Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Feb 15:3:e3.
doi: 10.1017/qpb.2021.17. eCollection 2022.

Quantitative modelling of fine-scale variations in the Arabidopsis thaliana crossover landscape

Yu-Ming Hsu  1   2   3 , Matthieu Falque  3 , Olivier C Martin  1   2   3
Affiliations

Quantitative modelling of fine-scale variations in the Arabidopsis thaliana crossover landscape

Yu-Ming Hsu et al. Quant Plant Biol. .

Abstract

In, essentially, all species where meiotic crossovers (COs) have been studied, they occur preferentially in open chromatin, typically near gene promoters and to a lesser extent, at the end of genes. Here, in the case of Arabidopsis thaliana, we unveil further trends arising when one considers contextual information, namely summarised epigenetic status, gene or intergenic region size, and degree of divergence between homologs. For instance, we find that intergenic recombination rate is reduced if those regions are less than 1.5 kb in size. Furthermore, we propose that the presence of single nucleotide polymorphisms enhances the rate of CO formation compared to when homologous sequences are identical, in agreement with previous works comparing rates in adjacent homozygous and heterozygous blocks. Lastly, by integrating these different effects, we produce a quantitative and predictive model of the recombination landscape that reproduces much of the experimental variation.

Keywords: chromatin state; epigenetic features; recombination rate; sequence divergence.

PubMed Disclaimer

Conflict of interest statement

The authors declare no potential conflicts of interest.

Figures

None
Graphical abstract
Figure 1.
Figure 1.
The correlations between recombination rate and nine genomic or epigenomic features taken from somatic tissues (cf. titles). Each dot represents the values for a 100-kb bin. The x-axis shows the density of each feature, and the y-axis is the recombination rate based on a total of 17,077 crossovers from the Col-0-Ler F2 population. Dots in red, blue or green are for bins located in arms, pericentromeric regions or the transition regions between arms and pericentromeric regions, respectively. The black curves are fits to polynomials of degree 4 (function lm(y ~ poly(x,4)) of the statistical package R). R 2 corresponds to the fraction of explained variance when using the polynomial as predictor (equation (2)). To ensure that the points fill most of the space, the scale in the main part of each panel is a zoom to display only 95% of the data, cutting the 2.5% extremities on both sides of the x-axes in all these plots. Insets show the data in the whole range.
Figure 2.
Figure 2.
Relations between our 10 chromatin states, genes, intergenic regions and recombination rate. (a) The top pie chart shows the genome-wide occupation percentages of each of the 10 states. ‘SV’ refers to low synteny regions or structural variations between Col-0 and Ler. The characteristics of the nine other states are: state 1 (intragenic, transcription starting site (TSS)), state 2 (intergenic, proximal promoter), state 3 (intragenic, coding sequence), state 4 (intergenic, distal promoter), state 5 (intergenic, H3K27me3 rich), state 6 (intergenic, transcription termination site (TTS)), state 7 (intragenic, long genes), state 8 (heterochromatic, AT rich) and state 9 (heterochromatic, GC rich). The lower pie chart shows the percentage of crossover occurrences identified in the 10 states. (b) Two plots, giving respectively the profiles of cumulated fractions of occurrences of the 10 different states (top) and the recombination rate pattern (bottom) in cM per Mb, along gene bodies and their 3-kb flanking regions. In the absence of SV, the entire 3-kb flanking region was used, otherwise it was truncated. The gene body goes from the TSS to the TTS as given in TAIR 10. Only non-transposable element coding genes satisfying the synteny filter have been included in the analysis. For the gene body region, the x-axis represents relative position, that is the distance from the TSS divided by the distance between TTS and TSS. That procedure allows one to pool genes of different sizes. For the flanking regions, x-axis represents position relative to the TSS or TTS in kb. The blue curve at the bottom is the predicted recombination rate when using the chromatin state profiles at the top together with the genome-wide recombination rates derived from (a). (c) Two plots as in (b) but now for the intergenic regions. Again, the blue curve is the predicted recombination rate when using the chromatin state profiles at the top together with the genome-wide recombination rates derived from (a). The legend in the middle of (b) and (c) indicates the corresponding chromatin state of each color used in plotting the chromatin-state profiles.
Figure 3.
Figure 3.
The relationship between the size of intergenic regions and their average recombination rate. These bar charts were constructed using all intergenic regions, but in the bottom, the regions were divided into three categories according to the transcription orientations of the two flanking genes, corresponding to convergent, divergent and parallel transcriptions. In all cases, the x-axis gives the size of the intergenic regions in kb, and the y-axis gives the corresponding averaged recombination rate (cM/Mb). Binning of the intergenic region sizes was applied every 500 bases up to a total size of 10 kb. For example, the leftmost bin covers intergenic regions of size 0–0.5 kb. However, we also include a rightmost bar on each chart to cover intergenic regions of sizes larger than 10 kb. Error bars are errors on the mean computed by the jackknife method (only the top segments are displayed). In both top and bottom figures, the blue curves give the predicted recombination rates using the genome-wide recombination rates of the 10 chromatin states as obtained from Figure 2a. The red curves show the predicted recombination rates when one includes the modulation based on the size of the intergenic regions as specified in equation (4).
Figure 4.
Figure 4.
The relationship between recombination rate and single nucleotide polymorphism (SNP) density. The Col-0 genome was decomposed into bins of 100 kb. For each cross starting with that of Rowan et al. (2019), SNPs and crossovers (COs) were inferred from reads produced using the F2 populations by mapping to the Col-0 genome. SNP density and recombination rates were then determined for each bin and displayed as a scatter plot. The five additional crosses are from Blackwell et al. (2020). The continuous red curves are fits when using the function (a + b x) exp(−cx) so as to maximise the log likelihood. To filter out the high SNP density regions that are expected to causally repress recombination, we restricted the analysis to SNP densities in the first two quantiles. All crosses show a reduced recombination rate at low SNP density and the likelihood ratio test allows us to reject the hypothesis H0 that ‘b = 0’, corresponding to no such suppressive effect (p-values shown for each cross and computed using the chi-square distribution with one degree of freedom).
Figure 5.
Figure 5.
Experimental and predicted recombination landscapes of chromosome 1. Landscapes using 100 kb bins obtained from the Rowan et al. (2019) dataset (red) and predicted from our calibrated model based on chromatin states (blue) with 15 parameters. Inset: a zoom in the right arm. For landscapes of all chromosomes, see Supplementary Figure S9.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Arbeithuber, B. , Betancourt, A. J. , Ebner, T. , & Tiemann-Boege, I. (2015). Crossovers are associated with mutation and biased gene conversion at recombination hotspots. Proceedings of the National Academy of Sciences of the United States of America, 112, 2109–2114. 10.1073/pnas.1416622112 - DOI - PMC - PubMed
    1. Barth, S. , Melchinger, A. E. , Devezi-Savula, B. , & Lübberstedt, T. (2001). Influence of genetic background and heterozygosity on meiotic recombination in Arabidopsis thaliana . Genome, 44, 971–978. 10.1139/g01-094 - DOI - PubMed
    1. Bauer, E. , Falque, M. , Walter, H. , Bauland, C. , Camisan, C. , Campo, L. , Meyer, N. , Ranc, N. , Rincent, R. , Schipprack, W. , Altmann, T. , Flament, P. , Melchinger, A. E. , Menz, M. , Moreno-González, J. , Ouzunova, M. , Revilla, P. , Charcosset, A. , Martin, O. C. , & Schön, C.-C. (2013). Intraspecific variation of recombination rate in maize. Genome Biology, 14, R103. 10.1186/gb-2013-14-9-r103 - DOI - PMC - PubMed
    1. Berardini, T. Z. , Reiser, L. , Li, D. , Mezheritsky, Y. , Muller, R. , Strait, E. , & Huala, E. (2015). The Arabidopsis information resource: Making and mining the “gold standard” annotated reference plant genome. Genesis, 53, 474–485. 10.1002/dvg.22877 - DOI - PMC - PubMed
    1. Blackwell, A. R. , Dluzewska, J. , Szymanska-Lejman, M. , Desjardins, S. , Tock, A. J. , Kbiri, N. , Lambing, C. , Lawrence, E. J. , Bieluszewski, T. , Rowan, B. , Higgins, J. D. , Ziolkowski, P. A. , & Henderson, I. R. (2020). MSH2 shapes the meiotic crossover landscape in relation to interhomolog polymorphism in Arabidopsis. The EMBO Journal, 39, e104858. 10.15252/embj.2020104858 - DOI - PMC - PubMed

LinkOut - more resources

-