Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2013 Dec 20;8(12):e85075.
doi: 10.1371/journal.pone.0085075. eCollection 2013.

Cytological Studies of Human Meiosis: Sex-Specific Differences in Recombination Originate At, or Prior To, Establishment of Double-Strand Breaks

Free PMC article

Cytological Studies of Human Meiosis: Sex-Specific Differences in Recombination Originate At, or Prior To, Establishment of Double-Strand Breaks

Jennifer R Gruhn et al. PLoS One. .
Free PMC article


Meiotic recombination is sexually dimorphic in most mammalian species, including humans, but the basis for the male:female differences remains unclear. In the present study, we used cytological methodology to directly compare recombination levels between human males and females, and to examine possible sex-specific differences in upstream events of double-strand break (DSB) formation and synaptic initiation. Specifically, we utilized the DNA mismatch repair protein MLH1 as a marker of recombination events, the RecA homologue RAD51 as a surrogate for DSBs, and the synaptonemal complex proteins SYCP3 and/or SYCP1 to examine synapsis between homologs. Consistent with linkage studies, genome-wide recombination levels were higher in females than in males, and the placement of exchanges varied between the sexes. Subsequent analyses of DSBs and synaptic initiation sites indicated similar male:female differences, providing strong evidence that sex-specific differences in recombination rates are established at or before the formation of meiotic DSBs. We then asked whether these differences might be linked to variation in the organization of the meiotic axis and/or axis-associated DNA and, indeed, we observed striking male:female differences in synaptonemal complex (SC) length and DNA loop size. Taken together, our observations suggest that sex specific differences in recombination in humans may derive from chromatin differences established prior to the onset of the recombination pathway.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.


Figure 1
Figure 1. Genome-wide mean MLH1 values in human males and females.
MLH1 foci were used as a marker for meiotic recombination events. (A) shows a representative pachytene oocyte and (B) a representative pachytene spermatocyte, that were immunostained for SYCP3 (in red), a component of the synaptonemal complex; CREST (in blue), detecting centromeric regions; and MLH1 foci (in green), detecting crossovers. (C) In total 4660 spermatocytes from 56 males (white) and 2038 oocytes from 63 females (black) were examined. Mean MLH1 counts (± S.E.) were significantly lower in males than females (49.09 ± 0.07 vs. 69.25 ± 0.29; t=92.5, p<0.0001) and the range was narrower in males than females (30-66 vs. 27-119). (D) Mean number (± S.E.) of MLH1 foci per cell for individual male and female samples, demonstrating the lack of overlap between the sexes, and the increased variation in individual female cases by comparison with males.
Figure 2
Figure 2. Chromosome-specific MLH1 values in males and females.
Ten representative large, medium and small chromosomes were identified by FISH and chromosome-specific MLH1 values determined. For all ten chromosomes, mean values were lower in males (white) than females (black), and for nine of the ten the differences were statistically significant: chromosome 1 (10 males, n of cells =139; 7 females, n=83; t=14.1; p<0.0001), 6 (5 males, n=174; 3 females, n=30; t=8.2; p<0.0001), 13 (4 males, n=139; 8 females, n=109; t=8.4; p<0.0001), 14 (4 males, n=142; 4 females, n=70; t=3.2; p<0.005), 15 (6 males, n=130; 3 females, n=45; t=3.3; p<0.005), 16 (12 males, n=204; 9 females, n=63; t=8.8; p<0.0001), 18 (2 males, n=187; 7 females, n=100; t=6.9; p<0.0001), 21 (10 males, n=302; 11 females, n=218; t=7.1; p<0.0001), and 22 (10 males, n=313; 11 females, n=161; t=4.4; p<0.0001). The difference did not reach significance for chromosome 9 (5 males, n=58; 2 females, n=34; t=1.9; p=0.064).
Figure 3
Figure 3. Chromosome-specific MLH1 localization patterns in males and females.
The chromosomal locations of MLH1 foci were determined using the same cells as in Figure 2 for ten representative large, medium and small chromosomes. Each chromosome arm was arbitrarily divided into five equal regions – centromeric, proximal, interstitial, distal, and telomeric – and the distribution of MLH1 foci recorded for both chromosome arms for metacentric and sub-metacentric chromosomes or for the q-arm only of acrocentric chromosomes. The distribution differed significantly between females (black) and males (white) for seven of the ten chromosomes: 1 (χ2=24.9; p<0.005), 6 (χ2=24.8; p<0.005), 13 (χ2=13.8; p=0.01), 16 (χ2=32.1; p<0.0001), 18 (χ2=47.7; p<0.0001), 21 (χ2=22.3; p<0.0001), and 22 (χ2=20.8; p<0.0001). However, sex-specific differences were not evident for chromosomes 9 (χ2=15.1; p=0.088), 14 (χ2=5.3; p=0.262), or 15 (χ2=7.8; p=0.101).
Figure 4
Figure 4. Spacing between adjacent MLH1 foci in males and females.
Inter-focal distances, calculated as the percent of the length of the synaptonemal complex between adjacent MLH1 foci, were determined using the same cells as in Figure 2; male data are depicted in white, female data in black. To obtain sufficient numbers of cells for direct male:female comparisons, we restricted our analysis to chromosomes having the same number of MLH1 foci in males and females; i.e., for chromosome 1 we analyzed cells in which the chromosome exhibited four MLH1 foci and for chromosomes 13, 14, 16, 18 and 22, cells in which the relevant chromosome exhibited two MLH1 foci. Thus, for chromosome 1, we made three measurements of inter-focal distances per cell, while for chromosomes 13, 14, 16, 18 and 22 we made a single measurement of inter-focal distance per cell. For chromosomes 6, 9, 15 and 21 we had a limited number of cells with the same number of MLH1 foci in both sexes; thus, these chromosomes were excluded from the analysis. For each chromosome, inter-focal distances were binned (by % value) into ten groups. The distribution of categories of inter-focal distances was significantly different between males and females for each of the six chromosomes: 1 (χ2=51.7; p<0.0001), 13 (χ2=26.7; p<0.0005), 14 (χ2=30.6; p<0.0001), 16 (χ2=31.9; p<0.0001), 18 (χ2=50.0; p<0.0001), and 22 (χ2=48.8; p<0.0001).
Figure 5
Figure 5. Comparison of SC length and DNA loop size in males and females.
Male data are in white, female data in black. (A) Chromosome-specific SC lengths were determined for cells scored in Figure 2 and striking sex-specific differences were evident on all ten chromosomes analyzed: 1 (t=8.9; p<0.0001), 6 (t=23.1; p<0.0001), 9 (t=12.8; p<0.0001), 13 (t=26.7; p<0.0001), 14 (t=30.8; p<0.0001), 15 (t=24.1; p<0.0001), 16 (t=21.7; p<0.0001), 18 (t=28.6; p<0.0001), 21 (t=36.4; p<0.0001), and 22 (t=35.4; p<0.0001). (B, C) Three individual chromosomes (1, 16 and 21) were analyzed for DNA loop size, using the deflection of FISH paint probes from the SC as a surrogate for loop size. For chromosomes 1 and 16, we measured the width of the FISH signal at the centromere and three points on each chromosome arm, and averaged the seven values. For chromosome 21, loop size was taken as the average of three measurements, one at the centromere and two on the long arm. (B) Blow-up image of a portion of a representative pachytene stage oocyte, labeled with DAPI (blue) and a chromosome 1 paint probe (red). White bars represent the seven individual DNA loop measurements, three from each chromosome arm and one at the centromere. The centromere was identified using CREST prior to FISH. (C) DNA loop size means were significantly greater in males for each chromosome; i.e., for chromosome 1 (2 males, n of cells=24; 3 females, n=23; t=15.2; p<0.0001), for 16 (2 males, n=39; 3 females, n=32; t=20.8; p<0.0001) and for 21 (2 males, n=37; 2 females, n=43; t=16.0; p<0.0001).
Figure 6
Figure 6. Genome-wide RAD51 values in males and females.
RAD51 foci were used as a surrogate for DSBs and the number of foci in leptotene stage cells determined [44 cells from 3 males (white triangles) and 39 cells from 5 females (black circles)]. The male and female distributions were virtually non-overlapping, with almost all spermatocytes exhibiting fewer than 200 RAD51 foci and most oocytes more than 200 foci. Further, the mean number of RAD51 foci (± S.E.) was significantly lower in males than in females (134.07 ± 5.47 and 250.28 ± 10.21, respectively; t=10.4; p<0.0001).
Figure 7
Figure 7. Synaptic initiation complexes (SCISs) in zygotene meiocytes.
(A) Representative zygotene spermatocyte (left), with antibodies against SYCP3 (detecting the axial/lateral elements of the SC) in red and CREST antiserum-positive signals (recognizing centromeric regions) in blue. Intense red signals indicate points where the axial/lateral elements have merged, consistent with full synapsis (i.e., SCISs). SCISs are located at, or near the telomeres, an arrangement typical for human males. The center panel shows a blow-up of a partially-synapsed bivalent (circled) and the right panel provides a schematic of the bivalent, with the synapsed regions at the ends of the arms and the proximal regions (including the centromeres) asynapsed. (B, C) Representative images of zygotene stage oocytes (left), with the axial/lateral element protein SYCP3 in red, the transverse element protein SYCP1 in green and centromere-associated CREST in blue; merged SYCP3/SYCP1 signals (yellow) indicate regions of synapsis (i.e., SCISs). Center panels provide blow-ups of partially synapsed bivalents (circled) and the right panel schematics, demonstrating the presence of multiple SCISs per chromosomes and in (C), co-localization of the centromere and one of the SCISs.

Similar articles

See all similar articles

Cited by 30 articles

See all "Cited by" articles


    1. Hassold T, Hunt P (2001) To err (meiotically) is human: the genesis of human aneuploidy. Nat Rev Genet 2: 280–291. doi:10.1038/35066065. PubMed: 11283700. - DOI - PubMed
    1. Hassold T (1986) Chromosome Abnormalities in Human Reproductive Wastage. Trends Genet 2: 105–110. doi:10.1016/0168-9525(86)90194-0. - DOI
    1. Hassold T, Abruzzo M, Adkins K, Griffin D, Merrill M et al. (1996) Human aneuploidy: Incidence, origin, and etiology. Environ Mol Mutagen 28: 167–175. Available online at: doi:10.1002/(SICI)1098-2280(1996)28:3<167:AID-EM2>3.0.CO;2-B. - DOI - PubMed
    1. Nagaoka SI, Hassold TJ, Hunt PA (2012) Human aneuploidy: mechanisms and new insights into an age-old problem. Nat Rev Genet 13: 493–504. doi:10.1038/nrg3245. PubMed: 22705668. - DOI - PMC - PubMed
    1. Ross LO, Maxfield R, Dawson D (1996) Exchanges are not equally able to enhance meiotic chromosome segregation in yeast. Proc Natl Acad Sci U S A 93: 4979–4983. doi:10.1073/pnas.93.10.4979. PubMed: 8643515. - DOI - PMC - PubMed

Publication types