Allelic differences within DNA polymerase delta subunit 1 correlate with geminivirus resistance in diverse plants

Abstract

Identifying loci conferring resistance to geminiviruses is an on-going priority in diverse crop species. Multiple geminivirus resistance quantitative trait loci (QTLs) and genes have been described, including most recently DNA polymerase delta subunit 1 (POLD1) in both cassava and tomato. From this, we hypothesized that POLD1-mediated resistance is present in a broad range of plant species. An analysis of multiple species with published geminivirus resistance QTLs was done to identify species with POLD1 variation consistent with known resistance alleles. Further, allele mining of over 7,000 accessions across 10 different genera, from both dicots and monocots, identified additional substitutions in key regions of the POLD1 protein as possible novel resistance alleles. These results provide evidence that POLD1 variation is a source for geminivirus resistance in diverse plants including cotton, soybean, squash, wheat, and maize.

Keywords: DNA polymerase delta 1; Plant Genetics and Genomics; QTL; allele mining; geminivirus; virus resistance.

Fig. 1.

Geminivirus resistance QTLs previously described in a) cassava, b) tomato, c) soybean, and d) squash. The location of key markers is labeled, and genes within the QTL region are either in light gray (forward strand) or dark gray (reverse strand). The location of the POLD1 gene is labelled. The size of the QTL reported in the corresponding citation is in parentheses. The 2 QTL plots from soybean are drawn to the same scale; similarly, the 2 QTL plots for Cu. moschata are at the same scale to one another.

Fig. 2.

Location of mutations affecting the POLD1 protein in diverse plant species. a) Mutations in POLD1 in geminivirus-resistant plant varieties. Red: positions affected by mutations identified as resistance alleles underlying resistance in cassava and tomato; light blue: positions affected by mutations found in POLD1 present under geminivirus resistance QTLs in soybean and squash; orange: positions affected by mutations found in public sequencing data for resistant cotton and through targeted sequencing of the resistant accession S. chilense LA2779. All mutations are found within the N-terminal domain, catalytic center of the palm motif, and the finger motifs. b) Mutations identified in POLD1 from allele mining of over 7,000 varieties from 10 different genera, analyzed for predicted effect on function using SIFT. Mutations above the POLD1 diagram: SNPs predicted to negatively affect protein function; in blue and teal are high confidence SNPs and in gray are low confidence mutations. All mutations below the diagram are predicted to be tolerated by the POLD1 protein. Teal mutations are found in the same areas of the protein structure as all mutations in a); bolded mutations are at a position also found in a). For both a) and b), each SNP is labeled with a 2-letter code for the source species and coordinates for the position within the POLD1 protein sequence of the source species. Mutations found in geminivirus-resistant accessions are indicated with an asterisk (*). At, A. thaliana; Cc, Caj. cajan; Cl, Citrullus lanatus; Cm, Cu. moschata; Cmu, Citrullus mucosospermus; Gh, G. hirsutum; Gm, Gl. max; Me, Manihot esculenta; Mt, Me. truncatula; Sh, S. huaylasense f. glabratum; Sl, S. lycopersicum; Sp, Solanum peruvianum; Ta, Triticum aestivum; Vu, V. umbellata; Zm, Z. mays; Sc, S. chilense; Sc, S. chmielewskii.

Fig. 3.

Model of the 3D protein structure of POLD1 from Homo sapiens (PDB ID: 6NTZ). Double-stranded DNA is in beige in the center of the protein structure. All residues from Fig. 2a are in red; teal residues from Fig. 2b are in teal; blue residues in Fig. 2b are in blue; low confidence residues from Fig. 2b are in light cyan; residues from Fig. 2b not predicted to affect protein structure are in dark gray.