Research ArticleResearch Article

Receptor-like Genes in the Major Resistance Locus of Lettuce Are Subject to Divergent Selection

Blake C. Meyers, Katherine A. Shen, Pejman Rohani, Brandon S. Gaut, Richard W. Michelmore

Published November 1998. DOI: https://doi.org/10.1105/tpc.10.11.1833

Abstract

Disease resistance genes in plants are often found in complex multigene families. The largest known cluster of disease resistance specificities in lettuce contains the RGC2 family of genes. We compared the sequences of nine full-length genomic copies of RGC2 representing the diversity in the cluster to determine the structure of genes within this family and to examine the evolution of its members. The transcribed regions range from at least 7.0 to 13.1 kb, and the cDNAs contain deduced open reading frames of ~5.5 kb. The predicted RGC2 proteins contain a nucleotide binding site and irregular leucine-rich repeats (LRRs) that are characteristic of resistance genes cloned from other species. Unique features of the RGC2 gene products include a bipartite LRR region with >40 repeats. At least eight members of this family are transcribed. The level of sequence diversity between family members varied in different regions of the gene. The ratio of nonsynonymous (Ka) to synonymous (Ks) nucleotide substitutions was lowest in the region encoding the nucleotide binding site, which is the presumed effector domain of the protein. The LRR-encoding region showed an alternating pattern of conservation and hypervariability. This alternating pattern of variation was also found in all comparisons within families of resistance genes cloned from other species. The Ka/Ks ratios indicate that diversifying selection has resulted in increased variation at these codons. The patterns of variation support the predicted structure of LRR regions with solvent-exposed hypervariable residues that are potentially involved in binding pathogen-derived ligands.

INTRODUCTION

Plant disease resistance is often inherited as single Mendelian resistance genes that determine the reaction to specific pathogen avirulence genes. These genes fall into several mechanistic and structural classes (Michelmore, 1995; Baker et al., 1997). Genes encoding similar amino acid motifs are found in diverse plant species and are effective against a wide range of pathogens, including viruses, bacteria, nematodes, and fungi. The most common class of cloned genes encodes proteins containing a nucleotide binding site (NBS) and leucine-rich repeats (LRRs; Staskawicz et al., 1995; Bent, 1996). These domains are often components of signal transduction proteins (Kobe and Deisenhofer, 1994; Traut, 1994), which supports the hypothesis that these genes encode receptors and may act early in a signal transduction pathway (reviewed in Baker et al., 1997).

Genetic studies have determined that resistance genes are often members of complex loci comprised of linked resistance specificities (Pryor, 1987; Crute and Pink, 1996). Molecular data from at least 10 families of resistance genes, including loci from tomato, lettuce, rice, flax, and Arabidopsis, indicate that these loci frequently contain arrays of related genes. Sequencing of the tomato Cf-4 and Cf-9 haplotypes, which confer resistance to Cladosporium fulvum, demonstrated the presence of five closely related members in each genotype spanning ~35 kb (Parniske et al., 1997). The Xa21 gene from rice, which confers resistance to the bacterial pathogen Xanthomonas oryzae pv oryzae, belongs to a multigene family containing at least eight members distributed over ~230 kb (Williams et al., 1996). DNA gel blot analysis indicates that the M locus may contain ⩾15 members of a multigene family contained within <1 Mb (Anderson et al., 1997). In Arabidopsis, eight RPP5 homologs are clustered over 90 kb (Bevan et al., 1998).

Plants are challenged by rapidly evolving pathogen populations and must be able to evolve new resistance specificities to detect virulent variants. However, little is known about the mechanisms that have influenced the evolution of both individual plant resistance genes and the multigene families that contain such genes. High levels of meiotic instability have been detected in some resistance gene clusters, particularly in the Rp1 complex of maize (Sudupak et al., 1993; Hulbert, 1997). At the Cf-4/9 locus of tomato, pairing between dissimilar haplotypes may increase variation by stimulating unequal intragenic recombination (Parniske et al., 1997). Intragenic recombination has probably resulted in variation in the LRR-encoding region of the L6 and M genes in flax (Ellis et al., 1995; Anderson et al., 1997). The multigene nature of resistance loci may facilitate meiotic instability in a heterozygous state. Published models for the generation of novel resistance gene specificities propose recombination, gene conversion, and unequal crossing over as the primary mechanisms in generating haplotype diversity (Shepherd and Mayo, 1972; Pryor, 1987; Richter et al., 1995; Hammond-Kosack and Jones, 1997).

The major cluster of resistance genes (the Dm3 locus) of lettuce is the most complex and largest family of plant resistance genes characterized to date. Genetic analysis of different lettuce genotypes has demonstrated >10 resistance specificities at this locus, most of which are Dm genes, encoding resistance to lettuce downy mildew (Bremia lactucae; Farrara et al., 1987; T. Nakahara and R.W. Michelmore, unpublished data). The Dm3 haplotype in cultivar Diana contains at least 24 diverse resistance gene candidate (RGC) sequences distributed over ~3.5 Mb (Meyers et al., 1998). Genomic bacterial artificial chromosome (BAC) clones containing 22 members of the RGC2 gene family (RGC2A to RGC2W) have been identified and mapped in the region encompassing Dm3 (Meyers et al., 1998). Limited genomic sequencing of two RGC2 sequences detected the presence of both NBS and LRR motifs (Shen et al., 1998). The family exhibits a high level of sequence divergence between members in the NBS region (Meyers et al., 1998). Deletion mutant mapping data and the molecular analysis of two additional mutants have identified one member of the RGC2 family as Dm3 (Okubara et al., 1997; Meyers et al., 1998; D.B. Chin, R. Arroyo-Garcia, B.C. Meyers, K.A. Shen, and R.W. Michelmore, unpublished data).

In this study, we analyzed the complete sequences of nine of the 24 genes from the Dm3 cluster, including all members of the subfamily most closely related to Dm3 and several of the more divergent members of the family. This analysis demonstrated that the genes clustered at the Dm3 locus are among the largest thus far reported for plants. They have multiple introns, one of which varies greatly in size. The LRR-encoding region seems to be bipartite and contains a polymorphic, compound trinucleotide simple sequence repeat in the open reading frame. Most of the genes studied were transcribed and contained intact open reading frames. Regions of hypervariability were identified in regions encoding amino acids in the LRR that may comprise a solvent-exposed surface. Sequence diversity within these regions may affect ligand binding and therefore contribute to the evolution of novel specificities.

RESULTS

Choice of RGC2 Copies for Sequencing

A total of nine RGC2 family members were selected for sequencing; these included the candidate Dm3 gene, members of a subfamily closely related to Dm3, and additional members to sample diversity in the family. Both mutant analysis and mapping data indicate that copy RGC2B is Dm3 (Okubara et al., 1997; Meyers et al., 1998). Genetic complementation with RGC2B is currently under way. Three RGC2 copies, RGC2C, RGC2D, and RGC2S, were selected because they comprise the subfamily most closely related to RGC2B. This subfamily shares a set of markers that reside in intron 3 and are lacking in other family members, including the low-copy markers AM14 (Anderson et al., 1996), IPCR800, and the microsatellite MSAT15-34 (Okubara et al., 1997; Meyers et al., 1998). In addition, we sequenced five divergent RGC2 copies to determine the degree and nature of evolutionary changes that have occurred within the RGC2 family. Prior sequence analysis had revealed that the regions encoding the NBS of RGC2A, RGC2N, RGC2J, RGC2K, and RGC2O were only 61 to 74% identical to each other and RGC2B. Mapping and sequence analysis indicate that these sequences include the range of sequence diversity and physical positions observed within the family (Meyers et al., 1998).

Analysis and Structure of the RGC2 Genes

The size and genomic structure of RGC2 genes were determined by analysis of genomic and cDNA sequences. Rapid amplification of cDNA ends (RACE; Frohman et al., 1988) products was obtained using primers in exon 2, 5′ of the region encoding the NBS in RGC2A and RGC2B (Table 1). The initiation of the cDNA occurred ~700 bp upstream of the first coding exon (Figure 1). However, there is an intron of 59 and 85 bp in the 5′ untranslated region (5′ UTR) in the mature mRNA of RCG2A and RGC2B, respectively. A 3.4-kb 3′ RACE product was isolated using a primer located at the end of exon 2 (Table 1). This cDNA was identical to the exon sequences of RGC2C and included a 283-bp 3′ UTR and a poly(A) tail. RACE using primers designed 5′ to or within the 3′ UTR did not amplify any larger products, indicating that we had identified the 3′ end of the gene.

The genomic structure of the nine RGC2 genes was similar. All genes, except for RGC2D (described below), had eight exons. Intron–exon splice boundaries were identified by comparison to the 5′ and 3′ RACE products described above and by computer analysis of the genomic sequence to predict putative splice sites (Hebsgaard et al., 1996). Computer analysis predicted splice sites in all genes at the same locations as those identified by comparisons of cDNA and genomic sequences. The predicted mRNA is ~5.9 kb. The open reading frame accounted for 5274 to 5757 bp of the genomic sequence, encoding a predicted protein of 1758 to 1919 amino acids. The length of the complete genomic sequence varied from 7 kb (RGC2O) to 13.6 kb (RGC2J) (Figure 1 and Table 2). Therefore, the RGC2 genes are among the largest genes thus far reported in plants. Most of the size difference between copies was due to differences in intron 3 (Figure 1 and Table 2). Comparisons between the sequences of the RGC2 genes and the insertion site of a T-DNA that destroyed Dm3 activity (Okubara et al., 1997) demonstrated that the insertion had occurred in intron 3 of RGC2B.

View this table:
Table 1.

Oligonucleotide Primers Used in This Study

Figure 1.
Figure 1.

Structure of Nine RGC2 Genes.

The coding region starts from the ATG, as marked. The 5′ untranslated leader sequence was identified in a subset of the genes by analysis of RACE products (see text). Dotted lines in the 5′ and 3′ regions of some copies indicate that genomic sequence was obtained but the intron–exon boundaries were not determined. RGC2A and RGC2K contain repeats as indicated (arrowheads). RGC2A and RGC2N contain stop codons at the positions indicated (asterisks); RGC2N and RGC2D contain deletions resulting in frameshift mutations at the positions indicated by the open triangle. The size of the intron 3 in RGC2D could not be determined, as indicated by the question mark (see text). The position of the poly(A) site identified in a 3′ RACE product of RGC2D is shown.

The predicted amino acid sequence of the RGC2B protein contains several distinct motifs as well as regions with no obvious homologies. The N-terminal and C-terminal regions of the protein have no significant similarities to sequences in the databases, including other disease resistance genes (Figures 2A and 2F). Unlike the class of NBS-LRR resistance genes, which includes the tobacco gene N, the flax gene L6, and the Arabidopsis gene RPP5, we found no homology to the N terminus of the Toll–interleukin-1 homology domain (TIR domain; Baker et al., 1997; Hammond-Kosack and Jones, 1997; Parker et al., 1997). No leucine zipper motifs were identified; this motif has been predicted from the nucleotide sequence of the resistance genes RPS2, RPM1, and Prf (Bent et al., 1994; Mindrinos et al., 1994; Grant et al., 1995; Salmeron et al., 1996). Adjacent to the N-terminal region, the RGC2B protein contains an NBS that is identifiable by the presence of the conserved P loop domain with the sequence GMGGVGKT, which is followed by four other characteristic motifs (Figure 2B). The sequence and spacing of these motifs and the position of the NBS in the protein are consistent with other known plant resistance genes (Hammond-Kosack and Jones, 1997). A short region with no significant similarity to other known genes separates the NBS from a C-terminal LRR region (Figure 2C).

The C-terminal two-thirds of the predicted protein is rich in leucine and other aliphatic residues and comprises a series of irregular repeats (Figures 2D and 2E). The LxxLxxaxaxxCxxaxxa (where x is any amino acid and a is a conserved aliphatic amino acid) consensus of RGC2 LRRs is more closely related to the predicted cytoplasmic LRR consensus LxxLxxLxLxx(N/C/T)x(x)LxxIPxxaxx than to the extracytoplasmic consensus LxxLxxLxLxxNxLxGxIPxxLx (Jones and Jones, 1997). However, the RGC2 LRRs are degenerate in comparison with this consensus and vary in length (Figures 2D and 2E). RGC2 genes encode ~20 LRRs 5′ of intron 3 and ~21 LRRs 3′ of intron 3. The highly variable intron bisects the LRR region and defines a bipartite configuration in which the C-terminal region exhibits a more evident alternating pattern of hypervariable and conserved amino acids (see below). The total of ~41 LRRs is larger than any previously reported LRR region (Kobe and Deisenhofer, 1994). Several regions with few aliphatic residues and a poor match to known LRR consensus sequences interrupt the LRR region. It is possible that these are “loop-out” regions, providing some sort of a molecular hinge between LRR regions, as proposed for the Cf-4 and Cf-9 resistance genes of tomato (Jones and Jones, 1997).

View this table:
Table 2.

Sizes of Coding Regions and Introns in Nine RGC2 Genes Sequenceda

One region in exon 5, encoding residues that do not match the LRR consensus, contains a small, compound, in-frame trinucleotide repeat designated MSATE6. This sequence is a derivative of the consensus (ACA)xACGAAGGGG(TCT)y and encodes polythreonine, a three–amino acid intervening sequence, and an adjacent stretch of polyserine (Figure 2E). The microsatellite MSATE6 is hypervariable among RGC2 copies and was quite useful for mapping and detecting transcripts of particular members of the RGC2 gene family (Meyers et al., 1998). In RGC2K, the microsatellite is (ACA)2AAGGCA-(TCT)2, representing the minimal repeat size observed in the RGC2 family. The largest array, (ACA)5ACGAAGGGG-(TCT)21, is in RGC2J. The function of this region in the protein is not known. However, this microsatellite sequence is the site of differences in half of the nine RGC2 copies sequenced: a 1.2-kb deletion in RGC2D, two large direct repeats in RGC2K, a 45-bp deletion in RGC2O, and a stop codon that occurs just 5′ of the microsatellite in RGC2A (Figure 1).

Transcript Analysis of the RGC2 Family

Six of the nine genes contain complete open reading frames; however, a variety of mutations indicated that the remaining three are pseudogenes. RGC2A contains a nonsense mutation in exon 5, 3 bp 5′ of the microsatellite MSATE6. The RGC2A microsatellite allele is missing from the cDNA, indicating that the 3′ end of this gene is not expressed. RGC2N contains a nonsense mutation in exon 2, ~2.2 kb downstream of the start codon, as well as a 1-bp deletion in exon 4 (Figure 1). Comparisons of the genomic sequence of RGC2D to other copies revealed an ~1.3-kb deletion that fused exon 5 to exon 6 and introduced a frameshift. This fusion in exon 5 occurred 51 bp 5′ of the microsatellite MSATE6 and eliminated parts of both exons, all of intron 5, and the microsatellite. A 3′ RACE product was obtained that was identical to the 5′ coding sequence of RGC2D but contained a poly(A) tail 176 bp 3′ of exon 3 (Figure 1). Intron donor and acceptor splice sites are present at both ends of intron 3 in RGC2D. However, we sequenced almost 6 kb from both ends of intron 3, although polymerase chain reaction (PCR) failed to amplify across the predicted gap in the sequence. Therefore, either intron 3 of RGC2D is too large to be amplified by PCR (at least ~10 kb total) or the two ends of the gene are rearranged with respect to each other.

Members of the RGC2 family that are transcribed were identified by analyzing RACE products and assaying the MSATE6 microsatellite in cDNA. Sequences of four RGC2 family members, RGC2B, RGC2C, RGC2D, and RGC2N, were identified as RACE products. Microsatellite MSATE6 (as described above) was amplified from at least seven copies by using a cDNA template, including an additional four, RGC2J, RGC2I, RGC2E, and RGC2S, which had not been identified by RACE analysis (Figure 3). Interestingly, the largest allele of MSATE6 from RGC2J is transcribed. The MSAT data, together with the sequenced RACE products, indicate that at least eight RGC2 copies are transcribed.

Genomic Comparisons between RGC2 Family Members

Large variations in exon size were observed only in the LRR-encoding regions. Throughout the coding regions, there were small indels of between one and seven codons that maintained intact open reading frames. Greater variation resulted in changes in the number of encoded LRRs. Relative to other RGC2 genes, RGC2A and RGC2N contain a direct repeat in exon 3 (80% nucleotide identity) that encodes approximately two LRRs (Figure 1). RGC2K contains a direct repeat of 480 bp in exon 5 (78% nucleotide identity) that encodes approximately six LRRs (Figure 1).

Figure 2.
Figure 2.

Amino Acid Sequence for the Predicted Full-Length Transcript of RGC2B.

The amino acid sequence is shown in single-letter code and is divided into six regions.

(A) The N terminus.

(B) The nucleotide binding site.

(C) A connecting region.

(D) The N-terminal LRR region.

(E) The C-terminal LRR region.

(F) The C terminus.

The conserved motifs of the NBS and the microsatellite MSATE6 are underlined in (B) and (E), respectively. LRRs have been aligned according to the consensus sequence given at bottom, which approximates the consensus for cytoplasmic LRRs (Jones and Jones, 1997); “a” indicates the positions of aliphatic amino acids. The positions of introns are shown by diamonds. Aliphatic and cysteine residues are in red and blue, respectively.

Intron positions but not sizes were found to be conserved between copies. RGC2 genes have five introns in the coding region, one in the 5′ UTR and one in the 3′ UTR. In the coding region, all introns are 3′ to the NBS-encoding region. Most introns range in length from 59 to 1815 bp. Extreme size variation was found to occur in intron 3, whereas the other introns show less variation (Table 2). The smallest intron 3 was 363 bp in RGC2O. This intron was four to 16 times longer in other family members; intron 3 was 6097 bp in RGC2J.

The similarities between intron 3 sequences are limited to regions adjacent to the splice sites (Figure 4). Sequence comparisons between the copies indicated that the ends of the intron have a high degree of similarity (>70% identical); however, this is lost within 400 to 500 bp. The middle of the introns contain DNA with no homology to distantly related RGC2 copies or to any sequences in the databases. Intron 3 was found to be closely related and to exceed 5 kb in four RGC2 genes: RGC2B, RGC2C, RGC2D, and RGC2S. Exon sequences indicate that RGC2O is also closely related to these four copies; however, intron 3 in RGC2O is <8% of intron 3 in RGC2B, the smallest of the above four genes. In the more diverse genes (RGC2K, RGC2A, and RGC2J), intron 3 varies from 1 to 6.1 kb, yet these introns have little sequence similarity to each other and to the RGC2B subfamily. Although genome expansion in intergenic regions of many plant species has been attributed to insertions of transposable elements (SanMiguel et al., 1996), no sequences homologous to known transposable elements were found in intron 3. We also were unable to identify terminal repeats characteristically associated with long terminal repeat elements or miniature inverted-repeat transposable elements that are often present in plant genes (Wessler et al., 1995).

Figure 3.
Figure 3.

MSATE6 from BAC, Genomic, and cDNA Templates.

Primers were designed from RGC2A to amplify a microsatellite marker from exon 5. MSATE6 was amplified from genomic DNA of cultivar Diana, cDNA of cultivar Diana, and 12 BAC clones each containing a single copy of RGC2. Numbers to the right designate individual bands. The cDNA lacked band 4, which represents the RGC2A sequence from which the primers were designed, indicating no contamination of the cDNA with genomic DNA. No BAC clones containing band 2 were identified (Meyers et al., 1998). Amplification of the microsatellite was not expected from the more divergent members of the RGC2 family in lettuce genomic DNA because of mismatches at the priming sites; however, microsatellites were amplified from BAC templates containing the divergent members.

Sequence conservation outside of the coding region was detected only for closely related genes. We obtained from 0.8 to 4.1 kb of genomic sequence 5 ′ to the ATG for RGC2A, RGC2B, RGC2D, RGC2K, RGC2O, and RGC2N. Sequence similarity was >96% between two sets of closely related copies: RGC2A and RGC2N, and RGC2B and RGC2S. Little sequence similarity was found between other sequences, indicating a high degree of divergence in both intron 3 and upstream sequences. Beyond the 3′ end of the open reading frame, 0.3 to 2.1 kb of sequence information was obtained for RGC2B, RGC2K, and RGC2J; again, the sequences outside of these divergent coding regions were unrelated. However, a probe from 4 kb 3′ of RGC2B hybridized with at least 11 members of the RGC2 family (Meyers et al., 1998); therefore, there may be regions of sequence conservation beyond the 3′ end of the genes.

Comparisons of Nucleotide Substitution Patterns in Different Regions of Resistance Genes

A comparison between the aligned deduced amino acid sequences revealed an alternating pattern of variable and conserved amino acids in the LRR region. This pattern was more pronounced in the C-terminal half of the LRR region, which is encoded 3′ of intron 3. The hypervariable amino acids in each repeat are positioned around two conserved aliphatic amino acid sites in the consensus xx(a)x(a)xx. In the porcine ribonuclease inhibitor, these amino acids form parallel β sheets flanked by β turns (Kobe and Deisenhofer, 1994; Jones and Jones, 1997); these comprise a solvent-exposed surface that interacts with the ligand (Kobe and Deisenhofer, 1995).

Frequencies of nonsynonymous (Ka) and synonymous (Ks) nucleotide substitutions and Ka/Ks ratios were calculated for five different regions of the open reading frame: the 5′ end, the NBS-encoding region, the spacer between the NBS- and LRR-encoding regions, the 5′-encoded LRR region, and the 3′-encoded LRR region (Figure 5). Similar analyses in mammalian genes involved in pathogen recognition have detected higher rates of nonsynonymous than synonymous substitution in ligand binding regions (Hughes and Nei, 1988; Tanaka and Nei, 1989). Ka/Ks ratios <1 may result from the elimination of most nonsynonymous substitutions through purifying selection. Ka/Ks ratios >1 indicate diversifying selection (Li, 1997). When the complete open reading frames of RGC2 genes are compared, Ks exceeds Ka, indicating conservation of the gene as a whole (Figure 5). The NBS-encoding region was the most highly conserved portion of the gene, with an average Ka/Ks ratio of 0.374 (Figures 5 and 6). Statistical analysis using a G test strongly rejected the null hypothesis for neutral evolution (33 of 36 pairwise comparisons were significantly <1 at P < 0.000001). The 5′ and NBS-LRR spacer regions also had Ka/Ks ratios <1 (Figures 5 and 6A to 6F).

Figure 4.
Figure 4.

Sequence Similarity between Intron 3 of Nine RGC2 Copies.

Pairwise comparisons were performed using intron 3 sequences from nine RGC2 copies. Sequences demonstrating the greatest similarity within intron 3 were placed together. Regions with 65 to 90% similarity between copies are shown in light gray; regions with >90% similarity are dark gray. Unshaded regions are <65% similar. RGC2D contains a gap in the sequence (?; see text).

Figure 5.
Figure 5.

Ka and Ks Values among RGC2 Genes.

Values were calculated for nonsynonymous (Ka) and synonymous (Ks) substitutions in the protein coding regions of the gene. Values were calculated for 36 pairwise comparisons and averaged. The range of the values is given below the Ka and Ks averages to indicate the diversity of the sequences compared. The Ka/Ks ratio was calculated by averaging the ratio for each comparison. Individual Ka and Ks values are plotted in Figure 6.

The LRR-encoding region of RGC2 genes had unusual substitution patterns. The two halves of the LRR-encoding region separated by intron 3 were considered independently because of the more pronounced pattern of variability in the 3′ portion. In the porcine ribonuclease inhibitor, the aliphatic residues in the xx(a)x(a)xx consensus are buried in the hydrophobic core of the protein and do not interact with the ligand (Kobe and Deisenhofer, 1994, 1995). Therefore, the Ka/Ks ratio was calculated for the nucleotides encoding the xx(a)x(a)xx region of the LRR repeats in RGC2, omitting the codons for the conserved aliphatic positions. The Ka/Ks ratio of the codons corresponding to the xx(a)x(a)xx amino acids of the C-terminal LRR region was significantly >1 in 11 of 36 comparisons, indicating that these residues are under divergent selection (Figure 6 and Table 3). The same positions in the 5′ end of the encoded LRR showed elevated Ka/Ks ratios, but the average was not >1. The Ka/Ks ratio was calculated separately for the LRR-encoding sequence between encoded xx(a)x(a)xx motifs, designated the “intervening residues.” The Ka/Ks ratios for the intervening residues were 0.553 in the 5′ and 0.635 in the 3′ regions (Figure 5). The ratios for many pairwise comparisons were significantly <1, indicating purifying rather than divergent selection. In a G test, 33 of 36 pairwise comparisons were significant at the P < 0.01 level for the 5′-encoded LRR, and 19 of 36 pairwise comparisons were significant at the P < 0.05 level for the 3′-encoded LRR; the lower number of significant comparisons in the 3′-encoded LRR reflects a low proportion of variable sites in this region. The statistically highly significant pattern of Ka/Ks ratios is evidence for a conserved backbone alternating with arrays of solvent-exposed β-sheet surfaces that are under diversifying selection in the LRR region of RGC2 proteins.

To determine whether alternating patterns of variability are a common feature of LRR regions in other plant disease resistance genes, we calculated the Ka/Ks ratios for the LRR- and putative effector-encoding regions of three types of LRR-containing resistance genes from other plant species. Alignments were made within the I2C and Mi families of tomato (Ori et al., 1997; Milligan et al., 1998), the Xa21 family of rice (Song et al., 1997), and between the L6 and M genes of flax (Lawrence et al., 1995; Anderson et al., 1997). The I2C, Mi, and Xa21 families represent paralogs within localized clusters of genes. L and M are homoeologous loci derived from an ancient polyploidization event (Ellis et al., 1995). The Ka/Ks ratios were calculated for the same regions as for RGC2, except that the LRR-encoding regions were not split in two because no bipartite structure was apparent. Comparisons were also made to data for the Cf-4/9 cluster from tomato (Parniske et al., 1997).

In each comparison, the Ka/Ks ratios indicated that all regions are under purifying selection except for the xx(a)x(a)xx residues of the LRRs (Figure 6 and Table 4). In every comparison, the codons for the xx(a)x(a)xx residues had a Ka/Ks ratio >1, indicating diversifying selection, whereas the different putative effector-encoding regions for each type of resistance gene had Ka/Ks ratios of <1, indicating selection for conservation. Many of these comparisons were significant only at the 5 to 10% level (Table 4). However, they are consistent with the highly significant data for RGC2. Xa21 paralogs have been grouped into two distinct classes based on sequence similarity (Song et al., 1997). Only comparisons within the Xa21 class of paralogs (i.e., B [the functional Xa21 gene], D, and F) showed elevated Ka/Ks ratios, and only one of these comparisons is significant at the 5% level (Table 4). Comparisons including other Xa21 family members (the A2 class) had Ka/Ks ratios <1 (data not shown), possibly indicating that these genes are not functional in disease resistance.

Figure 6.
Figure 6.

Synonymous and Nonsynonymous Substitution Frequencies in the RGC2 Family and Other Plant Disease Resistance Genes.

(A) 5′ and spacer regions of RGC2 genes.

(B) Regions encoding putative effector domains.

(C) RGC2 5′-encoded LRR.

(D) RGC2 3′-encoded LRR.

(E) The LRR-encoding region of other resistance gene families.

Ks and Ka substitutions were calculated for pairwise comparisons within resistance gene families. These values were plotted in the form (Ks, Ka). The diagonal corresponds to Ks = Ka, representing neutral evolution; points above this line provide evidence for diversifying selection, without implying statistical significance. Points below the diagonal suggest selection for conservation.

Wang et al. (1998) found evidence for diversifying selection only in the Xa21B/Xa21D comparison by using comparisons of the entire LRR-encoding region. Elevated Ka/Ks ratios previously have been reported for the Cf-4/9 genes in tomato (Parniske et al., 1997); however, in our reanalysis of these sequences, most pairwise comparisons are not significant (Table 4), suggesting that only some members of the Cf-4/9 family are currently under diversifying selection. In summary, evidence for diversifying selection of varying levels of significance was found in all comparisons within LRR-containing families of plant disease resistance genes. In each case, this selection was acting on the region that may comprise a solvent-exposed surface in other LRR-containing proteins.

DISCUSSION

The Dm3 downy mildew resistance locus of lettuce is composed of at least 24 diverse copies that span an estimated 3.5 Mb (Meyers et al., 1998). This locus is larger and more complex than are clusters of resistance genes described to date. Sequencing of nine full-length genomic copies demonstrated that the RGC2 NBS-LRR genes are among the largest and most diverse plant resistance genes. The distribution of variable amino acids and patterns of nucleotide substitution support a model for divergent selection acting on amino acid residues that comprise the putative ligand binding surfaces.

RGC2 Genes Are Similar to but Distinct from Other NBS-LRR Resistance Genes

The RGC2 genes are similar to many of the other known plant disease resistance genes but have several distinct features. The RGC2 genes encode polypeptides of 1758 to 1919 amino acids; these are among the largest of any disease resistance genes known in plants. Xa1 in rice, conferring resistance to X. o. oryzae, is 1802 amino acids (Yoshimura et al., 1998); Prf in tomato, required for resistance to Pseudomonas syringae pv tomato, is 1824 amino acids (Salmeron et al., 1996). The remainder of known NBS-LRR–type resistance products in plants vary from 909 (RPS2, Bent et al., 1994; Mindrinos et al., 1994) to 1361 (RPP5, Parker et al., 1997) amino acids. The N terminus of the RGC2 proteins comprises a short region with no similarity to proteins in databases, including other resistance gene products. The NBS region is similar in size and motifs to other resistance gene products. A short region that again has no similarity to sequences in the databases separates the NBS from the LRR domain.

The C-terminal region of RGC2 proteins contains many LRRs. The RGC2 consensus sequence is related to cytoplasmic LRR proteins (Kobe and Deisenhofer, 1994; Jones and Jones, 1997), although it is more degenerate and more variable in length. The total of >40 LRRs spanning 1249 to 1380 amino acids is larger than any previously reported LRR region (Kobe and Deisenhofer, 1994; Jones and Jones, 1997). It is considerably larger than LRR regions found in similarly sized resistance gene products; those of Xa1, RPP5, and Prf include ~558 (the number of LRRs cannot be determined because the structure of the Xa1 gene is atypical), 575 (21 LRRs), and 417 (18 LRRs) amino acids, respectively. Therefore, each LRR domain of the RGC2 proteins encoded either side of intron 3 (~20 and ~21 LRRs) is of similar size to the entire LRR region encoded by other NBS-LRR–type resistance genes.

We identified a hypervariable compound microsatellite within the coding region of RGC2 genes. Although members of this family containing different sizes of this repeat are transcribed, it is not known whether this repeat influences gene function. Some animal genes tolerate hypervariable trinucleotide repeats in coding sequences; repeats encoding polyglutamine have been identified within animal receptors involved in growth and development, but the function of these repeats also is not known (Edwards et al., 1991). Trimeric repeats within transcribed sequences also have been identified as the cause of dysfunctional alleles in several human genes, including the fragile-X syndrome and myotonic dystrophy in humans (Fu et al., 1991; Brook et al., 1992). In both cases, phenotypically normal individuals may have many repeats (<46 for fragile-X or <27 for myotonic dystrophy; Fu et al., 1991; Brook et al., 1992); expansion beyond a threshold results in a dysfunctional genotype. It remains to be determined whether the size of the repeat affects the activity or the recognition specificity of RGC2 genes. Expansion or contraction of the microsatellite sequence could alter the spacing of binding surfaces determined by LRRs flanking the repeat.

View this table:
Table 3.

Subset of Comparisons within the RGC2 Family That Show an Alternating Pattern of Positive and Purifying Selectiona

View this table:
Table 4.

Ka/Ks Ratios in Different Regions of Plant Resistance Genes

Intron position and number are conserved between RGC2 family members but differ from other resistance genes. RGC2 genes have seven introns, with five in the coding region. I2C, RPS2, and RPM1 lack introns (reviewed in Hammond-Kossack and Jones, 1997). Xa1 has three introns, of which two are in the coding region 5′ to the region encoding the NBS (Yoshimura et al., 1998). Three intron positions are shared among the TIR-NBS-LRR class of resistance genes N, L6, and RPP5 (Parker et al., 1997). Therefore, the size of the LRR region and the position of the introns indicate that RGC2 genes are members of a family of resistance genes that is distinct from those characterized to date.

Diversity in Intron 3 Suggests Distinct Lineages of RGC2 Genes

Most of the variation in the size of RGC2 genes is due to intron 3, which ranged from <400 bp to >6 kb. The disparate sequences and sizes of this intron suggest a complex evolutionary history. The diversity of sequences indicates that intron 3 has evolved independently in different lineages of the RGC2 gene family. Without knowing the progenitor or ancestral gene sequence, we cannot determine whether the variation in size was due to insertions or deletions. The numerous indels in intron 3 had no homologs in the databases, terminal inverted repeats characteristic of transposable elements, obvious secondary structure, or duplications of sequences flanking the indels (Wessler et al., 1995; Bennetzen, 1996). Therefore, although the Dm3 region contains transposable elements and retrotransposable elements in the intergenic regions (Meyers et al., 1998), we found no evidence for such elements within the RGC2 gene sequences. Consequently, the mechanisms generating the variation in intron size are not apparent.

The degree of sequence divergence in the intron might influence meiotic pairing and hence the frequency of unequal crossing over between paralogs. Pairing between more diverse sequences would tend to be repressed and result in decreased levels of recombination and gene conversion. Consequently, a high level of sequence diversity would be maintained, and individual members would tend to evolve independently. Genetic analysis of the Rp1 disease resistance cluster of maize indicates that recombination is lower between more distantly related haplotypes (Sudupak et al., 1993). In the Cf-4/9 resistance gene locus of tomato, it has been proposed that dissimilar intergenic regions suppress mispairing in homozygotes (Parniske et al., 1997). The coding region of RGC2 is more than twice as large as that of the genes in the Cf-4/9 cluster, and divergent intron sequences in the middle of the gene would be expected to affect pairing behavior. The consistent sequence diversity observed among members of the RGC2 family supports the hypothesis that there is little sequence exchange to homogenize these genes.

Selective Influences Differ across the RGC2 Gene

Nucleotide substitution patterns in the RGC2 family vary across the gene, particularly within the region encoding the LRRs. Synonymous substitutions have occurred at a higher frequency in the 5′ end of RGC2. Nonsynonymous substitutions were found at a significantly lower frequency than that of the synonymous substitutions, particularly within the NBS-encoding region; therefore, the low Ka/Ks ratio of the NBS-encoding region, which is the putative effector region, indicates that it has undergone the highest level of purifying selection within the gene. Within the region encoding the LRR domain, the putative ligand binding domain, there was an alternating pattern of conserved and variable amino acids. This was particularly evident in the 3′ end of the RGC2 LRR-encoding region. The conserved regions correspond to amino acids that may form a structural backbone of the LRR; the hypervariable amino acids are predicted to form β sheets that are involved in ligand binding (Kajava et al., 1995; Kobe and Deisenhofer, 1995; Jones and Jones, 1997). Furthermore, the Ka/Ks ratios in the putative ligand binding surfaces of the 3′-encoded LRRs were >1, implying that divergent selection occurs at these positions. This pattern was found in the LRR-encoding region of genes from diverse plant species that confer resistance to pathogens that include fungi, bacteria, and nematodes. These genes represent three of four described structural classes for plant resistance genes (reviewed in Baker et al., 1997), indicating that the LRR sequence of these disparate resistance gene products is influenced by natural selection in a similar manner. This pattern is further evidence that the LRR region of resistance gene products is composed of a conserved backbone with variation localized in solvent-exposed surfaces.

Because patterns of amino acid variability and calculations of nucleotide substitutions evaluate groups of amino acids or nucleotides, they are inherently limited and cannot identify particular amino acids or LRRs critical to ligand binding. Mutations that alter receptor activity and ligand binding could occur anywhere in the gene. Also, some LRRs or portions of the LRR region may have a greater functional role in ligand binding than others. LRRs in the C-terminal half of Cf proteins are highly conserved, and the hypervariable residues are localized to the N-terminal half of the protein (Parniske et al., 1997; Thomas et al., 1997). In RGC2 genes, the 3′ half of the LRR-encoding region was the more variable and showed higher Ka/Ks ratios than did the 5′-encoded LRR region. The detection of a hypervariable region within the LRR in all types of resistance genes analyzed suggests some biological significance to this pattern. However, further experimentation is necessary to confirm and refine conclusions derived from Ka and Ks calculations. In flax, domain swaps between alleles of the L gene indicate that the LRR is an important determinant of specificity (Ellis et al., 1997). Domain swaps between homologs followed by site-directed mutagenesis will focus on delineating regions of RGC2 genes critical for specificity determination.

All classes of LRR-containing resistance gene products contain hypervariable surfaces within the putative receptor domain. The statistically significant evidence for divergent selection acting on many RGC2 genes indicates that they must have been active in recognition of pathogens. This is consistent with our expression data. It is in contrast to the Cf-4/9 data in which many comparisons result in Ka/Ks ratios ⩽1 (Figure 6), indicating that numerous copies have not been under recent divergent selection. A large number of variable LRRs, particularly as found in RGC2, could increase binding opportunities or form multiple LRR subdomains. Such diversity in LRRs could increase the breadth of protein–ligand interactions and provide the flexibility for plants to coevolve with diverse pathogens.

Few proteins seem to be subject to diversifying selection, as indicated by a Ka/Ks ratio >1. Analyses of nucleotide substitution frequencies have detected evidence for diversifying selection in only 17 of 3595 families of sequences, and more than half of the 17 are antigenic surface proteins of parasites and viruses (Endo et al., 1996). The selective advantage of elevated occurrences of nonsynonymous substitutions has been most studied in the antigen recognition site (ARS) of class I mammalian major histocompatability complex (MHC) genes and the complementarity-determining region (CDR) of Ig genes (Hughes and Nei, 1988; Tanaka and Nei, 1989; Nei et al., 1997). Both the ARS and the CDR are responsible for recognizing and binding a wide range of potential ligands; evidence for diversifying selection in these genes suggests that variation is evolutionarily advantageous. The proposed ligand binding region of the RGC2 gene products and other plant resistance gene products also appears to be under diversifying selection. The evolution of new recognition specificities in MHC and Ig genes involves variation in the individual amino acids as well as recombination and gene conversion. The relative importance of each of these sources of variation in the evolution of plant resistance genes remains to be determined.

Mechanisms Involved in the Evolution of Resistance Genes

Models describing the evolution of plant resistance genes have proposed recombination and gene conversion as the primary forces that generate diversity within these multigene families (Shepherd and Mayo, 1972; Pryor, 1987; Richter et al., 1995; Hammond-Kosack and Jones, 1997). The models assume that these events result in the rapid evolution of novel resistance specificities to counteract variable pathogen populations. Unequal crossing over is clearly involved in the generation of duplicated arrays of resistance genes. The size of multigene families at resistance loci varies between haplotypes (Parniske et al., 1997; D.T. Lavelle and R.W. Michelmore, unpublished data). The Cf-2 locus of tomato includes two functional genes that are >99.9% identical, suggesting recent sequence duplication (Dixon et al., 1996). Novel specificities may also result from recombination or gene conversion. Recombination and unequal crossing over at the Rp1 resistance gene complex of maize are involved in meiotic instability and the generation of new specificities (Sudupak et al., 1993; Richter et al., 1995; Hulbert, 1997). Sequence analysis of the Cf-4/9 locus of tomato indicates that sequence exchange has occurred between gene family members, resulting from either recombination or gene conversion (Dixon et al., 1996; Parniske et al., 1997). Also, mutants identified at the M locus of flax may have resulted from intragenic recombination (Anderson et al., 1997). Recombination is therefore involved in alterations in the copy number of resistance genes and in the generation of novel resistance specificities.

The importance of single base changes in the evolution of plant disease resistance genes may have been overlooked. There is now evidence for diversifying selection in the predicted β-sheet portion of the LRR consensus in all types of LRR-encoding resistance genes (Parniske et al., 1997; Wang et al., 1998; this study). The nonsynonymous substitutions that have accumulated in the region encoding the putative ligand binding domain may substantially affect recognition specificity. Recombination between alleles and paralogs could result in enhanced or novel binding properties by shuffling LRRs between related genes. The relative importance of recombination versus single base changes depends on the rate at which each of these events occurs. Although unequal crossing over has been measured at some resistance gene clusters (Richter et al., 1995), there are few data on the point mutation rates in these genes. The high level of sequence diversity in the RGC2 family indicates that recombination and gene conversion are not homogenizing these sequences and are infrequent events (this study; Meyers et al., 1998). Therefore, novel amino acid substitutions in the solvent-exposed surfaces of the LRR region may be more important than intergenic recombination and gene conversion in the rapid evolution of novel specificities.

METHODS

Sequencing of RGC2 Copies

Genomic sequences of the RGC2 genes and flanking regions were obtained using a combination of several methods. All RGC2 copies were sequenced from the bacterial artificial chromosome (BAC) clones on which they were originally identified (Meyers et al., 1998). RGC2A and RGC2B copies were primarily sequenced using a primer walking strategy. Primers were synthesized by Gibco Life Technologies (Grand Island, NY). When possible, primers generated during the sequencing of RGC2A and RGC2B were used to sequence further RGC2 copies; additional primers for the more divergent members were synthesized as required. Some sequencing was performed using a modified version of the polymerase chain reaction (PCR)–based long-distance sequencer method (Hagiwara and Harris, 1996). To simplify the Hagiwara and Harris (1996) method, we performed a standard PCR, with 35 cycles, annealing for 30 sec at 58°C, and extension for 2 min at 72°C. PCR fragments were purified before sequencing by exonuclease I/shrimp alkaline phosphatase treatment (U.S. Biochemical Corp.) to remove unincorporated deoxynucleotide triphosphates and excess primers. Restriction digests, ligations, and agarose gel analysis of DNA fragments were performed according to standard protocols (Sambrook et al., 1989).

DNA Sequencing and Analysis

DNA sequencing was performed using an ABI 377 automated sequencer (Applied Biosystems Inc., Foster City, CA) and the PRISM Ready Reaction DyeDeoxy Terminator cycle sequencing kit (Applied Biosystems Inc.) with custom primers or standard Sp6, T7, M13 (-21), or M13 reverse primers. Sequence data were evaluated using Sequencher (GeneCodes, Ann Arbor, MI) for contig assembly and sequence editing. Splice site analysis and intron–exon boundaries were determined by comparison with cDNA sequences and by using the software program NetPlantGene available at www.cbs.dtu.dk/services/NetPgene/ (Hebsgaard et al., 1996). GeneDoc (www.cris.com/~ketchup/genedoc.shtml), DNAstar (Lasergene, Madison, WI), and Genetics Computer Group (Madison, WI) software packages were used for multiple sequence alignments and sequence comparisons. Nucleotide substitution rates were calculated by the method of Li (1993) in the Diverge program in the Genetics Computer Group software package. A 2 × 2 contingency table G test was used to test for the significance of differences in synonymous and nonsynonymous substitution rates (Zhang et al., 1997). The values for the 2 × 2 contingency table were estimated by using the model of Nei and Gojobori (1986). Ka and Ks values were plotted with the Freelance Graphics program (Lotus, Cambridge, MA). Phylogenetic studies were performed with PAUP* version 4.0 (Sinauer Associates, Sunderland, MA).

cDNA Analysis

RNA was isolated from lettuce (Lactuca sativa) cultivar Diana via the procedure of Jones et al. (1995). First-strand cDNA was synthesized by use of Superscript reverse transcriptase from 1 μg of total RNA, as specified by the manufacturer (Gibco Life Technologies). 5′ and 3′ rapid amplification of cDNA ends (RACE; Frohman et al., 1988) was performed using the Marathon kit (Clontech, Palo Alto, CA), according to the manufacturer's instructions, with primers designed from the predicted open reading frames observed in the genomic sequence.

GenBank accession numbers are as follows: RGC2A, AF072268; RGC2B, AF072267; RGC2C, AF072269; RGC2D, AF072270; RGC2J, AF072271; RGC2K, AF072272; RGC2N, AF072273; RGC2O, AF072274; and RGC2S, AF072275.

ACKNOWLEDGMENTS

We thank Dean O. Lavelle for technical assistance. This work was supported by the U.S. Department of Agriculture National Research Initiative Competitive Grant Program (Grant No. 95-37300-1571). Partial support for B.C.M. was provided by a National Science Foundation graduate research fellowship.

Footnotes

  • ↵1 Current address: DuPont Agricultural Biotechnology, Delaware Technology Park, Newark, DE 19714.

  • Received June 12, 1998.
  • Accepted September 14, 1998.
  • Published November 1, 1998.

REFERENCES

  1. ↵
    1. Anderson P.A.,
    2. Okubara P.A.,
    3. Arroyo-Garcia R.,
    4. Meyers B.C.,
    5. Michelmore R.W.
    (1996). Molecular analysis of irradiation-induced and spontaneous deletion mutants at a disease resistance locus in Lactuca sativa. Mol. Gen. Genet. 251, 316–325.
    OpenUrlPubMed
  2. ↵
    1. Anderson P.A.,
    2. Lawrence G.J.,
    3. Morrish B.C.,
    4. Ayliffe M.A.,
    5. Finnegan E.J.,
    6. Ellis J.G.
    (1997). Inactivation of the flax rust resistance gene M associated with loss of a repeated unit within the leucine-rich repeat coding region. Plant Cell 9, 641–651.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Baker B.,
    2. Zambryski P.,
    3. Staskawicz B.,
    4. Dinesh-Kumar S.P.
    (1997). Signaling in plant–microbe interactions. Science 276, 726–733.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    1. Bennetzen J.L.
    (1996). The contributions of retroelements to plant genome organization, function and evolution. Trends Microbiol. 4, 347–353.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Bent A.F.
    (1996). Plant disease resistance genes: Function meets structure. Plant Cell 8, 1757–1771.
    OpenUrlFREE Full Text
  6. ↵
    1. Bent A.F.,
    2. Kunkel B.N.,
    3. Dahlbeck D.,
    4. Brown K.L.,
    5. Schmidt R.,
    6. Giraudat J.,
    7. Leung J.,
    8. Staskawicz B.J.
    (1994). RPS2 of Arabidopsis thaliana: A leucine-rich repeat class of plant disease resistance genes. Science 265, 1856–1860.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Bevan M.,
    2. et al.
    (1998). Analysis of 1.9 Mb of contiguous sequence from chromosome 4 of Arabidopsis thaliana. Nature 391, 485–488.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Brook J.D.,
    2. McCurrach M.E.,
    3. Harley H.G.,
    4. Buckler A.J.,
    5. Church D.,
    6. Aburatani H.,
    7. Hunter K.,
    8. Stanton V.P.,
    9. Thirion J.P.,
    10. Hudson T.,
    11. Sohn R.,
    12. Zemelman B.,
    13. Snell R.G.,
    14. Rundle S.A.,
    15. Crow S.,
    16. Davies J.,
    17. Shelbourne P.,
    18. Buxton J.,
    19. Jones C.,
    20. Juvonen V.,
    21. Johnson K.,
    22. Harper P.S.,
    23. Shaw D.J.,
    24. Housman D.E.
    (1992). Molecular basis of myotonic dystrophy: Expansion of a trinucleotide (CTG) repeat at the 3′ end of a transcript encoding a protein kinase family member. Cell 68, 799–808.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Crute I.R.,
    2. Pink D.A.C.
    (1996). Genetics and utilization of pathogen resistance in plants. Plant Cell 8, 1747–1755.
    OpenUrlFREE Full Text
  10. ↵
    1. Dixon M.S.,
    2. Jones D.A.,
    3. Keddie J.S.,
    4. Thomas C.M.,
    5. Harrison K.,
    6. Jones J.D.G.
    (1996). The tomato Cf-2 disease resistance locus comprises two functional genes encoding leucine-rich repeat proteins. Cell 84, 451–459.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Edwards A.,
    2. Civitello A.,
    3. Hammond H.A.,
    4. Caskey C.T.
    (1991). DNA typing and genetic mapping with trimeric and tetrameric tandem repeats. Am. J. Hum. Genet. 49, 746–756.
    OpenUrlPubMed
  12. ↵
    1. Ellis J.G.,
    2. Lawrence G.J.,
    3. Finnegan E.J.,
    4. Anderson P.A.
    (1995). Contrasting complexity of two rust resistance loci in flax. Proc. Natl. Acad. Sci. USA 92, 4185–4188.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Ellis J.G.,
    2. Lawrence G.J.,
    3. Ayliffe M.,
    4. Anderson P.,
    5. Collins N.,
    6. Finnegan J.,
    7. Frost D.,
    8. Luck J.,
    9. Pryor T.
    (1997). Advances in the molecular genetic analysis of the flax–flax rust interaction. Annu. Rev. Phytopathol. 35, 271–291.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Endo T.,
    2. Ikeo K.,
    3. Gojobori T.
    (1996). Large-scale search for genes upon which positive selection may operate. Mol. Biol. Evol. 13, 685–690.
    OpenUrlAbstract
  15. ↵
    1. Farrara B.,
    2. Illott T.W.,
    3. Michelmore R.W.
    (1987). Genetic analysis of factors for resistance to downy mildew (Bremia lactucae) in lettuce (Lactuca sativa). Plant Pathol. 36, 499–514.
    OpenUrlCrossRef
  16. ↵
    1. Frohman M.A.,
    2. Dush M.K.,
    3. Martin G.R.
    (1988). Rapid production of full-length cDNAs from rare transcripts by amplification using a single gene-specific primer. Proc. Natl. Acad. Sci. USA 85, 8998–9002.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Fu Y.H.,
    2. Kuhl D.P.A.,
    3. Pizzuti A.,
    4. Pieretti M.,
    5. Sutcliffe J.S.,
    6. Richards S.,
    7. Verkerk A.J.M.H.,
    8. Holden J.J.A.,
    9. Fenwick R.G.,
    10. Warren S.T.,
    11. Oostra B.A.,
    12. Nelson D.L.,
    13. Caskey C.T.
    (1991). Variation of the CGG repeat at the fragile-X site results in genetic instability: Resolution of the Sherman paradox. Cell 67, 1047–1058.
    OpenUrlCrossRefPubMed
  18. ↵
    1. Grant M.R.,
    2. Godiard L.,
    3. Straube E.,
    4. Ashfield T.,
    5. Lewald J.,
    6. Sattler A.,
    7. Innes R.W.,
    8. Dangl J.L.
    (1995). Structure of the Arabidopsis RPM1 gene enabling dual specificity disease resistance. Science 269, 843–846.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Hagiwara K.,
    2. Harris C.C.
    (1996). ‘Long distance sequencer’ method: A novel strategy for large DNA sequencing projects. Nucleic Acids Res. 24, 2460–2461.
    OpenUrlFREE Full Text
  20. ↵
    1. Hammond-Kosack K.E.,
    2. Jones J.D.G.
    (1997). Plant disease resistance genes. Annu. Rev. Plant Physiol. Plant Mol. Biol. 48, 575–607.
    OpenUrlCrossRef
  21. ↵
    1. Hebsgaard S.M.,
    2. Korning P.G.,
    3. Tolstrup N.,
    4. Engelbrecht J.,
    5. Rouze P.,
    6. Brunak S.
    (1996). Splice site prediction in Arabidopsis thaliana DNA by combining local and global sequence information. Nucleic Acids Res. 24, 3439–3452.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Hughes A.L.,
    2. Nei M.
    (1988). Pattern of nucleotide substitution at major histocompatability complex class I loci reveals overdominant selection. Nature 335, 167–170.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Hulbert S.H.
    (1997). Structure and evolution of the Rp1 complex conferring rust resistance in maize. Annu. Rev. Phytopathol. 35, 293–310.
    OpenUrlCrossRefPubMed
  24. ↵
    1. Jones A.,
    2. Davies H.M.,
    3. Voelker T.A.
    (1995). Palmitoyl–acyl carrier protein (ACP) thioesterase and the evolutionary origin of plant acyl-ACP thioesterases. Plant Cell 7, 359–371.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Jones D.,
    2. Jones J.D.G.
    (1997). The role of leucine-rich repeat proteins in plant defenses. Adv. Bot. Res. Adv. Plant Pathol. 24, 89–167.
  26. ↵
    1. Kajava A.V.,
    2. Vassart G.,
    3. Wodak S.J.
    (1995). Modeling of the three-dimensional structure of proteins with the typical leucine-rich repeats. Structure 3, 867–877.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Kobe B.,
    2. Deisenhofer J.
    (1994). The leucine-rich repeat: A versatile binding motif. Trends Bio. Sci. 19, 415–421.
  28. ↵
    1. Kobe B.,
    2. Deisenhofer J.
    (1995). A structural basis of the interactions between leucine-rich repeats and protein ligands. Nature 374, 183–186.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Lawrence G.J.,
    2. Finnegan E.J.,
    3. Ayliffe M.A.,
    4. Ellis J.G.
    (1995). The L6 gene for flax rust resistance is related to the Arabidopsis bacterial resistance gene RPS2 and the tobacco viral resistance gene N. Plant Cell 7, 1195–1206.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Li W.H.
    (1993). Unbiased estimation of the rates of synonymous and nonsynonymous substitution. J. Mol. Evol. 36, 96–99.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Li W.H.
    (1997). Molecular Evolution. (Sunderland, MA: Sinauer Associates).
  32. ↵
    1. Meyers B.C.,
    2. Chin D.B.,
    3. Shen K.A.,
    4. Sivaramakrishnan S.,
    5. Lavelle D.O.,
    6. Zhang Z.,
    7. Michelmore R.W.
    (1998). The major resistance gene cluster in lettuce is highly duplicated and spans several megabases. Plant Cell 10, 1817–1832.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    1. Michelmore R.W.
    (1995). Molecular approaches to manipulation of disease resistance genes. Annu. Rev. Phytopathol. 15, 393–427.
    OpenUrlCrossRef
  34. ↵
    1. Milligan S.B.,
    2. Bodeau J.,
    3. Yaghoobi J.,
    4. Kaloshian I.,
    5. Zabel P.,
    6. Williamson V.M.
    (1998). The root knot nematode resistance gene Mi from tomato is a member of the leucine zipper, nucleotide binding, leucine-rich repeat family of plant genes. Plant Cell 10, 1307–1319.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Mindrinos M.,
    2. Katagiri F.,
    3. Yu G.L.,
    4. Ausubel F.M.
    (1994). The A. thaliana disease resistance gene RPS2 encodes a protein containing a nucleotide-binding site and leucine-rich repeats. Cell 78, 1089–1099.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Nei M.,
    2. Gojobori T.
    (1986). Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol. Biol. Evol. 3, 418–426.
    OpenUrlAbstract
  37. ↵
    1. Nei M.,
    2. Gu X.,
    3. Sitnikova T.
    (1997). Evolution by the birth-and-death process in multigene families of the vertebrate immune system. Proc. Natl. Acad. Sci. USA 94, 7799–7806.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Okubara P.A.,
    2. Arroyo-Garcia R.,
    3. Shen K.A.,
    4. Mazier M.,
    5. Kim S.J.,
    6. Yang C.-H.,
    7. Michelmore R.W.
    (1997). A transgenic mutant of Lactuca sativa (lettuce) with a T-DNA tightly linked to loss of downy mildew resistance. Mol. Plant-Microbe Interact. 10, 970–977.
    OpenUrlCrossRefPubMed
  39. ↵
    1. Ori N.,
    2. Eshed Y.,
    3. Paran I.,
    4. Presting G.,
    5. Aviv D.,
    6. Tanksley S.,
    7. Zamir D.,
    8. Fluhr R.
    (1997). The I2C family from the wilt disease resistance locus I2 belongs to the nucleotide binding, leucine-rich repeat superfamily of plant resistance genes. Plant Cell 9, 521–532.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Parker J.E.,
    2. Coleman M.J.,
    3. Szabò V.,
    4. Frost L.N.,
    5. Schmidt R.,
    6. Van der Biezen E.A.,
    7. Moores T.,
    8. Dean C.,
    9. Daniels M.J.,
    10. Jones J.D.G.
    (1997). The Arabidopsis downy mildew resistance gene RPP5 shares similarity to the Toll and interleukin-1 receptors with N and L6. Plant Cell 9, 879–894.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Parniske M.,
    2. Hammond-Kosack K.E.,
    3. Golstein C.,
    4. Thomas C.M.,
    5. Jones D.A.,
    6. Harrison K.,
    7. Wulff B.B.H.,
    8. 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, 821–832.
    OpenUrlCrossRefPubMed
  42. ↵
    1. Pryor T.
    (1987). The origin and structure of fungal disease resistance in plants. Trends Genet. 3, 157–161.
  43. ↵
    1. Richter T.E.,
    2. Pryor T.J.,
    3. Bennetzen J.L.,
    4. Hulbert S.H.
    (1995). New rust resistance specificities associated with recombination in the Rp1 complex in maize. Genetics 141, 373–381.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    1. Salmeron J.M.,
    2. Oldroyd G.E.D.,
    3. Rommens C.M.T.,
    4. Scofield S.R.,
    5. Kim H.-S.,
    6. Lavelle D.T.,
    7. Dahlbeck D.,
    8. Staskawicz B.J.
    (1996). Tomato Prf is a member of the leucine-rich repeat class of plant disease resistance genes and lies embedded within the Pto kinase gene cluster. Cell 86, 123–133.
    OpenUrlCrossRefPubMed
  45. ↵
    1. Sambrook J.,
    2. Fritisch E.F.,
    3. Maniatis T.
    (1989). Molecular Cloning: A Laboratory Manual, 2nd ed. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press).
  46. ↵
    1. SanMiguel P.,
    2. Tikhonov A.,
    3. Jin Y.K.,
    4. Motchoulskaia N.,
    5. Zakharov D.,
    6. Melake-Berhan A.,
    7. Springer P.S.,
    8. Edwards K.J.,
    9. Lee M.,
    10. Avramova Z.,
    11. Bennetzen J.L.
    (1996). Nested retrotransposons in the intergenic regions of the maize genome. Science 274, 765–768.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    1. Shen K.A.,
    2. Meyers B.C.,
    3. Islam-Faridi N.,
    4. Stelly D.M.,
    5. Michelmore R.W.
    (1998). Resistance gene candidates identified using PCR with degenerate oligonucleotide primers map to resistance gene clusters in lettuce. Mol. Plant-Microbe Interact. 11, 815–823.
    OpenUrlCrossRefPubMed
  48. ↵
    1. Shepherd K.W.,
    2. Mayo G.M.E.
    (1972). Genes conferring specific plant disease resistance. Science 175, 375–380.
    OpenUrlAbstract/FREE Full Text
  49. ↵
    1. Song W.-Y.,
    2. Pi L.-Y.,
    3. Wang G.-L.,
    4. Gardner J.,
    5. Holsten T.,
    6. Ronald P.C.
    (1997). Evolution of the rice Xa21 disease resistance gene family. Plant Cell 9, 1279–1287.
    OpenUrlAbstract/FREE Full Text
  50. ↵
    1. Staskawicz B.J.,
    2. Ausubel F.M.,
    3. Baker B.J.,
    4. Ellis J.G.,
    5. Jones J.D.G.
    (1995). Molecular genetics of plant disease resistance. Science 268, 661–667.
    OpenUrlAbstract/FREE Full Text
  51. ↵
    1. Sudupak M.A.,
    2. Bennetzen J.L.,
    3. Hulbert S.H.
    (1993). Unequal exchange and meiotic instability of disease-resistance genes in the Rp1 region of maize. Genetics 133, 119–125.
    OpenUrlAbstract/FREE Full Text
  52. ↵
    1. Tanaka T.,
    2. Nei M.
    (1989). Positive Darwinian selection observed at the variable region genes of immunoglobulins. Mol. Biol. Evol. 6, 447–459.
    OpenUrlAbstract
  53. ↵
    1. Thomas C.M.,
    2. Jones D.A.,
    3. Parniske M.,
    4. Harrison K.,
    5. Balint-Kurti P.J.,
    6. Hatzixanthis K.,
    7. Jones J.D.G.
    (1997). Characterization of the tomato Cf-4 gene for resistance to Cladosporium fulvum identifies sequences that determine recognitional specificity in Cf-4 and Cf-9. Plant Cell 9, 2209–2224.
    OpenUrlAbstract/FREE Full Text
  54. ↵
    1. Traut T.W.
    (1994). The functions and consensus motifs of nine types of peptide segments that form different types of nucleotide binding-sites. Eur. J. Biochem. 222, 9–19.
    OpenUrlPubMed
  55. ↵
    1. Wang G.-L.,
    2. Ruan D.-L.,
    3. Song W.-Y.,
    4. Sideris S.,
    5. Chen L.,
    6. Pi L.-Y.,
    7. Zhang S.,
    8. Zhang Z.,
    9. Fauquet C.,
    10. Gaut B.S.,
    11. Whalen M.C.,
    12. Ronald P.C.
    (1998). Xa21D encodes a receptor-like molecule with a leucine-rich repeat domain that determines race-specific recognition and is subject to adaptive evolution. Plant Cell 10, 765–779.
    OpenUrlAbstract/FREE Full Text
  56. ↵
    1. Wessler S.R.,
    2. Bureau T.E.,
    3. White S.E.
    (1995). LTR-retrotransposons and MITEs: Important players in the evolution of plant genomes. Curr. Opin. Genet. Dev. 5, 814–821.
    OpenUrlCrossRefPubMed
  57. ↵
    1. Williams C.E,
    2. Wang B.,
    3. Holsten T.E.,
    4. Scambray J.,
    5. Dasilva F.D.G.,
    6. Ronald P.C.
    (1996). Markers for selection of the rice Xa21 disease resistance gene. Theor. Appl. Genet. 93, 1119–1122.
    OpenUrlCrossRefPubMed
  58. ↵
    1. Yoshimura S.,
    2. Yamanouchi U.,
    3. Katayose Y.,
    4. Toki S.,
    5. Wang Z.X.,
    6. Kono I.,
    7. Kurata N.,
    8. Yano M.,
    9. Iwata N.,
    10. Sasaki T.
    (1998). Expression of Xa1, a bacterial blight-resistance gene in rice, is induced by bacterial inoculation. Proc. Natl. Acad. Sci. USA 95, 1663–1668.
    OpenUrlAbstract/FREE Full Text
  59. ↵
    1. Zhang Z.,
    2. Khumar S.,
    3. Nei M.
    (1997). Small-sample tests of episodic adaptive evolution: A case study of primate lysozymes. Mol. Biol. Evol. 14, 1335–1338.
    OpenUrlCrossRefPubMed
Back to top

Table of Contents

Print
Download PDF
Email Article
Citation Tools
Request Permissions
Receptor-like Genes in the Major Resistance Locus of Lettuce Are Subject to Divergent Selection
Blake C. Meyers, Katherine A. Shen, Pejman Rohani, Brandon S. Gaut, Richard W. Michelmore
The Plant Cell Nov 1998, 10 (11) 1833-1846; DOI: 10.1105/tpc.10.11.1833
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

In this issue

View this article with LENS