Mining the secretome of root-knot nematodes for cell wall modifying proteins

E.H.A. Roze

Research output: Thesisinternal PhD, WU

Abstract

The products of parasitism genes in nematodes must be secreted to reach their targets at the nematode-plant interface. These nematode secretory proteins are therefore recognised to play an important role in the nematode-plant interaction and as a result have been subject of intense study for years. In this thesis, the results on the identification of members of the parasitome in plant-parasitic nematodes, with a focus on the root-knot nematode species Meloidogyne chitwoodi and M. incognita, are described and discussed. In Chapter 1 a general introduction to the subject is given, which provides background on the main issues of this thesis. In Chapter 2 the first overview of the secretome and candidate parasitome members from M. chitwoodi are reported, which was at the start of the project a molecularly unexplored root-knot nematode species. As a starting point, all publicly available expressed sequence tags (ESTs) were used, i.e. 12,218, from three life stages of M. chitwoodi. A pipeline of bio-informatics algorithms was applied to screen this dataset for secretory proteins, which resulted in the identification of 398 secretome members. Two-third of these 398 sequences has no significant similarity to other known proteins in the nr database.
In this study ‘parasitome’ is defined as ‘the products of parasitism genes secreted by a range of secretory organs in the nematode, including the oesophageal, amphidial, and rectal glands, the intestine, and the hypodermis’. To point out candidate parasitism gene products within the M. chitwoodi secretome, the site of expression of the most abundant secretome members was localised using the in situ hybridisation procedure. We found that at least eight out of the twenty most abundantly represented members of the secretome are specifically expressed in the single-celled oesophageal glands. One tag hybridised in the dorsal oesophageal gland, seven in the two subventral oesophageal glands, two in the intestine, and one tag hybridised to the tail tip in the proximity of the phasmids. Four sequences showed similarity to putative parasitism genes from other nematode species, whereas seven represented pioneering sequences. In case of three oesophageal gland specific parasitism genes, the predicted subcellular localization in host plant cells, following cleavage of the leader peptide for protein secretion, was nuclear.
The results of this study, and those described in chapters 3, 4 and 5, show that the generation of ESTs indeed forms an invaluable resource for gene discovery, in this case (candidate) parasitism genes (McCarter et al., 2000). In particular, our findings obtained with in situ hybridisation microscopy make clear that with a combination of ESTs from high quality cDNA libraries and bioinformatics it is not necessary to use pre-selection methods to enrich libraries for transcripts from the gland region of a nematode. These pre-selections were based on tissue specific hybridizations (e.g. suppressive subtractive hybridization; Huang et al., 2004a), micro-aspiration of mRNA from specific regions in the parasite (Gao et al., 2003; Huang et al., 2003; Wang et al., 2001), and signal sequence trapping (Wang et al., 2001). Compared to these methods, EST generation from whole nematode libraries followed by bioinformatics analyses is relatively simple and more time efficient. Similarly, differential display methods, as were used by us and others in the past, seem less efficient than EST analysis. The consequence of using whole nematode cDNA libraries, without pre-selection, is that our strategy is most effective with larger numbers of ESTs. Five years ago the costs involved in the generation of ESTs were still a rate limiting factor. However, the generation of ESTs from (plant-)parasitic nematodes is an ongoing process which currently operates at a high pace. Fortunately, the ESTs that are being produced are made publicly available by efforts such as the Parasitic Nematode Sequencing Project (for an overview, see http://www.nematode.net/Species.Summaries/index.php (Wylie et al., 2004)). Our strategy may be further improved by using signal peptide prediction software which is trained on datasets built from nematode secretory proteins (e.g. Caenorhabditis elegans), since there is evidence that the composition of the signal peptides may vary between eukaryotic phyla (Pearson et al., 2005).
Our strategy to identify the products of parasitism genes forms a basis for further functional analysis of parasite proteins involved in the nematode-host interaction. Our efforts were focused on the genes involved in the very early stages of parasitism by infective second stage juveniles, including host invasion and suppression of early defence responses in the host. In order to extend this basis, it would be interesting to enrich the EST dataset analysed in Chapter 2 with ESTs from later parasitic juvenile stages of M. chitwoodi. The same analyses can then be applied to these ESTs in order to point out candidate parasitism genes which are for example involved in giant cell formation and maintenance. Alternatively, or in addition, candidate parasitism genes can be pointed out by in silico comparison of the ESTs (and their abundances) between different developmental stages. In this way, EST analysis also forms an alternative for the technically more complicated differential display methods such as cDNA-AFLP and micro array analyses (De Boer et al., 2002b; Qin et al., 2000).
In Chapter 3 and Chapter 4, the focus is on the repertoire of cell wall modifying proteins (CWMPs) which were identified from M. chitwoodi. These proteins form a subset of the parasitome of plant parasitic nematodes and play an important role during the intercellular migration of root-knot nematodes through host plant roots. Our study resulted in the identification of the most elaborate repertoire of CWMPs found in a single plant-parasitic nematode species so far. It includes -1,4-endoglucanases (cellulases), a -1,4-endoxylanase, pectate lyases, polygalacturonases, a cellulose binding protein (all described in Chapter 3) and expansin-like proteins (Chapter 4).
Four β-1,4-endoglucanases were identified from M. chitwoodi and named Mc-eng-1 to Mc-eng-4. Full-length sequences were obtained for Mc-eng-1 to Mc-eng-3. The encoded cellulases were found to belong to glycoside hydrolase family 5 (GHF5) and were all predicted to contain an N-terminal signal peptide for secretion. In situ hybridisation localised the site of expression of Mc-eng-3 in the subventral oesophageal secretory glands of M. chitwoodi pre-parasitic J2-s. Attempts to obtain the full length sequence of Mc-eng-4 failed, but sequence similarity search results with the partial sequence revealed that this sequence encodes a cellulase that consists of a catalytic domain and a carbohydrate binding module (CBM). These results indicate that M. chitwoodi produces cellulases with two different domain architectures, either a single catalytic domain (Mc-ENG-1, Mc-ENG-2, and Mc-ENG-3) or a catalytic domain linked to a CBM (Mc-ENG-4). Our phylogenetic analysis of root-knot nematode cellulases, including Mc-ENG-1, Mc-ENG-2, and Mc-ENG-3, showed that there are probably no cellulases that are unique for one (group of) nematode species, but that there are several lines of cellulases common to root-knot nematodes.
The site of Mc-cbp-1 (cellulose binding protein) expression was localised to the subventral oesophageal secretory glands of M. chitwoodi pre-parasitic J2-s. The Mc-CBP-1 protein consists of a bacterial family 2 CBM, which can mediate the binding to the cellulose substrate and/or may facilitate the non-hydrolytic disruption of cellulose fibres (Din et al., 1991; Levy et al., 2002; Shoseyov et al., 2006; Tomme et al., 1998). The CBM of Mc-CBP-1 is C-terminally linked to a domain of 80 amino acids that shares some similarity with Fn3-like domains. The latter domains are involved in protein-protein interactions, and in addition, the 80 amino acid N-terminal domain is rich in charged amino acids which are typically found on the hydrophilic surfaces of proteins to interact with other proteins. We hypothesise that Mc-CBP-1 is involved in the facilitation of cellulose hydrolysis, either as a docking station for CBM-lacking cellulases to bind to cellulose or by modification of the cellulose surface, or both. On the other hand, recent work by Gaulin et al. (2006) suggests that cellulose binding domains from the oomycete Phytophthora parasitica may also act as an elicitor of defence responses in plants (Gaulin et al., 2006). It is not yet clear if this is a general phenomenon of cellulose binding domains released by plant-pathogens and to what extent nematode cellulose binding proteins induce, either directly or indirectly through cell wall modifications, defence responses.
Since Meloidogyne species migrate intercellularly through the roots of a host plant, they go through the middle lamella which is rich in pectic polysaccharides. The complete degradation of pectins into monomers requires the combined action of several types of pectinolytic enzymes. Root-knot nematodes do not feed on carbohydrates released from pectin degradation and thus do not require the complete degradation of plant cell wall pectins. In contrast, root-knot nematodes actually make use of a controlled and local degradation of the pectins in the middle lamella only and will benefit most likely from hydrolases and lyases (depolymerases), which only cleave the backbone of pectin (Tamaru and Doi, 2001), to weaken the intercellular bondings between cells. Both the pectate lyases and polygalacturonases identified from M. chitwoodi in this study belong to this group of pectin degrading enzymes.
Four pectate lyases, named Mc-pel-1 to Mc-pel-4, were identified from M. chitwoodi invasive J2-s. In case of Mc-pel-1 and Mc-pel-2, full-length sequences were obtained, whereas Mc-pel-3 and Mc-pel-4 were represented by partial cDNA sequences in the M. chitwoodi ppJ2 cDNA library. Both Mc-pel-1 and Mc-pel-2 have an N-terminal signal peptide for secretion. In situ hybridisation performed on M. chitwoodi invasive J2-s revealed a specific expression confined to the subventral oesophageal secretory glands in case of Mc-pel-1. Based on similarity search results, the (putative) pectate lyases are all considered to be member of pectate lyase family III, which also comprises the pectate lyases from other plant-parasitic nematodes. Vanholme et al. (2007) showed with RNA interference (RNAi) knock-down experiments that a moderate but significant reduction in transcripts of a pectate lyase leads to strongly reduced infectivity of cyst nematodes (Vanholme et al., 2007). Cyst nematodes are not particularly focused on a stealthy invasion through the middle-lamella of host cells. Instead, they use a brute force approach during plant invasion and leave behind a trail of destructed cells. Stealthily invading root-knot nematodes are predicted to have stronger requirement for pectin degrading enzymes than cyst nematodes, however, there is no experimental data from RNAi knock-down experiments available at the moment to support this prediction.
A family of at least four pectate lyases was found in M. chitwoodi, while others reported similar findings in earlier studies on M. incognita. The pectate substrate of pectate lyase is essentially built from repeating units of galacturonic acid with varying degrees of methylation. The actual composition of the oligogalacturonate units changes during the life time of a plant, and varies between different plant species. The overall topology of pectate lyases is the same, the core consists of β-strands forming a right-handed parallel β-helix, which suggests a conserved mode of action (D'Ovidio et al., 2004). In an attempt to explain multiple pectate lyase gene families in plant-pathogens, Herron et al. (2000) proposed that plant-pathogens with broad host ranges have multiple isozymes of pectate lyases with similar catalytic properties, but that recognize differently composed and decorated oligogalacturonate units (Herron et al., 2000). This proposal now seems to find support in the repertoire of pectate lyases of the polyphagous root-knot nematodes.
Root-knot nematodes have evidently explored other venues, besides the production of pectate lyases, to achieve degradation of the pectin component in host cell walls. In total, two polygalacturonases, named Mc-pg-1 and Mc-pg-2, were identified from invasive J2-s of M. chitwoodi. The full-length sequence of Mc-pg-1 was obtained and with in situ hybridisation performed on M. chitwoodi invasive J2-s, a specific expression confined to the subventral oesophageal secretory glands was revealed. Based on similarities with other exo-acting polygalaturonases, we believe that Mc-PG-1 codes for an exo-polygalacturonase. The polygalacturonase MI-PG-1, which was cloned from M. incognita is also classified as an exo-acting enzyme (Jaubert et al., 2002). Exo- and endo-polygalacturonases have different impacts on the cell wall integrity of plants. Endo-polygalacturonases are known to play a role in plant defence responses, e.g. through the release of elicitor-active oligogalacturonides (Cervone et al., 1989; D'Ovidio et al., 2004; Favaron et al., 1988). Biotrophic pathogens, like Meloidogyne species, must avoid inducing host defence responses and therefore it seems likely that they will profit from a cell wall degrading enzyme repertoire that does not release oligogalacturonides. Experimental biochemical evidence for the classification of polygalacturonases from M. chitwoodi and M. incognita as exo-acting enzymes is still lacking. It will be interesting to investigate the enzymatic properties of these nematode polygalacturonases in order to establish to which class of polygalacturonases they belong.
Most of the research on the role of cell wall degrading enzymes in plant-pathogen interactions is done on pectin degrading enzymes in pathogens of dicotyledons. Pectins, however, represent only a minor fraction of the cell wall components in monocots, which mainly include hemicelluloses in their cell wall matrices (Carpita, 1996). A major part of the hemicellulose fraction of monocotyledonous cell walls consists of substituted xylan polymers. It is therefore expected that xylan-degrading enzymes have an important role in promoting virulence of pathogens of monocots; possibly equivalent to the role of pectinolytic enzymes for pathogens of dicots (Beliën et al., 2006 and references herein).
We identified two novel xylan-degrading enzymes in two root-knot nematode species that are virulent pathogens of both dicotyledons and monocotyledons, i.e. M. chitwoodi and M. incognita (described in Chapter 3 and 5). These xylan-degrading enzymes are β-1,4-endoxylanases, which are capable of hydrolysing substituted xylan polymers into fragments of random size. The β-1,4-endoxylanase from M. incognita, named Mi-xyl1, is the first functional β-1,4-endoxylanase of animal origin. Both Mi-xyl1 and the β-1,4-endoxylanase from M. chitwoodi, Mc-xyl-1, were found to be expressed in the subventral oesophageal gland cells of the nematode and the encoded proteins were predicted to have an N-terminal signal peptide for secretion. Based on similarity search results we consider MI-XYL1 and Mc-XYL-1 enzymes that belong to GHF5. These two β-1,4-endoxylanases share 39% identity in their catalytic domain. Further comparison between the two xylanases showed that Mc-XYL-1 has an ancillary stretch of 76 amino acids at its C-terminus which is, based on sequence similarity, a putative xylan binding module (XBM).
So far, results of Southern blot analysis and EST database screenings did not show evidence of endoxylanases in nematode species that have specialized on dicots (M. hapla, G. rostochiensis and G. pallida), in spite of thousands of ESTs in public sequence databases. Based on our findings, we hypothesise that the production of β-1,4-endoxylanase in plant parasitic nematodes is correlated with parasitism on monocots. RNA interference studies as mentioned above may provide information about the requirement of β-1,4-endoxylanase for infectivity on monocots. In addition, it would be interesting to investigate if the specialist nematodes of monocots indeed have a bias towards xylan degradation in their repertoire of cell wall modifying proteins. Unfortunately, nematode specialists on graminaceous monocots, such as the cereal cyst nematode Heterodera avenea, have not been included in the scope of molecular nematology so far. Therefore, there is currently no sequence information available of such nematode species.
Chapter 4 deals with the expansin-like proteins identified in Meloidogyne spp., with the emphasis on those from M. chitwoodi. Expansins form a diverse protein superfamily in plants and play a role in various biological processes in which re-arrangement of plant cell wall polysaccharides is involved (Cosgrove, 2000a). Plant expansins lack hydrolytic activity and are proposed to weaken non-covalent interactions between cellulose and hemicellulose polymers (McQueen-Mason and Cosgrove, 1994). They seem to act synergistically with cellulases by making the plant cell wall polysaccharides more accessible to enzymatic attack (Cosgrove, 2000a). A similar synergistic action of expansins and cell wall degrading enzymes secreted by plant-pathogens may facilitate their invasion of the host.
The distribution of expansins was believed to be restricted to land plants (Cosgrove, 2000a), but recently, a small number of expansin-like sequences have been identified from other organisms (Darley et al., 2003; Laine et al., 2000; Saloheimo et al., 2002). In addition, a functional β-expansin, Gr-EXPB1, was found to be secreted by the plant-parasitic nematode Globodera rostochiensis (Kudla et al., 2005; Qin et al., 2004). This latter sequence was used to query assembled ESTs from M. chitwoodi and found in total four expansin-like sequences, named Mc-EXP1 to Mc-EXP4. These M. chitwoodi sequences harbour most of the signature motifs of α- and β-expansins, but since we were not able to produce active recombinant protein to test expansin activity of the proteins on plant tissues, the sequences are designated as expansin-like proteins.
Expansin-like Mc-EXP1 is represented by 23 ESTs in the EST dataset from M. chitwoodi ppJ2-s and belongs to the top 20 of most abundantly expressed members of the secretome of this developmental stage of M. chitwoodi (Chapter 2 of this thesis). Both Mc-EXP1 and Mc-EXP2 are multi-domain proteins. They consist of an N-terminal signal peptide for secretion, followed by either a CBM (Mc-EXP1) or a Lysin-motif domain (LysM; Mc-EXP2) both linked to a C-terminal expansin-like domain. Compared to the domain structure of plant expansins, the expansin-like domain and the polysaccharide-binding domain (CBM) are in reverse orientation. Thus, the domain structure of Mc-EXP1 is similar to that of Gr-EXPB1. The sequence identity between the C-termini of Mc-EXP1 and Mc-EXP2 and plant expansins ranged from 30% to 35%. Notably, for both expansin-like proteins highest sequence similarity is with α-expansins from plants. Conclusive evidence on the type of activity of Mc-EXP1 and Mc-EXP2, i.e. do they resemble more the α- or -expansins, requires the production of active recombinant protein and subsequent cell wall extension assays on the different types of cell walls. Unfortunately, we have not been able to achieve heterologous expression of Mc-EXPs in plants, which would have allowed us to study the type of expansin activity of the proteins.
Probes designed on Mc-EXP1 localised the gene transcription in the subventral oesophageal secretory glands of ppJ2-s of M. chitwoodi. The same localisation was found for the Mc-EXP1 and Mc-EXP2 proteins, which was investigated by immunofluorescence microscopy with specific antisera to Mc-EXP1 and Mc-EXP2. A strong fluorescence, with a granular pattern, was observed in the subventral oesophageal gland extensions and ampullae and to a lesser extent in the gland lobes. The presence of a signal peptide for secretion and the localisation of the transcripts and protein in the subventral oesophageal secretory glands of M. chitwoodi ppJ2-s strongly suggest that expansin-like Mc-EXP1 is secreted by the nematode. The current data on the developmental expression and production of expansin-like Mc-EXP2 are less straightforward. Mc-EXP2 was identified in a library made from nematode eggs and could not be amplified from the cDNA library made from M. chitwoodi ppJ2-s. In addition, no in situ hybridisation signal was observed in M. chitwoodi ppJ2-s with an antisense probe spanning the N-terminal putative LysM domain. On the other hand, in immunofluorescence microscopy experiments, the Mc-EXP2 protein was localised specifically in the subventral oesophageal glands of M. chitwoodi ppJ2-s. Therefore, a developmental expression study needs to be performed on Mc-EXP2 in order to find out if the results obtained at the protein level are either confirmed or contradicted by results obtained at the transcriptional level.
The other two expansin-like proteins from M. chitwoodi, Mc-EXP3 and Mc-EXP4, only contain an expansin-like domain with an N-terminal signal peptide for secretion. Mc-EXP4 is represented by 55 ESTs from the ppJ2 stage of M. chitwoodi and herewith it belongs to the top 10 of most abundantly represented transcripts in the M. chitwoodi EST dataset (see Chapter 2 of this thesis). Despite its high abundance, no hybridisation signal was detected for Mc-EXP4 by in situ hybridisation in ppJ2-s of M. chitwoodi. Specific staining of the subventral oesophageal secretory glands in this developmental stage was observed in case of Mc-EXP3.
In addition to the expansin-like genes in M. chitwoodi, evidence was also found for the presence of expansin-like genes in five other root-knot nematode species, a root-lesion nematode species, and plant pathogenic oomycete and fungal species. It remains to be shown whether these organisms produce functional expansins, but our findings make us point at two things. Firstly, it appears that expansins do not occur in only a small number of organisms outside the plant kingdom, but instead, are widespread and likely to be involved in many plant-pathogen interactions and plant-microbe interactions in general. Secondly, the proposed and currently adopted nomenclature of the expansin superfamily might need revision. For, the designations ‘expansin’ (expansin A and expansin B) and ‘expansin-like’ (expansin-like family A and B) in this nomenclature are kept exclusively for plant proteins. Proteins from other organisms that share structural similarity with both domains of plant expansins are grouped in a separate ‘catch-all’ category and designated ‘expansin-related’, solely based on the fact that they do not originate from plants. When the expansin-like proteins identified in this study are indeed functional expansins, there would be no biological rationale behind the ‘catch-all’ group of expansin-related proteins.
A series of conserved cysteines (C) and the HFD motif around amino acid position 110 are used as the key signatures of the plant expansin family (Cosgrove, 2000b). A series of conserved cysteines is also present along the backbone of the root-knot nematode expansin-like domains. In case of the HFD motif, only the H and D residues are conserved in the expansin-like proteins of root-knot nematodes. A similar conservation was found in the β-expansin Gr-EXPB1 from G. rostochiensis, for which cell wall expansion activity on type II primary cell walls was found (Kudla et al., 2005; Qin et al., 2004). These findings suggest that the functional significance of the HFD motif is obscure and can be clarified by the biochemical characterization of root-knot nematode expansin-like proteins, e.g. by site directed mutagenesis followed by activity assays.
It seems likely that the CWMP repertoire of M. chitwoodi mirrors its wide host plant range since the latter represents a broad diversity of cell wall polysaccharides. The CWMPs identified from M. chitwoodi enable the migrating nematode to cleave the backbone of all major types of plant cell wall polysaccharides and to modify their interactions. The subventral oesophageal gland specific expression and presence of predicted secretion signal peptides suggest that the CWMPs are secreted from the nematode. Hereby, this enzyme complex facilitates the intercellular migration of M. chitwoodi in the host plant. Post-transcriptional gene silencing by soaking nematode juveniles in double-stranded RNA (RNA interference) was successfully applied to assess the importance of β-1,4-endoglucanases and pectate lyases for infection of plant roots by cyst nematodes (Chen et al., 2005; Vanholme et al., 2007). Similar RNA interference experiments targeting CWMP encoding genes from M. chitwoodi, followed by infection tests on monocots and dicots, may reveal the influence of the individual CWMPs on the host range of M. chitwoodi.
On the other hand, our findings raise the questions how it is possible that a nematode can secrete such a variety of CWMPs without i) inflicting any detectable damage to the cells along the migratory track and ii) being noticed by the host plant? An explanation for the fact that plant cells remain intact might be that the cell wall modifying enzymes that are found to be produced by the nematode only cleave the backbone of plant cell wall polysaccharides. As mentioned above, root-knot nematodes do not feed on carbohydrates released from plant cell wall degradation and thus do not require the complete degradation of plant cell wall polysaccharides. In contrast, root-knot nematodes only seem to weaken the intercellular bondings between cells by local secretion of the CWMPs through the stylet into the middle lamella resulting in controlled degradation. This controlled degradation might also be a way to prevent the production of elicitor-active oligogalacturonides involved in plant defence responses. Alternatively or in addition, the nematode might constantly repress the host defence mechanisms during intercellular migration through the roots.
In relation to these questions it is interesting to study whether the nematode can ‘sense’, either inside or outside the plant root system or both, with which type of plant (cell wall) it is dealing. Following the perception of cell wall type signals, the nematode could adapt its repertoire of CWMPs to it. This would enable the nematode to avoid the production of CWMPs that do not act on the available substrate, thus saving valuable energy and reducing the chance of being detected by the host. Potato root exudates are for example known to induce the secretion of proteins by cyst nematodes (Smant et al., 1997). One way to test whether host plant root exudates already influence the production of CWMPs at the transcriptional level would be to treat eggs or hatched nematodes with different root exudates followed by real-time quantitative PCR analysis. One could for example hatch nematodes in water and in root exudates obtained from a monocot and a dicot and compare the transcriptional levels of a certain CWMP.
Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Wageningen University
Supervisors/Advisors
  • Bakker, Jaap, Promotor
  • Smant, Geert, Co-promotor
Award date15 Feb 2008
Place of Publication[S.l.]
Publisher
Print ISBNs9789085048589
Publication statusPublished - 2008

Fingerprint

Meloidogyne chitwoodi
expansins
root-knot nematodes
cell walls
Nematoda
expressed sequence tags
pectate lyase
proteins
endo-1,4-beta-xylanase
polygalacturonase
signal peptide
plant parasitic nematodes
cellulases
pectins
Liliopsida
parasitism
carbohydrate binding
enzymes
cellulose
xylan

Keywords

  • meloidogyne
  • plant parasitic nematodes
  • pathogenesis-related proteins
  • protein secretion
  • cell wall components
  • gene expression
  • nucleotide sequences
  • genetic mapping

Cite this

@phdthesis{99e9c2bcdec94ad9a4139259a4869394,
title = "Mining the secretome of root-knot nematodes for cell wall modifying proteins",
abstract = "The products of parasitism genes in nematodes must be secreted to reach their targets at the nematode-plant interface. These nematode secretory proteins are therefore recognised to play an important role in the nematode-plant interaction and as a result have been subject of intense study for years. In this thesis, the results on the identification of members of the parasitome in plant-parasitic nematodes, with a focus on the root-knot nematode species Meloidogyne chitwoodi and M. incognita, are described and discussed. In Chapter 1 a general introduction to the subject is given, which provides background on the main issues of this thesis. In Chapter 2 the first overview of the secretome and candidate parasitome members from M. chitwoodi are reported, which was at the start of the project a molecularly unexplored root-knot nematode species. As a starting point, all publicly available expressed sequence tags (ESTs) were used, i.e. 12,218, from three life stages of M. chitwoodi. A pipeline of bio-informatics algorithms was applied to screen this dataset for secretory proteins, which resulted in the identification of 398 secretome members. Two-third of these 398 sequences has no significant similarity to other known proteins in the nr database. In this study ‘parasitome’ is defined as ‘the products of parasitism genes secreted by a range of secretory organs in the nematode, including the oesophageal, amphidial, and rectal glands, the intestine, and the hypodermis’. To point out candidate parasitism gene products within the M. chitwoodi secretome, the site of expression of the most abundant secretome members was localised using the in situ hybridisation procedure. We found that at least eight out of the twenty most abundantly represented members of the secretome are specifically expressed in the single-celled oesophageal glands. One tag hybridised in the dorsal oesophageal gland, seven in the two subventral oesophageal glands, two in the intestine, and one tag hybridised to the tail tip in the proximity of the phasmids. Four sequences showed similarity to putative parasitism genes from other nematode species, whereas seven represented pioneering sequences. In case of three oesophageal gland specific parasitism genes, the predicted subcellular localization in host plant cells, following cleavage of the leader peptide for protein secretion, was nuclear. The results of this study, and those described in chapters 3, 4 and 5, show that the generation of ESTs indeed forms an invaluable resource for gene discovery, in this case (candidate) parasitism genes (McCarter et al., 2000). In particular, our findings obtained with in situ hybridisation microscopy make clear that with a combination of ESTs from high quality cDNA libraries and bioinformatics it is not necessary to use pre-selection methods to enrich libraries for transcripts from the gland region of a nematode. These pre-selections were based on tissue specific hybridizations (e.g. suppressive subtractive hybridization; Huang et al., 2004a), micro-aspiration of mRNA from specific regions in the parasite (Gao et al., 2003; Huang et al., 2003; Wang et al., 2001), and signal sequence trapping (Wang et al., 2001). Compared to these methods, EST generation from whole nematode libraries followed by bioinformatics analyses is relatively simple and more time efficient. Similarly, differential display methods, as were used by us and others in the past, seem less efficient than EST analysis. The consequence of using whole nematode cDNA libraries, without pre-selection, is that our strategy is most effective with larger numbers of ESTs. Five years ago the costs involved in the generation of ESTs were still a rate limiting factor. However, the generation of ESTs from (plant-)parasitic nematodes is an ongoing process which currently operates at a high pace. Fortunately, the ESTs that are being produced are made publicly available by efforts such as the Parasitic Nematode Sequencing Project (for an overview, see http://www.nematode.net/Species.Summaries/index.php (Wylie et al., 2004)). Our strategy may be further improved by using signal peptide prediction software which is trained on datasets built from nematode secretory proteins (e.g. Caenorhabditis elegans), since there is evidence that the composition of the signal peptides may vary between eukaryotic phyla (Pearson et al., 2005). Our strategy to identify the products of parasitism genes forms a basis for further functional analysis of parasite proteins involved in the nematode-host interaction. Our efforts were focused on the genes involved in the very early stages of parasitism by infective second stage juveniles, including host invasion and suppression of early defence responses in the host. In order to extend this basis, it would be interesting to enrich the EST dataset analysed in Chapter 2 with ESTs from later parasitic juvenile stages of M. chitwoodi. The same analyses can then be applied to these ESTs in order to point out candidate parasitism genes which are for example involved in giant cell formation and maintenance. Alternatively, or in addition, candidate parasitism genes can be pointed out by in silico comparison of the ESTs (and their abundances) between different developmental stages. In this way, EST analysis also forms an alternative for the technically more complicated differential display methods such as cDNA-AFLP and micro array analyses (De Boer et al., 2002b; Qin et al., 2000). In Chapter 3 and Chapter 4, the focus is on the repertoire of cell wall modifying proteins (CWMPs) which were identified from M. chitwoodi. These proteins form a subset of the parasitome of plant parasitic nematodes and play an important role during the intercellular migration of root-knot nematodes through host plant roots. Our study resulted in the identification of the most elaborate repertoire of CWMPs found in a single plant-parasitic nematode species so far. It includes -1,4-endoglucanases (cellulases), a -1,4-endoxylanase, pectate lyases, polygalacturonases, a cellulose binding protein (all described in Chapter 3) and expansin-like proteins (Chapter 4). Four β-1,4-endoglucanases were identified from M. chitwoodi and named Mc-eng-1 to Mc-eng-4. Full-length sequences were obtained for Mc-eng-1 to Mc-eng-3. The encoded cellulases were found to belong to glycoside hydrolase family 5 (GHF5) and were all predicted to contain an N-terminal signal peptide for secretion. In situ hybridisation localised the site of expression of Mc-eng-3 in the subventral oesophageal secretory glands of M. chitwoodi pre-parasitic J2-s. Attempts to obtain the full length sequence of Mc-eng-4 failed, but sequence similarity search results with the partial sequence revealed that this sequence encodes a cellulase that consists of a catalytic domain and a carbohydrate binding module (CBM). These results indicate that M. chitwoodi produces cellulases with two different domain architectures, either a single catalytic domain (Mc-ENG-1, Mc-ENG-2, and Mc-ENG-3) or a catalytic domain linked to a CBM (Mc-ENG-4). Our phylogenetic analysis of root-knot nematode cellulases, including Mc-ENG-1, Mc-ENG-2, and Mc-ENG-3, showed that there are probably no cellulases that are unique for one (group of) nematode species, but that there are several lines of cellulases common to root-knot nematodes. The site of Mc-cbp-1 (cellulose binding protein) expression was localised to the subventral oesophageal secretory glands of M. chitwoodi pre-parasitic J2-s. The Mc-CBP-1 protein consists of a bacterial family 2 CBM, which can mediate the binding to the cellulose substrate and/or may facilitate the non-hydrolytic disruption of cellulose fibres (Din et al., 1991; Levy et al., 2002; Shoseyov et al., 2006; Tomme et al., 1998). The CBM of Mc-CBP-1 is C-terminally linked to a domain of 80 amino acids that shares some similarity with Fn3-like domains. The latter domains are involved in protein-protein interactions, and in addition, the 80 amino acid N-terminal domain is rich in charged amino acids which are typically found on the hydrophilic surfaces of proteins to interact with other proteins. We hypothesise that Mc-CBP-1 is involved in the facilitation of cellulose hydrolysis, either as a docking station for CBM-lacking cellulases to bind to cellulose or by modification of the cellulose surface, or both. On the other hand, recent work by Gaulin et al. (2006) suggests that cellulose binding domains from the oomycete Phytophthora parasitica may also act as an elicitor of defence responses in plants (Gaulin et al., 2006). It is not yet clear if this is a general phenomenon of cellulose binding domains released by plant-pathogens and to what extent nematode cellulose binding proteins induce, either directly or indirectly through cell wall modifications, defence responses. Since Meloidogyne species migrate intercellularly through the roots of a host plant, they go through the middle lamella which is rich in pectic polysaccharides. The complete degradation of pectins into monomers requires the combined action of several types of pectinolytic enzymes. Root-knot nematodes do not feed on carbohydrates released from pectin degradation and thus do not require the complete degradation of plant cell wall pectins. In contrast, root-knot nematodes actually make use of a controlled and local degradation of the pectins in the middle lamella only and will benefit most likely from hydrolases and lyases (depolymerases), which only cleave the backbone of pectin (Tamaru and Doi, 2001), to weaken the intercellular bondings between cells. Both the pectate lyases and polygalacturonases identified from M. chitwoodi in this study belong to this group of pectin degrading enzymes. Four pectate lyases, named Mc-pel-1 to Mc-pel-4, were identified from M. chitwoodi invasive J2-s. In case of Mc-pel-1 and Mc-pel-2, full-length sequences were obtained, whereas Mc-pel-3 and Mc-pel-4 were represented by partial cDNA sequences in the M. chitwoodi ppJ2 cDNA library. Both Mc-pel-1 and Mc-pel-2 have an N-terminal signal peptide for secretion. In situ hybridisation performed on M. chitwoodi invasive J2-s revealed a specific expression confined to the subventral oesophageal secretory glands in case of Mc-pel-1. Based on similarity search results, the (putative) pectate lyases are all considered to be member of pectate lyase family III, which also comprises the pectate lyases from other plant-parasitic nematodes. Vanholme et al. (2007) showed with RNA interference (RNAi) knock-down experiments that a moderate but significant reduction in transcripts of a pectate lyase leads to strongly reduced infectivity of cyst nematodes (Vanholme et al., 2007). Cyst nematodes are not particularly focused on a stealthy invasion through the middle-lamella of host cells. Instead, they use a brute force approach during plant invasion and leave behind a trail of destructed cells. Stealthily invading root-knot nematodes are predicted to have stronger requirement for pectin degrading enzymes than cyst nematodes, however, there is no experimental data from RNAi knock-down experiments available at the moment to support this prediction. A family of at least four pectate lyases was found in M. chitwoodi, while others reported similar findings in earlier studies on M. incognita. The pectate substrate of pectate lyase is essentially built from repeating units of galacturonic acid with varying degrees of methylation. The actual composition of the oligogalacturonate units changes during the life time of a plant, and varies between different plant species. The overall topology of pectate lyases is the same, the core consists of β-strands forming a right-handed parallel β-helix, which suggests a conserved mode of action (D'Ovidio et al., 2004). In an attempt to explain multiple pectate lyase gene families in plant-pathogens, Herron et al. (2000) proposed that plant-pathogens with broad host ranges have multiple isozymes of pectate lyases with similar catalytic properties, but that recognize differently composed and decorated oligogalacturonate units (Herron et al., 2000). This proposal now seems to find support in the repertoire of pectate lyases of the polyphagous root-knot nematodes. Root-knot nematodes have evidently explored other venues, besides the production of pectate lyases, to achieve degradation of the pectin component in host cell walls. In total, two polygalacturonases, named Mc-pg-1 and Mc-pg-2, were identified from invasive J2-s of M. chitwoodi. The full-length sequence of Mc-pg-1 was obtained and with in situ hybridisation performed on M. chitwoodi invasive J2-s, a specific expression confined to the subventral oesophageal secretory glands was revealed. Based on similarities with other exo-acting polygalaturonases, we believe that Mc-PG-1 codes for an exo-polygalacturonase. The polygalacturonase MI-PG-1, which was cloned from M. incognita is also classified as an exo-acting enzyme (Jaubert et al., 2002). Exo- and endo-polygalacturonases have different impacts on the cell wall integrity of plants. Endo-polygalacturonases are known to play a role in plant defence responses, e.g. through the release of elicitor-active oligogalacturonides (Cervone et al., 1989; D'Ovidio et al., 2004; Favaron et al., 1988). Biotrophic pathogens, like Meloidogyne species, must avoid inducing host defence responses and therefore it seems likely that they will profit from a cell wall degrading enzyme repertoire that does not release oligogalacturonides. Experimental biochemical evidence for the classification of polygalacturonases from M. chitwoodi and M. incognita as exo-acting enzymes is still lacking. It will be interesting to investigate the enzymatic properties of these nematode polygalacturonases in order to establish to which class of polygalacturonases they belong. Most of the research on the role of cell wall degrading enzymes in plant-pathogen interactions is done on pectin degrading enzymes in pathogens of dicotyledons. Pectins, however, represent only a minor fraction of the cell wall components in monocots, which mainly include hemicelluloses in their cell wall matrices (Carpita, 1996). A major part of the hemicellulose fraction of monocotyledonous cell walls consists of substituted xylan polymers. It is therefore expected that xylan-degrading enzymes have an important role in promoting virulence of pathogens of monocots; possibly equivalent to the role of pectinolytic enzymes for pathogens of dicots (Beli{\"e}n et al., 2006 and references herein). We identified two novel xylan-degrading enzymes in two root-knot nematode species that are virulent pathogens of both dicotyledons and monocotyledons, i.e. M. chitwoodi and M. incognita (described in Chapter 3 and 5). These xylan-degrading enzymes are β-1,4-endoxylanases, which are capable of hydrolysing substituted xylan polymers into fragments of random size. The β-1,4-endoxylanase from M. incognita, named Mi-xyl1, is the first functional β-1,4-endoxylanase of animal origin. Both Mi-xyl1 and the β-1,4-endoxylanase from M. chitwoodi, Mc-xyl-1, were found to be expressed in the subventral oesophageal gland cells of the nematode and the encoded proteins were predicted to have an N-terminal signal peptide for secretion. Based on similarity search results we consider MI-XYL1 and Mc-XYL-1 enzymes that belong to GHF5. These two β-1,4-endoxylanases share 39{\%} identity in their catalytic domain. Further comparison between the two xylanases showed that Mc-XYL-1 has an ancillary stretch of 76 amino acids at its C-terminus which is, based on sequence similarity, a putative xylan binding module (XBM). So far, results of Southern blot analysis and EST database screenings did not show evidence of endoxylanases in nematode species that have specialized on dicots (M. hapla, G. rostochiensis and G. pallida), in spite of thousands of ESTs in public sequence databases. Based on our findings, we hypothesise that the production of β-1,4-endoxylanase in plant parasitic nematodes is correlated with parasitism on monocots. RNA interference studies as mentioned above may provide information about the requirement of β-1,4-endoxylanase for infectivity on monocots. In addition, it would be interesting to investigate if the specialist nematodes of monocots indeed have a bias towards xylan degradation in their repertoire of cell wall modifying proteins. Unfortunately, nematode specialists on graminaceous monocots, such as the cereal cyst nematode Heterodera avenea, have not been included in the scope of molecular nematology so far. Therefore, there is currently no sequence information available of such nematode species. Chapter 4 deals with the expansin-like proteins identified in Meloidogyne spp., with the emphasis on those from M. chitwoodi. Expansins form a diverse protein superfamily in plants and play a role in various biological processes in which re-arrangement of plant cell wall polysaccharides is involved (Cosgrove, 2000a). Plant expansins lack hydrolytic activity and are proposed to weaken non-covalent interactions between cellulose and hemicellulose polymers (McQueen-Mason and Cosgrove, 1994). They seem to act synergistically with cellulases by making the plant cell wall polysaccharides more accessible to enzymatic attack (Cosgrove, 2000a). A similar synergistic action of expansins and cell wall degrading enzymes secreted by plant-pathogens may facilitate their invasion of the host. The distribution of expansins was believed to be restricted to land plants (Cosgrove, 2000a), but recently, a small number of expansin-like sequences have been identified from other organisms (Darley et al., 2003; Laine et al., 2000; Saloheimo et al., 2002). In addition, a functional β-expansin, Gr-EXPB1, was found to be secreted by the plant-parasitic nematode Globodera rostochiensis (Kudla et al., 2005; Qin et al., 2004). This latter sequence was used to query assembled ESTs from M. chitwoodi and found in total four expansin-like sequences, named Mc-EXP1 to Mc-EXP4. These M. chitwoodi sequences harbour most of the signature motifs of α- and β-expansins, but since we were not able to produce active recombinant protein to test expansin activity of the proteins on plant tissues, the sequences are designated as expansin-like proteins. Expansin-like Mc-EXP1 is represented by 23 ESTs in the EST dataset from M. chitwoodi ppJ2-s and belongs to the top 20 of most abundantly expressed members of the secretome of this developmental stage of M. chitwoodi (Chapter 2 of this thesis). Both Mc-EXP1 and Mc-EXP2 are multi-domain proteins. They consist of an N-terminal signal peptide for secretion, followed by either a CBM (Mc-EXP1) or a Lysin-motif domain (LysM; Mc-EXP2) both linked to a C-terminal expansin-like domain. Compared to the domain structure of plant expansins, the expansin-like domain and the polysaccharide-binding domain (CBM) are in reverse orientation. Thus, the domain structure of Mc-EXP1 is similar to that of Gr-EXPB1. The sequence identity between the C-termini of Mc-EXP1 and Mc-EXP2 and plant expansins ranged from 30{\%} to 35{\%}. Notably, for both expansin-like proteins highest sequence similarity is with α-expansins from plants. Conclusive evidence on the type of activity of Mc-EXP1 and Mc-EXP2, i.e. do they resemble more the α- or -expansins, requires the production of active recombinant protein and subsequent cell wall extension assays on the different types of cell walls. Unfortunately, we have not been able to achieve heterologous expression of Mc-EXPs in plants, which would have allowed us to study the type of expansin activity of the proteins. Probes designed on Mc-EXP1 localised the gene transcription in the subventral oesophageal secretory glands of ppJ2-s of M. chitwoodi. The same localisation was found for the Mc-EXP1 and Mc-EXP2 proteins, which was investigated by immunofluorescence microscopy with specific antisera to Mc-EXP1 and Mc-EXP2. A strong fluorescence, with a granular pattern, was observed in the subventral oesophageal gland extensions and ampullae and to a lesser extent in the gland lobes. The presence of a signal peptide for secretion and the localisation of the transcripts and protein in the subventral oesophageal secretory glands of M. chitwoodi ppJ2-s strongly suggest that expansin-like Mc-EXP1 is secreted by the nematode. The current data on the developmental expression and production of expansin-like Mc-EXP2 are less straightforward. Mc-EXP2 was identified in a library made from nematode eggs and could not be amplified from the cDNA library made from M. chitwoodi ppJ2-s. In addition, no in situ hybridisation signal was observed in M. chitwoodi ppJ2-s with an antisense probe spanning the N-terminal putative LysM domain. On the other hand, in immunofluorescence microscopy experiments, the Mc-EXP2 protein was localised specifically in the subventral oesophageal glands of M. chitwoodi ppJ2-s. Therefore, a developmental expression study needs to be performed on Mc-EXP2 in order to find out if the results obtained at the protein level are either confirmed or contradicted by results obtained at the transcriptional level. The other two expansin-like proteins from M. chitwoodi, Mc-EXP3 and Mc-EXP4, only contain an expansin-like domain with an N-terminal signal peptide for secretion. Mc-EXP4 is represented by 55 ESTs from the ppJ2 stage of M. chitwoodi and herewith it belongs to the top 10 of most abundantly represented transcripts in the M. chitwoodi EST dataset (see Chapter 2 of this thesis). Despite its high abundance, no hybridisation signal was detected for Mc-EXP4 by in situ hybridisation in ppJ2-s of M. chitwoodi. Specific staining of the subventral oesophageal secretory glands in this developmental stage was observed in case of Mc-EXP3. In addition to the expansin-like genes in M. chitwoodi, evidence was also found for the presence of expansin-like genes in five other root-knot nematode species, a root-lesion nematode species, and plant pathogenic oomycete and fungal species. It remains to be shown whether these organisms produce functional expansins, but our findings make us point at two things. Firstly, it appears that expansins do not occur in only a small number of organisms outside the plant kingdom, but instead, are widespread and likely to be involved in many plant-pathogen interactions and plant-microbe interactions in general. Secondly, the proposed and currently adopted nomenclature of the expansin superfamily might need revision. For, the designations ‘expansin’ (expansin A and expansin B) and ‘expansin-like’ (expansin-like family A and B) in this nomenclature are kept exclusively for plant proteins. Proteins from other organisms that share structural similarity with both domains of plant expansins are grouped in a separate ‘catch-all’ category and designated ‘expansin-related’, solely based on the fact that they do not originate from plants. When the expansin-like proteins identified in this study are indeed functional expansins, there would be no biological rationale behind the ‘catch-all’ group of expansin-related proteins. A series of conserved cysteines (C) and the HFD motif around amino acid position 110 are used as the key signatures of the plant expansin family (Cosgrove, 2000b). A series of conserved cysteines is also present along the backbone of the root-knot nematode expansin-like domains. In case of the HFD motif, only the H and D residues are conserved in the expansin-like proteins of root-knot nematodes. A similar conservation was found in the β-expansin Gr-EXPB1 from G. rostochiensis, for which cell wall expansion activity on type II primary cell walls was found (Kudla et al., 2005; Qin et al., 2004). These findings suggest that the functional significance of the HFD motif is obscure and can be clarified by the biochemical characterization of root-knot nematode expansin-like proteins, e.g. by site directed mutagenesis followed by activity assays. It seems likely that the CWMP repertoire of M. chitwoodi mirrors its wide host plant range since the latter represents a broad diversity of cell wall polysaccharides. The CWMPs identified from M. chitwoodi enable the migrating nematode to cleave the backbone of all major types of plant cell wall polysaccharides and to modify their interactions. The subventral oesophageal gland specific expression and presence of predicted secretion signal peptides suggest that the CWMPs are secreted from the nematode. Hereby, this enzyme complex facilitates the intercellular migration of M. chitwoodi in the host plant. Post-transcriptional gene silencing by soaking nematode juveniles in double-stranded RNA (RNA interference) was successfully applied to assess the importance of β-1,4-endoglucanases and pectate lyases for infection of plant roots by cyst nematodes (Chen et al., 2005; Vanholme et al., 2007). Similar RNA interference experiments targeting CWMP encoding genes from M. chitwoodi, followed by infection tests on monocots and dicots, may reveal the influence of the individual CWMPs on the host range of M. chitwoodi. On the other hand, our findings raise the questions how it is possible that a nematode can secrete such a variety of CWMPs without i) inflicting any detectable damage to the cells along the migratory track and ii) being noticed by the host plant? An explanation for the fact that plant cells remain intact might be that the cell wall modifying enzymes that are found to be produced by the nematode only cleave the backbone of plant cell wall polysaccharides. As mentioned above, root-knot nematodes do not feed on carbohydrates released from plant cell wall degradation and thus do not require the complete degradation of plant cell wall polysaccharides. In contrast, root-knot nematodes only seem to weaken the intercellular bondings between cells by local secretion of the CWMPs through the stylet into the middle lamella resulting in controlled degradation. This controlled degradation might also be a way to prevent the production of elicitor-active oligogalacturonides involved in plant defence responses. Alternatively or in addition, the nematode might constantly repress the host defence mechanisms during intercellular migration through the roots. In relation to these questions it is interesting to study whether the nematode can ‘sense’, either inside or outside the plant root system or both, with which type of plant (cell wall) it is dealing. Following the perception of cell wall type signals, the nematode could adapt its repertoire of CWMPs to it. This would enable the nematode to avoid the production of CWMPs that do not act on the available substrate, thus saving valuable energy and reducing the chance of being detected by the host. Potato root exudates are for example known to induce the secretion of proteins by cyst nematodes (Smant et al., 1997). One way to test whether host plant root exudates already influence the production of CWMPs at the transcriptional level would be to treat eggs or hatched nematodes with different root exudates followed by real-time quantitative PCR analysis. One could for example hatch nematodes in water and in root exudates obtained from a monocot and a dicot and compare the transcriptional levels of a certain CWMP.",
keywords = "meloidogyne, plantenparasitaire nematoden, pathogenesis-gerelateerde eiwitten, eiwitsecretie, celwandstoffen, genexpressie, nucleotidenvolgordes, genetische kartering, meloidogyne, plant parasitic nematodes, pathogenesis-related proteins, protein secretion, cell wall components, gene expression, nucleotide sequences, genetic mapping",
author = "E.H.A. Roze",
note = "WU thesis, no. 4388",
year = "2008",
language = "English",
isbn = "9789085048589",
publisher = "s.n.",
school = "Wageningen University",

}

Roze, EHA 2008, 'Mining the secretome of root-knot nematodes for cell wall modifying proteins', Doctor of Philosophy, Wageningen University, [S.l.].

Mining the secretome of root-knot nematodes for cell wall modifying proteins. / Roze, E.H.A.

[S.l.] : s.n., 2008. 150 p.

Research output: Thesisinternal PhD, WU

TY - THES

T1 - Mining the secretome of root-knot nematodes for cell wall modifying proteins

AU - Roze, E.H.A.

N1 - WU thesis, no. 4388

PY - 2008

Y1 - 2008

N2 - The products of parasitism genes in nematodes must be secreted to reach their targets at the nematode-plant interface. These nematode secretory proteins are therefore recognised to play an important role in the nematode-plant interaction and as a result have been subject of intense study for years. In this thesis, the results on the identification of members of the parasitome in plant-parasitic nematodes, with a focus on the root-knot nematode species Meloidogyne chitwoodi and M. incognita, are described and discussed. In Chapter 1 a general introduction to the subject is given, which provides background on the main issues of this thesis. In Chapter 2 the first overview of the secretome and candidate parasitome members from M. chitwoodi are reported, which was at the start of the project a molecularly unexplored root-knot nematode species. As a starting point, all publicly available expressed sequence tags (ESTs) were used, i.e. 12,218, from three life stages of M. chitwoodi. A pipeline of bio-informatics algorithms was applied to screen this dataset for secretory proteins, which resulted in the identification of 398 secretome members. Two-third of these 398 sequences has no significant similarity to other known proteins in the nr database. In this study ‘parasitome’ is defined as ‘the products of parasitism genes secreted by a range of secretory organs in the nematode, including the oesophageal, amphidial, and rectal glands, the intestine, and the hypodermis’. To point out candidate parasitism gene products within the M. chitwoodi secretome, the site of expression of the most abundant secretome members was localised using the in situ hybridisation procedure. We found that at least eight out of the twenty most abundantly represented members of the secretome are specifically expressed in the single-celled oesophageal glands. One tag hybridised in the dorsal oesophageal gland, seven in the two subventral oesophageal glands, two in the intestine, and one tag hybridised to the tail tip in the proximity of the phasmids. Four sequences showed similarity to putative parasitism genes from other nematode species, whereas seven represented pioneering sequences. In case of three oesophageal gland specific parasitism genes, the predicted subcellular localization in host plant cells, following cleavage of the leader peptide for protein secretion, was nuclear. The results of this study, and those described in chapters 3, 4 and 5, show that the generation of ESTs indeed forms an invaluable resource for gene discovery, in this case (candidate) parasitism genes (McCarter et al., 2000). In particular, our findings obtained with in situ hybridisation microscopy make clear that with a combination of ESTs from high quality cDNA libraries and bioinformatics it is not necessary to use pre-selection methods to enrich libraries for transcripts from the gland region of a nematode. These pre-selections were based on tissue specific hybridizations (e.g. suppressive subtractive hybridization; Huang et al., 2004a), micro-aspiration of mRNA from specific regions in the parasite (Gao et al., 2003; Huang et al., 2003; Wang et al., 2001), and signal sequence trapping (Wang et al., 2001). Compared to these methods, EST generation from whole nematode libraries followed by bioinformatics analyses is relatively simple and more time efficient. Similarly, differential display methods, as were used by us and others in the past, seem less efficient than EST analysis. The consequence of using whole nematode cDNA libraries, without pre-selection, is that our strategy is most effective with larger numbers of ESTs. Five years ago the costs involved in the generation of ESTs were still a rate limiting factor. However, the generation of ESTs from (plant-)parasitic nematodes is an ongoing process which currently operates at a high pace. Fortunately, the ESTs that are being produced are made publicly available by efforts such as the Parasitic Nematode Sequencing Project (for an overview, see http://www.nematode.net/Species.Summaries/index.php (Wylie et al., 2004)). Our strategy may be further improved by using signal peptide prediction software which is trained on datasets built from nematode secretory proteins (e.g. Caenorhabditis elegans), since there is evidence that the composition of the signal peptides may vary between eukaryotic phyla (Pearson et al., 2005). Our strategy to identify the products of parasitism genes forms a basis for further functional analysis of parasite proteins involved in the nematode-host interaction. Our efforts were focused on the genes involved in the very early stages of parasitism by infective second stage juveniles, including host invasion and suppression of early defence responses in the host. In order to extend this basis, it would be interesting to enrich the EST dataset analysed in Chapter 2 with ESTs from later parasitic juvenile stages of M. chitwoodi. The same analyses can then be applied to these ESTs in order to point out candidate parasitism genes which are for example involved in giant cell formation and maintenance. Alternatively, or in addition, candidate parasitism genes can be pointed out by in silico comparison of the ESTs (and their abundances) between different developmental stages. In this way, EST analysis also forms an alternative for the technically more complicated differential display methods such as cDNA-AFLP and micro array analyses (De Boer et al., 2002b; Qin et al., 2000). In Chapter 3 and Chapter 4, the focus is on the repertoire of cell wall modifying proteins (CWMPs) which were identified from M. chitwoodi. These proteins form a subset of the parasitome of plant parasitic nematodes and play an important role during the intercellular migration of root-knot nematodes through host plant roots. Our study resulted in the identification of the most elaborate repertoire of CWMPs found in a single plant-parasitic nematode species so far. It includes -1,4-endoglucanases (cellulases), a -1,4-endoxylanase, pectate lyases, polygalacturonases, a cellulose binding protein (all described in Chapter 3) and expansin-like proteins (Chapter 4). Four β-1,4-endoglucanases were identified from M. chitwoodi and named Mc-eng-1 to Mc-eng-4. Full-length sequences were obtained for Mc-eng-1 to Mc-eng-3. The encoded cellulases were found to belong to glycoside hydrolase family 5 (GHF5) and were all predicted to contain an N-terminal signal peptide for secretion. In situ hybridisation localised the site of expression of Mc-eng-3 in the subventral oesophageal secretory glands of M. chitwoodi pre-parasitic J2-s. Attempts to obtain the full length sequence of Mc-eng-4 failed, but sequence similarity search results with the partial sequence revealed that this sequence encodes a cellulase that consists of a catalytic domain and a carbohydrate binding module (CBM). These results indicate that M. chitwoodi produces cellulases with two different domain architectures, either a single catalytic domain (Mc-ENG-1, Mc-ENG-2, and Mc-ENG-3) or a catalytic domain linked to a CBM (Mc-ENG-4). Our phylogenetic analysis of root-knot nematode cellulases, including Mc-ENG-1, Mc-ENG-2, and Mc-ENG-3, showed that there are probably no cellulases that are unique for one (group of) nematode species, but that there are several lines of cellulases common to root-knot nematodes. The site of Mc-cbp-1 (cellulose binding protein) expression was localised to the subventral oesophageal secretory glands of M. chitwoodi pre-parasitic J2-s. The Mc-CBP-1 protein consists of a bacterial family 2 CBM, which can mediate the binding to the cellulose substrate and/or may facilitate the non-hydrolytic disruption of cellulose fibres (Din et al., 1991; Levy et al., 2002; Shoseyov et al., 2006; Tomme et al., 1998). The CBM of Mc-CBP-1 is C-terminally linked to a domain of 80 amino acids that shares some similarity with Fn3-like domains. The latter domains are involved in protein-protein interactions, and in addition, the 80 amino acid N-terminal domain is rich in charged amino acids which are typically found on the hydrophilic surfaces of proteins to interact with other proteins. We hypothesise that Mc-CBP-1 is involved in the facilitation of cellulose hydrolysis, either as a docking station for CBM-lacking cellulases to bind to cellulose or by modification of the cellulose surface, or both. On the other hand, recent work by Gaulin et al. (2006) suggests that cellulose binding domains from the oomycete Phytophthora parasitica may also act as an elicitor of defence responses in plants (Gaulin et al., 2006). It is not yet clear if this is a general phenomenon of cellulose binding domains released by plant-pathogens and to what extent nematode cellulose binding proteins induce, either directly or indirectly through cell wall modifications, defence responses. Since Meloidogyne species migrate intercellularly through the roots of a host plant, they go through the middle lamella which is rich in pectic polysaccharides. The complete degradation of pectins into monomers requires the combined action of several types of pectinolytic enzymes. Root-knot nematodes do not feed on carbohydrates released from pectin degradation and thus do not require the complete degradation of plant cell wall pectins. In contrast, root-knot nematodes actually make use of a controlled and local degradation of the pectins in the middle lamella only and will benefit most likely from hydrolases and lyases (depolymerases), which only cleave the backbone of pectin (Tamaru and Doi, 2001), to weaken the intercellular bondings between cells. Both the pectate lyases and polygalacturonases identified from M. chitwoodi in this study belong to this group of pectin degrading enzymes. Four pectate lyases, named Mc-pel-1 to Mc-pel-4, were identified from M. chitwoodi invasive J2-s. In case of Mc-pel-1 and Mc-pel-2, full-length sequences were obtained, whereas Mc-pel-3 and Mc-pel-4 were represented by partial cDNA sequences in the M. chitwoodi ppJ2 cDNA library. Both Mc-pel-1 and Mc-pel-2 have an N-terminal signal peptide for secretion. In situ hybridisation performed on M. chitwoodi invasive J2-s revealed a specific expression confined to the subventral oesophageal secretory glands in case of Mc-pel-1. Based on similarity search results, the (putative) pectate lyases are all considered to be member of pectate lyase family III, which also comprises the pectate lyases from other plant-parasitic nematodes. Vanholme et al. (2007) showed with RNA interference (RNAi) knock-down experiments that a moderate but significant reduction in transcripts of a pectate lyase leads to strongly reduced infectivity of cyst nematodes (Vanholme et al., 2007). Cyst nematodes are not particularly focused on a stealthy invasion through the middle-lamella of host cells. Instead, they use a brute force approach during plant invasion and leave behind a trail of destructed cells. Stealthily invading root-knot nematodes are predicted to have stronger requirement for pectin degrading enzymes than cyst nematodes, however, there is no experimental data from RNAi knock-down experiments available at the moment to support this prediction. A family of at least four pectate lyases was found in M. chitwoodi, while others reported similar findings in earlier studies on M. incognita. The pectate substrate of pectate lyase is essentially built from repeating units of galacturonic acid with varying degrees of methylation. The actual composition of the oligogalacturonate units changes during the life time of a plant, and varies between different plant species. The overall topology of pectate lyases is the same, the core consists of β-strands forming a right-handed parallel β-helix, which suggests a conserved mode of action (D'Ovidio et al., 2004). In an attempt to explain multiple pectate lyase gene families in plant-pathogens, Herron et al. (2000) proposed that plant-pathogens with broad host ranges have multiple isozymes of pectate lyases with similar catalytic properties, but that recognize differently composed and decorated oligogalacturonate units (Herron et al., 2000). This proposal now seems to find support in the repertoire of pectate lyases of the polyphagous root-knot nematodes. Root-knot nematodes have evidently explored other venues, besides the production of pectate lyases, to achieve degradation of the pectin component in host cell walls. In total, two polygalacturonases, named Mc-pg-1 and Mc-pg-2, were identified from invasive J2-s of M. chitwoodi. The full-length sequence of Mc-pg-1 was obtained and with in situ hybridisation performed on M. chitwoodi invasive J2-s, a specific expression confined to the subventral oesophageal secretory glands was revealed. Based on similarities with other exo-acting polygalaturonases, we believe that Mc-PG-1 codes for an exo-polygalacturonase. The polygalacturonase MI-PG-1, which was cloned from M. incognita is also classified as an exo-acting enzyme (Jaubert et al., 2002). Exo- and endo-polygalacturonases have different impacts on the cell wall integrity of plants. Endo-polygalacturonases are known to play a role in plant defence responses, e.g. through the release of elicitor-active oligogalacturonides (Cervone et al., 1989; D'Ovidio et al., 2004; Favaron et al., 1988). Biotrophic pathogens, like Meloidogyne species, must avoid inducing host defence responses and therefore it seems likely that they will profit from a cell wall degrading enzyme repertoire that does not release oligogalacturonides. Experimental biochemical evidence for the classification of polygalacturonases from M. chitwoodi and M. incognita as exo-acting enzymes is still lacking. It will be interesting to investigate the enzymatic properties of these nematode polygalacturonases in order to establish to which class of polygalacturonases they belong. Most of the research on the role of cell wall degrading enzymes in plant-pathogen interactions is done on pectin degrading enzymes in pathogens of dicotyledons. Pectins, however, represent only a minor fraction of the cell wall components in monocots, which mainly include hemicelluloses in their cell wall matrices (Carpita, 1996). A major part of the hemicellulose fraction of monocotyledonous cell walls consists of substituted xylan polymers. It is therefore expected that xylan-degrading enzymes have an important role in promoting virulence of pathogens of monocots; possibly equivalent to the role of pectinolytic enzymes for pathogens of dicots (Beliën et al., 2006 and references herein). We identified two novel xylan-degrading enzymes in two root-knot nematode species that are virulent pathogens of both dicotyledons and monocotyledons, i.e. M. chitwoodi and M. incognita (described in Chapter 3 and 5). These xylan-degrading enzymes are β-1,4-endoxylanases, which are capable of hydrolysing substituted xylan polymers into fragments of random size. The β-1,4-endoxylanase from M. incognita, named Mi-xyl1, is the first functional β-1,4-endoxylanase of animal origin. Both Mi-xyl1 and the β-1,4-endoxylanase from M. chitwoodi, Mc-xyl-1, were found to be expressed in the subventral oesophageal gland cells of the nematode and the encoded proteins were predicted to have an N-terminal signal peptide for secretion. Based on similarity search results we consider MI-XYL1 and Mc-XYL-1 enzymes that belong to GHF5. These two β-1,4-endoxylanases share 39% identity in their catalytic domain. Further comparison between the two xylanases showed that Mc-XYL-1 has an ancillary stretch of 76 amino acids at its C-terminus which is, based on sequence similarity, a putative xylan binding module (XBM). So far, results of Southern blot analysis and EST database screenings did not show evidence of endoxylanases in nematode species that have specialized on dicots (M. hapla, G. rostochiensis and G. pallida), in spite of thousands of ESTs in public sequence databases. Based on our findings, we hypothesise that the production of β-1,4-endoxylanase in plant parasitic nematodes is correlated with parasitism on monocots. RNA interference studies as mentioned above may provide information about the requirement of β-1,4-endoxylanase for infectivity on monocots. In addition, it would be interesting to investigate if the specialist nematodes of monocots indeed have a bias towards xylan degradation in their repertoire of cell wall modifying proteins. Unfortunately, nematode specialists on graminaceous monocots, such as the cereal cyst nematode Heterodera avenea, have not been included in the scope of molecular nematology so far. Therefore, there is currently no sequence information available of such nematode species. Chapter 4 deals with the expansin-like proteins identified in Meloidogyne spp., with the emphasis on those from M. chitwoodi. Expansins form a diverse protein superfamily in plants and play a role in various biological processes in which re-arrangement of plant cell wall polysaccharides is involved (Cosgrove, 2000a). Plant expansins lack hydrolytic activity and are proposed to weaken non-covalent interactions between cellulose and hemicellulose polymers (McQueen-Mason and Cosgrove, 1994). They seem to act synergistically with cellulases by making the plant cell wall polysaccharides more accessible to enzymatic attack (Cosgrove, 2000a). A similar synergistic action of expansins and cell wall degrading enzymes secreted by plant-pathogens may facilitate their invasion of the host. The distribution of expansins was believed to be restricted to land plants (Cosgrove, 2000a), but recently, a small number of expansin-like sequences have been identified from other organisms (Darley et al., 2003; Laine et al., 2000; Saloheimo et al., 2002). In addition, a functional β-expansin, Gr-EXPB1, was found to be secreted by the plant-parasitic nematode Globodera rostochiensis (Kudla et al., 2005; Qin et al., 2004). This latter sequence was used to query assembled ESTs from M. chitwoodi and found in total four expansin-like sequences, named Mc-EXP1 to Mc-EXP4. These M. chitwoodi sequences harbour most of the signature motifs of α- and β-expansins, but since we were not able to produce active recombinant protein to test expansin activity of the proteins on plant tissues, the sequences are designated as expansin-like proteins. Expansin-like Mc-EXP1 is represented by 23 ESTs in the EST dataset from M. chitwoodi ppJ2-s and belongs to the top 20 of most abundantly expressed members of the secretome of this developmental stage of M. chitwoodi (Chapter 2 of this thesis). Both Mc-EXP1 and Mc-EXP2 are multi-domain proteins. They consist of an N-terminal signal peptide for secretion, followed by either a CBM (Mc-EXP1) or a Lysin-motif domain (LysM; Mc-EXP2) both linked to a C-terminal expansin-like domain. Compared to the domain structure of plant expansins, the expansin-like domain and the polysaccharide-binding domain (CBM) are in reverse orientation. Thus, the domain structure of Mc-EXP1 is similar to that of Gr-EXPB1. The sequence identity between the C-termini of Mc-EXP1 and Mc-EXP2 and plant expansins ranged from 30% to 35%. Notably, for both expansin-like proteins highest sequence similarity is with α-expansins from plants. Conclusive evidence on the type of activity of Mc-EXP1 and Mc-EXP2, i.e. do they resemble more the α- or -expansins, requires the production of active recombinant protein and subsequent cell wall extension assays on the different types of cell walls. Unfortunately, we have not been able to achieve heterologous expression of Mc-EXPs in plants, which would have allowed us to study the type of expansin activity of the proteins. Probes designed on Mc-EXP1 localised the gene transcription in the subventral oesophageal secretory glands of ppJ2-s of M. chitwoodi. The same localisation was found for the Mc-EXP1 and Mc-EXP2 proteins, which was investigated by immunofluorescence microscopy with specific antisera to Mc-EXP1 and Mc-EXP2. A strong fluorescence, with a granular pattern, was observed in the subventral oesophageal gland extensions and ampullae and to a lesser extent in the gland lobes. The presence of a signal peptide for secretion and the localisation of the transcripts and protein in the subventral oesophageal secretory glands of M. chitwoodi ppJ2-s strongly suggest that expansin-like Mc-EXP1 is secreted by the nematode. The current data on the developmental expression and production of expansin-like Mc-EXP2 are less straightforward. Mc-EXP2 was identified in a library made from nematode eggs and could not be amplified from the cDNA library made from M. chitwoodi ppJ2-s. In addition, no in situ hybridisation signal was observed in M. chitwoodi ppJ2-s with an antisense probe spanning the N-terminal putative LysM domain. On the other hand, in immunofluorescence microscopy experiments, the Mc-EXP2 protein was localised specifically in the subventral oesophageal glands of M. chitwoodi ppJ2-s. Therefore, a developmental expression study needs to be performed on Mc-EXP2 in order to find out if the results obtained at the protein level are either confirmed or contradicted by results obtained at the transcriptional level. The other two expansin-like proteins from M. chitwoodi, Mc-EXP3 and Mc-EXP4, only contain an expansin-like domain with an N-terminal signal peptide for secretion. Mc-EXP4 is represented by 55 ESTs from the ppJ2 stage of M. chitwoodi and herewith it belongs to the top 10 of most abundantly represented transcripts in the M. chitwoodi EST dataset (see Chapter 2 of this thesis). Despite its high abundance, no hybridisation signal was detected for Mc-EXP4 by in situ hybridisation in ppJ2-s of M. chitwoodi. Specific staining of the subventral oesophageal secretory glands in this developmental stage was observed in case of Mc-EXP3. In addition to the expansin-like genes in M. chitwoodi, evidence was also found for the presence of expansin-like genes in five other root-knot nematode species, a root-lesion nematode species, and plant pathogenic oomycete and fungal species. It remains to be shown whether these organisms produce functional expansins, but our findings make us point at two things. Firstly, it appears that expansins do not occur in only a small number of organisms outside the plant kingdom, but instead, are widespread and likely to be involved in many plant-pathogen interactions and plant-microbe interactions in general. Secondly, the proposed and currently adopted nomenclature of the expansin superfamily might need revision. For, the designations ‘expansin’ (expansin A and expansin B) and ‘expansin-like’ (expansin-like family A and B) in this nomenclature are kept exclusively for plant proteins. Proteins from other organisms that share structural similarity with both domains of plant expansins are grouped in a separate ‘catch-all’ category and designated ‘expansin-related’, solely based on the fact that they do not originate from plants. When the expansin-like proteins identified in this study are indeed functional expansins, there would be no biological rationale behind the ‘catch-all’ group of expansin-related proteins. A series of conserved cysteines (C) and the HFD motif around amino acid position 110 are used as the key signatures of the plant expansin family (Cosgrove, 2000b). A series of conserved cysteines is also present along the backbone of the root-knot nematode expansin-like domains. In case of the HFD motif, only the H and D residues are conserved in the expansin-like proteins of root-knot nematodes. A similar conservation was found in the β-expansin Gr-EXPB1 from G. rostochiensis, for which cell wall expansion activity on type II primary cell walls was found (Kudla et al., 2005; Qin et al., 2004). These findings suggest that the functional significance of the HFD motif is obscure and can be clarified by the biochemical characterization of root-knot nematode expansin-like proteins, e.g. by site directed mutagenesis followed by activity assays. It seems likely that the CWMP repertoire of M. chitwoodi mirrors its wide host plant range since the latter represents a broad diversity of cell wall polysaccharides. The CWMPs identified from M. chitwoodi enable the migrating nematode to cleave the backbone of all major types of plant cell wall polysaccharides and to modify their interactions. The subventral oesophageal gland specific expression and presence of predicted secretion signal peptides suggest that the CWMPs are secreted from the nematode. Hereby, this enzyme complex facilitates the intercellular migration of M. chitwoodi in the host plant. Post-transcriptional gene silencing by soaking nematode juveniles in double-stranded RNA (RNA interference) was successfully applied to assess the importance of β-1,4-endoglucanases and pectate lyases for infection of plant roots by cyst nematodes (Chen et al., 2005; Vanholme et al., 2007). Similar RNA interference experiments targeting CWMP encoding genes from M. chitwoodi, followed by infection tests on monocots and dicots, may reveal the influence of the individual CWMPs on the host range of M. chitwoodi. On the other hand, our findings raise the questions how it is possible that a nematode can secrete such a variety of CWMPs without i) inflicting any detectable damage to the cells along the migratory track and ii) being noticed by the host plant? An explanation for the fact that plant cells remain intact might be that the cell wall modifying enzymes that are found to be produced by the nematode only cleave the backbone of plant cell wall polysaccharides. As mentioned above, root-knot nematodes do not feed on carbohydrates released from plant cell wall degradation and thus do not require the complete degradation of plant cell wall polysaccharides. In contrast, root-knot nematodes only seem to weaken the intercellular bondings between cells by local secretion of the CWMPs through the stylet into the middle lamella resulting in controlled degradation. This controlled degradation might also be a way to prevent the production of elicitor-active oligogalacturonides involved in plant defence responses. Alternatively or in addition, the nematode might constantly repress the host defence mechanisms during intercellular migration through the roots. In relation to these questions it is interesting to study whether the nematode can ‘sense’, either inside or outside the plant root system or both, with which type of plant (cell wall) it is dealing. Following the perception of cell wall type signals, the nematode could adapt its repertoire of CWMPs to it. This would enable the nematode to avoid the production of CWMPs that do not act on the available substrate, thus saving valuable energy and reducing the chance of being detected by the host. Potato root exudates are for example known to induce the secretion of proteins by cyst nematodes (Smant et al., 1997). One way to test whether host plant root exudates already influence the production of CWMPs at the transcriptional level would be to treat eggs or hatched nematodes with different root exudates followed by real-time quantitative PCR analysis. One could for example hatch nematodes in water and in root exudates obtained from a monocot and a dicot and compare the transcriptional levels of a certain CWMP.

AB - The products of parasitism genes in nematodes must be secreted to reach their targets at the nematode-plant interface. These nematode secretory proteins are therefore recognised to play an important role in the nematode-plant interaction and as a result have been subject of intense study for years. In this thesis, the results on the identification of members of the parasitome in plant-parasitic nematodes, with a focus on the root-knot nematode species Meloidogyne chitwoodi and M. incognita, are described and discussed. In Chapter 1 a general introduction to the subject is given, which provides background on the main issues of this thesis. In Chapter 2 the first overview of the secretome and candidate parasitome members from M. chitwoodi are reported, which was at the start of the project a molecularly unexplored root-knot nematode species. As a starting point, all publicly available expressed sequence tags (ESTs) were used, i.e. 12,218, from three life stages of M. chitwoodi. A pipeline of bio-informatics algorithms was applied to screen this dataset for secretory proteins, which resulted in the identification of 398 secretome members. Two-third of these 398 sequences has no significant similarity to other known proteins in the nr database. In this study ‘parasitome’ is defined as ‘the products of parasitism genes secreted by a range of secretory organs in the nematode, including the oesophageal, amphidial, and rectal glands, the intestine, and the hypodermis’. To point out candidate parasitism gene products within the M. chitwoodi secretome, the site of expression of the most abundant secretome members was localised using the in situ hybridisation procedure. We found that at least eight out of the twenty most abundantly represented members of the secretome are specifically expressed in the single-celled oesophageal glands. One tag hybridised in the dorsal oesophageal gland, seven in the two subventral oesophageal glands, two in the intestine, and one tag hybridised to the tail tip in the proximity of the phasmids. Four sequences showed similarity to putative parasitism genes from other nematode species, whereas seven represented pioneering sequences. In case of three oesophageal gland specific parasitism genes, the predicted subcellular localization in host plant cells, following cleavage of the leader peptide for protein secretion, was nuclear. The results of this study, and those described in chapters 3, 4 and 5, show that the generation of ESTs indeed forms an invaluable resource for gene discovery, in this case (candidate) parasitism genes (McCarter et al., 2000). In particular, our findings obtained with in situ hybridisation microscopy make clear that with a combination of ESTs from high quality cDNA libraries and bioinformatics it is not necessary to use pre-selection methods to enrich libraries for transcripts from the gland region of a nematode. These pre-selections were based on tissue specific hybridizations (e.g. suppressive subtractive hybridization; Huang et al., 2004a), micro-aspiration of mRNA from specific regions in the parasite (Gao et al., 2003; Huang et al., 2003; Wang et al., 2001), and signal sequence trapping (Wang et al., 2001). Compared to these methods, EST generation from whole nematode libraries followed by bioinformatics analyses is relatively simple and more time efficient. Similarly, differential display methods, as were used by us and others in the past, seem less efficient than EST analysis. The consequence of using whole nematode cDNA libraries, without pre-selection, is that our strategy is most effective with larger numbers of ESTs. Five years ago the costs involved in the generation of ESTs were still a rate limiting factor. However, the generation of ESTs from (plant-)parasitic nematodes is an ongoing process which currently operates at a high pace. Fortunately, the ESTs that are being produced are made publicly available by efforts such as the Parasitic Nematode Sequencing Project (for an overview, see http://www.nematode.net/Species.Summaries/index.php (Wylie et al., 2004)). Our strategy may be further improved by using signal peptide prediction software which is trained on datasets built from nematode secretory proteins (e.g. Caenorhabditis elegans), since there is evidence that the composition of the signal peptides may vary between eukaryotic phyla (Pearson et al., 2005). Our strategy to identify the products of parasitism genes forms a basis for further functional analysis of parasite proteins involved in the nematode-host interaction. Our efforts were focused on the genes involved in the very early stages of parasitism by infective second stage juveniles, including host invasion and suppression of early defence responses in the host. In order to extend this basis, it would be interesting to enrich the EST dataset analysed in Chapter 2 with ESTs from later parasitic juvenile stages of M. chitwoodi. The same analyses can then be applied to these ESTs in order to point out candidate parasitism genes which are for example involved in giant cell formation and maintenance. Alternatively, or in addition, candidate parasitism genes can be pointed out by in silico comparison of the ESTs (and their abundances) between different developmental stages. In this way, EST analysis also forms an alternative for the technically more complicated differential display methods such as cDNA-AFLP and micro array analyses (De Boer et al., 2002b; Qin et al., 2000). In Chapter 3 and Chapter 4, the focus is on the repertoire of cell wall modifying proteins (CWMPs) which were identified from M. chitwoodi. These proteins form a subset of the parasitome of plant parasitic nematodes and play an important role during the intercellular migration of root-knot nematodes through host plant roots. Our study resulted in the identification of the most elaborate repertoire of CWMPs found in a single plant-parasitic nematode species so far. It includes -1,4-endoglucanases (cellulases), a -1,4-endoxylanase, pectate lyases, polygalacturonases, a cellulose binding protein (all described in Chapter 3) and expansin-like proteins (Chapter 4). Four β-1,4-endoglucanases were identified from M. chitwoodi and named Mc-eng-1 to Mc-eng-4. Full-length sequences were obtained for Mc-eng-1 to Mc-eng-3. The encoded cellulases were found to belong to glycoside hydrolase family 5 (GHF5) and were all predicted to contain an N-terminal signal peptide for secretion. In situ hybridisation localised the site of expression of Mc-eng-3 in the subventral oesophageal secretory glands of M. chitwoodi pre-parasitic J2-s. Attempts to obtain the full length sequence of Mc-eng-4 failed, but sequence similarity search results with the partial sequence revealed that this sequence encodes a cellulase that consists of a catalytic domain and a carbohydrate binding module (CBM). These results indicate that M. chitwoodi produces cellulases with two different domain architectures, either a single catalytic domain (Mc-ENG-1, Mc-ENG-2, and Mc-ENG-3) or a catalytic domain linked to a CBM (Mc-ENG-4). Our phylogenetic analysis of root-knot nematode cellulases, including Mc-ENG-1, Mc-ENG-2, and Mc-ENG-3, showed that there are probably no cellulases that are unique for one (group of) nematode species, but that there are several lines of cellulases common to root-knot nematodes. The site of Mc-cbp-1 (cellulose binding protein) expression was localised to the subventral oesophageal secretory glands of M. chitwoodi pre-parasitic J2-s. The Mc-CBP-1 protein consists of a bacterial family 2 CBM, which can mediate the binding to the cellulose substrate and/or may facilitate the non-hydrolytic disruption of cellulose fibres (Din et al., 1991; Levy et al., 2002; Shoseyov et al., 2006; Tomme et al., 1998). The CBM of Mc-CBP-1 is C-terminally linked to a domain of 80 amino acids that shares some similarity with Fn3-like domains. The latter domains are involved in protein-protein interactions, and in addition, the 80 amino acid N-terminal domain is rich in charged amino acids which are typically found on the hydrophilic surfaces of proteins to interact with other proteins. We hypothesise that Mc-CBP-1 is involved in the facilitation of cellulose hydrolysis, either as a docking station for CBM-lacking cellulases to bind to cellulose or by modification of the cellulose surface, or both. On the other hand, recent work by Gaulin et al. (2006) suggests that cellulose binding domains from the oomycete Phytophthora parasitica may also act as an elicitor of defence responses in plants (Gaulin et al., 2006). It is not yet clear if this is a general phenomenon of cellulose binding domains released by plant-pathogens and to what extent nematode cellulose binding proteins induce, either directly or indirectly through cell wall modifications, defence responses. Since Meloidogyne species migrate intercellularly through the roots of a host plant, they go through the middle lamella which is rich in pectic polysaccharides. The complete degradation of pectins into monomers requires the combined action of several types of pectinolytic enzymes. Root-knot nematodes do not feed on carbohydrates released from pectin degradation and thus do not require the complete degradation of plant cell wall pectins. In contrast, root-knot nematodes actually make use of a controlled and local degradation of the pectins in the middle lamella only and will benefit most likely from hydrolases and lyases (depolymerases), which only cleave the backbone of pectin (Tamaru and Doi, 2001), to weaken the intercellular bondings between cells. Both the pectate lyases and polygalacturonases identified from M. chitwoodi in this study belong to this group of pectin degrading enzymes. Four pectate lyases, named Mc-pel-1 to Mc-pel-4, were identified from M. chitwoodi invasive J2-s. In case of Mc-pel-1 and Mc-pel-2, full-length sequences were obtained, whereas Mc-pel-3 and Mc-pel-4 were represented by partial cDNA sequences in the M. chitwoodi ppJ2 cDNA library. Both Mc-pel-1 and Mc-pel-2 have an N-terminal signal peptide for secretion. In situ hybridisation performed on M. chitwoodi invasive J2-s revealed a specific expression confined to the subventral oesophageal secretory glands in case of Mc-pel-1. Based on similarity search results, the (putative) pectate lyases are all considered to be member of pectate lyase family III, which also comprises the pectate lyases from other plant-parasitic nematodes. Vanholme et al. (2007) showed with RNA interference (RNAi) knock-down experiments that a moderate but significant reduction in transcripts of a pectate lyase leads to strongly reduced infectivity of cyst nematodes (Vanholme et al., 2007). Cyst nematodes are not particularly focused on a stealthy invasion through the middle-lamella of host cells. Instead, they use a brute force approach during plant invasion and leave behind a trail of destructed cells. Stealthily invading root-knot nematodes are predicted to have stronger requirement for pectin degrading enzymes than cyst nematodes, however, there is no experimental data from RNAi knock-down experiments available at the moment to support this prediction. A family of at least four pectate lyases was found in M. chitwoodi, while others reported similar findings in earlier studies on M. incognita. The pectate substrate of pectate lyase is essentially built from repeating units of galacturonic acid with varying degrees of methylation. The actual composition of the oligogalacturonate units changes during the life time of a plant, and varies between different plant species. The overall topology of pectate lyases is the same, the core consists of β-strands forming a right-handed parallel β-helix, which suggests a conserved mode of action (D'Ovidio et al., 2004). In an attempt to explain multiple pectate lyase gene families in plant-pathogens, Herron et al. (2000) proposed that plant-pathogens with broad host ranges have multiple isozymes of pectate lyases with similar catalytic properties, but that recognize differently composed and decorated oligogalacturonate units (Herron et al., 2000). This proposal now seems to find support in the repertoire of pectate lyases of the polyphagous root-knot nematodes. Root-knot nematodes have evidently explored other venues, besides the production of pectate lyases, to achieve degradation of the pectin component in host cell walls. In total, two polygalacturonases, named Mc-pg-1 and Mc-pg-2, were identified from invasive J2-s of M. chitwoodi. The full-length sequence of Mc-pg-1 was obtained and with in situ hybridisation performed on M. chitwoodi invasive J2-s, a specific expression confined to the subventral oesophageal secretory glands was revealed. Based on similarities with other exo-acting polygalaturonases, we believe that Mc-PG-1 codes for an exo-polygalacturonase. The polygalacturonase MI-PG-1, which was cloned from M. incognita is also classified as an exo-acting enzyme (Jaubert et al., 2002). Exo- and endo-polygalacturonases have different impacts on the cell wall integrity of plants. Endo-polygalacturonases are known to play a role in plant defence responses, e.g. through the release of elicitor-active oligogalacturonides (Cervone et al., 1989; D'Ovidio et al., 2004; Favaron et al., 1988). Biotrophic pathogens, like Meloidogyne species, must avoid inducing host defence responses and therefore it seems likely that they will profit from a cell wall degrading enzyme repertoire that does not release oligogalacturonides. Experimental biochemical evidence for the classification of polygalacturonases from M. chitwoodi and M. incognita as exo-acting enzymes is still lacking. It will be interesting to investigate the enzymatic properties of these nematode polygalacturonases in order to establish to which class of polygalacturonases they belong. Most of the research on the role of cell wall degrading enzymes in plant-pathogen interactions is done on pectin degrading enzymes in pathogens of dicotyledons. Pectins, however, represent only a minor fraction of the cell wall components in monocots, which mainly include hemicelluloses in their cell wall matrices (Carpita, 1996). A major part of the hemicellulose fraction of monocotyledonous cell walls consists of substituted xylan polymers. It is therefore expected that xylan-degrading enzymes have an important role in promoting virulence of pathogens of monocots; possibly equivalent to the role of pectinolytic enzymes for pathogens of dicots (Beliën et al., 2006 and references herein). We identified two novel xylan-degrading enzymes in two root-knot nematode species that are virulent pathogens of both dicotyledons and monocotyledons, i.e. M. chitwoodi and M. incognita (described in Chapter 3 and 5). These xylan-degrading enzymes are β-1,4-endoxylanases, which are capable of hydrolysing substituted xylan polymers into fragments of random size. The β-1,4-endoxylanase from M. incognita, named Mi-xyl1, is the first functional β-1,4-endoxylanase of animal origin. Both Mi-xyl1 and the β-1,4-endoxylanase from M. chitwoodi, Mc-xyl-1, were found to be expressed in the subventral oesophageal gland cells of the nematode and the encoded proteins were predicted to have an N-terminal signal peptide for secretion. Based on similarity search results we consider MI-XYL1 and Mc-XYL-1 enzymes that belong to GHF5. These two β-1,4-endoxylanases share 39% identity in their catalytic domain. Further comparison between the two xylanases showed that Mc-XYL-1 has an ancillary stretch of 76 amino acids at its C-terminus which is, based on sequence similarity, a putative xylan binding module (XBM). So far, results of Southern blot analysis and EST database screenings did not show evidence of endoxylanases in nematode species that have specialized on dicots (M. hapla, G. rostochiensis and G. pallida), in spite of thousands of ESTs in public sequence databases. Based on our findings, we hypothesise that the production of β-1,4-endoxylanase in plant parasitic nematodes is correlated with parasitism on monocots. RNA interference studies as mentioned above may provide information about the requirement of β-1,4-endoxylanase for infectivity on monocots. In addition, it would be interesting to investigate if the specialist nematodes of monocots indeed have a bias towards xylan degradation in their repertoire of cell wall modifying proteins. Unfortunately, nematode specialists on graminaceous monocots, such as the cereal cyst nematode Heterodera avenea, have not been included in the scope of molecular nematology so far. Therefore, there is currently no sequence information available of such nematode species. Chapter 4 deals with the expansin-like proteins identified in Meloidogyne spp., with the emphasis on those from M. chitwoodi. Expansins form a diverse protein superfamily in plants and play a role in various biological processes in which re-arrangement of plant cell wall polysaccharides is involved (Cosgrove, 2000a). Plant expansins lack hydrolytic activity and are proposed to weaken non-covalent interactions between cellulose and hemicellulose polymers (McQueen-Mason and Cosgrove, 1994). They seem to act synergistically with cellulases by making the plant cell wall polysaccharides more accessible to enzymatic attack (Cosgrove, 2000a). A similar synergistic action of expansins and cell wall degrading enzymes secreted by plant-pathogens may facilitate their invasion of the host. The distribution of expansins was believed to be restricted to land plants (Cosgrove, 2000a), but recently, a small number of expansin-like sequences have been identified from other organisms (Darley et al., 2003; Laine et al., 2000; Saloheimo et al., 2002). In addition, a functional β-expansin, Gr-EXPB1, was found to be secreted by the plant-parasitic nematode Globodera rostochiensis (Kudla et al., 2005; Qin et al., 2004). This latter sequence was used to query assembled ESTs from M. chitwoodi and found in total four expansin-like sequences, named Mc-EXP1 to Mc-EXP4. These M. chitwoodi sequences harbour most of the signature motifs of α- and β-expansins, but since we were not able to produce active recombinant protein to test expansin activity of the proteins on plant tissues, the sequences are designated as expansin-like proteins. Expansin-like Mc-EXP1 is represented by 23 ESTs in the EST dataset from M. chitwoodi ppJ2-s and belongs to the top 20 of most abundantly expressed members of the secretome of this developmental stage of M. chitwoodi (Chapter 2 of this thesis). Both Mc-EXP1 and Mc-EXP2 are multi-domain proteins. They consist of an N-terminal signal peptide for secretion, followed by either a CBM (Mc-EXP1) or a Lysin-motif domain (LysM; Mc-EXP2) both linked to a C-terminal expansin-like domain. Compared to the domain structure of plant expansins, the expansin-like domain and the polysaccharide-binding domain (CBM) are in reverse orientation. Thus, the domain structure of Mc-EXP1 is similar to that of Gr-EXPB1. The sequence identity between the C-termini of Mc-EXP1 and Mc-EXP2 and plant expansins ranged from 30% to 35%. Notably, for both expansin-like proteins highest sequence similarity is with α-expansins from plants. Conclusive evidence on the type of activity of Mc-EXP1 and Mc-EXP2, i.e. do they resemble more the α- or -expansins, requires the production of active recombinant protein and subsequent cell wall extension assays on the different types of cell walls. Unfortunately, we have not been able to achieve heterologous expression of Mc-EXPs in plants, which would have allowed us to study the type of expansin activity of the proteins. Probes designed on Mc-EXP1 localised the gene transcription in the subventral oesophageal secretory glands of ppJ2-s of M. chitwoodi. The same localisation was found for the Mc-EXP1 and Mc-EXP2 proteins, which was investigated by immunofluorescence microscopy with specific antisera to Mc-EXP1 and Mc-EXP2. A strong fluorescence, with a granular pattern, was observed in the subventral oesophageal gland extensions and ampullae and to a lesser extent in the gland lobes. The presence of a signal peptide for secretion and the localisation of the transcripts and protein in the subventral oesophageal secretory glands of M. chitwoodi ppJ2-s strongly suggest that expansin-like Mc-EXP1 is secreted by the nematode. The current data on the developmental expression and production of expansin-like Mc-EXP2 are less straightforward. Mc-EXP2 was identified in a library made from nematode eggs and could not be amplified from the cDNA library made from M. chitwoodi ppJ2-s. In addition, no in situ hybridisation signal was observed in M. chitwoodi ppJ2-s with an antisense probe spanning the N-terminal putative LysM domain. On the other hand, in immunofluorescence microscopy experiments, the Mc-EXP2 protein was localised specifically in the subventral oesophageal glands of M. chitwoodi ppJ2-s. Therefore, a developmental expression study needs to be performed on Mc-EXP2 in order to find out if the results obtained at the protein level are either confirmed or contradicted by results obtained at the transcriptional level. The other two expansin-like proteins from M. chitwoodi, Mc-EXP3 and Mc-EXP4, only contain an expansin-like domain with an N-terminal signal peptide for secretion. Mc-EXP4 is represented by 55 ESTs from the ppJ2 stage of M. chitwoodi and herewith it belongs to the top 10 of most abundantly represented transcripts in the M. chitwoodi EST dataset (see Chapter 2 of this thesis). Despite its high abundance, no hybridisation signal was detected for Mc-EXP4 by in situ hybridisation in ppJ2-s of M. chitwoodi. Specific staining of the subventral oesophageal secretory glands in this developmental stage was observed in case of Mc-EXP3. In addition to the expansin-like genes in M. chitwoodi, evidence was also found for the presence of expansin-like genes in five other root-knot nematode species, a root-lesion nematode species, and plant pathogenic oomycete and fungal species. It remains to be shown whether these organisms produce functional expansins, but our findings make us point at two things. Firstly, it appears that expansins do not occur in only a small number of organisms outside the plant kingdom, but instead, are widespread and likely to be involved in many plant-pathogen interactions and plant-microbe interactions in general. Secondly, the proposed and currently adopted nomenclature of the expansin superfamily might need revision. For, the designations ‘expansin’ (expansin A and expansin B) and ‘expansin-like’ (expansin-like family A and B) in this nomenclature are kept exclusively for plant proteins. Proteins from other organisms that share structural similarity with both domains of plant expansins are grouped in a separate ‘catch-all’ category and designated ‘expansin-related’, solely based on the fact that they do not originate from plants. When the expansin-like proteins identified in this study are indeed functional expansins, there would be no biological rationale behind the ‘catch-all’ group of expansin-related proteins. A series of conserved cysteines (C) and the HFD motif around amino acid position 110 are used as the key signatures of the plant expansin family (Cosgrove, 2000b). A series of conserved cysteines is also present along the backbone of the root-knot nematode expansin-like domains. In case of the HFD motif, only the H and D residues are conserved in the expansin-like proteins of root-knot nematodes. A similar conservation was found in the β-expansin Gr-EXPB1 from G. rostochiensis, for which cell wall expansion activity on type II primary cell walls was found (Kudla et al., 2005; Qin et al., 2004). These findings suggest that the functional significance of the HFD motif is obscure and can be clarified by the biochemical characterization of root-knot nematode expansin-like proteins, e.g. by site directed mutagenesis followed by activity assays. It seems likely that the CWMP repertoire of M. chitwoodi mirrors its wide host plant range since the latter represents a broad diversity of cell wall polysaccharides. The CWMPs identified from M. chitwoodi enable the migrating nematode to cleave the backbone of all major types of plant cell wall polysaccharides and to modify their interactions. The subventral oesophageal gland specific expression and presence of predicted secretion signal peptides suggest that the CWMPs are secreted from the nematode. Hereby, this enzyme complex facilitates the intercellular migration of M. chitwoodi in the host plant. Post-transcriptional gene silencing by soaking nematode juveniles in double-stranded RNA (RNA interference) was successfully applied to assess the importance of β-1,4-endoglucanases and pectate lyases for infection of plant roots by cyst nematodes (Chen et al., 2005; Vanholme et al., 2007). Similar RNA interference experiments targeting CWMP encoding genes from M. chitwoodi, followed by infection tests on monocots and dicots, may reveal the influence of the individual CWMPs on the host range of M. chitwoodi. On the other hand, our findings raise the questions how it is possible that a nematode can secrete such a variety of CWMPs without i) inflicting any detectable damage to the cells along the migratory track and ii) being noticed by the host plant? An explanation for the fact that plant cells remain intact might be that the cell wall modifying enzymes that are found to be produced by the nematode only cleave the backbone of plant cell wall polysaccharides. As mentioned above, root-knot nematodes do not feed on carbohydrates released from plant cell wall degradation and thus do not require the complete degradation of plant cell wall polysaccharides. In contrast, root-knot nematodes only seem to weaken the intercellular bondings between cells by local secretion of the CWMPs through the stylet into the middle lamella resulting in controlled degradation. This controlled degradation might also be a way to prevent the production of elicitor-active oligogalacturonides involved in plant defence responses. Alternatively or in addition, the nematode might constantly repress the host defence mechanisms during intercellular migration through the roots. In relation to these questions it is interesting to study whether the nematode can ‘sense’, either inside or outside the plant root system or both, with which type of plant (cell wall) it is dealing. Following the perception of cell wall type signals, the nematode could adapt its repertoire of CWMPs to it. This would enable the nematode to avoid the production of CWMPs that do not act on the available substrate, thus saving valuable energy and reducing the chance of being detected by the host. Potato root exudates are for example known to induce the secretion of proteins by cyst nematodes (Smant et al., 1997). One way to test whether host plant root exudates already influence the production of CWMPs at the transcriptional level would be to treat eggs or hatched nematodes with different root exudates followed by real-time quantitative PCR analysis. One could for example hatch nematodes in water and in root exudates obtained from a monocot and a dicot and compare the transcriptional levels of a certain CWMP.

KW - meloidogyne

KW - plantenparasitaire nematoden

KW - pathogenesis-gerelateerde eiwitten

KW - eiwitsecretie

KW - celwandstoffen

KW - genexpressie

KW - nucleotidenvolgordes

KW - genetische kartering

KW - meloidogyne

KW - plant parasitic nematodes

KW - pathogenesis-related proteins

KW - protein secretion

KW - cell wall components

KW - gene expression

KW - nucleotide sequences

KW - genetic mapping

M3 - internal PhD, WU

SN - 9789085048589

PB - s.n.

CY - [S.l.]

ER -