Introduction

Finger millet (Eleusine coracana (L.) Gaertn.) (2n = 4, x = 36), sub-species coracana, belongs to the family Poaceae and the genus Eleusine in the tribe Eragrostideae. Finger millet commonly known as ragi (India), bulo (Uganda), wimbi (Swahili), and tellebun (Sudan) and is an important cereal crop for subsistence agriculture in dry areas of Eastern Africa, India, and Sri Lanka. It has nutritional qualities superior to that of rice and is on par with that of wheat [1]. It is rich in protein (6–13 %) and calcium (0.3–0.4 %) and serves as an important staple food for rural populations in developing tropical countries where calcium deficiency and anaemia are widespread [2]. Plant diseases are the greatest deterrents to crop production worldwide. Among them, Magnaporthe grisea (anamorph: Pyricularia oryzae), a filamentous ascomycetes fungus, causes blast disease in economically important crops like rice. One of the major biotic constraints in the finger millet crop is the high seed yield loss (50 %) due to blast disease caused by the pathogenic fungus, M. grisea. This demands genetic improvement of finger millet by transferring plant resistant genes against fungus [3]. Fungus affects leaves, fingers, neck and discolors the seeds. Hence, major efforts have been devoted for understanding the molecular mechanisms of genetic resistance and incorporating them into breeding programs to avoid yield loss caused by pathogen.

The analysis of DNA sequence variation is of major importance in genetic studies. In this context, molecular markers are a useful tool for assaying genetic variation, and have greatly enhanced the genetic analysis of crop plants. Microsatellites or simple sequence repeat (SSR) markers have been useful for integrating the genetic, physical and sequence-based physical maps in plant species, and simultaneously have provided molecular breeders with an efficient tool to link phenotypic and genotypic variation. The disadvantages of SSR markers are that they are often tedious and costly cloning and enrichment procedures required for their generation [4–6]. Microsatellites developed from expressed sequence tags (ESTs), popularly known as EST–SSRs or genic microsatellites, represent functional molecular markers as a putative function for a majority of such markers can be deduced by database searches and other in silico approaches. Furthermore, EST–SSR markers are expected to possess high inter specific transferability as they belong to relatively conserved genic regions of the genome. The EST databases have become particularly attractive resources for such In-silico mining, as was demonstrated in, e.g., citrus [7], coffee [8], and particularly in the cereals [9], which can be further effectively being used in diversity studies [10], linkage map construction [11], QTL mapping studies [12] and marker assisted breeding programmes for disease resistance.

Disease resistance is frequently governed by specific recognition between pathogen AVIRULENCE (Avr) genes and corresponding plant disease RESISTANCE (R) genes. This type of gene for gene interaction usually is accompanied by a hypersensitive response leading to the restriction of pathogen growth [13]. Forty-eight R-genes have been cloned from numerous plant species (Arabidopsis, rice, wheat, etc.) using map-based cloning and transposon tagging [14]. The nucleotide binding site—leusine rich repeat (NBS–LRR) disease resistance genes belong to a large and diverse super family of genes, which can be subdivided based on characteristic N-terminal features of their products. Conserved amino acid sequence motifs in the NBS domain have been widely used to isolate and classify NBS–LRR encoding genes [15–18]. The NBS–LRR class resistance genes are proposed to act as receptors in signal transduction pathways that are triggered in response to pathogen attack. The NBS region of characterized R-genes and R-gene analogs (RGAs) contains several highly conserved motifs, in spite of the diversity of pathogens against which they act. Comparative genomics is a powerful tool for genome analysis, annotation with an objective to understand the detailed process of evolution at the gross level and to translate DNA sequence data into proteins of known functions. The rationale here is that DNA sequences encoding important cellular functions are likely to be conserved between species than the non-coding sequences. Thus the present study aimed at (1) Identifying the type, frequency and distribution of EST–SSRs of finger millet; (2) Comparative genomic analysis of NBS–LRR regions of finger millet with rice.

Materials and methods

Plant materials and genomic DNA isolation

For comparative genomic analysis, a set of 16 finger millet genotypes (ten resistant and six susceptible) were taken for the polymorphism analysis of NBS–LRR regions for blast resistance. The details of the genotypes used in the study along with the blast scoring data were given in Table 1. The genomic DNA of different accessions of finger millet was isolated by standard method [19], quantified and analyzed on agarose gel electrophoresis [20].

Table 1 The genotypes used in the study with their blast scoring data
Full size table

EST data mining

By July 2012 a total of 1956 EST sequences were available for finger millet in NCBI (GenBank) database. The downloaded sequences were obtained in FASTA format for sequence assembly and SSR analysis. The SSR identification was done using one pipeline tool “Websat” software [21]. Primers designed to flanking sequences using the Websat software which uses the Primer3 software. Five classes of SSRs, that is, di-, tri-, tetra-, penta-, and hexa nucleotide repeats were targeted for identification using this tool. Two types of search criteria were taken for identification of EST–SSRs. The first one was: the minimum number of repeats was 6 for dinucleotide, 4 for trinucleotide, and three for tetra, penta and hexanucleotides. While the second criterion was changed for tetra, penta and hexa with a minimum number of repeats as two was considered. The main parameters for primer designing were: GC content of 40–60 %, annealing temperature (Tm) of 50–60 °C and expected amplified products size of 100–450 bp. All other parameters were set to default values.

SSR amplification and detection

PCR was performed in 20 μL reaction volume containing 2 μL 10× buffer having 15 mM MgCl2, 0.2 μM of each forward and reverse primer, 2 μL of 2 mM dNTPs, 0.2 μL of 0.5 U/μL of Taq DNA polymerase (Invitrogen, USA), and about 50 ng of template DNA. Amplifications were performed in a Thermocycler (MJ Research, USA) programmed for an initial denaturation of 4 min at 95 °C followed by 35 cycles of 30 s at 95 °C, 30 s at (different annealing temperatures for different primers), 1.0 min at 72 °C, and a final extension of 10 min at 72 °C, and hold at 4 °C. The PCR products were fractioned on 3.5 % agarose gel. The electrophoresis was held at 100 volts for 3 h at room temperature. Gels were stained with ethidium bromide and visualized using Bio Imaging System (SynGene, USA) and the primers that amplified and showed detectable length polymorphism were identified. Only those bands that were clear and reproducible were scored for data analysis. Molecular weight of the bands was estimated using 100 bp DNA ladder as standard.

Strategy of comparative analysis of NBS–LRR regions

The EST sequences of finger millet were retrieved from the NCBI website. These sequences were used for BLASTn analysis and identified the Oryza sativa sequences producing high similarity at an E value more than 4e-13. Then identified the positions of these finger millet and rice EST sequences on the rice chromosome maps in the gramene website (www.gramene.org) and compared with the positions of different rice blast genes positions.

Sequencing by ABI 3130XL genetic analyzer

All primer pairs were initially tested via PCR and agarose gel analysis aimed at identifying those pairs producing single amplicons. The primers which produced single amplicons and showed polymorphism between resistant and susceptible genotypes were used for further analysis. The PCR products were purified using a QIAquick® PCR purification kit (Qiagen Inc., Valencia, CA, USA) according to the manufacturer’s protocol. Dideoxy cycle sequencing was performed using the chain-termination method and an ABI Prism Big Dye reaction kit (ver. 3.1) according to the manufacturer’s protocols (Applied Biosystems). The sequencing products were run on an ABI 3130XL genetic analyzer. Sequence editing and assembly of the contigs were performed using Sequencher 4.10. For sequence comparisons among the sequences of genotypes, BioEdit (ver. 7.0.5.3; [22]) with the Clustal W multiple alignment option was used and then adjusted manually by the authors. The PCR products were sequenced from both ends and the resulting termination products were analyzed on an ABI 3130XL genetic analyzer. The two resulting sequence traces derived from opposite ends of each amplicon were analyzed and aligned with standard DNA analysis software Phred and Phrap (http://www.phrap.org/).

Results and discussion

SSR mining and EST-SSRs frequency and distribution in finger millet

Simple sequence repeats (SSRs) were searched among the 1956 EST sequences of finger millet available in the NCBI website using the software Websat. The search was performed by considering two types of criteria. The first criteria includes repeat motif as: di- (with a repeat count n ≥ 6 repeat units), tri- (n ≥ 4), tetra- (n ≥ 3), penta- (n ≥ 3) and hexa- (n ≥ 3) nucleotides. According to this criteria, a total of 599 (30.6 %) out of 1956 ESTs had SSRs, of which 545 SSR primer pairs were designed. The remaining ESTs had SSRs at very close to the 5′ end or 3′ end of the sequences hence primer pairs could not be designed. Different workers used different criteria for finding the SSRs in the EST sequences. Recently, Reddy et al. [23] found that 324 ESTs had SSRs and developed 132 primer pairs out of the then available ESTs (1927) by using the criteria as n ≥ 5 for di-, tri-, tetra-, penta-, and hexa- repeats. However, we have taken the most common criteria used by several workers in different crops [24]. Thirty-two ESTs had more than two SSRs. As expected, the most frequent type of microsatellites corresponded to trimeric SSRs (320, i.e. represents 53.5 % of the detected SSRs and 16.4 % of total ESTs). The percentage of different SSR repeat motifs present in finger millet has been presented in Fig. 1. Similar findings have been reported by Victoria et al. [24], where they also identified trimer repeat motifs were the most frequent repeats among green algae and mosses; monocots and dicots. In case of medicinal plants also trimer motifs are the most abundant repeats. Zheng et al. [25] found 55 % of the ESTs of Chinese medicinal plant (Epimedium sagittatum (Sieb. Et Zucc.) Maxim) consisted of trimeric repeats. This showed that trimers are the most frequent motifs among crop species. Reddy et al. [23] found dimeric repeats as the most frequent among all the repeats which may be due to that, minimum of repeats was given five for finding the SSRs. In the present study, trimers were followed by the dimeric (102, 17 % of detected SSRs) and tetrameric (75, 12.5 % of detected SSRs) microsatellite repeats. The frequency of pentameric and hexameric SSRs was lower, representing only 23 (3.9 %) and 25 (4.2 %) of the detected microsatellites respectively. However, Reddy et al. [23] found only 11 penta nucleotides and no hexa nucleotide repeats since they used minimum criteria as 5 repeats. So, these results showed that minimum three repeats is sufficient for penta- and hexa- nucleotide repeat SSRs as evidenced in other crops [24, 25]. The second criteria have given different results in comparison to first criteria. According to the second criteria, the repeat motif used as follows: penta- (with a repeat count n ≥ 2 repeat units) and hexa- (n ≥ 2) nucleotides and the remaining di-, tri- and tetra repeats were same as first criteria. According to this criteria, a total of 1248 (64 %) out of 1956 ESTs had microsatellites, and ninety-five ESTs had more than two microsatellites which were higher than the first criteria. The most frequent type of microsatellites observed corresponding to pentameric SSRs (765, i.e. represents 39.1 % of the total ESTs) which was followed by hexameric (630, 32.2 % of the total ESTs).

Fig. 1
figure 1

The details of the repeat motifs (in percentage) present in finger millet

Full size image

Among the dimeric SSRs, GA was found as the most common motif, followed by AG repeat motif and similar results were found by Reddy et al. [23]. GA was the most abundant motif which was in accordance with the reports in other crops [26–28]. However, Maia et al. [29], found AG/CT and GA/TC were the most frequent dimeric repeat motifs in the crops belonged to the family Poaceae. In case of Algae species, the most frequent dimer motif was AC/GT and CA/TG and in vascular plants like rice AG and GA was the most frequent dimer motif [24]. This showed that among the lower plant species AC and CA; and among higher plant species GA and AG were the frequent dimer motifs. The GA/CT motif represent codons GAG, AGA, UCU, and CUC, in mRNA which translate into the amino acids Arg, Glu, Ala, and Leu, respectively. In proteins, Ala and Leu are present in higher frequencies which results in the abundance of GA/CT motifs in EST sequences. Among trimeric SSRs, the most common motifs were CGG, followed by CTC, CCG and GCA and also supported by Reddy et al. [23]. In rice, CCG/CGG is the predominant trimeric motif and it appears to be high in the members of the grass family [27]. The motifs CCG/CGG, CGC/GCG, and GCC/GGC seem to be less common in other families [29]. For the three other classes, the most common SSR types corresponded to GAGC (for tetrameric SSRs), AAGAG (for pentameric SSRs), and CAGCTC (for hexameric SSRs). The details of the repeat type, frequency distribution have been presented in Table 2. Among the grasses, 0.85 % of all motifs were either CCTC/GAGG, AGGA/TCCT, or CATC/GATG. Other reports have shown ACGT as the most abundant in barley [30] and AAAG/CTTT and AAGG/CCTT in perennial ryegrass [31]. The pentamer repeat AAGAG is the most common repeat in finger millet and also in solanaceous crops [29]. These results showed that most of the higher crops share the common microsatellite repeat motifs.

Table 2 The frequency, type and distribution details of EST-SSRs of finger millet
Full size table

Comparative genomic analysis of NBS-LRR of finger millet with rice

Comparative genomics has progressed the discovery and understanding of orthologues, and also it has brought to light many fast evolving ‘orphan’ genes of unknown function and evolutionary history. In Eleusine species, comparative analysis provides an opportunity to study rapid genome changes, function of the newly isolated gene sequences and their further utilization in molecular breeding assisted crop improvement programmes for developing high yielding, stress resistant genotypes. Srinivasachary et al. [32] studied the comparative genomic analysis of finger millet and rice genomes through molecular markers and showed that 85 % synteny present between these two crops. In this concern, there is a need to exploit the comparative genomic strategy to identify the molecular markers associated with important agronomic, physiological traits of finger millet. The blast fungus M. grisea causes major economic drain and main pathological threats to rice and finger millet crop over a greater part of the world. In the present study, 45 EST sequences of NBS–LRR region for disease resistance (including blast resistance) in finger millet were retrieved from the NCBI website. The EST sequences were clustered using the MEGA4 [33] software for finding the similar sequences. The MEGA4 analysis could group all the 45 EST sequences into five major clusters viz., A, B, C, D and E (Fig. 2). The major cluster A further divided into two minor clusters A1 and A1. The cluster A contains the 12 ESTs, cluster B 10 ESTs and cluster C, D and E contained 12, 7 and 4 EST sequences respectively. To identify the homology between finger millet and rice chromosomes, BLASTn analysis was performed with these EST sequences in the NCBI website and selected those had E value more than 4e–13. The finger millet ESTs clustered under ‘A’ exhibited homology to the rice EST sequences of AY337926, AY337898 and AF392824 (Table 3). To know the map locations of the finger millet EST sequences on rice chromosome map, these were used for BLASTn search engine of gramene website and obtained their positions on the rice chromosome maps and also same strategy was applied for rice EST sequences. The ESTs of finger millet clustered under C were showed homology with rice PiKh contig (GU258508) and four other sequences (DQ272576, AF220745, Y09812, AF074892) where both the finger millet and rice ESTs were hit on the eleventh chromosome. These results indicated that PiKh gene orthologs may be playing important role for the blast resistance in finger millet also. In a similar way, the finger millet EST sequences (GU301915, EU075236, EU075221) of cluster D showed homology to the rice ESTs (AB430853, DD461353 (both Pi21)) with a putative function related to Pi21 gene influencing the blast resistance and all were hit on the same rice chromosome, i.e. fourth.

Fig. 2
figure 2

Clustering pattern of EST sequences of NBS–LRR regions of finger millet generated by MEGA4

Full size image
Table 3 In-silico comparative analysis of NBS–LRR regions of finger millet with rice
Full size table

However, the ESTs of finger millet clustered in ‘A’ were hit on the 6th chromosome of rice, while the homologous rice EST sequences were hit on the 11th chromosome. Similarly, the ESTs of finger millet clustered in B were hit on the 2nd chromosome of rice, while the homologous rice EST sequences were hit on the 11th chromosome. Most of the ESTs of rice sequences were hit on the 11th chromosome; however finger millet ESTs were hit on 11th, 6th and 2nd chromosomes. This showed that the finger millet blast resistance may be governed by the orthologous regions of rice blast genes. There was high synteny between rice and finger millet chromosomes as reported by Srinivasachary et al. [32] and the same was authenticated by our results.

Functional marker based characterization and sequencing analysis

A total of 22 SSRs were identified, and 15 SSR loci were designed from sixteen GenBank accessions representing the NBS–LRR region of rice which showed similarity with the finger millet EST sequences. Of the 15, two primer pairs were not amplified among the selected resistant and susceptible genotypes of finger millet. The finger millet genotypes used in the present study were thoroughly characterized for years together for their response to blast resistance (Table 1). Hence, for the present study, 13 SSR primers were used for molecular characterization of ten resistant and six susceptible finger millet genotypes. Out of the thirteen SSR loci, eight were found polymorphic (61 %). The details of the polymorphic primers have been given in Table 4. The SSR loci FMBLEST5 designed from the rice EST contig (DQ272576.1) could able to clearly differentiate the susceptible and resistant genotypes (Fig. 3). The alleles which were unique to the resistant and susceptible genotypes were further used for sequencing and analysis by purifying the PCR product. A total of two sequences were obtained from two different resistant genotypes and one from the susceptible genotype with a Phred score above 40. The Clustal W was performed to know the similarity among the sequences and found high similarity among the resistant genotype sequences. However, sequence from the susceptible genotype did not show any similarity with the other two sequences and also did not hit with the rice resistant EST sequences after BLASTn analysis. Recently, Panwar et al. [34] studied the functional markers based molecular characterization of resistance gene analogs encoding NBS–LRR disease resistance proteins from a large collection of finger millet genotypes for association of NBS sequences with the blast disease resistance and susceptibility caused by blast fungus. The study established genetic relationships among the closely related genotypes and thus help build a more complete picture of the molecular characterization of finger millet genotypes. Such a characterization of genetic variation within natural populations and among breeding lines proves to be crucial for the effective blast management in finger millet or any other target crops. Reddy et al. [35] isolated resistance gene homologues from finger millet (Eleusine coracana L.) using degenerate oligonucleotide primers designed to the conserved regions of the nucleotide binding site of the previously cloned plant disease resistance genes. Of the 107 clones sequenced, 41 showed homology to known R-genes, and are denoted as EcRGHs (Eleusine coracana resistance gene homologues), while 11 showed homology to pollen signalling proteins (PSiPs), and are denoted as EcPSiPs (Eleusine coracana pollen signalling proteins).

Table 4 The details of the polymorphic primers found among the resistant and susceptible finger millet genotypes
Full size table
Fig. 3
figure 3

Molecular profiling of susceptible and resistant finger millet genotypes based on the primer FMBLEST5 (sequencing was done with VR708 (Susceptible), VL 333 (R), VL 324 (R) genotypes)

Full size image

The alleles sequenced from the resistant (three alleles) and susceptible genotypes (one allele) were analyzed by performing BLASTn analysis to find the homologous sequences (sequences yet to submit to NCBI). The allelic sequences from resistant genotypes could find high similarity with the rice NBS-LRR regions (some were Pi genes of blast resistance), while the sequence obtained from the susceptible allele did not show any similarity with the any NBS-LRR region. The sequences obtained from the resistant alleles were hit on 11th chromosome of rice and the rice ESTs also hit on the 11th rice chromosome. Along with In-silico analysis, our results also showed the eleventh chromosomal region sequences of rice could be playing important role in the blast disease resistance in finger millet. The nucleotide sequence was further analyzed by conserved domain (CD) search available in the NCBI website and resulted in identification of the NB-ARC (nucleotide binding-APAF-1, R proteins, and CED-4) domain in our sequence. The NB-ARC domain is a novel signalling motif shared by plant resistance gene products and regulators of cell death in animals. This region includes a nucleotide-binding (NB) domain, consisting of kinase 1a (P-loop), 2 and 3a motifs [36], and several other short conserved motifs with unknown function. R gene products are key components in plant defence, which appears macroscopically as rapid localised host cell death at the site of pathogen ingress [37]. This hypersensitive response (HR) is a form of programmed cell death that is thought to impede further infection. The amino acid sequence of finger millet resistant genotype was further compared with the previously cloned plant disease resistance genes and it contained the characteristic NBS motifs of kinase-2 and kinase 3a of plant R-genes (Fig. 4) confirming that the sequences characterized in the present study belong to the NBS-LRR gene super family. Reddy et al. [35] also noticed kinase-2 and kinase-3a motifs in all the EcRGHs isolated from finger millet (Fig. 4). The NBS-LRR genes are usually grouped into two different subfamilies [38]; subfamily I contains the TIR element (Toll-interleukin-1 receptor-like domain) and has been found only in dicots, whereas the subfamily II which lacks the TIR domain (called non-TIR) was found in both monocots and dicots. The sequences converted to amino acid sequences also had a tryptophan residue (W) at the end of the kinase-2 motif, which indicates that these belong to the non-TIR sub-class of R-genes. The last amino acid residue of the kinase-2 domain can be used to predict with 95 % accuracy whether an R-gene belongs to the TIR–NBS–LRR class or the non-TIR–NBS–LRR family; conservation of tryptophan (W) at this location is tightly linked to the non-TIR class of R-genes (RPS2, RPS5 and RPP8) of A. thaliana, whereas conservation of aspartic acid reside (D) or its uncharged derivative asparagines (N) is characteristic of the TIR class of R-genes (N and L6) [39, 40]. Thus the present results showed that comparative genomic analysis of finger millet with rice genome sequences will be very much useful in exploiting the molecular marker studies for identification of the genes responsible for blast resistance, their mapping and for further introgression through marker assisted breeding approaches for enhancing the blast resistance in finger millet genotypes of locally well adopted germplasm.

Fig. 4
figure 4

The kinase2 and kinase3a motifs of NB–ARC domain present in the sequence of finger millet genotype as obtained from CD domain search

Full size image