmasiri N, Estelle M, Voinnet O, D806–D813. Jones JD. 2006. A plant miRNA contributes to antibacterial resistance by Zhao M, Ding H, Zhu J-K, Zhang F, Li W-X. 2011. Involvement of miR169 in repressing auxin signaling. Science 312: 436–439. the nitrogen-starvation responses in Arabidopsis. New Phytologist 190: 906– Nogueira FT, Madi S, Chitwood DH, Juarez MT, Timmermans MC. 2007. 915. Two small regulatory RNAs establish opposing fates of a developmental axis. Zhao B, Ge L, Liang R, Li W, Ruan K, Lin H, Jin Y. 2009. Members of miR- Genes & Development 21: 750–755. 169 family are induced by high salinity and transiently inhibit the NF-YA Nuthikattu S, McCue AD, Panda K, Fultz D, DeFraia C, Thomas EN, Slotkin transcription factor. BMC Molecular Biology 10:29–29. RK. 2013. The initiation of epigenetic silencing of active transposable elements Zhao W, Li Z, Fan J, Hu C, Yang R, Qi X, Chen H, Zhao F, Wang S. is triggered by RDR6 and 21-22 nucleotide small interfering RNAs. Plant 2015. Identification of jasmonic acid-associated microRNAs and Physiology 162: 116–131. characterization of the regulatory roles of the miR319/TCP4 module Portillo M, Cabrera J, Lindsey K, Topping J, Andres MF, Emiliozzi M, Oliveros under root-knot nematode stress in tomato. Journal of Experimental Botany JC, Garcia-Casado G, Solano R, Koltai H et al. 2013. Distinct and conserved 66: 4653–4667. transcriptomic changes during nematode-induced giant cell development in Zhao CZ, Xia H, Frazier TP, Yao YY, Bi YP, Li AQ, Li MJ, Li CS, Zhang BH, tomato compared with Arabidopsis: a functional role for gene repression. New Wang XJ. 2010. Deep sequencing identifies novel and conserved microRNAs Phytologist 197: 1276–1290. in peanuts (Arachis hypogaea L.). BMC Plant Biology 10:3. Portillo M, Lindsey K, Casson S, Garcia-Casado G, Solano R, Fenoll C, Escobar Zhou X, Wang G, Zhang W. 2007. UV-B responsive microRNA genes in C. 2009. Isolation of RNA from laser-capture-microdissected giant cells at Arabidopsis thaliana. Molecular Systems Biology 3: 103–103. early differentiation stages suitable for differential transcriptome analysis. Molecular Plant Pathology 10: 523–535. R Development Core Team 2008. R: A language and environment for Supporting Information statistical computing. Vienna Austria: R Foundation for Statistical Computing. URL http://www.R-project.org [accessed 01 September 2015]. Additional supporting information may be found in the online Rambani A, Rice JH, Liu J, Lane T, Ranjan P, Mazarei M, Pantalone V, Stewart version of this article. CN Jr, Staton M, Hewezi T. 2015. The methylome of soybean roots during the compatible interaction with the soybean cyst nematode. Plant Physiology 168: 1364–1377. Fig. S1 Expression pattern of pMIR390a::GUS and pTAS3a:: Replogle A, Wang J, Bleckmann A, Hussey RS, Baum TJ, Sawa S, Davis EL, GUS in noninfected plant roots and in roots infected with Wang X, Simon R, Mitchum MG. 2011. Nematode CLE signaling in M. incognita. Arabidopsis requires CLAVATA2 and CORYNE. Plant Journal 65: 430–440.

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 16 Research Phytologist

Fig. S2 Pie charts indicating the number of 24 nt sequences that Table S3 Putative target genes identified for all the differentially are exclusive, up-or downregulated in gall and control root expressed known miRNAs libraries. Table S4 Putative target genes identified for the differentially Fig. S3 Distribution of repeat associated small interfering RNA expressed novel miRNAs (rasiRNAs) exclusive of galls and roots among the five Arabidop- sis chromosomes. Please note: Wiley Blackwell are not responsible for the content or functionality of any supporting information supplied by the Table S1 Sequence of primers used for the analysis of miRNA authors. Any queries (other than missing material) should be target genes by real time q-PCR directed to the New Phytologist Central Office.

Table S2 Number of reads found in the six libraries generated of known miRNAs found in miRBase for Arabidopsis

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust bs_bs_banner

MOLECULAR PLANT PATHOLOGY DOI: 10.1111/mpp.12114

Technical advance NEMATIC: a simple and versatile tool for the in silico analysis of plant–nematode interactions

JAVIER CABRERA1, REGLA BUSTOS2, BRUNO FAVERY3, CARMEN FENOLL1 AND CAROLINA ESCOBAR1,* 1Facultad de Ciencias Ambientales y Bioquímica, Universidad de Castilla-La Mancha, Avenida de Carlos III s/n, 45071 Toledo, Spain 2Centro de Biotecnología y Genómica de Plantas U.P.M. I.N.I.A., Campus de Montegancedo, 28223, Pozuelo de Alarcón (Madrid), Spain 3Institut Sophia Agrobiotech, Université de Nice Sophia Antipolis, INRA UMR 1355 CNRS UMR 7254, 400 route des chappes, BP 167, 0690 Sophia Antipolis Cedex, France

Keywords: functional classification, galls, giant cells, in silico SUMMARY analysis, plant–nematode interactions, syncytia, transcriptome Novel approaches for the control of agriculturally damaging comparison. nematodes are sorely needed. Endoparasitic nematodes complete their life cycle within the root vascular cylinder, inducing special- ized feeding cells: giant cells for root-knot nematodes and syncytia for cyst nematodes. Both nematodes hijack parts of the INTRODUCTION transduction cascades involved in developmental processes, or partially mimic the plant responses to other interactions with Root-knot and cyst nematodes (RKNs and CNs) are obligate sed- microorganisms, but molecular evidence of their differences and entary plant endoparasites that produce serious economic losses commonalities is still under investigation. Transcriptomics has all over the world (Nicol et al., 2011). RKNs and CNs induce their been used to describe global expression profiles of their interac- specialized feeding cells, termed giant cells (GCs) and syncytia, tion with Arabidopsis, generating vast lists of differentially respectively, which they use for nourishment until reproduction expressed genes. Although these results are available in public within the root vascular cylinder (reviewed in Jones and Goto, databases and publications, the information is scattered and dif- 2011; Sobczak and Golinowski, 2011). Both nematode feeding cell ficult to handle. Here, we present a rapid, visual, user-friendly and (NFC) types are multinucleated, have dense cytosol and suffer easy to handle spreadsheet tool, called NEMATIC (NEMatode– endoreduplication (de Almeida Engler et al., 2012). GCs induced Arabidopsis Transcriptomic Interaction Compendium; http:// by RKNs are immersed in galls, ‘pseudo-organs’ caused by root www.uclm.es/grupo/gbbmp/english/nematic.asp). It combines swelling and tissue hyperplasia (Bird, 1996). GCs develop from existing transcriptomic data for the interaction between vascular cell precursors suffering repeated mitosis with aborted Arabidopsis and plant-endoparasitic nematodes with data from cytokinesis (Caillaud et al., 2008). In contrast, syncytia develop different transcriptomic analyses regarding hormone and cell cycle from vascular cells, mainly procambial cells, which fuse and are regulation, development, different plant tissues, cell types and incorporated into the syncytium (Hussey and Grundler, 1998). various biotic stresses. NEMATIC facilitates efficient in silico Although cytological and histological studies have described clear studies on plant–nematode biology, allowing rapid cross- structural and morphological differences between both NFCs comparisons with complex datasets and obtaining customized (Jones, 1981), molecular evidence of their differences and com- gene selections through sequential comparative and filtering monalities is still under investigation. Most of this evidence comes steps. It includes gene functional classification and links to utilities from the study of Arabidopsis promoter trap lines infected with from several databases.This data-mining spreadsheet will be valu- either Heterodera schachtii (CN) or Meloidogyne incognita (RKN), able for the understanding of the molecular bases subjacent to some showing common or contrasting activation patterns in both feeding site formation by comparison with other plant systems, NFCs (Barthels et al., 1997; Goddijn et al., 1993). Other compari- and for the selection of genes as potential tools for sons have also been established over the years based mainly on biotechnological control of nematodes, as demonstrated in the the differential expression patterns of individual genes (reviewed experimentally confirmed examples provided. in de Almeida Engler and Favery, 2011; Gheysen and Fenoll, 2002; Gheysen and Mitchum, 2009). Over the last decade, microarray technology has been useful to elucidate the global expression profiles of Arabidopsis–RKN and *Correspondence: Email: [email protected] Arabidopsis–CN interactions (Barcala et al., 2010; Escobar et al.,

© 2013 BSPP AND JOHN WILEY & SONS LTD 1 2 J. CABRERA et al.

2011; Fuller et al., 2007; Jammes et al., 2005; Puthoff et al., 2003; biotic stresses, as well as the functional classification of genes and Szakasits et al., 2009). Vast lists of Arabidopsis differentially links to different gene utilities from several databases, in a spread- expressed genes (DEGs) identified from comparisons between sheet tool called NEMATIC (NEMatode–Arabidopsis Transcripto- infected and uninfected roots are scattered across public data- mic Interaction Compendium; http://www.uclm.es/grupo/gbbmp/ bases and publications (reviewed in Escobar et al., 2011). Initially, english/nematic.asp). We provide three different examples for the transcriptomic analyses were performed for whole roots or for use of this tool for a better understanding of its utility. We expect nematode-induced organs, such as galls or roots containing that NEMATIC, which will be periodically updated, will be useful to syncytia (Fuller et al., 2007; Jammes et al., 2005; Puthoff et al., address questions and hypotheses regarding plant–nematode 2003). The combination of transcriptomic analysis with newly interactions in a simple and rapid manner, and therefore will developed cell biology techniques, such as laser capture facilitate knowledge advancement in this field. microdissection and cytoplasm microaspiration, represented a step forward in the understanding of the specifically differentiated RESULTS NFCs (Barcala et al., 2010; Szakasits et al., 2009). These approaches included not only Arabidopsis, but other plant species, NEMATIC, a simple and interactive tool for such as Glycine max (Alkharouf et al., 2006; Ithal et al., 2007; high-throughput biological data analysis centred on Khan et al., 2004; Klink et al., 2007, 2009), tomato (Portillo et al., plant–nematode interactions 2013) and rice (Kyndt et al., 2012). In parallel, microarray technol- ogy has been used in different fields of plant sciences to generate A spreadsheet tool (http://www.uclm.es/grupo/gbbmp/english/ wide-ranging transcriptomic information on the genes involved in nematic.asp) was generated using Excel 2007 software for rapid, key processes, such as hormone responses (Nemhauser et al., high-throughput analysis of biological data centred on plant– 2006), cell cycle regulation (Menges et al., 2003) and root devel- nematode interactions. It is composed of seven different sheets opment (Brady et al., 2007), generating complex datasets grouped into three different clusters according to their utilities. included in large databases. It is believed that endoparasitic nematodes capture part of the 1. ‘Plant–Nematode’ and ‘Custom Selection’ sheets: searching, transduction cascades involved in developmental processes, such selecting and filtering the genes of interest as lateral root formation, or partially mimic the plant responses In the ‘Plant–Nematode’ sheet, DEGs from Arabidopsis–nematode to other interactions with microorganisms, i.e. symbiosis or interactions and from transcriptomes considered to be of special Agrobacterium tumefaciens (Barcala et al., 2010; J. Cabrera et al., interest were collected and fused in a single table. A dropdown unpublished data; Damiani et al., 2012; Grunewald et al., 2009; menu was implemented to allow selection among the gene sets Mathesius, 2003). Commonalities and differences among compiled in the tool. (i) A total of 9688 Arabidopsis loci found to transcriptomes of the plant–nematode interactions included in be differentially expressed in any of the Arabidopsis–nematode different databases could be very useful for understanding the transcriptomes compiled (Table 1; Fig. 1). These are laser micro- molecular bases of plant–nematode interactions. Yet, many of the dissected GCs at 3 days post-inoculation (dpi) (Barcala et al., data are scattered, heterogeneous and difficult to handle in order 2010), hand-sectioned galls at 3, 7, 14 and 21 dpi (Barcala et al., to compare global transcriptional signatures of distinct plant– 2010; Jammes et al., 2005), microaspirated syncytia at 5 and nematode interactions. The creation of visual and user-friendly 15 dpi (treated together as originally described, because very few Excel-based tools that facilitate the handling of the extensive transcripts were different between the two time points; Szakasits amounts of genomic data based on the available transcriptomes et al., 2009) and roots containing syncytia at 3 dpi (Puthoff et al., would constitute a great advance to obtain conclusions from these 2003). The log2 values shown were calculated by comparing the comparative holistic approaches for different biological processes infected material with the corresponding uninfected control roots, (Ogata et al., 2009; Parizot et al., 2010). all described in the references (Table S1, see Supporting Informa- Thus, we developed a rapid, visual, user-friendly and easy to tion). Only experiments with more than two independent biologi- handle spreadsheet tool compiling existing transcriptomic data for cal replicates were considered, i.e. more than two independent the interaction between Arabidopsis and plant-endoparasitic chips hybridized. (ii) All genes classified by MapMan (Thimm et al., nematodes.These range from gall transcriptomes at different devel- 2004) into 34 functional categories. (iii) A total of 7002 up- and opmental stages (Barcala et al., 2010; Jammes et al., 2005) and down-regulated genes after exogenous treatment with abscisic root segments infected with CNs (Puthoff et al., 2003) to the acid (ABA), gibberellins (GL), auxin (indole-3-acetic acid, transcriptomes of isolated NFCs (GCs,Barcala et al., 2010;syncytia, IAA), ethylene (acetyl-CoA carboxylase, ACC), cytokinins (CK), Szakasits et al., 2009). We also integrated data from different brassinosteroids (BR) and jasmonate (JA) (Nemhauser et al., transcriptomic analyses regarding hormone and cell cycle regula- 2006). (iv) A compiled list of 2620 transcriptional regulators tion, development, different plant tissues and cells, and different classified into 30 families (Mitsuda and Ohme-Takagi, 2009)

MOLECULAR PLANT PATHOLOGY © 2013 BSPP AND JOHN WILEY & SONS LTD A plant–nematode interactions transcriptomic tool 3

Table 1 Common differentially expressed (DE) genes between pairs of plant–nematode transcriptomes compiled in the spreadsheet tool NEMATIC.

3-dpi 3-dpi 7-dpi 14-dpi 21-dpi 5 + 15-dpi Roots containing Number of common DE genes GCs galls galls galls galls syncytia 3-dpi syncytia

3-dpi GCs (Barcala et al., 2010) – 120 127 211 223 529 6 3-dpi galls (Barcala et al., 2010) – – 91 131 147 248 15 7-dpi galls (Jammes et al., 2005) – – – 642 741 738 17 14-dpi galls (Jammes et al., 2005) –––– 1398 921 20 21-dpi galls (Jammes et al., 2005) – – – – – 1102 25 5 + 15-dpi syncytia (Szakasits et al., 2009) –––– – – 57 Roots containing 3-dpi syncytia (Puthoff et al., 2003) –––– – – – dpi, days post-inoculation; GC, giant cell. originating from four representative databases of Arabidopsis: A filter button was added to the header of each transcriptome RARTF (rarge.gsc.riken.jp/rartf/) (Iida et al., 2005), AGRIS and information columns, facilitating the filtering of genes (arabidopsis.med.ohio-state.edu/AtTFDB/) (Davuluri et al., 2003), depending on the objectives of each user through the colour code DATF (datf.cbi.pku.edu.cn/) (Guo et al., 2005) and PlnTFDB or text filters. Filters were additive, i.e. it was possible to select (plntfdb.bio.uni-potsdam.de/v2.0/index.php?sp_id=ATH) (Riano- sets of genes fulfilling more than one condition among the Pachon et al., 2007). (v) A total of 1033 up-regulated genes during transcriptomes, e.g. up-regulated in a specific transcriptome, but G1, G2, M and S cell cycle phases (Menges et al., 2003; Parizot down-regulated by a certain hormone. et al., 2010). (vi) A total of 4790 genes enriched in 31 specific root In addition, a summary table showing the number of genes cell types: lateral root cap, columella, non-hair cells, hair cells, from the query found in each transcriptome was retrieved, as well cortex, endodermis, quiescent centre, stele, mature pericycle, xylem as the percentages of these genes relative to the total number of pole pericycle, phloem pole pericycle, lateral root primordia initials, genes in the plant–nematode set. maturing xylem cells, protoxylem and two-thirds of metaxylem, pericycle, protoxylem, metaxylem, phloem, developing proto- 2. ‘Anatomy’, ‘Cell Types’, ‘Development’ and ‘Biotic Stress’ phloem and metaphloem, and companion cells (Brady et al., sheets: selecting and filtering genes by Genevestigator 2007). expression values Another option, named ‘Custom Selection’ was included in the Data from experiments included in the Genevestigator database, dropdown menu, by which it was possible to select a customized performed on ATH1 Affymetrix 22K arrays (Hruz et al., 2008), for list of genes (Fig. 1) previously introduced into the ‘Custom Selec- each of the DEGs in any plant–nematode transcriptome (9688 tion’ sheet. genes), were collected. Thus, the corresponding Affymetrix probes The output generated after selection in the dropdown menu for each AGI code and the expression values of different experi- showed information on all the DEGs in any plant–nematode ments divided in terms of anatomy, biotic stress, cell types and transcriptome compiled. The output was organized into 25 development could be easily accessed for the previously filtered columns grouped into seven clusters: (i) seven columns with the genes on plant–nematode interactions by pasting them into the log2 values obtained from the reference publications for each gene first column of these sheets (Fig. 1; Hruz et al., 2008). selected (Barcala et al., 2010; Jammes et al., 2005; Puthoff et al., The ‘Anatomy’ sheet included experiments from 59 plant 2003; Szakasits et al., 2009); (ii) one column with the Arabidopsis tissues/organs of seedlings, inflorescences, leaves and roots. The Genome Initiative (AGI) code description from TAIR10; (iii) one ‘Cell Types’ sheet included experiments from 20 different cell column with the corresponding MapMan category (Thimm et al., types. The ‘Development’ sheet included experiments from ger- 2004); (iv) 14 columns with hormone regulation (Nemhauser minated seeds to mature siliques (nine stages). The ‘Biotic Stress’ et al., 2006); (v) four columns with the cell cycle phase in which a sheet included experiments involving fungi infection/exposure gene was expressed (Menges et al., 2003; Parizot et al., 2010); (vi) (Alternaria brassicicola, Botrytis cinerea, Blumeria graminis, one column with the transcription factor (TF) family to which the Erysiphe or Golovinomyces cichoracearum, E. orontii, Gigaspora gene belonged (Mitsuda and Ohme-Takagi, 2009); and (vii) enrich- rosea), insects (Bemisia tabaci, Myzus persicae), oomycetes ment in a certain root cell type (Brady et al., 2007; Fig. 1). Each (Phytophthora infestans, P. parasitica), bacteria (Pseudomonas row corresponded to a specific AGI code.The information was only syringae, Escherichia coli), viruses (Cabbage leaf curl virus, shown if the gene was differentially expressed in any plant– Turnip mosaic virus) and nematodes (M. incognita and nematode transcriptome compiled. The table cells were condition- H. schachtii) distributed in 118 columns. A custom selection of ally coloured to facilitate visualization according to the fold different categories from each group was possible. The corre- change, with red indicating induced, green repressed and grey not sponding Affymetrix probes for each AGI code in the input list differentially expressed or not represented. were also provided.

© 2013 BSPP AND JOHN WILEY & SONS LTD MOLECULAR PLANT PATHOLOGY 4 J. CABRERA et al.

Fig. 1 Flow chart representing an overview of the information captured and the putative different input data and the outputs obtained in the spreadsheet.

MOLECULAR PLANT PATHOLOGY © 2013 BSPP AND JOHN WILEY & SONS LTD A plant–nematode interactions transcriptomic tool 5

For the categories of Anatomy, Cell Types and Development, the dropdown menu and, in a few seconds, the expression values in table obtained displayed the average expression value of all 3-, 7-, 14- and 21-dpi galls of the 2620 TFs from the four different samples calculated from those annotated as belonging to this databases compiled (RARTF, AGRIS, DATF and PlnTFDB) were category in Genevestigator. For the Biotic Stress category, the gene obtained. expression responses were calculated as log2 ratios between the Many plant TFs are usually regulated by phytohormones. experimental and control samples.The resulting values thus reflect Among all the phytohormones described, auxin has been the most the up- or down-regulation of genes. studied in the interaction between Arabidopsis and plant- A colour code was facilitated according to the intensity of the endoparasitic nematodes, particularly centred on auxin transport expression of a certain gene in a category (see Experimental and accumulation in NFCs (Grunewald et al., 2009; reviewed by Procedures S1, Supporting Information, for a detailed description Goverse and Bird, 2011). However, very few examples of auxin- of the colour code).The filter buttons implemented in each column regulated TFs are known, and the knowledge of their putative allowed the filtering of genes by their expression values. functions in either galls or syncytia formation or maintenance is still scarce (J. Cabrera et al., unpublished data; Grunewald et al., 3. ‘Gene Info’ sheet: compiled information about filtered genes 2008). Therefore, in this context, as a second step, those TFs that A ‘Gene Info’ sheet was also provided to show information about were up-regulated in response to exogenous auxin in the study any group of selected DEGs in the plant–nematode transcriptomes of Nemhauser et al. (2006) were selected after filtering by ‘up- compiled and previously filtered, by pasting them into the first regulated by auxins’. column of the sheet. For each AGI code, the TAIR description, In the third step, among those TFs up-regulated by auxins MapMan category, subcellular localization, microRNA regulation, filtered previously, onlyTFs up-regulated in any of the developmen- hormone regulation, KEGG pathways, InterPro domains and Salk tal stages of galls recorded were subsequently selected. From the mutant lines (Fig. 1) were provided. A system of filters could also initial 2620 TFs, a list of 13 genes was generated (Table 2). Among be used in this sheet to select genes fulfilling certain restrictions. them are LBD16 and WRKY23 (Table 2), shown to be induced in Putative functions, interaction networks, cross-species homology, galls using reporter lines (J. Cabrera et al., unpublished data; co-expression analysis, mutant and transgenic lines available, etc., Grunewald et al., 2008, respectively). Moreover, functional studies could also be easily obtained by clicking in the link provided for showed that the suppression of LBD16 function resulted in infec- rapid access to 19 databases: TAIR, KEGG, GRAMENE, AGRIS, tion resistance to M. javanica in Arabidopsis, similar to the reduc- PLAZA, NCBI, EnsemblPlants, COP, PPDB, Genevestigator, tion in infection observed for H. schachtii in loss-of-function lines e-NORTHERN, ATTED-II, eFPBrowser, AtGenExpress, PUBMED, of WRKY23 when compared with the wild-type (J. Cabrera et al., GeneMania, AHD2.0, Salk and MapMan. unpublished data; Grunewald et al., 2008). These experimental Gene information concerning descriptions, InterPro domains, results validate the output obtained after application of the differ- KEGG pathways, subcellular localizations or microRNA regulation ent set-up filters from NEMATIC (http://www.uclm.es/grupo/ was downloaded from the TAIR FTP download section (http:// gbbmp/english/nematic.asp), as experimental data demonstrated www.arabidopsis.org/download/index.jsp). Salk mutant lines that both TFs are induced by auxins, and during plant–nematode were obtained from the Salk homozygote T-DNA collection (http:// interactions, reinforcing the utility of the selection. In addition, signal.salk.edu/cgi-bin/homozygotes.cgi) RAP2.6L, an ethylene responsive factor homologue of RAP2.6, with a functional role in the H. schachtii-Arabidopsis interaction (Ali et al., 2013), was detected by NEMATIC (http://www.uclm.es/ Example of utility 1: TFs induced by auxins in galls grupo/gbbmp/english/nematic.asp) as also induced in galls. formed by Meloidogyne spp. Other TFs up-regulated by auxins and in galls from the output In this example, one of the most interesting advantages offered by list, such as LBD33 and WRKY45, belong to the same families as this tool is shown.The user can score for genes fulfilling previously LBD16 and WRKY23, respectively (Table 2). In parallel with designed restrictions or limits through the subsequent use of LBD16 function, LBD33 has been shown to mediate lateral root filters in the data compiled in the spreadsheet tool. organogenesis in combination with LBD18 (Berckmans et al., Transcriptional regulators are expected to play a significant role 2011). Another TF in the list, LAX2, regulates auxin influx into the in the cell differentiation processes that occur during the estab- cell, together with LAX1 and LAX3 (Yang et al., 2006), and is lishment of RKNs and CNs (reviewed by Gheysen and Mitchum, expressed in lateral root primordia and emerging lateral root, but 2009), as they are key integrators modifying downstream gene not in syncytia at any time during infection (Lee et al., 2011). expression to generate diverse plant cell responses.The expression However, a related member, LAX3, interacts with 19C07, an effec- profiles of genes encoding Arabidopsis TFs in galls induced by tor from CNs, to control feeding site development. In addition to RKNs could be easily obtained in NEMATIC. As a first step, we these TFs, seven other interesting TFs selected by our search have selected the option ‘Transcriptional Regulation//All (2620)’ in the not yet been related, either directly or indirectly, to the plant–

© 2013 BSPP AND JOHN WILEY & SONS LTD MOLECULAR PLANT PATHOLOGY 6 J. CABRERA et al.

Table 2 Genes filtered with the NEMATIC Log ratio in gall transcriptomes 2 spreadsheet fulfilling three restrictions: (i) 3-dpi galls 7-dpi galls 14-dpi galls 21-dpi galls classified as plant transcription factors in any four different databases (RARTF, AGRIS, DATF AT1G29950 – – – 0.7 and PlnTFDB); (ii) induced by auxins in AT2G21050 ( ) ––– 1.4 LAX2 Nemhauser et al. (2006); and (iii) up-regulated AT2G42430 (LBD16) – 1.1 – – in any of the gall developmental stages AT2G47260 (WRKY23) 1.4 – – – AT3G01970 (WRKY45) – 0.9 – – recorded. In bold, those transcription factors AT3G60530 (GATA4) – 0.7 – – (TFs) experimentally studied during AT4G37790 (HAT22) – – 0.8 – plant–nematode interactions. AT5G06080 (LBD33) 1.4 – – – AT5G13330 (Rap2.6L) 1.1 – – 0.8 AT5G25190 (ESE3)–––0.7 AT5G47370 (HAT2) – 1.2 1.2 1.8 AT5G48150 (PAT1)–––0.9 AT5G65670 (IAA9)–––0.6 dpi, days post-inoculation. nematode interaction (Table 2). For example, HAT2 is a TF pref- A bar chart summarizing the data obtained for each group of erentially expressed in the early stages of lateral root formation, hormone-responsive genes is shown in Fig. 2. The predominant and has been suggested to be involved in auxin-induced hormone-responsive genes from the up-regulated genes that morphogenesis (Sawa et al., 2002). The roles of these TFs in galls are enriched in GCs relative to the exogenous application of induced by RKNs deserve further investigation. phytohormones were those up-regulated by IAA, whereas those regulated by CK were enriched in the up-regulated genes of Example of utility 2: phytohormone-related expression syncytia (Fig. 2). The proportion of ACC up-regulated genes was profiles in NFCs low among the up-regulated genes of syncytia and those Phytohormones are signalling molecules that regulate plant responding to CK in GCs. Strikingly, IAA- and CK-repressed genes development and cell differentiation, active processes occurring were also predominant in GCs and syncytia relative to the pro- during NFC formation (reviewed in Goverse and Bird, 2011). portion of genes after exogenous treatment (Fig. 2). This suggests Therefore, changes in hormone levels in NFCs are likely to that the repression of auxin and/or CK-regulated genes may be mediate, in part, the drastic transcriptional changes observed crucial in both feeding site types. Most studies so far have that are necessary for the successive establishment of the nema- focused on genes regulated by IAA or CK that are up-regulated in tode. The transcriptional data obtained by Nemhauser et al. feeding sites. Therefore, NEMATIC is useful to glimpse the main (2006), i.e. those genes up- or down-regulated after the exog- differences found between the hormone transcriptional balances enous addition of ABA, GL, IAA, ACC, CK, BR and JA, obtained of the two types of NFC. We noted a high percentage of IAA under the same experimental conditions and in the same labora- up-regulated genes (26%) in GCs, but of CK up-regulated genes tory, were compiled in NEMATIC (http://www.uclm.es/grupo/ (21%) in syncytia (Fig. 2; Table S2). Although, in syncytia, there gbbmp/english/nematic.asp). These sets of genes can be selected were more genes induced than repressed (121 and 84, respec- in the dropdown menu, facilitating the visualization of hormone- tively; Table S2), in GCs, there were 12.5 times more genes down- regulated genes across the plant–nematode interactions. As regulated than up-regulated by CK (25 and two genes, an example, we can deepen our understanding of hormone respectively; Table S2), therefore showing a clear repression transduction pathways and their regulatory networks underlying (Fig. 2) in GCs. An interesting and novel finding was the obser- NFC formation (GCs and syncytia). First, we select from the vation that IAA- and CK-repressed genes are predominantly dropdown menu the groups of up- or down-regulated genes for repressed respect to the experiment of Nemhauser et al. (2006) each hormone (Fig. 2). Second, making use of the filters placed in both feeding sites. above the GCs and syncytia transcriptomes, we filter by colour Although further expression and functional analysis should (red for up-regulated; green for down-regulated) those genes be performed to validate these results, NEMATIC (http:// from the plant–nematode transcriptomes with the same regula- www.uclm.es/grupo/gbbmp/english/nematic.asp) allowed us to tion as the selected group of hormone-regulated genes (Fig. 2; obtain, in a rapid manner, a general view of hormone-related Table S2, see Supporting Information). In addition, a summary transcriptional regulation during the plant–nematode interaction. table indicating the number of genes found in each NFC Moreover, the selection of genes responsive to hormones less transcriptome is shown each time we introduce a set of genes studied in plant–nematode interactions, such as BR or GL, could from the dropdown menu. also be of great interest.

MOLECULAR PLANT PATHOLOGY © 2013 BSPP AND JOHN WILEY & SONS LTD A plant–nematode interactions transcriptomic tool 7

Fig. 2 Right panel: percentage of hormone-responsive genes from each nematode feeding cell transcriptome responding to a specific hormone. Red bars indicate up-regulated genes and green bars indicate down-regulated genes. GCs, giant cells. Syncytia. Hormone treatments, transcript profile after exogenous treatment of the hormones indicated following Nemhauser et al. (2006). Left panel: flow chart showing the two-step procedure used to obtain the data represented in the right-hand graphs.

following steps can filter genes depending on their regulation in Example of utility 3: selection of ‘a top list of genes’ different Arabidopsis organs, developmental stages and/or cell after sequential filtering types, together with their expression patterns during all biotic One of the main problems encountered after data analysis from interactions contained in Genevestigator, hormonal regulation, holistic gene expression experiments is the huge amount of data etc. (Fig. 1). Finally, detailed and integrated information from dif- obtained. There is an increased need to filter these output data in ferent Arabidopsis databases can be obtained for each gene, order to select some of these genes to raise hypotheses that could allowing the selection of desired groups of genes, based on the lead to direct biotechnological purposes. With NEMATIC (http:// information of transgenic and mutant lines available, orthologous www.uclm.es/grupo/gbbmp/english/nematic.asp), it is possible to genes, genes co-expressed in various biological processes, protein obtain a particular list of genes with a ‘suitable selected’ expres- information, interactomes, references from functional analyses, sion profile during plant–nematode interactions. These output etc., leading to the easy selection of a reduced number of genes data, followed by a detailed analysis of gene expression patterns (Fig. 1) for further experimental work. during plant development and different biotic interactions, can be To facilitate the understanding of this type of analysis, an subsequently filtered, thus providing a sound aid to the selection example is given using NEMATIC (http://www.uclm.es/grupo/ and design of biotechnology-based tools for nematode control. For gbbmp/english/nematic.asp) applied to a particular comparison of this purpose, we provided a set of independent filter-based tools genes up-regulated in GCs (Table S3, see Supporting Information) that could be used for these multiple comparisons. The first step is for a hypothetical biotechnological application. We follow the to filter genes through comparisons with the plant–nematode initial hypothesis: ‘Genes fulfilling the following requirements: (i) interaction expression data available in Arabidopsis (Fig. 1). The up-regulated in GCs; (ii) with a fold change in GCs and syncytia

© 2013 BSPP AND JOHN WILEY & SONS LTD MOLECULAR PLANT PATHOLOGY 8 J. CABRERA et al. higher than 2.5; and (iii) barely expressed in selected root cell genes in the plant–nematode transcriptomes in a rapid manner. types and root parts’. Using the 310 up-regulated genes in GCs as The query list can be either a personal set of genes of interest or the query list, obtained from the dropdown menu in the ‘Plant– a list of those predefined in the spreadsheet tool, e.g. those Nematode’ sheet, we filtered (making use of the filter by text regulated by hormones (Nemhauser et al., 2006), those preferen- button above the column) those genes with a fold change ≥ 2.5 in tially expressed in each cell type of the root (Brady et al., 2007), GCs and syncytia (eight genes; Table S3). From these eight genes, those from different cell cycle phases (Menges et al., 2003), those those with very low expression (≤0.1) in roots were filtered (fil- from each MapMan category or those encoding TFs (Mitsuda tering by the colour code; in the ‘Anatomy’ sheet: roots, primary and Ohme-Takagi, 2009). The output list shows the expression root, root tip, meristematic zone, elongation zone, root hair zone, profiles of the selected genes in the Arabidopsis–nematode stele, pericycle or lateral root; in the ‘Cell types’ sheet: protoplasts transcriptomes, as well as hormone regulation, cell cycle informa- from any root tissue; two genes; Table S3). Therefore, from a list of tion, TF family and the specific root cell type for the DEGs. The 310 genes, only two genes fulfilled the desired expression require- compilation of these transcriptomes and the use of filters in the ments. Thereafter, we can consult the information provided for spreadsheet output, with their multiple combinations, becomes a these two genes from the different databases compiled in the powerful and useful tool to select candidate genes within the vast ‘Gene Info’ sheet (and, if necessary, apply new filters based on the amount of existing data. In an easy manner, genes fulfilling dif- MapMan category, hormone regulation, microRNA information, ferent criteria across several transcriptomes can be extracted. It is KEGG pathway, etc.). At4g10270 encodes a wound-responsive important to note that all query and filter steps can be performed protein and At2g28950 encodes an expansin (AtEXPA6), whose in a user-friendly, easy and intuitive way, making this tool useful expression in syncytia and GCs has been validated and confirmed for all scientists in the field with no bioinformatics background. by semi-quantitative reverse transcription-polymerase chain reac- NEMATIC is centred in the model plant Arabidopsis thaliana. tion (RT-PCR) by Wieczorek et al. (2006) and Barcala et al. (2010). However, microarray experiments have been performed to study These genes could possibly be used for biotechnological purposes plant–nematode interactions in other plant species, such as in various ways. For example, their promoters, expected to be Glycine max (Alkharouf et al., 2006; Ithal et al., 2007; Khan et al., active in both feeding cell types, but with no or low activity in the 2004; Klink et al., 2007, 2009), tomato (Portillo et al., 2013) and rest of the root, could drive the expression of cytotoxic or anti- rice (Kyndt et al., 2012). We have focused NEMATIC exclusively in nematode proteins in rootstocks. Arabidopsis as a starting point, because there were accessible data from two different plant–nematode groups (RKNs and CNs), with specific information at the tissue and cellular level, and the DISCUSSION great majority of the input data came from our own group. In In the last decade, microarray technology has been used to reveal addition, Arabidopsis is a model for dicotyledonous plant species the molecular aspects underlying the interaction between the with easy access to different databases providing massive com- model plant Arabidopsis thaliana and the plant-endoparasitic plementary information. However, NEMATIC can be edited easily nematodes Meloidogyne spp. and Heterodera spp.A large amount to introduce new transcriptomes obtained from other plant of data and extensive lists of DEGs have been generated (Escobar systems if the user is interested. NEMATIC aims to be the starting et al., 2011; Kyndt et al., 2012; Portillo et al., 2013).These data are point for the construction of a full database with all transcriptomic scattered in their corresponding publications or in databases, (currently available or future) data related to plant–nematode making it difficult to cross-compare the expression profiles among interactions in different plant species, such as rice and several transcriptomes. It is also arduous to determine how a group of legumes. genes of interest is regulated in the different plant–nematode The examples provided in the present work highlight the poten- transcriptomes. Moreover, other transcriptomes have been tial of the NEMATIC spreadsheet tool. In a few seconds, it is released, such as those corresponding to cell cycle or hormone possible to select a short list of TFs induced by auxins and regulation, which could be of great interest for cross-comparison up-regulated in galls (Example 1, Table 2). Among them, we with those of plant–nematode interactions. However, the available detected genes already validated in galls (WRKY23 and LBD16)or tools for these holistic comparisons are not easy to handle. studied in the context of CNs (LAX2), but we also extracted a new Similar to the spreadsheet developed by Parizot et al. (2010) in set of candidate TF-coding genes with similar characteristics, from the context of lateral root formation, we present here a simple and the vast amount of existing data, paving the way for future studies useful tool, NEMATIC (http://www.uclm.es/grupo/gbbmp/english/ on the molecular basis of the plant–nematode interaction nematic.asp), which compiles transcriptomic data on plant– (Table 2). nematode interactions and adds other interesting transcriptomes NEMATIC is also a good option for a general overview of how and groups of genes, facilitating rapid data handling. With genes from particular functional categories perform during the NEMATIC, it is possible to obtain the expression profiles of a list of plant–nematode interaction. In the second example provided, we

MOLECULAR PLANT PATHOLOGY © 2013 BSPP AND JOHN WILEY & SONS LTD A plant–nematode interactions transcriptomic tool 9 showed the trends of genes regulated by hormones in RKN and CN ACKNOWLEDGEMENTS feeding cells (Fig. 2); a similar analysis could be performed with genes from other categories of interest, such as the cell wall, cell This work was supported by the Spanish Government (AGL2010-17388 to cycle or secondary metabolism. Novel results could also be CE and CSD2007-057 to CF) and the Castilla-La Mancha Government obtained easily, although they require further validation, such as (PCI08-0074-0294 to CF and CE). JC was supported by a fellowship from the Ministry of Education, Spain. The authors have no conflicts of interest the large proportion of auxin down-regulated genes in NFCs to declare. (Fig. 2; Table S2). The proportion of auxin-regulated genes in NFCs is important with regard to both up- and down-regulated genes, REFERENCES but 136 and 54 auxin down-regulated genes (Nemhauser et al., Ali, M.A., Abbas, A., Kreil, D.P. and Bohlmann, H. (2013) Overexpression of the 2006) were also down-regulated in syncytia and GCs, respectively, transcription factor RAP2.6 leads to enhanced callose deposition in syncytia and whereas 61 and 20 were up-regulated (Table S3). As the study of enhanced resistance against the beet cyst nematode Heterodera schachtii in Arabidopsis roots. BMC Plant Biol. 13, 47. this hormone in relation to the plant–nematode interaction has Alkharouf, N.W., Klink, V.P., Chouikha, I.B., Beard, H.S., MacDonald, M.H., Meyer, been centred in the up-regulated genes, the study of the many S., Knap, H.T., Khan, R. and Matthews, B.F. (2006) Timecourse microarray analyses reveal global changes in gene expression of susceptible Glycine max (soybean) roots down-regulated genes identified using NEMATIC may change during infection by Heterodera glycines (soybean cyst nematode). Planta, 224, 838– the view of auxin-regulated genes during plant–nematode 852. interactions. de Almeida Engler, J. and Favery, B. (2011) The plant cytoskeleton remodelling in nematode induced feeding sites. In: Genomics and Molecular Genetics of Plant– In the third example, we showed that the use of conventional Nematode Interactions (Jones, J., Gheysen, G. and Fenoll, C., eds), pp. 369–394. Excel filters opens up a pathway to obtain many different combi- Dordrecht: Springer. nations of genes fulfilling various requirements. For instance, it is de Almeida Engler, J., Kyndt, T., Vieira, P., Van Cappelle, E., Boudolf, V., Sanchez, V., Escobar, C., De Veylder, L., Engler, G., Abad, P. and Gheysen, G. (2012) CCS52 possible to identify rapidly genes oppositely regulated in RKNs and DEL1 genes are key components of the endocycle in nematode-induced feeding and CNs, those unique for each interaction or those commonly sites. Plant J. 72, 185–198. Barcala, M., Garcia, A., Cabrera, J., Casson, S., Lindsey, K., Favery, B., regulated. Although the genes and the results obtained with this García-Casado, G., Solano, R., Fenoll, C. and Escobar, C. (2010) Early tool must be studied functionally, it may represent a good starting transcriptomic events in microdissected Arabidopsis nematode-induced giant cells. point to select candidate genes for future validation (Table S3). Plant J. 61, 698–712. Barthels, N., van der Lee, F.M., Klap, J., Goddijn, O.J., Karimi, M., Puzio, P., In conclusion, databases compiling data on transcriptomics Grundler, F.M., Ohl, S.A., Lindsey, K., Robertson, L., Robertson, W.M., from plant–nematode interactions are still lacking. The friendly Van Montagu, M., Gheysen, G. and Sijmons, P.C. (1997) Regulatory sequences of Arabidopsis drive reporter gene expression in nematode feeding structures. Plant and visual spreadsheet tool NEMATIC contains the data available Cell, 9, 2119–2134. in complex datasets from the main plant-endoparasitic nematodes Berckmans, B., Vassileva, V., Schmid, S.P.C., Maes, S., Parizot, B., Naramoto, S., interacting with the model plant Arabidopsis. This is the first time Magyar, Z., Alvim Kamei, C.L., Koncz, C., Bögre, L., Persiau, G., De Jaeger, G., Friml, J., Simon, R., Beeckman, T. and De Veylder, L. (2011) Auxin-dependent cell that a genome-wide resource for data mining, which facilitates cycle reactivation through transcriptional regulation of Arabidopsis E2Fa by lateral efficient in silico studies on plant–nematode biology, has been organ boundary proteins. Plant Cell, 23, 3671–3683. Bird, D.M. (1996) Manipulation of host gene expression by root-knot nematodes. made available. The main utilities of the tool are rapid cross- J. Parasitol. 82, 881–888. comparisons with other complex datasets and the acquisition of Brady, S.M., Orlando, D.A., Lee, J.Y., Wang, J.Y., Koch, J., Dinneny, J.R., Mace, D, customized gene selections through sequential comparative and Ohler, U. and Benfey, P.N. (2007) A high-resolution root spatiotemporal map reveals dominant expression patterns. Science, 318, 801–806. filtering steps. Therefore, NEMATIC is an efficient and easy to Caillaud, M.C., Dubreuil, G., Quentin, M., Perfus-Barbeoch, L., Lecomte, P., de handle instrument for the selection of genes with biotechnological Almeida Engler, J., Abad, P., Rosso, M.N. and Favery, B. (2008) Root-knot potential in the control of plant-endoparasitic nematodes, and in nematodes manipulate plant cell functions during a compatible interaction. J. Plant Physiol. 165, 104–113. the basic molecular understanding of plant–nematode interac- Damiani, I., Baldacci-Cresp, F., Hopkins, J., Andrio, E., Balzergue, S., Lecomte, P., tions, accelerating collective progress in the functional genomics Puppo, A., Abad, P., Favery, B. and Hérouart, D. (2012) Plant genes involved in harbouring symbiotic rhizobia or pathogenic nematodes. New Phytol. 194, 511–522. of plant–nematode interactions. Davuluri, R.V., Sun, H., Palaniswamy, S.K., Matthews, N., Molina, C., Kurtz, M. and Grotewold, E. (2003) AGRIS: Arabidopsis gene regulatory information server, an information resource of Arabidopsis cis-regulatory elements and transcription factors. BMC Bioinformatics, 4, 25. EXPERIMENTAL PROCEDURES Escobar, C., Brown, S. and Mitchum, M.G. (2011) Transcriptomic and proteomic analysis of the plant response to nematode infection. In: Genomics and Molecular NEMATIC (http://www.uclm.es/grupo/gbbmp/english/nematic.asp) has Genetics of Plant–Nematode Interactions (Jones, J., Gheysen, G. and Fenoll, C., eds), been constructed using Microsoft Excel 2007 and can also be opened pp. 157–174. Dordrecht: Springer. Fuller, V.L., Lilley, C.J., Atkinson, H.J. and Urwin, P.E. (2007) Differential gene using Microsoft Excel 2010 for Windows XP,Windows 7 or Mac Os X. The expression in Arabidopsis following infection by plant-parasitic nematodes spreadsheet has been saved in the ‘Office Excel 2007 Binary’ file format Meloidogyne incognita and Heterodera schachtii. Mol. Plant Pathol. 8, 2–4. (.xlsb extension), which minimizes its size. The .xlsb extension format is Gheysen, G. and Fenoll, C. (2002) Gene expresion in nematode feeding sites. Annu. suitable to store files that contain a large amount of data and can be read Rev. Phytopathol. 40, 191–219. Gheysen, G. and Mitchum, M.G. (2009) Molecular insights in the susceptible plant rapidly. More detailed information on the use and content of NEMATIC is response to nematode infection. In: Cell Biology of Plant Nematode Parasitism (Berg, described in Experimental Procedures S1. R.H. and Taylor, C.G., eds), pp. 45–82. Heidelberg: Springer.

© 2013 BSPP AND JOHN WILEY & SONS LTD MOLECULAR PLANT PATHOLOGY 10 J. CABRERA et al.

Goddijn, O.J., Lindsey, K., van der Lee, F.M., Klap, J.C. and Sijmons, P.C. (1993) Nicol, J.M., Turner, S.J., Coyne, D.L. and Nijs, L. (2011) Current nematode threats to Differential gene expression in nematode-induced feeding structures of transgenic world agriculture. In: Genomics and Molecular Genetics of Plant–Nematode Interac- plants harbouring promoter-gusA fusion constructs. Plant J. 4, 863–873. tions (Jones, J., Gheysen, G. and Fenoll, C., eds), pp. 21–45. Dordrecht: Springer. Goverse, A. and Bird, D. (2011) The role of plant hormones in nematode feeding Ogata, Y., Sakurai, N., Aoki, K., Suzuki, H., Okazaki, K., Saito, K. and Shibata, D. cell formation. In: Genomics and Molecular Genetics of Plant–Nematode (2009) KAGIANA: an excel-based tool for retrieving summary information on Interactions (Jones, J., Gheysen, G. and Fenoll, C., eds), pp. 325–348. Dordrecht: Arabidopsis genes. Plant Cell Physiol. 50, 173–177. Springer. Parizot,B.,De Rybel,B. and Beeckman,T. (2010) VisuaLRTC:a new view on lateral root Grunewald, W., Karimi, M., Wieczorek, K., Van de Cappelle, E., Wischnitzki, E., initiation by combining specific transcriptome data sets. Plant Physiol. 153, 34–40. Grundler, F., Inzé, D., Beeckman, T. and Gheysen, G. (2008) A role for AtWRKY23 Portillo, M., Cabrera, J., Lindsey, K., Topping, J., Andrés, M.F., Emiliozzi, M., in feeding site establishment of plant-parasitic nematodes. Plant Physiol. 148, 358– Oliveros, J.C., García-Casado, G., Solano, R., Koltai, H., Resnick, N., Fenoll, C. 368. and Escobar, C. (2013) Distinct and conserved transcriptomic changes during Grunewald, W., van Noorden, G., Van Isterdael, G., Beeckman, T., Gheysen, G. nematode-induced giant cell development in tomato compared with Arabidopsis: a and Mathesius, U. (2009) Manipulation of auxin transport in plant roots during functional role for gene repression. New Phytol. 197, 1276–1290. Rhizobium symbiosis and nematode parasitism. Plant Cell, 21, 2553–2562. Puthoff, D.P., Nettleton, D., Rodermel, S.R. and Baum, T.J. (2003) Arabidopsis gene Guo, A., He, K., Liu, D., Bai, S., Gu, X., Wei, L. and Luo, J. (2005) DATF: a database expression changes during cyst nematode parasitism revealed by statistical analyses of Arabidopsis transcription factors. Bioinformatics, 21, 2568–2569. of microarray expression profiles. Plant J. 33, 911–921. Hruz, T., Laule, O., Szabo, G., Wessendorp, F., Bleuler, S., Oertle, L., Oertle, L., Riano-Pachon, D.M., Ruzicic, S., Dreyer, I. and Mueller-Roeber, B. (2007) PlnTFDB: Oertle, L., Widmayer, P., Gruissem, W. and Zimmermann, P. (2008) an integrative plant transcription factor database. BMC Bioinformatics, 8, 42. Genevestigator v3: a reference expression database for the meta-analysis of Sawa, S., Ohgishi, M., Goda, H., Higuchi, K., Shimada, Y., Yoshida, S. and Koshiba, transcriptomes. Adv. Bioinformatics, 2008, 420747. T. (2002) The HAT2 gene, a member of the HD-Zip gene family, isolated as an auxin Hussey, R.S. and Grundler, F.M. (1998) Nematode parasitism of plants. In: Physiology inducible gene by DNA microarray screening, affects auxin response in Arabidopsis. and Biochemistry of Free-living and Plant Parasitic Nematodes (Perry, R.N. and Plant J. 32, 1011–1022. Wright, J., eds), pp. 213–243. Oxford: CAB International Press. Sobczak, M. and Golinowski, W. (2011) Cyst nematodes and syncytia. In: Genomics Iida, K., Seki, M., Sakurai, T., Satou, M., Akiyama, K., Toyoda, T., Konagaya, A. and and Molecular Genetics of Plant–Nematode Interactions (Jones, J., Gheysen, G. and Shinozaki, K. (2005) RARTF: database and tools for complete sets of Arabidopsis Fenoll, C., eds), pp. 61–82. Dordrecht: Springer. transcription factors. DNA Res. 12, 247–256. Szakasits, D., Heinen, P., Wieczorek, K., Hofmann, J., Wagner, F., Kreil, D.P., Ithal, N., Recknor, J., Nettleton, D., Maier, T., Baum, T.J. and Mitchum, M.G. (2007) Sykacek, P., Grundler, F.M. and Bohlmann, H. (2009) The transcriptome of Developmental transcript profiling of cyst nematode feeding cells in soybean roots. syncytia induced by the cyst nematode Heterodera schachtii in Arabidopsis roots. Mol. Plant–Microbe Interact. 20, 510–525. Plant J. 57, 771–784. Jammes, F., Lecomte, P., de Almeida-Engler, J., Bitton, F., Martin-Magniette, M.L., Thimm, O., Blasing, O., Gibon, Y., Nagel, A., Meyer, S., Kruger, P., Selbig, J., Müller, Renou, J.P., Abad, P. and Favery, B. (2005) Genome-wide expression profiling L.A., Rhee, S.Y. and Stitt, M. (2004) MAPMAN: a user-driven tool to display of the host response to root-knot nematode infection in Arabidopsis. Plant J. 44, genomics data sets onto diagrams of metabolic pathways and other biological 447–458. processes. Plant J. 37, 914–939. Jones, M.G.K. (1981) Host cell responses to endoparasitic nematode attack: structure Wieczorek, K., Golecki, B., Gerdes, L., Heinen, P., Szakasits, D., Durachko, D.M., and function of giant cells and syncytia. Ann. Appl. Biol. 97, 353–372. Cosgrove, D.J., Kreil, D.P., Puzio, P.S., Bohlmann, H. and Grundler, F.M. (2006) Jones, M.G.K. and Goto, D.B. (2011) Root-knot nematodes and giant cells. In: Expansins are involved in the formation of nematode-induced syncytia in roots of Genomics and Molecular Genetics of Plant–Nematode Interactions (Jones, J., Arabidopsis thaliana. Plant J. 48, 98–112. Gheysen, G. and Fenoll, C., eds), pp. 83–102. Dordrecht: Springer. Yang, Y., Hammes, U.Z., Taylor, C.G., Schachtman, D.P. and Nielsen, E. (2006) High- Khan, R., Alkharouf, N., Beard, H., MacDonald, M., Chouikha, I., Meyer, S., affinity auxin transport by the AUX1 influx carrier protein. Curr. Biol. 16, 1123–1127. Grefenstette, J., Knap, H. and Matthews, B. (2004) Microarray analysis of gene expression in soybean roots susceptible to the soybean cyst nematode two days post invasion. J. Nematol. 36, 241–248. SUPPORTING INFORMATION Klink, V.P., Overall, C.C., Alkharouf, N.W., MacDonald, M.H. and Matthews, B.F. (2007) Laser capture microdissection (LCM) and comparative microarray expression Additional Supporting Information may be found in the online analysis of syncytial cells isolated from incompatible and compatible soybean (Glycine max) roots infected by the soybean cyst nematode (Heterodera glycines). version of this article at the publisher’s web-site: Planta, 226, 1389–1409. Klink, V.P., Hosseini, P., Matsye, P., Alkharouf, N.W. and Matthews, B.F. (2009) A Table S1 Experimental conditions and statistical analysis used in gene expression analysis of syncytia laser microdissected from the roots of the the experiments on plant–nematode interactions compiled in Glycine max (soybean) genotype PI 548402 (Peking) undergoing a resistant reaction NEMATIC as described by the authors in the reference papers. after infection by Heterodera glycines (soybean cyst nematode). Plant Mol. Biol. 71, 525–567. Table S2 Genes differentially expressed under seven different Kyndt, T., Denil, S., Haegeman, A., Trooskens, G., Bauters, L., Van Criekinge, W., hormone treatments of Arabidopsis plants (Nemhauser et al., De Meyer, T. and Gheysen, G. (2012) Transcriptional reprogramming by root knot and migratory nematode infection in rice. New Phytol. 196, 887–900. 2006), which are also differentially expressed in plant-nematode Lee, C., Chronis, D., Kenning, C., Peret, B., Hewezi, T., Davis, E.L., Baum, T.J., feeding cells, listed in 14 independent sheets (up- and down- Hussey, R., Bennett, M. and Mitchum, M.G. (2011) The novel cyst nematode regulated). Arabidopsis Genome Initiative (AGI) codes for the effector protein 19C07 interacts with the Arabidopsis auxin influx transporter LAX3 to control feeding site development. Plant Physiol. 155, 866–880. genes from each described treatment are provided. Mathesius, U. (2003) Conservation and divergence of signalling pathways between Table S3 Flow chart and details of the filtering steps to obtain roots and soil microbes—the Rhizobium–legume symbiosis compared to the devel- opment of lateral roots, mycorrhizal interactions and nematode-induced galls. Plant genes fulfilling the following criteria: (i) up-regulated in giant cell Soil, 255, 105–119. (GCs); (ii) with a fold change in GCs and syncytia of greater than Menges, M., Hennig, L., Gruissem, W. and Murray, J.A. (2003) Genome-wide gene 2.5; and (iii) barely expressed in selected root cell types and root expression in an Arabidopsis cell suspension. Plant Mol. Biol. 53, 423–442. Mitsuda, N. and Ohme-Takagi, M. (2009) Functional analysis of transcription factors parts. in Arabidopsis. Plant Cell Physiol. 50, 1232–1248. Experimental Procedures S1 Document containing extended Nemhauser, J.L., Hong, F. and Chory, J. (2006) Different plant hormones regulate similar processes through largely nonoverlapping transcriptional responses. Cell, and more detailed information on the use, content and data 126, 467–475. sources of NEMATIC.

MOLECULAR PLANT PATHOLOGY © 2013 BSPP AND JOHN WILEY & SONS LTD CHAPTER 3: Molecular parallelisms between lateral roots and giant cell and gall formation Chapter 1: Introduction

Chapter 2: Holistic Studies A general overview of the plant nematode interaction Chapter 3: LBD16 & Lateral Roots and the research Chapter 4: 3D Reconstruction performed in this field

Developmental pathways Holistic studies performed mediated by hormones in the susceptible and and altered during the resistant interactions with plant- nematode root- knot nematodes interactions

A method for the 3D phenotping of the giant cells. Insights into their volumes and morphological features

Transcriptomic and The transcriptomes of molecular mechanisms are early developing giant cells common between lateral from tomato and root and giant cells and arabidopsis showed a gall formation. LBD16 is a common massive and key gene in both functional gene repression processes

A tool to facilitate the selection of genes frm the data generated on trasncriptomes from plant- nematode interactions and other fields

miRNA and tasiRNA Regulation of LBD16 co- pathways for gene regulated genes differs silencing are functional in between giant cells and galls and the rasiRNAs are syncytia. Differences in highly induced. Molecular the balance parallelisms with lateral Auxins/Cytokinins between roots both feeding cells Aim of the chapter.

The creation of the tool NEMATIC allowed us the comparison between the up- regulated genes in early developing giant cells and galls and those genes found as characteristics of the different root cell types accordingly to the transcriptomes performed by Dr. Philip Benfey’s group from cell-sorted root cells. The results presented in this chapter in the article “A role for LATERAL ORGAN BOUNDARIES- DOMAIN 16 during the interaction Arabidopsis–Meloidogyne spp. provides a molecular link between lateral root and root-knot nematode feeding site development” showed that the transcriptomes of giant cells and galls were significantly enriched in those genes characteristic of undifferentiated root cells, as those from the quiescent centre and from the xylem pole pericycle cells dividing to form a new lateral root primordia (founder cells). Moreover, we showed that the LBD16, a key transcription factor for the regulation of the lateral root formation mediated by auxins in the founder cells, is necessary for a correct giant cell and gall formation and that it is induced by nematode secretions.

We demonstrated that LBD16 has a role during giant cell and gall formation by Meloidogyne spp. in Arabidopsis whilst this gene seems not to have a relevant role in the syncytia induced by cyst nematodes. To further study the differences found in the expression and role of LBD16 in both feeding cells, in the article “Genes co-regulated with LBD16 in nematode feeding sites inferred from in silico analysis show similarities to regulatory circuits mediated by the auxin/cytokinin balance in Arabidopsis” we found that LBD16 and its co-regulated genes are integrated in signalling cascades mediated by the auxin/cytokinin balance in Arabidopsis Research

A role for LATERAL ORGAN BOUNDARIES-DOMAIN 16 during the interaction Arabidopsis–Meloidogyne spp. provides a molecular link between lateral root and root-knot nematode feeding site development

Javier Cabrera1, Fernando E. Dıaz-Manzano1, Marıa Sanchez1, Marie-No€elle Rosso2, Teresa Melillo3, Tatsuaki Goh4, Hidehiro Fukaki4, Susana Cabello5, Julia Hofmann5, Carmen Fenoll1 and Carolina Escobar1 1Facultad de Ciencias Ambientales y Bioquımica, Universidad de Castilla-La Mancha, Av. Carlos III s/n, E-45071 Toledo, Spain; 2INRA, Aix-Marseille Universite, UMR 1163, Biotechnologie des Champignons Filamenteux, F-13009 Marseille, France; 3Istituto per la Protezione de lle Piante, CNR, 70126 Bari, Italy; 4Department of Biology, Graduate School of Science, Kobe University, 1-1 Rokkodai, Kobe 657-8501, Japan; 5Division of Plant Protection, Department of Crop Sciences, University of Natural Resources and Applied Life Sciences, Konrad Lorenz Str. 24, Tulln a. d. Donau A-3430, Austria

Summary Author for correspondence:  Plant endoparasitic nematodes induce the formation of their feeding cells by injecting effec- Carolina Escobar tors from the esophageal glands into root cells. Although vascular cylinder cells seem to be Tel: +34 925268800 ext. 5476 involved in the formation of root-knot nematode (RKN) feeding structures, molecular evi- Email: [email protected] dence is scarce. We address the role during gall development of LATERAL ORGAN Received: 10 February 2014 BOUNDARIES-DOMAIN 16 (LBD16), a key component of the auxin pathway leading to the Accepted: 21 March 2014 divisions in the xylem pole pericycle (XPP) for lateral root (LR) formation.  Arabidopsis T-DNA tagged J0192 and J0121 XPP marker lines, LBD16 and DR5::GUS New Phytologist (2014) 203: 632–645 promoter lines, and isolated J0192 protoplasts were assayed for nematode-dependent doi: 10.1111/nph.12826 gene expression. Infection tests in LBD16 knock-out lines were used for functional analysis.  Key words: Arabidopsis, auxins, giant cells J0192 and J0121 lines were activated in early developing galls and giant cells (GCs), resem- (GCs), Lateral Organ Boundaries-Domain 16 bling the pattern of the G2/M-transition specific ProCycB1;1:CycB1;1(NT)-GUS line. LBD16 (LBD16), lateral root, root-knot nematodes was regulated by auxins in galls as in LRs, and induced by RKN secretions. LBD16 loss of func- (RKNs), xylem pole pericycle (XPP). tion mutants and a transgenic line with defective XPP cells showed a significantly reduced infection rate.  The results show that genes expressed in the dividing XPP, particularly LBD16, are impor- tant for gall formation, as they are for LR development.

Introduction procambial or pericycle cells, whereas males develop in syncytia from pericycle cells (Golinowski et al., 1996; Sobczak et al., Plant parasitic nematodes represent a major threat to agriculture 1997). From these initial cells, the syncytium incorporates neigh- (Bird et al., 2009) and the need to develop nematode resistance is boring cells that fuse via dissolution of their cell walls along plas- intensifying with the banning of pesticides (Atkinson et al., modesmata (Sobczak & Golinowski, 2009). Based on 2012). Nematodes induce the formation of sophisticated feeding histological observations, it is believed that giant cells (GCs) structures inside the root that operate as physiological sinks to induced by RKNs are most commonly formed from parenchy- supply nutrients to the nematode (Perry & Moens, 2011). matic cells within the stele that surround the nematode head and Among the most damaging groups of endoparasites are the root- that the primary effector molecules stimulating GC formation knot nematodes (RKNs; Meloidogyne spp.) and the cyst nema- are esophageal gland secretions (Berg et al., 2009). Among these, todes (CNs; Heterodera spp. and Globodera spp.). Juveniles are Mi16D10, which interacts with nuclear plant targets, such as attracted by root chemical signals that modify nematode behavior SCARECROW-like transcription factors, may function in the and gene expression, penetrate the root and establish in the vascu- extensive transcriptional reprogramming that accompanies GC lar cylinder, transforming selected root cells into their feeding ontogenesis (Quentin et al., 2013). Selected cells undergo mitosis sites (Perry & Moens, 2011; Teillet et al., 2013). In Arabidopsis, with aborted cytokinesis, becoming enlarged, multinucleated, females of Heterodera schachtii usually develop in syncytia from endoreduplicated cells that feed the nematode (Caillaud et al.,

632 New Phytologist (2014) 203: 632–645 Ó 2014 The Authors www.newphytologist.com New Phytologist Ó 2014 New Phytologist Trust New Phytologist Research 633

2008; Gheysen & Mitchum, 2009; Almeida-Engler & Favery, infection stages. Furthermore, functional studies indicate that 2011). Concurrently, root pericycle and cortical cells proliferate LBD16 is a key molecular player in XPP cells during GC and gall around the nematode, increasing the girth of the root and form- development. ing a gall (Berg et al., 2009). There are parallelisms between lateral root (LR) development, Materials and Methods RKN gall formation and the establishment of Rhizobium symbio- nts, based on similar expression patterns of genes encoding tran- Nematode populations scription factors and cell cycle regulators (Mathesius, 2003). This, together with the common activation patterns of tagged Meloidogyne javanica (MIK; Portillo et al., 2009) and M. arenaria lines in LRs and galls (Barthels et al., 1997) and the pericycle cell (obtained from Dr M. F. Andres, CSIC, Spain) were multiplied divisions that accompany nematode establishment, suggests a in vitro on cucumber (Cucumis sativus cv Hoffmanns Giganta) relationship between LR emergence and RKN feeding site forma- grown in 0.3% Gamborg medium (Gamborg et al., 1968) with tion. However, clear molecular evidence is still lacking. Arabid- 3% sucrose. Heterodera schachtii population (from Dr J. Hof- opsis LRs originate exclusively from pericycle founder cells mann, BOKU, Austria) in mustard seedlings (Sinapsis alba cv Al- located opposite the xylem poles (xylem pole pericycle (XPP) batros). Egg hatching was stimulated in sterile water for RKNs cells). Members of the LATERAL ORGAN BOUNDARIES- and in sterile 3 mM ZnCl2 for CNs. DOMAIN (LBD) transcription factor family have been shown to play a crucial role in LR development (Okushima et al., 2007; Plant material, growth conditions and nematode Bell et al., 2012), being highly conserved across plant species. inoculation The transcription of LBD16, LBD18 and LBD29 is activated by AUXIN RESPONSE FACTOR 7(ARF7) and ARF19 in the auxin Arabidopsis thaliana (L.) Heynh Columbia-0 (Col-0) was the signaling cascade, leading to the emergence of a new LR (Okushi- background of all the transgenic lines. For expression analysis, we ma et al., 2007; Lee et al., 2009). The localized activity of LBD16 used J0121, J0192 (Laplaze et al., 2005), DR5::GUS (Ulmasov and its related LBDs is involved in the symmetry breaking of LR et al., 1997) and two different versions of pLBD16::GUS carrying founder cells for LR initiation (Goh et al., 2012). Interestingly, either a 1500-bp or a 2500-bp promoter sequence (Laplaze et al., LBD16, LBD17, LBD18 and LBD29 have been identified as 2005; Okushima et al., 2007). Col-0, lbd16-1, 35S:LBD16- important regulators that function during callus induction, SRDX (Okushima et al., 2007), pLBD16:LBD16-SRDX (Goh thereby establishing a molecular link between auxin signaling and et al., 2012), J0121 and J0121>>DTA (Laplaze et al., 2005) were plant regeneration programs (Fan et al., 2012). used for the infection tests. The T-DNA insertion mutant lbd16- Auxin is a key regulatory factor that causes changes in gene 1 (SALK_095791) was genotyped by PCR to identify homozy- expression upon nematode infection (Goverse & Bird, 2011). gous plants (Supporting Information Fig. S1). The seeds were Direct evidence for this role of auxin is the finding that promot- surface-sterilized and sown in modified Gamborg B5 medium ers containing ‘ARF binding sites’, such as the promoter of an (Escobar et al., 2003). auxin responsive gene from the Gretchen Hagen 3 family of For expression analysis, three independent experiments were soybean (GH3) and the synthetic auxin-inducible promoter, performed with 20–30 plants per line each. The plates were inoc- DR5, are activated at early stages of gall and syncytium develop- ulated 12 d after germination with freshly hatched M. javanica, ment (Hutangura et al., 1999; Karczmarek et al., 2004). Addi- M. arenaria (Escobar et al., 2003) or H. schachtii (Golinowski tionally, Meloidogyne incognita infection did not progress in the et al., 1996) juveniles and carefully examined every 12 h (Portillo tomato (Solanum lycopersicum) auxin-insensitive mutant dia- et al., 2013). Galls or syncytia were hand-dissected for micros- geotropica, dgt (Richardson & Price, 1982). However, consistent copy inspection at selected developmental time-points. and solid findings based on genetic analysis are only available for For functional analysis, three independent infection tests were CNs, in which a range of Arabidopsis mutants in auxin transport, performed with a minimum of 60 individual plants per line per homeostasis and responses have been tested (e.g. mutations at the experiment (Fig. S2). The plates were inoculated with 7–10 AXR2 locus, PIN-FORMED 1 and PIN-FORMED 2/auxin M. javanica, M. arenaria or H. schachtii juveniles per main root. efflux carrier component 2: axr2, pin1-1 and pin2/eir1-1, A thin temperate (30–37°C) film of 0.3% agarose in Gamborg respectively Goverse et al., 2000). To date, a direct molecular medium was placed over the roots to facilitate nematode penetra- relationship between LR development, gall and GC development tion. Plants were vertically grown to avoid early appearance of and auxin signaling has not yet been described. LRs. The plates were systematically examined under the stereo- This work shows that the promoter of LBD16, which encodes microscope at 10 d post infection (dpi) to determine the number a transcription factor involved in LR initiation, is active from of infections per plant. very early stages of gall development, regulated by auxins in galls To estimate GC size in Col-0, J0121, 35S::LBD16-SRDX and and induced by nematode secretions. XPP marker lines (J0192 J0121>>DTA, galls were hand-dissected at 15 dpi, fixed and and J0121) were also activated in galls with a pattern resembling embedded (Barcala et al., 2010). The sections were stained with that of the G2-M transition marker ProCycB1;1:CycB1;1(NT)- 1% toluidine blue in 1% borax solution (TAAB). The TrakEM2 GUS in XPP dividing cells. The activation pattern of the auxin (Cardona et al., 2012) plug-in for FIJI (Schindelin et al., 2012) sensor DR5 paralleled that of LBD16 only during the early was used to measure GC areas. Two representative galls from each

Ó 2014 The Authors New Phytologist (2014) 203: 632–645 New Phytologist Ó 2014 New Phytologist Trust www.newphytologist.com New 634 Research Phytologist

genotype were entirely sectioned into 2-lm sections. The 10 sec- A Leica TCS SP2 confocal laser scanning microscope was used tions in which the GCs showed the greatest expansion in each gall for the detection of GFP expression. Most galls were also hand- À were selected to quantify the area occupied by the GCs (< 100 sectioned and immediately stained with 0.5 lgml 1 propidium GCs sections scored) as well as to obtain the mean area and stan- iodide (PI) in phosphate-buffered saline (PBS) for 5 min, or dard errors among the sections as in Portillo et al. (2013). freshly embedded in 5% low melting agarose and subsequently Statistical analysis of the infection parameters and GC areas sectioned. Vibroslices were stained with PI under the same condi- was performed using ANOVA and the Schefee test using the SPSS tions, and immediately observed under the confocal microscope. package (IBM, Armonk, NY, USA) with P < 0.05. The emission spectra were set to 515, 600–700 and 617 nm for GFP, chloroplast autofluorescence and PI, respectively. A long- pass 500-nm dichroic beam splitter was used. Meloidogyne incognita secretions and protoplast isolation Stylet secretions were collected from freshly hatched J2s of In silico analysis of GC and gall transcriptomes M. incognita after induction in resorcinol as described by Rosso et al. (1999). The secreted proteins were separated by ion The genes that were up-regulated in 3-dpi GCs and gall tran- exchange chromatography on a High Q anion exchange resin scriptomes (Barcala et al., 2010) were manually compared with (Bio-Rad, Melville, NY, USA) in Bis-Tris (20 mM; pH 6). The the sets of enriched genes in each root cell type described by fraction eluted with 0.4 M NaCl, dialyzed with Bis-Tris 20 mM, Brady et al. (2007). To identify the transcripts enriched in GCs pH 6, and concentrated by ultrafiltration using YM50- and and galls, the v2 test was used with a significance level of YM3-kDa membranes, was tested in protoplasts prepared from P < 0.05. The abundance of each group in the Arabidopsis J0192 Arabidopsis leaves (Fig. S8d). genome was used as the background. Mesophyll protoplasts were isolated from Arabidopsis J0192 and DR5::GFP using a modified method described by Sheen Results (2002). Leaves from 4-wk-old plants grown in vitro were col- lected, cut into 0.5–1-mm2 strips, transferred to 0.5 M mannitol GCs and galls show transcriptional similarities with and plasmolyzed for 1 h. They were transferred to a cell wall undifferentiated root cell types digestion solution containing 0.8% Cellulase R-10 (Serva, Hei- delberg, Germany) and 0.2% Macerozyme (Serva) in 0.4 M It has been established that the CN H. schachtii induces syncy- mannitol. Protoplasts were collected, washed and incubated in tium formation from root procambial or pericycle cells (Goli- À Murashige and Skoog medium supplemented with 4.4 g l 1 vita- nowski et al., 1996; Sobczak et al., 1997). However, the root cell À À À mins (Sigma), 20 g l 1 sucrose, 0.6 g l 1 MES, 0.05 mg l 1 kine- types contributing to gall and/or GC formation are still a matter À tin and 40 mg l 1 ampicillin with different concentrations of of debate. To determine if specific root cell types shared molecu- 2,4-dichlorophenoxyacetic acid (2,4-D) or in a medium lacking lar determinants with early-developing galls and GCs, we per- auxin but containing nematode secretions (5–7 ll). As a negative formed pairwise comparisons between the up-regulated genes in control, the same medium without auxins or secretions was used. RKN feeding structures at 3 dpi (Barcala et al., 2010), and genes The fluorescence background was assessed with Col-0 plants. characteristically expressed in each of the 16 root cell types avail- Protoplasts were maintained in a growth chamber at 25°Cin able from Brady et al. (2007). Only the transcripts that were darkness. Three to four independent experiments with 200–300 enriched in the lateral root primordium (LRP) and quiescent cen- intact protoplasts per treatment in each experiment were scored ter cells were overrepresented in both galls and GCs after v2 for the percentage of GFP-expressing protoplasts. This ranged analyses (P < 0.05; see the Materials and Methods section; Fig. 1; from 0% in all the control samples to 25–50% in the auxin- and Table S1). Transcripts characteristic of developing xylem were secretion-treated samples in each independent experiment. The overrepresented specifically in GCs, whereas transcripts from the micrographs shown are representative of the intact protoplast endodermis were enriched in galls but not in GCs (Fig. 1; Table after each treatment. S1). These results suggest that the commonly induced genes in both galls and GCs show transcriptional profile similarities to those observed in undifferentiated root cell types, such as those Histological analysis of GUS and GFP expression from the LRP and the quiescent center. For GUS staining, hand-dissected galls were prefixed in 0.5% glutaraldehyde for 5 min under moderate vacuum and incubated Two marker lines for LRP initial cells are activated in GCs for 5–8 h in a solution containing 5 mM EDTA (pH 8), 0.05% and galls Triton X-100, 0.05 mM K3Fe(CN)6, 0.05 mM K4Fe(CN)6 and À 1mgml 1 X-GlcA in 50 mM sodium phosphate buffer. Galls Two Arabidopsis GAL4-GFP enhancer trap lines, J0192 and were photographed under a Leica Mz125 stereomicroscope J0121 (Laplaze et al., 2005), were inoculated with M. javanica. (Leica Microsystems, Wetzlar, Germany) or a Nikon eclipse 90i Line J0192 is an early and very specific marker line for XPP cells microscope (Nikon Corp., Tokyo, Japan). For tissue localization that divide to form a new LRP before its emergence, whereas of GUS expression, galls were fixed, sectioned at 2 lm and pro- J0121 is specific for two to three root pericycle cell files adjacent cessed as in Barcala et al. (2010). to the xylem poles along the entire root starting at the elongation

New Phytologist (2014) 203: 632–645 Ó 2014 The Authors www.newphytologist.com New Phytologist Ó 2014 New Phytologist Trust New Phytologist Research 635

Fig. 1 Schematic representation of root cell types with transcriptomic similarity to developing galls and giant cells (GCs) induced by Meloidogyne javanica. Specific Arabidopsis root cell transcriptomes (Brady et al., 2007) were compared with the up-regulated genes in 3-d post infection (dpi) GCs and gall transcriptomes (Barcala et al., 2010). Overrepresented gene overlaps were identified by v2 analysis (P < 0.05). A color code is indicated for the overrepresented root cell transcriptomes in galls and GCs. The table outlines the root marker cell lines used, showing v2 and P-values; P < 0.05 indicates overrepresentation or enrichment. pSE, protophloem sieve elements; mSE, metaphloem sieve element; CC, companion cells. zone (Laplaze et al., 2005). In the J0192>>GFP line, GFP expres- sion is driven by the promoter region of LBD16 (At2g42430; Fig. 2 Two independent pericycle markers are active in Arabidopsis galls Laplaze et al., 2005). We monitored the time-courses of and giant cells (GCs) formed by Meloidogyne javanica at very early >> developmental stages. Confocal images of galls stained with propidium J0192 GFP and J0121>>GFP expression in young galls iodide (red fluorescence) showing GFP (green) at 2, 4 and 6 d post (Fig. 2). We also performed in vivo monitoring by following the infection (dpi) from lines J0192>>GFP and J0121>>GFP (a, c, e and b, d, f, same galls along the consecutive infection stages (Fig. S3). Both respectively) are presented. (g, h) Semithin sections of J0192>>GUS (g) lines were activated as early as 1 dpi (Fig. S3). The J0192>>GFP and J0121>>GUS at 6 dpi (h). Asterisks, GCs; N, nematode. Bars, 100 lm. line showed GFP expression in the center of the gall at 2 dpi (Fig. 2a) and by 4 and also 6 dpi, the entire vascular cylinder inside the gall was expressing GFP (Fig. 2c,e; all channels for the A closer inspection of hand-dissected infected roots during the confocal images in Fig. 2 are shown in Fig. S4). Semithin sections migration-establishment phase (12–24 h post inoculation (hpi)) of 7-dpi galls from the GAL4-GUS enhancer trap line revealed a pair of contiguous pericycle cells expressing GFP near J0192>>GUS also showed a strong staining inside the vascular the nematode head (Fig. 3a, a1, a2), which closely resembled the cylinder, according to GFP expression, extended from the pericy- induction observed during LRP emergence in the pericycle cle layer inwards, including the GCs (Fig. 2g). (Fig. 3e). J0192 was frequently activated in XPP cells on both

Ó 2014 The Authors New Phytologist (2014) 203: 632–645 New Phytologist Ó 2014 New Phytologist Trust www.newphytologist.com New 636 Research Phytologist

Fig. 3 Xylem pericycle pole (XPP) cells in Arabidopsis galls induced by Meloidogyne javanica express molecular markers of lateral root (LR) initials and the G2/M cell cycle phase. (a, a1, a2, b, b1, c, c1, c2, d, d1) Confocal images of galls stained with propidium iodide (red fluorescence) showing GFP (green) from line J0192>>GFP, which marks the identity of LR initials. (a, a1) Galls at 12–24 h post infection (hpi). (a2) Detailed view of the XPP, xylem (X) and nematode (N) head with a close-up image of the overlap with the transmission image of (a1) as indicated. (b, b1, c, c1, c2, d, d1) Galls at 24–48 hpi (b, b1), 60–72 hpi (c, c1, c2) and 84–96 hpi (d, d1). (e–h) Confocal images of lateral root primordia (LRPs) at different stages from uninfected roots of line J0192>>GFP. (i–k) Transmission images of galls infected at 12, 24–36 and 72 hpi from line ProCycB1;1:CycB1;1(NT)-GUS, which marks cells in G2/M. White arrows indicate the activated cells at the XPPs and the direction of dividing cells relative to the root vascular cylinder. Bars: 100 lm, except for enlarged photos (a1, a2, b1, b1, c2 and d1), 25 lm.

sides of the vascular cylinder inside the gall (Fig. 3b, b1), contrary vascular cylinder, whereas the adjacent root portions maintained to the induction in only one side of the XPP during LR forma- the expression in two to three files of XPP cells (Fig. 2b). GFP tion (Fig. 3e,f). As the galls developed and more cells formed, the expression increased toward the center of the gall during the GFP signal increased toward the vascular cylinder (Fig. 3c, c1), infection (4 and 6 dpi; Fig. 2d,f). Accordingly, staining of whereas in the LRP divisions occurred toward the cortex J0121>>GUS galls at 6 dpi showed a strong signal in all the cells (Fig. 3e–h). The labeled vascular cylinder cells exhibited abnor- inside the gall, including the GCs (Fig. 2h), similar to mal irregular division planes (see close-up image in Fig. 3c1) and J0192>>GUS. the number of labeled cell layers increased during gall develop- The ProCycB1;1:CycB1;1(NT)-GUS reporter line (Colon- ment until they had filled the entire vascular cylinder as early as Carmona et al., 1999) allows the detection of cells undergoing 48–60 hpi (Fig. 3d, d1). The restricted induction of J0192, a the G2/M transition, which are actively engaged in cell division. well-defined marker line for LRP initials, inside the galls at early CycB1;1(NT)-GUS expression was observed at both sides of the infection stages seems to correlate with the results from the in XPP during nematode migration and establishment (12–24 hpi; silico analysis. Fig. 3i). At 24–36 hpi, cells in the XPP were also clearly stained Similar results were obtained for the J0121>>GFP line. At at both sides of the vascular cylinder (Fig. 3j) until a strong GUS 1 dpi, GFP expression in pericycle cells inside the gall was similar stain was eventually detected in all the vascular cylinder cells to that of the adjacent root segments (Fig. S3a, a1). At 2 dpi, inside the gall (72 hpi; Fig. 3k). Therefore, the temporal pattern when the gall began to swell, additional GFP spread toward the of divisions in the gall resembled the expression pattern observed

New Phytologist (2014) 203: 632–645 Ó 2014 The Authors www.newphytologist.com New Phytologist Ó 2014 New Phytologist Trust New Phytologist Research 637

(a) (b) (c) (d)

(e) (f) (g) (h)

(i) (j) (k)

(l) (m) (n) (o)

Fig. 4 LATERAL ORGAN BOUNDARIES-DOMAIN 16 (LBD16)::GUS and the synthetic auxin-inducible promoter, DR5::GUS activation patterns show strong similarities only at early infection stages of Meloidogyne javanica in Arabidopsis. Galls from LBD16::GUS showed a strong stain in the center of the gall at 3 and 7 d post infection (dpi) (a, b) that decreased at 13 and disappeared at 21 dpi (c, d). At the early stages of infection (3 and 7 dpi), the activation patterns of DR5::GUS were similar (e, f). At medium-to-late stages of infection (13 and 21 dpi), a clear signal was still present in DR5::GUS (g, h). Semithin sections of galls from LBD16::GUS at 4 dpi (i) and from DR5::GUS at 4 dpi (j) showed GUS staining in giant cells (GCs) and in the small dividing cells inside the vascular cylinder. A comparative graph of the positive galls in both lines (k) shows that the percentage of LBD16::GUS-positive galls dropped to 8% at 15 dpi and eventually to 0% at 21 dpi, whereas the GUS staining for DR5 expression was maintained in over 80% of the galls until 29 dpi (dark blue bars, % positive galls for DR5; light blue bars, % positive galls for LBD16). Noninfected control roots for LBD16::GUS and DR5::GUS lines showed a clear GUS signal in lateral root primordials (LRPs) (m, o), and only a positive signal in the root tip of DR5::GUS, but not of LBD16::GUS (n, l). Asterisks, GCs. Bars: (a– h) 200 lm; (i, j) 50 lm; (l, m–o) 500 lm. in the J0192>>GFP and J0192>>GUS lines, which began at However, at 11–13 dpi the intensity of the GUS stain in the both sides of the XPP in the gall and eventually filled up the pLBD16::GUS galls decreased, and some of the galls no longer entire vascular cylinder inside the gall. expressed GUS (Fig. 4c,k); yet, DR5-driven GUS expression remained strong (Fig. 4g,k). In completely expanded 21-dpi galls of the pLBD16::GUS line, we seldom observed any GUS LBD16, a key regulator in the auxin signaling pathway expression (Fig. 4d), but most 21-dpi galls in DR5::GUS leading to LR formation, is expressed in galls during early/ remained GUS-positive (Fig. 4h). A comparative graph of the middle gall development GUS-positive galls in both lines (Fig. 4k) shows that at 2 dpi In the J0192>>GFP line, GFP expression is driven by the pro- all galls tested from both lines showed GUS staining, but the moter region of LBD16 (At2g42430; Laplaze et al., 2005), percentage of LBD16::GUS-positive galls dropped to 9% at encoding a transcription factor expressed specifically during the 15 dpi, whereas DR5 expression was maintained until 29 dpi first stages of LRP formation and induced by auxins (Okushi- (80%). A clear signal was observed in all cell types inside the ma et al., 2007; Lee et al., 2009). A transgenic line containing vascular cylinder from the pericycle inward in stained 2500 bp of the LBD16 promoter region fused to GUS pLBD16::GUS gall sections, including the GCs at 4 dpi (pLBD16::GUS; Okushima et al., 2007) was infected with (Fig. 4i), similar to the J0192>>GUS line (Figs 2, 4) and the M. javanica, and its expression pattern was compared with that DR5::GUS line at 4 dpi (Fig. 4j). Uninfected control plants of of the ‘auxin sensor’ synthetic promoter DR5 (Ulmasov et al., the pLBD16::GUS line only expressed GUS in the LRP 1997; Fig. 4). It showed strong expression at 3 dpi in the cen- (Fig. 4l,m), and DR5::GUS line and those of the DR5::GUS ter of the gall (Fig. 4a), similar to DR5::GUS (Fig. 4e). The line only expressed GUS in the LRP and the root tip (Fig. 4n, expression was strong at 7 dpi in both marker lines (Fig. 4b,f). o). Thus, pLBD16::GUS and DR5::GUS were expressed in

Ó 2014 The Authors New Phytologist (2014) 203: 632–645 New Phytologist Ó 2014 New Phytologist Trust www.newphytologist.com New 638 Research Phytologist

M. javanica galls at early infection stages in the same gall tis- LBD16 is regulated by auxins during gall development sues and cells; however, at middle/late infection stages, they showed opposite patterns. Additionally, the pLBD16 activation LBD16 is a key regulator in the auxin signaling pathway leading pattern observed after nematode infection validated the results to the formation of new LRs (Okushima et al., 2007). To further obtained with the enhancer trap line J0192 (Figs 2, 3). investigate its role in galls, 3-dpi galls from pLBD16::GUS plants A shorter, 1500-bp version of the LBD16 promoter mimicked were incubated for 4 d on medium containing 300 lM a-(phenyl the expression pattern obtained for the 2500-bp promoter in galls ethyl-2-one)-indole-3-acetic acid (PEO-IAA), an auxin (Fig. S5; Laplaze et al., 2005). However, the intensity of the GUS stain was weaker than that observed with the 2500-bp promoter under the same experimental conditions. We found that only one putative ARF binding element was present in the 1500-bp pro- moter version (Fig. S5e; top), whereas up to three ARF elements were present in the 2500-bp promoter (Fig. S5e; bottom), which could explain the higher intensity of the signal. Our results obtained with the DR5::GUS line and M. javanica are somewhat contradictory to those of Karczmarek et al. (2004) with M. incognita, who showed that DR5 was active at very early stages (18–24 hpi) around the nematode head but decreased at 3–5 dpi and eventually disappeared from 10 dpi onwards. However, Absmanner et al. (2013) showed DR5 signal until late infection stages. In our case, GUS signal was also maintained until late infection stages (> 29 dpi; Figs 4, S6). To rule out the possibility that GUS protein stability could influ- ence our results, transgenic lines carrying DR5::GFP were inocu- lated because GFP has a shorter half-life than GUS (de Ruijter et al., 2003). Strong GFP fluorescence was detected at 11, 15, 21 and 28 dpi in the center of the galls, which matches the GUS staining pattern obtained with the DR5::GUS line inside the vascular cylinder including the GCs (Fig. S7). These results confirm that the galls formed by M. javanica in Arabidopsis contained auxins as late as 29 dpi, when the gall is completely developed.

Fig. 5 LATERAL ORGAN BOUNDARIES-DOMAIN 16 (LBD16)is regulated by auxins and responds to nematode secretions. Two day post infection (dpi) LBD16::GUS (2500-bp version) Arabidopsis galls infected with Meloidogyne javanica were transferred to plates containing either 300 lM a-(phenyl ethyl-2-one)-indole-3-acetic acid (PEO-IAA) (Hayashi et al., 2008) in dimethyl sulfoxide (DMSO) or only DMSO, which was used as a control. GUS activity was examined 4 d later. (a, b) A clear signal was detected in both the lateral root primordial (LRP) (a) and the galls (b) in the DMSO treatment. (a1, b1) PEO-IAA treatment led to the disappearance of GUS signal in the LRP (a1) and in the center of the gall (b1). (c, c1, d, d1, e, e1) Representative protoplasts of the J0192>>GFP line incubated with buffer alone (c, c1), nematode (Meloidogyne incognita) secretions (d, d1) or the synthetic auxin 2,4-D (e, e1). Right panels are confocal images showing GFP (green) and chloroplast autofluorescence (red). The left panels show the corresponding transmission images. Bars: (a, a1, b, b1) 200 lm; (c, c1, d, d1, e, e1) 20 lm. For the protoplast analysis, three to four independent experiments were performed per treatment and a total of 200–300 protoplasts counted in each of the replicates per treatment. The proportion of active protoplasts was variable (25–50%) with either auxins or nematode secretions, but no GFP was ever detected in buffer-treated controls. For an overview image of protoplast with the different treatments, see Supporting Information Fig. S8.

New Phytologist (2014) 203: 632–645 Ó 2014 The Authors www.newphytologist.com New Phytologist Ó 2014 New Phytologist Trust New Phytologist Research 639

(a) (b)

Fig. 6 LATERAL ORGAN BOUNDARIES- DOMAIN 16 (LBD16) function is crucial for Arabidopsis gall development. (a) Meloidogyne javanica gall formation in the loss-of-function lbd16-1 mutant l showed a significant reduction in the nematode infection rate compared with Columbia (Col- (c) (d) 0). (b, c) Similar results were obtained with three independent lines of 35S::LBD16-SRDX (b) and pLBD16::LBD16-SRDX (c). (d) A significant reduction in the infection rate was also observed in two independent 35S:: LBD16-SRDX lines tested with Meloidogyne arenaria. Statistical analysis was performed with three independent experiments per line and at least 60 plants per experiment using ANOVA and the Schefee test; significant differences with Col-0: *, P < 0.05; values are means Æ SE. antagonist that inhibits the auxin signaling pathway by binding significant reduction in the number of infections per root to the SCFTIR1/AFBs ubiquitin–ligase complex (Hayashi et al., compared with Col-0 (Fig. 6a; 36%; P < 0.05). Three indepen- 2008). The expression of the LBD16 promoter in the young LRP dent 35S::LBD16-SRDX lines (Fig. 6b) containing the LBD16 (Fig. 5a, a1) and in 6-dpi galls (Fig. 5b, b1) was inhibited com- coding sequence fused to the transcriptional repressor domain pared with the control, indicating that its expression in galls is at SRDX driven by the 35S promoter also showed a strong reduc- least partially regulated by auxins. tion (42–52%). The same protein fusion driven by the native LBD16 promoter, which is active in the LR initial cells at the XPP, showed a reduction in the infection rate that was slightly LBD16 is activated by nematode secretions in protoplasts lower than that in lines with the 35S promoter, but significant Freshly isolated leaf protoplasts from the line J0192>>GFP were (22%; Fig. 6c; P < 0.05). Those successfully established nema- incubated with buffer, auxin or fractionated secretions extracted todes in 35S::LBD16-SRDX lines formed smaller galls with con- from Meloidogyne incognita juveniles. Representative protoplasts sistently less expanded GCs than in controls at 14 dpi (three- or for each treatment are shown in Fig. 5 (see Fig. S8 for an over- four-fold reduction in the mean total area occupied by the GCs; view and the Materials and Methods section). GFP signal was Fig. 7; P < 0.05; see the Materials and Methods section). There- clearly detected in protoplasts incubated with secretions in an fore, the function of LBD16 is crucial for proper gall formation auxin-free medium (Fig. 5d, d1), whereas no signal was ever and GC development. detected in the control with only buffer (Fig. 5c, c1). Protoplasts pLBD16::GUS and DR5::GUS plants infected with incubated in a medium containing a synthetic auxin (2,4-dichlo- M. arenaria mimicked the activation pattern shown during infec- rophenoxyacetic acid; Fig. 5e, e1) also showed clear signal. The tion with M. javanica (Fig. S6). The LBD16-SRDX lines most auxin inducibility of LBD16 is in accordance with previous severely affected by infection with M. javanica (Fig. 6) also reports (Okushima et al., 2007; Lee et al., 2009) and with our showed a significant decrease in the infection rate with results in planta (Fig. 4). These data provide a possible explana- M. arenaria (almost 54% compared with the wild type; Fig. 6d; tion for LBD16 induction in the pericycle cells when the nema- P < 0.05). Therefore, LBD16 function is conserved in galls tode is inside the root vascular cylinder, and are compatible with formed by both RKN species, in accordance with their similar the presence of an auxin maximum in the gall. expression patterns.

The role of LBD16 in the LR initial cells at the XPP is crucial LBD16 is not active in the feeding sites of CNs and does for proper gall formation and GC development not show a clear functional role during syncytium development Meloidogyne javanica infection tests with different LBD16 loss-of- function lines were carried out (Figs 6, S5). lbd16-1, a homozy- Syncytia induced by CNs are known to develop from pericycle gous insertion mutant line that was characterized for LR cells (Golinowski et al., 1996; Sobczak et al., 1997) and auxins formation (SALK_095791; Okushima et al., 2007), exhibited a have also been shown to play a role during CN infection

Ó 2014 The Authors New Phytologist (2014) 203: 632–645 New Phytologist Ó 2014 New Phytologist Trust www.newphytologist.com New 640 Research Phytologist

(a) (c)

Fig. 7 LATERAL ORGAN BOUNDARIES- DOMAIN 16 (LBD16) function is crucial for the development of giant cells (GCs) induced by Meloidogyne javanica in Arabidopsis. (a, (b) (d) b) Toluidine staining of semithin (2-lm) sections of galls from Columbia (Col-0) (a) and the 35S::LBD16-SRDX line (b) shown at the same magnification. Asterisks, giant cells (GCs); N, nematode. Bars, 50 lm. (c) Graph of the average GC area from representative galls in both lines; N = 20 sections of 2 lm from each of the lines tested. (d) Gall width. Bars represent the average width of galls per line, as indicated, from three independent experiments (N = 15). Significant differences: *, P < 0.05; values are means Æ SE.

(Goverse & Bird, 2011). Therefore, we analyzed the activation (a) (b) pattern of LBD16 and its putative function during infection with H. schachtii and syncytium establishment. The 2500-bp LBD16:: GUS line did not stain positively for GUS inside the feeding cells at any of the stages of syncytium development (Fig. 8a), nor dur- ing syncytium expansion (6 dpi; Fig. 8b), or in a well-established syncytium (11 dpi; Fig. 8c). However, this line showed clear GUS staining in the LRP in the same roots, as in the noninfected control roots (see blue signal in Fig. 4m) (Fig. 8a,b arrows). In (c) (d) agreement with this result, three independent 35S::LBD16-SRDX lines showed no differences in the infection rate compared with the wild type (Fig. 8d). Thus, the lines with the most severe phe- notype during RKN infection (Fig. 6) did not show any signifi- cant effect on CNs, suggesting that molecular components other than LBD16, either those from the pericyle or those involved in LR formation, may be important for syncytium formation. Fig. 8 LATERAL ORGAN BOUNDARIES-DOMAIN 16 (LBD16) expression and functional analysis do not support a role for LBD16 in syncytium Genetic ablation experiments reinforce the importance of development in Arabidopsis. (a–c) The LBD16::GUS line showed no signal molecular components in the XPP for gall and GC in syncytia formed by Heterodera schachtii at 3, 6 or 11 d post infection development (dpi; arrows). Only the lateral root primordial (LRP) in (a) and (b) showed GUS staining (arrows). LR, lateral root; N, nematode; Sy, syncytia. (d) To further investigate the importance of molecular compo- Infection tests of three independent 35S:LBD16-SRDX lines showed no nents in the XPP during gall formation, the J0121>>GFP line significant effect on the infection rate (number of syncytia per main root and length) relative to Columbia (Col-0). This result was based on scores was used in genetic ablation experiments using the diphtheria of syncytia formed over the length of the main root, and values are toxin chain A (DTA) gene. F1 plants from crosses between means Æ SE. Statistical analysis was performed with three independent two homozygous lines, J0121>>GFP and UAS-DTA experiments per line; at least 50 plants per experiment were tested. (J0121>>DTA), were assayed for infection with M. javanica. ANOVA and the Schefee test were performed (P < 0.05). Bars, 200 lm. These F1 plants are unable to form LRs, confirming that the XPP cells are necessary for LRP development (Laplaze et al., 2005). The results showed a drastic decrease (> 80%) in gall galls formed in the J0121>>DTA line compared with the con- formation in the J0121>>DTA plants (Fig. 9b), and 70% of trol (Figs 7d, 9d–f; see the Materials and Methods section). the plants failed to host any infection (Fig. 9c) as compared These results confirmed the importance of molecular compo- with the J0121>>GFP control line. Moreover, a major reduc- nents in the XPP, where LBD16 is activated, during nematode tion in galls and GC size (> 50%) was observed in the few establishment in gall and GC development.

New Phytologist (2014) 203: 632–645 Ó 2014 The Authors www.newphytologist.com New Phytologist Ó 2014 New Phytologist Trust New Phytologist Research 641

(a) (c)

Fig. 9 Genetic ablation of xylem pericycle (b) pole (XPP) cells interferes with gall and giant cell (GC) development induced by Meloidogyne javanica in Arabidopsis. F1 plants from crosses between two homozygous lines, J0121>>GFP and a UAS- DTA line (J0121>>DTA), which express the diphtheria toxin chain A in the domain of the tagged LATERAL ORGAN BOUNDARIES- DOMAIN 16 (LBD16) gene, were used. (a) Seedlings of the J0121>>GFP control line and the J0121>>DTA line were inoculated 5 d after germination when they had identical root phenotypes. (b) Percentage of (d) galls per main root in control J0121>>GFP and J0121>>DTA lines. A significant > reduction in infection ( 80%) was observed. (f) Statistical analysis was performed with three independent experiments per line (with a total of at least 217 plants per line) using ANOVA and the Schefee test; significant differences: *, P < 0.05; values are means Æ SE. (c) The percentage of plants with more than one gall was also considerably lower in J0121>>DTA compared with J0121>>GFP. (d, e) Semithin (e) sections (4 lm) from 14 d post infection (dpi) galls stained with toluidine blue in control J0121>>GFP (d) and ablated J0121>>DTA (e) lines; asterisks, GCs; N, nematode. Bars, 50 lm. (f) In a comparison of the two lines, the average GC area showed a significant reduction of > 50% in the J0121>>DTA line compared with the control J0121>>GFP (P < 0.05); N = 20 sections of 2 lm from each of the lines tested; significant differences: *, P < 0.05; values are means Æ SE.

Discussion importance of specific gene expression in the XPP for nematode feeding site formation, and identify one likely molecular compo- Plant endoparasitic nematodes transform host root cells into their nent: the transcription factor, LBD16, involved in both LR for- feeding sites (Gheysen & Mitchum, 2009). Heterodera schachtii mation and gall development following auxin responses. nematodes select a single procambial or pericycle cell for initiation of their syncytia in Arabidopsis roots (Golinowski et al., 1996; Galls and GCs show transcriptional similarities to Sobczak et al., 1997). No such certainty exists for the ontogeny of undifferentiated root cell types RKN galls and GCs, although root pericycle and cortical cells pro- liferate around the nematode, contributing to the formation of the We found that the transcriptomes of 3-dpi galls and GCs were galls (Berg et al., 2009). Division of pericycle cells is necessary for enriched in genes characteristic of two root cell types: LR initials LR formation, where LBD16 participates in the auxin signaling and quiescent center cells (Fig. 1). These genes have varied func- cascade leading to the division of specific XPP cells to form the tions, such as cell cycle regulation and cytoskeleton or cell wall new organ (Goh et al., 2012). We address the expression pattern remodeling (Table S1). Some encode transcription factors, such and the functional role of LBD16 in galls. Our results confirm the as HSFB4, which are characteristic of the quiescent center

Ó 2014 The Authors New Phytologist (2014) 203: 632–645 New Phytologist Ó 2014 New Phytologist Trust www.newphytologist.com New 642 Research Phytologist

transcriptome (Table S1). Other genes related to cell wall relaxa- (Peret et al., 2009) also divide during early gall formation, and tion, such as the expansin AtEXPA6 induced in galls and GCs in are consistent with our in silico analysis, which indicated enrich- transcriptomic analyses and by q-PCR (Barcala et al., 2010), are ment in LR initial cell genes in gall and GC transcriptomes also characteristic of the LRP (Table S1). (Fig. 1). Thus, we confirmed that there is a high molecular simi- The resistance to M. incognita of the auxin-insensitive tomato larity between gall formation and LR formation, except for the mutant dgt (Richardson & Price, 1982), which lacks LRs, as well subtle differences already mentioned. as the similarities in the gene expression of plant transcription fac- The importance of XPP-specific genes during infection was tors (e.g. Medicago truncatula homologues to PHANTASTICA also shown in the genetic ablation experiments using the DTA and Class I knotted, Mt-phan and Mt-knox-1, respectively) and gene (which inhibits protein synthesis) driven by the enhancer cell cycle regulators during the development of galls, LRs and nod- trap line J0121. The J0121>>DTA line showed a clear reduction ules, point to a correlation between these processes (Goverse et al., in infection rates, and most of the plants were unable to host any 2000; Mathesius, 2003; Moreno-Risueno et al., 2010). Addition- galls (70%). Additionally, the development of galls and GCs was ally, 39 out of 103 promoter tag lines displaying a distinct clearly impaired (Figs 7, 9). However, other transduction path- response to nematode infection also exhibited activity at LR initia- ways or redundant molecular components may also be involved tion sites (Barthels et al., 1997). Furthermore, the characteristic in gall and GC development because some galls were still formed genes expressed in LR initials were overrepresented in GCs and when XPP cells were severely compromised. galls in a comparative in silico transcriptomic analysis based on sta- A novel view of cell reprogramming has been reported by tistics (Fig. 1). Taken together, these data reinforce the similarity Sugimoto et al. (2010), showing that calli form from the differen- between the two processes. Molecular confirmation of this parallel tiation of pericycle-like cells with meristematic features (large is inferred from the analysis described in the next sections. nuclei, small vacuoles and dense cytoplasm) present in the organ. This occurs through the ectopic activation of an LR developmen- tal program that produced root meristem-like tissue. Whether The specific activation of two XPP marker lines during RKN gall development is also related to this process of callus formation infection connects LR initial cell identity with gall and GC through the activation of an LR initiation-like program remains formation to be elucidated. Interestingly, LBD transcription factors are key Two XPP marker lines, J0121 and J0192, showed strong and dis- regulators in the callus induction process (Fan et al., 2012) as tinct GFP expression in the galls formed by M. javanica.In their ectopic expression triggers spontaneous callus formation but J0192, a marker induced specifically in those XPP cells compe- their suppression inhibits this process (Fan et al., 2012). More- tent to become LR founder cells, the pericycle on both sides of over, the formation of a highly organized LR or unorganized the vascular cylinder often started to express GFP when the nema- callus depends on LBD abundance (Fan et al., 2012). tode reached the vascular cylinder during migration and establish- Our findings on the importance of LBD16 in early gall devel- ment (Figs 2, 3, S3, S4). Yet in LRs, the J0192>>GFP signal was opment are consistent with its role in LR and callus development. distributed in XPP cell rows only on one side of the vascular cyl- LBD16 is specifically expressed in the LRP only during the early inder and in an alternate manner (Fig. 3; Moreno-Risueno et al., stages of development and is regulated by auxins (Laplaze et al., 2010). In both marker lines, the signal progressed inwards in galls 2005). In plants infected with M. javanica and M. arenaria, and outwards in LRs, and its position(s) with respect to the vascu- LBD16 expression was detected very early after infection, 1 dpi lar cylinder (frequently on two sides in galls and only on one side up to 11–15 dpi (Figs 2–4, S1, S2, S6) and was regulated by aux- in LRs) differed between the two processes (Laplaze et al., 2005; ins in galls, as shown by its inhibition by PEO-IAA, similar to Figs 2, 3). Abnormal irregular division planes were frequently LRP (Fig. 5; Okushima et al., 2007; Lee et al., 2009). Although observed inside the labeled cells of the gall vascular cylinder its expression at early stages appeared to correlate with the pres- (close-up in Fig. 3). By contrast, in LRs anticlinal divisions ence of auxins in the same cell types, as shown by DR5::GUS,at occurred initially, followed by periclinal divisions that formed the later stages the mere presence of auxins in the gall was not LRP (Fig. 3; Lavenus et al., 2013). Eventually, the entire vascular sufficient to activate LBD16 expression (Figs 4, S6). Additionally, cylinder inside the gall was labeled, including the GCs (Fig. 2). the presence of auxin-response elements ‘ARF binding sites’ The infection tests performed with the LBD16 mutant lines in the LBD16 promoter sequence correlated with the intensity of (the gene that drives GFP expression in J0192) confirm a role for the GUS signal (Fig. S5). However, both promoter versions were LBD16 in the dividing XPP during gall development (Figs 6, 7). active only during the early stages of infection. These results may A significant reduction (P < 0.05) of at least 20% in gall forma- indicate the necessity for a threshold level of auxins in the gall to tion relative to the wild-type controls was observed in the three allow LBD16 expression, which would mimic the scenario that independent approaches used to investigate its role. In addition, occurs during the first divisions of LR development. However, the GUS expression pattern of the marker line ProCycB1;1:CycB1;1 the absence of GUS signal in the LBD16::GUS syncytia (where (NT)-GUS, active only during the G2/M transition, mimicked DR5 was highly activated; Karczmarek et al., 2004) suggests that that of J0192 in XPP cells during early nematode establishment other signals apart from auxins could be contributing to early and in most cells inside the vascular cylinder of the gall as the activation and/or late repression of LBD16 in galls and GCs, and infection progressed (Figs 2, 3). These results suggest that the that different molecular components in pericycle cells may partic- founder cells contained in the XPP that divide to form a new LR ipate in the formation of syncytia as compared with galls.

New Phytologist (2014) 203: 632–645 Ó 2014 The Authors www.newphytologist.com New Phytologist Ó 2014 New Phytologist Trust New Phytologist Research 643

The involvement of auxins in nematode feeding site develop- cascades leading to LBD16 expression through other molecular ment is broadly accepted (Goverse & Bird, 2011). Here, we show components has yet to be elucidated. However, the activation of that the auxin sensor DR5 (fused to either GUS or GFP) is active individual protoplasts suggests that nematode secretions function until the late stages (45 dpi) of M. javanica and M. arenaria infec- in a cell-autonomous manner to activate LBD16 (Figs 5, S8). In tion (Figs 4, S6). This suggests a putative role for auxins not only this context, it is possible that nematode secretions directly or indi- during the initial infection stages (nematode establishment) but rectly activate XPP cells to start proliferation, as they appear on also in feeding site maintenance. Some genes induced in nema- both sides of the vascular cylinder (Fig. 3). Support for this comes tode feeding sites until late infection stages have auxin response from the irregular cell divisions observed in developing galls of the cis elements in their promoters (Wang et al., 2007; Karczmarek J0192>>GFP line (Fig. 3). One of the best characterized secreted et al., 2008; Wieczorek et al., 2008; Swiecicka et al., 2009). Most peptides from RKNs, 16D10, showed a high similarity with the of the relevant studies related to the role of auxin in nematode C-terminal conserved motif of the plant CLAVATA3/ESR (CLE) feeding site formation were performed using CNs and were protein family (Huang et al., 2006; Teillet et al., 2013). focused on alterations in auxin transport (Goverse & Bird, Overexpression of CLE peptides in Arabidopsis caused strikingly 2011). However, auxin transduction pathways that trigger the irregular anticlinal divisions (symmetric and asymmetric) in the activation of nematode-induced genes have not been well studied. pericycle (Meng et al., 2012). Additionally, the establishment of One of the few exceptions is a member of the transcription factor asymmetry in Arabidopsis LR founder cells is regulated by LBD16 family WRKY (WRKY23). The response of WRKY23 to auxins is and other LBD proteins (Goh et al., 2012). Thus, nematodes controlled by the Aux/IAA protein SOLITARY-ROOT (SLR/ might use similar mechanisms for gall formation. IAA14) (Grunewald et al., 2008). It is induced by both RKN and In conclusion, we have demonstrated the important role in gall CN, and loss-of-function mutants were more resistant to infec- and GC formation of a molecular component specific for the tion by CNs (Grunewald et al., 2009). However, other plant- or XPP LR initial cells, that is, LBD16. As we have shown that nematode-derived signals could participate in its activation RKN feeding sites share a clear common transducer with LR for- (Grunewald et al., 2009). This would partially agree with the mation, this supports the view that RKN may at least partially activation of LBD16 by nematode secretions (see next section). ‘hijack’ plant transduction pathways leading to LR formation. However, contrary to what was observed for WRKY23, LBD16 is LBD16 showed a specific function during RKN establishment not activated by CNs, and CN infection was not affected in and not during the establishment of CNs, was regulated by aux- LBD16 loss-of-function lines (Fig. 8). This result agrees with ins in galls and was induced by nematode secretions. Finally, we transcriptional data from microaspirated H. schachtii syncytia, described peculiarities of cell reprogramming during gall forma- showing that LBD16 was down-regulated (Szakasits et al., 2009). tion similar to those occurring during LR and callus formation, As mentioned in the introduction, the pericycle is a key cell type where LBDs play a major role in the control of both for the formation of the syncytium, reinforcing the idea that developmental programs. molecular processes occurring in the pericycle during cyst nema- tode infection are not mediated by the same molecular compo- Acknowledgements nents as in galls, for example LBD16. We thank Lucıa Munoz~ for her technical assistance, Ana Belen Yuste for Fig. S3 and Dr Laplaze for the J0192, J0121>>DTA Nematode secretions activate LBD16 in protoplasts and J0121 seeds. We also thank Dr Janice de Almeida and Dr Gil- We investigated the possibility that nematode secretions could bert Engler for their confocal microscopy training. This work was trigger changes in pericycle cells by inducing the expression of supported by grants from the Spanish Government to C.E. LBD16, as XPP cells near the nematode head were frequently acti- (AGL2010-17388) and C.F. (CSD2007-057) and the Castilla-La vated (Fig. 3). We demonstrated that, similarly to exogenous aux- Mancha Government to C.F. and C.E. (PCI08-0074-0294). ins, secretions from M. incognita juveniles were able to induce LBD16 expression in Arabidopsis leaf protoplasts (Figs 5, S8). References Therefore, the presence of the nematode itself could cause a local increase in auxin concentrations in the XPP. Auxin compounds Absmanner B, Stadler R, Hammes UZ. 2013. Phloem development in have been identified in nematode secretions (De Meutter et al., nematode-induced feeding sites: the implications of auxin and cytokinin. Frontiers in Plant Science 4: 241. 2005; Goverse & Bird, 2011). Alternatively, infective nematodes Almeida-Engler J, Favery B. 2011. The plant cytoskeleton remodelling in may manipulate auxin homeostasis in an indirect way by releasing nematode induced feeding sites. In: Jones J, Gheysen G, Fenoll C, eds. enzymes that interfere with local active auxin concentrations in Genomics and molecular genetics of plant–nematode interactions. Dordrecht, the plants. In this respect, functional analysis of a CN gene encoding a Netherlands: Springer, 369–394. secreted chorismate mutase, an enzyme involved in the conversion Atkinson HJ, Lilley CJ, Urwin PE. 2012. Strategies for transgenic nematode control in developed and developing world crops. Current Opinion in of the auxin precursor tryptophan, suggests that nematodes may Biotechnology 23: 251–256. be able to redirect the biosynthesis of auxins in planta (Doyle & Barcala M, Garcıa A, Cabrera J, Casson S, Lindsey K, Favery B, Garcıa-Casado Lambert, 2003; Goverse & Bird, 2011). G, Solano R, Fenoll C, Escobar C. 2010. Early transcriptomic events in Whether LBD16 induction by nematode secretions is triggered microdissected Arabidopsis nematode-induced giant cells. Plant Journal 61: – directly by auxins or is caused by the activation of signaling 698 712.

Ó 2014 The Authors New Phytologist (2014) 203: 632–645 New Phytologist Ó 2014 New Phytologist Trust www.newphytologist.com New 644 Research Phytologist

Barthels N, van der Lee FM, Klap J, Goddijn OJ, Karimi M, Puzio P, Grundler Grunewald W, van Noorden G, Van Isterdael G, Beeckman T, Gheysen FM, Ohl SA, Lindsey K, Robertson L et al. 1997. Regulatory sequences of G, Mathesius U. 2009. Manipulation of auxin transport in plant roots Arabidopsis drive reporter gene expression in nematode feeding structures. during Rhizobium symbiosis and nematode parasitism. Plant Cell 21: Plant Cell 9: 2119–2134. 2553–2562. Bell EM, Lin WC, Husbands AY, Yu L, Jaganatha V, Jablonska B, Mangeon A, Hayashi K, Tan X, Zheng N, Hatate T, Kimura Y, Kepinski S, Nozaki H. 2008. Neff MM, Girke T, Springer PS. 2012. Arabidopsis lateral organ boundaries Small-molecule agonists and antagonists of F-box protein-substrate interactions negatively regulates brassinosteroid accumulation to limit growth in organ in auxin perception and signaling. Proceedings of the National Academy of boundaries. Proceedings of the National Academy of Sciences, USA 109: Sciences, USA 105: 5632–5637. 21146–21151. Huang G, Dong R, Allen R, Davis EL, Baum TJ, Hussey RS. 2006. A root-knot Berg RH, Fester T, Taylor CG. 2009. Development of the root-knot nematode nematode secretory peptide functions as a ligand for a plant transcription feeding cell. In: Berg RH, Taylor CG, eds. Cell biology of plant nematode factor. Molecular Plant-Microbe Interactions 19: 463–470. parasitism. Heidelberg, Germany: Springer, 115–152. Hutangura P, Mathesius U, Jones MGK, Rolfe BG. 1999. Auxin induction is a Bird D, Opperman CH, Williamson VM. 2009. Plant infection by root-knot trigger for root gall formation caused by root-knot nematodes in white clover nematode. In: Berg RH, Taylor CG, eds. Cell biology of plant nematode and is associated with the activation of the flavonoid pathway. Functional Plant parasitism. Heidelberg, Germany: Springer, 1–14. Biology 26: 221–231. Brady SM, Orlando DA, Lee J-Y, Wang JY, Koch J, Dinneny JR, Mace D, Karczmarek A, Fudali S, Lichocka M, Sobczak M, Kurek W, Janakowski S, Ohler U, Benfey PN. 2007. A high-resolution root spatiotemporal map reveals Roosien J, Golinowski W, Bakker J, Goverse A et al. 2008. Expression of two dominant expression patterns. Science 318: 801–806. functionally distinct plant endo-beta-1,4-glucanases is essential for the Caillaud MC, Dubreuil G, Quentin M, Perfus-Barbeoch L, Lecomte P, de compatible interaction between potato cyst nematode and its hosts. Molecular Almeida Engler J, Abad P, Rosso MN, Favery B. 2008. Root-knot nematodes Plant-Microbe Interactions 21: 791–798. manipulate plant cell functions during a compatible interaction. Journal of Karczmarek A, Overmars H, Helder J, Goverse A. 2004. Feeding cell Plant Physiology 165: 104–113. development by cyst and root-knot nematodes involves a similar early, local Cardona A, Saalfeld S, Schindelin J, Arganda-Carreras I, Preibisch S, Longair and transient activation of a specific auxin-inducible promoter element. M, Tomancak P, Hartenstein V, Douglas RJ. 2012. TrakEM2 software for Molecular Plant Pathology 5: 343–346. neural circuit reconstruction. PLoS ONE 7: e38011. Laplaze L, Parizot B, Baker A, Ricaud L, Martiniere A, Auguy F, Franche C, Colon-Carmona A, You R, Haimovitch-Gal T, Doerner P. 1999. Nussaume L, Bogusz D, Haseloff J. 2005. GAL4-GFP enhancer trap lines for Spatio-temporal analysis of mitotic activity with a labile cyclin–GUS fusion genetic manipulation of lateral root development in Arabidopsis thaliana. protein. Plant Journal 20: 503–508. Journal of Experimental Botany 56: 2433–2442. De Meutter J, Tytgat T, Prinsen E, Gheysen G, Van Onckelen H, Gheysen G. Lavenus J, Goh T, Roberts I, Guyomarc HS, Lucas M, De Smet I, 2005. Production of auxin and related compounds by the plant parasitic Fukaki H, Beeckman T, Bennett M, Laplaze L. 2013. Lateral root nematodes Heterodera schachtii and Meloidogyne incognita. Communications in development in Arabidopsis: fifty shades of auxin. Trends in Plant Science Agricultural and Applied Biological Sciences 70:51–60. 18: 450–458. Doyle EA, Lambert KN. 2003. Meloidogyne javanica chorismate mutase 1 Lee HW, Kim NY, Lee DJ, Kim J. 2009. LBD18/ASL20 regulates lateral root alters plant cell development. Molecular Plant-Microbe Interactions 16: formation in combination with LBD16/ASL18 downstream of ARF7 and 123–131. ARF19 in Arabidopsis. Plant Physiology 151: 1377–1389. Escobar C, Barcala M, Portillo M, Almoguera C, Jordano J, Fenoll C. 2003. Mathesius U. 2003. Conservation and divergence of signalling pathways between Induction of the Hahsp17.7G4 promoter by root-knot nematodes: roots and soil microbes – the Rhizobium–legume symbiosis compared to the involvement of heat-dhock rlements in promoter sctivity in giant cells. development of lateral roots, mycorrhizal interactions and nematode-induced Molecular Plant-Microbe Interactions 16: 1062–1068. galls. Plant and Soil 255: 105–119. Fan M, Xu C, Xu K, Hu Y. 2012. LATERAL ORGAN BOUNDARIES Meng L, Buchanan BB, Feldman LJ, Luan S. 2012. CLE-like (CLEL) peptides DOMAIN transcription factors direct callus formation in Arabidopsis control the pattern of root growth and lateral root development in regeneration. Cell Research 22: 1169–1180. Arabidopsis. Proceedings of the National Academy of Sciences, USA 109: Gamborg OL, Miller RA, Ojima K. 1968. Nutrient requirements of 1760–1765. suspension cultures of soybean root cells. Experimental Cell Research 50: Moreno-Risueno MA, Van Norman JM, Moreno A, Zhang J, Ahnert SE, 151–158. Benfey PN. 2010. Oscillating gene expression determines competence for Gheysen G, Mitchum MG. 2009. Molecular insights in the susceptible plant periodic Arabidopsis root branching. Science 329: 1306–1311. response to nematode infection. Development of the root-knot nematode Okushima Y, Fukaki H, Onoda M, Theologis A, Tasaka M. 2007. ARF7 and feeding cell. In: Berg RH, Taylor CG, eds. Cell biology of plant nematode ARF19 regulate lateral root formation via direct activation of LBD/ASL genes parasitism. Heidelberg, Germany: Springer, 45–82. in Arabidopsis. Plant Cell 19: 118–130. Goh T, Joi S, Mimura T, Fukaki H. 2012. The establishment of asymmetry in Peret B, Larrieu A, Bennett MJ. 2009. Lateral root emergence: a difficult birth. Arabidopsis lateral root founder cells is regulated by LBD16/ASL18 and related Journal of Experimental Botany 60: 3637–3643. LBD/ASL proteins. Development 139: 883–893. Perry RN, Moens M. 2011. Introduction to plant–parasitic nematodes; modes of Golinowski W, Grundler FMW, Sobczak M. 1996. Changes in parasitism. In: Jones J, Gheysen G, Fenoll C, eds. Genomics and molecular the structure of Arabidopsis thaliana during female development of genetics of plant–nematode interactions. Dordrecht, the Netherlands: Springer, the plant-parasitic nematode Heterodera schachtii. Protoplasma 194: 3–20. 103–116. Portillo M, Cabrera J, Lindsey K, Topping J, Andres MF, Emiliozzi M, Oliveros Goverse A, Bird D. 2011. The role of plant hormones in nematode feeding cell JC, Garcia-Casado G, Solano R, Koltai H et al. 2013. Distinct and conserved formation. In: Jones J, Gheysen G, Fenoll C, eds. Genomics and molecular transcriptomic changes during nematode-induced giant cell development in genetics of plant–nematode interactions. Dordrecht, the Netherlands: tomato compared with Arabidopsis: a functional role for gene repression. New Springer, 325–348. Phytologist 197: 1276–1290. Goverse A, Overmars H, Engelbertink J, Schots A, Bakker J, Helder J. 2000. Portillo M, Lindsey K, Casson S, GarcIA-Casado G, Solano R, Fenoll C, Both induction and morphogenesis of cyst nematode feeding cells are mediated Escobar C. 2009. Isolation of RNA from laser-capture-microdissected giant by auxin. Molecular Plant-Microbe Interactions 13: 1121–1129. cells at early differentiation stages suitable for differential transcriptome Grunewald W, Karimi M, Wieczorek K, Van de Cappelle E, Wischnitzki E, analysis. Molecular Plant Pathology 10: 523–535. Grundler F, Inze D, Beeckman T, Gheysen G. 2008. A role for AtWRKY23 Quentin M, Abad P, Favery B. 2013. Plant parasitic nematode effectors target in feeding site establishment of plant-parasitic nematodes. Plant Physiology 148: host defense and nuclear functions to establish feeding cells. Frontiers in Plant 358–368. Science 4: 53.

New Phytologist (2014) 203: 632–645 Ó 2014 The Authors www.newphytologist.com New Phytologist Ó 2014 New Phytologist Trust New Phytologist Research 645

Richardson L, Price NS. 1982. Host–parasite relationships of Meloidogyne Supporting Information incognita and the dia-geotropica tomato mutant. Journal of Nematology 14: 465–466. Additional supporting information may be found in the online Rosso M-N, Favery B, Piotte C, Arthaud L, De Boer JM, Hussey RS, Bakker J, version of this article. Baum TJ, Abad P. 1999. Isolation of a cDNA encoding a b-1,4-endoglucanase in the root-knot nematode Meloidogyne incognita and expression analysis during plant parasitism. Molecular Plant-Microbe Interactions 12: 585–591. Fig. S1 Genotyping the lbd16-1 (SALK_095791) insertion line de Ruijter NCA, Verhees J, van Leeuwen W, van der Krol AR. 2003. Evaluation to identify homozygous plants. and comparison of the GUS, LUC and GFP reporter system for gene expression studies in plants. Plant Biology 5: 103–115. Fig. S2 Infection test design for LBD16 loss-of-function lines Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, and their root phenotypes. Preibisch S, Rueden C, Saalfeld S, Schmid B et al. 2012. Fiji: an open-source platform for biological-image analysis. Nature Methods 9: 676–682. Sheen J. 2002. A transient expression assay using Arabidopsis mesophyll protoplasts. Fig. S3 Representative in vivo time-courses of individual galls [WWW document] URL http://molbio.mgh.harvard.edu/sheenweb/protocols/ from lines J0121 and J0192 sequentially monitored. AtprotoRL04.pdf [accessed 5 February 2011]. Sobczak M, Golinowski W. 2009. Structure of cyst nematode feeding sites. In: Fig. S4 All confocal channel photographs (transmission, red for Berg RH, Taylor CG, eds. Cell biology of plant nematode parasitism. Heidelberg, Germany: Springer, 153–188. PI, green for GFP) corresponding to the representative samples Sobczak M, Golinowski W, Grundler FW. 1997. Changes in the structure of shown in Fig. 2. Arabidopsis thaliana roots induced during development of males of the plant parasitic nematode Heterodera schachtii. European Journal of Plant Pathology Fig. S5 GUS activity of LBD16::GUS lines carrying two different – 103: 113 124. versions of the LBD16 promoter infected with M. javanica. Sugimoto K, Jiao Y, Meyerowitz EM. 2010. Arabidopsis regeneration from multiple tissues occurs via a root development pathway. Developmental Cell 18: 463–471. Fig. S6 Activation patterns of DR5 and LBD16 after M. arenaria Swiecicka M, Filipecki M, Lont D, Van Vliet J, Qin L, Goverse A, Bakker J, infection. Helder J. 2009. Dynamics in the tomato root transcriptome on infection with the potato cyst nematode Globodera rostochiensis. Molecular Plant Pathology 10: Fig. S7 Time-course of the auxin sensors DR5::GFP and DR5:: 487–500. Szakasits D, Heinen P, Wieczorek K, Hofmann J, Wagner F, Kreil DP, Sykacek GUS throughout the M. javanica life cycle. P, Grundler FM, Bohlmann H. 2009. The transcriptome of syncytia induced by the cyst nematode Heterodera schachtii in Arabidopsis roots. Plant Journal Fig. S8 Overview image of protoplasts of the J0192>>GFP line 57: 771–784. incubated with buffer and secretions and protein content in Teillet A, Dybal K, Kerry BR, Miller AJ, Curtis RHC, Hedden P. 2013. M. incognita secretion fractions. Transcriptional changes of the root-knot nematode Meloidogyne incognita in response to Arabidopsis thaliana root signals. PLoS ONE 8: e61259. Ulmasov T, Murfett J, Hagen G, Guilfoyle TJ. 1997. Aux/IAA proteins repress Table S1 List of genes from the differential transcriptomes of expression of reporter genes containing natural and highly active synthetic galls and GCs compared with those characteristic of each root cell auxin response elements. Plant Cell 9: 1963–1971. type Wang X, Replogle AMY, Davis EL, Mitchum MG. 2007. The tobacco Cel7 gene promoter is auxin-responsive and locally induced in nematode feeding sites of heterologous plants. Molecular Plant Pathology 8: 423–436. Please note: Wiley Blackwell are not responsible for the content Wieczorek K, Hofmann J, Blochl A, Szakasits D, Bohlmann H, Grundler or functionality of any supporting information supplied by the FM. 2008. Arabidopsis endo-1,4-beta-glucanases are involved in the authors. Any queries (other than missing material) should be formation of root syncytia induced by Heterodera schachtii. Plant Journal directed to the New Phytologist Central Office. 53: 336–351.

Ó 2014 The Authors New Phytologist (2014) 203: 632–645 New Phytologist Ó 2014 New Phytologist Trust www.newphytologist.com This article was downloaded by: [Universidad Castilla La Mancha] On: 23 April 2015, At: 07:42 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Plant Signaling & Behavior Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/kpsb20 Genes co-regulated with LBD16 in nematode feeding sites inferred from in silico analysis show similarities to regulatory circuits mediated by the auxin/cytokinin balance in Arabidopsis Javier Cabreraa, Carmen Fenolla & Carolina Escobara a Facultad de Ciencias Ambientales y Bioquímica; Universidad de Castilla-La Mancha; Toledo, Spain Accepted author version posted online: 09 Feb 2015.

Click for updates

To cite this article: Javier Cabrera, Carmen Fenoll & Carolina Escobar (2015) Genes co-regulated with LBD16 in nematode feeding sites inferred from in silico analysis show similarities to regulatory circuits mediated by the auxin/cytokinin balance in Arabidopsis, Plant Signaling & Behavior, 10:3, e990825, DOI: 10.4161/15592324.2014.990825 To link to this article: http://dx.doi.org/10.4161/15592324.2014.990825

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions SHORT COMMUNICATION Plant Signaling & Behavior 10:3, e990825; March 2015; © 2015 Taylor & Francis Group, LLC Genes co-regulated with LBD16 in nematode feeding sites inferred from in silico analysis show similarities to regulatory circuits mediated by the auxin/cytokinin balance in Arabidopsis

Javier Cabrera, Carmen Fenoll, and Carolina Escobar*

Facultad de Ciencias Ambientales y Bioquímica; Universidad de Castilla-La Mancha; Toledo, Spain

Keywords: Arabidopsis, auxins, cytokinins, callus, cyst-nematodes, giant cells (GCs), lateral organ boundaries-domain 16 (LBD16), lateral root, root-knot nematodes (RKNs), syncytia

Plant endoparasitic nematodes, root-knot and cyst nematodes (RKNs and CNs) induce within the root vascular cylinder transfer cells used for nourishing, termed giant cells (GCs) and syncytia. Understanding the molecular mechanisms behind this process is essential to develop tools for nematode control. Based on the crucial role in gall development of LBD16, also a key component of the auxin pathway leading to the divisions in the xylem pole pericycle during lateral root (LR) formation, we investigated genes co-regulated with LBD16 in different transcriptomes and analyzed their similarities and differences with those of RKNs and CNs feeding sites (FS). This analysis confirmed LBD16 and its co-regulated genes, integrated in signaling cascades mediated by auxins during LR and callus formation, as a particular feature of RKN-FS distinct to CNs. However, LBD16, and its positively co-regulated genes, were repressed in syncytia, suggesting a selective down- regulation of the LBD16 auxin mediated pathways in CNs-FS. Interestingly, cytokinin-induced genes are enriched in syncytia and we encountered similarities between the transcriptome of shoot regeneration from callus, modulated by cytokinins, and that of syncytia. These findings establish differences in the regulatory networks leading to both FS formation, probably modulated by the auxin/cytokinin balance.

Plant endoparasitic nematodes (PPNs), root-knot and cyst root (LR) formation.8 We found that Arabidopsis XPP marker nematodes (RKNs and CNs) induce their feeding cells from vas- lines (J0121 and J0192; both induced specifically in the xylem cular cell precursors within the root. The feeding cells from pole pericycle of roots9) were activated in early developing galls/ RNKs (Meloidogyne spp.), called giant cells (GCs), are abnor- GCs. The activation pattern of pLBD16::GUS resembled that of mally large cells induced by nematode effectors in the vascular the G2/M-transition specific ProCycB1;1:CycB1;1(NT)-GUS line cylinder that undergo repeated mitosis with aborted cytokine- and DR5::GUS in early infection stages. LBD16 was regulated by sis,1-3 transforming them into transfer-like cells that the nema- auxins not only in LR, but also in galls, and induced by RKN tode uses as sinks of nutrients from the plant.4 Additionally, secretions in protoplasts. Infection tests of several independent vascular cells around the GCs divide and increase in number, LBD16 loss-of-function lines showed a significant reduction in 7 Downloaded by [Universidad Castilla La Mancha] at 07:42 23 April 2015 and cortex cells become hypertrophic forming the typical galls or the infection rate of M. javanica and M. arenaria. The promoter knots in the roots associated to the infection of these nematodes. of LBD16 was not induced within the syncytia formed by H. In the case of CNs (Heterodera spp and Globodera spp.), a syncy- schachtii in Arabidopsis and, accordingly, the most severe loss of tium is formed inside the roots by gradual incorporation of function line did not show an evident effect during the infection neighboring vascular cells.5 The development of the PPNs inside with CNs.7 The specific activation of 2 XPP marker lines and the the root is entirely dependent on the development of the GCs or function of LBD16 during RKN infection connect LR initial syncytia. Therefore, knowledge of the molecular mechanisms cells identity with gall and GC formation. behind this process is required to develop biotechnological tools To further investigate the putative role of LBD16 in Arabi- against these plagues that cause high economic losses around the dopsis galls and GCs, we have here identified the 50 genes with world.6 the highest positive, and the 50 genes with the highest negative We have previously addressed the role during gall develop- correlation values (according to their Pearson correlation coeffi- ment of LBD16, 7 a key component of the auxin pathway leading cients) with LBD16 across all the “Perturbations” microarray to the divisions in the xylem pole pericycle (XPP) during lateral experiments released at August 2014 from Genevestigator (i.e.,

*Correspondence to: Carolina Escobar; Email: [email protected] Submitted: 09/29/2014; Revised: 10/22/2014; Accepted: 10/24/2014 http://dx.doi.org/10.4161/15592324.2014.990825

www.tandfonline.com Plant Signaling & Behavior e990825-1 those genes with a positive correlation in different experiments co-regulated genes and DE in the RKN and CN transcriptomes, were considered positively co-regulated with LBD16, and those 16 were up-regulated by exogenous auxin treatment,13 with a negative correlation were considered co-regulated nega- (Table S1) reinforcing the importance of LBD16 in the galls tively with LBD16). We then downloaded their expression values auxin- mediated signaling. The fact that LBD16 and auxin (log2 values, referred always to the controls used in each experi- inducible co-regulated genes were repressed in syncytia ment) in the different transcriptomes available from Arabidopsis (Table S1) suggested a selective down- regulation of the LBD16 infected with RKNs and cyst nematodes, using our previously auxin mediated pathways in the CNs feeding sites, in contrast to designed tool NEMATIC.10 LBD16 was up-regulated in the RKNs. These results are in accordance to the absence of pheno- early-medium stages of GC and gall development7,11 while it was type in LBD16 loss-of-function plants infected with Heterodera down-regulated in microaspirated syncytia.12 From the 50 posi- spp7 regardless of other auxin-regulated genes induced in syncytia tively co-regulated genes from Genevestigator, 20 were differen- (for a review see refs. 3, 14). Similar results were obtained for the tially expressed (DE) also in the plant-nematode transcriptomes genes co-regulated negatively with LBD16, as they were found (Table S1). As expected, due to the opposite regulation pattern mainly down-regulated in galls and GCs while they showed a sig- of LBD16 in syncytia as compared to galls and GCs, most of the nificant up-regulation in the CNs feeding sites (Table S2). Strik- positively co-regulated genes with LBD16 were also upregulated ingly, from the 41 negatively co-regulated genes that were in galls and GCs, but downregulated in syncytia. Among them, upregulated in syncytia, 28 belonged to the category of photosyn- we found other LBD genes co-regulated with LBD16, i.e., thesis from Mapman,15 (Table S2) a characteristic of the syncytia LBD17, LBD18 or LBD33. Additionally, from the 20 LBD16 transcriptome that is different to that of GCs.12,16 In agreement to this, differentiation of plas- tids occurs in the syncytium.12 Therefore, the expression pat- terns and the molecular net- work of co-regulated genes with LBD16 in other biologi- cal processes seems also con- served during the RKNs feeding site development. To deepen in the putative role of LBD16 and its co-regu- lated genes in PPN feeding sites, we compared their expres- sion values in GCs, galls and syncytia with the expression values that they showed in the 3287 transcriptomes deposited in the “Perturbations” section of Genevestigator.17 The most similar expression profile for the pool of genes co-regulated with LBD16 in RKN feeding Downloaded by [Universidad Castilla La Mancha] at 07:42 23 April 2015 sites was found in the transcrip- tome of Arabidopsis explants incubated on solid callus induc- ing media (Fig. 1 and S1), while those co-regulated with syncytia showed similar expres- sion values to the transcrip- Figure 1. A simplified model showing the general results of in silico comparisons of the GCs and galls transcrip- tomes of callus in the process tomes highlighting similarities to that of hormone-regulated genes and those co-expressed with LBD16. The of shoot regeneration (Fig. 1 most similar expression profile for the pool of genes co-regulated with LBD16 in galls and GCs was found in the and S2). In this last study, transcriptome of Arabidopsis explants incubated on solid callus inducing media while those co-regulated in syn- LBD16 and LBD18 were cytia showed similar expression values to the transcriptomes of callus in the process of shoot regeneration indi- cated in the diagram. GUS activity of a DR5::GUS line on a 4 dpi gall showing intense staining in GCs. Syncytia downregulated as they are in show abundant chlorophyll fluorescence under fluorescence microscopy indicative of the presence of chloro- syncytia, suggesting that they plasts. N, nematode; all giant cells are marked with an asterisk. Drawings of callus in white boxes and their con- should be repressed during nections to the syncytia, GCs and galls microarray data are based on in silico analysis from the microarray the developmental shift from experiments in Genevestigator released at August 2014 (Fig S1 and S2). callus to shoot development.

e990825-2 Plant Signaling & Behavior Volume 10 Issue 3 Moreover, during shoot regeneration from callus, induction of mediated by auxins that are shared also by LR and callus develop- photosynthetic genes is described18 similarly to syncytia.12 These ment. In contrast, LBD16 and its positively co-regulated genes results reinforce the molecular link between callus and gall forma- were repressed in syncytia. However, cytokinins induced genes tion suggested in Cabrera et al.7 with LBD16 and other LBD are enriched in syncytia transcriptome, showing also clear similar- genes as key players during these processes.19 However, the down ities with the transcriptome of shoot regeneration from callus that regulation of LBD16 and co-regulated genes in syncytia suggest is modulated preferentially by cytokinins. Those findings estab- that this pathway should be repressed during syncytia develop- lish differences in the regulatory networks leading to both feeding ment, probably to favor developmental programs related to shoot sites formation probably modulated by the auxin/cytokinin regeneration from callus. Interestingly, the balance between auxins balance. andcytokininsiscrucialforthemaintenanceorprogressionto shoot from callus.20 In this respect, the proportion of cytokinin up-regulated genes in syncytia is much higher than those induced Disclosure of Potential Conflicts of Interest by auxins,10 which is indispensable for the shoot regeneration No potential conflicts of interest were disclosed. from callus.19 This is also in accordance to the fact that plastids develop in syncytia,12 as plastid development is induced in dark- ness in the presence of cytokinins,21 a similar situation that might Funding be happening in syncytia. This is also in accordance to the fact This work was supported by the Spanish Government that genes encoding chloroplast proteins were among the most (AGL2010-17388 and AGL2013-48787, to C. Escobar, strongly upregulated genes in early forming syncytia.12 However, a CSD2007-057 to C. Fenoll, PLANT-KBBE PLANT-043, higher proportion of auxin than cytokinin inducible genes is neces- PCIN-2013-053 to C. Fenoll and C. Escobar). Javier Cabrera sary for callus induction,18 equivalent to that of early developing was supported by a fellowship from the Ministry of Education, Arabidopsis GCs.10 Moreover, the presence of chloroplast-like Spain. We acknowledge Prof. Dr. Florian M. W. Grundler and structures inside 7dpi dark-grown GCs induced by M. graminicola Dr. Shahid M. Siddique from Bonn University for kindly pro- in rice has been recently demonstrated. This reinforces the idea vide us with the syncytia fluorescence microscopy picture from that a dinamic disturbance of hormone homeostasis, probably the Figure 1. balance between auxin and cytokinin is important during the development and maintenance of both feeding cells.22 In conclusion, in silico data confirm LBD16 and its co-regu- Supplemental Material lated genes as a hallmark of the RKN feeding sites compared to Supplemental data for this article can be accessed on the syncytia. Those genes are integrated in signaling cascades publisher’s website.

References BOUNDARIES-DOMAIN 16 during the interaction 13. Nemhauser JL, Hong F, Chory J. Different plant hor- 1. Huang CS, Maggenti AR. Mitotic aberrations and Arabidopsis-Meloidogyne spp. provides a molecular link mones regulate similar processes through largely non- nuclear changes of developing giant cells in Vicia faba between lateral root and root-knot nematode feeding site overlapping transcriptional responses. Cell 2006; caused by root knot nematode, Meloidogyne javanica. development. New Phytol 2014; 203:632-45; 126:467-75; PMID:16901781; http://dx.doi.org/ Phytopathology 1969; 59:447-55. PMID:24803293; http://dx.doi.org/10.1111/nph.12826 10.1016/j.cell.2006.05.050 2. Caillaud MC, Lecomte P, Jammes F, Quentin M, Pag- 8. Goh T, Joi S, Mimura T, Fukaki H. The establishment 14. Quentin M, Hewezi T, Damiani I, Abad P, Baum T, notta S, Andrio E, de Almeida Engler J, Marfaing N, of asymmetry in Arabidopsis lateral root founder cells Favery B. How pathogens affect root structure. In: Gounon P, Abad P, et al. MAP65–3 microtubule-asso- is regulated by LBD16/ASL18 and related LBD/ASL Crespi M ed. Root Genomics and Soil Interactions. ciated protein is essential for nematode-induced giant proteins. Development 2012; 139:883-93; Blackwell Publishing Ltd., Oxford, UK, 2012. cell ontogenesis in Arabidopsis. Plant Cell 2008; PMID:22278921; http://dx.doi.org/10.1242/dev. 15. Thimm O, Blasing O, Gibon Y, Nagel A, Meyer S, € 20:423-37; PMID:18263774 071928 Kruger P, Selbig J, Muller LA, Rhee SY, Stitt, M. 3. Kyndt T, Vieira P, Gheysen G, De Almeida-Engler J. 9. Laplaze L, Parizot B, Baker A, Ricaud L, Martiniere A, MAPMAN: a user-driven tool to display genomics data Auguy F, Franche C, Nussaume L, Bogusz D, Haseloff J. sets onto diagrams of metabolic pathways and other Downloaded by [Universidad Castilla La Mancha] at 07:42 23 April 2015 Nematode feeding sites: unique organs in plant roots. Planta 2013; 238:807-18; PMID:23824525; http://dx. GAL4-GFP enhancer trap lines for genetic manipulation biological processes. Plant J 2004; 37:914-39; doi.org/10.1007/s00425-013-1923-z of lateral root development in Arabidopsis thaliana. J Exp PMID:14996223; http://dx.doi.org/10.1111/j.1365- 4. Cabrera J, Barcala M, Fenoll C, Escobar C. Transcrip- Bot 2005; 56(419):2433-42; PMID:16043452; http:// 313X.2004.02016.x tomic signatures of transfer cells in early developing dx.doi.org/10.1093/jxb/eri236 16. Barcala M, Garcia A, Cabrera J, Casson S, Lindsey K, nematode feeding cells of Arabidopsis focused on auxin 10. Cabrera J, Bustos R, Favery B, Fenoll C, Escobar C. Favery B, Garcıa-Casado G, Solano R, Fenoll C, Esco- and ethylene signaling. Front Plant Sci 2014; 5:107; NEMATIC: a simple and versatile tool for the in silico bar, C. Early transcriptomic events in microdissected PMID:24715895; http://dx.doi.org/10.3389/fpls. analysis of plant-nematode interactions. Mol. Plant Arabidopsis nematode-induced giant cells. Plant J 2014.00107 Pathol 2014; 15:627-36; PMID:24330140; http://dx. 2010; 61:698-712; PMID:20003167 5. Sobczak M, Golinowski W. Structure of cyst nematode doi.org/10.1111/mpp.12114 17. Hruz T, Laule O, Szabo G, Wessendorp F, Bleuler S, feeding sites. In: Berg RH, Taylor CG, eds. Cell Biol- 11. Jammes F, Lecomte P, de Almeida-Engler J, Bitton F, Oertle L, Widmayer P, Gruissem W, Zimmermann P. ogy of Plant Nematode Parasitism. Heidelberg, Ger- Martin-Magniette ML, Renou JP, Abad P, Favery B. Genevestigator v3: a reference expression database for many: Springer Verlag, 2009; pp 153-88. Genome-wide expression profiling of the host response the meta-analysis of transcriptomes. Adv Bioinformat- 6. Nicol JM, Turner SJ, Coyne DL, Nijs L. Current nem- to root-knot nematode infection in Arabidopsis. Plant J ics 2008; 2008:420747; PMID:19956698; http://dx. atode threats to world agriculture. In: Jones J, Gheysen 2005; 44:447-58; PMID:16236154; http://dx.doi.org/ doi.org/10.1155/2008/420747 G, Fenoll C, eds. Genomics and Molecular Genetics of 10.1111/j.1365-313X.2005.02532.x 18. Ping C, Lall S, Nettleton D, Howell SH. Gene expres- Plant–Nematode Interactions. Dordrecht: Springer 12. Szakasits D, Heinen P, Wieczorek K, Hofmann J, sion programs during shoot, root, and callus develop- Verlag, 2011; pp 21-45. Wagner F, Kreil DP, Sykacek P, Grundler FM, Bohl- ment in Arabidopsis tissue culture. Plant Physiol 2006; 7. Cabrera J, Dıaz-Manzano F, Sanchez M, Rosso MN, mann H. The transcriptome of syncytia induced by the 141:620-37; PMID:16648215; http://dx.doi.org/ Melillo T, Goh T, Fukaki H, Cabello S, Hofmann J, cyst nematode Heterodera schachtii in Arabidopsis roots. 10.1104/pp.106.081240 Fenoll C, et al. A role for LATERAL ORGAN Plant J 2009; 57:771-84; PMID:18980640 19. Fan M, Xu C, Xu K, Hu Y. LATERAL ORGAN BOUNDARIES DOMAIN transcription factors direct

www.tandfonline.com Plant Signaling & Behavior e990825-3 callus formation in Arabidopsis regeneration. Cell Res 21. Chory J, Reinecke D, Sim S, Washburn T, Brenner Van Criekinge W, De Meyer T, Kyndt T. Transcrip- 2012; 22:1169-80; PMID:22508267; http://dx.doi. M. A role for cytokinins in de-etiolation in Arabidop- tional analysis through RNA sequencing of giant cells org/10.1038/cr.2012.63 sis (det mutants have an altered response to cytoki- induced by Meloidogyne graminicola in rice roots. J 20. Skoog F, Miller CO. Chemical regulation of growth nins). Plant Physiol 1994; 104:339-47; Exp Bot. 2013; 64(12):3885-98. doi: 10.1093/jxb/ and organ formation in plant tissue cultured in vitro. PMID:12232085 ert219 Symp Soc Exp Biol 1957; 11:118-31; 22. Ji H, Gheysen G, Denil S, Lindsey K, Topping JF, PMID:13486467 Nahar K, Haegeman A, De Vos WH, Trooskens G, Downloaded by [Universidad Castilla La Mancha] at 07:42 23 April 2015

e990825-4 Plant Signaling & Behavior Volume 10 Issue 3 CHAPTER 4: A new method for the phenotyping of the giant cells Chapter 1: Introduction

Chapter 2: Holistic Studies A general overview of the plant nematode interaction Chapter 3: LBD16 & Lateral Roots and the research Chapter 4: 3D Reconstruction performed in this field

Developmental pathways Holistic studies performed mediated by hormones in the susceptible and and altered during the resistant interactions with plant- nematode root- knot nematodes interactions

A method for the 3D phenotping of the giant cells. Insights into their volumes and morphological features

Transcriptomic and The transcriptomes of molecular mechanisms are early developing giant cells common between lateral from tomato and root and giant cells and arabidopsis showed a gall formation. LBD16 is a common massive and key gene in both functional gene repression processes

A tool to facilitate the selection of genes frm the data generated on trasncriptomes from plant- nematode interactions and other fields

miRNA and tasiRNA Regulation of LBD16 co- pathways for gene regulated genes differs silencing are functional in between giant cells and galls and the rasiRNAs are syncytia. Differences in highly induced. Molecular the balance parallelisms with lateral Auxins/Cytokinins between roots both feeding cells Aim of the chapter.

The identification of susceptibility genes from tomato and Arabidopsis like TPX-1 or LBD16, respectively, which mutation yielded a reduction in the number of infections, prompted the necessity of finding a reproducible method to test their impact on giant cell development and/or formation. There was no standardized method in the literature to compare the size, neither the morphological features of the giant cells within the galls generated in different plant lines. A simplified and standardized method was established in the article entitled “Phenotyping nematode feeding sites: three- dimensional reconstruction and volumetric measurements of giant cells induced by root-knot nematodes in Arabidopsis”. We demonstrated the validity of our method in the detection of differences between loss of function and control lines. Moreover, for the first time, we obtained the three dimensional reconstruction and the volumes occupied by the giant cells induced by the nematode inside the gall during the progress of the infection. Research

Methods Phenotyping nematode feeding sites: three-dimensional reconstruction and volumetric measurements of giant cells induced by root-knot nematodes in Arabidopsis

Javier Cabrera1, Fernando E. Dıaz-Manzano1, Marta Barcala1, Ignacio Arganda-Carreras2, Janice de Almeida-Engler3, Gilbert Engler3, Carmen Fenoll1 and Carolina Escobar1 1Facultad de Ciencias Ambientales y Bioquımica, Universidad de Castilla-La Mancha, Av. Carlos III s/n, E-45071 Toledo, Spain; 2AgroParisTech, UMR1318, Institut Jean-Pierre Bourgin, 78026 Versailles, France; 3Institut National de la Recherche Agronomique, UMR 1355 ISA/Centre National de la Recherche Scientifique, UMR 7254 ISA/Universite de Nice-Sophia Antipolis, UMR ISA, 400 Route des Chappes, 06903 Sophia-Antipolis, France

Summary Author for correspondence:  The control of plant parasitic nematodes is an increasing problem. A key process during the Carolina Escobar infection is the induction of specialized nourishing cells, called giant cells (GCs), in roots. Tel: +34 925268800 ext. 5476 Understanding the function of genes required for GC development is crucial to identify targets Email: [email protected] for new control strategies. We propose a standardized method for GC phenotyping in differ- Received: 21 July 2014 ent plant genotypes, like those with modified genes essential for GC development. Accepted: 19 November 2014  The method combines images obtained by bright-field microscopy from the complete serial sectioning of galls with TRAKEM2, specialized three-dimensional (3D) reconstruction software New Phytologist (2015) for biological structures. doi: 10.1111/nph.13249  The volumes and shapes from 162 3D models of individual GCs induced by Meloidogyne javanica in Arabidopsis were analyzed for the first time along their life cycle. A high correla- Key words: Arabidopsis, giant cells (GCs), tion between the combined volume of all GCs within a gall and the total area occupied by all phenotyping, root-knot nematodes (RKN), the GCs in the section/s where they show maximum expansion, and a proof of concept from three-dimensional reconstruction. two Arabidopsis transgenic lines (J0121 ≫ DTA and J0121 ≫ GFP) demonstrate the reliability of the method.  We phenotyped GCs and developed a reliable simplified method based on a two-dimen- sional (2D) parameter for comparison of GCs from different Arabidopsis genotypes, which is also applicable to galls from different plant species and in different growing conditions, as thickness/transparency is not a restriction.

encloses them. In order to precisely determine the role of specific Introduction genes during GC differentiation and/or maintenance, it is crucial The agricultural impact of plant parasitic root-knot nematodes to phenotype these feeding cells and to compare those formed in (RKN), particularly of Meloidogyne spp., is increasing with the wild-type plants with those formed in plants with modified gene progressive banning of many effective chemicals used for their functions. control. RKN are widespread around the world and can infect GC phenotyping has been based on observations of histological thousands of plant species. As the durability of natural resistance sections from nematode-induced galls embedded in resins, mostly ® â genes is in question (Abad et al., 2003), there is an urgent need to Araldite (Sigma-Aldrich) or Technovit (Heraeus Kulzer search for alternative strategies for their control, such as those GmbH, Wehrheim, Germany), using bright-field optics and based on biotechnology (Fuller et al., 2008; Lilley et al., 2011; images collected with digital cameras (de Almeida Engler et al., Atkinson et al., 2012). Hence, assessing the function of genes 2012; Kyndt et al., 2013). To date, data regarding three-dimen- putatively involved in crucial processes for the plant–nematode sional (3D) shape, volume, and changes in these parameters dur- interaction is essential. One of these processes is the formation of ing nematode infection are scarce. This is mainly because of the specialized nourishing cells, named giant cells (GCs; Jones & complex structure in which GCs are embedded within the galls Goto, 2011), within the vasculature of the host roots which that are surrounded by other tissues, and it is also due to the become hyperplasic and swell, forming a gall or root-knot that presence of expanding sedentary nematodes within the galls,

Ó 2015 The Authors New Phytologist (2015) 1 New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 2 Research Phytologist

which limits accessibility of the GCs for optical imaging. Another many researchers. Moreover, the cell walls of epidermal and corti- complex, inaccessible structure is the shoot apical meristem cal cell layers in mature galls are exceptionally light absorbing, (SAM), for which clear 3D reconstructions have been obtained by greatly limiting the performance of a two-photon microscope. A confocal laser scanning microscopy with different methods (Kayes more recently developed type of microscopy called selective plane & Clark, 1998; Grandjean et al., 2004; Chakraborty et al., illumination microscopy (SPIM) enables the generation of 3D 2013), but for which cell volume data have only been calculated representations of large but relatively transparent biological sam- using serial histological sections (Vanhaeren et al., 2010). By con- ples. Use of SPIM allows a significant improvement in live tissue trast, high-quality volumetric data have been obtained by confocal preservation during extensive time-lapse imaging, although there laser scanning microscopy for other plant structures, for example are still limitations in visualizing large specimens such as mature the root tip meristem (Vanhaeren et al., 2010). Similar to the galls. Therefore, the limits set by the ‘transport mean free path’ SAM, GCs present difficulties in terms of accessibility as they are of a photon in the gall tissues, as a consequence of absorption situated in the center of the gall, where the root cortical cells and scattering of light, prevent deep 3D imaging with confocal, appear to be hypertrophied and the vascular tissues that surround two-photon or SPIM microscopy. the GCs present abundant tiny asymmetrically dividing cells In this work, we describe a novel approach based on standard resulting from continuous cell division as the infection progresses microscopy to obtain 3D reconstructions of GCs induced by (Bird, 1961; Bleve-Zacheo & Melillo, 1997). Quantitative data Meloidogyne javanica in Arabidopsis roots at early to late develop- on GC shape and size are limited and rely on methods based on ment stages (3, 5, 7, 9, 11, 21 and 40 d post infection (dpi)). The two-dimensional (2D) parameters. Thus, no volumetric data are method combines images obtained by conventional light micros- yet available. The physiological status of GCs is closely related to copy from the complete serial sectioning of galls with software their morphological features, for instance the presence of cell wall specialized in the 3D reconstruction of biological structures, ingrowths and wall thickenings. However, currently no standard- TRAKEM2 (Cardona et al., 2012). We provide for the first time ized methods have been adapted to accurately phenotype GCs. complete 3D models for all individual GCs within galls and addi- The methods that have been used are based on the measurement tional important morphological features related to their volume of GC diameters over 2D sections (Gal et al., 2006; Das et al., and shape. These results established a reference for the develop- 2008; de Almeida Engler et al., 2012), the areas of individual ment of a simplified method for GC size comparison, based on GCs (Banora et al., 2011; Vieira et al., 2012, 2013; Antonino de 2D images taken from gall sections. This was possible because we Souza et al., 2013; Iberkleid et al., 2013) or the GC pool area as a detected a high correlation between the combined volume of all whole (Vovlas et al., 2005; Wasson et al., 2009; Portillo et al., GCs within a gall and the total area occupied by all the GCs in 2013; Cabrera et al., 2014b), deduced from semi-thin sections. the section(s) where they showed their maximum expansion. We Using these methodologies, differences in GC shape and size were demonstrate that this standardized method can be used for com- found between wild type and loss-of-function or overexpressing parison of GCs from different Arabidopsis lines and propose that lines. However, these methods are based exclusively on 2D imag- it can also be applied to galls from different plant species and in ing techniques (Kyndt et al., 2013), which have limitations in different growing conditions. accurately determining the final GC shapes and volumes. Three-dimensional representations of the surface of GCs have Materials and Methods been obtained based on scanning electron microscopy (Jones & Dropkin, 1976; Orion & Wergin, 1982), which does not allow Nematode populations volume measurements, as only images of the surface are recorded. Three-dimensional representations can be produced from optical Meloidogyne javanica Treub, 1885 (Portillo et al., 2009) was main- sectioning of live or fixed tissue samples by confocal laser scan- tained in vitro on cucumber (Cucumis sativus L.) plants grown at ning microscopy. However, the image quality rapidly decreases 28°C in the dark in 0.3% Gamborg medium (Gamborg et al., when analyzing very thick specimens. In order to improve confo- 1968) supplemented with 3% sucrose. Egg hatching was stimu- cal imaging of GCs, clarifying methods have been applied to lated in sterile water for 3–4d. enhance undisturbed light penetration, thereby permitting visual- ization and 3D reconstruction of entire galls, GCs and their Plant material, growth conditions and nematode nuclei, based on propidium iodide staining (Vieira et al., 2012). inoculation However, accurate 3D reconstructions of galls were only obtained down to a depth of 200 lm using Arabidopsis as a host Arabidopsis thaliana (L.) Heynh Columbia-0 (Col-0) plants were (Vieira et al., 2012). Larger galls, which are typical of late devel- used throughout this study. Seeds were surface-sterilized with opmental stages, as well as galls induced in different host plants 30% commercial bleach, washed, and sown in 0.3% Gamborg did not allow efficient confocal imaging even after extensive clear- medium (Gamborg et al., 1968) supplemented with 1.5% ing, as a consequence of excessive light absorption and scattering. sucrose. For stratification, plates were kept at 4°C for 2 d and Two-photon excitation microscopy can generate images deeper thereafter the plates were kept vertically in a growth chamber at inside complex specimens than standard confocal microscopy as 25°C, 60% relative humidity and in a long-day photoperiod. a result of long wavelength excitation and more efficient emission The plates were inoculated 5 d later just behind each root tip with light detection. However, this equipment is still not available to 7–10 M. javanica juveniles per main root. All plants showed only

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 3 a primary root when inoculated, therefore having a high growth steps of incubation in an Araldite : acetone solution: 2 h in a 1 : 3 homogeneity to avoid individual differences. Plants were care- solution, overnight incubation in a 1 : 1 solution and three steps fully examined every 12 h under a Leica Mz125 stereomicroscope of 1 h in a 3 : 1 solution. Finally, the galls were oriented longitu- to establish a penetration and infection timeline, resulting in a dinally in silicone molds filled with 100% Araldite and main- maximum error of 12 h when assessing gall age (Barcala et al., tained at 60°C for 48 h for the polymerization of the Araldite. 2010). Galls were hand-dissected at 3, 5, 7, 9, 11, 21 and 40 dpi All steps were performed at 4°C. only from the primary root for maximum homogeneity. Gall sectioning and image capture Gall fixation and embedding Galls at 3, 5, 7, 9, 11, 21 and 40 dpi were fully sectioned at 2 lm Galls were rinsed twice for 10 min in sodium phosphate buffer with a diamond knife in an ultramicrotome (Microm HM360; (10 mM; pH 7) and fixed overnight in 3% glutaraldehyde. Dehy- Thermo Scientific, Waltham, MA, USA). Correlative longitudi- dration of the galls was carried out in three steps of 15 min in nal sections were carefully recovered and placed one by one on increasing concentrations of ethanol (30, 50 and 70%), two steps glass slides. Sections were stained for 5 min with 1% toluidine of 1 h in 90% ethanol, three steps of 1 h in 100% ethanol and, blue in 1% borax solution (TAAB) at 40°C. High-quality micro- À finally, two steps of 15 min in 100% acetone. Subsequently, galls graphs (0.2 lm pixel 1) of the 2-lm sections containing GCs were embedded in Araldite (Sigma-Aldrich) following several were obtained at 910 magnification under a light microscope (Nikon Eclipse 90i) equipped with a digital camera (Nikon Dxm Table 1 The average number of giant cells (GCs) scored in each 1200c). TRAKEM2 software provided together with the Fiji image developmental stage, and the average number of sections obtained for a processing toolbox was downloaded from http://fiji.sc/Fiji and complete sectioning of every GC within a gall induced by Meloidogyne javanica in Arabidopsis thaliana roots used for sample alignment and 3D gall reconstruction. Most of the pre-established parameters of the software were maintained Days post Average no. for all the samples reconstructed, with the exception of: pixel infection (dpi) No. of GCs of sections width (0.2 lm), pixel height (0.2 lm) (thus, resolution is À 0.2 lm pixel 1 on the x/y-axis), voxel depth (2 lm), steps per 32020 51925scale (5), feature descriptor size (8), maximum alignment error 71951(50) and iterations for mesh smoothing (15). Measurements of 93343volume, area, maximum diameter or surface area of the GCs were 11 26 59 calculated by TRAKEM2 depending on the number of pixels 21 23 74 occupied by each GC. 40 22 99

(a) (b) (c)

Fig. 1 Three basic steps of section processing prior to the three-dimensional (3D) reconstruction of a gall induced by Meloidogyne javanica in Arabidopsis. Five representative sections of the whole 2-lm section series (indicated by numbers) from a gall stained with toluidine are shown. (a) Consecutive nonaligned images. (b) Consecutive aligned images. (c) Consecutive aligned images with giant cells (GCs) labeled. The color code for each GC was maintained in all the consecutive sections as indicated. See Supporting Information Video S1 for a detailed view of the process.

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 4 Research Phytologist

Fig. 2 Serial sectioning of an Arabidopsis gall induced by Meloidogyne javanica with aligned images and colored giant cells (GCs) (numbers indicated at the top left of the images). Two-micrometer sections are stained with toluidine. This method enables detailed characterization of cell morphology across the series (left panels, only toluidine staining; right panels, colored cells). It avoids misinterpretation arising from the observation of single isolated sections or a small group of sections. The blue cell in section 49 could be misinterpreted in section 57 as three independent GCs. The pink cell that in section 49 is distant from the nematode appears to be in direct contact in section 57. See Supporting Information Video S2 for a detailed view.

Table 2 Volume data for the giant cells (GCs) induced by Meloidogyne javanica in Arabidopsis thaliana roots analyzed in each developmental stage

Giant cell volume (lm3 9 1000)

Giant cells scored 3 dpi 5 dpi 7 dpi 9 dpi 11 dpi 21 dpi 40 dpi

1 51.08 429.04 264.24 1706.60 1222.61 1266.25 898.34 2 78.10 103.77 477.94 434.32 777.58 706.38 862.16 3 64.98 208.38 280.76 675.00 372.13 32.39 1578.19 4 34.52 297.64 491.99 414.31 239.64 1002.97 1475.63 5 33.10 181.72 802.90 711.56 1524.61 755.42 653.03 6 44.03 79.24 157.66 450.44 169.88 239.34 5515.04 7 119.83 153.66 422.48 208.10 7.08 461.10 432.55 8 56.92 35.45 705.58 159.68 42.41 4284.63 2470.65 9 44.00 81.19 946.48 759.53 66.31 7178.51 5025.32 10 197.45 47.31 118.08 881.92 690.28 1467.47 4290.73 11 35.04 174.20 223.68 550.18 825.08 2438.88 5380.71 12 23.44 14.27 556.00 1035.09 103.00 1811.96 7484.93 13 19.99 22.52 364.17 119.77 194.19 4417.29 6426.85 14 22.36 13.59 188.85 259.01 113.74 1942.30 14 537.59 15 16.87 17.44 392.07 427.42 261.62 822.25 980.73 16 63.01 15.25 186.02 547.25 142.68 679.39 1273.03 17 48.43 16.57 238.31 197.75 169.92 534.44 1886.69 18 46.60 11.90 564.68 100.13 907.22 985.97 984.86 19 83.69 29.14 301.89 163.82 1609.97 2646.93 1354.74 20 12.19 133.19 2887.99 543.58 1177.22 21 69.74 2213.99 523.17 3200.48 22 91.36 1138.09 2600.88 464.22 23 396.02 3875.18 1362.92 24 196.98 1924.14 25 126.28 1847.33 26 633.60 1252.40 27 210.35 28 238.47 29 73.83 30 83.45 31 325.51 32 154.63 33 590.57 Average volumes 54.78 101.70 404.41 397.75 945.35 1682.80 3106.99

dpi, days post infection. New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 5

b; Video S1). This process was repeated consecutively down Results and Discussion through all the sections to the end of the stack. The alignment of consecutive sections made it easier to define each GC within a GC labeling from gall serial sections using TRAKEM2 gall and, by making use of the area list tool, to label it manually Herein, we used a simple technique to obtain a 3D overview of with a color code which was used in all the consecutive sections RKN-induced GCs, analyzing their phenotypic characteristics in which that particular GC could be distinguished (Fig. 1c; and the volume that they occupy during nematode life cycle pro- Video S1). Detailed step-by-step video tutorials on the use of gression. A total of 2240 longitudinal histological sections (2 lm TRAKEM2 created by the authors in collaboration with the soft- thick) from 162 GCs in seven different stages of gall development ware developers are available at http://fiji.sc/TrakEM2_tutorials. (3, 5, 7, 9, 11, 21 and 40 dpi) (Table 1) were processed. The This alignment proved to be particularly useful to avoid misin- average number of sections obtained from the complete section- terpretation of GC number and size, as individual GCs could be ing of every GC within a gall increased during development traced along the entire gall. In this way, multiple GCs can be (from 20 at 3 dpi to 99 at 40 dpi; Table 1). The stack of light unambiguously identified, allowing one to distinguish between microscope photographs obtained was imported into TRAKEM2 multiple GCs and those individual cells that in some 2D sections (http://fiji.sc/TrakEM2), software that allows 3D modeling from seemed to be more than one as a consequence of the presence of histological serial sections (Fig. 1a; Supporting Information interspersed vascular cells. See, for instance, the blue GC in Fig. 2 Video S1). TRAKEM2 has been successfully used in other research which turned out to represent a single cell with protrusions fields, as in the reconstruction of neuronal circuits in Drosophila (Fig. 2, blue GC; Video S2). The GC position with respect to the (Cardona et al., 2012). It is publicly available for download nematode (Fig. 2, pink GC) could also be clearly determined; for together with the Fiji image processing toolbox (an open source example, a GC that in a given section seemed to be far away from image processing package based on IMAGEJ (US National the nematode was found to be adjacent to the parasite when seen Institutes of Health, Bethesda, MD, USA), widely used for in some of the subsequent sections in which this cell appeared microscopy image processing; Schindelin et al., 2012; Schneider (Fig. 2, pink GC). Usually, GCs contained a zone with an et al., 2012), in contrast to commercial software previously used elongated tube-like structure in the area adjacent to the nema- in 3D reconstructions in Arabidopsis (Vanhaeren et al., 2010). tode, in contrast with the more rounded and expanded shapes For optimal GC recognition, all sections were rotated and trans- that they presented in the area more distant from the nematode lated to perfectly align them onto each other by making use of (Figs 2, 4i–l). This strongly suggests that all GCs have been in the tools for free affine transforms provided by TRAKEM2. The contact with the nematode at some stage of the gall development. high similarity between consecutive sections allowed TRAKEM2 Thus, we determined the exact number of GCs in each gall, again to analyze their pixels for similar features and rigidly transform avoiding misinterpretations arising from those sections where (rotate and/or translate) the upper image for an optimal match GCs were interspersed with other cells. We found that the num- over the consecutive image on the bottom (compare Fig. 1a and ber of GCs was not correlated with the stage of development

(a) (b)

Fig. 3 Volume evolution, sphericity and surface area to volume ratio (SA : V) of giant cells (GCs) formed by Meloidogyne javanica in Arabidopsis along infection time. (a) Volume of each of the individual 162 GCs reconstructed. (b) The average volume occupied by all the GCs corresponding to the same developmental stage. (a, b) The x-axes in the two graphs indicate galls at different (c) (d) infection stages: at 3 (n = 20), 5 (n = 19), 7 (n = 19), 9 (n = 33), 11 (n = 26), 21 (n = 23) and 40 d post infection (dpi; n = 22); in (b) bars indicate Æ SE. (c) The ratio of the total surface area (SA) to the final volume (V) occupied by single GCs (SA : V ratio). The ratio becomes smaller as the volume increases. (d) Sphericity (Ψ) measurement of the GCs at the different stages of development. Note that sphericity is different in each GC at all developmental stages indicated. Vp, volume of the GC; Ap, surface area of the GC.

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 6 Research Phytologist

(Table 1). These observations reinforce the idea that, once the the feeding cells are fully expanded. Accordingly, Class I small nematode is sedentary within the vascular cylinder, differentia- heat-shock proteins (sHSPs) are highly abundant in developed tion of those selected cells into GCs is in progress and almost no GCs as compared with the rest of the gall, correlating with the new GCs will be induced at later infection stages. This suggests activation of several sHSP promoters in the medium–late develop- that the nematode moves its head from one cell to another for mental stages. These chaperones may be related to the mainte- nutrient uptake until it has completed its life cycle (Bleve-Zacheo nance of proper protein folding, preventing the aggregation of & Melillo, 1997; Grundler & B€ockenhoff, 1997). abundantly synthesized proteins in highly active cells (Escobar In summary, using TRAKEM2 we aligned histological sections et al., 2003; Barcala et al., 2008). Thus, the coincidence in time of from galls induced by RKN in Arabidopsis, facilitating the identifi- low SA : V ratios and high metabolic activities is apparently a con- cation and observation of the GCs within the gall and reducing the tradiction, but the extensive formation of wall ingrowths at multiple sources of misinterpretation regarding their number and medium–late developmental stages could be a response to a clear position. This method also allowed the easy collection of accurate functional requirement to compensate for the decrease in the qualitative and quantitative data from histological sections that are SA : V ratio as the cells expands. Thus, an increase in the effective essential for GC phenotyping, as shown in the following sections. solute exchange area is achieved through the differentiation of the GCs into transfer cells (Jones & Northcote, 1972; Jones & Gun- ning, 1976; Cabrera et al., 2014a). Yet, the increase in their Three-dimensional reconstruction and volume estimation plasma membrane surface area may be up to 20-fold (Jones & of GCs

After aligning and assigning a color code to each GC, 3D models (a) (b) (c) (d) of GCs were generated using the IMAGEJ 3D Viewer (Schmid et al., 2010), integrated in TRAKEM2 (Fig. S1). All the 3D mod- els obtained in this work can be visualized and handled online at https://sketchfab.com/jcabrerachaves/models. In addition to the visualization of the morphological characteristics of the cells, one (e) (f) (g) (h) of the most valuable applications of the 3D reconstruction is the extraction of volumetric data (Table 2). We measured the vol- umes of 162 GCs obtained from 2240 independent sections at seven different stages (3, 5, 7, 9, 11, 21 and 40 dpi). They ranged from 7081 lm3 (11 dpi) to 14 537 585 lm3 (40 dpi; Table 2). (i) (j) Interestingly, the volume of individual GCs did not always corre- late well with the stage of gall development (Table 2; Fig. 3a). However, the average volume occupied by all the GCs as a pool within a gall at each infection time showed a clear tendency to increase as the infection progressed, with a positive correlation between the two variables (see exponential line tendency in Fig. 3b). This suggests that individual GCs grow asynchronously during gall development, perhaps as a result of mechanical restric- tions or of differential stimulation by the nematode. GC volumes at the early stages of development were more homogenous than the volumes of older GCs (Fig. 3a,b). Individual variability may (k) (l) buffer differences between some developmental stages; for exam- ple, there was a large increment in size from 9 to 11 dpi, but the size of the GCs did not vary much from 3 to 5 dpi or from 7 to 9 dpi. All these findings emphasize the importance of the GC pool within a gall, as a whole functional ‘pseudo-organ’. The surface area and the surface area to volume ratio (SA : V) for each GC were obtained after the reconstruction (Fig. 3c). The diffusion rate of solutes is proportional to the surface area, and larger cells of similar shape will have slower diffusion rates than smaller cells. Thus, the higher the cell SA : V ratio, the more effec- Fig. 4 Representative samples of reconstructed giant cell (GC) shapes. tive solute exchange should be, and therefore a higher metabolic (a–h) Three-dimensional (3D) reconstruction showing the highly diverse activity could be maintained. The SA : V ratio of GCs decreased morphology of the GCs at (a) 21, (b–f) 7 or (g–h) 5 d post infection (dpi). – – as size increased, from values above 0.4 in the smallest cells to val- Note the remarkably irregular shapes in (a h). (i l) Three-dimensional reconstruction of GCs together with the nematode. A protruding GC end ues under 0.1 in the largest ones (Fig. 3a). Studies of their protein next to the nematode head is always observed. The shape of the content indicated that the maximum metabolic activity within nematode corresponds to the cavity left in each section by the nematode the GC coincides with the egg-laying stage (Bird, 1961), when body, as it is normally retracted during the embedding process.

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 7

Goto, 2011). In this respect, for the functional characterization of during in vivo monitoring of galls harboring GFP fusions (de Al- particular genes during gall formation, it might be worth examin- meida Engler et al., 2004). ing the cell wall ingrowths even if the GC size is not altered in The diameter of GCs within Arabidopsis galls may range from loss-of-function lines as compared with controls, as suggested by 100 to 500 lm at late developmental stages (40 dpi). They can Cabrera et al. (2014a). be even larger in galls within other plant hosts such as cowpea To further investigate GC morphology, we also calculated the (Vigna unguiculata (L.) Walp.) or tomato (Solanum lycopersicum sphericity of GCs at the different developmental stages studied. L.) (Gal et al., 2006; Das et al., 2008), hampering the generation Sphericity, a measure of the roundness of an object, is defined as of high-quality 3D projections of a complete GC for accurate the ratio of the surface area of a sphere (having the same volume volume quantification using 3D confocal projections of serial as the object of interest) to the surface area of the object; hence optical sections. Thus, 3D reconstructions of GCs with the the sphericity of a sphere should be 1 (Girshovitz & Shaked, method presented here proved to be very useful for accurate mea- 2012). Our results showed that the shape of GCs, measured in surement of GC volumes and identification of their morphologi- terms of their sphericity, seemed to be the same throughout gall cal characteristics. This also represents a step forward in the development (Fig. 3d). No spherical cells were found in our phenotyping of feeding cells, providing valuable and standardized study, as values of sphericity ranged between 0.4 and 0.7 for all information on the number, shape, position and volume of GCs the GCs characterized (Fig. 3d). In fact, the shape of some GCs and minimizing misinterpretations arising from the use of 2D was clearly more elongated than spherical (Figs 4, S1). As, for the images. The method could easily be applied to galls of different same volume, spherical cells will have slower diffusion rates than plant species having different thicknesses, providing that the tis- elongated cells (Young, 2006), the shape of GCs could be another sues are well fixed to maintain gall morphology in the sections. way to compensate for the low SA : V ratio of GCs at late devel- opmental stages. The low sphericity values were also consistent Correlation between pooled GC area in single sections and with the presence of abundant protuberances and crevices which total GC volume within a gall: a simplified standardized produced an irregular shape and prevented GCs from becoming phenotyping method spherical, as clearly illustrated in the 2D sections and 3D models (Figs 2, 4a–h, S1). This might be attributable to mechanical con- The abovementioned results demonstrate that GC phenotyping straints imposed by the presence and feeding activity of the nema- provides valuable information in addition to the currently per- tode and other vascular cells which push and pull the GCs as galls formed infection tests. While methods for studying nematode develop. However, some common features were identified among resistance or susceptibility with infection tests are fairly well the GCs, such as an elongated protuberance from the main body understood and established, no standardized methods exist for the of the GC adjacent to the nematode head (Fig. 2), which is more detailed phenotyping of GCs. In our study, the volumes of the easily appreciated in the 3D reconstruction (Fig. 4i–l). These pro- GCs induced by an RKN within a gall have been accurately mea- tuberances might be associated with the accumulation of micro- sured for the first time; this method provides an effective means of tubules and organelles close to the nematode head observed analyzing in depth the phenotypic differences between GCs.

Table 3 Compilation of studies on the interaction between root-knot nematodes and plants

Days post Number of Method Reference Species infection (dpi) samples

Comparison of the size of the GCs by de Almeida Engler et al. (2004) Arabidopsis (Arabidopsis thaliana)40 – visual observation from gall sections. Caillaud et al. (2008) 10 No measurements made Clement et al. (2009) 7, 21 de Almeida Engler et al. (2012) 7, 21 Pegard et al. (2005) Pepper (Capsicum annuum L.) 5 Anwar & McKenry (2007) Cotton (Gossypium hirsutum L.) – Souza Ddos et al. (2011) Tobacco (Nicotiana tabacum L.) 8 Measurement of GC diameter from sections Gal et al. (2006) Tomato (Solanum lycopersicum L.) 42 10 galls Das et al. (2008) Cowpea (Vigna unguiculata L.) 5, 9, 14, 19, 21 3 GCs Measurement of the area occupied by Banora et al. (2011) Arabidopsis (Arabidopsis thaliana) 7, 14, 21 60 GCs individual GCs in gall sections Vieira et al. (2012) 14, 18, 40 25 GCs Vieira et al. (2013) 7, 14, 21, 40 60 GCs Iberkleid et al. (2013) Tomato (Solanum lycopersicum L.) 5, 15, 28 50 GCs Antonino de Souza et al. (2013) Tobacco (Nicotiana tabacum L.) 14 21 GCs Measurement of the total area occupied Vovlas et al. (2005) Chickpea (Cicer arietinum L.) 30 24 sections by the pool of GCs within gall sections Portillo et al. (2013) Tomato (Solanum lycopersicum L.) 15 20 sections Cabrera et al. (2014a) Arabidopsis (Arabidopsis thaliana) 14 20 sections

Differences in the giant cells (GC) size between wild type and loss-of-function or overexpressing lines inferred from two-dimensional sections were investigated with different methods. The methods used to phenotype the GCs are explained in the first column. The plant species, dpi and number of GCs or galls scored are indicated in the other columns.

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 8 Research Phytologist

Recent studies (summarized in Table 3) investigated differ- and 3D structure of GCs. However, 3D reconstruction of GCs ences in GC size between wild type and loss-of-function or over- within a gall is a tedious, demanding and time-consuming task expressing lines by inference from 2D gall sections. Our results because of the need to process many sections per gall (Table 1; presented here demonstrate that there are multiple sources of Figs 1, 2; Videos S1, S2). Therefore, a simpler method is still misinterpretation of data from 2D gall sections, such as those needed. related to the number, position and shape of GCs (Fig. 2). Sev- To this end, we investigated the correlation between GC vol- eral of these studies did not provide quantitative measurement of umes obtained in this work and other parameters used to phe- the GCs, but compared their sizes by direct, qualitative observa- notype the GCs. The maximum diameter, provided by the tions of 2D sections (Table 3). Other studies collected various software after reconstruction, showed a poor Pearson correlation quantitative data from gall sections. Three different parameters index of 0.61 (P < 0.05) (Fig. 5a), whereas a much higher corre- were used: the maximum diameter of the GCs, the maximum lation (0.93; P < 0.05) was obtained between the volume and area occupied by individual GCs and the maximum area of all the area of the individual GCs measured in the section where GCs as a pool in selected gall sections (Table 3). The present they showed maximum expansion (considering maximum study has shown that GC morphology is extremely irregular expansion as the section where a specific GC showed its largest (Figs 4, S1). Thus, inferring differences in GC size from 1D or area from all the sections in which it was colored; Fig. 5b). The 2D parameters such as the diameter and area seems not to pro- maximum correlation indexes were obtained between the vide sufficient accuracy. By contrast, the volume data obtained in volume of the total pool (considering the sum of all the GCs) this study allowed us to determine with high accuracy the size contained in a gall and the area of the GC pool in the section

(a)

(b)

(c)

(d)

Fig. 5 Giant cell (GC) diameter, area and volume correlations. Correlations are shown between (a) the diameter and the volume of individual GCs, (b) the volume and the area of maximum expansion of individual GCs, (c) the volume of the total GC pool within a gall and the GC pool area in the section showing its maximum expansion, and (d) the volume of the GC pool within a gall and the average GC pool area in the 10 sections where GCs showed their maximum expansion. For all Pearson correlation coefficients shown, P < 0.05. Left panels, sections with those cells measured colored; middle panels, three- dimensional reconstructions; right panels, graphs representing the correlations described.

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 9 showing maximum expansion (0.97; P < 0.05; Fig. 5c). Interest- demonstrated here (Fig. 3a; Table 2) are not taken into account. ingly, a similarly high correlation value (0.97; P < 0.05) was also If the whole volume occupied by the GC pool within a gall is not obtained between the volume of the GC pool contained in a measured, this may introduce significant errors. Nevertheless, the gall and the average area of the pool of GCs from the 10 sec- choice of the section with maximum area occupied by the GC tions where the pool of GCs showed their maximum expansion pool also would require the measurement of all the sections of a (Fig. 5d). gall to avoid misinterpretations leading to errors. Extrapolation of the volumes from the 2D parameters pre- For all these reasons, we propose a simplified method that sented some technical difficulties. For instance, the identification minimizes these problems and speeds up the processing consider- of sections where GCs show their maximum expansion area is ably. The method is based on a 2D parameter that shows a high difficult if a previous labeling and alignment of the correlative correlation with the final volume of the GC pool. We showed sections obtained from a gall is not performed. In some of the ref- that the average area occupied by all the GCs within a gall from erences given in Table 3, only two to three GCs per gall were the 10 sections with the maximum expanded GC pool had a high chosen for area measurement. If individual GCs are selected, the correlation with the pooled GC volume (Fig. 5d). By using an striking differences in volume among GCs of the same age algorithm in which the areas of two of these 10 sections with

(a)

(b)

(c) (d)

Fig. 6 Schematic representation of the simplified method for giant cell (GC) size phenotyping. Serial sections of the two main types of gall (a, b) detected in this study (one-dome or two-dome shape). The 10 sections with the largest GC expanded area are colored. (c) Correlation between the average GC pool area of the two sections of 4 lm with maximum GC expansion (y-axis) and their volume (x-axis) was high. (d) The same as (c) but with sections of 8 lm. Pearson correlation coefficients: P < 0.05.

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 10 Research Phytologist

maximum expanded GC pool are randomly selected with 1000 were sectioned at 2 lm (shown with a high area : volume correla- iterations (Wichman & Hill, 1982), we obtained high correla- tion; Figs 5d, 6c). Similar results were obtained for 8-lm sections tions between the two parameters (pooled GC area and volume): (Fig. 6d) and for 5–7-lm sections (data not shown). Thus, by a minimum correlation index of 0.968 and a maximum of 0.972. selecting and measuring the area of the GC pool in the two The identification of two random sections among those 10 sec- sections with the highest GC expansion from either 4- or 8-lm- tions with the largest area occupied by the GC pool within a thick sections (easily identified at first glance, as the differences completely sectioned gall can be performed easily, accurately and between sections are greater than at 2 lm), the correlation index with minimum errors. To further facilitate large-scale GC pheno- with the GC volume was maintained at 0.97 (P < 0.05) in both typing, the number of sections obtained from a gall should be cases (Fig. 6c,d). This simplified method that we propose reduces minimized, as processing an entire gall into 2-lm-thick sections the number of gall sections necessary for large-scale phenotyping is tedious work. Using 4–8-lm-thick sections, the number of sec- while maintaining a very high accuracy. The method is based on tions for each gall would dramatically decrease (two to four times a 2D parameter (the area of the GC pool within a gall) that we less than at 2 lm); for example, from 100 sections to 50 or 25. demonstrated to be proportional to the size measured as the total Moreover, differences in the area occupied by the GC pool GC volume developed by M. javanica inside the gall. Therefore, among thicker sections are easily determined by simple visual this method will be useful for comparison of GC phenotypes in inspection. Fig. 6 shows schematically the reduction in the num- terms of their size among different plant genotypes and/or plant ber of sections obtained from two different types of gall (a and species. b), which were artificially classified depending on their GC area distribution. In Fig. 6(a), the sections with the maximum GC A proof of concept: simplified GC phenotyping method area are correlative, whereas in (b), there are two groups of sec- applied to a specific case in Arabidopsis tions with maximum areas. Those shapes were selected among the 162 GCs scored, as shown in more detail in Fig. S2. When a In order to confirm the utility of the simplified method, we gall is processed in 4-lm-thick sections, and sections with the compared the size of GCs from galls induced by M. javanica in maximum GC area are selected, at least two of the sections will the transgenic line J0121 ≫ DTA (where GCs are genetically be among those 10 with the maximum GC area from the gall if it ablated by the expression of the diphtheria toxin A in the GCs)

(a)

(b)

Fig. 7 A proof of concept: simplified giant cell (GC) phenotyping method applied to a specific case. (a) Toluidine-stained 4-lm representative sections of galls from each Arabidopsis line as indicated: J0121 ≫ GFP and J0121 ≫ DTA. Bars, 50 lm. Asterisks indicate GCs (b) Histograms indicate the average volume occupied by the GC pool from two representative galls of each line as (c) indicated (Æ SE). (c) The average volume and the average area of the two largest sections per gall for each genotype are also indicated in the small table underneath. dpi, days post inoculation.

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 11 with that in the control line J0121 ≫ GFP (expressing GFP two lines (c. 2-fold) was obtained as when using their total GC within the GCs) [Correction added after online publication 22 volumes (Fig. 7c). This is expected from the high correlation (Pear- January 2015: the preceding sentence has been corrected]. In a son correlation coefficient R = 0.97; P < 0.05) between the two previous work, we found that the GC pool in J0121 ≫ DTA was parameters that we found (see previous section). Thus, we con- smaller than that in the control line J0121 ≫ GFP, based on the firmed that a 2D measurement (area of the GC pool from the two GC pooled area measured in 4-lm-thick single sections (Cabrera sections with the maximum expansion area of the GCs) selected by et al., 2014b); however, 3D reconstructions were not generated. visual inspection is a valid parameter to use to compare GC sizes In the present work, we reconstructed J0121 ≫ DTA GCs from between two different plant genotypes. The shapes of the individ- 14-dpi galls and from the control J0121 ≫ GFP (Fig. 7a) using ual GCs from the J0121 ≫ GFP control line and the 4-lm sections. The average volume occupied by the GC pool J0121 ≫ DTA line (Fig. 8a,b) were similar, and no differences in was at least 2-fold larger in the control line than in J0121 ≫ DTA the typically irregular shape were evident. (Fig. 7b). When the GC pool areas from the two sections with the This example further confirms that the simple, fast and easy maximum GC expansion (Fig. 7c) among all sections obtained standardized method proposed here is very effective for detection from each gall were compared, the same difference between the of phenotypic differences between GCs from different

(a)

(b)

Fig. 8 Three-dimensional reconstruction of individual giant cells (GCs) from two Arabidopsis genotypes. Figure shows the different shapes of the GCs reconstructed from (a) the J0121 ≫ GFP control line, and (b) the J0121 ≫ DTA line. Note that all GCs from both lines show the typically irregular shape described for the wild-type cells in Supporting Information Fig. S1.

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com New 12 Research Phytologist

Arabidopsis lines, with high confidence that the results obtained Banora MY, Rodiuc N, Baldacci-Cresp F, Smertenko A, Bleve-Zacheo T, et al. are representative of the differences in the total volume occupied Mellilo MT, Karimi M, Hilson P, Evrard JL, Favery B 2011. Feeding by the GCs. cells induced by phytoparasitic nematodes require gamma-tubulin ring complex for microtubule reorganization. PLoS Pathogens 7: e1002343. In conclusion, there is an increasing need for standardized Barcala M, Garcia A, Cabrera J, Casson S, Lindsey K, Favery B, Garcia-Casado methods to accurately phenotype GCs, particularly when evaluat- G, Solano R, Fenoll C, Escobar C. 2010. Early transcriptomic events in ing gene function in mutant or transgenic plants. Here, we microdissected Arabidopsis nematode-induced giant cells. Plant Journal 61: – obtained for the first time 3D models and volumes of 162 GCs 698 712. from galls induced in Arabidopsis by M. javanica. These data Barcala M, Garcia A, Cubas P, Almoguera C, Jordano J, Fenoll C, Escobar C. 2008. Distinct heat-shock element arrangements that mediate the heat shock, provided the basis on which to develop a simple method based but not the late-embryogenesis induction of small heat-shock proteins, correlate on 2D images of thin and thick gall sections that showed a high with promoter activation in root-knot nematode feeding cells. Plant Molecular correlation to the volume of the reconstructed GCs. The utility Biology 66: 151–164. Bird AF. 1961. The ultrastructure and histochemistry of a nematode-induced of the simplified method was confirmed in a particular transgenic – line compared with a control line (J0121 ≫ DTA versus giant cell. Journal of Biophysical and Biochemical Cytology 11: 701 715. ≫ Bleve-Zacheo T, Melillo LT. 1997. The biology of giant cells. In: Fenoll C, J0121 GFP). Therefore, we showed that the method proposed Grundler FMW, Ohl SA, eds. Cellular and molecular aspects of plant–nematode permits data to be extracted with high confidence with respect to interactions. Dordrecht, the Netherlands: Kluwer Academic, 65–79. the volume of cells, considering the complexity and highly irregu- Cabrera J, Barcala M, Fenoll C, Escobar C. 2014a. Transcriptomic signatures of lar shape of GCs. Putative methods for reconstruction and vol- transfer cells in early developing nematode feeding cells of Arabidopsis focused ume measurements using confocal microscopy, two-photon on auxin and ethylene signaling. Frontiers in Plant Science 5: 107. Cabrera J, Diaz-Manzano FE, Sanchez M, Rosso MN, Melillo T, Goh T, Fukaki microscopy or SPIM would have to be adapted to the gall thick- H, Cabello S, Hofmann J, Fenoll C et al. 2014b. A role for LATERAL ness and optical characteristics of the different tissues in the galls ORGAN BOUNDARIES-DOMAIN 16 during the interaction of each plant species, which represents a serious limitation. By Arabidopsis-Meloidogyne spp. provides a molecular link between lateral root – contrast, the simple method that we developed, using Arabidopsis and root-knot nematode feeding site development. New Phytologist 203: 632 as a model, could easily be extended to different plant species and 645. Caillaud M-C, Lecomte P, Jammes F, Quentin M, Pagnotta S, Andrio E, de growing conditions as it is independent of gall thickness or trans- Almeida Engler J, Marfaing N, Gounon P, Abad P et al. 2008. MAP65-3 parency. microtubule-associated protein is essential for nematode-induced giant cell ontogenesis in Arabidopsis. Plant Cell 20: 423–437. Cardona A, Saalfeld S, Schindelin J, Arganda-Carreras I, Preibisch S, Longair Acknowledgements M, Tomancak P, Hartenstein V, Douglas RJ. 2012. TrakEM2 software for neural circuit reconstruction. PLoS ONE 7: e38011. This work was supported by grants from the Spanish Govern- Chakraborty A, Perales MM, Reddy GV, Roy-Chowdhury AK. 2013. Adaptive ment to C.E. (AGL2010-17388) and C.F. (CSD2007-057 and geometric tessellation for 3D reconstruction of anisotropically developing cells PCIN-2013-053) and the Castilla-La Mancha Government to in multilayer tissues from sparse volumetric microscopy images. PLoS ONE 8: C.F. and C.E. (PCI08-0074-0294). We are very grateful to Dr e67202. Clement M, Ketelaar T, Rodiuc N, Banora MY, Smertenko A, Engler G, Abad Albert Cardona for his valuable help and training with the P, Hussey PJ, de Almeida Engler J. 2009. Actin-depolymerizing TRAKEM2 program and to Marıa Luisa Garcıa for her technical factor2-mediated actin dynamics are essential for root-knot nematode infection assistance with the sectioning. of Arabidopsis. Plant Cell 21: 2963–2979. Das S, DeMason DA, Ehlers JD, Close TJ, Roberts PA. 2008. Histological characterization of root-knot nematode resistance in cowpea and its relation to – References reactive oxygen species modulation. Journal of Experimental Botany 59: 1305 1313. Abad P, Favery B, Rosso MN, Castagnone-Sereno P. 2003. Root-knot nematode Escobar C, Barcala M, Portillo M, Almoguera C, Jordano J, Fenoll C. 2003. parasitism and host response: molecular basis of a sophisticated interaction. Induction of the Hahsp17.7G4 promoter by root-knot nematodes: Molecular Plant Pathology 4: 217–224. involvement of heat-shock elements in promoter activity in giant cells. de Almeida Engler J, Kyndt T, Vieira P, Van Cappelle E, Boudolf V, Sanchez V, Molecular Plant–Microbe Interactions 16: 1062–1068. Escobar C, De Veylder L, Engler G, Abad P et al. 2012. CCS52 and DEL1 Fuller VL, Lilley CJ, Urwin PE. 2008. Nematode resistance. New Phytologist genes are key components of the endocycle in nematode-induced feeding sites. 180:27–44. Plant Journal 72: 185–198. Gal TZ, Aussenberg ER, Burdman S, Kapulnik Y, Koltai H. 2006. Expression of de Almeida Engler J, Van Poucke K, Karimi M, De Groodt R, Gheysen G, a plant expansin is involved in the establishment of root knot nematode Engler G, Gheysen G. 2004. Dynamic cytoskeleton rearrangements in giant parasitism in tomato. Planta 224: 155–162. cells and syncytia of nematode-infected roots. Plant Journal 38:12–26. Gamborg OL, Miller RA, Ojima K. 1968. Nutrient requirements of suspension Antonino de Souza JD Jr, Ramos Coelho R, Tristan Lourenco I, da Rocha cultures of soybean root cells. Experimental Cell Research 50: 151–158. Fragoso R, Barbosa Viana AA, Pepino Lima, de Macedo L, Mattar da Silva Girshovitz P, Shaked NT. 2012. Generalized cell morphological parameters MC, Gomes Carneiro RM, Engler G et al. 2013. Knocking-down Meloidogyne based on interferometric phase microscopy and their application to cell life incognita proteases by plant-delivered dsRNA has negative pleiotropic effect on cycle characterization. Biomedical Optics Express 3: 1757–1773. nematode vigor. PLoS ONE 8: e85364. Grandjean O, Vernoux T, Laufs P, Belcram K, Mizukami Y, Traas J. 2004. In Anwar SA, McKenry MV. 2007. Variability in reproduction of four populations vivo analysis of cell division, cell growth, and differentiation at the shoot apical of Meloidogyne incognita on six cultivars of cotton. Journal of Nematology 39: meristem in Arabidopsis. Plant Cell 16:74–87. 105–110. Grundler FMW, Bockenhoff€ A. 1997. Physiology of nematode feeding and Atkinson HJ, Lilley CJ, Urwin PE. 2012. Strategies for transgenic nematode feeding sites. In: Fenoll C, Grundler FMW, Ohl SA, eds. Cellular and control in developed and developing world crops. Current Opinion in molecular aspects of plant–nematode interactions. Dordrecht, the Netherlands: Biotechnology 23: 251–256. Kluwer Academic Publishers, 107–119.

New Phytologist (2015) Ó 2015 The Authors www.newphytologist.com New Phytologist Ó 2015 New Phytologist Trust New Phytologist Research 13

Iberkleid I, Vieira P, de Almeida Engler J, Firester K, Spiegel Y, Horowitz SB. Vanhaeren H, Gonzalez N, Inze D. 2010. Hide and seek: uncloaking the 2013. Fatty acid-and retinol-binding protein, Mj-FAR-1 induces tomato host vegetative shoot apex of Arabidopsis thaliana. Plant Journal 63: 541–548. susceptibility to root-knot nematodes. PLoS ONE 8: e64586. Vieira P, Engler G, de Almeida Engler J. 2012. Whole-mount confocal imaging Jones MGK, Dropkin VH. 1976. Scanning electron microscopy in of nuclei in giant feeding cells induced by root-knot nematodes in Arabidopsis. nematode-induced giant transfer cells. Cytobios 15: 149–161. New Phytologist 195: 488–496. Jones MGK, Goto DB. 2011. Root-knot nematodes and giant cells. In: Jones J, Vieira P, Escudero C, Rodiuc N, Boruc J, Russinova E, Glab N, Mota M, De Gheysen G, Fenoll C, eds. Genomics and molecular genetics of plant–nematode Veylder L, Abad P, Engler G et al. 2013. Ectopic expression of Kip-related interactions. Dordrecht, the Netherlands: Springer, 83–102. proteins restrains root-knot nematode-feeding site expansion. New Phytologist Jones MGK, Gunning BES. 1976. Transfer cells and nematode induced giant 199: 505–519. cells in Helianthemum. Protoplasma 87: 273–279. Vovlas N, Rapoport HF, Jimenez Diaz RM, Castillo P. 2005. Differences in Jones MGK, Northcote DH. 1972. Multinucleate transfer cells induced in coleus feeding sites induced by root-knot nematodes, Meloidogyne spp., in chickpea. roots by the root-knot nematode, Meloidogyne arenaria. Protoplasma 75: 381– Phytopathology 95: 368–375. 395. Wasson AP, Ramsay K, Jones MG, Mathesius U. 2009. Differing Kayes JM, Clark SE. 1998. CLAVATA2, a regulator of meristem and organ requirements for flavonoids during the formation of lateral roots, nodules development in Arabidopsis. Development 125: 3843–3851. and root knot nematode galls in Medicago truncatula. New Phytologist 183: Kyndt T, Vieira P, Gheysen G, de Almeida-Engler J. 2013. Nematode feeding 167–179. sites: unique organs in plant roots. Planta 238: 807–818. Wichman BA, Hill ID. 1982. Algorithm AS 183: an efficient and portable Lilley CJ, Wang D, Atkinson HJ, Urwin PE. 2011. Effective delivery of a pseudo-random number generator. Applied Statistics 31: 188–190. nematode-repellent peptide using a root-cap-specific promoter. Plant Young KD. 2006. The selective value of bacterial shape. Microbiology and Biotechnology Journal 9: 151–161. Molecular Biology Reviews 70: 660–703. Orion D, Wergin WP. 1982. Chloroplast differentiation in tomato root galls induced by the root knot nematode Meloidogyne incognita. Journal of Nematology 14:77–83. Supporting Information Pegard A, Brizzard G, Fazari A, Soucaze O, Abad P, Djian-Caporalino C. 2005. Histological characterization of resistance to different root-knot nematode Additional supporting information may be found in the online species related to phenolics accumulation in Capsicum annuum. Phytopathology version of this article. 95: 158–165. Portillo M, Cabrera J, Lindsey K, Topping J, Andres MF, Emiliozzi M, Oliveros Fig. S1 The full 3D models or reconstructions of the 162 indi- JC, Garcia-Casado G, Solano R, Koltai H et al. 2013. Distinct and conserved transcriptomic changes during nematode-induced giant cell development in vidual GCs scored in this study. tomato compared with Arabidopsis: a functional role for gene repression. New Phytologist 197: 1276–1290. Fig. S2 Area distribution of the GC pool from representative Portillo M, Lindsey K, Casson S, Garcia-Casado G, Solano R, Fenoll C, Escobar reconstructed galls used in the study. C. 2009. Isolation of RNA from laser-capture-microdissected giant cells at early differentiation stages suitable for differential transcriptome analysis. Molecular Plant Pathology 10: 523–535. Video S1 Video file explaining the three basic steps carried out Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, during 3D reconstruction of GCs. Preibisch S, Rueden C, Saalfeld S, Schmid B et al. 2012. Fiji: an open-source platform for biological-image analysis. Nature Methods 9: 676–682. Video S2 Video file showing the accurate identification of the Schmid B, Schindelin J, Cardona A, Longair M, Heisenberg M. 2010. A GC number during the labeling process, avoiding misinter- high-level 3D visualization API for Java and ImageJ. BMC Bioinformatics 11: 274. pretations. Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to ImageJ: 25 years of image analysis. Nature Methods 9: 671–675. Please note: Wiley Blackwell are not responsible for the content Souza Ddos S, de Souza JD Jr, Grossi-de-Sa M, Rocha TL, Fragoso RR, Barbosa or functionality of any supporting information supplied by the et al. AE, de Oliveira GR, Nakasu EY, de Sousa BA, Pires NF 2011. Ectopic authors. Any queries (other than missing material) should be expression of a Meloidogyne incognita dorsal gland protein in tobacco accelerates the formation of the nematode feeding site. Plant Science 180: 276–282. directed to the New Phytologist Central Office.

Ó 2015 The Authors New Phytologist (2015) New Phytologist Ó 2015 New Phytologist Trust www.newphytologist.com Integrated Discussion Giant cells induced by Meloidogyne spp. root-knot nematodes inside the vascular cylinder of the host plant roots constitute an astonishing biological model regarding the fine-tuned interaction between two eukaryotic organisms. The nematode needs to alter plant cells to feed, to develop and finally to complete its life cycle and the plant try to counteract the physiological damage. Adjustments occur at the molecular level within cells that differentiate from vascular root cells into highly specialized feeding cells that become completely different morphologically and metabolically from their stem cells. These nematodes sorted the plant defences by manipulating the metabolism of the plant with their secretions. Understanding the molecular mechanisms underlying this interaction becomes a challenge from the biological point of view. It requires the study on different fields like the cell metabolism, the hormonal regulation, the secondary metabolism, the mechanisms of cell expansion, the cell cycle regulation, the meristem maintenance, the epigenetic regulation of the transcription among others (Cabrera et al., 2015a; Escobar et al., 2015).

It is highly plausible to think that the differentiation of regular vascular root cells into these highly specialised feeding structures is accompanied by vast changes in their gene expression. This has been the main topic of the research developed in this thesis, to deepen into the knowledge of the changes in the gene expression that allow the differentiation of the giant cells and/or galls from root cells. In this respect, we obtained the transcriptomes of the giant cells and galls, pseudo organs where they are imbibed, both induced by M. javanica in two model plants, Arabidopsis and tomato (Barcala et al., 2010; Portillo et al., 2013). We considered important to differentiate between the transcriptomes of hand dissected galls and the transcriptomes of the giant cells induced by the nematode inside the gall as only the giant cells eventually nourish the nematode. We combined complex cell biology techniques as laser capture microdissection with microarrays to obtain the transcriptomes of giant cells and galls at 3 days post infection in Arabidopsis (Barcala et al., 2010) and the transcriptomes of giant cells at 3 and 7 days post infection and galls at 1, 3, 7 and 14 days post infection in tomato (Portillo et al., 2013). The transcriptomes of the giant cells and the galls, were different. In tomato, as in Arabidopsis, less than half of giant cell differentially expressed genes at 3 and 7 days post infection were also detected as differentially expressed in galls at the same developmental stage (Portillo et al., 2013). Moreover, we could observe a dilution effect as those genes with the highest fold changes (with positive or negative tendencies) in giant cells were commonly found as differentially expressed in the galls, but with lower fold changes than in giant cells, while those genes with the smallest fold changes were mostly distinctive from giant cells. This has been already demonstrated by other authors with other plant or nematode species (Damiani et al., 2012; Ji et al., 2013) confirming the importance of the giant cells isolation to obtain their specific transcriptome.

We observed a general down- regulation among those genes differentially expressed, in early developing giant cells in tomato, fact that was already observed in Arabidopsis where 851 of 1161 genes were down- regulated (Barcala et al., 2010). Moreover, when the transcriptomes of the giant cells from both species, tomato and arabidopsis, were compared with a reciprocal blast, we observed that most of the commonly regulated genes were down-regulated in both species while only a few genes were commonly up- regulated (Portillo et al., 2013). This was very notorious in the case of the secondary metabolism genes such as those from the synthesis of phenylpropanoids, defence related genes such as those encoding peroxidases that were commonly down- regulated in both species. One example is the peroxidase TPX- down- regulated in the giant cells of Arabidopsis and tomato which repression seemed to have a role during the infection. Accordingly, the expression of this peroxidase in the Mi-1 resistant line Motelle, was high,while in the susceptible line it was down- regulated. Moreover, the over- expression of TPX-1 in the susceptible background Moneymaker yielded a reduction in the number of infection or egg/masses and the giant cells formed in this line were smaller than those formed in the control line (Portillo et al., 2013).

Therefore, the transcriptomes of the giant cells induced by M. javanica in Arabidopsis and tomato were distinct to that of galls and an exclusive characteristic of giant cells was the shut-down of defence-related genes. This gene repression seems to play a role for the success of the infection. A general gene repression previous to the differentiation to a specific cell type has been widely study in other biological systems in animals (Boyer et al., 2005; Loh et al., 2006) and plants (Berdasco et al., 2008; Yadav et al., 2013). This down- regulation could be achieved by processes that regulate gene expression such as differential regulation of transcription factors (Yadav et al., 2013)or by epigenetic changes as modifications in the state of the DNA methylation (Berdasco et al., 2008). Gene silencing could be at least partly achieved also by the participation of small RNAs that silence the genes at both translational, transcriptional level and also mediateepigenetic mechanisms (Axtell, 2013). Hence, to investigate the putative function of the small RNAs in the massive gene silencing observed in the transcriptome of giant cells, we performed 6 independent sRNAs libraries by using the Illumina Solexa sequencing by synthesis technology (Cabrera et al., 2015b). The total numbers of sequences obtained for the 6 libraries were around 10M and 99% of these sequences corresponded to sequences between 18 and 29 nucleotides, most of them ranging from 20 to 24 nucleotides, which is the typical length of sRNAs processed by DICER in Arabidopsis. The percentage of sequences with 24 nucleotides was consistently higher in the three replicates of galls while the percentage of sequences with 21 nucleotides, which is the typical length of miRNAs, was higher in the control roots (Cabrera et al., 2015b). The putative role of both classes of sRNAs in the gene silencing observed in giant cells was further studied. From the 338 known miRNAs in Arabidopsis, we found a total of 62 miRNAs differentially expressed between galls and uninfected roots. Accordingly to the lower abundance of 21 nucleotides sequences in galls than in controls, we found only 11 miRNAs induced against the 51 miRNAs that showed down- regulation in the galls. Large families of miRNAs as miR166 and miR169 were consistently down- regulated in galls. In general, the 50% of the down- regulated genes presented FCs under -3-fold, while the up- regulated genes were mostly under +3-fold. Therefore, there is also a strong trend to the down- regulation in the miRNAs (Cabrera et al., 2015b).

We deepen in the study of the miRNA390a, one of the most abundant induced miRNA in galls. MiRNA390 is induced by auxins in the xylem pole pericycle cells and triggers the formation of another small RNA, the TAS3-derived trans-acting short- interfering RNAs (tasiRNAs; Marin et al., 2010). TasiRNA3a target the degradation of ARF2, ARF3 and ARF4 mRNAs and allow in this way lateral root growth (Marin et al., 2010). The activation pattern of the pMIR390a::GUS further confirmed our sequencing results, as it was specifically induced in early developing galls and giant cells (Cabrera et al., 2015b). Its expression overlapped with that of the promoter of the gene encoding the tasiRNA3a in the founder cells of the xylem pole pericycle cells and in the centre of the galls. We showed as well that the mature forms of both sRNAs were activated in galls by qPCR (miRNA390a was induced in galls) and by the use of a sensor line (tasiRNA3a; Cabrera et al., 2015b). Infection tests with mutant lines with a reduce production of both sRNAs showed a significant decrease in the number of infections and, moreover, giant cell size was reduced in the case of the mutant mir390a (Cabrera et al., 2015b; Cabrera et al., 2015c). For the first time we described the activation and function of these two sRNAs silencing pathways, miRNAs and tasiRNAs, during the development of the galls and giant cells induced by root-knot nematodes in plants (Cabrera et al., 2015b).

The second main difference in the sRNA population between galls and giant cells was that the proportion of 24 nucleotides sequences in the galls libraries was consistently higher than in the root libraries (Cabrera et al., 2015b). 24 nucleotides mostly correspond to repeat associated small interfering RNA (rasiRNA) that are related with the gene silencing at the DNA level mediating epigenetic processes (Pumplin and Voinnet, 2013). A differential expression analysis between galls and control roots showed 56455 sequences differentially expressed (P<0.05) in galls as compared to control roots and from them a 76% were either exclusive or up- regulated in galls (Cabrera et al., 2015b). These results compared to those relative to the sRNAs of 20-21 nt described above, mostly down-regulated in galls, indicated a contrasted regulation between the 20-21 nt sRNAs and those 24 nt sRNAs in galls. Therefore, the results presented, indicated that the abundance of the 24nt sRNAs (most probable rasiRNAs), constitutes a galls hallmark that looks promising to explain the general down- regulation found in the early-developing giant cells transcriptomes as they actively participate in the RNA-directed DNA methylation. This epigenetic mechanism has been already implicated in other stresses and processes (Matzke and Mosher, 2014)

We generated a huge amount of transcriptomic data coming from the microarrays analysis and the sRNAs sequencing experiments, becoming difficult to classify the information and to compare them with other transcriptomes. It was also arduous to select target genes for further functional studies among those genes differentially expressed in the transcriptomes of plant- nematode interaction. Hence, we developed NEMATIC, an Excel based database in which we have compiled around 9000 thousand genes that have been described as differentially expressed in different Arabidopsis interactions with endoparasitic nematodes (Cabrera et al., 2014a). We also, added valuable information from other publications to aid the understanding of the plant-nematode the interaction, i.e. those genes differentially expressed under different hormonal treatments, or those enriched in different cell cycle phases or root cells, as well as all the transcriptional regulators described in Arabidopsis and all the genes categorized according to Mapman into functional categories. Additionally, the expression values in thousands of transcriptomes from several conditions for all the genes differentially expressed in the microarrays were downloaded from Genevestigator (Cabrera et al., 2014a).

Different proofs of concepts for the use of NEMATIC were performed along my PhD, one of them was to try to identify transcriptomic signatures from different root cell types that might be also present in the transcriptome of giant cells. For that the up- regulated genes in giant cells and galls at 3days post infection in Arabidopsis were compared with those genes that were characteristic of the different root cell types (Cabrera et al., 2014b). The up- regulated genes in giant cells and galls were characteristic of undifferentiated cells of the root such as those from the quiescent center and the xylem pole pericycle cells that divide to originate a new lateral root primordium. Furthermore, the resistance to M. incognita of the auxin-insensitive tomato mutant dgt (Richardson & Price, 1982), which lack lateral roots, as well as the similarities in the gene expression of plant transcription factors (e.g. Medicago truncatula homologues to PHANTASTICA and Class I knotted, Mt-phan and Mt-knox-1, respectively) and cell cycle regulators during the development of galls, lateral roots and nodules, point to a correlation between these processes (Goverse et al., 2000; Mathesius, 2003; Moreno-Risueno et al., 2010). Additionally, 39 out of 103 promoter tag lines displaying a distinct response to nematode infection also exhibited activity at lateral root initiation sites (Barthels et al., 1997). Taken together, these data reinforce the similarity between the two processes. Hence, we selected two enhancer trap marker lines for the xylem pole pericycle cells (xylem pole pericycle) that divide to produce a lateral root primordial (Cabrera et al., 2014b). The line J0192 is specific for the xylem pole pericycle cells dividing to generate a new lateral root and the J0121 is a marker of the xylem pole pericycle along all the root and is also active during the first divisions of the lateral root primordia. When these lines were infected with M. javanica, GFP in both lines was expressed in the whole galls in addition to the xylem pole pericycle (Cabrera et al., 2014b). J0192 expression started in the xylem pole pericycle in a couple of cells near to the nematode head and eventually, the whole center of the gall was expressing GFP. The expression was different to that obtained during the lateral root development as we could found expression in both sides of the xylem pole pericycle and additionally, the GFP spreads from the pericyle inwards to the center of the galls. In contrast, during lateral root formation the signal starts only in one side of the pericycle and increases from the pericycle outwards (Cabrera et al., 2014b).

In the marker line J0192 the GFP expression is driven by the promoter of the gene LBD16 (Laplaze et al., 2005). We observed the same expression than J0192 in a line in which a region of 2500 bp of the promoter of LBD16 is fused to GUS (Cabrera et al., 2014b). GUS expression was found at 3 and 7 days post infection in the centre of the galls and decreased at 14 days post infection. In the same way that the previously mentioned MIR390a (Marín et al., 2010), LBD16 is an auxin responsive gene that has been demonstrated to have a role during the lateral root formation and is activated by auxin response factors (Okushima et al., 2007). To check whether the induction of LBD16 was due to the presence of auxins in the galls we compare the induction of the LBD16 with that of the auxin marker DR5. Both lines were activated in the center of the galls at early infection stages, the induction of DR5 was maintained in the time up to 40 days post infection (Cabrera et al., 2014b). Moreover, when we treated the promoter LBD16 line with an auxin signaling inhibitor, the GUS expression for LBD16 disappeared. Thus, the LBD16, a key transcription factor, during the lateral root formation (Okushima et al., 2007) is activated in the galls by auxins (Cabrera et al., 2014b). We also tested whether the nematodes produce any secretion that might trigger LBD16 promoter activation in protoplasts of the line J0192<< GFP as a pair of cells started to react in the pericycle soon after the nematode penetration in the vascular cylinder. As expected, the LBD16 was activated by the presence of auxins as a positive control, but surprisingly it wasactivated as well by nematode secretions (Cabrera et al., 2014b). Moreover, knock-out lines for the LBD16 challenged with M. javanica yielded up to a 60% of reduction in the infection (Cabrera et al., 2014b). Additionally, the giant cells inside the galls formed were smaller than those from the control line, as showed by our giant cell phenotyping method after measuring the area of the giant cells (Cabrera et al., 2014b; Cabrera et al., 2015c). Even a higher reduction on the percentage of infection was obtained when we used a genetic ablation line based on the J0121 marker line expressing the diphtheria toxin. Almost no infection was obtained in this line and the giant cells developed in these galls were significantly smaller than those from the control (Cabrera et al., 2014b). Additionally, no induction of LBD16 could be detected at any developmental stage of the syncytia formation and consequently, the mutant lines seemed not to affect the nematode reproduction (Cabrera et al., 2014b). These indicated that LBD16 is functionally relevant for root- knot nematodes but not for cyst nematodes.

Therefore, the xylem pole pericycle cells are crucial during the formation of the galls and the giant cells. Furthermore, LBD16 and miRNA390a integrated in signaling cascades mediated by auxin (Okushima et al., 2007; Marin et al., 2010) are important players during the formation of the lateral root galls and giant cells (Cabrera et al., 2014b; Cabrera et al., 2015b).

We also used NEMATIC to try to understand molecular differences regarding gene expression patterns between the feeding sites of two different groups of endoparasitic nematodes, the giant cells and the syncytia (Cabrera et al., 2014a; Cabrera et al., 2015d). In one hand, we compared the transcriptomes of giant cells (Barcala et al., 2010) and microaspirated syncytia (Szakasits et al., 2009) with the genes differentially expressed under different hormone treatment generated by Joan Chory lab (Nemhauser et al., 2007). While the giant cells seem to have more genes induced by auxins, in the case of syncytia there were more genes induced by cytokinins (Cabrera et al., 2014a). Moreover, we downloaded from Genevestigator the genes commonly or oppositely regulated with LBD16 in thousands of transcriptomes (Cabrera et al., 2015d). The commonly regulated genes were mostly up- regulated in galls and down- regulated in syncytia and the opposite regulated genes were mostly down- regulated in galls and up- regulated in syncytia (Cabrera et al., 2015dThe expression profiles of these genes in giant cells, galls and microaspirated syncytia with other transcriptomes were compared. The analysis showed that the expression profiles in giant cells and galls were more similar to that of undifferentiated callus cells while the expression profiles in syncytia were more similar to callus differentiating to shoots. This in silico analysis pointed again to the putative role of auxins in the galls and the importance of cytokinin in the syncytia (Cabrera et al., 2015d).

In conclusion, we found in our transcriptomic analyses a massive gene repression in the giant cells that seemed to be conserved among different species and could have a role for the proper establishment of the nematode feeding cells. At least partly, this massive gene silencing could be explained by the participation of sRNAs such as miRNAs, tasiRNAs and rasiRNAs, as our sequencing and functional analyses showed. A classification tool, NEMATIC, based on massive information from transcriptional profiles during the plant-nematode interaction assisted to identify that signatures of undifferentiated cells from roots as the quiescent center and lateral root primordia showed common characteristics to that of galls/giant cells. A strong parallelism between the lateral root development mediated by auxins in those founder cells from the xylem pole pericycle and the development of the galls and giant cells was also described by functional and expression analysis of lateral root primordia marker lines, among them a crucial transducer of lateral root formation, LBD16. Finally, the comparison of the co-regulated genes with LBD16 in galls and giant cells to that of other transcriptomes opened a new research field, as there were also high similarities to that of callus formation, a new organogenesis process that is also exogenously induced by altering the auxin/cytokinin balance. CONCLUSIONS 1. Gene reprogramming in the nematode feeding cells, giant cells, induced by root- knot nematodes in the roots differs from that of the galls where they are imbibed in tomato and Arabidopsis. 2. Gene repression is a hallmark of the giant cells transcriptomes, being common to tomato and Arabidopsis. Genes related to plant defence or to the secondary metabolism were found commonly and significantly down- regulated in the giant cell transcriptomes from both species, but not in galls. 3. Gene repression in the giant cells induced in Arabidopsis is at least partly mediated by regulatory small RNAs. The role of miRNA390a and the trans- acting small interfering RNA3a during galls formation and giant cell development was demonstrated. 24 nucleotides long small RNAs showed a significant up- regulation in the galls which may be mediating gene repression operating by epigenetic regulation. 4. The processes of lateral root development and gall and/or giant cell formation share molecular pathways mediated by auxins. In particular, two key components of the molecular mechanisms mediated by auxins leading to the formation of lateral roots, LBD16 and miRNA390a, were identified as crucial during the formation of the galls and giant cells induced by Meloidogyne javanica in Arabidopsis. 5. The auxin/cytokinin balance seems to be an important regulatory signal to control gene expression patterns in giant cells shown by in silico analysis of different transcriptomes. It also pointed to marked differences between syncytia and GCs.