Evolution of pathogen response genes associated with increased disease susceptibility during adaptation to an extreme drought in a Brassica rapa plant population

Abstract

Background: Pathogens are key components in natural and agricultural plant systems. There is evidence of evolutionary changes in disease susceptibility as a consequence of climate change, but we know little about the underlying genetic basis of this evolution. To address this, we took advantage of a historical seed collection of a Brassica rapa population, which we previously demonstrated evolved an increase in disease susceptibility to a necrotrophic fungal pathogen following a drought.

Results: Previously, we combined a resurrection experiment with genome-wide sequencing of 124 pooled ancestral and descendant plants. Here, using these previously generated sequence data (Franks et al. in Mol Ecol 25(15):3622-3631, 2016), we show that well-characterized necrotrophic fungal pathogen response (NFPR) genes have evolved, as indicated by changes in allele frequency, between ancestors and descendants, with several of them identified as extreme F_ST outliers. The jasmonic acid (JA) signaling pathway in particular seems to underlie the evolution of disease susceptibility, in addition to its well characterized role in plastic disease response. We identify a list of 260 genes that are both NFPR genes and are differentially expressed in response to drought, based on publicly available data. We present evidence that five of these genes evolved between ancestors and descendants, suggesting that the drought acted as the evolutionary driver, and that the accompanying increase in disease susceptibility may have been a consequence of genetic pleiotropy.

Conclusions: Our study provides evidence that for this population, standing variation in NFPR genes is affected by natural selection related to climate change. Our results reveal potentially important candidates that may underlie trait evolution in both crops and natural systems. Additionally, this trade-off between adaptation to biotic and abiotic stresses is an example of how climate change can have diverse and unexpected consequences.

Keywords: Alternaria brassicae; Brassica rapa; Climate change; Drought; Necrotrophic fungal pathogen response; Resurrection approach.

Fig. 1

Differentiation at multiple loci across the genome in a Brassica rapa population after seven years of drought, as well as historical measures of selection. a F_ST, an estimate of allelic differentiation between the 1997 and 2004 populations, was calculated using 100 kb sliding windows and then averaged for each gene (dots shown). Genes reported in the literature as involved in response to necrotrophic fungal infection (A. brassicicola) are shown in black. All other genes are shown in grey. For necrotrophic fungal pathogen response genes, significantly differentiated genes are labelled with red points and their gene ID. A LOESS trend line (span = 0.03) is shown for all genes (blue) and just for necrotrophic fungal pathogen response genes (red). Tajima’s D, a statistic estimating the effect of non-random processes from the site frequency spectrum, shows historical selection across the genome for the b ancestral population and c descendant populations. Analysis was conducted using a 100 kb sliding window (dots shown are windows; LOESS trend line in blue). Red bars highlight regions containing necrotrophic pathogen related genes with a significant high F_ST that are also in an area with reduced Tajima’s D (visual inspection). Regions with reduced Tajima’s D may have been subject to a historical selective sweep

Fig. 2

Mapping of SNPs and designation of synonymous and non-synonymous (outlined in black box) SNPs in our two top candidate necrotrophic fungal pathogen response genes in Brassica rapa that evolved over seven years of drought. Shown are alignments between B. rapa reference V. 1.18, ancestral B. rapa and descendant B. rapa for our 2 top candidate genes in top and bottom panels, with exons and non-coding SNPs shown in green. Nucleotide identities in coding regions are indicated below SNP sites and identified by the following colors: red for A, bright blue for C, orange for G, and purple for T. Heterozygous sites are indicated by the presence of different adjacent colors and indicated by including both nucleotides below heterozygous sites (e.g. A/T). Allele frequencies are not indicated. Undetermined regions are shown in dark blue. a Bra030295 had 19 SNPs, 5 of which are in exons, and 2 of which are non-synonymous (black boxes), indicating candidate functional polymorphisms. b Bra038107 had 5 segregating sites, all in the single exon, with 3 non-synonymous SNPs (black boxes), indicating candidate functional polymorphisms

Fig. 3

Venn diagram showing gene databases used in this study for annotation including necrotrophic fungal pathogen response genes (NFPR), drought response genes (Drought), genes which evolved genome-wide (Sig Genes), and pleiotropic genes (overlap between NFPR and drought response) out of a total 35,202 genes