Figure 1.
A. XY. The Y short arm (Yp) gene complement of an XY male (represented to scale in the magnified view) comprises seven single copy genes, two duplicated genes and one multi copy gene. The pseudoautosomal region (PAR) located distally on the Y long arm mediates pairing and crossing over with the X PAR during meiosis to generate the XY sex bivalent. B–D. The diminishing Yp gene complements for the three XO male mouse models that lack the Y long arm. B. XSxraO. The Yp-derived Sxra sex-reversal factor, attached distal to the X PAR provides an almost complete Yp gene complement. C. XSxrbO. The Sxra-derived deletion variant Sxrb has a 1.3 Mb deletion (ΔSxr-b) removing 6 single copy genes and creating a Zfy2/1 fusion gene spanning the deletion breakpoint (†). D. XOSry. This model has only one Y chromosome gene, namely the testis determinant Sry provided as an autosomally located transgene. E. Y*X. This mini sex-chromosome is an X chromosome with a deletion from just proximal to Amelx to within the DXHXF34 repeat adjacent to the X centromere. † represents the deletion breakpoint. This X chromosome derivative has a complete PAR that can pair with the PAR of XSxra, XSxrb or X to form a ‘minimal sex bivalent’. Scale bar for magnified views is 150 kb.
Figure 2.
Efficiency of XY synapsis in the XY*X males with varying Yp complements.
Spread pachytene spermatocytes from 6 week old XY and XEY*XSry testes stained with antibodies against SYCP3 (green) and CREST (red). Frames show PAR-PAR sex chromosome synapsis in XY and XEY*XSry males; the arrow points to an unsynapsed Y*X chromosome in an XEY*XSry male.
Table 1.
X-Y pairing efficiency in pachytene spermatocytes.
Figure 3.
Zfy1 and Zfy2 promote meiosis-II in the presence of the sex chromosome pairing partner Y*X.
Data collected after DNA quantitation of spermatids using DAPI fluorescence intensity measurement on SYCP3-labelled testis cell spreads. Pooled data expressed as percentages are shown for each genotype (n = 4). Key: in black, models with robust apoptotic elimination of MI spermatocytes with X univalents; in white, XO models with markedly reduced apoptotic response; striped, XY*X models with markedly reduced apoptotic response in which the frequency was adjusted to remove spermatids derived from MI spermatocytes that had not formed an X-Y*X bivalent by PAR-PAR synapsis (see Table S1). A. Percentage of haploid round spermatids found in testis of 6 week old XO and XY*X males with various Yp chromosome gene contents. The data for the two XO male genotypes derive from Vernet et al., 2012 [14]. The Yp-derived Sxrb (which includes a Zfy2/1 fusion gene encoding a ZFY1-like protein) and Sxra (which includes Zfy1 and Zfy2) promote meiosis II in the presence of Y*X; Sxra is substantially more effective than Sxrb. B. Percentage of haploid round spermatids found in testis of 6 week old XEY*XSry males with X-linked Zfy transgene additions. Zfy1, and to a greater extent Zfy2, promote meiosis II. NS, Non significant; *p≤0.05; **p≤0.01; ***p≤0.001.
Figure 4.
The mouse Zfy and Zfx genes are transcribed in interphasic secondary spermatocytes.
Representative images of interphasic secondary spermatocyte nuclei are shown hybridized with RNA FISH probes specific for Zfy1, Zfy2 or Zfx (arrows, top panels). Interphasic secondary spermatocytes were distinguished from diploid spermatids by staining spread spermatogenic cells from 6-week old XY males with an antibody against SYCP3 (red, top panels). The appropriate localization of the RNA FISH probe to the encoding genes was confirmed by DNA FISH (arrows, bottom panels). Nuclei are stained with DAPI (blue). X- or Y-bearing secondary spermatocytes are respectively represented by an X or a Y next to the cell.
Table 2.
X and Y-linked ‘Zf’ gene expression by RNA-FISH in spread interphasic secondary spermatocytes from 6 week old XY male.
Figure 5.
Zfx over-expression promotes meiosis II.
A. Spread cells found in testis of 6-week old XEY*XSry males without or with Zfx transgene. Pachytene (Pa), diploid spermatid (St d) and haploid spermatids (St h) nuclei are stained with DAPI (top panel) and higher magnifications are shown additionally labelled with γH2AFX, and SYCP3 antibodies (bottom panel). B. Percentage of haploid round spermatids found in mice from panel A (see also Table S2D). Key: in black, XEY*XSry,Zfx transgenic males have a robust apoptotic elimination of the ∼25% of MI spermatocytes that have an X univalent (see Figure S5); striped, XEY*XSry males have a markedly reduced apoptotic response so the frequency was adjusted as detailed in Table S1. The addition of the Zfx transgene significantly increases (***p≤0.00001) the proportion of haploid round spermatids.
Figure 6.
Zfy2 acidic domain is a much more potent transactivator than other ‘Zf’ acidic domains.
Levels of β-galactosidase induced by the Gal4-DNA-binding domain on its own (pGB-CEN6; negative control) or fused to an acidic domain from one of six different ZF isoforms from human (hs) or mouse (mm). Among the mouse sex-linked genes, Zfy2 has a substantially more potent activation domain than Zfy1, and Zfx is significantly less potent than Zfy1. mm ZFA derives from the autosomal Zfa gene that originated from a retroposed X transcript. *p≤0.05; **p≤0.01.
Figure 7.
A summary of the meiotic outcome in XO and XY*X males with varying Yp gene content.
Throughout this figure the thickness of the arrows indicates the proportion of cells progressing from one step to the next and the cheeses at the bottom represent the proportion of haploid and diploid spermatids. Size of the cheese indicates the relative success of the different models in meiosis completion. A. XO models. In XEOSry and XESxrbO males the majority of spermatocytes complete meiosis I because of the reduced apoptotic response at MI due to the absence of Zfy2 [11]. Zfx expression is likely responsible for the residual apoptotic response (see Figure S5). The majority of spermatocytes then arrest at the interphase between meiosis I and meiosis II. This could be a consequence of the prior triggering of the MI SAC by the univalent X, the reduced apoptotic response due to the absence of Zfy2, or the lack of a Yp gene or genes that promotes meiosis II. In XSxraO males there is a very efficient apoptotic elimination of spermatocytes at MI so that very few complete meiosis I and this results in a 97% reduction in the number of spermatids. This precludes any firm conclusion as to a role for Yp genes for completion of meiosis II because the apoptotic elimination may have had a bias towards removing MI cells that were otherwise destined to arrest at the following interphase. B. XY*X models. In these models the spermatocytes that form a sex bivalent circumvent the MI SAC/apoptotic response and complete meiosis I. This reveals that Sxra strongly promotes meiosis II, thus confirming that a gene or genes on Yp promotes meiosis II. Surprisingly Sxrb, which did not promote meiosis II in XESxrbO males, does so now that the apoptotic response is circumvented by formation of an X-Y*X sex bivalent. C. The XY*XSry ‘Zf’ transgene addition models. These transgene additions revealed that Zfy1 and Zfy2 are the genes on Yp that promote meiosis II with Zfy2 the more effective. Sxrb includes the Zfy2/1 fusion gene that encodes a ZFY1-like protein, whereas Sxra includes Zfy1 and Zfy2, thus explaining the more potent effect of Sxra in promoting meiosis II. Zfx over-expression also promotes meiosis II (Figure 5) making it likely that the endogenous Zfx also does so to some degree.
Figure 8.
A combined MI SAC and G2/MII checkpoint model to explain the consequences of the male-specific apoptotic response to spermatocytes with univalent chromosomes at MI.
To illustrate the model we consider the consequences of these two checkpoint responses in XSxraO, XY*XSxra and XY males, all of which have a complete ‘Zf’ gene complement. Red arrows denote DNA damage. The cheeses at the bottom represent the proportion of haploid and diploid spermatids. A. In XSxraO males each MI spermatocyte will have a univalent X that is expected to trigger the MI SAC and cause a brief delay in MI progression. We propose that this delay is detected by the surrounding Sertoli cell, which initiates a robust Zfy2+Zfx-dependent apoptotic response. In order to explain the mix of diploid and haploid spermatids originating from the very few surviving MI spermatocytes we propose: 1) Rare MI spermatocytes complete meiosis I with apoptotic DNA damage that triggers a G2/MII checkpoint in the subsequent interphase and blocks progression to MII – these interphasic secondary spermatocytes then enter spermiogenesis to form diploid spermatids. [The number of such cells was elevated in the XO models lacking Zfy2 in Figure 7]. 2) In rare cases some MI spermatocytes evade the MI SAC and apoptosis by achieving bipolar attachment to the spindle – these complete meiosis I and meiosis II to form haploid spermatids. B. In XY*XSxra males (25% of MI spermatocytes with univalent X and Y*X) and XY males (5% of MI spermatocytes with univalent X and Y) the MI cells with univalents will follow the pathways 1) and 2) above, although the likelihood of both univalents achieving bipolar attachment to the spindle will be much lower. The MI cells that have formed a sex bivalent will progress through both divisions to form haploid spermatids (unless they have a pair of autosomal univalents in which case pathways 1 and 2 apply).