About Us | Help Videos | Contact Us | Subscriptions
 

The Plant Genome - Article

 

 

This article in TPG

  1. Vol. 4 No. 1, p. 55-64
    OPEN ACCESS
     
    Received: Nov 11, 2010


 View
 Download
 Alerts
 Permissions
 Share

doi:10.3835/plantgenome2010.11.0024

The Rsv3 Locus Conferring Resistance to Soybean Mosaic Virus is Associated with a Cluster of Coiled-Coil Nucleotide-Binding Leucine-Rich Repeat Genes

  1. Su Jeoung Suh,
  2. Brian C. Bowman,
  3. Namhee Jeong,
  4. Kiwoung Yang,
  5. Christin Kastl,
  6. Sue A. Tolin,
  7. M.A. Saghai Maroof and
  8. Soon-Chun Jeong 
  1. S.J. Suh, N. Jeong, K. Yang, and S.C. Jeong, Bio-Evaluation Center, Korea Research Institute of Bioscience and Biotechnology, Cheongwon, Chungbuk 363-883, Republic of Korea; B.C. Bowman, C. Kastl, and M.A. Saghai Maroof, Dep. of Crop & Soil Environmental Sciences, Virginia Tech, Blacksburg, VA, 20461; S.A. Tolin, Dep. of Plant Pathology, Physiology & Weed Science, Virginia Tech, Blacksburg, VA 24061; Su Jeong Suh and Brian C. Bowman contributed equally to this work

Abstract

The Soybean mosaic virus (SMV) resistance locus, Rsv3, previously mapped between markers A519F/R and M3Satt in the soybean molecular linkage group B2 (chromosome 14), has been characterized by examination of the soybean genome sequence. The 154 kbp interval encompassing Rsv3 contains a family of closely related coiled-coil nucleotide-binding leucine-rich repeat (CC-NB-LRR) genes. Tightly linked to this region are additional CC-NB-LRR genes and several leucine-rich repeat receptor-like kinase (LRR-RLK) genes, thereby indicating that members of both multigene families constitute a heterogeneous cluster at the Rsv3 chromosomal region. To further confirm the sequence and genetic map concordance, we developed 16 markers from the genomic sequence including predicted CC-NB-LRR genes and their flanking sequences. Mapping of the resultant markers in three populations showed parallel alignment between the genetic and sequence maps in the Rsv3-containing region. Phylogenetic analysis of five CC-NB-LRR genes including a pseudogene showed they were highly similar to each other and formed a subclade within a CC-NB-LRR gene clade with representatives from several plant families including legume species. These results demonstrate that the Rsv3 locus is associated with this cluster of CC-NB-LRR genes, thereby suggesting that the Rsv3 gene most likely encodes a member of this gene family. In addition, information from this study should facilitate marker-assisted selection and pyramiding of resistance genes.


Abbreviations

    Avr, avirulence; CC, coiled-coil; GI, gene identification; LRR, leucine-rich repeat; LL, ‘L29’ × ‘Lee68’; LS, ‘L29’ × ‘Sowon’; MLG, molecular linkage group; NB, nucleotide-binding; PCR, polymerase chain reaction; R genes; resistance genes; RLK, receptor-like kinase; SMV, Soybean mosaic virus; TIR, Toll/interleukin-1 receptor; ‘TL, Tousan’ 140 × ‘lee68’

In plants, most characterized disease resistance genes (R genes) encode nucleotide-binding leucine-rich repeat (NB-LRR) proteins characterized by nucleotide-binding (NB) and leucine-rich repeat (LRR) structures as well as variable amino- and carboxy-terminal domains (Collier and Moffett, 2009). These genes tend to cluster in short chromosomal regions (Michelmore and Meyers, 1998). On the basis of their N-termini, two subfamilies of NB-LRR resistance proteins are known: the first is characterized by the Toll/interleukin-1 receptor (TIR)-domain homologous to the Drosophila Toll and mammalian interleukin-1 receptors and the second is characterized by a coiled-coil (CC) structure. Truncated versions of NB-LRR genes exist that encode proteins lacking either a domain near the N-terminus of NB or the LRR region or that consist only of a TIR domain. The two subfamilies cluster separately in phylogenetic analysis using their NB domains (McHale et al., 2006).

Resistance genes mediate dominant resistance to pathogens possessing corresponding avirulence (Avr) genes (Collier and Moffett, 2009; Jones and Dangl, 2006). Single dominant R genes typically respond differentially to pathogen strains or races. The typical dominant resistance response is associated with several defense-related events, including rapid induction of reactive oxygen species, phytoalexin accumulation, and activation of salicylic acid biosynthesis and pathogenesis-related genes, which often results in localized necrotic response (Hammond-Kosack and Jones, 1996).

Rsv3, one of three known genes that confer resistance to the Soybean mosaic virus (SMV) in soybean [Glycine max (L.) Merr.], is unlike the well-characterized Rsv1 alleles in terms of the patterns of resistance to seven SMV strain groups (G1–G7 classified on the basis of their virulence; Cho and Goodman, 1979). Various alleles of the Rsv1 locus generally confer extreme resistance to the lower numbered (G1 through G4) strain groups and condition necrotic or mosaic reactions to higher numbered (G5 through G7) groups (Chen et al., 1991). Rsv3 alleles from diverse soybean cultivars including ‘Columbia’, ‘Hardee’, ‘Tousan 140’, and ‘Harosoy’ confer extreme resistance to the higher numbered strain groups (G5 through G7) and condition stem-tip necrosis and/or mosaic symptoms to the lower numbered groups (Tu and Buzzell, 1987; Buzzell and Tu, 1989; Bowers et al., 1992; Gunduz et al., 2002; Ma et al., 2002). Although stem-tip necrosis was proposed as a representative symptom conditioned by Rsv3 in a line derived from Columbia × Harosoy at the time of the first description of this gene (Tu and Buzzell, 1987; Buzzell and Tu, 1989), this symptom has not been observed in ‘L29’, a ‘Williams’ isoline derived from Hardee (Bernard et al., 1991; Gunduz et al., 2000). Pyramiding the Rsv3 gene from L29 with Rsv1 conferred resistance to all strains of SMV, demonstrating the value of this gene for developing durable SMV resistant soybean lines (Saghai Maroof et al., 2008). Rsv1 is associated with a NB-LRR gene cluster on the soybean molecular linkage group (MLG) F (chromosome 13) (Hayes et al., 2004) where multiple disease resistance genes have been identified and there are multiple NB-LRR clusters (Innes et al., 2008; Wawrzynski et al., 2008). Jeong et al. (2002) mapped the Rsv3 locus between markers A519F/R and M3Satt on MLG B2 (chromosome 14). Although the disease responses of the Rsv3 gene against SMV, including extreme resistance and stem-tip necrosis, were typical of those conditioned by NB-LRR genes, the molecular nature of the Rsv3 gene is largely unknown. Interestingly, one end sequence of the restriction fragment length polymorphism marker M1a, which is closely linked to the Rsv3 gene, was reported (Jeong et al., 2002) to contain an LRR consensus sequence highly similar to that of the extracellular LRR domain of resistance genes, Cf-9 and Xa21 (Jones et al., 1994; Song et al., 1995).

In 2008, preliminary soybean whole genome shotgun sequence assembly was released (version “Glyma0”) by the USDOE-Joint Genome Institute Community Sequencing Program (www.phytozome.net/soybean.php [verified 7 Jan. 2011]) and then an improved version Glyma1.0 was released and reported in 2010 (Schmutz et al., 2010). Integration of the soybean sequence and physical maps with the dense genetic marker map would allow the association of mapped phenotypic effectors with the causal DNA sequence (Jackson et al., 2006). In this study, using a sequence-based marker development strategy in three populations, we determined that the Rsv3 gene cosegregates with a cluster of the coiled-coil nucleotide-binding leucine-rich repeat (CC-NB-LRR) resistance genes, which is located in the middle of a heterogeneous cluster containing multiple CC-NB-LRR and leucine-rich repeat receptor-like kinase (LRR-RLK) genes.


Materials and Methods

Plant Genetic Materials and Disease Reactions

A BC3F2 population of 188 individuals from a cross between L29 (Rsv3) and ‘Sowon’ (rsv3) (hereafter referred to as the LS population) was used to investigate the genetic linkage relationship between Rsv3 and molecular markers. This LS population was previously used to develop an Rsv3-linked sequence-based marker, and disease reactions of its F2 individuals were determined by inoculation with the SMV strain G6 (Yu et al., 2005).

To substantiate the genetic relationship between Rsv3 and the microsatellite markers, two additional populations were used: an F2 population of 183 individuals from a cross between L29 (Rsv3) and ‘Lee68’ (rsv3) (hereafter referred to as the LL population) and an F2 population of 61 individuals from a cross between Tousan 140 (Rsv3) and Lee68 (rsv3) (hereafter referred to as the TL population). These LL and TL populations were previously used to locate the Rsv3 gene in the context of soybean molecular linkage groups (Jeong et al., 2002). The LL and TL populations were previously screened for resistance to SMV strain G7, as described by Jeong et al. (2002) and Gunduz et al. (2002). Briefly, SMV G7 cultures were maintained on the soybean cultivar York, which is resistant to SMV G1 but susceptible to SMV G7, to ensure a uniform source of inoculum. Twenty plants for each F2:3 line were inoculated at the unifoliate stage using a carborundum rub method and scored at 14 and 28 d post inoculation. Plants were designated resistant if no symptoms were present and susceptible if mosaic symptoms appeared.

Alignment of Sequences of Rsv3-Linked Markers against the Soybean Whole Genome Sequence

To locate the Rsv3 region in the soybean whole genome sequence, BLASTN searches of sequences of molecular markers that had been mapped near the Rsv3 locus were performed initially against the whole genome shotgun sequence release Glyma0 (http://www.phytozome.net). Sequence analysis was subsequently repeated against Glyma1.0. The markers used for BLASTN searches included A519F/R, M3Satt, M1a, Satt063, Satt560, and Gm-r-Z20a (Jeong et al., 2002; Yu et al., 2005). The predicted gene models from the region delimited by these markers were retrieved from the soybean gene annotation database (accessible at Phytozome v5.0, http://www.phytozome.net, accessed April 2010) for further analysis. Open-reading frame and conserved protein domains were obtained from the Glyma1.0 annotations with the help of the “GBrowse” function of Phytozome (Stein et al., 2002).

Marker Development

To confirm the genetic and physical concordance of the region of the soybean genome sequence corresponding to the predicted Rsv3-residing region, several novel markers were generated from a single NB-LRR gene (Glyma14 g38500.1) and microsatellite repeat sites in the sequence region delimited by A519F/R and M3Satt and from a single LRR-RLK gene (Glyma14 g38650.1) and microsatellite repeat sites flanking the region. For the NB-LRR gene and LRR-RLK gene sequences, primer sets were designed to amplify genomic DNA of the parental soybean lines Sowon and L29 to detect sequence polymorphisms between them (Supplemental Table S1). Polymerase chain reaction (PCR) products were prepared for sequencing by excising a band of expected size from an agarose gel followed by purification by an Accuprep Gel Purification Kit (Bioneer, Daejeon, Korea). When necessary, a given PCR product was subcloned into a plasmid and multiple clones were sequenced. Primers were designed using the Primer3 (http://frodo.wi.mit.edu/primer3/input.htm [verified 7 Jan. 2011]) program. Amplification by means of microsatellite primer sets (Supplemental Table S2) was performed as described by Jeong and Saghai Maroof (2004). Polymerase chain reaction products were resolved by using 3% agarose or 6.5% polyacrylamide gel electrophoresis. Sequence analysis was performed using the BioEdit program (Hall, 1999). Linkage analysis of the markers was performed using MapMaker 3.0b (Lander et al., 1987).

Phylogenetic Analysis

The 1600 NB sequences used in phylogenetic analyses to subdivide NB-LRR proteins from the diverse plant texa into the functionally distinct TIR-domain-containing and CC-domain-containing subfamilies (McHale et al., 2006) were downloaded to serve as a local protein database. BLASTP searches of NB-LRR proteins located at the Rsv3-residing chromosomal region were performed against the local protein database using the “Local Blast” option implemented in BioEdit (Hall, 1999). The NB sequence of NB-LRR proteins near or cosegregating with the Rsv3 locus, the hit sequences, and their close relatives were used to construct a gene tree.

Full-length amino acid sequences of 194 LRR-RLKs and their phylogeny, representing most of the LRR-RLK genes in the Arabidopsis thaliana (L.) Heynh. genome (Gou et al., 2010), were downloaded to serve as a local protein database. TBLASTN searches of LRR-RLK proteins located near the Rsv3 locus were performed as above against this database (Hall, 1999).

For multiple sequence alignment and phylogenetic analysis, protein sequences were analyzed by using ClustalW and the Neighbor-joining and bootstrap methods implemented in MEGA 4 (Kumar et al., 2008). The weighing matrix used for ClustalW alignment was BLOSUM with the penalty of gap opening 10 and gap extension 0.2. The bootstrap consensus trees were inferred from 1000 replicates.


Results

Phytozome Annotation Map

Sequences of molecular markers that have been mapped in three Rsv3-segregating populations (Jeong et al., 2002; Yu et al., 2005) were positioned on soybean chromosome 14 sequence (pseudomolecule) using BLASTN searches against the soybean genome sequence database. The sequential order of the markers determined by genetic maps was concordant to the physical positions of the markers on the soybean chromosome 14 sequence (Fig. 1). Inspection of the soybean gene annotation database revealed that the Rsv3 chromosomal region contains multiple members of two gene families: NB-LRR and LRR-RLK (Fig. 2; see Table 1 for the gene and marker annotation). Therefore, the sequence region (154 kbp) between A519F/R and M3Satt that brackets the Rsv3 locus on the soybean MLG B2 (chromosome 14) (Jeong et al., 2002; Yu et al., 2005) and the surrounding regions were further analyzed by sequence comparison and sequence-based marker development.

Figure 1.
Figure 1.

Genetic and sequence (physical) maps in the vicinity of the Soybean mosaic virus resistance gene, Rsv3, on the soybean chromosome 14 (molecular linkage group B2). Markers were mapped in the BC3F2 population ‘L29’ (Rsv3) × ‘Sowon’ (rsv3) (LS), the F2 population L29 (Rsv3)’ × ‘Lee68’ (rsv3) (LL), and the F2 population ‘Tousan 140’ (Rsv3) × Lee68 (rsv3) (TL). The resultant genetic maps were aligned with each other and then were aligned with a map constructed from the positions of the markers on the sequence of soybean chromosome 14. Values on the left side of the genetic maps are map distances in cM. Values on the left side of the sequence (physical) map are physical distance from Satt063 in kbp. The bar on the left side of the sequence map indicates the Rsv3 locus predicted on the basis of comparison of the current genetic maps. Satt063, A519-derived markers, Rsv3, M3Satt, and S156h are connected by dotted lines to show the Rsv3-containing chromosomal region.

 
Figure 2.
Figure 2.

Sequence map of the soybean chromosome 14 in the vicinity of the soybean Rsv3 gene presumed to be located between molecular markers A519F/R and M3Satt. (A) Sequence map and gene annotations of chromosome 14 between 45,750 and 49,000 kbp positions. Locations of markers are indicated above the chromosome line and locations of nucleotide-binding leucine-rich repeat (NB-LRR) (filled triangle) and leucine-rich repeat receptor-like kinase (LRR-RLK) (open triangle) genes are indicated below the line. Genes other than the NB-LRR and LRR-RLK genes are not presented. (B) Sequence map and gene annotations of the chromosome 14 between A519F/R and M3Satt. The predicted NB-LRR genes are indicated by filled rectangular arrows with orientations (from 5′ to 3′), the predicted LRR-RLK genes by open rectangular arrows, and the other genes by hatched rectangular arrows. Newly developed markers between A519F/R and M3Satt are indicated below the chromosome line. Numbers below the chromosome line are the annotated genes numbered consecutively from A519F/R to M3Satt. The five NB-LRR genes: 2, Glyma14 g38500.1; 3, Glyma14 g38510.1; 4, Glyma14 g38540.1; 5: Glyma14 g38560.1; 8, Glyma14 g38590.1 (pseudogene; highlighted by *). The other six genes: 1, Glyma14 g38490.1; 6, Glyma14 g38570.1; 7, Glyma14 g38580.1; 9, Glyma14 g38600.1; 10, Glyma14 g38610.1; 11, Glyma14 g38620.1.

 

View Full Table | Close Full ViewTable 1.

Gene and marker annotations of soybean chromosome 14 between 45,750,000 bp and 49,000,000 bp.

 
Gene name Marker name Position Gene annotation
Glyma14g36630.1 45988517..45993144 Leucine-rich repeat receptor-like kinase
Glyma14g36660.1 46001094..46004267 Leucine-rich repeat receptor-like kinase
Glyma14g36810.1 46091207..46093272 Leucine-rich repeat receptor-like kinase
Satt063 46705813..46705956
S156v 46726494..46727093
S156w 46772356..46772567
Glyma14g37590.1 46863817..46868269 Leucine-rich repeat receptor-like kinase
Glyma14g37630.1 46908176..46914671 Leucine-rich repeat receptor-like kinase
A516 46983593..46984402
A593 46983608..46984399
GME 47034693..47035079
Glyma14g37860.1 47141015..47143938 Coiled-coil nucleotide-binding leucine-rich repeat protein
Gm-r-Z20a 47214338..47215135
S156t 47253102..47253291
GMC 47466562..47466808
Glyma14g38390.1 47508501..47512747 Leucine-rich repeat receptor-like kinase
Glyma14g38420.1 47522200..47532508 Nucleotidyltransferase, putative
S156l 47531967..47532194
Glyma14g38430.1 47535945..47538954 Expansin 45, endoglucanase-like
S156g 47550957..47551193
Glyma14g38460.1 47566439..47569354 Transcription factor RF2b, putative
S156i 47570884..47571083
Glyma14g38490.1 47615588..47619943 Transcriptional factor B3, putative
A519F/R 47616322..47617670
S156a 47623366..47623586
S156b 47623609..47623789
S156c 47624441..47624659
NB500pro1 47628799..47628967
NB500pro2 47628799..47628967
Glyma14g38500.1 47630519..47633501 Coiled-coil nucleotide-binding leucine-rich repeat protein
Glyma14g38510.1 47648153..47651727 Coiled-coil nucleotide-binding leucine-rich repeat protein
Glyma14g38540.1 47670152..47672833 Coiled-coil nucleotide-binding leucine-rich repeat protein
Glyma14g38560.1 47691826..47695095 Coiled-coil nucleotide-binding leucine-rich repeat protein
Glyma14g38570.1 47701161..47706197 DNA double-strand break repair RAD50 ATPase
S156e 47709881..47710153
Glyma14g38580.1 47721318..47725350 Cinnamate 4-hydroxylase, putative
Glyma14g38590.1 47728270..47730998 Coiled-coil nucleotide-binding leucine-rich repeat protein (partial pseudogene)
BarcSoySSR_14_1417 47738893..47739065
Glyma14g38600.1 47741179..47745790 Translation initiation factor IF5
Glyma14g38610.1 47758657..47760317 AP2 domain transcription factor
Glyma14g38620.1 47764961..47770195 Ubiquitin-conjugating enzyme E2, putative
M3Satt 47769743..47770460
Glyma14g38630.1 47775271..47778947 Leucine-rich repeat receptor-like kinase
M1a 47778362..47779054
S156h 47814988..47815220
Glyma14g38640.1 47800987..47804143 Root phototropism protein, putative
Glyma14g38650.1 47816591..47827883 Leucine-rich repeat receptor-like kinase
RLK650.p1 47823861..47824095
Glyma14g38670.1 47836300..47846561 Leucine-rich repeat receptor-like kinase
Satt560 47849427..47849504
Glyma14g38700.1 47872961..47876633 Coiled-coil nucleotide-binding leucine-rich repeat protein
Glyma14g38740.1 47897628..47899939 Coiled-coil nucleotide-binding leucine-rich repeat protein
Glyma14g39180.1 48293479..48297011 Leucine-rich repeat receptor-like kinase
Glyma14g39290.1 48411264..48415224 Leucine-rich repeat receptor-like kinase
Glyma14g39550.1 48641632..48644930 Leucine-rich repeat receptor-like kinase
Glyma14g39690.1 48721778..48724358 Leucine-rich repeat receptor-like kinase
Gene annotation information was retrieved from the soybean gene annotation database, Glyma1.0 (accessible at Phytozome v5.0, http://www.phytozome.net [verified 13 Jan. 2011]). Only nucleotide-binding leucine-rich repeat (NB-LRR) and leucine-rich repeat receptor-like kinase (LRR-RLK) genes are presented for the chromosomal region outside of the region delimited by S1561 and RLK650.p1.

Sequence Evaluation of the Chromosomal Region between A519F/R and M3Satt through New Marker Development

To further substantiate that the soybean genome sequence region delimited by A519F/R and M3Satt is correctly assembled and corresponds to the predicted Rsv3-residing region, six new markers were developed based on the regional sequence information. Four markers, S156a, S156b, S156c, and S156e, were microsatellite based and two (NB500pro1 and NB500pro2) were designed from the gene model Glyma14 g38500.1 (see below). The locations of the markers were confirmed by mapping in three populations: LS, LL, and TL. The microsatellite markers S156a and S156b cosegregated with Rsv3 in the LS population (Fig. 1). The markers S156a, S156b, and S156c mapped 0.3 cM away from Rsv3 and S156e cosegregated with Rsv3 in the LL population (Fig. 1). In the TL population, markers S156b and S156c cosegregated with Rsv3 (Fig. 1). Marker BARCSOYSSR_14_1417 (hereafter referred to as BS1417), which was in silico identified from the soybean genome sequence (Song et al., 2010), was mapped 0.3 cM away from Rsv3 in the LS population and cosegregated with Rsv3 in the LL population (Fig. 1).

Of the four NB-LRR genes located between A519F/R and M3Satt, Glyma14 g38500.1 was genetically mapped by using two markers generated from its promoter region. A set of primers was designed to PCR amplify the promoter and 5′-end-coding region (Supplemental Table S1). The PCR products amplified from the soybean parental lines using this primer set gave an expected size of 2.3 kbp and their end sequences were aligned, as expected, to the Glyma14 g38500.1 sequence region with greater than 99% similarity. Then, sequences of promoter parts of the PCR products from the parental lines L29 and Sowon were determined and then aligned. The comparison showed several single nucleotide polymorphic sites. Two of the polymorphic sites were used to generate the sequence-based markers NB500pro1 and NB500pro2 (Supplemental Table S2). The markers cosegregated with Rsv3 in the LS population (Fig. 1).

Sequence Evaluation of the Chromosomal Regions Flanking the A519F/R to M3Satt Region through New Marker Development

Alignment of the soybean genome sequence north of the Rsv3-containing chromosomal region delimited by A519F/R and M3Satt was examined through mapping microsatellite markers generated from its corresponding sequence region. Three microsatellite markers, S156g, S156i, and S156l, were generated from the A519F/R-flanking sequence within 100 kbp (Supplemental Table S2). S156g and S156i cosegregated in the LS population with A519F/R, and S156l mapped 0.3 cM away from A519F/R (Fig. 1). S156g was 0.2 cM away from A519F/R in the LL population. Markers S156g and S156i cosegregated with A519 in the TL population. Additional microsatellite markers located over 100 kbp away from A519F/R were developed: GMC, S156t, GME, S156v, and S156w (Supplemental Table S2). All of the additional markers mapped in the three populations at the genetic locations predicted by the soybean genome sequence.

Alignment of the soybean genome sequence south of the Rsv3-containing chromosomal region delimited by A519F/R and M3Satt was examined by mapping microsatellite markers generated from its corresponding sequence region. One microsatellite marker, S156h, generated from the M3Satt-flanking sequence was within 100 kbp. Marker S156h mapped 0.5 cM south of M3Satt in the LS population, cosegregated with Rsv3 in the LL population, and cosegregated with M3Satt in the TL population (Fig. 1). One of the LRR-RLK genes, Glyma14 g38650.1, which is located near the M1a-hit Glyma14 g38630.1 gene, was genetically mapped using a marker generated from one of its intron–exon junction regions. A set of primers was designed to PCR amplify the intron–exon region (Supplemental Table S1). Sequences of the PCR products from L29 and Sowon were aligned, as expected, to the Glyma14 g38650.1 sequence region with greater than 99% similarity. One sequence-based marker RLK650.p1, which was generated from a T/C single nucleotide polymorphism site between L29 and Sowon, cosegregated with S156h and Satt560 in the LS population (Fig. 1).

Construction of a Phylogenetic Tree of NB-LRR Genes

Examination of the list of the gene models for the sequence region between A519F/R and M3Satt on Gm14 indicated that the region contains four full-length NB-LRR genes and one NB-LRR pseudogene, which are members of the disease resistance gene superfamily (Table 1). An additional three NB-LRR genes were observed outside of the A519F/R to M3Satt sequence region (Fig. 2). The NB sequences of the NB-LRR proteins near or cosegregating with the Rsv3 locus, the sequences BLAST-hit against the NB sequence database created by McHale et al. (2006), and their close relatives were used to construct a gene tree. The previously described full-length soybean CC-NB-LRR proteins were also included to determine their relationships with the Rsv3-associated CC-NB-LRR proteins: 3gG2, 5gG3, and 6gG9 associated with the Rsv1 locus on chromosome 13 (Hayes et al., 2004), Rpg1-b encoded by Rpg1-b on chromosome 13 (Ashfield et al., 2004), and Rps1-k-1 and Rps1-k-2 associated with the Rps1-k locus on chromosome 3 (Gao and Bhattacharyya, 2008). Overall amino-acid-sequence identity between the Rsv3-associated NB-LRR proteins and the previously described soybean CC-NB-LRR proteins was less than 25% and amino-acid-sequence identity between their N-terminal domains ranged from 20 to 50%. The NB sequences of the five NB-LRR genes (including the one pseudogene) found in the sequence between A519F/R and M3Satt are highly similar to each other and formed a subclade in the Neighbor-joining tree (Fig. 3). The subclade was designated as the Rsv3-associated NB in Fig. 3. The NB sequences of two (Glyma14 g38700.1 and Glyma14 g38740.1) of the three NB-LRR genes outside of the A519F/R to M3Satt sequence region are sisters to the Rsv3-associated NB subclade. Collectively, the seven NB sequences inside and outside of the A519F/R to M3Satt sequence region formed a well-supported monophyletic group with a bootstrap support of 98%. Branch lengths indicated that the Rsv3-associated NBs are probably the consequence of recent duplications. We defined the clade as the GmCC-NB I (the red box in Fig. 3). The gene model Glyma14 g37860.1 is an outlier out of the three NB-LRR genes outside of the A519F/R to M3Satt sequence region (below the red box in Fig. 3). In BLAST searches, N-terminal and C-terminal parts of the NB domain in Glyma14 g37860.1 best hit, respectively, different groups of NB sequences that belong to two distantly related nonlegume clades, in the phylogenetic tree of McHale et al. (2006). The best-hit sequences (GenBank gene identification [GI] number 16933577 and GenBank GI number 53680944) for each of the two groups were included in the present phylogenetic tree. The six previously described full-length-cloned soybean CC-NB-LRR proteins formed a monophyletic clade, which was defined as the GmCC-NB II (the blue box in Fig. 3). Thus, the results from phylogenetic analysis of NB-LRR genes indicate that the Rsv3-associated CC-NB-LRR genes appear to be members of a novel CC-NB-LRR class that has not been functionally characterized in soybean.

Figure 3.
Figure 3.

Neighbor-joining phylogenetic tree of nucleotide-binding leucine-rich repeat (NB-LRR) genes. The tree was constructed using only the nucleotide-binding (NB) sequences. The NB sequence of NB-LRR proteins near or cosegregating with the Rsv3 locus, the hit sequences and their close relatives retrieved from BLASTP searches against the 1600 NB sequences collected by McHale et al. (2006) with the NB-LRR proteins in the vicinity of the Rsv3 locus, and the full-length-cloned soybean coiled-coil nucleotide-binding leucine-rich repeat (CC-NB-LRR) proteins (3 gG2, 5 gG3, 6 gG9, Rpg1-b, Rps1-k-1, and Rps1-k-2) were used to construct a gene tree. The designations of sequences except the soybean sequences are as used by McHale et al. (2006). The soybean sequences are in bold. The tree was rooted using human APAF1 protein. The scale bar represents five amino acid differences. The boxed regions represent two soybean CC-NB-LRR subclades GmCC-NB I and GmCC-NB II. Medic sati, Medicago sativa L.; Theob caca, Theobroma cacao L.; Citru sine, Citrus sinensis (L.) Osbeck; Gossy barb, Gossypium barbadense L.; Pisum sati, Pisum sativum L.; Ponci trif, Citrus trifoliata L.; Solan phur, Solanum phureja × Solanum stenotomum; Glyci max, Glycine max (L.) Merr.; Homo sapi, Homo sapiens.

 

The Cluster of NB-LRR Genes Cosegregating with Rsv3 is Located in the Middle of an LRR-RLK Gene Cluster

The restriction fragment length polymorphism marker M1a is tightly linked to Rsv3 (TL in Fig. 1), and the end sequence of M1a contains the extracellular LRR domain (Jeong et al., 2002). Because the consensus sequence of the M1a LRR is identical to that of the extracellular LRR resistance genes Cf-9 and Xa21 (Jones et al., 1994; Song et al., 1995) and because the clustering of disease resistance genes has been reported in many plants (Michelmore and Meyers, 1998), it was hypothesized that the Rsv3 region might contain a member of the extracellular LRR class of disease resistance genes (Jeong et al., 2002). The BLASTN search of the M1a end sequence (GenBank accession no. AF348333) against the soybean genome sequence identified significantly several chromosomal regions, including Gm11 (32.6 Mbp position), Gm18 (4.4 Mbp), Gm02 (45.6 Mbp), Gm14 (47.8 Mbp), and Gm20 (23.1 Mbp), with an E value smaller than 1.3e−113. The M1a sequence showed the highest similarity (99.9% identity) to a part of Gm11 and 80.6% similarity to a part of Gm14. The M1a-hit Gm14 sequence was located near the M3Satt locus and outside of the sequence region delimited by M3Satt and A519F/R (Fig. 2A and 2B). The M1a-containing full-length gene (Glyma14 g38630.1; Glyma1.0 release at http://www.phytozome.net [verified 13 Jan. 2011]) is a member of the LRR-RLK gene family. Interestingly, members of this gene family are repeated 13 times in the vicinity of the Rsv3 locus (Fig. 2; Table 1), thereby supporting the previous thought that extracellular LRR domain-containing sequences might be clustered in this region of the chromosome (Jeong et al., 2002). However, none of these genes appears to be located between M3Satt and A519F/R. The predicted full-length amino acid sequences of the 13 LRR-RLK genes were compared using TBLASTN searches against a local protein database, which contains 194 A. thaliana LRR-RLK sequences reported by Gou et al. (2010). All the sequences hit the A. thaliana LRR-RLK sequences with an E value smaller than 3e–20. However, the G. max sequences dispersed into eight subfamilies of the phylogenetic tree constructed using the 194 A. thaliana LRR-RLK sequences (Table 2).


View Full Table | Close Full ViewTable 2.

Comparison of the 13 soybean leucine-rich repeat receptor-like kinase (LRR-RLK) genes in the vicinity of Rsv3 to their Arabidopsis thaliana (L.) Heynh. homologues.

 
Query sequence Subject sequence E value Subfamily
Glyma14g36630 AT5G58300 0 LRR III
Glyma14g36660 AT1G72180 3e–20 LRR XI
Glyma14g36810 AT1G56140 3e–40 LRR VIII-2
Glyma14g37590 AT1G74360 3e–20 LRR X
Glyma14g37630 AT4G18640 e−114 LRR VI
Glyma14g38390 AT4G29450 4e–10 LRR I
Glyma14g38630 AT5G58300 0 LRR III
Glyma14g38650 AT1G06840 0 LRR VIII-1
Glyma14g38670 AT1G06840 0 LRR VIII-1
Glyma14g39180 AT1G56145 6e−71 LRR VIII-2
Glyma14g39290 AT1G66150 0 LRR IX
Glyma14g39550 AT1G48480 0 LRR III
Glyma14g39690 AT1G56120 7e−52 LRR VIII-2
A. thaliana gene that showed the best match in a TBLASTN search against the local protein database, which contains 194 Arabidopsis LRR-RLK sequences reported by Gou et al. (2010), using a predicted soybean LRR-RLK amino acid sequence.
Name of subfamily in the phylogenetic tree constructed using the 194 A. thaliana LRR-RLK sequences, to which the A. thaliana gene homologous to the soybean gene belongs.


Discussion

More than 40 R genes have been functionally characterized over the past two decades, the majority of which belong to the NB-LRR family (Lukasik and Takken, 2009). The NB-LRR genes tend to cluster in many plant genomes (Michelmore and Meyers, 1998). In this study, we showed that a cluster of the four NB-LRR genes is cosegregating with the Rsv3 locus in our three mapping populations segregating for Rsv3. Despite the lack of physical mapping, parallel alignment between the genetic maps (constructed using public and novel markers) and the genome sequence map (constructed by placing the marker sequences on the soybean genome sequence) are strong evidence that the NB-LRR genes or their variants (as Williams 82 is SMV susceptible) are candidate(s) for Rsv3. Furthermore, none of the other types of genes in the sequence region delimited by A519F/R and M3Satt have been reported to be involved in classical disease resistance response mechanisms (Table 1). Although some members of the LRR-RLK gene family have been reported to be disease resistance genes (Parniske and Jones, 1999; Song et al., 1997), our genetic mapping results indicated that these sequences are outside of the sequence region delimited by A519F/R and M3Satt and are unlikely candidate genes for Rsv3.

The Rsv3-residing chromosomal region is of great interest with respect to the evolution of multigene clusters because members of the CC-NB-LRR and LRR-RLK multigene families constitute a heterogeneous cluster. Nucleotide-binding leucine-rich repeat (NB-LRR) or LRR-RLK genes often occur in clusters that consist of several copies of homologous gene sequences arising from a single gene subfamily (simple clusters) or colocalized gene sequences derived from two or more unrelated subfamilies (complex clusters) and may also contain unrelated single genes interspersed between the homologs (Shiu and Bleecker, 2001; Friedman and Baker, 2007). It has been suggested that intergenic unequal crossover and intragenic mispairing contribute to altered gene copy number within the cluster (e.g., Parniske and Jones, 1999; Kuang et al., 2004, 2005; for a review, see Friedman and Baker, 2007). Although it has been reported that a single Prf gene, a member of the NB-LRR superfamily, is embedded within a cluster of five Pto kinase homologs (Salmeron et al., 1996), colocalization of multiple members of both of the two disease resistance gene superfamilies has not been reported to the best of our knowledge. Surprisingly, the members of the two gene superfamilies appear to be interspersed with each other in the Rsv3-residing chromosomal region on Gm14. Our results suggest that the NB-LRR genes in the Rsv3-containing chromosomal region likely arose from a single gene subfamily (Fig. 3) and that the LRR-RLK genes likely arose from eight unrelated subfamilies reported by Gou et al. (2010) (Table 2). Feature(s) of this chromosomal region that resulted in the coevolution of the two gene clusters at the same chromosomal region or interactions between the two gene clusters during the evolution remain unclear. Interestingly, our examination of the duplication blocks at the SoyBase Browser (http://soybase.org [verified 13 Jan. 2011]) indicated that only LRR-RLK genes cluster on the region on chromosome 2 that is homeologous to the Rsv3 region (data not shown). Although our genetic analysis clearly excluded an Rsv3 candidacy of LRR-RLK genes, the possibility of interaction between NB-LRR and LRR-RLK proteins in conferring resistance to SMV at the Rsv3 locus cannot be dismissed in light of the case of the Prf and Pto interaction (Salmeron et al., 1996). Further analyses including crosses and sequence comparison between soybean cultivars and heterologous expression will help resolve these issues.

The LRR domain structure in the NB-LRR genes would lead to the expectation that, in determining recognition specificity, either Avr proteins or recognition cofactors bind to this domain. At the same time, an accumulating body of evidence suggests that the N-terminal domains of NB-LRR proteins also play a role in Avr recognition (reviewed by Collier and Moffett, 2009). Rsv1, one of the three known SMV resistance genes, is associated with the CC-NB-LRR gene cluster on the soybean chromosome 13 (MLG F) (Jeong et al., 2001; Hayes et al., 2004), the NB domains of which belong to the CC-NB-LRR superfamily (Ashfield et al., 2004). Because the findings of this study suggest that Rsv3 may also encode a member of the CC-NB-LRR gene family, which is distantly related to the Rsv1 locus-associated CC-NB-LRR genes, it is hypothesized that the different disease responses of Rsv1 and Rsv3 against a spectrum of SMV strains may be due to the different structures of the CC-NB-LRR genes. It will be interesting to further elucidate what structural difference between the Rsv1-associated and Rsv3-associated CC-NB-LRR genes make the Rsv1 and Rsv3 genes confer different disease responses to the same SMV strain.

Recently the elicitors or pathogenic determinants governing the disease reactions of Rsv1- or Rsv3-genotype soybeans to different SMV strains have been identified. Chimeric clones, constructed by exchanging genomic sequences from virulent and avirulent SMV strains, were used to map the region within the SMV genome that induces a defense response in Rsv1- and Rsv3-containing cultivars. It was shown that the helper component-protease (HC-Pro) and P3 proteins are independently recognized by Rsv1-genotype soybean and elicit the extreme resistance phenotype (Hajimorad et al., 2006; Eggenberger et al., 2008). Similarly, the cytoplasmic inclusion (CI) protein was shown to be the elicitor or pathogenic determinant recognized by Rsv3-genotype soybean (Seo et al., 2009; Zhang et al., 2009) and that a single amino acid substitution was responsible for strains that can avoid Rsv3 recognition to become avirulent (Seo et al., 2009). Thus, two lines of evidence, phylogenetic difference between Rsv1 and Rsv3 shown in this work and distinct viral genomic regions as elicitors and as determinants of pathogenicity, demonstrate the complexity of the interactions in the soybean–SMV pathosystem that are yet to be elucidated.

Alleles of R genes confer different resistance reactions to pathogens (e.g., Jones and Dangl, 2006). For example, when inoculated with SMV strain G1, Rsv1-n, an allele of Rsv1, confers a severe or lethal necrotic reaction that is a typical resistance reaction but is not desirable in a commercial cultivar (Tucker et al., 2009). To remove or to replace these genes in a soybean cultivar, markers located in the middle of the resistance gene clusters, which span over several centiMorgans in many cases, would be essential for marker-assisted selection of desirable line(s) from a breeding population. Among the three known R genes conferring resistance to SMV, several studies have developed molecular markers tightly linked to Rsv1 (e.g., Gore et al., 2002; Shi et al., 2010) and Rsv4 (Hwang et al., 2006; Saghai Maroof et al., 2010). In particular, Saghai Maroof et al. (2010) used the soybean genome sequence to develop numerous Rsv4-linked molecular markers as well as to show that the Rsv4 gene likely belongs to a new class of resistance genes. The molecular markers developed in this study, and prediction of the molecular nature of the Rsv3 gene, should provide additional tools for pyramiding SMV-resistance genes to obtain durable SMV-resistant soybean cultivars and for elucidating the structure and function of the Rsv3 gene.

Acknowledgments

This work was supported by a grant from the BioGreen 21 Project (code no. 20080401034011), Rural Development Administration, the Republic of Korea, and, in part, by the KRIBB Research Initiative Program. The work was also supported in part by the Virginia Agricultural Experiment Station. We thank two anonymous reviewers for their constructive comments on an earlier version of the manuscript.

 

References

Footnotes

  • All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.



Files:

Comments
Be the first to comment.



Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.