Identification of test isolate
Morphological characteristics viz., hyphae, conidia, conidiophores, conidia size and conidiophore foot cells were studied on the detached leaves of host using stereo zoom microscope (Fig.1; Table 2). Attanayake et al. 2010 [7] described two groups of powdery mildew infected pea pathogen in combination of morphological and molecular characters. PCR amplification revealed an amplicon of approximately ~560bp, (Fig. 2) which was further gel purified and lyophilized before sequencing. BLAST analysis of the sequences of test isolates P-1 (from pea) and P-2 (from clover) indicated that both the strains were placed in the phylogenetic lineage occupied by the genus Erysiphe along with species, pisi and trifolii, respectively (Fig.2). The 18S rRNA sequence strain has been deposited in the NCBI GeneBank (Accession number KX455922 and KX455923, respectively).
Screening for resistant pea lines
Previously, screening was done and selected 3 resistant lines were crossed with JI-2302 (er1) and JI-2480 (er2) in 8 cross combinations viz., JI-2480 x Acacia, JI-2480 x PMR-10, JI-2480 x EC-381866-1, JI-2480 x Lincoln, JI-2302 x Acacia, JI-2302 x PMR-10, JI-2302 x EC-381866-1 and JI-2302 x Lincoln under net-house and green house and description of infection was observed [6]. Resistant was governed in maximum cultivars due to presence of er1 gene (Table 3). We therefore, select er1 gene for further studies.
Amplification studies
A total of 50 pea lines were used for RNA extraction (Fig.3). cDNA prepared from RNA was further amplified by specific primers mentioned in material and methods. To achieve this many times repeated PCRs were carried out in all the samples to standardize the protocol. Out of 50 lines, amplification was possible with primer 3F and 3R which produced 40 amplicons of variable size (250-300 bp) targeted that the er1 gene was present only in these lines. (Table 4; Fig 3).
In silico analysis of gene sequences
BLAST N search for homology of all the sequences of er1 gene corresponds to the gene present in Pisum sativum MLO1 (MLO1) mRNA, complete cds; homology queries values ≥ 90% and E values near 0 for Nucleotide Blast analysis. Phylogenetic analyses separated the pea accessions into 3 groups. The major group A constituted by 32 accessions and the remaining 8 accessions were grouped as 6 and 2 in B and C group, respectively. Obtained tree was then saved to newick format and Fig Tree program was used for tree illustration (Fig.4) (http://tree.bio.ed.ac.uk/software/figtree/ ).
No. of haplotypes
A total of 11 haplotypes were obtained and the frequency of haplotypes ranged from 1 to 24. Hap-1 was the most abundant haplotype representing 24 genotypes including the reference genotype (FJ463618.1). The remaining haplotypes were represented by single genotype except Hap- 4 which is represented by 3 genotypes (Table 5). The Hap-1 having 23 genotypes showed 100 per cent similarity with reference genotype (FJ463618.1), hence does not have any base substitution w.r.t PsMLO1, whereas, Hap-2, 3, 4, 5, 6, 7, 8, 9, 10 and 11 have 9, 5, 1, 6, 6, 16, 15, 14, 13 and 6 base substitutions, respectively.
Polymorphic sites
The analysis of polymorphic sites was carried out using DNAsp VI, and total of 36 sequences were used with a total of 198 variable sites. The 47 polymorphic sites included 18 Singleton variable sites, out of which 17 were with two variants and one was with three variants. There were 29 Parsimony informative sites out of which 26 were with two variants and three with three variants. The analysis of polymorphic sites was further studied in detail using multiple sequence alignment. Multiple sequence alignment was carried out in MEGA-(software) using the tool Clustal W [8]. Each haplotype of evaluated resistant pea genotypes were compared with the reference genotype for any site with replacement, deletion and addition (Fig. 5).
Haplotype diversity and Tajima’s Test
Haplotype diversity was calculated using DNAsp VI where haplotype (gene) diversity was of 0.5571 and Nucleotide diversity (per site) of 0.01606 (Table 6). Tajima’s test also found to be statistically significant with Tajima's D -2.09021 at P < 0.05. A median-joining network inferred from 40 set of sequences with 33 no. of active haplotypes was drawn. A value of zero was set for epsilon (e=0) to calculate sparse networks quickly, or incrementally. Maximum no. of mutations (29) was found at character 941 and least no. of mutations (1) were ranged at characters 992-1190 (Fig. 6).