Mettl3-mediated m6A regulates spermatogonial differentiation and meiosis initiation

Abstract

METTL3 catalyzes the formation of N⁶-methyl-adenosine (m⁶A) which has important roles in regulating various biological processes. However, the in vivo function of Mettl3 remains largely unknown in mammals. Here we generated germ cell-specific Mettl3 knockout mice and demonstrated that Mettl3 was essential for male fertility and spermatogenesis. The ablation of Mettl3 in germ cells severely inhibited spermatogonial differentiation and blocked the initiation of meiosis. Transcriptome and m⁶A profiling analysis revealed that genes functioning in spermatogenesis had altered profiles of expression and alternative splicing. Our findings provide novel insights into the function and regulatory mechanisms of Mettl3-mediated m⁶A modification in spermatogenesis and reproduction in mammals.

Figure 1

Mettl3 is essential for male fertility and spermatogonial differentiation during spermatogenesis. (A) Confocal immunofluorescence detection of METTL3 by staining of the Mettl3^Ctrl and Mettl3^cKO testes at postnatal day 6 (P6). PLZF was co-stained to indicate the location of the undifferentiated spermatogonia. The DNA was stained with DAPI. White circles denote the Mettl3 null spermatogonia. Scale bar, 10 μm. (B) Morphological analysis of the 8-week-old Mettl3^Ctrl and Mettl3^cKO testes. Scale bar, 2 mm. (C) Testis weight of the 8-week-old Mettl3^Ctrl and Mettl3^cKO mice. Student's t-test, error bars indicate standard error of measurement (SEM). ^***P < 0.001, n = 8. (D) Hematoxylin eosin (H&E) staining of Mettl3^Ctrl and Mettl3^cKO testes at postnatal day 8 (P8), postnatal day 10 (P10), postnatal day 12 (P12) and 8 weeks showed that the spermatogonial differentiation was inhibited in Mettl3 knockout testes. Red arrows indicate the representative stages of the spermatocytes. A, type A spermatogonia; In, intermediate spermatogonia; B, type B spermatogonia; L, leptotene spermatocytes; Z, zygotene spermatocytes; P, pachytene spermatocytes. Left panel, P8, P10, P12, scale bar, 20 μm; 8 weeks, scale bar, 100 μm. Right panel, P8, P10, P12, scale bar, 5 μm; 8 weeks, scale bar, 20 μm. (E) Immunofluorescence co-staining of PLZF and DDX4 in Mettl3^Ctrl and Mettl3^cKO testes at P8. Scale bar, 20 μm. (F) Statistics results of DDX4-positive but PLZF-negative cells in Mettl3^Ctrl and Mettl3^cKO testes at P8. At least 100 tubules were counted from 3 different mice. Student's t-test, error bars indicate SEM. ^***P < 0.001. (G) Immunofluorescence staining of KIT in Mettl3^Ctrl and Mettl3^cKO testes at P8. Scale bar, 20 μm. (H) Statistics of Kit-positive cells in Mettl3^Ctrl and Mettl3^cKO testes at P8. At least 100 tubules were counted from three different mice. Student's t-test, error bars indicate SEM. ^***P < 0.001. (I) Immunofluorescence staining of KIT in Mettl3^Ctrl and Mettl3^cKO testes at P12. Scale bar, 20 μm. (J) Statistics of KIT-positive cells in Mettl3^Ctrl and Mettl3^cKO testes at P12. At least 100 tubules were counted from 3 different mice. Student's t-test, error bars indicate SEM. ^***P < 0.001.

Figure 2

Mettl3 deletion causes a severe defect in meiosis during spermatogenesis. (A) Immunohistochemical staining of STRA8 in Mettl3^Ctrl and Mettl3^cKO testes at P10 and P12. Left panel, scale bar, 100 μm. Right panel, scale bar, 5 μm. Statistics of the ratio of STRA8-positive tubule in Mettl3^Ctrl and Mettl3^cKO testes at P10 and P12. At least 750 tubules were counted from 3 different mice. Student's t-test, error bars indicate SEM. ^***P < 0.001. Statistics of the number of STRA8-positive cells per tubule in Mettl3^Ctrl and Mettl3^cKO testes at P10 and P12. At least 50 tubules were counted from 3 different mice. Student's t-test, error bars indicate SEM. ^***P < 0.001. (B) Immunofluorescence staining of SYCP3 in Mettl3^Ctrl and Mettl3^cKO testes at P12. DNA was stained with DAPI. Scale bar, 20 μm. (C-E) Immunofluorescence staining for detection of SYCP3 and RAD51 on nuclear surface spreads of leptotene (C), zygotene-like (D) and pachytene (E) spermatocytes from Mettl3^Ctrl and Mettl3^cKO testes at P12. Scale bar, 5 μm. (F) Statistics of the proportion of leptotene, zygotene/zygotene-like, pachytene spermatocytes in Mettl3^Ctrl and Mettl3^cKO testes at P12. At least 1 000 cells were counted from 3 different mice. Student's t-test, error bars indicate SEM. ^***P < 0.001. (G) Statistics of the proportion of γ-H2AX expression patterns including only sex-body positive, partially positive and positive in Mettl3^Ctrl and Mettl3^cKO spermatocytes at P12. At least 550 cells were counted from 3 different mice. Student's t-test, error bars indicate SEM. ^***P < 0.001. (H) Immunofluorescence staining of TUNEL in Mettl3^Ctrl and Mettl3^cKO testes at P12. The DNA was stained with DAPI. Scale bar, 50 μm. (I) Statistics of number of TUNEL-positive cells per tubule in Mettl3^Ctrl and Mettl3^cKO testes at P12. At least 50 tubules were counted from 3 different mice. Student's t-test, error bars indicate SEM. ^**P < 0.01.

Figure 3

Mettl3 deletion alters expression pattern of genes involved in mouse spermatogenesis. (A) Heatmap showing the differentially expressed genes between Mettl3^Ctrl and Mettl3^cKO testes at P12 and their function enrichment. Two-fold expression difference and P = 0.05 as the cutoff. The enriched Gene Ontology (GO) terms of the biological process category in down or upregulated genes in Mettl3^cKO samples were shown on the right panels. (B) Violin plot of expression levels of germ cell specifically expressed genes in Mettl3^Ctrl and Mettl3^cKO samples at P6 and P12. The log2-transformed expression level (FPKM) was used for drawing the plot. ^*** means the P < 10e-5, as calculated by the Wilcoxon signed-rank paired test. (C) The unsupervised hierarchical cluster analysis for the expression of germ cell specifically expressed genes in Mettl3^Ctrl and Mettl3^cKO samples at P6 and P12. (D) The interaction network showing genes involved in spermatogenesis regulation. The downregulated genes in Mettl3^cKO testes which had been validated by qRT-PCR marked in blue. The regulation relationships were produced by pathway studio; the positive, negative regulation among these genes was marked as arrow, T-head respectively, while the regulation with unknown effects between two genes were connected directly. (E, F) qRT-PCR validation of downregulated genes involved in spermatogenesis in Mettl3^cKO testes at P6 (E) and P12 (F). Student's t-test was performed and data was shown as mean ± SEM. of three independent experiments. ^*P < 0.05, ^*P < 0.01, ^**P < 0.001; ns, no significance.

Figure 4

Mettl3-mediated m⁶A regulates alternative splicing of genes involved in spermatogenesis. (A) Schematics and statistics of the five types of alternative splicing events in Mettl3^Ctrl and Mettl3^cKO testes at P6 and P12. The differential alternative splicing events were identified by rMATS and the FDR = 0.05 were used as cutoff. Each group sample included two replicates. (B) The logo plot for bases adjacent to m⁶A sites. All 12 516 putative m⁶A residues were used to enrich the m⁶A motif and the plot was produced by WebLogo. (C) Distribution of m⁶A sites across the length of mRNA transcripts. (D) Transcriptome-wide distribution of m⁶A sites. Pie charts showing the percentage of m⁶A sites in distinct RNA sequence types: 5′ UTR, CDS, Stop codon and 3′ UTR. (E) The average m⁶A site frequency per gene in all genes and in differentially spliced genes in Mettl3^Ctrl testes at P12. (F) The cumulative frequency curve for the alternative exon inclusion ratio in Mettl3^Ctrl and Mettl3^cKO testes at P12. The exon inclusion level was calculated by alternative exon coverage divided by constitutive exon coverage and transformed by log2. The exons containing m⁶A sites in Mettl3^cKO testes showed lower coverage level than in Mettl3^Ctrl testes (P < 0.01, Wilcoxon signed-rank paired test). No significant difference was exhibited between exons without m⁶A modification. (G, H) Distribution of RNA-seq reads of Cdk11b (G) and Syce2 (H) in Mettl3^Ctrl and Mettl3^cKO testes samples, the alternative exon of each gene was marked in grey and shadow. The m⁶A site location was marked as a red triangle. The exon skipping event was validated by RT-PCR (right panel). (I) The interaction network showing genes involved in spermatogenesis regulation. The genes containing alternative exons which had been validated by RT-PCR were marked in red. The regulation relationships were produced by Pathway Studio; the positive, negative regulation among these genes was marked as arrow, T-head respectively, while the regulation with unknown effects between two genes was connected directly.