Meiotic recombination is the process by which chromosomes swap segments during the formation of pollen and egg cells, creating the novel allele combinations that fuel evolution and agriculture alike. The number and position of these crossovers, or exchanges of genetic material between homologous chromosomes, are tightly controlled by chromosome architecture, a phenomenon called crossover interference, and by the choice of DNA repair pathway. Yet recombination is also famously plastic in response to the environment, and temperature is among the strongest modulators known in plants. Previous work had shown that in barley, temperature redistributes crossovers between proximal and distal chromosome regions, with similar though more moderate remodelling reported in wheat. In Arabidopsis itself, both warming and cooling can elevate crossover frequency. What remained unclear was how much of this response varies within a species, which genes are responsible, and how temperature connects mechanistically to the choice between the two biochemical routes that generate crossovers.

Those two routes are central to the new findings. In most eukaryotes, including plants, crossovers are produced by two pathways: the interference-sensitive ZMM, or class I, route, and the interference-insensitive MUS81-EME1, or class II, route. Class I crossovers are spaced apart along chromosomes by interference, which prevents exchanges from occurring too close together, while class II crossovers are largely free of such constraints. The researchers began by surveying nine genetically diverse Arabidopsis accessions crossed to lines carrying fluorescent seed reporters that allow crossover frequencies to be measured in two defined intervals on chromosome 3: a subtelomeric interval called 420, near the chromosome end, and a pericentromeric interval called 3.9, near the centromere. When the hybrids were grown at 12 degrees Celsius instead of the standard 21 degrees, most accessions showed a consistent signature: crossover frequency rose in the subtelomeric interval and fell in the pericentromeric one, indicating a redistribution of exchanges toward the chromosome arms.

The magnitude of the shift varied considerably from cross to cross, ranging from a 7 to 32 percent increase in the 420 interval and an 11 to 25 percent decrease in the 3.9 interval. Notably, the response did not correlate with the spring temperatures of the accessions’ native habitats or with their flowering time, consistent with the idea that most modern Arabidopsis accessions belong to a cosmopolitan lineage shaped by post-glacial expansion and human dispersal. Two accessions, however, stood out. Ler-0 and Se-0 showed no significant crossover changes at either interval when crossed to the reporter background, revealing that cold-induced recombination remodelling depends on genetic background and is not universal across the species. Intriguingly, when the researchers backcrossed the reporters into a mostly Ler genetic background, the subtelomeric cold response came back even stronger than in the reference Col-0 accession, roughly 2.5-fold greater, suggesting that Col and Ler achieve their contrasting behaviours through distinct underlying mechanisms.

To see the whole picture, the team turned to genome-wide crossover mapping. They generated large F2 populations from two hybrids, Col-420 × Per-1, which responds strongly to cold, and Col-420 × Ler, which does not, and genotyped hundreds of individual plants by whole-genome sequencing to locate crossovers as switches between parental SNP genotypes. In the responsive cross, cold produced a clear genome-wide redistribution: crossovers per individual rose on chromosome arms from 4.25 to 5.03 on average, while pericentromeric crossovers trended downward, and the landscape shifted visibly toward the distal ends of chromosomes. In the Col × Ler cross, by contrast, nothing changed significantly at any scale. Crucially, analysis of the coefficient of coincidence, a measure of interference, showed that interference patterns were essentially identical across temperatures and genotypes, extending over distances of up to roughly 8 megabases. Cold, in other words, was not weakening interference; it was changing which pathway produced the crossovers.

The genetic dissection that followed pinpointed the culprit. Quantitative trait locus mapping in the Col × Ler F2 populations, using subtelomeric crossover frequency as the trait, revealed a major locus on chromosome 4, dubbed rQTL4, whose effect intensified dramatically in the cold. At 21 degrees Celsius the locus explained 20.1 percent of the variance in crossover frequency; at 12 degrees it explained 48.6 percent, with the LOD score jumping from 8.48 to 23.14. A second locus on chromosome 1, likely corresponding to the known recombination modifier HEI10, showed only modest temperature modulation. The critical experiment came from F6 lines that differed exclusively at a 15.6-kilobase interval on chromosome 4 encompassing SNI1 and four neighbouring genes, fixed either for the Col or the Ler haplotype while the rest of the genome was identical. Cold increased subtelomeric recombination more than twofold more strongly in lines carrying the Ler version of the interval, and a significant genotype-by-temperature interaction confirmed that this small stretch of DNA governs the differential cold response.

SNI1, it turns out, is the Arabidopsis orthologue of NSE5, a component of the SMC5/6 complex, a conserved structural maintenance of chromosomes machine increasingly implicated in meiotic recombination and in chromosome architecture. The team tested the gene’s role directly using sni1-1, a hypomorphic allele that reduces but does not abolish SNI1 function. In Col inbreds, wild-type plants showed the canonical cold response, but sni1-1 mutants eliminated it entirely: heterozygotes measured 21.3 centimorgans in the 420 interval at 21 degrees and 22.2 at 12 degrees, while homozygotes measured 31.1 at both temperatures. Reducing SNI1 dosage, in other words, fixed the recombination landscape in a constitutively cold-like, subtelomere-biased state that no longer responded to temperature. Conversely, in the buffered Col × Ler hybrid, lowering SNI1 dosage unmasked and amplified cold responsiveness, with subtelomeric crossovers rising by up to nearly 6 centimorgans at 12 degrees in sni1-1 homozygotes. Genome-wide sequencing of sni1-1 populations confirmed the pattern, showing significant cold-driven increases in total, arm and subtelomeric crossover counts that were absent in wild-type hybrids.

The mechanistic link to pathway choice came from cytology in zip4 mutants, which are defective in the class I pathway and therefore reveal class II activity directly. In the Col background, zip4 chiasmata, the cytological manifestations of crossovers, nearly doubled at 12 degrees, rising from 1.15 to 2.04 per cell, demonstrating that cold promotes class II crossover formation. In zip4 sni1-1 double mutants, chiasmata were already elevated at 21 degrees, reaching levels comparable to cold-treated zip4, and did not increase further in the cold. The interpretation is elegant: SNI1 normally inhibits class II crossover formation at warm temperatures, and cold increases class II output primarily by relieving this SNI1-dependent constraint. In Ler, the cold-driven rise in zip4 chiasmata from 0.86 to 2.73 was likewise abolished by the sni1-1 mutation, while in hybrids, adequate SNI1 activity buffered the response and reduced dosage unmasked it. Inter-homologue polymorphism, the sequence differences between chromosomes from different parents, appears to dampen class II crossovers, which may explain why hybrid responses were consistently weaker than those of inbreds.

The study also probed chromosome architecture using super-resolution microscopy of the synaptonemal complex, the protein scaffold that zips homologous chromosomes together during meiosis. In Col, cold shortened total synaptonemal complex length from 217.2 to 172.1 micrometres, and sni1-1 mutants displayed a constitutively short complex even at 21 degrees, measuring 163.2 micrometres. Yet the relationship proved subtle rather than deterministic. Ler increased class II crossovers without detectable complex shortening, and Col × Ler hybrids shortened their complexes in the cold without any corresponding crossover response in the assayed intervals. The authors therefore interpret total complex length as a broad readout of SNI1/SMC5/6-dependent chromosome organization rather than a causal variable that directly dictates crossover outcome. Structural modelling added a final layer: the Ler and Se-0 alleles of SNI1 carry a unique I235V amino acid substitution at the interface with the partner protein ASAP1, plus four promoter variants that raise transcript levels, a compound haplotype that may explain why these accessions buffer the cold response despite expressing more SNI1 messenger RNA.

The implications stretch well beyond Arabidopsis. Because cold shifts crossovers toward chromosome arms without disrupting interference, modulating SNI1 dosage or allele state could allow breeders to bias genetic shuffling toward gene-rich distal regions under mild cold, avoiding the recombination-poor pericentromeric drag that limits the recovery of novel combinations. The allele-by-temperature interaction uncovered here argues for haplotype-aware breeding strategies in which the choice of SNI1 variant and growth regime are deployed together predictably across germplasm. More broadly, the work provides a concrete molecular mechanism for a long-standing evolutionary idea: that environmentally modulated recombination can accelerate adaptation by rebalancing the number and placement of crossovers, thereby altering the supply of selectable variation. As climate change shifts temperature regimes across growing seasons, a cold-sensitive brake on the class II pathway may prove to be one of the levers by which plants, and potentially the crops we depend on, tune their own genetic futures.

Subject of Research: Temperature-dependent control of meiotic crossover pathway choice by the SMC5/6 component SNI1 in Arabidopsis

Article Title: SNI1/NSE5 is a cold-sensitive brake on class II crossovers in Arabidopsis

Article References: SNI1/NSE5 is a cold-sensitive brake on class II crossovers in Arabidopsis. (n.d.). https://doi.org/10.1038/s41477-026-02400-x

Image Credits: AI Generated

DOI: 10.1038/s41477-026-02400-x

Keywords: meiotic recombination, crossovers, SNI1, SMC5/6 complex, Arabidopsis thaliana, temperature response, class II crossovers, synaptonemal complex, QTL mapping, subtelomeric recombination, plant genetics, climate adaptation