NGS-BSA analysis places Ra locus on chromosome 4
To genetically characterize the novel Ra locus we determined previously, a BSA assay was conducted using the parents (male, BF145; female, F58050), and the R-pool and S-pool samples. High throughput sequencing produced ~102, 91, 158, and 131 million trimmed and filtered reads from BF145, F58050, R pool and S pool, respectively. The average sequencing depth was about 25× for parents and 33× for bulked samples, and the on-target mapping rate to the DM reference ranged from 84.6% to 89.2% in the four samples, indicating high-quality data suitable for downstream analysis. GATK variant calling identified a total of 1,542,634 variants, including 1,365,306 SNPs and 177,328 InDels, between the R and S DNA bulks.
Δ(SNP-index) analysis between the bulks revealed one unequal region spanning 23.32 to 34.97 Mb on chromosome 4 of the DM potato reference genome (Fig. 1), surpassing the 95% confidence interval threshold. However, unlike diploid populations, no distinct peak was observed in this study. Moreover, a region spanning 11.65 Mb was too extensive to precise gene mapping or marker development. Clearly, this approach alone couldn’t identify or fine-map the projected Ra locus in the tetraploid potato materials.
We thereafter explored variant enrichment approach, as described by Strachan et al. (2019), successfully used in both diploid and tetraploid potatoes for gene mapping and cloning (Strachan et al. 2019; Witek et al. 2016). As ERA was controlled by a single dominant gene in the BC1 population (Huang et al. 2021), we deduced that the alleles of resistant and susceptible plants should be Rararara and rararara, respectively. Thus, the 19,238 SNPs and 2,300 InDels matching these expected ratios were retained. SNPs and InDels in a 2 Mb interval using a 10 kb sliding window were then computed for each chromosome, as shown in Fig. 2 and Table S1, total 53.8% SNPs and 42.7% InDels were enriched on chromosome 4, respectively. Notably, the clustered SNPs/InDels located at 6.05-11.97 Mb and 39.01-46.51 Mb on chromosome 4 (Fig. S1), were potentially linked to the ER to PVA. Complementing the Δ (SNP-index) analysis, total three regions (region Ⅰ, 6.05-11.97 Mb; region Ⅱ, 23.32-34.97 Mb; region Ⅲ, 39.01-46.51 Mb) on chromosome 4 were tentatively identified as candidate regions associated with the Ra locus.
InDel markers define the Ra locus to a 1.33Mb interval on chromosome 4 in potato
To verify the three candidate intervals, a set of 80 InDel primers was developed and utilized to screen for polymorphism among the two parents and two bulks. Eventually, total 56 InDel markers exhibited polymorphism (Table S2), and region Ⅰ (6.05-11.97 Mb), region Ⅱ (23.32-34.97 Mb) and region Ⅲ (39.01-46.51 Mb) contained 30, 17, and 9 InDel markers, respectively. Upon completing the initial assessment, these polymorphic markers were applied to the segregating BC1 population. In the chi-squared test by SPSS 20.0, the segregation ratio of 39 markers fitted the 1:1 ratio and was subsequently used to construct a genetic linkage map for the resistant parent BF145 by TetraploidMap (InDel markers in Table S2). The map covered 63.36 cM, integrating 39 InDel markers, with an average distance of 1.41 cM between markers (Fig. 3). Integrated with the phenotypic data of the BC1 population, assigned as “0” for susceptible or “1” for resistant, the Ra locus was tentatively placed within 40.82 cM between the markers M8-83 and M10-8 using TetraploidMap (Fig. 3).
To further confirm the mapping results, the graphical genotyping was employed to validate the NGS-BSA outcomes. As shown in Fig. 4, a total of 26 markers were determined to be closely linked to ERA in the presumptive interval based on the NGS-BSA and the genetic linkage analyses, among them, 11 were in region Ⅰ (6.05-11.97 Mb), 8 in region Ⅱ (23.32-34.97 Mb), and 7 markers in region Ⅲ (39.01-46.51 Mb). Graphical genotyping assigned “1” (green) for alleles that originated from the resistant parent BF145, or “0” (blue) for alleles that originated from the susceptible parent F58050 (Fig. 4). Assessing the phenotypes of all progeny clones identified 15 recombinants in the BC1 population by using these markers, of which 5 were resistant and 10 were susceptible. In region Ⅲ (39.01-46.51 Mb), the genotype of four susceptible recombinants (BC007, BC045, BC103, and BC148) was assigned “1” by the eight markers mentioned above in this region (Fig. 4), indicating the absence of the Ra locus in this region. Similarly, in region Ⅱ (23.32-34.97 Mb), the susceptible recombinant BC007 was still assigned “1” by all eight markers in this region, reflecting the deviation of the Ra locus from this region. Overall, three recombinants (BC045, BC103, and BC148) in this region were exchanged between markers M25903 and M33897_230. In region Ⅰ (6.05-11.97 Mb), the recombinants BC007 and BC073 were exchanged between markers M8-83 and M10-8. Based on the outcome of the resistance test, the Ra locus was mapped to marker M8-83 and M10-8. In view of all above analysis, the region Ⅰ (6.05-11.97 Mb) is suggested to be the true candidate region for the Ra locus.
The genetic locations of 39 markers on the chromosome 4 were compared with their physical locations on the DM potato reference genome (Fig. S2). The genetic locations of most markers on the chromosome 4 are consistent with the physical locations of the markers on the DM potato reference genome (Fig. S2). Hence, the genetic linkage map for the resistant parent BF145 exhibits good collinearity with the genome. Based on the genetic map, the most effective markers that were used to delimit the Ra interval were M8-83 and M10-8. According to the physical location of the flanked markers (M8-83 and M10-8) on the potato DM reference genome, the Ra gene might be harbored in the genomic region of 8.78-10.11 Mb (1.33 Mb) on chromosome 4.
Development of Ra InDel marker for marker-assisted selection
The distances of between the flanked markers M8-83 and M10-8 from Ra were 0.74 cM and 0.73 cM, respectively (Fig. 3), showing a strong agreement of 98.6% and 99.3%, respectively, with the phenotyping results in the BC1 population. The closely linked InDel marker M10-8 was chosen to evaluate its association with PVA resistance across various potato cultivars/advanced breeding clones (n = 43). The marker was present in 24 clones, with 20 of them exhibiting ER to PVA by graft-inoculation. Among the 19 M10-8-absent clones, 6 were determined to be ER to PVA (Fig 5, Table 1, Table S3). Overall, a 76.7% concordance was obtained between the marker and the phenotype, accompanied by a correlation coefficient of 0.443** (P < 0.01).
Table 1 Evaluation of molecular markers M10-8 in 43 potato cultivars and advanced breeding lines
Molecular marker M10-8
|
Response to PVA infection
|
Clones
|
Total
|
Percentage (%)
|
1
|
ER
|
20
|
33
|
76.7
|
0
|
S
|
13
|
1
|
S
|
4
|
10
|
23.3
|
0
|
ER
|
6
|
ER, extreme resistance; S, Susceptible; 0/1, absence/presence of the molecular marker M10-8.