A MANDA C OTTAGE 1 AND P ETER U RWIN 2 **
15.2. Genetic Engineering for Nematode Resistance
Use of Natural Resistances
Chapter 14 examines resistance and resistant cultivars in detail, so the transfer of cloned resistance genes is summarized only briefly here. Disease resistance in plants was thought to be mediated by two complementary genes; a pathogen avirulence gene (Avr) and a host resistance gene (R). This gene for gene hypothesis was first pro- posed over 50 years ago by Harold Flor, from studying host responses to flax rust fungus (Flor, 1971). It was thought that the host R-gene product acts as a receptor to the pathogen’s Avr-gene product, which functions as a ligand; binding of Avr to R activates plant defence mechanisms. This elicitor–receptor model may function in some resistance mechanisms but in recent years research has shown a decided lack of directR–Avr interaction. The guard hypothesis proposes that the R proteins associ- ate with the cellular targets of the Avr proteins (Dangl and Jones, 2001). These cel- lular targets could be proteins involved in plant defence, for example, or be required to provide pathogen nutrition. The guard hypothesis proposes that when a pathogen Avr-gene product binds to a target in a resistant plant cell, this complex is recognized by the R protein, which initiates the plant’s defence. Therefore, in a susceptible host no R protein is present and the plant target is unguarded from the pathogen virulence elicitor resulting in disease. The R protein may be constitutively bound to the target but become unbound and active upon binding of the Avr product to the target. A fur- ther model (the decoy model) proposes that R proteins do not recognize Avr protein targets directly but proteins that resemble them; in the absence of an R-gene these decoys do not contribute to pathogen fitness, but in its presence elicit plant defence mechanisms (van der Hoorn and Kamoun, 2008). The decoy theory arose as it appears that a limited number of host cellular targets are targeted by several pathogen effectors. The major class of plant R-genes encode proteins containing a nucleotide- binding and leucine-rich repeat (NB–LRR). Nearly half the genes encoding NB–LRR proteins in Arabidopsis thaliana are single copy; thus, divergent evolution has not been necessary to provide effective resistance possibly due to the structural constraints of the pathogen elicitors imposed by the binding of plant targets. Transferring R-genes to susceptible hosts would seem to offer an effective and long-term solution to disease and as early as 1993 a patent protecting sequences conferring resistance to Heterodera schachtii in sugar beet was submitted (Sandal et al., 1993). Numerous nematode R- genes have since been identified (see Chapter 14) and those that have been utilized in engineering resistance include Hs1 pro (Cai et al., 1997), Mi-1 (Milligan et al., 1998),
Hero (Ernst et al., 2002) and Gpa2 (van der Vossen et al., 2000). The gene products ofMi-1, Hero and Gpa2 belong to the leucine-zipper, nucleotide-binding site (NBS), LRR-containing class of R proteins and are probably located in the cell cytoplasm, whilst the Hs1 pro gene encodes an extracellular protein. The first of these R-genes to be cloned was Hs1 (pro-1) and was identified as conferring resistance to H. schachtii in a wild species of beet, Beta procumbens, and conferred resistance to a susceptible beet variety in hairy root culture (Cai et al., 1997). The second R-gene to be identified was the tomato Mi gene, which confers resistance to M. incognita and several other species of Meloidogyne; it had been cloned as part of a larger DNA fragment (Milligan et al., 1998), of which the Mi1.2 element was shown to be involved not only in resist- ance to nematodes but also to potato aphids. The potato Gpa2 R-gene was initially isolated from a cluster of R-genes at the Gpa2 locus; at least two of the genes in the cluster are active, one of which (Rx1) confers resistance to potato virus X and the other, Gpa2, confers resistance to the potato cyst nematode Globodera pallida (van der Vossen et al., 2000).
The tomato resistance gene Hero was isolated from a region containing 14 homologous genes, of which eight appear to be functional genes. Hero encodes an NBS–LRR type R protein with a leucine zipper motif within the N-terminal domain (Ernstet al., 2002). Hero is unusual amongst NBS–LRR type R proteins as it contains a negatively-charged region between LRRs 4 and 5; the resulting peptide sequence is predicted to form a coil, the likely function of which is pathogen recognition (Ernst et al., 2002). Hero confers resistance to a number of cyst nematodes but predominantly Globodera spp. and resistance levels of 95% have been observed for G. rostochiensis.
Cyst nematodes fail to progress to maturity in resistant plants as, although a syncytium is initiated by juvenile nematodes, it quickly becomes encapsulated by necrotic cells leading to its degradation (Sobczak et al., 2005). This resistance mechanism is not associated with the rapid hypersensitive response (HR) initiated by many R-gene products, such as that initiated by the Mi-1 gene where necrosis of cells adjacent to the migratory tract and feeding cell occurs within 12 h following root invasion.
Despite the identification of a number of R-genes conferring nematode resistance there have been relatively few successes using these genes as part of an engineered resistance. In particular, many confer resistance to one species only and even to a specific pathotype. Transfer of R-genes to a different host species has resulted in loss of resistance. For example, transfer of Hero A (from tomato) into potato resulted in no significant resistance to Globodera spp. (Sobczak et al., 2005); similarly, transfer of Mi-1.2 (from tomato) to tobacco resulted in no significant resistance to Meloidogyne species (reviewed in Williamson, 1998). A further limiting factor of Mi resistance is that it is temperature-sensitive and breaks down over 28°C, limiting its use to temperate climes.
Two studies have identified entirely novel plant-nematode resistance mecha- nisms; one comprises altered host gene expression, the other altered protein regula- tion. The first of these studies identified resistance conferred by gene copy number variation, resulting in overexpression of three disparate genes encoding an amino acid transporter, an a-SNAP protein and a WI12 (wound-inducible domain) protein (Cooket al., 2012). These genes are located at the rhg1-b locus, which confers resist- ance to the soybean cyst nematode H. glycines. In susceptible soybean varieties one copy of the genes is found per haploid genome but ten tandem copies are found in therhg1-b haplotype. Overexpression of all three genes was shown to be required for
Rhg1-mediatedH. glycines resistance. The mechanism by which overexpression con- fers nematode resistance is unknown, but sequence comparisons and structure predic- tions of the encoded proteins suggests that they may combine to produce an unfavourable environment for nematodes; overexpression of the amino acid trans- porter may result in altered auxin biosynthesis or distribution, possibly impacting on feeding cell initiation and maintenance, the a-SNAP protein may have a role in exo- cytosis, thus increased abundance may result in removal of products required by the nematode for development or for feeding cell maintenance, while the WI12 protein may be involved in the production of phenazine-like compounds, which are toxic to nematodes (Cook et al., 2012).
The second study identified resistance conferred by altered regulation of a serine hydroxymethyltransferase (Shmt) (Liu et al., 2012). The Shmt encoding gene was identified at the Rhg4 (resistance to H. glycines 4) locus and was shown to confer nematode resistance. Shmt has an essential role in one carbon folate metabolism and while alleles of Rhg4 encode functional enzymes, polymorphisms are predicted to disrupt Shmt regulation. Folate one-carbon metabolism is essential for nematode feeding cell development and maintenance and loss of Shmt regulation leading to nematode resistance may occur via a number of mechanisms as a consequence of folate deficiency in syncytia; nematode starvation may result with subsequent degeneration of the syncytia, or a hypersensitive response may be generated to the folate-deficient syncytia leading to nematode death (folate deficiency has previously been shown to induce apoptosis of mammalian cells). Further explanations of the resistance mechanism are that the compound produced by the altered Shmt acts either as a nematicide or effector protein, triggering resistance signalling pathways (Liuet al., 2012).
Both studies offer insights into further engineered solutions for nematode resist- ance; overexpression of a combination of genes by either increasing copy number or altering transcript levels using promoters to produce elevated levels of transcripts and, hence, protein products can create an environment noxious to nematodes. The same can be achieved by removing regulatory elements from proteins, which would result in the production of compounds necessary for nematode parasitism being less abundant or functionally altered.
15.2.1. BroadeningR-gene resistance
There is evidence to suggest that the products of different R-genes share common downstream signalling cascades, the intermediates of which could be engineered to respond to a wider range of nematode species. Research has identified some of these convergent signalling molecules; for example, RAR1 was identified as a com- mon convergence point in signalling pathways initiated by fungi and viruses (Liu et al., 2002), SGT1 was shown to interact with Rar1 (Azevedo et al., 2002) and the heat shock protein HSP90 was identified as their molecular chaperone (Takahashi et al., 2003). However, only HSP90 was shown to be essential for Mi-1 mediated resistance to aphids and nematodes (Bhattarai et al., 2007). With an increase in the understanding of how resistance genes exert their function, it may be possible to broaden the specificity of a resistance or increase its durability by directed modification.
15.2.2. An alternative strategy of using natural resistances
In theory, determining Avr structure and utilizing ligand–receptor binding modelling may suggest engineering solutions that confer less receptor specificity, thus enabling R-gene mediated response to a wider range of nematode species. However, this approach may not be possible as a putative Avr-gene product (MAP-1) thought to interact with Mi was found exclusively in avirulent populations of M. incognita and otherMeloidogyne species (Semblat et al., 2001). The presence of this protein only in avirulent populations discounts R protein modification as no ligand would appear to exist in virulent populations. However, this may not be the situation for other Avr- gene products, which may differ between virulent and avirulent species by amino acid sequence, thus offering a means of receptor modification to achieve control. RBP-1 is a SPRY domain containing protein secreted by G. pallida, which elicits an HR response in potato plants expressing the R-geneGpa2. RBP-1 is highly polymorphic and initiates a resistance response only if a proline residue is present at position 187 in the SPRY domain. However, whilst the LRR domain of Gpa2 recognizes RBP-1 variants and initiates activation of the HR response, the interaction of Gpa2 with RBP-1 is entirely dependent on a third protein RanGap2 (Sacco et al., 2009).
The tacit assumption is that recognition of nematode Avr effector proteins triggers plant NB–LRR protein mediated cell death. However, in one case the opposite is true;
theG. rostochiensis effector protein SPRYSEC-19 was shown to associate with the LRR domain of a SW5 resistance gene in tomato; this association did not result in pro- grammed cell death and thus resistance, but conversely was shown to suppress it (Postma et al., 2012). This was true for other (SW5B, Rx1, Gpa2 and RGH10) but not all NB–
LRR resistance proteins (where presence of SPRYSEC-19 did result in cell death). This new insight into nematode effector protein function offers a further means of control as targeting nematode effector proteins that suppress cell death would enhance resistance.
Currently very little is definitively established as regards Avr–R-gene product inter- actions; however, genomic sequences are now published for M. incognita and M. hapla, providing new opportunities to identify Avr and other genes involved in parasitism.