Figure 1.
Expression and localization of BAZ1A in male germ cells.
(A, B) Equal amounts of low-speed centrifuged, nuclear-enriched extracts from a panel of mouse tissues (8.5 µg protein) (A) or cultured human (U2OS and HeLa) and mouse (B16 and ear fibroblasts (EF)) cells (25 µg) (B) were immunoblotted with anti-BAZ1A, SNF2H and BAZ1B antibodies. Approximate molecular masses: BAZ1A = 178 kDa, SNF2H = 122 kDa, BAZ1B = 171 kDa and ACTIN = 42 kDa. +/+, wild-type; −/−, Baz1a−/−. (C) Immunofluorescence of testis sections with BAZ1A and γH2AX antibodies. (i–iii) Testis sections at various stages (indicated by uppercase Roman numerals). Bar = 20 µm. (i′–iii′) Magnification of areas indicated in left panels. Bar = 10 µm. Arrows point to sex bodies. Note that anti-γH2AX staining along the heads of elongating spermatids at stage X-XI and the focal signal in round spermatids at stage VIII is non-specific due to cross-reactivity with the developing acrosome. ES, elongating spermatid; RS, round spermatid; Z, zygotene spermatocyte; EP, early pachytene spermatocyte; S, Sertoli cell; P, pachytene spermatocyte; L, leptotene spermatocyte; pL, preleptotene spermatocyte; M, peritubular myoid cell; D, diplotene spermatocyte. (D) Schematic summarizing BAZ1A expression during spermatogenesis, with darker shades of red indicating intensity of BAZ1A staining. Micrographs are from insets in Figure S2B; panels a–d are co-stained with HP1β, a marker of heterochromatin (see Figure S2B and text for further explanation); panels e–h display higher exposures of the BAZ1A staining. RS, round spermatid; ES, elongating spermatid; CS, condensing spermatid. Abbreviations of meiotic stages as in (C). See text for full description of staining patterns. (E) Coimmunoprecipitation (IP) of SNF2H and BAZ1A from wild-type whole testis lysates. (F) Immunofluorescence of testis sections with BAZ1A and SNF2H antibodies. Magnifications of the indicated cells are shown. SP, spermatogonia; *autofluorescent red blood cells; bar = 20 µm. (G) BAZ1A immunofluorescence on squashed spermatocyte nuclei at different stages of meiotic prophase I as indicated by staining for the axial element component SYCP3. Bar = 5 µm.
Figure 2.
(A) Schematic of generation of Baz1a conditional allele. (i) Partial Baz1a genomic locus. (ii) Targeting vector. (iii) Conditional allele. (iv) Deletion allele following Cre-mediated recombination of the loxP sites (open triangles) flanking exon six. Black triangles, FRT sites; N, PGK-neo; TK, hsv-thymidine kinase; P1, P2 and P3, PCR primers. Not to scale. (B) PCR genotyping of tail DNA. M, marker. (C) Immunoblotting of BAZ1A immunoprecipitated from whole testis lysates. Mock IP, no antibody; No Lysate, IP from buffer. (D) Anti-BAZ1A immunofluorescence on testis sections from wild-type and mutant mice. Bar = 20 µm. (E–G) Immunoflurescence with indicated antibodies on squash preparations of diplotene spermatocyte nuclei. Bar = 10 µm.
Figure 3.
Baz1a mutants can repair meiotic DSBs.
(A, B) Squash preparations of spermatocyte nuclei showing accumulation of γH2AX in response to meiotic DSBs, and its disappearance from autosomes as DSBs are repaired. Arrows, sex bodies; bar = 10 µm. (C) DMC1 focus counts on spread spermatocyte nuclei pooled from two sets of wild-type and mutant mice. Nuclei were staged by co-staining for the axial element protein SYCP3. Early zygotene was the only stage in which a statistically significant difference (P<0.05) was observed (P value from t-test; n = number of nuclei). (D) Spread chromosomes from Baz1a−/− pachytene spermatocyte showing MLH1 foci. (E) MLH1 focus counts pooled from two sets of wild-type and mutant littermates (P value from t-test; n = number of nuclei). (F) Primary/primordial and total follicles in Baz1a−/− and heterozygote littermate controls. Ovaries were dissected from one Baz1a−/− or two Baz1a+/− females at three months of age and stained with anti-MVH antibodies to detect oocytes. Baz1a+/− mice displayed no apparent phenotype in any of the cell types analyzed in this study, so they served as normal controls for these experiments. Each point is the number of follicles of the indicated type per histological section. (P values from Welch's t test; n = number of ovary sections.) (G, H) Anti-MVH-stained ovary sections from three-month old mice of the indicated genotypes. Examples at various stages of folliculogenesis are indicated: A, antral follicle; S, secondary follicle; P, primary/primordial follicle. Bar = 100 µm.
Figure 4.
Repair of endogenously and exogenously induced DSBs in BAZ1A-deficient somatic cells.
(A) Representative FACS plots of dissociated thymocytes and splenocytes expressing CD4 and CD8, implying completion of V(D)J recombination. (B) Quantification of thymocytes and splenocytes expressing CD4 and CD8 in Baz1a−/− as percent of wild-type (WT) littermates analyzed in parallel. N = 3 littermate pairs. (C) Representative FACS plots of B cells stimulated to undergo CSR to IgG1 or IgG3. (D) Cultured B cells stimulated to undergo CSR to IgG1 or IgG3 quantified in Baz1a−/− as a percentage of wild type. N = 3 littermate pairs. (E) DR-GFP reporter at the Pim1 locus used to assay HR. Cleavage of the I-SceI site by the I-SceI endonuclease followed by repair by HR leads to GFP+ cells. (F, G) HR is not significantly different in primary cultures of MEFs (F) or ear fibroblasts (G) from Baz1a−/−Pim1+/Drgfp or control (Baz1a+/+Pim1+/Drgfp or Baz1a+/−Pim1+/Drgfp) mice. MEFs were derived from individual embryos from a total of two Baz1a−/−Pim1+/Drgfp mice and three controls. For ear fibroblasts, three independent experiments were performed using cultures derived from individual mice. Error bars in B, D, F, G are mean ± s.d.
Figure 5.
Quantitative and morphological defects of Baz1a−/− sperm.
(A) Bright-field images of sperm heads overlaid with DAPI fluorescence (blue). Bar = 2 µm. (B) Bright-field images of sperm. Bar = 5 µm. (C–D) Total and motile sperm counts from total epididymides of two mice of each genotype (error bars, mean ± s.d. of triplicate counts). (E–H) Hematoxylin and eosin stained sections of caput (E, F) and cauda (G, H) epididymides. Bar = 20 µm.
Figure 6.
Abnormal spermatogenesis in the absence of Baz1a.
(A, B) Periodic Acid-Schiff and hematoxylin stained sections of stage IX seminiferous tubules showing sperm retention (arrows) in the mutant. Bar = 20 µm. (C) Mutant tubule with multi-nucleate round spermatids (arrows). Bar = 20 µm. (D) Quantification of multi-nucleate cells (error bars, mean ± s.d.). (E, F) Transmission electron microscopy of mutant testis sections showing a multi-nucleate round spermatid (E) and elongating spermatids (F). Arrows indicate the degree of acrosome stretching. N, nucleus; V, vacuole; bar = 2 µm. (G, H) TUNEL staining of testis sections. Bar = 100 µm. (I, J) Quantification of TUNEL-positive cells per tubule (I) (p value from t-test) and percentage of tubules with ≥1 TUNEL-positive cell per tubule (J) (p value from Fisher's exact test). Error bars, mean ± s.d.
Figure 7.
Baz1a is dispensable for spermatogenesis-associated changes in chromatin protein composition.
(A) Testis cells were fractionated into cytoplasmic (cyt), nucleoplasmic (nuc), chromatin bound (chr) and insoluble (insol) extracts and immunoblotted for a panel of testis-specific histone variants. α-tubulin is a cytoplasmic marker. (B) Immunofluorescence on testis sections with anti-acetyl histone H4 (Ac-H4) antibody. Bar = 50 µm. (C) HCl-extracted proteins from sonication resistant (Res.) and sensitive (Sen.) spermatids were precipitated with 4% and then 20% trichloroacetic acid (TCA), separated by acid/urea PAGE and stained with Coomassie (top panel). A duplicate gel was used for western blotting with antibodies against the indicated proteins (bottom panels).
Figure 8.
Wide-spread transcriptional perturbations in Baz1a-deficient spermatids.
(A) Differential expression in normal spermatids vs. spermatocytes. Each point plots expression of a microarray probe averaged between samples. Red, significantly up regulated in spermatids (≥2-fold, q = 0.05); blue, significantly down regulated; gray, not significantly different. Dashed lines indicate two-fold change. (B) Differential expression between Baz1a−/− and control spermatocytes (i) or spermatids (ii). Plotting and color code as in A. (C) Proportional area Venn diagrams showing overlap of differentially expressed probe sets. “Developmentally regulated” probes were differentially expressed in spermatids compared to spermatocytes (≥2-fold, q = 0.05); “mis-regulated” probes were differentially expressed in Baz1a−/− vs. control. (D, E) Heat maps of expression levels in all samples of developmentally regulated and mis-regulated probes. The 500 most differentially expressed probes were chosen from comparison of spermatids vs. spermatocytes in controls (“developmentally regulated”, D) or Baz1a−/− vs. control spermatids (“mis-regulated”, E). Samples were clustered hierarchically and probes were grouped by k-means clustering (k of 7 and 10 for D and E, respectively). Mouse number and Baz1a genotype of samples are indicated (+ = control; − = Baz1a−/−). (F) Box and whisker plots of expression changes in Baz1a−/− vs. control spermatocytes for the indicated probe sets. Boxes show median and interquartile range, whiskers extend from minimum to maximum values; p values are from Wilcoxon rank sum tests.