Background

In rice more than 70 diseases caused by fungi, bacteria, viruses and nematodes are prevalent (Oryza sativa). The most devastating of them are the ones caused by Magnaporthe grisea (rice blast), Xanthomonas oryzae pv. oryzae (bacterial leaf blight, BLB) and Rhizoctonia solani (sheath blight). Improved agricultural practices, nutritional supplements, application of fungicides, bactericides and resistant cultivars had been used for disease control but no durable solution was available due to the breakdown of the resistance by high pathogenic variability. Hence, the search for resistant rice genotypes, particularly among the landraces, is in progress. According to Harlan [1] the extensive diverse array of rice landraces available worldwide are probable storehouses for novel alleles for many qualitative and quantitative traits. Harlan’s study emphasized that each landrace has certain unique properties or characteristics; such as early maturity, adaptation to particular soil types, resistance or tolerance to biotic and abiotic stresses, and in the end usage of the grains. India is home to many such unique landraces and the ones found in the ecological hotspots of the Indo-Burma region, and the Indian states of West Bengal, Assam, Nagaland, Mizoram and Manipur deserve special mention [2].

BLB caused by the vascular pathogen Xanthomonas oryzae pv. oryzae (Xoo) is one of the most serious diseases leading to crop failure in rice growing countries including Korea, Taiwan, Philippines, Indonesia, Thailand, India and China. Xanthomonas (from two Greek words; xanthos, meaning ‘yellow’, and monas, meaning ‘entity’) is a large genus of gram-negative and yellow-pigmented bacteria. Xoo enters rice leaf typically through the hydathodes at the leaf margin, multiplies in the intercellular spaces of the underlying epithelial tissue, and moves to the xylem vessels to cause systemic infection [3, 4].

Genes conferring resistance to the major classes of plant pathogens have been isolated from a variety of plant species and are termed ‘R genes’ [5]. Comparison of the structural features and the sequences of the predicted proteins from the cloned ‘R genes’ from various plants have led to the identification of common domains which are conserved and show little variation. These conserved domains can be divided into five broad classes. They are the nucleotide-binding domain (NBD), the leucine rich repeat domain (LRR), the coiled coil domain (CC), the serine/threonine protein kinase domain and the detoxifying enzymes [5]. A total of 38 [6] BLB resistance genes (R genes) have been identified in rice, including Xa1, Xa2, Xa3/Xa26, Xa4, xa5, Xa6, Xa7, xa8, xa9, Xa10, Xa11, Xa12, xa13, Xa14, xa15, Xa16, Xa17, Xa18, xa19, xa20, Xa21, Xa22(t), Xa23, xa24(t), xa25/Xa25(t), Xa25, xa26(t), Xa27, xa28(t), Xa29(t), Xa30 (t), xa31(t), Xa32(t), xa33(t), xa34(t), Xa35(t), Xa36(t). The recessive resistance genes include xa5, xa8, xa9, xa13, xa15, xa19, xa20, xa24, xa25/Xa25(t), xa26(t), xa28(t), xa31(t), xa33(t), and xa34(t). Of the 37, 10 BLB resistance (R) genes have been mapped on rice chromosomes 4 (Xa1, Xa2, Xa12, Xa14 and Xa25), chromosome 5 (xa5), chromosome 6 (Xa7), chromosome 8 (xa13), and chromosome 11 (Xa3, Xa4, Xa10, Xa21, Xa22, and Xa23). The chromosomal locations for the rest of the BLB resistance genes still remain elusive. These R genes are known to act in a gene-for-gene manner and are the main sources for genetic improvement of rice for resistance to Xoo. Ten of the recessive R genes; xa5[7], xa8[8], xa13[9], xa24[10], xa26, xa28[11] and xa32[12] occur naturally and confer race-specific resistance. The other 3, xa15[13], xa19 and xa20[14], were created by mutagenesis and each confers a wide spectrum of resistance to Xoo[11, 13, 15].

Six BLB resistance genes, Xa1, xa5, Xa21, Xa21(A1), Xa26 and Xa27, have been cloned, sequenced and characterized. In 1967, Sakaguchi [16] identified Xa1 conforming a high level of specific resistance to race 1 strains of Xoo in Japan and mapped it on rice chromosome 4. The gene xa5 is a naturally occurring mutation that is most commonly found in the Aus-Boro group of rice varieties from the Bangladesh region of Asia [7, 17]. The predicted protein product of Xa21 carries LRRs in the extracellular region and a serine/threonine kinase domain in the cytoplasm [18]. Xa21 is a member of a multigene family located on rice chromosome 11 [18, 19]. Seven Xa21 gene family members, designated A1, A2, B, C, D, E, and F, were cloned and grouped into two classes based on DNA sequence similarity [18]. Xa26 is a dominant gene coding for a LRR receptor kinase protein. It is mapped to the long arm of chromosome 11 [11, 20] and was found in cultivar Mingui 63 which showed resistance against a number of Xoo strains both at seedling and at adult stage suggesting that it was not developmentally regulated [14]. The Xa27 locus of rice conferred resistance to diverse strains of Xoo, including PXO99A, a strain isolated from rice variety IRBB27 by map-based cloning. Xa27 is an intron-less gene and encodes a protein of 113 amino acids.

Natural selection in the ecological niches of the world has generated landraces that are highly diverse for various quality, quantity and disease resistance traits controlling loci. It is important to identify and maintain this polymorphism to widen the genetic base of the commercially cultivated varieties and to reduce pathogen pressure. According to Glaszman et al. [21] study of local sequence variation reveals the multiple examples of mutation that have taken place due to adaptation towards specific drifts and selection pressure. This adaptive neo diversity superimposes on the ancestral diversity inherited from wild relatives and forms an important section in the passport data of various accessions. It is a tedious task to put the existing natural variation to commercial use. As a step towards that process Nordborg and Weigel [22] suggested the use of genome-wide association (GWA) mapping which associates the phenotype of interest to DNA sequence variation present in an individual’s genome determined by polymorphism at various loci. GWA mapping gives much higher resolution than linkage mapping because they involve studying associations in natural populations and reflect adaptive recombination events. This kind of mapping is very useful in self fertilized species like A. thaliana and rice [23]. Further, in view of the challenge of assessing the diversity in large germplasm collections, the core collection concept was developed wherein diversity analysis will first be concentrated on a representative manageable sample before extending the study to a broad range of accessions [24]. Such programs have been undertaken for rice and chickpea. In accordance with such postulates the objective of this study is to analyze a small set of phenotypically variable rice accessions from BLB disease hotspot for getting a birds-eye view of the existing diversity in 6 BLB resistant gene loci of those accessions.

Reports of diversity analysis of the BLB resistance genes are available. Ullah et al.,[24] identified the presence of the genes Xa4, xa5, Xa7, and xa13 in 52 basmati landraces and five basmati cultivars using Polymerase Chain Reaction (PCR) based methods. They also found that the gene Xa7 was most prevalent among the cultivars and landraces while the genes xa5 and xa13 were confined to landraces only. Ten basmati landraces from their study had multiple resistance genes. Arif et al.,[25] identified the BLB resistance gene Xa4 in 49 Pakistani rice lines. Lee et al.,[11] identified three rice cultivars with resistance to various Phillipino Xoo strains. The cultivar Nep Bha Bong had a new recessive gene, designated as xa26(t) for moderate resistance to races 1, 2, and 3 and resistance to race 5. The cultivar Arai Raj had a dominant gene designated as Xa27(t) for resistance to race 2. The cultivar Lota Sail had a recessive gene designated as xa28(t) for resistance to race 2. Bimolata [26] analyzed the sequence variation in the functionally important domains of Xa27 across the Oryza species and found synonymous and non-synonymous mutations in addition to a number of InDels in non-coding regions of the gene. To the best of our knowledge, there is no report available on diversity of BLB resistance loci of rice landraces from the Indian states of Assam, Arunachal Pradesh, Nagaland, Mizoram, Manipur, Tripura and West Bengal.

In this study 34 pairs of primers were designed from conserved domains of the six BLB resistance genes; Xa1, Xa5, Xa21, Xa21(A1), Xa26 and Xa27. The designed primer pairs were used to generate PCR based polymorphic DNA profiles to detect and elucidate the genetic diversity of the six genes in the 22 rice accessions collected from West Bengal and the North Eastern States of India.

Methods

Plant materials

A total of 22 rice genotypes, including landraces and check genotypes, were collected from rice research stations in India. The names of the accessions, source, category and disease phenotype are given in Table 1.

Table 1 Name of the landraces, their source, category, disease phenotype and number of accessions
Full size table

Designing primers from conserved domains of 6 BLB resistance genes

Thirty four pairs of primers were designed from publicly available sequences (NCBI) of conserved domains of 6 BLB resistance genes using the software BatchPrimer3 (http://probes.pw.usda.gov/batchprimer3). The conserved domains are: P loop, kinase 2, trans-membrane domain and LRR domain of the Xa1 gene; TF IIA domain of the xa5 gene; receptor kinase domain of the Xa26 gene; the total DNA sequence of the Xa27 gene; signal, LRR, charged and kinase domain of the Xa21 gene; and LRR, SNAP O11 and kinase domain of the Xa21(A1) gene. These primer pairs were named according to the initials of the first author and the corresponding author and were numbered from BDTG1 to BDTG34. The primer pairs were designed only from the exons such that the length of the amplified products was limited to 500 to 700 base pairs. Details of the primer names, respective resistance genes, representing protein domains, original genotypes from which the resistance genes were identified, number of exons and introns, chromosomal location in base pairs (bp) of each primer pairs and the expected length of the amplification product from the original genotype in base pairs (bp) are given in Table 2.

Table 2 Details of the primers used
Full size table

Isolation of rice genomic DNA and PCR amplification

Total genomic DNA was isolated from ten 3 day-old rice seedlings using the method of Walbot [27] with modifications. The DNA was PCR amplified using a protocol standardized in our lab and used in our previous paper [28].

Polyacrylamide gel electrophoresis and allele scoring

The PCR products were resolved in 6% polyacrylamide gel using the procedure described by Sambrook et al. [29]. The gel staining, visualization and assignment of alleles were done according to protocols in our previous paper [28]. Null alleles were assigned when no amplification product was generated [30]. When an allele was found in less than 5% of the germplasms under study, it was designated as rare [31].

Calculation of polymorphism information content (PIC) value

The polymorphism information content (PIC) value for the primer pairs was calculated using the formula given by Anderson et al.[32] for self pollinated species

PIC i =1– ∑ i = 1 n P 2 ij ,

where Pij is the frequency of the jth allele for the ith marker.

Genetic diversity analysis using PCR amplification profiles

A genetic similarity matrix between all possible combinations of pairs of rice accessions was made using Jaccard’s co-efficient [33] and the NTSYS-pc software package, version 2.02e, [34]. This similarity matrix was used to make a phylogenetic tree using the Unweighted Pair-Group Method of Arithmetic average (UPGMA) and Neighbor-Joining (NJoin) module of the NTSYS-pc. Support for clusters was evaluated by bootstrap analysis using WinBoot software [35] through generating 1,000 samples by re-sampling with replacement of characters within the combined 1/0 data matrix.

Sequencing and analysis of rare alleles

The DNA was eluted from the bands of rare alleles using QIAquick Gel Extraction Kit following manufacturer’s protocol. The eluted DNA was sequenced through outsourcing and the sequences were submitted to NCBI. For finding the homology and conserved domains, the sequences were BLAST [36] searched against the non-redundant database of NCBI using default parameters. Apart from NCBI BLAST, homology search for the obtained sequences were done using the “blastn” option of the Rice Annotation Database (http://rice.plantbiology.msu.edu).

Results

Analysis of PCR profiles

The summary of the data of the PCR profiles of the 22 accessions using the 34 pairs of primers is given in Table 3. All the 34 primer pairs produced polymorphic profiles and a total of 140 alleles were identified including 41 rare alleles. There were no unique alleles detected. The number of alleles ranged from 2 to 8 with an average of 4.06 alleles/primer pair. The primer pairs amplifying various regions of the LRR domain (Table 2) on an average produced 4.6 alleles/primer pair. Primer pairs amplifying the regions of kinase domain on an average produced 3.8 alleles/primer pair.

Table 3 Maximum and minimum band length, number of alleles (T), null (N) and rare (R) alleles with names of genotypes and PIC values for each primer pair
Full size table

The PIC value ranged from 0.16 for the least informative primer pair BDTG1 to 0.79 for the most informative primer pairs BDTG18, BDTG26 and BDTG29. The average PIC value was 0.56/primer pair.

Diversity in the six loci in this set of rice accession

The diversity generated by the 34 primer pairs in this set of rice accession is given in Additional file 1: Table S1. Briefly the highest variation was found in the locus Xa21(A1) between 4802 bp to 5082 bp (exon1, LRR domain, BDTG26) and between 5763 bp to 6173 bp (exon 1, LRR domain, BDTG29); and in the Xa26 locus between 4574 bp to 5141 bp (exon1, BDTG18), producing 8 alleles each. Three regions in the locus Xa21, from 8 bp to 208 bp (exon 1, the Signal domain, BDTG20), from 260 bp to 760 bp (exon 2, the LRR domain, BDTG21) and from 1279 bp to 1880 bp (exon 2, the LRR domain, BDTG23) produced 7 alleles each. Although the Xa27 locus was small, 392 bp long (1518 bp to 1909 bp), the primer pairs BDTG19 generated 6 alleles including 3 rare 2 null alleles. The next most variable region was in the Xa1 locus between 27182 bp to 27917 bp (exon 4, LRR domain, BDTG10), which produced 5 alleles. The region of TFIIA domain from 406048 bp to 406306 bp of locus xa5 (exon1, BDTG 11) produced 4 alleles including one rare allele and one null allele. The other most variable regions identified within the different loci are given in Additional file 1: Table S1.

Genetic diversity within the different categories of landraces

The West Bengal accessions produced a total of 107 alleles with an average of 3.15 alleles/primer pair. In this group, the highest number of alleles was generated by the primer pair BDTG23, while only one allele each was produced by BDTG6 and BDTG31. The North Eastern accessions produced a total of 100 alleles with an average of 2.94 alleles/primer pair. While the highest number of 6 alleles was generated by BDTG29, only 1 allele each was produced by BDTG8 and BDTG26. The check varieties comprised of one resistant and one susceptible accession. Out of the 41 rare alleles, 8 were produced by the resistant West Bengal landrace Bhasamanik and 7 each were produced by the resistant landraces Raghusail and Bangladeshi Patnai. Four rare alleles were identified in the Assamese aromatic landrace Lal binni and 2 rare alleles each were identified in the landraces Aijong, IC524526, IC524502 and Gobindobhog.

Dendrogram from the genetic similarity values

In the dendrogram the similarity between the rice accessions ranged from 18% to 89% and on this basis they were divided into 2 major clusters A and B (Figure 1). Cluster A separated out at 18% level of similarity and consisted of Raghusail and Bhasamanik, both of which were resistant accessions from West Bengal. Cluster B was subdivided into 4 different sub clusters. Cluster 1 segregated out at 53% level of similarity and included the North Eastern accessions Aijong, Boro Chhaiyamora, IC524502, IC524526 and Lal Binni along with the West Bengal accession Gobindobhog. All the accessions in cluster 1 were BLB susceptible. Cluster 2 segregated at 28.8% level of similarity and included the West Bengal accessions Dudherswar, Bangladeshi Patnai, Talmari, Bangalaxmi and Taraori Basmati, all of which were susceptible. Cluster 3 separated out at 28% level of similarity and consisted of Morianghou, Kala boro dhan, Buhrimtui and Bhu from the North Eastern States along with Chamormoni – an accession from West Bengal. Cluster 4 consisted of the accessions Pusa Basmati 1, the susceptible check TN1, Kataribhog - a resistant accession from West Bengal and IR72 - a resistant check.

Figure 1
figure 1

Dendrogram showing genetic relationship among 22 rice accessions based on Jaccard's genetic similarity matrix derived from 140 alleles at 6 BLB resistance gene loci. The major clusters are indicated on the left margin and the sub-clusters are indicated on the right margin.

Full size image

Homology searches for the sequences of the rare alleles

A total of forty one rare alleles were sequenced. Of these, 40 were submitted to and were assigned accession numbers by NCBI. The accession numbers of the sequences, details of sequence homology, and the details of the conserved domains corresponding to each sequence is given in Table 4. Fifteen of the sequences were from the North Eastern accessions and 25 sequences were from the West Bengal accessions. BLAST searches using the NCBI database revealed that six rare alleles from the North East were homologous to sequences of BLB resistance genes of Oryza sativa japonica. Three of the rare alleles were homologous with sequences of the Xa21 gene of O. longistaminata. Two rare alleles were homologous to sequences of Xa1 and Xa21(A1) gene of O. sativa indica and one rare allele each was homologous to the Xa21 gene sequence from O. rufipogon and Xa27 gene sequence of O. officinalis ecotype IC203740. The rare alleles from HR806765 and JM426578 from the North Eastern landraces Buhrimtui and Aijong respectively did not show any homology to the existing database.

Table 4 Details of the sequenced rare alleles obtained from this study and homology searches with NCBI
Full size table

Out of the 25 rare alleles from the West Bengal landraces, Raghusail and Bhasamanik contributed 8 rare alleles each and Bangladeshi Patnai contributed 7 alleles. Eight rare alleles were homologous to sequences from O. longistaminata and 7 rare alleles were homologous to O. sativa indica sequences from the NCBI database. Five rare alleles each were homologous to sequences of O. rufipogon and O. sativa indica.

Results of homology search using the Rice Genome Annotation Project (RGAP) Database are given in Table 5. The name of the locus which produced the most significant match, description of the matched locus, E-value and details of the Pfam hits are shown in the table. According to this database most of the rare alleles were homologous to sequences of receptor kinase like proteins.

Table 5 Details of the sequenced rare alleles obtained from this study and homology searches with Rice annotation Database
Full size table

Identification of conserved domains and retrotransposons from the DNA sequences of rare alleles using NCBI and rice genome annotation project database

A total of 23 conserved domains were identified from the 40 rare alleles. The details of homology search and the conserved domain corresponding to the sequence of each rare allele is given in Table 4. Fifteen of the domains were homologous to LRRs. These domains included receptor like kinases (found in 9 sequences), LRR N-terminal domains (found in 4 sequences) and Leucine-Rich Repeats ribonuclease inhibitor (RI)-like subfamily (found in 2 sequences).

The sequences with accession numbers HR575926 and HR575924 (both derived from landrace Raghusail) and HR806763 (derived from landrace IC524526) were homologous to the NB-ARC domain-containing protein having a Pfam hit with BED zinc finger domain (zf-BED). According to Arvind [37] BED-type zinc-finger domain [named after BAEF (boundary element-associated factor) [38] and DREF (DNA replication-related element-binding factor), [39] is found in the Oryza Xa1 gene. HR614233 and HR575927 were significantly homologous to transcription initiation factor IIA gamma chain, having a Pfam hit with TFIIA_gamma_N. Another sequence HR614234 was homologous to aspartic proteinase nepenthesin-1 precursor having a Pfam hit with Asp or Aspartic proteases family. Sequence JM426580 was significantly homologous to mRNA sequence of the gene Xa27 of Oryza sativa indica. The sequences HR806767 and HR806746 were found to have conserved domains homologous to sugar transferase, and NodB domain of the carbohydrate esterase 4 superfamily.

Conserved domain searches using the Rice Annotation Database revealed the presence of mobile DNA elements within the sequence of 4 of the rare alleles. HQ832768, the sequence of a rare allele from the West landrace Bhasamanik was homologous to an unclassified retrotransposon protein having a Pfam hit of Plant_tran or plant tranposases. The sequence HR806765 from the Mizoram landrace Buhrimtui showed homology with a putative transposon protein, CACTA, En/Spm sub-class of Oryza sativa subsp. japonica. According to UniProt database this transposon protein has a molecular function of helicase and hydrolase. JM426578 from the Assam landrace Aijong was significantly homologous to a putative retrotransposon protein of the Ty3-gypsy type. HR806755 from the Assam landrace Lal Binni was significantly homologous to a putative unclassified retrotransposon protein.

Discussion

The Eastern and North Eastern regions of India are one of the richest reserves of bio-diversity in the country [40]. The inherent variation in the ecotypes of rice, spontaneously evolved in the Eastern State of West Bengal was high enough for scientists to group them as Oryza sativa var. benghalensis, at one time [41]. DNA-based markers like SSR and RAPD have been used extensively for the study of such inherent genetic diversity in rice. The results of these studies were also used for unambiguous identification of germplasm and their protection under the trade related intellectual property rights (TRIPS) of the World Trade Organization (WTO). The accessions used in this study were selected from a larger collection to include as much variability as possible based on the agro-morphological data and SSR polymorphism analysis done previously in our laboratory [28, 42]. As a follow up of those studies, we aim to extend the search for genetic variability specific to various quality traits and disease resistance abilities. Information on the diversity of disease resistance loci is important to the plant breeders for the identification of diverse donors with major genes and partial resistance. In this preliminary assessment we have tried to find the genetic diversity within six cloned BLB resistance genes in a set of 22 diverse rice accessions using PCR based methods. Even though the sample size is small (22 accessions) it includes accessions of rice varieties from 5 Indian states both aromatic and non-aromatic along with traditional and evolved basmatis and checks.

The PCR profiles of all the 34 primer pairs were clear and consistent. Stutter bands, which were minor products amplified in PCR that has lower intensity than the main allele and normally lacks or has extra repeat units were also present in the profiles of most of the primer pairs [43]. The null alleles were probably due to mutations in the binding region of one or both of the primers, thereby inhibiting primer annealing [30]. The presence of 140 alleles in the 22 accessions indicates high genetic diversity within the 6 BLB resistance gene loci. Analyzing the phenotype-genotype association after actual disease inoculation is requisite for confirming whether the identified rare alleles have any impact on BLB resistance or they are new alleles for BLB resistance. Moreover, the sample size being 22 accessions only, an identified rare allele might no longer be rare after the inclusion of more accessions.

It can be seen from the dendrogram that there was no state-wise or geographical segregation of the accessions based on the obtained polymorphism data. However cluster 1 and 3 consists mostly of the accessions from the North Eastern States. There was some degree of segregation based on whether the accessions were resistant or susceptible. The two resistant landraces from West Bengal, Raghusail and Bhasamanik segregated into a separate major cluster (major cluster A). These two landraces were about 39% similar amongst themselves. The dendrogram also shows instances where susceptible and resistant cultivars have been grouped together. The resistant accessions Kataribhog and IR72 have 89% similarity amongst themselves and they are grouped into cluster 4 along with two susceptible accessions TN1 and Pusa Basmati 1. Another resistant cultivar Bangladeshi patnai is 42% similar to a susceptible, but very popular table rice variety, Dudherswar. Future similar studies incorporating more accessions will confirm whether the alleles generated by the designed primers used here are actually able to segregate accessions on the basis of disease phenotype. Future efforts should concentrate on DNA sequencing, Multiple Sequence alignment and association mapping of all the involved alleles to identify possible linkages between the DNA sequence and the disease phenotype. For improving disease resistance of the aromatic accessions parents may be chosen from major cluster A and B.

According to Zhao et al. [44] most of the knowledge about the genetic architecture of complex traits in rice is based on traditional quantitative trait locus (QTL) linkage mapping using bi-parental populations, which though informative but are not suitable to investigate the genomic potential and tremendous phenotypic variation of the more than 120,000 accessions available in public germplasm repositories. This can only be achieved by documentation of genomic variation at specific loci controlling complex traits using specific genomic region based primers rather than random primers. This variation then has to be coupled with association mapping, a method popularly known as GWA. The information regarding the diversity of domains of the 6 BLB resistant loci obtained in this study is the first step towards such mapping programs. Rather than sequencing all the alleles obtained, only the rare alleles were sequenced in this study. Hence we could not establish any association between the DNA sequence and the resistant and susceptible accessions. For this sequencing of all the alleles and its correlation with disease phenotype are required and these are areas open for future investigation. If such associations can be found, then those will be the forerunner of GWA mapping for BLB resistance loci. In addition to the usual domains like LRR, TFIIA and BED-type zinc-finger, homologies to other conserved domains were also found in this study. The sequence HR806767 was homologous to a sugar transferase domain. Members of sugar transferase family are similar to the pfam00534 Glycosyl transferases group 1 domain. Glycosyltransferases can transfer single or multiple activated sugars to a range of plant molecules, resulting in the glycosylation of plant compounds and plays a key role in in the regulation of plant growth, development and in defense responses to stress environments [45]. Sequence HR806746 is homologous to a Catalytic NodB homology domain of the carbohydrate esterase 4 superfamily. This family catalyzes the N- or O-deacetylation of substrates such as acetylated chitin, peptidoglycan, and acetylated xylan, respectively [46]. The sequence HR614234 is homologous to aspartic proteinase nepenthesin-1 precursor. The Oryza sativa constitutive disease resistance 1 (OsCDR1) gene product is an aspartic proteinase that has been implicated in disease resistance signaling. This apoplastic enzyme is a member of the group of ‘atypical’ plant aspartic proteinases [47]. These unusual conserved domains within the rare alleles can be the result of local adaptation. Evaluation of the exact role of these unusual motifs in BLB resistance could be done with the help of disease inoculation and assessment of the disease phenotype. However that was beyond the scope of this study and has been left for future studies.

Transposable elements (TEs) were detected in the DNA sequence of 4 rare alleles. Transposable elements (TEs) are fundamental role players in the variation and adaptive evolution of plant genomes [48–50]. Grass genomes are reported to have active retrotransposons [51]. LTR retrotransposons constitute a major portion of the rice genome [52]. Retrotansposons are activated during stress, wounding and pathogen attack [53, 54]. For example transcription of the tobacco retrotransposon Tnt1 could be induced by pathogens and microbial elicitors, as well as by abiotic factors, [55–57]. Moreover Tnt1 insertion could change host gene splicing [58]. A group of LTR retrotransposons was found near the genes encoding the NPR1 disease resistance-activating factor and a heat-shock-factor-(HSF-) like protein in sugarbeet hybrid US H20 [59]. The TEs in this study were found mostly in landraces from the North East or from West Bengal BLB resistant landrace. The probable role of these identified tranaposable elements in this study are yet to be investigated.

Conclusion

As the name implies, conserved domains of genes are thought to possess little variation. However, this study finds that there is high genetic variability even within the conserved domains of BLB resistance genes in a small set of 22 rice accessions. Environmental stresses including high rainfall, humidity, varied topography and altitude, heavy natural selection pressures of diseases and pests, together with introductions over time and space from adjoining countries like Bhutan, China, Myanmar and Bangladesh; introgression from the wild and weedy relatives, tribal preferences and rituals have been instrumental in the development of this diversity [60]. The inclusion of more genotypes from remote ecological niches and hotspots holds more promise for further allele mining. Future studies should concentrate on DNA sequencing of all the alleles obtained in this study to bring out possible differences between susceptible and resistance accessions. Association mapping after disease inoculation will help to bring out the linkage between the alleles and disease phenotype. Such kind of mapping will be the stepping stone towards genome wide association mapping for BLB resistant loci. Search for transposable elements in the BLB resistance gene loci of the North eastern and resistant rice accessions, and elucidation of their function should form another area of interest.