SEARCH

SEARCH BY CITATION

Keywords:

  • NB-ARC-LRR;
  • virulence spectrum;
  • race specificity;
  • intragenic allele pyramiding;
  • powdery mildew;
  • wheat

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Some plant resistance genes occur as allelic series, with each member conferring specific resistance against a subset of pathogen races. In wheat, there are 17 alleles of the Pm3 gene. They encode nucleotide-binding (NB-ARC) and leucine-rich-repeat (LRR) domain proteins, which mediate resistance to distinct race spectra of powdery mildew. It is not known if specificities from different alleles can be combined to create resistance genes with broader specificity. Here, we used an approach based on avirulence analysis of pathogen populations to characterize the molecular basis of Pm3 recognition spectra. A large survey of mildew races for avirulence on the Pm3 alleles revealed that Pm3a has a resistance spectrum that completely contains that of Pm3f, but also extends towards additional races. The same is true for the Pm3b and Pm3c gene pair. The molecular analysis of these allelic pairs revealed a role of the NB-ARC protein domain in the efficiency of effector-dependent resistance. Analysis of the wild-type and chimeric Pm3 alleles identified single residues in the C-terminal LRR motifs as the main determinant of allele specificity. Variable residues of the N-terminal LRRs are necessary, but not sufficient, to confer resistance specificity. Based on these data, we constructed a chimeric Pm3 gene by intragenic allele pyramiding of Pm3d and Pm3e that showed the combined resistance specificity and, thus, a broader recognition spectrum compared with the parental alleles. Our findings support a model of stepwise evolution of Pm3 recognition specificities.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Plants evolved two lines of defence against pathogen infections (Dodds and Rathjen, 2010). The first is the basal resistance that relies on pre-formed physical and chemical barriers, and an immune system that induces defence responses upon the detection of pathogen-associated molecular patterns (PAMPs; Schwessinger and Zipfel, 2008). To fight microbes that produce effector molecules suppressing this resistance, plants evolved a second line of defence, which detects the presence or the action of pathogen effectors (Chisholm et al., 2006; Jones and Dangl, 2006). This recognition is mediated by the products of resistance (R) genes, which induce a strong resistance reaction that stops the infection. Such a pathogen is avirulent on the host, and the detected effector is an avirulence (Avr) factor. The pairwise interaction between R and Avr gene products is characterized genetically as gene-for-gene resistance (Flor, 1971). Although R proteins recognize and respond to a wide variety of pathogen-derived effectors, they are built from a very limited set of modular domains (Dangl and Jones, 2001). These include a nucleotide-binding (NB) domain followed by ARC1, ARC2 (thus NB-ARC) and a C-terminal leucine-rich-repeat (LRR) domain. The ARC subdomains were named after their presence in the human apoptotic protease-activating factor 1 (APAF-1), R proteins and the Caenorhabditis elegans Death-4 (CED-4) protein (van der Biezen and Jones, 1998). NB-ARC domains are highly similar in structure to mammalian NACHT domains (Albrecht and Takken, 2006). In the plant R proteins I-2, Mi-1 and N, they were shown to bind and hydrolyse ATP (Tameling et al., 2002; Ueda et al., 2006). These findings, together with further structure–function analyses, indicate that the NB-ARC domain works as a reversible molecular switch (Takken et al., 2006; Rairdan and Moffett, 2007; Danot et al., 2009; Lukasik and Takken, 2009).

The LRR domain is a tandem array of repeats that are typically 20–29 amino acids long (Kobe and Kajava, 2001). Each repeat contains a conserved motif with the consensus sequence LxxLxLxxN(Cx)xL (Wei et al., 2008). Crystal structures of LRR domains revealed that the second conserved leucine and adjacent residues in the consensus sequence form a short β-strand, and that the β-strands of the different LRRs are arranged in parallel (Enkhbayar et al., 2004). These parallel β-strands form a β-sheet that lines the concave face of a horseshoe-shaped structure. The first five of the x-residues are exposed on the concave surface, and several studies demonstrated their involvement in the binding of interaction partners (reviewed in Bella et al., 2008). In plant R proteins, these residues are thought to mediate recognition specificity, as they were shown to be highly variable and under diversifying selection (e.g. Ellis et al., 2000; Mondragon-Palomino et al., 2002; Seeholzer et al., 2010). Direct evidence for a role in specificity determination comes from domain swap and mutagenesis experiments (reviewed by DeYoung and Innes, 2006; Dunning et al., 2007). However, there is emerging evidence that the LRR domain is also involved in diverse intra- and intermolecular interactions, which are not directly implicated in pathogen recognition, but contribute to R protein activity regulation (Jones and Takemoto, 2004; McDowell and Simon, 2006; Lukasik and Takken, 2009).

Many R genes confer resistance only to a subset of all existing pathogen races. Among them is the multiallelic Pm3 locus from hexaploid wheat (Triticum aestivum L.), which confers race-specific resistance to wheat powdery mildew (Blumeria graminis f.sp. tritici). PM3 belongs to the subgroup of NB-LRR proteins encoding an N-terminal coiled-coiled (CC) domain. In the modern bread wheat gene pool, it occurs in seven, functionally distinct, true alleles, Pm3a–Pm3g, which have been molecularly isolated (Yahiaoui et al., 2004; 2006; Srichumpa et al., 2005). Recently, the alleles Pm3k–Pm3t were cloned from tetraploid wheat species and hexaploid wheat land races (Bhullar et al., 2009; 2010; Yahiaoui et al., 2009). The Pm3 resistance alleles are highly similar in sequence to the susceptible allele Pm3CS, which also represents the consensus sequence of all resistance alleles (Yahiaoui et al., 2006).

To improve disease resistance of plants, it is advantageous to make use of R genes conferring broad spectrum resistance. The artificial extension of the recognition spectrum was successful in the Rx protein conferring viral resistance in potato (Solanum tuberosum) (Farnham and Baulcombe, 2006). In flax (Linum usitatissimum), detailed molecular studies were performed on R alleles L5, L6 and L7, which show overlapping resistance specificities to flax rust (Melampsora lini) (Luck et al., 2000; Dodds et al., 2006; Wang et al., 2007). Recombinant genes were found to confer resistance to only a subset of the rust strain recognized by either of the parental L alleles (Ellis et al., 1999, 2007; Luck et al., 2000). Apart from the flax alleles, relatively little is known about the resistance spectra and their overlap for different R alleles from a specific locus. Therefore, it is not clear if natural allelic specificities might be combined. Here, we tested a large set of powdery mildew isolates for recognition by Pm3a–Pm3g alleles to determine whether there are natural examples for extensions in recognition capacities. Indeed, two Pm3 pairs were identified where one allele recognized all pathogen races that are also recognized by the second allele, but this one allele extended the recognition spectrum to an additional set of mildew isolates. To elucidate the molecular mechanism leading to these functional differences, a series of domain swap experiments was performed. The very low number of sequence polymorphisms between functionally different Pm3 alleles makes them an ideal system to study the molecular basis of race specificity. Based on the results obtained, we investigated if it is possible to rationally design Pm3 genes with broadened disease resistance by combining specificities from different Pm3 alleles. We could demonstrate that intragenic allele pyramiding of Pm3d and Pm3e leads to a functional gene with dual resistance specificities, thus achieving an extended resistance spectrum.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Two pairs of Pm3 alleles with narrow and extended resistance spectra

We carried out a virulence survey for Pm3a–Pm3g alleles based on 102 isolates in 2007 to determine the avirulence spectra of wheat powdery mildew isolates present in agricultural ecosystems in Switzerland. The data obtained were combined with the results of earlier studies from 1992 to 1998. In total, data from more than 710 powdery mildew isolates tested on differential lines for the Pm3a–Pm3d alleles, from 494 isolates tested on Pm3f lines, and from 102 isolates for the Pm3e and Pm3g lines were available. We analysed whether one of the Pm3 alleles represents a natural example with an extended resistance spectrum compared with another Pm3 gene. In this case, there would be no isolate showing virulence on the broad-spectrum allele and avirulence on the corresponding narrow-spectrum allele, and so we specifically searched for missing combinations of virulence/avirulence on the different Pm3 alleles. We found numerous isolates avirulent on Pm3a, and either avirulent on Pm3f (AvrPm3a/AvrPm3f) or virulent on Pm3f [AvrPm3a/avrPm3f; note that ‘avr’ with a small ‘a’ indicates the absence of the corresponding avirulence (Avr) factor], as well as isolates virulent on both alleles (Figure 1a; Table S1a). However, no isolates were found with the combination of avirulence on Pm3f and virulence on Pm3a. A similar pattern was observed for virulence/avirulence on the allelic pair Pm3b and Pm3c, where all powdery mildew isolates avirulent on Pm3c were also avirulent on Pm3b, and none were avirulent on Pm3c and virulent on Pm3b (AvrPm3c/avrPm3b; Figure 1b; Table S1b). These observations were confirmed by a re-examination of earlier publications describing powdery mildew infection tests on Pm3 alleles. In these publications, isolates scored as avirulent on Pm3f or Pm3c were never virulent on Pm3a or Pm3b wheat lines, respectively (Briggle, 1969; Huang and Röder, 2004; Huang et al., 2004). Based on the frequencies of virulence on Pm3a (8.1%) and Pm3b (7.9%), and of avirulence on Pm3f (35.0%) and Pm3c (47.5%), in our studied mildew population, we expected to observe 14 isolates virulent on Pm3a and avirulent on Pm3f (out of 494 isolates), and 27 isolates virulent on Pm3b and avirulent on Pm3c (out of 710 isolates). The absence of these (a)virulence combinations is very unlikely to be the result of statistical fluctuation (χ2 test, P < 0.001 for both pairs), but can be explained by an extended resistance spectrum of Pm3a and Pm3b compared with Pm3f and Pm3c.

image

Figure 1. Pm3a/Pm3f and Pm3b/Pm3c avirulence gene combinations found in powdery mildew isolates. (a) All 173 isolates avirulent on Pm3f were also avirulent on Pm3a, and thus are included in the total of 454 isolates avirulent on Pm3a. (b) Similarly, all 337 isolates avirulent on Pm3c were also avirulent on Pm3b. No isolates were found that were virulent on Pm3a and avirulent on Pm3f (avrPm3a/AvrPm3f) or virulent on Pm3b and avirulent on Pm3c (avrPm3b/AvrPm3c).

Download figure to PowerPoint

Polymorphisms both in the ARC2 and LRR domains of PM3A are required for full PM3A-mediated resistance

We hypothesized that particular sequence segments in the allelic pairs Pm3a/Pm3f and Pm3b/Pm3c could be responsible for the shared resistance spectrum. We first analysed the sequence differences between the Pm3a and Pm3f alleles, which are more similar to each other than to any other Pm3 allele. The encoded proteins PM3A and PM3F exclusively share, among all PM3 allelic variants, two residues each in LRRs 2, 13 and 14, and another five residues in LRR 27 (Figure 2a); with PM3B, they have in common two polymorphic residues in LRR 1 (Figure 2a) and a sequence block in the spacer region, which connects the NB-ARC and the LRR domain (Figure 2b). PM3A and PM3F differ from each other by two clearly delimited, polymorphic sequence blocks: one in LRRs 19–22 (13 amino acid differences) and one in the ARC2 domain and spacer region (19 amino acid differences in the ARC2, and one in the spacer). Therefore, we addressed the question of how these two segments determine the observed differences in the resistance spectra of the two alleles.

image

Figure 2.  Amino acid polymorphisms in the PM3APM3G proteins. (a) The first sequence (PM3CS) lists the consensus residues at the polymorphic sites of the leucine-rich-repeat (LRR) domain (vertical numbers give the amino acid positions in the PM3 protein). Below the PM3CS sequence, the polymorphic residues and the corresponding PM3 proteins are indicated. Dots represent residues identical to those in PM3CS and deletions are shown as dashes. Amino acids at the x-positions of the LxxLxLxx motif are highlighted in red. The LRR numbers are given underneath. (b) Polymorphic sites in the NB-ARC domains of PM3 proteins are listed below the corresponding PM3CS residues. Spacer is the region connecting the NB-ARC domain with the LRR domain. *The NB-ARC and spacer sequences of PM3CPM3E and PM3G are identical to PM3CS.

Download figure to PowerPoint

We constructed three chimeras between Pm3a and Pm3f (Figure 3). The first two, Pm3a-fARC and Pm3f-aARC, were produced in vitro by reciprocally exchanging the ARC-encoding sequences between Pm3a and Pm3f. Similarly, the replacement of the LRR-encoding sequence of Pm3a with the corresponding sequence of Pm3f yielded Pm3a-fLRR19–22. This third construct is identical to Pm3f-aARC, except for a threonine at position 543 (T543) in the spacer region, which is shared with PM3A, but not with PM3F (Figures 2b and 3). These and all further constructs used in this study were functionally tested in a transient expression system (Schweizer et al., 1999), where they were biolistically delivered into leaf epidermal cells of the wheat line Chancellor that does not carry an endogenous copy of Pm3 (Yahiaoui et al., 2004). The leaves were subsequently infected with a specific powdery mildew isolate and the infection level was quantified as a haustorium index (HI) that gives the percentage of susceptible interactions in transformed cells. The Pm3a/Pm3f chimeras were tested with a powdery mildew isolate that distinguishes the reaction of Pm3a from Pm3f. Based on wheat seedling infection results, we chose isolate 97028, which is avirulent on Pm3a (HI of 17%) and significantly more virulent on Pm3f (49% HI; Student’s t-test, P < 0.001; Figure 3). As a susceptible control, we used Pm3CS (76% HI), the naturally occurring susceptible Pm3 allele that has the consensus sequence of the resistance alleles Pm3a–Pm3g, and is equal to an empty vector control (Yahiaoui et al., 2006). All three chimeric constructs showed a resistance level intermediate between Pm3a and Pm3f (32, 24 and 25% HI; Student’s t-test, P < 0.01; Figure 3). This indicates that: (i) the PM3A-specific sequence in both the ARC2 domain and the LRRs 19–22 contribute independently to the increased resistance of PM3A compared with PM3F; (ii) only their combination leads to the resistance level of PM3A; and (iii) that the single amino acid difference between PM3F-AARC and PM3A-FLRR19–22 in the spacer region has no significant effect on the resistance level (Student’s t-test, P = 0.116). As controls, we tested the chimeric constructs with powdery mildew isolates 96224 and 07201, which are avirulent and virulent, respectively, on both Pm3a and Pm3f (Figure 3). They were not significantly different from Pm3a and Pm3f when tested with isolate 96224 (Student’s t-test, P > 0.4), nor from Pm3CS when tested with isolate 07201 (Student’s t-test, P > 0.06). Thus, the conclusions above are not the result of artefacts like instability or autoactivity of the chimeric proteins.

image

Figure 3.  Polymorphisms both in the ARC2 domain and in leucine-rich-repeats (LRRs) 19–22 of PM3A are required for full PM3A-mediated resistance. Left panel: Schematic diagram of wild-type and chimeric PM3 proteins. On top, the domain structure is indicated (not drawn to scale). Bars represent polymorphic amino acids compared with the consensus sequence PM3CS, which was used as a susceptible control (blue bars, polymorphisms in the NB-ARC domain; black bars, polymorphisms in the spacer region; red bars, polymorphisms in the x-residues of the LxxLxLxx motif of the LRRs; yellow bars, all other polymorphisms in LRRs). Right panel: Constructs encoding the proteins shown in the left panel were driven by the CaMV 35S promoter and tested by transient expression in susceptible wheat leaves. Leaves were biolistically transformed, challenged with a single powdery mildew isolate, and resistance responses were evaluated by microscopic analyses. The powdery mildew isolate used is indicated below the graph, and the presence (Avr) or absence (avr) of the relevant avirulence genes, as deduced from leaf segment infection tests, is stated. The result is indicated with the haustorium index, which gives the percentage of compatible interactions (values report the mean of three independent experiments and error bars give the standard deviation; for comprehensive statistical analysis see Table S2). The relevant significant differences at *P < 0.05, **P < 0.01 and ***P < 0.001 are indicated.

Download figure to PowerPoint

The NB-ARC domain of PM3 proteins controls the efficiency of effector-dependent resistance

Interestingly, the two broad-spectrum proteins PM3A and PM3B both have a distinct NB-ARC domain sharing an identical ARC2 sequence, whereas the NB-ARC domains encoded by Pm3c–Pm3g and Pm3CS are identical (Figure 2b). In particular, PM3A and PM3B share all polymorphic residues in the ARC2 domain. The latter protein has two additional amino acid differences in the NB, and 19 differences in the ARC1 domain compared with all other PM3 proteins. We wanted to investigate whether the characteristic NB-ARC domain of PM3B also contributes to the difference in recognition between PM3B and PM3C. Several powdery mildew isolates avirulent on Pm3b and virulent on Pm3c in seedling infection tests were avirulent on transiently transformed cells expressing Pm3c. We made a similar observation for Pm3a and Pm3f, where there was no isolate showing full virulence on Pm3f and avirulence on Pm3a in the transient assay (Figure 3). This might be because of the nature of the transient assay in which the genes are overexpressed. In fact, this result indicates that PM3C confers similar resistance specificity as PM3B when it is present at higher levels. Thus, it was not possible to directly relate the presence of the PM3B-specific NB-ARC to the functional difference of PM3B compared with PM3C. Instead, we performed domain-swap experiments using the PM3B NB-ARC domain to test its contribution to the resistance function of other Pm3 genes and to Pm3b-specific resistance.

We reciprocally exchanged the NB-ARC encoding region of Pm3b and Pm3d to construct the recombinant genes Pm3b-dNB-ARC and Pm3d-bNB-ARC (Figure 4a). These chimeras were functionally tested with the isolates 97011 (avirulent on Pm3d; virulent on Pm3b) and 96229 (avirulent on Pm3b; virulent on Pm3d), which discriminate Pm3b- from Pm3d-dependent resistance. Like Pm3b, Pm3b-dNB-ARC was not functional against isolate 97011. However, it was significantly less effective than Pm3b against isolate 96229 (40% HI compared with 11% HI; Student’s t-test, P < 0.01). This indicates that polymorphisms in the NB-ARC of PM3B enhance the PM3B-dependent resistance response. The second construct, Pm3d-bNB-ARC, showed significant quantitative differences compared with Pm3d (Student’s t-test, < 0.05): the resistance to 97011 was stronger (an HI reduction from 31 to 16%), and the susceptibility to 96229 was decreased from 82 to 53% HI. This result raises the question whether the PM3B-specific NB-ARC domain causes a weak autoactivation of PM3D. Therefore, we challenged leaves transiently expressing Pm3d-bNB-ARC with the isolate 07016, which is virulent on both Pm3b and Pm3d. This resulted in an HI of 77% (Figure 4a), demonstrating that Pm3d-bNB-ARC is not autoactive.

image

Figure 4.  Polymorphisms in the NB-ARC domain of PM3B enhance resistance activity. (a) Chimeric constructs of Pm3b and Pm3d were transiently expressed in wheat and tested for resistance against powdery mildew isolates 97011 and 96229, which discriminate Pm3b- from Pm3d-dependent resistance. The construct Pm3d-bNB-ARC was also challenged with a virulent control isolate (07016). (b) A domain swap between Pm3b and Pm3CS was functionally analysed using isolates 96229 and 96224. The most relevant polymorphic residues are specified by the single letter code above the drawings. The designations and experimental procedures are the same as described in Figure 3. Values report the mean of three (isolates 97011 and 96229) or two (isolate 07016) independent experiments, and error bars give the standard deviation. For comprehensive statistical analysis see Tables S3, S4, and S5.

Download figure to PowerPoint

To further test the hypothesis of whether the PM3B NB-ARC domain enhances the resistance response, we made the construct Pm3CS-bNB-ARC by replacing the consensus NB-ARC domain of the susceptible protein PM3CS with the NB-ARC of PM3B. We tested it functionally by using isolate 96229 and an isolate avirulent on all seven alleles Pm3a–Pm3g (96224; Figure 4b). Pm3CS-bNB-ARC did not confer resistance to either of the isolates. This indicates that the NB-ARC domain of PM3B does not contribute to the recognition of isolate 96229 (and 96224). The lower HI of PM3B compared with PM3B-DNB-ARC, and of PM3D-BNB-ARC compared with PM3D, upon challenge with isolate 96229 (Figure 4a) can be explained by a higher resistance protein activity caused by the PM3B NB-ARC domain. It is likely that the LRR domain of PM3D (but not of PM3CS; Figure 4b) weakly senses the presence of isolate 96229, but that the presence of the PM3B NB-ARC in PM3D-BNB-ARC activates the protein sufficiently to allow the triggering of a (still weak) resistance response. A role of the PM3B NB-ARC in the activity, but not the resistance specificity, of PM3 proteins is also consistent with the increased resistance of PM3D-BNB-ARC compared with PM3D to isolate 97011 (Figure 4a). Considering the sequence similarity of the NB-ARC in PM3B and PM3A (Figure 2b), it is tempting to speculate that PM3A NB-ARC also increases the resistance activity, causing the reduced HI of PM3A compared with PM3A-FARC and of PM3F-AARC, and PM3A-FLRR19-22 compared with PM3F, when challenged with isolate 97028 (Figure 3).

Modelling polymorphic residues of PM3A and PM3B in the NB-ARC domain structure

To better understand the impact of the amino acid polymorphisms between the PM3A/B- and the PM3CS-type of NB-ARC domains, we determined their position relative to conserved sequences in other R proteins. Therefore, the consensus sequence PM3CS was added to the structure-based multiple sequence alignment of the NB-ARC domains of different R proteins published in van Ooijen et al. (2008b). On PM3CS, the positions of the polymorphic amino acids present in PM3A and/or PM3B were marked (Figure S1). Only four of them aligned with residues conserved in the majority of the R proteins, and the underlying substitutions in PM3A/PM3B were conservative (V362M, V396I, T400S, F466I). Furthermore, the polymorphic positions do not map to gain- and loss-of-function positions described in the other R proteins, nor corresponded to positions predicted to be involved in ADP binding. Based on the alignment, we constructed a protein structure model of PM3CS using the crystal structure of human APAF-1 (Riedl et al., 2005) as a template in order to find the 3D position of polymorphic sites in the protein. Alignment gaps were substituted by loop modelling of PM3CS. Remarkably, polymorphic sites are evenly distributed in the ARC1 (Figure 5a), whereas 16 of the 18 variant amino acids located in the ARC2 cluster on one side of the subdomain (Figure 5b). Their side chains point to the outside of the subdomain, with only three exceptions (P463, F466, S495), and two side chains have interdomain contacts with the NB domain (L483, E485). Furthermore, there are two loops in the PM3CS sequence that are considerably longer compared with APAF-1, and six polymorphic sites locate there. These observations derived from the protein structure model of PM3CS indicate an important role of one side of the ARC2 subdomain for molecular interaction and signalling.

image

Figure 5.  Polymorphic amino acids of PM3A and PM3B locate mostly on one side of the ARC2 subdomain. A protein structure model of the PM3CS sequence was constructed to localize variant amino acids in 3D. Residues that correspond to polymorphic amino acids in PM3B are indicated in orange; those present in both PM3A and PM3B are highlighted in red. The NB domain is coloured in cyan, the ARC1 subdomain is coloured in marine and the ARC2 subdomain is coloured in dark blue. Bound ADP is represented as sticks in CPK atom colours. (a) View of the complete NB-ARC domain. Variant amino acids are evenly spread in the ARC1 subdomain. (b) Alternative view of the ARC2 subdomain with transparent surface visualization to show the 3D distribution of polymorphic sites. Of 18 polymorphic sites, 16 locate to one side of the subdomain, and only three of those are completely buried.

Download figure to PowerPoint

Polymorphic residues in both the N- and C-terminal LRRs of PM3B, PM3C and PM3D are necessary for resistance specificity

PM3A and PM3F have identical sequences both in the N-terminal and the C-terminal part of the LRR domain (Figure 2a). PM3B and PM3C differ in LRRs 1–4, but share the polymorphic residue I1309 in LRR 26 compared with PM3CS. In contrast, the remaining PM3 proteins show unique polymorphic residues in the C-terminal part of the LRR (LRRs 26–28). Thus, only the allelic pairs Pm3a/Pm3f and Pm3b/Pm3c encode identical C-terminal LRRs. We hypothesized that this sequence identity is the basis of their overlap in the resistance spectrum. Therefore, we studied the dependence of PM3B and PM3C function on the shared polymorphic residue I1309. Pm3b and Pm3c were recombined with Pm3CS, resulting in the constructs Pm3b-CSLRR26, Pm3c-CSLRR26 and Pm3CS-b/cLRR26 (Figure 6a). These constructs were transiently expressed and the transformed cells were challenged with the powdery mildew isolates 96224 and, partly, 97019, which are avirulent on Pm3b and Pm3c (Figure 6a). Pm3b-CSLRR26 and Pm3c-CSLRR26 were both partially compromised in resistance function. The residual resistance activity was effector dependent, as they were not significantly different from Pm3CS (Student’s t-test, P > 0.2) when challenged with the virulent isolates 07016 and 07296, respectively. Pm3CS-b/cLRR26 also showed partial resistance to 96224 (49% HI), but full susceptibility to 97019 (85% HI; Figure 6a). These results demonstrate that I1309 is functionally important in both PM3B and PM3C. Its replacement in PM3C and PM3B by the conserved methionine of PM3CS leads to a decreased resistance level, but I1309 alone (construct Pm3CS-b/cLRR26) is not sufficient to confer complete Pm3b- or Pm3c-dependent resistance. In PM3C-CSLRR26, the residual recognition capacities are the result of sequence polymorphisms in the N-terminal LRRs. One of these residues, R588, is also shared by PM3B, as well as by PM3A and PM3F (Figure 2a). As we have previously shown that the NB-ARC domain does not contribute to the recognition of isolate 96224 (Figure 4b), the residual recognition capacities of PM3B-CSLRR26 (Figure 6a) must result from the two polymorphisms in LRR 1 and/or polymorphic residues in the spacer region.

image

Figure 6.  Polymorphic residues in both C- and N-terminal LRRs contribute to Pm3c- and Pm3d-specific resistance. (a) The residue I1309 plays a similar role in PM3B- and PM3C-mediated resistance. (b) The swapping of the C-terminal LRRs of PM3B and PM3D disrupts their resistance specificity. Note that the wild-type controls Pm3b and Pm3d in (b) are identical to those in Figure 4(a), as Pm3b, Pm3d, Pm3b-dNB-ARC, Pm3d-bNB-ARC, Pm3b-dLRR22,28 and Pm3d-b/cLRR26 were tested in parallel. The designations and experimental procedures are the same as described in Figure 3. Values report the mean of three (isolates 96224, 97019, 97011 and 96229) or two (isolates 07016 and 07296) independent experiments, and error bars give the standard deviation. For comprehensive statistical analysis see Tables S4, S6 and S7.

Download figure to PowerPoint

We conclude that polymorphic amino acids at both N- and C-terminal ends of the LRR are necessary to confer full Pm3c- and possibly also Pm3b-specific resistance. This is reminiscent of our previous studies on PM3D, which was also functionally dependent on polymorphisms in both C- and N-terminal LRRs (Yahiaoui et al., 2006). In these studies, both the replacement of the only polymorphic residue in the N-terminal LRRs (W659 in LRR 4), or the replacement of C-terminal polymorphisms (R1155 in LRR 22 and R1358 in LRR 28) by conserved residues of PM3CS led to a complete loss of resistance function. To further analyse the role of amino acid polymorphisms in the N- and C-terminal LRRs of PM3 proteins, we reciprocally exchanged the last eight LRRs of PM3B and PM3D in which they differ by three amino acids in the LRRs 22, 26 and 28 (Figure 6b). The resulting constructs, Pm3b-dLRR22,28 and Pm3d-b/cLRR26, failed to confer resistance to the two tested isolates, 97011 and 96229. The chimera Pm3b-dLRR22,28 showed residual resistance activity (40% HI) to isolate 96229, similar to Pm3b-CSLRR26, which also caused a weak resistance response to isolate 96224 (Figure 6a). Also, in this case, the residual resistance was effector-dependent, as Pm3b-dLRR22,28 showed full susceptibility to isolate 97011, and is most probably mediated by polymorphisms in the spacer region and/or the N-terminal LRRs of the PM3B protein. The failure of Pm3b-dLRR22,28 to mediate resistance to isolate 97011 confirms that Pm3d-dependent resistance depends on both N-terminal and C-terminal residues of the LRR domain.

Intragenic allele pyramiding of Pm3d and Pm3e leads to a functional gene with dual resistance specificities

We hypothesized that it should be possible to generate a gene with multiple recognition specificities by combining polymorphic residues of different functional alleles residing in the C-terminal LRR domain. To test if such pyramiding is possible in principle, we considered only combinations of Pm3 alleles that: (i) have identical NB-ARC sequences to avoid possible interfering effects of this domain; (ii) functionally identical N-terminal LRRs (ideally identical in sequence); and (iii) have the polymorphic amino acids in different LRRs to circumvent sterical changes that might inhibit proper folding or alter a putative interaction surface. The alleles Pm3d and Pm3e fulfilled these conditions: PM3D and PM3E differ from each other by only three amino acids in the C-terminal LRRs 22, 27 and 28 (Figure 2). We made the construct Pm3d+e that combines all polymorphic sites of PM3D and PM3E compared with PM3CS (Figure 7a). Its function was tested by transient transformation and challenged with the differential isolates 97019 (avirulent on Pm3e; virulent on Pm3d) and DB Asosan (avirulent on Pm3d; virulent on Pm3e). Results of this assay showed that the chimera PM3D+E conferred resistance to both isolates at the same level as the original PM3D and PM3E (Figure 7a). The control experiment with the virulent isolate 94202 confirmed that the resistance mediated by Pm3d+e was effector dependent.

image

Figure 7.  Intragenic allele pyramiding resulted in construct Pm3d+e, which conferred both Pm3d- and Pm3e-dependent resistance. (a) In the transient expression assays, the haustorium index of hybrid construct Pm3d+e was not significantly different (n.s.; Student’s t-test, P > 0.4) from the one of Pm3e after infection with isolates 97019 and 94202, and from the one of Pm3d after infection with isolate DB Asosan. The designations and experimental procedures are the same as described in Figure 3. Values report the mean of three independent experiments and error bars give the standard deviation. For comprehensive statistical analyses see Table S8. (b) Representative pictures of infection phenotypes of leaf segments from control lines (Bobwhite SH 98 26, Kolibri, W150) and transgenic T2Pm3d+e plants infected with powdery mildew isolates 97019, DB Asosan and 94202. For Pm3d+e, two leaf segments of each of the two independent transformation events 6 and 13 are shown.

Download figure to PowerPoint

To confirm the dual function of the Pm3d/Pm3e pyramid, we stably transformed wheat line Bobwhite SH 98 26, which does not carry an endogenous Pm3 copy, with Pm3d+e under the control of the maize ubiquitin promoter. In two independent, segregating T1 and T2 families, the presence of the transgene co-segregated with the resistance to the previously used isolates 97019 (virulent on line Kolibri carrying Pm3d; avirulent on line W150 carrying Pm3e), and to DB Asosan (avirulent on Kolibri; virulent on W150), as inferred from Southern blot analysis and leaf segment infection tests (Figure 7b). All plants tested showed susceptibility to isolate 94202, demonstrating that PM3D+E is not autoactive, but confers race-specific resistance (Figure 7b). The T2 generation was resistant when challenged with the isolates 09003 (virulent on Kolibri; avirulent on W150) and Ken 2-5 (avirulent on Kolibri; virulent on W150), thus confirming the results with the isolates 97019 and DB Asosan. The reproducibility of the transient expression assay results (Figure 7a) in stable transgenic plants demonstrates that Pm3d+e indeed represents a functional allele pyramid. Thus, based on the detailed analysis of the functional role of individual subdomains of the PM3 protein by a series of domain-swap experiments, it was possible to successfully predict a chimeric allele with pyramided resistance specificities.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Virulence analyses in pathogen populations detect natural examples of Pm3 alleles with enlarged recognition spectra

We used a large set of powdery mildew isolates to characterize the resistance spectrum of Pm3 alleles. These studies indicated that Pm3a and Pm3b represent alleles with an extended resistance spectrum compared with Pm3f and Pm3c, respectively. Studies in flax showed that the flax rust resistance genes L5, L6 and L7 have overlapping resistance specificities, which are based on the recognition of the same Avr gene, AvrL567 (Dodds et al., 2004; 2006). L6-mediated resistance is more effective (no fungal growth) than L7-dependent resistance (low level of rust sporulation), and in contrast to L7, L6 is not suppressed by a fungal inhibitor, because of the L6-specific polymorphisms in the TIR domain (Luck et al., 2000; Dodds et al., 2006). L5 shows a reduced recognition repertoire compared with L6 and L7, but the underlying molecular basis is not yet described. However, recombinant alleles showed novel resistance specificities by the loss of recognized rust strains, compared with the parental alleles. These specificities are determined by one or two amino acids in a C-terminal LRR (L6L11RV; Dodds et al., 2006), or by five or fewer amino acids in the TIR-NB domains (RL10-1, RL10-2/3, L2-L10Sph; Ellis et al., 1999; Luck et al., 2000).

The study of the Pm3 alleles with enlarged resistance spectra allowed us to identify determinants of different powdery mildew isolate recognition spectra, and to propose hypotheses on the underlying molecular mechanism. We found that the NB-ARC domains in PM3B and PM3A increase resistance activity. The PM3A-specific sequence in C-terminal LRRs was shown to contribute to the higher resistance of PM3A compared with PM3F, possibly by contributing to a higher binding affinity to AvrPM3A. On the pathogen side, the quantitative difference in resistance intensity within the Pm3a/Pm3f and Pm3b/Pm3c pairs might be caused by slightly different biochemical properties of the different AVR proteins, resulting in differential binding affinities. The cloning of the AvrPm3 genes will offer the prospects of testing these hypotheses experimentally.

The ARC2 domain modulates PM3 activity

The results of constructs with reciprocally swapped NB-ARC domains of PM3A and PM3B point to a role of the NB-ARC domain in controlling the efficiency of effector-dependent resistance. PM3A and PM3B share all polymorphic residues in the ARC2 domain. As both NB-ARC domains cause a comparable HI reduction (Pm3a compared with Pm3a-fARC, and Pm3f-aARC and Pm3a-fLRR19-22 compared with Pm3f, Figure 3; Pm3b compared with Pm3b-dNB-ARC, and Pm3d-bNB-ARC compared with Pm3d; Figure 4a), it is likely that the polymorphisms in the ARC2 domain specifically cause this alteration. The ARC2 domain of R proteins was described as the regulatory element that transduces pathogen perception by the LRR domain into R-protein activation (Tameling et al., 2006). It was suggested that the LRR domain binds directly to the ARC1 domain, with an alteration of this interaction upon effector binding that would be transmitted by the ARC2 domain (Rairdan and Moffett, 2006; van Ooijen et al., 2008a). A 3D structural model shows that the PM3A- and PM3B-specific amino acids locate mainly on one side of the ARC2 subdomain (Figure 5b). Together with the predominant position of the variant residues within long loops and on the surface, our results suggest an important role of that region for interdomain interactions, facilitating changes in the LRR-ARC1 interaction. Alternatively, these polymorphic residues could also increase the affinity with other factors required for downstream signalling or stabilize a certain protein conformation required for signalling activity or for pathogen perception, leading to a signal more intense in time and/or amplitude.

The role of the N- and C-terminal LRRs in PM3 race specificity

Pathogen recognition specificity of PM3 proteins is determined by their LRR domain. This was also found in other LRR-containing plant R proteins (e.g. Ellis et al., 1999; 2007; Dodds et al., 2001; Wulff et al., 2001; 2009; Shen et al., 2003; Rairdan and Moffett, 2006; Zhou et al., 2006). Additional sequences outside the LRR domain were also reported to be involved in recognition specificity (Luck et al., 2000). Studies on some mammalian NACHT-LRR proteins (NLPs) have revealed two different roles of the LRR domain: the N-terminal LRRs modulate activation, whereas C-terminal LRRs are responsible for bacterial recognition (Inohara and Nunez, 2003; Tanabe et al., 2004). Possibly, these different subdomain functions are conserved in plant NB-ARC-LRR proteins (Belkhadir et al., 2004; Lukasik and Takken, 2009). Domain swaps of flax L6/L11, barley (Hordeum vulgare) Mla1/Mla6 and potato Rx/Gpa2, as well as mutational analysis of Rx, have shown that recognition specificity is determined by C-terminal LRRs alone (L6, Mla6), or may also involve N-terminal LRRs (Shen et al., 2003; Farnham and Baulcombe, 2006; Rairdan and Moffett, 2006; Ellis et al., 2007). In PM3, there is evidence for a major role of the C-terminal LRRs in recognition specificity. Sequence analysis and our previous experiments (Yahiaoui et al., 2006) revealed that Pm3e and Pm3g specificity is determined exclusively by putative solvent-exposed residues of the LxxLxLxx motif in the C-terminal LRRs. Our domain-swap experiments showed that polymorphic residues in the LRRs 19–22 are required for Pm3a-specific resistance (Figure 3). However, PM3B, PM3C (this study) and PM3D (Yahiaoui et al., 2006) also depend functionally on the N-terminal LRR polymorphisms. In addition, polymorphic residues in the N-terminal LRRs of PM3B and PM3C confer residual, race-specific, resistance responses. In these two proteins, both N- and C-terminal LRRs might contribute to the recognition-mediating molecular interactions. Finally, it should be noted that a functional combination of N- and C-terminal LRRs also needs a fitting ARC2 domain for optimal function: PM3B-DNB-ARC and PM3D-BNB-ARC are still functioning race-specifically, but they lose full resistance activity. This is reminiscent of findings from the tobacco N-like proteins (Gao et al., 2007).

The majority of the LRR domains (mainly from bacterial, animal or human proteins) that have been co-crystallized with their ligands show binding in the concave LRR face (Bella et al., 2008). In the PM3 proteins studied, polymorphic residues in the C-terminal LRRs occur exclusively in the non-conserved x-positions (marked in Figure 2a) of the putative LxxLxLxx motif on the predicted concave surface of the LRR domain. The only exception are polymorphic amino acids in the sequence block of LRRs 19–22 in PM3A, which is thought to be of ancient origin and derived from gene conversion (Yahiaoui et al., 2006). We consider it likely that these polymorphic x-positions in C-terminal LRRs interact directly with corresponding AVR proteins. This hypothesis is supported by the high level of diversifying selection in the PM3 proteins (Yahiaoui et al., 2006), which is assumed to be characteristic for direct protein–protein interactions (Wang et al., 2007). In the case of PM3B and PM3C, it is possible that polymorphic residues in the N-terminal LRRs are also involved in AVR binding, similar to the structural model for the flax AvrL567–L5 interaction, where the N- and C-terminal LRRs of L5 contribute to the AVR binding (Wang et al., 2007).

Intragenic pyramiding of Pm3 allelic resistance specificities: towards artificial evolution of broad spectrum resistance

The LRR-mediated ligand binding appears to involve single residues on the concave LRR surface that function in a cooperative manner (Bella et al., 2008; Herrin et al., 2008; Velikovsky et al., 2009). In flax L5 and L6, single amino acids were shown to have quantitative and qualitative effects on the R protein–Avr protein interaction (Wang et al., 2007). Based on these results, a model of stepwise evolution of R and Avr genes was proposed. It states that a new R gene might evolve from a gene by an initial mutation that confers weak resistance. Subsequently, the accumulation of further mutations would lead to a strong resistance response. Our data suggest that such a stepwise evolution led to the extended, broad-spectrum resistance of Pm3a and Pm3b compared with Pm3f and Pm3c, respectively.

The flax L5/L6/L7 alleles and the Pm3a/Pm3f and Pm3b/Pm3c alleles represent natural examples for broad- and narrow-spectrum R genes. In Arabidopsis thaliana, RPP1-WsB detects four alleles of the Hyaloperonospora parasitica avirulence gene ATR1, whereas its paralogue RPP1-Nd recognizes only one of them (Rehmany et al., 2005). Recombinant L alleles of flax were reported to confer narrower recognition spectra compared with the parental alleles (Ellis et al., 1999; 2007; Luck et al., 2000), and recombination between homologues of Cf4/9 (tomato) and between paralogues of Rp1 (maize) created novel resistance specificities (Parniske et al., 1997; Smith and Hulbert, 2005). A first example for an artificial gene with broadened resistance spectrum comes from the mutational analysis of Rx, where in three cases a single amino acid change led to a broader resistance against potato virus X, and against a distantly related poplar mosaic virus (Farnham and Baulcombe, 2006). The pyramiding of Pm3d and Pm3e now represents an example for a designed R gene with a broader recognition spectrum. Thus, the molecular analysis of R proteins with overlapping spectra of specificity might lead to an improved understanding of the evolution of broad-spectrum resistance. The ultimate applied goal of such work would be the possibility to rationally design broad-spectrum R genes.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Fungal strains

Our wheat powdery mildew (B. graminis f.sp. tritici) isolates originate from the former mildew collections of Agroscope Reckenholz-Tänikon ART (http://www.art.admin.ch) and INRA Rennes (http://www.rennes.inra.fr), from USDA-ARS, North Carolina State University, Raleigh (http://www.ars.usda.gov/saa/psru) and from our own isolate collection. The (a)virulences relevant for transient expression assays and seedling infection experiments are listed in Table 1.

Table 1.   Powdery mildew isolates used for seedling infection experiments and transient expression assays
IsolateAvirulenceaVirulenceb
  1. aFor each isolate, only the avirulence (Avr) genes relevant for this work are listed.

  2. bFor each isolate, only the absent avirulence genes (avr) relevant for our work are listed.

07016avrPm3b, avrPm3c, avrPm3d
07201avrPm3a, avrPm3f
07296avrPm3b, avrPm3c
09003AvrPm3eavrPm3d
94202avrPm3d, avrPm3e
96224AvrPm3a–AvrPm3g
96229AvrPm3bavrPm3d
97011AvrPm3davrPm3b
97019AvrPm3b AvrPm3c, AvrPm3eavrPm3d
97028AvrPm3aavrPm3f
DB AsosanAvrPm3davrPm3e
Ken 2-5AvrPm3davrPm3e

Wheat powdery mildew virulence profiling

Powdery mildew isolates were collected in the seasons 1992–1998 and in 2007 in the Swiss plateau with a spore trap, and single colony-derived isolates were propagated and stored on leaf segments, as previously described (Winzeler et al., 1991). Virulence for Pm3a–Pm3g was tested on a differential set of wheat lines and varieties using Asosan/8*Chancellor for Pm3a, Chul/8*Chancellor for Pm3b, Sonora/8*Chancellor for Pm3c, Kolibri for Pm3d, W150 for Pm3e (only tested in 2007), Michigan Amber/8*Chancellor for Pm3f (tested as of 1994) and Aristide for Pm3g (only in 2007). Wheat lines and varieties were grown and infected with the test isolates as described in Limpert et al. (1987). Three leaf segments on different plates were tested per line and isolate.

Construction of recombinant genes

The Pm3aPm3g and Pm3CS alleles have been cloned previously into the expression vector PGY1 (35S promoter and terminator; Yahiaoui et al., 2004; Srichumpa et al., 2005; Yahiaoui et al., 2006). All chimeric genes were constructed directly in PGY1 by using the unique restriction sites AflII, BclI or NsiI, indicated in Figure S2, and the flanking sites BamHI (5′ end) or SalI (3′ end), or by site-directed mutagenesis (construct Pm3d+e), as described in Appendix S1. All constructs were checked by restriction enzyme digests and DNA sequencing of allele-specific regions and junction sites.

Transient expression assay

Seven-day-old primary leaves of the susceptible wheat variety Chancellor were used for transient transformation. Particle bombardment was performed with the Biolistic PDS-1000/He System with the Hepta Adapter (Bio-Rad, http://www.bio-rad.com), following an adapted protocol of Douchkov et al. (2005). Per shot, 3 mg of gold particles (1-μm diameter; Bio-Rad) were coated with a mixture of 1.25 μg of pUbiGUS reporter plasmid (Schweizer et al., 1999) and 1.25 μg of the test plasmids (Pm3 wild-type or recombinant genes in PGY1). Four hours after bombardment, leaves were infected with powdery mildew at high density and kept for 44 h on plates with slightly open lids at 20°C, 16 h of light and 80% relative humidity. GUS staining (Schweizer et al., 1999), staining of the fungus with Coomassie blue (Schweizer et al., 1993) and microscopic evaluation (Yahiaoui et al., 2006) were performed as described previously.

In silico analysis

Sequence alignments were computed using muscle (http://www.drive5.com/muscle; Edgar, 2004). Shading of physicochemically conserved residues was produced by GeneDoc (http://www.psc.edu/biomed/genedoc). The secondary structure of PM3CS was predicted by PSIPRED (http://bioinf.cs.ucl.ac.uk/psipred). The secondary structure assignment of the PDB structure of APAF-1 (PDB identifier 1z6t, chain A) was obtained from the DSSP database (http://www.cmbi.kun.nl/gv/dssp). Structure-based protein sequence alignment was constructed including the PM3CS sequence as described in van Ooijen et al. (2008b). A protein structure model of PM3CS was obtained by submitting the pairwise alignment of PM3CS and APAF-1 to the HOMER-M web server (http://protein.bio.unipd.it/homer). Alignment gaps were substituted by loop modelling of PM3CS sequence positions 453–462 (GFILEYKEDS) and 486–491 (SKDYSG), as well as including positions 357, 358, 389 and 390 into the structure using the ModLoop web service (http://modbase.compbio.ucsf.edu/modloop; Fiser and Sali, 2003). The protein structure image of the model including the positions of polymorphic sites was illustrated using pymol (http://www.pymol.org).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank Gabriele Büsing, Gerhard Herren and Noel Immenhauser for stable wheat transformation and characterization of the transgenic lines, and Francis Parlange for help with powdery mildew virulence profiling. Röbi Dudler is acknowledged for helpful comments on the manuscript. This project was financially supported by grants from the Swiss National Science Foundation (31003A-127061/1) and by the German National Genome Research Network (NGFN).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Albrecht, M. and Takken, F.L.W. (2006) Update on the domain architectures of NLRs and R proteins. Biochem. Biophys. Res. Commun. 339, 459462.
  • Belkhadir, Y., Subramaniam, R. and Dangl, J.L. (2004) Plant disease resistance protein signaling: NBS-LRR proteins and their partners. Curr. Opin. Plant Biol. 7, 391399.
  • Bella, J., Hindle, K.L., McEwan, P.A. and Lovell, S.C. (2008) The leucine-rich repeat structure. Cell. Mol. Life Sci. 65, 23072333.
  • Bhullar, N.K., Street, K., Mackay, M., Yahiaoui, N. and Keller, B. (2009) Unlocking wheat genetic resources for the molecular identification of previously undescribed functional alleles at the Pm3 resistance locus. Proc. Natl Acad. Sci. USA, 106, 95199524.
  • Bhullar, N.K., Zhang, Z.Q., Wicker, T. and Keller, B. (2010) Wheat gene bank accessions as a source of new alleles of the powdery mildew resistance gene Pm3: a large scale allele mining project. BMC Plant Biol. 10, 88.
  • Van Der Biezen, E.A. and Jones, J.D.G. (1998) The NB-ARC domain: a novel signalling motif shared by plant resistance gene products and regulators of cell death in animals. Curr. Biol. 8, R226R227.
  • Briggle, L.W. (1969) Near-isogenic lines of wheat with genes for resistance to Erysiphe graminis f. sp. tritici. Crop Sci. 9, 7072.
  • Chisholm, S.T., Coaker, G., Day, B. and Staskawicz, B.J. (2006) Host-microbe interactions: shaping the evolution of the plant immune response. Cell, 124, 803814.
  • Dangl, J.L. and Jones, J.D.G. (2001) Plant pathogens and integrated defence responses to infection. Nature, 411, 826833.
  • Danot, O., Marquenet, E., Vidal-Ingigliardi, D. and Richet, E. (2009) Wheel of life, wheel of death: a mechanistic insight into signaling by STAND proteins. Structure, 17, 172182.
  • DeYoung, B.J. and Innes, R.W. (2006) Plant NBS-LRR proteins in pathogen sensing and host defense. Nat. Immunol. 7, 12431249.
  • Dodds, P.N. and Rathjen, J.P. (2010) Plant immunity: towards an integrated view of plant-pathogen interactions. Nat. Rev. Genet. 11, 539548.
  • Dodds, P.N., Lawrence, G.J. and Ellis, J.G. (2001) Six amino acid changes confined to the leucine-rich repeat beta-strand/beta-turn motif determine the difference between the P and P2 rust resistance specificities in flax. Plant Cell, 13, 163178.
  • Dodds, P.N., Lawrence, G.J., Catanzariti, A.M., Ayliffe, M.A. and Ellis, J.G. (2004) The Melampsora lini AvrL567 avirulence genes are expressed in haustoria and their products are recognized inside plant cells. Plant Cell, 16, 755768.
  • Dodds, P.N., Lawrence, G.J., Catanzariti, A.M., Teh, T., Wang, C.I.A., Ayliffe, M.A., Kobe, B. and Ellis, J.G. (2006) Direct protein interaction underlies gene-for-gene specificity and coevolution of the flax resistance genes and flax rust avirulence genes. Proc. Natl Acad. Sci. USA, 103, 88888893.
  • Douchkov, D., Nowara, D., Zierold, U. and Schweizer, P. (2005) A high-throughput gene-silencing system for the functional assessment of defense-related genes in barley epidermal cells. Mol. Plant Microbe Interact. 18, 755761.
  • Dunning, F.M., Sun, W., Jansen, K.L., Helft, L. and Bent, A.F. (2007) Identification and mutational analysis of Arabidopsis FLS2 leucine-rich repeat domain residues that contribute to flagellin perception. Plant Cell, 19, 32973313.
  • Edgar, R.C. (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 32, 17921797.
  • Ellis, J.G., Lawrence, G.J., Luck, J.E. and Dodds, P.N. (1999) Identification of regions in alleles of the flax rust resistance gene L that determine differences in gene-for-gene specificity. Plant Cell, 11, 495506.
  • Ellis, J., Dodds, P. and Pryor, T. (2000) The generation of plant disease resistance gene specificities. Trends Plant Sci. 5, 373379.
  • Ellis, J.G., Lawrence, G.J. and Dodds, P.N. (2007) Further analysis of gene-for-gene disease resistance specificity in flax. Mol. Plant Pathol. 8, 103109.
  • Enkhbayar, P., Kamiya, M., Osaki, M., Matsumoto, T. and Matsushima, N. (2004) Structural principles of leucine-rich repeat (LRR) proteins. Proteins, 54, 394403.
  • Farnham, G. and Baulcombe, D.C. (2006) Artificial evolution extends the spectrum of viruses that are targeted by a disease-resistance gene from potato. Proc. Natl Acad. Sci. USA, 103, 1882818833.
  • Fiser, A. and Sali, A. (2003) ModLoop: automated modeling of loops in protein structures. Bioinformatics, 19, 25002501.
  • Flor, H.H. (1971) Current status of the gene-for-gene concept. Annu. Rev. Phytopathol. 9, 275296.
  • Gao, J.S., Sasaki, N., Kanegae, H., Konagaya, K.I., Takizawa, K., Hayashi, N., Okano, Y., Kasahara, M., Matsushita, Y. and Nyunoya, H. (2007) The TIR-NBS but not LRR domains of two novel N-like proteins are functionally competent to induce the elicitor p50-dependent hypersensitive response. Physiol. Mol. Plant Pathol. 71, 7887.
  • Herrin, B.R., Alder, M.N., Roux, K.H., Sina, C., Ehrhardt, G.R.A., Boydston, J.A., Turnbough, C.L. and Cooper, M.D. (2008) Structure and specificity of lamprey monoclonal antibodies. Proc. Natl Acad. Sci. USA, 105, 20402045.
  • Huang, X.Q. and Röder, M.S. (2004) Molecular mapping of powdery mildew resistance genes in wheat: a review. Euphytica, 137, 203223.
  • Huang, X.Q., Hsam, S.L.K., Mohler, V., Röder, M.S. and Zeller, F.J. (2004) Genetic mapping of three alleles at the Pm3 locus conferring powdery mildew resistance in common wheat (Triticum aestivum L.). Genome, 47, 11301136.
  • Inohara, N. and Nunez, G. (2003) NODs: intracellular proteins involved in inflammation and apoptosis. Nat. Rev. Immunol. 3, 371382.
  • Jones, J.D.G. and Dangl, J.L. (2006) The plant immune system. Nature, 444, 323329.
  • Jones, D.A. and Takemoto, D. (2004) Plant innate immunity – direct and indirect recognition of general and specific pathogen-associated molecules. Curr. Opin. Immunol. 16, 4862.
  • Kobe, B. and Kajava, A.V. (2001) The leucine-rich repeat as a protein recognition motif. Curr. Opin. Struct. Biol. 11, 725732.
  • Limpert, E., Felsenstein, F.G. and Andrivon, D. (1987) Analysis of virulence in populations of wheat powdery mildew in Europe. J. Phytopathol. 120, 18.
  • Luck, J.E., Lawrence, G.J., Dodds, P.N., Shepherd, K.W. and Ellis, J.G. (2000) Regions outside of the leucine-rich repeats of flax rust resistance proteins play a role in specificity determination. Plant Cell, 12, 13671377.
  • Lukasik, E. and Takken, F.L.W. (2009) STANDing strong, resistance proteins instigators of plant defence. Curr. Opin. Plant Biol. 12, 427436.
  • McDowell, J.M. and Simon, S.A. (2006) Recent insights into R gene evolution. Mol. Plant Pathol. 7, 437448.
  • Mondragon-Palomino, M., Meyers, B.C., Michelmore, R.W. and Gaut, B.S. (2002) Patterns of positive selection in the complete NBS-LRR gene familiy of Arabidopsis thaliana. Genome Res. 12, 13051315.
  • Van Ooijen, G., Mayr, G., Albrecht, M., Cornelissen, B.J.C. and Takken, F.L.W. (2008a) Transcomplementation, but not physical association of the CC-NB-ARC and LRR domains of tomato R protein Mi-1.2 is altered by mutations in the ARC2 subdomain. Mol. Plant, 1, 401410.
  • Van Ooijen, G., Mayr, G., Kasiem, M.M.A., Albrecht, M., Cornelissen, B.J.C. and Takken, F.L.W. (2008b) Structure-function analysis of the NB-ARC domain of plant disease resistance proteins. J. Exp. Bot. 59, 13831397.
  • Parniske, M., Hammond-Kosack, K.E., Golstein, C., Thomas, C.M., Jones, D.A., Harrison, K., Wulff, B.B.H. and Jones, J.D.G. (1997) Novel disease resistance specificities result from sequence exchange between tandemly repeated genes at the Cf-4/9 locus of tomato. Cell, 91, 821832.
  • Rairdan, G.J. and Moffett, P. (2006) Distinct domains in the ARC region of the potato resistance protein Rx mediate LRR binding and inhibition of activation. Plant Cell, 18, 20822093.
  • Rairdan, G. and Moffett, P. (2007) Brothers in arms? Common and contrasting themes in pathogen perception by plant NB-LRR and animal NACHT-LRR proteins. Microbes Infect. 9, 677686.
  • Rehmany, A.P., Gordon, A., Rose, L.E., Allen, R.L., Armstrong, M.R., Whisson, S.C., Kamoun, S., Tyler, B.M., Birch, P.R.J. and Beynon, J.L. (2005) Differential recognition of highly divergent downy mildew avirulence gene alleles by RPP1 resistance genes from two Arabidopsis lines. Plant Cell, 17, 18391850.
  • Riedl, S.J., Li, W.Y., Chao, Y., Schwarzenbacher, R. and Shi, Y.G. (2005) Structure of the apoptotic protease-activating factor 1 bound to ADP. Nature, 434, 926933.
  • Schweizer, P., Gees, R. and Mosinger, E. (1993) Effect of jasmonic acid on the interaction of barley (Hordeum vulgare L.) with the powdery mildew Erysiphe graminis f.sp. hordei. Plant Physiol. 102, 503511.
  • Schweizer, P., Pokorny, J., Abderhalden, O. and Dudler, R. (1999) A transient assay system for the functional assessment of defense-related genes in wheat. Mol. Plant Microbe Interact. 12, 647654.
  • Schwessinger, B. and Zipfel, C. (2008) News from the frontline: recent insights into PAMP-triggered immunity in plants. Curr. Opin. Plant Biol. 11, 389395.
  • Seeholzer, S., Tsuchimatsu, T., Jordan, T., Bieri, S., Pajonk, S., Yang, W.X., Jahoor, A., Shimizu, K.K., Keller, B. and Schulze-Lefert, P. (2010) Diversity at the Mla powdery mildew resistance locus from cultivated barley reveals sites of positive selection. Mol. Plant Microbe Interact. 23, 497509.
  • Shen, Q.H., Zhou, F.S., Bieri, S., Haizel, T., Shirasu, K. and Schulze-Lefert, P. (2003) Recognition specificity and RAR1/SGT1 dependence in barley Mla disease resistance genes to the powdery mildew fungus. Plant Cell, 15, 732744.
  • Smith, S.M. and Hulbert, S.H. (2005) Recombination events generating a novel Rp1 race specificity. Mol. Plant Microbe Interact. 18, 220228.
  • Srichumpa, P., Brunner, S., Keller, B. and Yahiaoui, N. (2005) Allelic series of four powdery mildew resistance genes at the Pm3 locus in hexaploid bread wheat. Plant Physiol. 139, 885895.
  • Takken, F.L.W., Albrecht, M. and Tameling, W.I.L. (2006) Resistance proteins: molecular switches of plant defence. Curr. Opin. Plant Biol. 9, 383390.
  • Tameling, W.I.L., Elzinga, S.D.J., Darmin, P.S., Vossen, J.H., Takken, F.L.W., Haring, M.A. and Cornelissen, B.J.C. (2002) The tomato R gene products I-2 and Mi-1 are functional ATP binding proteins with ATPase activity. Plant Cell, 14, 29292939.
  • Tameling, W.I.L., Vossen, J.H., Albrecht, M., Lengauer, T., Berden, J.A., Haring, M.A., Cornelissen, B.J.C. and Takken, F.L.W. (2006) Mutations in the NB-ARC domain of I-2 that impair ATP hydrolysis cause autoactivation. Plant Physiol. 140, 12331245.
  • Tanabe, T., Chamaillard, M., Ogura, Y. et al. (2004) Regulatory regions and critical residues of NOD2 involved in muramyl dipeptide recognition. EMBO J. 23, 15871597.
  • Ueda, H., Yamaguchi, Y. and Sano, H. (2006) Direct interaction between the tobacco mosaic virus helicase domain and the ATP-bound resistance protein, N factor during the hypersensitive response in tobacco plants. Plant Mol. Biol. 61, 3145.
  • Velikovsky, C.A., Deng, L., Tasumi, S., Iyer, L.M., Kerzic, M.C., Aravind, L., Pancer, Z. and Mariuzza, R.A. (2009) Structure of a lamprey variable lymphocyte receptor in complex with a protein antigen. Nat. Struct. Mol. Biol. 16, 725730.
  • Wang, C.I.A., Guncar, G., Forwood, J.K. et al. (2007) Crystal structures of flax rust avirulence proteins AvrL567-A and -D reveal details of the structural basis for flax disease resistance specificity. Plant Cell, 19, 28982912.
  • Wei, T., Gong, J., Jamitzky, F., Heckl, W.M., Stark, R.W. and Rössle, S.C. (2008) LRRML: a conformational database and an XML description of leucine-rich repeats (LRRs). BMC Struct. Biol. 8, 47.
  • Winzeler, M., Streckeisen, P. and Fried, P.M. (1991) Virulence analysis of the wheat powdery mildew population in Switzerland between 1980 and 1989. In Integrated Control Of Cereal Mildews: Virulence Patterns And Their Change (Jørgensen, J.H. ed.). Røskilde, Denmark: Risø National Laboratory, pp. 1521.
  • Wulff, B.B.H., Thomas, C.M., Smoker, M., Grant, M. and Jones, J.D.G. (2001) Domain swapping and gene shuffling identify sequences required for induction of an Avr-dependent hypersensitive response by the tomato Cf-4 and Cf-9 proteins. Plant Cell, 13, 255272.
  • Wulff, B.B.H., Heese, A., Tomlinson-Buhot, L., Jones, D.A., De La Pena, M. and Jones, J.D.G. (2009) The major specificity-determining amino acids of the tomato Cf-9 disease resistance protein are at hypervariable solvent-exposed positions in the central leucine-rich repeats. Mol. Plant Microbe Interact. 22, 12031213.
  • Yahiaoui, N., Srichumpa, P., Dudler, R. and Keller, B. (2004) Genome analysis at different ploidy levels allows cloning of the powdery mildew resistance gene Pm3b from hexaploid wheat. Plant J. 37, 528538.
  • Yahiaoui, N., Brunner, S. and Keller, B. (2006) Rapid generation of new powdery mildew resistance genes after wheat domestication. Plant J. 47, 8598.
  • Yahiaoui, N., Kaur, N. and Keller, B. (2009) Independent evolution of functional Pm3 resistance genes in wild tetraploid wheat and domesticated bread wheat. Plant J. 57, 846856.
  • Zhou, B., Qu, S.H., Liu, G.F., Dolan, M., Sakai, H., Lu, G.D., Bellizzi, M. and Wang, G.L. (2006) The eight amino-acid differences within three leucine-rich repeats between Pi2 and Piz-t resistance proteins determine the resistance specificity to Magnaporthe grisea. Mol. Plant Microbe Interact. 19, 12161228.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Figure S1. Structure-based multiple sequence alignment of the NB-ARC domain of PM3CS with APAF-1, CED-4 and ten plant R proteins. The PM3CS sequence was added to the structure-based multiple sequence alignment published in van Ooijen et al. (2008), including gain- and loss-of-function positions in pink and yellow, respectively. Green stars mark amino-acid positions that are conserved within R proteins and APAF-1 at the ADP binding site. Amino acids that are polymorphic in the corresponding PM3B sequence are coloured in orange; those that are polymorphic in both PM3A and PM3B are in red. The secondary-structure prediction reveals that PM3CS carries the vast majority of the α-helices (blue) and β-stands (brown) present in APAF-1. Furthermore, the alignment indicates that the NB-ARC domain encoded by Pm3CS contains all motifs of the three subdomains NB, ARC1 and ARC2. Only the first motif, hhGRExE, seems to be absent in PM3CS.

Figure S2. Amino acid differences of the PM3A PM3G proteins. Polymorphic amino acids are indicated below the PM3CS sequence, which is identical to the consensus sequence of the seven functional PM3 proteins. The protein domains are indicated on the left. The predicted CC structure is underlined. Motifs conserved in the NB-ARC domains of NBS-LRR proteins are underlined and labelled on top. The x-positions of the LxxLxLxx motif are highlighted in red, and the conserved leucines or other hydrophobic residues are represented in bold. Residues of the ARC2 subdomain that form loops in the 3D model (Figure 5) are coloured in green, and side chains of polymorphic residues that point to the solvent are highlighted in light blue. The notional positions of the AflII, BclI and NsiI restriction sites in the corresponding DNA sequence used for constructing chimeric genes are indicated.

Table S1. Numbers of powdery mildew isolates with the different (a)virulence gene combinations for Pm3a/Pm3f (a) and Pm3b/Pm3c (b).

Table S2.P values of Student’s t-test on HI of constructs shown in Figure 3.

Table S3.P values of Student’s t-test on HI of constructs shown in Figure 4a.

Table S4.P values of Student’s t-test on HI of constructs shown in Figures 4a and 6a (virulent controls).

Table S5.P values of Student’s t-test on HI of constructs shown in Figure 4b.

Table S6.P values of Student’s t-test on HI of constructs shown in Figure 6a.

Table S7.P values of Student’s t-test on HI of constructs shown in Figure 6b.

Table S8.P values of Student’s t-test on HI of constructs shown in Figure 7.

Appendix S1. Supplementary text for Experimental Procedures.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
TPJ_4342_sm_Legends.doc44KSupporting info item
TPJ_4342_sm_AppendixS1.doc64KSupporting info item
TPJ_4342_sm_FiguresS1-S2.pdf2134KSupporting info item
TPJ_4342_sm_TablesS1-S8.pdf95KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.