Mammalian spermatogenesis at single-cell resolution downstream of genetic perturbations in meiotic double-strand break repair
Ipsita AgarwaliD, Brynelle Myers, Martin HoulardiD, Anjali Gupta HinchiD, Emmanuelle BitouniD, Simon Robert MyersiD
Abstract
Fertility in mammals relies on successful pairing of homologous chromosomes mediated by DNA double-strand breaks (DSBs). Here, we develop a system of mouse hybrids in which (a)symmetry of binding by the break-positioning protein PRDM9 to homologs varies over broad scales in the genome. Profiling transcription and chromatin in single nuclei, we trace how resulting delays in repair of variable subsets of meiotic DSBs propagate through spermatogenesis and drive large fertility differences in animals. We demonstrate that only asymmetry-generating mutations in PRDM9-binding motifs, not high average (~1%) divergence, disrupt chromosomal pairing. We observe substantial variation in animal-level sensitivity to asymmetry, and identify an interacting locus containing Dmc1 and Mei1 controlling (R2=0.64) this variation. Silencing of unpaired autosomes downstream of asynapsis and failure of normal sex chromosome silencing independently explain cell death in pachytene. Surprisingly, many cells with synaptic defects evade cell cycle arrest, and even those where physical division arrests still exhibit transcriptional progression to post-division states. Attrition of abnormal cells via arrest continues beyond the first division; nonetheless, cells that complete both meiotic divisions exhibit aneuploidy, especially of the sex chromosomes. This partly reflects de novo segregation errors explained by silencing of only chromosomes 16 and 19. Even "normal" euploid spermatids show crossovers redistributed at multi-megabase scales, indicating novel and potentially post-zygotic impacts of delays in meiotic DSB-repair. We thus elucidate cell-level and chromosome-specific impacts of regulatory variation in ~0.03% of the genome cascading through germline development, advancing our understanding of fertility and reproductive isolation.
02 /Specification & Classification
Click any tag to browse matching papers- Citations
- 0
- Organism
- Assay
- Research Focus
- Approach
In-depth summary
Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 2 (11 Sept 2026) · CC BY 4.0
Single-nucleus RNA+ATAC profiling of 32 mice shows asymmetric PRDM9 binding drives asynapsis, pachytene cell loss, aneuploidy and multi-megabase crossover redistribution, tuned by a chromosome 15 locus.
Asymmetric PRDM9 binding was known to cause asynapsis and hybrid male sterility, but it remained unclear whether divergence outside hotspots contributes, whether affected cells survive to transmit aneuploidy, and whether broad-scale crossover rates shift. Hybrids with locally varying PRDM9 symmetry at near-constant heterozygosity let each downstream meiotic consequence be quantified in single nuclei.
Key results
ShownAsynapsis depends on PRDM9 binding asymmetry rather than genome-wide divergence. CAST segments, where the B6 allele binds both homologs equally, rescue synapsis in cis. Chromosome silencing inferred from RNA counts works as a validated proxy for asynapsis.
In PWD/B6 regions the B6-derived PRDM9 allele bound the PWD vs B6 homologs at 80%/20%; in PWD/CAST regions it bound at 51%/49%. RNA-based silencing matched FISH asynapsis rates in PWD x B6 mice (98.7%). A cis exponential-decay model based on CAST length explained ~75% of the variance across chromosomes (r = 0.86, p < 2x10^-15). · PRDM9 binding at recombination hotspots in early…; Chromosome-specific transcriptional silencing in…; PRDM9 binding symmetry and a region containing…
SuggestedA chromosome 15 interval containing Dmc1 and Mei1 acts in trans to set how sensitive each animal is to PRDM9 asymmetry. CAST ancestry in this interval goes with stronger silencing, and likely more asynapsis.
Relative silencing efficacy across 23 mice was associated with an 18.8 Mb interval (63.5 - 82.4 Mb; r = 0.8, p = 1.9x10^-5; permutation p=0.002). The abstract reports R^2=0.64. The interval holds 309 genes, and Mei1 carries 13 non-synonymous B6/CAST changes. No clear allele-specific expression was seen for the candidate genes. · Abstract; PRDM9 binding symmetry and a region containing…; PRDM9 binding asymmetry predicts downstream…
ShownFertility loss runs through pachytene cell death. Autosomal asynapsis and X-silencing failure jointly predict late-pachytene survival, which alone predicts sperm count; testes weight also tracks a chromosome 11 locus. X-silencing failure follows genome-wide asymmetry, not autosomal competition.
The joint model predicted 79% of late-pachytene survival. Among upstream factors, only pachytene survival predicted sperm count and testes weight; a 6.5 Mb chromosome 11 locus (r = -0.73, p = 6.5 x 10^-5) added to testes weight prediction. X-silencing failure tracked genome-wide B6 proportion (ρ = 0.61; p = 0.002), not single-cell autosomal silencing (ρ = -0.06). · Similar, dynamic, chromatin alterations occur on…; PRDM9 binding asymmetry predicts downstream…
ShownSome cells escape pachytene arrest and divide while carrying aneuploidies. Autosomal MI errors arise in cis on asymmetric chromosomes, but the cause of MI X/Y missegregation is unclear. MII deletions of autosomes and X/Y appear driven in trans by asynapsis of chromosomes 16 and 19.
Over 15% of MI cells had sex chromosome abnormalities (vs 1% in B6 x CAST), and 89% of these were inferred to arrest at or before MII. MII cells showed 2% X/Y abnormalities. Asymmetric binding raised complete MI deletions (OR = 2.08; p < 0.001). Asynapsis of chromosomes 16 and 19 predicted MII autosomal deletions (R^2 = 0.92) and X/Y deletions (R^2 = 0.85). · Aneuploidy of the sex chromosomes and of…
ShownPRDM9 binding symmetry reshapes broad-scale crossover maps. On asynapsis-prone chromosomes, crossovers shift into CAST segments rich in symmetric hotspots, which the authors read as evidence that crossover designation begins early, before synapsis is complete.
245,368 crossovers were mapped in 24 mice. Crossover enrichment in CAST vs B6 segments was >1 on all autosomes except 9 and 14, which carry introgressed domesticus segments. Enrichment correlated with PWD x B6 asynapsis rates (r = 0.91; p = 7.7x10^-7) and reached ~2-fold on chromosomes 18 and 19, vs the 1.26-fold expected from extra DSBs. · (A)symmetric binding by PRDM9 reprograms…
Methods, data and reproducibility
10x Genomics single-nucleus multiome (RNA+ATAC) on FACS-enriched nuclei from male germline tissue. Pseudotime comes from Gibbs-sampled Dirichlet expression components; reads are assigned to strains via ancestry-informative SNPs; HMMs infer ploidy and crossovers; and a statistical MSUC model estimates per-chromosome silencing as a proxy for asynapsis.
- Male germline tissue from PWD x (B6 x CAST) hybrids carrying B6 and PWD Prdm9 alleles, with varying paternal CAST ancestry · mouse · n = 23
- Sterile PWD x B6 hybrids (B6 and PWD Prdm9 alleles) · mouse · n = 2
- Fertile B6 x CAST hybrids carrying the human Prdm9 allele · mouse · n = 2
- Parental strains B6 and CAST · mouse · n = B6 n=4; CAST n=1
- Single nuclei passing QC with paired RNA-seq and ATAC-seq reads across spermatogenesis · mouse · n = 203,869 cells
- Previously identified PWD- or B6-allele-controlled recombination hotspots with identified PRDM9 motifs · mouse · n = 12,175 hotspots
- Published FISH-based chromosome-specific asynapsis rates in PWD x B6 males, used for validation · mouse
- Published PRDM9-associated H3K4me3 asymmetry estimates at hotspots in PWD x B6 animals · mouse
- Methods & tools
- 10x Genomics multiome assay (snRNA-seq + snATAC-seq), FACS nuclei sorting, Gibbs sampling with Dirichlet priors (75 time-varying expression components), Linear-interpolation pseudotime ranking, Ancestry-informative SNP read assignment, Hidden Markov Model for ploidy and ancestry state, Hidden Markov Model for crossover mapping, Statistical MSUC silencing model with overdispersion, Exponential-decay cis asynapsis model, Permutation testing for genome-wide association, Principal components analysis
- Reproducibility gaps
- Raw and processed data are promised on publication. Analysis and figure code is to be posted on GitHub (agarwali/spermatogenesis_single_cell); no URL or release was given at posting.
Limitations (6)
- There is no direct measurement of DSB formation or synapsis. Asynapsis is inferred from transcriptional silencing, which slightly underestimates FISH-measured rates, possibly because silencing is incomplete or asynaptic cells are lost. · stated by the authors
- The causal variant in the chromosome 15 locus is not identified. Dmc1 and Mei1 are candidates only, and fine-mapping is left for future work. · stated by the authors
- The sparse ATAC data cannot resolve individual DSB sites or place crossovers at single hotspots. · stated by the authors
- MI cell counts are modest (2,944 cells), attrition of arrested cells is unknown, and the proxy for arrest may be imperfect. The cause of X/Y missegregation remains unresolved. · stated by the authors
- Locus mapping and conditional causal models rest on 23 PWD x (B6 x CAST) males with no replication cohort, so variance-explained estimates may be optimistic. · noted in this summary
- FACS enrichment of early/mid-meiotic nuclei alters stage proportions, which could bias the pachytene survival measure, defined as a proportion of cells, that is used to predict fertility. · noted in this summary
Who should read:Read this if you study meiotic recombination, PRDM9, hybrid sterility or MSCI, or want a single-nucleus multiome framework for reading synapsis, aneuploidy and crossovers from germ cells.
Reproducibility ledger
Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.
Sign in to file a reproducibility report.
Related papers
- A high-resolution map of non-crossover events reveals impacts of genetic diversity on mammalian meiotic recombination
Ran Li, Emmanuelle Bitoun, … Simon Robert Myers·Nature CommunicationsPeer-reviewed·Aug 29, 2019
- Modulation of Prdm9-controlled meiotic chromosome asynapsis overrides hybrid sterility in mice
S. Gregorova, V. Gergelits, … J. Forejt·bioRxiv·Dec 11, 2017
- Sexual dimorphism in the meiotic requirement for PRDM9: a mammalian evolutionary safeguard
Natalie R. Powers, Beth L. Dumont, … Tanmoy Bhattacharyya·Science AdvancesPeer-reviewed·Oct 23, 2020
- Ensuring meiotic DNA break formation in the mouse pseudoautosomal region
Laurent Acquaviva, Michiel Boekhout, … Scott Keeney·NaturePeer-reviewed·May 27, 2020
- Genomic and chromatin features shaping meiotic double-strand break formation and repair in mice
Shintaro Yamada, Seoyoung Kim, … Scott Keeney·Cell CyclePeer-reviewed·Aug 18, 2017
- Defects in Meiotic Recombination Delay Progression Through Pachytene in Mouse Spermatocytes
James H. Crichton, David Read, Ian R. Adams·ChromosomaPeer-reviewed·Jun 15, 2018