Skip to main content
PMC home page
Genetics logoLink to Genetics
. 2008 Dec;180(4):2267–2276. doi: 10.1534/genetics.108.095034

Two Adjacent Nucleotide-Binding Site–Leucine-Rich Repeat Class Genes Are Required to Confer Pikm-Specific Rice Blast Resistance

Ikuo Ashikawa *,1, Nagao Hayashi †, Hiroko Yamane ‡, Hiroyuki Kanamori ‡, Jianzhong Wu §, Takashi Matsumoto §, Kazuko Ono **, Masahiro Yano **
PMCID: PMC2600957  PMID: 18940787

Abstract

The rice blast resistance gene Pikm was cloned by a map-based cloning strategy. High-resolution genetic mapping and sequencing of the gene region in the Pikm-containing cultivar Tsuyuake narrowed down the candidate region to a 131-kb genomic interval. Sequence analysis predicted two adjacently arranged resistance-like genes, Pikm1-TS and Pikm2-TS, within this candidate region. These genes encoded proteins with a nucleotide-binding site (NBS) and leucine-rich repeats (LRRs) and were considered the most probable candidates for Pikm. However, genetic complementation analysis of transgenic lines individually carrying these two genes negated the possibility that either Pikm1-TS or Pikm2-TS alone was Pikm. Instead, it was revealed that transgenic lines carrying both of these genes expressed blast resistance. The results of the complementation analysis and an evaluation of the resistance specificity of the transgenic lines to blast isolates demonstrated that Pikm-specific resistance is conferred by cooperation of Pikm1-TS and Pikm2-TS. Although these two genes are not homologous with each other, they both contain all the conserved motifs necessary for an NBS–LRR class gene to function independently as a resistance gene.

PLANTS protect themselves against a wide variety of pathogens, such as viruses, bacteria, fungi, nematodes, and insects, through resistance (R) genes that recognize avirulence (Avr) genes in the pathogens. A number of R genes have been cloned and their structures characterized (Martin et al. 2003; Deyoung and Innes 2006; Takken et al. 2006). These R genes can be divided into several classes according to their structural features. The majority encode proteins belonging to the NBS–LRR class, which harbor nucleotide-binding site (NBS) and C-terminal leucine-rich repeat (LRR) motifs (Martin et al. 2003). The NBS domain in R proteins contains a number of conserved motifs, such as kinase 1a or P-loop, kinase 2, and kinase 3a, and this domain may affect R protein function through nucleotide binding, hydrolysis, and control of cell death. The LRR domain is generally thought to be the major determinant of recognition specificity for pathogen avirulence factors (Hulbert et al. 2001; Martin et al. 2003; Deyoung and Innes 2006; Takken et al. 2006). Supporting evidence has been provided, for example, by allelic comparisons and domain-swapping experiments between different alleles at the L and P loci of flax (Ellis et al. 1999; Dodds et al. 2001). However, other evidence supports the possibility that regions additional to the LRR are involved in resistance specificity. One example was provided by an analysis of the L6 and L7 genes for flax rust resistance (Ellis et al. 1999): although these two genes have distinct race-specific resistance, they contain identical NBS–LRR-encoding regions and differ only in the N-terminal Toll/interleukin-1 receptor (TIR) domain.

Rice blast, caused by Magnaporthe grisea, is one of the most devastating diseases of rice. There have been a number of studies identifying blast resistance genes, locating them on the rice chromosome, and (in some cases) cloning and characterizing them at the molecular level. Information obtained from these studies has been used, for example, to develop DNA markers to select blast resistant rice lines, to understand the molecular mechanisms underlying the specific host-pathogen recognition revealed by these blast resistance genes, and to learn the genome organization of R gene clusters and the evolution of complex R gene loci. To date, almost 40 rice blast resistance genes have been identified and mapped (Lin et al. 2007). Of these, 8 genes have been cloned: Pib (Wang et al. 1999), Pita (Bryan et al. 2000), Pid2 (Chen et al. 2006), Pi9 (Qu et al. 2006), Pi2 and Pizt (Zhou et al. 2006), Pi36 (Liu et al. 2007), and Pi37 (Lin et al. 2007). With the exception of Pid2, which was reported to encode a receptor-like kinase (Chen et al. 2006), these cloned genes belong to the NBS–LRR class of resistance genes.

Mapping efforts have indicated that many of the blast resistance genes are allelic or closely linked. For example, at the Piz locus on chromosome 6, at least four genes have been identified (Liu et al. 2002; Hayashi et al. 2004) and three of them, Pi9, Pi2, and Pizt, have been cloned (Qu et al. 2006; Zhou et al. 2006). Structural comparisons of these cloned genes have provided information on the DNA region within these genes responsible for determining their distinct resistance specificities. In addition, analysis of sequences at the Pi9 locus in a group of genetically distinct cultivars has revealed the complex and divergent genome organization of this locus (Zhou et al. 2007). Two blast resistance genes, Pita and Pita2, have been located at the Pita locus on chromosome 12 (Kiyosawa 1967; Rybka et al. 1997; Bryan et al. 2000). These two genes are interesting in terms of their resistance specificity: Pita2 has a broader resistance spectrum than Pita. That is, no M. grisea isolate has been found that is avirulent toward Pita but virulent toward Pita2 (Bryan et al. 2000). This resistance spectrum feature of Pita and Pita2 suggests that Pita2 blast specificity is conferred by a combination of Pita and at least one additional resistance gene (Bryan et al. 2000). Pita has been cloned (Bryan et al. 2000), while Pita2 has not been cloned, and it remains to be answered whether this combination hypothesis is valid.

Another example of a major allelic blast-resistance locus is provided by the Pik locus on chromosome 11, where at least five genes, Pik, Pikm, Pikh, Pikp, and Piks, have been identified (Kiyosawa 1968, 1978; Inukai et al. 1994; Hayashi et al. 2006; Li et al. 2007). The genes Pik and Pikm originated in the Chinese japonica cultivars To-To (Kiyosawa 1968) and Hokushi Tami (Kiyosawa 1978), respectively, whereas Pikh and Pikp originated in the indica cultivars Te-tep (Kiyosawa 1978) and Pusur (Kiyosawa 1969a), respectively. Piks is thought to have originated from a Japanese japonica cultivar (Kiyosawa 1969b). With the exception of Piks, the resistance spectra of these genes to distinct blast isolates are similar; in particular, the resistance specificities of Pik and Pikm mostly overlap each other. The relationship of the resistance spectra of these two genes resembles that of Pita and Pita2: Pikm has a broader resistance spectrum than Pik (Kiyosawa and Nomura 1988). Gene cloning and the subsequent structural characterization of the Pik and Pikm genes would provide the basis for understanding the molecular features responsible for this similar, but distinct, resistance specificity of these genes.

Toward the above-mentioned goal, we have embarked on the map-based cloning of Pik and Pikm. In this article, we report the cloning and characterization of the Pikm gene. Precise mapping and subsequent sequencing of the gene region showed that Pikm resided in a highly divergent genome region, where a large deletion and an insertion between the genomes of the resistant cultivar Tsuyuake and the susceptible cultivar Nipponbare were present. Cloning of the gene revealed a unique feature of Pikm: the resistance provided by Pikm is not conferred by a single gene, but rather by a combination of two NBS–LRR class genes, both of which reside adjacently at the Pikm locus.

MATERIALS AND METHODS

Plant materials and PCR primers:

For mapping of Pikm, we used an F2 segregating population derived from a cross between the blast-resistant cultivar Tsuyuake and the blast-susceptible line 99SL44 (Hayashi et al. 2006). The line 99SL44 is a Nipponbare-based chromosome substitution line in which a segment of chromosome 11 containing the Pikm locus is replaced by a corresponding chromosomal segment from an indica cultivar, Kasalath. For complementation analysis, we used the blast-susceptible cultivar Nipponbare as a host cultivar for transformation. To evaluate the resistance specificity of transgenic plants, as differential cultivars we used Tsuyuake (Pikm+), Kanto 51 (Pik+), K60 (Pikp+), and three cultivars, IRBLk-Ka (Pik+), IRBLkm-Ts (Pikm+), and IRBLkh-K3 (Pikh+), from Lijiangxintuanheigu monogenic lines (Tsunematsu et al. 2000).

We used PCR primer pairs to define the candidate genome region of Pikm (85H07S5, k2167, k4731, k3951, and k3952), to check the presence of transgenes in transgenic lines (Gene2TY1 and Gene1TY1), to examine expression of the transferred candidate genes in transgenic lines (RT13 and Gene1TY1), to obtain the expression profiles of the candidate genes in Tsuyuake (RRT5 and RRT17), and to perform 5′- and 3′-rapid amplification of cDNA ends (RACE) (RT30R, RT32R, RT23F, RT31F, RT21F, and RT4R). The sequences of these primers are listed in supplemental Table S1.

Construction of BAC library and sequencing:

Megabase-sized rice DNA was prepared as described by Zhang et al. (1995) from young leaves of Tsuyuake. A bacterial artificial chromosome (BAC) library was constructed by the conventional method, through partial DNA digest by HindIII, size fractionation of high-molecular-weight DNA in pulsed-field gel electrophoresis (CHFF, Bio-Rad Laboratories, Hercules, CA), vector ligation (pIndigo BAC-5, EPICENTRE Biotechnologies, Madison, WI), and transformation of high-molecular-weight DNA into Escherichia coli (DH10B strain). After preparation of the library, which contained 32,706 clones with an average insert size of 153 kb, positive clones covering the Pikm gene region were screened by using DNA markers tightly linked to Pikm. The BAC clones Ts18H12, Ts69H20, and Ts50A3 were selected and their sequences determined by using a shotgun strategy (Messing et al. 1981) as described before (International Rice Geome Sequencing Project 2005).

Sequence annotation and computational analysis of DNA:

The Rice Genome Automated Annotation System (RiceGAAS) (Sakata et al. 2002) was used to analyze genomic sequence data. This system integrates the programs GENSCAN, RiceHMM, FGENESH, and MZEF for finding putative gene regions and the homology-search-analysis programs Blast, HMMER, ProfileScan, and MOTIF for predicting the putative functions of genes. Pairwise comparisons between genomic or protein sequences were performed with the BLAST program (bl2seq) (http://blast.ncbi.nlm.nih.gov/bl2seq/wblast2.cgi) and the CLUSTALW program (http://clustalw.ddbj.nig.ac.jp/top-j.html). The theoretical isoelectric point (pI) and protein molecular weight were computed as described (http://br.expacy.org/tools/pi_tool.html). The coiled-coil structures of proteins were searched for by using COILS (http://www.ch.embnet.org/software/COILS_form.html).

Candidate gene cloning and complementation analysis:

Among genes predicted within the Pikm region from the Tsuyuake genomic sequence, we selected two putative genes, PiKm1-TS and PiKm2-TS, as candidate genes for Pikm. From the Tsuyuake BAC clone TS18H12, we used a high-fidelity Taq polymerase, PrimeStar (Takara, Tokyo) to amplify a 9.1-kb fragment, which contained the 1.9 kb of the Pikm1-TS 5′-untranslated region (UTR), the whole Pikm1-TS coding sequence, and the 0.9-kb Pikm1-TS 3′-UTR, and a 8.1-kb fragment, which contained the 2.1-kb Pikm2-TS 5′-UTR, the whole Pikm2-TS coding region, and the 2.7-kb Pikm2-TS 3′-UTR. The amplified products were individually inserted into the SmaI site of the binary vector pPZP2H-lac (Fuse et al. 2001) to form constructs for transformation. These constructs were validated by comparison of their insert sequences with the sequences of the corresponding regions in the BAC clone TS18H12 (accession no. AB462256).

We transformed the two constructs, each containing the Pikm1-TS or Pikm2-TS region, individually into Agrobacterium strain EHA101 and then infected Nipponbare callus with them by the method of Toki (1997). Primary transgenic plants (T0 plants) regenerated from hygromycin-resistant calluses were grown in an isolated greenhouse. The presence of the transformed DNA fragments in the T0 plants was checked by PCR assay with the Pikm1-TS-specific primer pair Gene2TY1 and the Pikm2-TS-specific primer pair Gene1TY1 (supplemental Table S1). The transgene copy number was evaluated by Southern hybridization analysis with an ECL Direct Labeling and Detection System (GE Healthcare, Buckinghamshire, UK). We bred the T1 progeny through self-pollination of the T0 plants. Some of these T0 and T1 plants were tested for reaction to blast infection. To develop lines harboring both the Pikm1-TS and Pikm-TS2 genes, we selected T1 plants that carried a single copy of either Pikm1-TS or Pikm2-TS and crossed a T1 plant having Pikm1-TS with another T1 plant having Pikm2-TS. Through these crosses, we obtained five distinct F1 plants that harbored both of the candidate genes. We evaluated the blast resistance of F2 plants bred from self-pollination of these F1 plants.

Evaluation of blast resistance:

The blast resistance of the transgenic lines, the parental cultivars, and the differential cultivars was evaluated by the method described by Hayashi et al. (2004). Briefly, we sprayed blast spores suspended in 0.02% Tween 20 onto plants and then placed these plants in a dew chamber for 22 hr at 25°. The plants were then transferred to a greenhouse and grown for 7 days before the disease reaction was examined. For the complementation analysis and gene expression analysis, we used blast isolate Ina 86-137, which is avirulent to the Pikm donor cultivar Tsuyuake but virulent to the transformation-recipient cultivar Nipponbare. Plants of Nipponbare and Tsuyuake were used, respectively, as susceptible and resistant controls. To evaluate the resistance specificity of transgenic plants, a set of five M. grisea isolates that discriminated Pikm from other blast-resistance genes was used.

Expression analysis:

Expression of transformed Pikm1-TS and Pikm2-TS in transgenic lines was examined by a RT–PCR assay using total RNAs isolated from transgenic lines as templates and RT13R (for detecting Pikm1-TS RNA) and Gene1TY1 (for detecting Pikm2-TS RNA) as primer pairs (supplemental Table S1). Total RNA was isolated with an RNeasy plant kit (QIAGEN, Valencia, CA). To completely eliminate DNA comtaminants, we treated isolated RNA with RNase-free DNase I (QIAGEN). First-strand cDNA was synthesized from ∼1 μl of total RNA by using a Superscript First-Strand synthesis system (Invitrogen, Carlsbad, CA), diluted 10 times with water, and used as a template for RT–PCR.

We used quantitative RT–PCR analysis to examine the expression profiles of candidate genes in Tsuyuake. We isolated total RNAs from seedling leaves of Tsuyuake collected 0, 0.5 (i.e., 12 hr), 1, 3, and 5 days after inoculation (DAI) with the blast isolate Ina 86-137. For negative control inoculation, we sprayed 0.02% Tween 20 without blast spores. The primer pairs RRT5 and RRT17 (supplemental Table S1) were used for detecting transcripts from Pikm1-TS and Pikm2-TS. Primers for rice actin were used as positive RT–PCR controls (Wang et al. 1999). Quantitative RT–PCR analysis was done with an ABI 7500 Real Time PCR System using SYBR Premix Ex Taq (Takara). Each experiment was performed at least twice.

Rapid amplification of cDNA ends:

To determine the 5′- and 3′-end sequences of the cDNAs, we performed RACE with a SMART RACE cDNA amplification kit (Clontech, Mountain View, CA). First-strand cDNA was synthesized from total RNA isolated from Tsuyuake. The 5′ RACE product of Pikm1-TS was PCR-amplified by using the synthesized first-strand cDNA as a template and the gene-specific primer RT23F and the universal primer A mix in the kit; this was followed by another PCR with the nested primer RT31F (supplemental Table S1). The 5′ RACE product of Pikm2-TS was amplified with RT30R and the universal primer A mix, followed by another PCR with the nested primer RT32R (supplemental Table S1). The 3′ RACE products of Pikm1-Ts and Pikm2-TS were amplified with the primers RT21F and RT4R, respectively, together with the universal primer A mix. These RACE products were cloned into pCR 2.1 vector (Invitrogen) for sequencing.

RESULTS

Identification of candidate genes for Pikm:

In our previous study, by performing linkage analysis using 2118 F2 segregating plants bred from a cross between Tsuyuake and 99SL44, we preliminarily located the gene Pikm in the close vicinity of the markers k2167 and k4731 (Hayashi et al. 2006). In our present study, to narrow down the candidate region, we developed a number of DNA markers in the genomic region around the above two markers by using Nipponbare genome sequence data (accession no. NC 008404), and we performed linkage analysis by using the same mapping population that we used for our previous study. This analysis delimited the candidate gene region into a 180-kb interval in the Nipponbare genome between the two markers 85H07S5 and k3952. Although six plants with a chromosomal recombination in this interval were found among the F2 plants, these recombinations occurred only around the two ends of this region: at the proximal end four recombinations were detected in a 23-kb interval next to marker 85H07S5, whereas at the distal end two recombinations were detected in a 8.3-kb interval next to marker k3952 (Figure 1A). In contrast, no recombinations were detected in the 160-kb interval flanked by the above two regions. This nonuniform distribution of the recombination sites at this locus hindered our efforts to narrow down the candidate region into a smaller interval. The annotation tools accompanying the RiceGAAS database predicted the presence of 19 putative genes within this region (Figure 1 and supplemental Figure S1). Of these genes, 6 were similar to the disease R genes cloned from various plant species. These 6 genes encoded NBS–LRR-class proteins, and we designated them Pikm1-NP to Pikm6-NP (Figure 1A). The clustering of these putative R genes in this small genome region indicates that Pikm is located within a typical R gene region (Bai et al. 2002).

Figure 1.—

Figure 1.—

Open in a new tab

Physical maps of the Pikm gene region and the structures of the two genes constituting Pikm. (A) Physical maps of the Pikm locus generated by using the Nipponbare (Pikm−) genome sequence (top) and Tsuyuake (Pikm+) sequence (bottom). The candidate gene region is delimited by the marker 85H07S5 from the proximal side and the marker k3951 from the distal side. The numbers above each of the maps represent distances in kilobases. The arrows below each map represent putative genes predicted by the annotation tools in RiceGAAS. The six boldface arrows on the Nipponbare map and the two boldface arrows on the Tsuyuake map represent putative R genes. (B) Structures of Pikm1-TS and Pikm2-TS. Dark-shaded boxes are exons, light-shaded boxes are 5′- or 3′-UTRs, and the lines are introns.

The cultivar Nipponbare is susceptible and hence its genome does not carry the gene Pikm. To identify candidate genes for Pikm, we attempted to obtain sequence data of the gene region from the Pikm-carrying cultivar Tsuyuake. First, we used the markers 85H07s5, k4731, and k3951 as probes to screen the Tsuyuake BAC library for BAC clones containing the Pikm region. We found that three BAC clones, Ts18H12, Ts69H20, and Ts50A13, covered the genome region delimited by the markers 85H07S5 and k3951. Comparison of the sequences of these Tsuyuake BAC clones (accession no. AB462256) with the Nipponbare sequence of the equivalent region revealed that the two cultivars had marked differences in their genome organization at this Pikm locus: the Tsuyuake genome lacked a 127-kb segment present in the Nipponbare genome, whereas next to this deletion there existed a 70-kb segment present only in Tsuyuake, but not in Nipponbare (Figure 1A). Owing to this insertion and deletion, the size of the Pikm candidate region in the Tsuyuake genome was reduced to 131 kb. Analysis of this Tsuyuake sequence by using the gene-finding programs in RiceGAAS predicted 16 genes within the region (supplemental Figure S1). Among these, two putative coding sequences, designated Pikm1-TS and Pikm2-TS, showed homology with the NBS–LRR class resistance genes. They were predicted to reside adjacent to each other, being transcribed in opposite directions. Because there were no other putative genes apart from these 2 that showed high levels of identity with the existing R gene, we considered Pikm1-TS and Pikm2-TS to be the most probable candidates for Pikm.

Blast resistance of transgenic plants harboring either PiKm1-TS or PiKm2-TS:

To examine whether the Pikm1-TS or Pikm2-TS gene conferred Pikm-specific blast resistance, we introduced these two candidate genes individually into the blast susceptible rice cultivar Nipponbare and evaluated the resistance of the transformed plants. We evaluated the blast resistance of some of the T0 plants. From a group of these plants, we randomly selected 12 plants to which Pikm1-TS had been introduced and 13 plants to which PiKm2-TS had been introduced. The presence of the respective transgenes in each of the tested plants was verified by PCR assay with the primers specific to Pikm1-TS and Pikm2-TS. Following inoculation with the blast isolate Ina 86-137, all of the tested plants harboring Pikm1-TS and most of the plants harboring Pikm2-TS (10 of the 13 plants) exhibited spindle-shaped lesions (data not shown), which are typically detected in blast susceptible cultivars, such as Nipponbare. This result did not support the possibility that either of the two candidates was the gene Pikm.

To confirm the above result, we examined the blast resistance of the T1 plants. We used a PCR assay to select those T1 plants that contained the transgenes. We inoculated blast spores onto a total of 21 T1 plants having Pikm1-TS (derived from 4 distinct T0 plants) and onto a total of 86 T1 plants having Pikm2-TS (derived from 11 distinct T0 plants). All of these tested transgenic lines showed no resistance to blast. To examine whether silencing of the introduced transgenes, if it was present, was responsible for this susceptible phenotype of the transgenic lines, we isolated RNA from some plants randomly selected from the tested plants and checked the expression of the integrated genes. The results of RT–PCR indicated that the two genes were actually expressed (data not shown). In light of these results in both the T0 and the T1 plants, we concluded that neither Pikm1-TS nor Pikm2-TS endowed blast resistance when only one of the two was introduced into the susceptible cultivar Nipponbare.

Blast resistance of transgenic plants harboring both Pikm1-TS and Pikm2-TS:

We then examined the possibility that a combination of the two candidate genes conferred Pikm-associated blast resistance. To perform this examination, we crossed T1 plants having Pikm1-TS with those having Pikm2-TS, and we obtained transgenic plants possessing both genes. The resultant F1 plants were self-pollinated. We examined the blast resistance of 106 F2 progeny. On the basis of the presence or absence of the two transgenes, these F2 plants were classified into four groups: 59 plants with both genes, 22 with only Pikm1-TS, 20 with only Pikm2-TS, and 5 with neither. The segregation pattern of these four types of F2 plants indicated that both of the transgenes were independently inherited. Following the inoculation with the blast isolate Ina 86-137, we observed a cosegregation between the presence of both transgenes and the resistant phenotype (Figure 2): all of the transgenic plants that carried both candidate genes were evaluated as resistant, whereas all of the plants that carried only one of the two transgenes, or neither of them, were susceptible. The group of resistant plants consisted mainly of highly resistant plants that exhibited no specks, or only several small brown specks, on their leaves (Figure 2A). The resistant group also included a small number of weakly resistant plants, which had a number of small brown specks or a small number of elliptical lesions. In contrast, like the transgenic lines that had neither of the two genes, plants with only one of the two genes were highly susceptible to blast inoculation. They developed typical spindle-shaped blast lesions on their leaves, and parts of the leaves were killed by the coalescence of lesions. This indicated that Pikm1-TS or Pikm2-TS, when present alone in a transgenic plant, did not confer partial blast resistance. Thus, we suggest that the complete resistance observed in the transgenic lines having both genes is due to cooperation of the two genes, rather than to an additive effect of partial resistance conferred by each of the two genes.

Figure 2.—

Figure 2.—

Open in a new tab

Complementation testing and molecular analysis of the transgenic lines. (A) Reaction to inoculation with the blast isolate Ina 86-137 in Tsuyuake, Nipponbare, and a set of F2 progeny (numbers 1–7) derived from a cross between a T1 plant harboring Pikm1-TS and another T1 plant harboring Pikm2-TS. The isolate Ina 86-137 is incompatible with the Pikm-donor cultivar Tsuyuake and compatible with the transformation-recipient cultivar Nipponbare. (B) Cosegregation of the resistance phenotype with the presence of both the Pikm1-TS gene and the Pikm2-TS gene. The presence of Pikm1-TS and Pikm2-TS in the transgenic lines is shown by the presence of a product from PCR amplification with the primers Gene2TY1 (top) and Gene1TY1 (bottom), respectively. The amplified products were separated through a 2.0% agarose gel. The transgenic lines numbered 3 and 6 contain both Pikm1-TS and Pikm2-TS, those numbered 2 and 4 contain only Pikm1-TS, those numbered 1 and 5 contain only Pikm2-TS, and that numbered 7 contains neither of the transgenes.

Resistance specificity of transgenic plants:

To verify whether transgenic plants that had a combination of Pikm1-TS and Pikm2-TS had the same resistance specificity as Pikm, we inoculated the transgenic lines and differential cultivars with a set of blast isolates that enabled us to distinguish Pikm from other resistance genes. The transgenic lines were susceptible to the isolate 1804-4, indicating that the transgenes belonged to the Pik family and not to another family such as Pib, Piz, or Pita (Table 1). The transgenic lines were susceptible to the isolate H05-72-1 and resistant to the isolates P2-b and Kyu92-22; this indicated that the transgene was not Pik, Pikp, or Pikh. These results provided experimental evidence that the combination of the two genes Pikm1-TS and Pikm2-TS had the same function as the blast resistance gene Pikm.

TABLE 1.

Disease reactions of transgenic plants containing Pikm1-TS and Pikm2-TS and of differential cultivars to Magnaporthe grisea isolates

Cultivar Ina 86-137 Kyu 92-22 1804-4 H05-72-1 P2-b
Nipponbare S S S S S
Transgenic plants R R S S R
Differential cultivara (Pikm) R R S S R
Differential cultivarb (Pik) R S S S R
K60 (Pikp) R S ND ND S
IRBLkh-K3 (Pikh) ND R S R ND
Open in a new tab

ND, resistance was not evaluated in this study; S, susceptible; R, resistant.

a

Tsuyuake and/or IRRLkm-Ts was used.

b

Kanto51 and/or IRRLk-Ka was used.

Structure of Pikm1-TS and Pikm2-TS:

To obtain the full-length cDNA sequences of Pikm1-TS and Pikm2-TS, we sequenced RACE products from the 5′- and 3′-ends of the transcribed regions of each gene and the overlapping intermediate RT–PCR fragments. The size and structure of both genes were determined by comparing the obtained full-length cDNA sequences with the genomic DNA sequences. Pikm1-TS (accession no. AB462324) contained a 3432-bp coding region, interrupted by two introns (119 and 2769 bp long) and flanked by a 62-bp 5′-UTR and a 190-bp 3′-UTR (Figure 1B). As shown in the multiple DNA sequence alignment of Pikm1-TS and Pikm2-TS (supplemental Figure S2), the structure of Pikm2-TS (accession no. 462325) was not homologous with that of Pikm1-TS: this second gene contained a 3066-bp coding region, interrupted by one intron (163 bp long) and flanked by a 109-bp 5′-UTR and a 283-bp 3′-UTR (Figure 1B).

Protein translations of the cDNA sequences revealed that Pikm1-TS encoded a 1143-residue NBS–LRR-class polypeptide (Figure 3) with a molecular weight of 126.85 kDa and a pI of 5.96. The NBS domain contained three sequence motifs characteristic to this domain, GLPGGGKTTI (beginning at residue 291), KKYLIVIDDI (beginning at residue 377), and GGRIIMTTRL (beginning at residue 405), corresponding to kinase 1a (P-loop), kinase 2, and kinase 3a, respectively. In the N-terminal region (residues 1–240), there exists a non-TIR (nT) motif of rice NBS–LRR genes (Bai et al. 2002), GLIAEVRELSYDLDDAVDDF (beginning at residue 74). A COILS analysis predicted the possible existence of a CC domain (P = 0.5) between residues 151 and 165. The interval between the kinase 1a and nT motifs (198 residues) was almost 70 residues larger than the interval of a typical rice NBS–LRR protein (∼130 residues), suggesting that a polypeptide sequence was inserted in this region of the Pikm1-TS-encoded protein. The C-terminal domain contained ∼16 imperfect LRR repeats, which were composed of ∼14% leucine. This LRR region contained six repeat units that showed a good match to the consensus sequence, LxxLxxLxLxxTxxLxxLPxxLxx. Following these LRR repeats was a 103-residue non-LRR region. By comparison, Pikm2-TS encoded a smaller polypeptide, a 1021-residue NBS–LRR-class protein (Figure 4) with a molecular weight of 114.57 kDa and a pI of 8.64. The NBS domain of this protein also contained the three motifs, kinase 1a (beginning at residue 212), kinase 2 (beginning at residue 330), and kinase 3a (beginning at residue 359). Upstream of the NBS domain, an nT motif, WMKQIRDIAYDVECIDDF (beginning at residue 70), was also found, but a CC domain was not predicted. The C-terminal region was smaller than that of the Pikm1-TS product and contained 13 imperfect LRR repeats (residues 610–960) composed of ∼17% leucine, but it did not contain a non-LRR region. The LRR repeats matching the consensus sequence were not as clearly discernible as in Pikm1-TS.

Figure 3.—

Figure 3.—

Open in a new tab

Deduced amino acid sequence of the Pikm1-TS protein. The three conserved motifs, i.e., kinase 1a, kinase 2, and kinase 3a, forming the NBS domain and the nT motif upstream of the NBS domain are in boldface type. The LRR repeat region and non-LRR region are shown detached from the rest of the sequence. The LRR region contains six repeat units that show a good match to the consensus sequence Lxx Lxx Lx LxxTxLxxLPxxLxx. In this LRR region, leucines and other residues matching the consensus sequence are shown in boldface type.

Figure 4.—

Figure 4.—

Open in a new tab

Deduced amino acid sequence of the Pikm2-TS protein. The three conserved motifs (i.e., kinase 1a, kinase 2, and kinase 3a) forming the NBS domain and the nT motif upstream of the NBS domain are in boldface type. The C-terminal LRR is shown detached from the rest of sequence. In the LRR region, leucines and other residues matching the consensus sequence LxxLxxLxLxxTxLxxLPxxLxx are shown in boldface type.

The susceptible cultivar Nipponbare contained the alleles of Pikm1-TS and Pikm2-TS, i.e., Pikm5-NP (accession no. ABA95385) and Pikm6-NP (accession no. NP_001068487), respectively. Pikm5-NP and Pikm6-NP, respectively, encode 1125- and 1044-residue NBS–LRR-class proteins. The overall gene product architectures of Pikm5-NP and Pikm6-NP resembled those of the products of the respective Tsuyuake alleles (supplemental Figures S3 and S4): the products of both of the Nipponbare alleles contained an nT motif in the N-terminal region, three conserved motifs in the NBS region, and imperfect LRR repeats, whereas the Pikm5-NP product contains a C-terminal non-LRR region but Pikm6-NP does not. In contrast to these homologies revealed in the overall molecular design, at the amino-acid-sequence level these two proteins were markedly different from the products of their Tsuyuake alleles, especially in the N-terminal and the NBS regions (supplemental Figures S3 and S4). Through alignment of the protein sequences of these allelic genes by using the BLASTP program, the amino-acid-sequence identities between Pikm5-NP and Pikm1-TS and between Pikm6-NP and Pikm2-TS were calculated to be as low as 59 and 76%, respectively.

Expression of Pikm1-TS and Pikm2-TS in Tsuyuake:

To examine whether Pikm1-TS and Pikm2-TS were expressed in response to the challenge of blast infection, we performed quantitative RT–PCR to specifically detect each of the transcripts from the two genes in Tsuyuake at five time points (0, 0.5, 1, 3, and 5 days) after blast inoculation. Expression of both Pikm1-TS and Pikm2-TS was detected even at 0 DAI, indicating that both of the genes were constitutively expressed. Following blast inoculation, expression of Pikm1-TS increased from 0.5 to 3 DAI, and then declined toward the original level by 5 DAI (Figure 5). This induction of expression in Pikm1-TS was not detected in the negative control inoculations, indicating that the observed induction of Pikm1-TS expression was due to the challenge of blast infection. In contrast, although expression of Pikm2-TS appeared to increase slightly from 0.5 to 3 DAI, the extent of the induction was small (Figure 5).

Figure 5.—

Figure 5.—

Open in a new tab

Expression profiles of Pikm1-TS and Pikm2-TS in the cultivar Tsuyuake inoculated with the blast isolate Ina 86-137. Expression of these genes was examined 0, 0.5 (i.e., 12 hr), 1, 3, and 5 days after inoculation, together with 5 days after inoculation without blast spores (negative control inoculation).

DISCUSSION

We cloned and characterized the rice blast resistance gene Pikm. High-resolution mapping of this gene, followed by complementation analysis and evaluation of the resistance specificity of the transgenic plants, demonstrated that the blast resistance of Pikm is conferred by a combination of two genes, Pikm1-TS and Pikm2-TS. Both genes encode non-TIR NBS–LRR-class proteins. Although the products of each of the two genes contain all of the necessary structural motifs to independently function as NBS–LRR-class R proteins, the presence of the two genes in a plant is indispensable for the exhibition of blast resistance.

Pikm1-TS and Pikm2-TS resided adjacently at the Pikm locus as a cluster (Figure 1), but their structures differed. First, they differed in the number and position of introns: both Pikm1-TS and Pikm2-TS contained an intron at the immediate N-terminal side of the sequence encoding the kinase 2 motif in the NBS domain. In addition to this, Pikm1-TS contained another intron upstream of the sequence encoding the NBS domain. Second, the Pikm1-TS product contained a C-terminal non-LRR region, whereas that of Pikm2-TS did not. Third, Pikm1-TS contained fairly well-conserved repeat units matching a consensus sequence in its LRR domain, whereas Pikm2-TS did not. All of the above-mentioned structural differences indicate that these two genes did not evolve from one another by a simple duplication event.

Sequence comparison showed a marked difference in the genome organization at the Pikm locus between Tsuyuake and Nipponbare: Tsuyuake lacked the 127-kb segment present in the Nipponbare genome, whereas it contained a 70-kb segment that was absent in Nipponbare. Thus, of the 131-kb Pikm candidate region in the Tsuyuake genome, only 61-kb stretches had corresponding DNA segments in the Nipponbare genome. High levels of divergence in genome structure has been reported at many R gene loci, such as the Pi9 locus in rice (Zhou et al. 2007), the Rp1 locus in maize (Smith et al. 2004), and the Rps2 (Mauricio et al. 2003), Rpp13 (Rose et al. 2004), and Rpp5 (Noel et al. 1999) loci in Arabidopsis. It has been proposed that the dynamics of genome organization and high polymorphism characteristic of the R gene loci may be essential for generating new resistance specificities in the process of evolution (Meyers et al. 2005). In this light, a comparative study of the genome structures at the Pikm locus in cultivars that have other Pik family genes would provide insight into the origin and evolution of allelic resistance genes at this locus.

Among the R genes characterized to date, we consider Pikm to be unique, because two independent NBS–LRR class genes are required to confer Pikm-specific resistance. To our knowledge, two examples have been reported in which expression of resistance requires two independent genetic determinants. The first example is the gene RPP2 in Arabidopsis (Sinapidou et al. 2004). Resistance to Cala2, specified by the RPP2 locus in Arabidopsis Col-0, requires two genes, RPP2A and RPP2B, both of which reside at the RPP2 locus and are NBS–LRR-class genes. Although a physical interaction between the products of these two genes was proposed (Sinapidou et al. 2004) for the resistance response of RPP2, no experimental evidence supporting this idea has yet been reported. These two genes contain differences in their structures: the RPP2B product contains all necessary conserved motifs to function as a NBS–LRR protein, whereas the RPP2A product lacks the kinase 1a motif in its NBS domain and its LRR domain appears too small; thus, RPP2A may not independently function as an NBS–LRR gene. This is in contrast to the case of the Pikm genes; although the structures of Pikm1-TS and Pikm2-TS were dissimilar, the products of both genes contained all of the necessary molecular motifs required for NBS–LRR proteins to function as R proteins. The second example is the Pto-resistance complex in tomato. Although the Pto gene complements resistance to the bacterium Pseudomonas syringae in susceptible tomato (Martin et al. 1993), another gene, Prf, which lies within the Pto cluster, is essential for effective resistance (Salmeron et al. 1996). Pto encodes a serine-threonine protein kinase, whereas Prf is an NBS–LRR-class gene. Although a number of models have been proposed for the possible interaction of those two genes, the epistatic relationship between the two has not been resolved. Recently, Mucyn et al. (2006) showed that these two genes constitutively interact with each other. They proposed that Prf acts as the transduction module and Pto acts in the complex as a regulatory subunit of Prf.

So far, the number of R genes identified that require more than two genetic determinants for their function is very limited, and information on the mechanism underlying the cooperative function of two such distinct genes is also limited. Cloning and characterization of other examples of such classes of resistance genes and follow-up studies that investigate how the products of the two genes interact would give us a better understanding of the function of these R genes.

The information we obtained on Pikm may help us to understand the resistance specificity of the genes belonging to the Pik family. As described earlier, the resistance spectra of most of the Pik family genes are similar. Pik and Pikm, especially, show a similar resistance response to distinct blast isolates. The resistance spectrum of Pikm is known to be broader than that of Pik, and no blast isolates that are incompatible with Pik but compatible with Pikm have been reported. This suggests that Pikm is composed of two, or more than two, genetic factors (Kiyosawa and Nomura 1988): one is Pik and the other is responsible for the resistance not covered by Pik, and the additive effect of these genes accounts for the broader resistance spectrum of Pikm. Our study showed that Pikm is not composed of a single genetic determinant, but that two genes are involved in its function. However, neither of the two seems to function as Pik, because transgenic lines that contained one of the two genes were susceptible to the blast isolate Ina 86-137, with which Pik is known to be incompatible. This result contradicts Kiyosawa's hypothesis. To understand the molecular mechanism underlying the difference in the resistance spectra of Pik and Pikm, cloning and characterization of Pik are needed as a next step.

Acknowledgments

The authors thank I. Ando and H. Satoh of the National Institute of Crop Science, Japan, for their kind technical advice on blast resistance evaluation; A. Shomura of the Institute of the Society for Techno-Innovation of Agriculture, Forestry, and Fisheries, Japan, for introducing the plasmids used for transformation into Agrobacterium; and K. Hayashi of the National Agricultural Research Center, Japan, for her cooperation in the genetic mapping of Pikm. This work was supported by a grant from the Ministry of Agriculture, Forestry and Fisheries of Japan (Green Technology Project QT-4005).

Sequence data from this article have been deposited with the EMBL/GenBank/DDBJ Data Libraries under accession nos. AB462256, AB462324, and AB462325.

References

  1. Bai, J., L. A. Pennill, J. Ning, S. W. Lee, J. Ramalingam et al., 2002. Diversity in nucleotide binding site-leucine-rich repeat genes in cereals. Genome Res. 12 1871–1884. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bryan, G. T., K. S. Wu, L. Farrall, Y. Jia, H. P. Hershey et al., 2000. A single amino acid difference distinguishes resistant and susceptible alleles of the rice blast resistance gene Pi-ta. Plant Cell 12 2033–2045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chen, X., J. Shang, D. Chen, C. Lei, Y. Zou et al., 2006. A B-lectin receptor kinase gene conferring rice blast resistance. Plant J. 46 794–804. [DOI] [PubMed] [Google Scholar]
  4. Deyoung, B. J., and R. W. Innes, 2006. Plant NBS-LRR proteins in pathogen sensing and host defense. Nat. Immunol. 7 1243–1249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Dodds, P. N., G. J. Lawrence and J. G. Ellis, 2001. Six amino acid changes confined to the leucine-rich repeat beta-strand/beta-turn motif determine the difference between the P and P2 rust resistance specificities in flax. Plant Cell 13 163–178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Ellis, J. G., G. J. Lawrence, J. E. Luck and P. N. Dodds, 1999. Identification of regions in alleles of the flax rust resistance gene L that determine differences in gene-for-gene specificity. Plant Cell 11 495–506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Fuse, T., T. Sasaki and M. Yano, 2001. Ti-plasmid vectors useful for functional analysis of rice genes. Plant Biotechnol. 18 219–222. [Google Scholar]
  8. Hayashi, K., N. Hashimoto, M. Daigen and I. Ashikawa, 2004. Development of PCR-based SNP markers for rice blast resistance genes at the Piz locus. Theor. Appl. Genet. 108 1212–1220. [DOI] [PubMed] [Google Scholar]
  9. Hayashi, K., H. Yoshida and I. Ashikawa, 2006. Development of PCR-based allele specific and InDel marker sets for nine rice blast resistance genes. Theor. Appl. Genet. 113 251–260. [DOI] [PubMed] [Google Scholar]
  10. Hulbert, S. H., C. A. Webb, S. M. Smith and Q. Sun, 2001. Resistance gene complexes: evolution and utilization. Annu. Rev. Phytopathol. 39 285–312. [DOI] [PubMed] [Google Scholar]
  11. International Rice Genome Sequencing Progect, 2005. The map-based sequence of the rice genome. Nature 436 793–800. [DOI] [PubMed] [Google Scholar]
  12. Inukai, T., R. J. Nelson, R. S. Zeigler, S. Sarkarung, D. J. Mackill et al., 1994. Allelism of blast resistance genes in near-isogenic lines of rice. Phytopathology 84 1278–1283. [Google Scholar]
  13. Kiyosawa, S., 1967. Inheritance of resistance of the rice variety Pi No. 4 to blast. Jpn. J. Breed. 17 165–172. [Google Scholar]
  14. Kiyosawa, S., 1968. Inheritance of blast-resistance in some Chinese rice varieties and their derivatives. Jpn. J. Breed. 18 193–204. [Google Scholar]
  15. Kiyosawa, S., 1969. a Inheritance of resistance of rice varieties to a Philippine fungus strain of Pyricularia oryzae. Jpn. J. Breed. 19 61–73. [Google Scholar]
  16. Kiyosawa, S., 1969. b Inheritance of blast-resistance in west Pakistani rice variety. Pusur. Jpn. J. Breed. 19 121–128. [Google Scholar]
  17. Kiyosawa, S., 1978. Identification of blast-resistance genes in some rice varieties. Jpn. J. Breed. 28 287–296. [Google Scholar]
  18. Kiyosawa, S, and K. Nomura, 1988. Disease resistance in crop plants and mutational breeding. Gamma Field Symposia No. 27. Institute of Radiation Breeding, NIAR, MAFF, Ibaraki, Japan.
  19. Li, L. Y., L. Wang, J. X. Jing, Z. Q. Li, F. Lin et al., 2007. The Pik-m gene, conferring stable resistance to isolates of Magnaporthe oryzae, was finely mapped in a crossover-cold region on rice chromosome 11. Mol. Breed. 20 179–188. [Google Scholar]
  20. Lin, F., S. Chen, Z. Que, L. Wang, X. Liu et al., 2007. The blast resistance gene Pi37 encodes a nucleotide binding site-leucine-rich repeat protein and is a member of a resistance gene cluster on rice chromosome 1. Genetics 177 1871–1880. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Liu, G., G. Lu, L. Zeng and G. L. Wang, 2002. Two broad-spectrum blast resistance genes, Pi9(t) and Pi2(t), are physically linked on rice chromosome 6. Mol. Genet. Genomics 267 472–480. [DOI] [PubMed] [Google Scholar]
  22. Liu, X., F. Lin, L. Wang and Q. Pan, 2007. The in silico map-based cloning of Pi36, a rice coiled-coil-nucleotide-binding site-leucine-rich repeat gene that confers race-specific resistance to the blast fungus. Genetics 176 2541–2549. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Martin, G. B., S. H. Brommonschenkel, J. Chunwongse, A. Frary, M. W. Ganal et al., 1993. Map-based cloning of a protein kinase gene conferring disease resistance in tomato. Science 262 1432–1436. [DOI] [PubMed] [Google Scholar]
  24. Martin, G. B., A. J. Bogdanove and G. Sessa, 2003. Understanding the function of plant disease resistance proteins. Annu. Rev. Plant Biol. 54 23–61. [DOI] [PubMed] [Google Scholar]
  25. Mauricio, R., E. A. Stahl, T. Korves, D. Tian, M. Kreitman et al., 2003. Natural selection for polymorphism in the disease resistance gene Rps2 of Arabidopsis thaliana. Genetics 163 735–746. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Messing, J., R. Crea and P. H. Seebung, 1981. A system for shotgun DNA sequencing. Nucleic Acids Res. 9 309–321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Meyers, B. C., S. Kaushik and R. S. Nandety, 2005. Evolving disease resistance genes. Curr. Opin. Plant Biol. 8 129–134. [DOI] [PubMed] [Google Scholar]
  28. Mucyn, T. S., A. Clemente, V. M. E. Andriotis, A. L. Balmuth, G. E. D. Oldroyd et al., 2006. The tomato NBARC-LRR protein Prf interacts with Pto kinase in vivo to regulate specific plant immunity. Plant Cell 18 2792–2806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Noel, L., T. L. Moores, E. A. Van Der Biezen, M. Parniske, M. J. Daniels et al., 1999. Pronounced intraspecific haplotype divergence at the RPP5 complex disease resistance locus of Arabidopsis. Plant Cell 11 2099–2111. [PMC free article] [PubMed] [Google Scholar]
  30. Qu, S., G. Liu, B. Zhou, M. Bellizzi, L. Zeng et al., 2006. The broad-spectrum blast resistance gene Pi9 encodes a nucleotide-binding site-leucine-rich repeat protein and is a member of a multigene family in rice. Genetics 172 1901–1914. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Rose, L.E., P. D. Bittner-Eddy, C. H. Langley, E. B. Holub, R. W. Michelmore et al., 2004. The maintenance of extreme amino acid diversity at the disease resistance gene, RPP13, in Arabidopsis thaliana. Genetics 166 1517–1527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Rybka, K., M. Miyamoto, I. Ando, A. Saito and S. Kawasaki, 1997. High resolution mapping of the indica-derived rice blast resistance genes II. Pita-2 and Pita and a consideration of their origin. Mol. Plant Microbe Interact. 10 517–524. [Google Scholar]
  33. Sakata, K., Y. Nagamura, H. Numa, B. A. Antonio, H. Nagasaki et al., 2002. RiceGAAS: an automated annotation system and database for rice genome sequence. Nucleic Acids Res. 30 98–102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Salmeron, J. M., G. E. D. Oldroyd, C. M. T. Rommens, S. R. Scofield, H. S. Kim et al., 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. [DOI] [PubMed] [Google Scholar]
  35. Sinapidou, E., K. Williams, L. Nott, S. Bahkt, M. Tor et al., 2004. Two TIR:NB:LRR genes are required to specify resistance to Peronospora parasitica isolate Cala2 in Arabidopsis. Plant J. 38 898–909. [DOI] [PubMed] [Google Scholar]
  36. Smith, S. M., A. J. Pryor and S. H. Hulbert, 2004. Allelic and haplotypic diversity at the Rp1 rust resistance locus of maize. Genetics 167 1939–1947. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Takken, F. L. W., M. Albrecht and W. I. L. Tameling, 2006. Resistance proteins: molecular switches of plant defense. Curr. Opin. Plant Biol. 9 383–390. [DOI] [PubMed] [Google Scholar]
  38. Toki, S., 1997. Rapid and efficient Agrobacterium-mediated transformation in rice. Plant Mol. Biol. Rep. 15 16–21. [Google Scholar]
  39. Tsunematsu, H., M. J. T. Yanoria, L. A. Ebron, N. Hayashi, I. Ando et al., 2000. Development of monogenic lines of rice for blast resistance. Breed. Sci. 50 229–234. [Google Scholar]
  40. Wang, Z. X., M. Yano, U. Yamanouchi, M. Iwamoto, L. Monna et al., 1999. The Pib gene for rice blast resistance belongs to the nucleotide binding and leucine-rich repeat class of plant disease resistance genes. Plant J. 19 55–64. [DOI] [PubMed] [Google Scholar]
  41. Zhang, H. B., X. Zhao, X. Ding, A. H. Paterson and R. A. Wing, 1995. Preparation of megabase-size DNA from plant nuclei. Plant J. 7 175–184. [Google Scholar]
  42. Zhou, B., S. Qu, G. Liu, M. Dolan, H. Sakai et al., 2006. The eight amino-acid differences within three leucine-rich repeats between Pi2 and Piz-t resistance proteins determine the resistance specificity to Magnaporthe grisea. Mol. Plant Microbe Interact. 19 1216–1228. [DOI] [PubMed] [Google Scholar]
  43. Zhou, B., M. Dolan, H. Sakai and G. L. Wang, 2007. The genomic dynamics and evolutionary mechanism of the Pi2/9 locus in rice. Mol. Plant Microbe Interact. 20 63–71. [DOI] [PubMed] [Google Scholar]

Articles from Genetics are provided here courtesy of Oxford University Press

RESOURCES