Meiotic crossovers are associated with open chromatin and enriched with Stowaway transposons in potato

Abstract

Background: Meiotic recombination is the foundation for genetic variation in natural and artificial populations of eukaryotes. Although genetic maps have been developed for numerous plant species since the late 1980s, few of these maps have provided the necessary resolution needed to investigate the genomic and epigenomic features underlying meiotic crossovers.

Results: Using a whole genome sequencing-based approach, we developed two high-density reference-based haplotype maps using diploid potato clones as parents. The vast majority (81%) of meiotic crossovers were mapped to less than 5 kb. The fine-scale accuracy of crossover detection was validated by Sanger sequencing for a subset of ten crossover events. We demonstrate that crossovers reside in genomic regions of "open chromatin", which were identified based on hypersensitivity to DNase I digestion and association with H3K4me3-modified nucleosomes. The genomic regions spanning crossovers were significantly enriched with the Stowaway family of miniature inverted-repeat transposable elements (MITEs). The occupancy of Stowaway elements in gene promoters is concomitant with an increase in recombination rate. A generalized linear model identified the presence of Stowaway elements as the third most important genomic or chromatin feature behind genes and open chromatin for predicting crossover formation over 10-kb windows.

Conclusions: Collectively, our results suggest that meiotic crossovers in potato are largely determined by the local chromatin status, marked by accessible chromatin, H3K4me3-modified nucleosomes, and the presence of Stowaway transposons.

Fig. 1

Construction of two high-density haplotype maps. a W4M6 and DMRH pseudo one-way testcross populations were used to construct high-density haplotype maps in potato. b Using short nucleotide variants (SNVs) to calculate linkage disequilibrium (LD) within windows of 50 SNVs, we used high frequency linked alleles to reconstruct the two segregating haplotypes of the heterozygous parent. The reconstructed haplotypes were then used to assign haplotypes to individual SNVs in the progeny. A single crossover between SNV 3 and SNV 4 is illustrated for individual 3. c Haplotype map of the W4M6 and DMRH populations. Red and blue segments reflect the two alleles segregating in the W4M6 population (US-W4 alleles) with yellow segments in between to illustrate large crossover intervals (lack of intervening haplotype-differentiating SNVs). The green and orange segments represent the two alleles segregating in the DMRH population (RH alleles)

Fig. 2

Validation of map construction and crossover resolution. a Seven deletion PCR markers were designed using the heterozygous parent US-W4. A random subset of individuals from the F₁ population (n = 56) was used to screen these markers to validate the SNV-based haplotype map. Here, we illustrate a crossover event between the coordinates of 64 and 66 Mb for F₁ individual 34 and PCR-based genotyping methodology for a subset of six genotypes. The genotype call is written above the bands for each individual as denoted by the F₁ individual ID. The only individual with a change in its genotype call (due to a crossover) is the F₁ individual 34. A single band within a lane represents a homozygote and two bands per lane indicates a heterozygote. Deletion markers were selected based on homozygosity in M6 and heterozygosity in US-W4. Red, haplotype 1 and genotype “A”; blue, haplotype 2 and genotype “B”. b Sanger sequencing of a fine-resolution crossover. The top two sequences are the two US-W4 haplotypes, differentiated by red and blue polymorphisms, respectively, while the bottom sequence is the result of Sanger sequencing a single crossover from one recombinant individual and is chimeric with respect to the two haplotypes

Fig. 3

Genomic features of meiotic crossovers. a Crossover counts per megabase observed in the W4M6 population (purple). b Crossover counts per megabase observed in the DMRH population (orange). c Recombination rate interpolated onto 100-kb windows using crossovers from the W4M6 and DMRH populations collectively. The red background of this track denotes the upper third quartile for recombination rate, while the white background encompasses 75–25% of recombination rates (second quartile), and blue highlighted regions indicate regions below the first quartile of recombination. d Gene density per megabase. Normalized SNV density from the W4M6 (e) and DMRH (f) populations per 100 kb. SNV counts per 100 kb were multiplied by the percentage of bases covered by reads in the 100-kb window and divided by the total number of reads within the window. g Count of DNA transposons (class II) per 100 kb. h Count of RNA transposons (class I) per 100 kb. i Cytosine methylation in CG (first track), CHG (second track), and CHH (third track) contexts expressed as the average methylation level per base, averaged across a 1-Mb window. The gray windows indicate regions in the top quartile of recombination rate. The top quartile of recombination rate is also illustrated as black bars in the center of the figure

Fig. 4

Genomic features associated with meiotic crossovers. a An example of the expected versus observed distributions of crossover counts for chromosome 1 of potato. b Frequencies of the two W4M6 haplotypes, blue and orange, overlaid with − log10 transformed P values from a chi-squared test for distorted segregation (black dots). The red dashed line across the plot represents the − log10 transformed significance threshold of P < 0.01. c Overlap analysis of crossovers with various genomic features (orange). Matched “cold regions” were also compared against different genomic features (blue), in addition to simulations of random regions (purple). Error bars represent the standard deviation of 10,000 simulations. If an interval overlapped more than one genomic feature, the interval was counted towards each overlapping feature. d Aggregated plot of crossover frequency relative to all potato genes. Genes were split into 50 windows and up- and downstream regions were split into 25-bp windows. Counts of crossovers were averaged across genes for each window. Orange, crossovers; blue, cold regions; gray, random regions. e Heatmap of gene ontology (GO) terms associated with crossover-overlapped, cold, and random genes. No terms were significant in either random or cold data sets, and thus the only GO terms shown are those that were significantly enriched for crossover-associated genes. P values are expressed as Benjamini–Hochberg corrected P values

Fig. 5

Meiotic crossovers are associated with fine-scale genomic and chromatin features. a Left panel: Spearman’s correlation matrix for recombination rate, DNase-seq, H3K4me3, and 5mC in CHH, CHG, and CG contexts across 100-kb windows on chromosome 12. Right panel: Heatmaps of CHH, CHG, CG, H3K4me3, and DNase-seq density overlaying recombination rate across 100-kb non-overlapping windows of chromosome 12. b Left panel: Normalized DNase-seq read counts over and flanking crossovers (orange), matched cold regions (blue), and random regions (gray). Right panel: Normalized H3K4me3 read counts over and flanking crossovers (orange), matched cold regions (blue), and random regions (gray). c Comparison of normalized DNase-seq and H3K4me3 read counts for genomic features overlapping crossovers, and simulations of the analogous features near (10–1000 kb) but not overlapping crossovers. Error bars on the simulated distributions are from 10,000 permutations. d Left panel: Normalized DNase-seq read counts over and flanking genes and their promoters that overlapped crossovers (orange) and permutations of genes that did not overlap genes (blue). Right panel: Normalized H3K4me3 read counts over and flanking genes and their promoters that overlapped crossovers (orange) and permutations of genes that did not overlap genes (blue). In both panels the light blue shading for non-crossover genes reflects a simulated distribution of 100 permutations, while the statistical tests in the text are based on 10,000 permutations. Genes were split into 50 windows, while flanking regions were split into 10-bp windows. Lines reflect averaged values across genes in each group (crossover-associated and random). *Significance at P < 0.05

Fig. 6

Meiotic crossovers are enriched with Stowaway MITE transposons. a Significantly enriched (before Bonferroni multiple testing correction) transposons between fine-resolution crossovers and matched cold regions. Top barplot: Proportion of significantly enriched transposons in cold regions (gray background) and crossovers (red background) out of all transposons overlapping each data set. Blue bars represent the portion the transposon takes in cold regions. Orange bars represent the portion the transposon takes in crossovers. Bottom barplot: −log10 transformed P values from Fisher’s exact test. Grey bars are for transposons enriched within cold regions, purple bars denote significance for transposons enriched within crossovers. Transformed P values are shown only for the data set in which they are significant. *Significance following Bonferroni correction at P < 0.05. b Aggregate plot of Stowaway element distribution relative to genes across the potato genome. Orange, normalized count of Stowaway elements (per bp window); blue, random control. c Heatmap of Stowaway elements per 100 kb on chromosome 12, on top of recombination rate interpolated onto 100-kb windows. d Promoters of genes from the upper quartile of recombination rate are more enriched with Stowaway elements than genes from crossover-poor regions. Orange, normalized count of Stowaway transposons in gene promoters of recombination rich regions; gray, normalized count of Stowaway elements within gene promoters from recombination poor regions. e Promoters carrying Stowaway elements underlie regions with significantly higher recombination rates (orange line), compared to simulated distributions (one million permutations) of nearby (within 500 kb) promoters lacking Stowaway elements (blue histogram). Y-axis represents the counts of observations with a mean cM/Mb value (X-axis)