Multigenerational paternal obesity enhances the susceptibility to male subfertility in offspring via Wt1 N6-methyladenosine modification

Abstract

There is strong evidence that obesity is a risk factor for poor semen quality. However, the effects of multigenerational paternal obesity on the susceptibility to cadmium (a reproductive toxicant)-induced spermatogenesis disorders in offspring remain unknown. Here, we show that, in mice, spermatogenesis and retinoic acid levels become progressively lower as the number of generations exposed to a high-fat diet increase. Furthermore, exposing several generations of mice to a high fat diet results in a decrease in the expression of Wt1, a transcription factor upstream of the enzymes that synthesize retinoic acid. These effects can be rescued by injecting adeno-associated virus 9-Wt1 into the mouse testes of the offspring. Additionally, multigenerational paternal high-fat diet progressively increases METTL3 and Wt1 N6-methyladenosine levels in the testes of offspring mice. Mechanistically, treating the fathers with STM2457, a METTL3 inhibitor, restores obesity-reduced sperm count, and decreases Wt1 N6-methyladenosine level in the mouse testes of the offspring. A case-controlled study shows that human donors who are overweight or obese exhibit elevated N6-methyladenosine levels in sperm and decreased sperm concentration. Collectively, these results indicate that multigenerational paternal obesity enhances the susceptibility of the offspring to spermatogenesis disorders by increasing METTL3-mediated Wt1 N6-methyladenosine modification.

Fig. 1. Multigenerational paternal HFD enhances susceptibility to spermatogenesis disorder in offspring.

F0 generation male mice were fed NC or HFD from 5 weeks to 15 weeks old, and then mated with NC-fed female mice to breed F1 generation. Similarly, a subset of the males in F1 generation were continued to be treated with HFD for 10 weeks, and mated with normal female to breed F2 generation. Male mice of F1 and F2 generations were exposed to CdCl₂ for 10 weeks, and named NC, NCD, HFD1D or HFD2D group respectively. All mice were euthanized at 15 weeks of age. a Experimental design flowchart of Cd-induced spermatogenesis impairment in mice with HFD-feeding. The black arrow indicated NC treatment. The purple arrow indicated HFD treatment. The red arrow indicated Cd treatment. b The fertility rate was calculated. n = 4 independent experiments, Degree of Freedom (DOF) = 11, F = 18.20, P = 0.0007. c The pregnancy rate was counted. n = 4 independent experiments, P = 0.3911. d, e Epididymal sperm counts were measured. n = 10 mice, DOF = 39, F = 37.68, P < 0.0001. f Sperm motility was recorded; n = 10 mice, DOF = 39, F = 14.01, P < 0.0001. g Testicular H&E staining. h The number of Testicular seminiferous tubules number at different stages were evaluated. n = 4 mice, DOF = 15, F = 30.25, P < 0.0001. n.s. not significant. *P < 0.05; **P < 0.01. In regard to Fig. 1b, e, f, h, statistical significance was evaluated by two-sided one-way ANOVA with post hoc LSD tests. Data are presented as mean ± SEM. Source data are provided with this paper.

Fig. 2. Multigenerational paternal HFD exacerbates environmental stress-impaired testicular germ cell development in offspring.

F0 generation male mice were fed NC or HFD from 5 weeks to 15 weeks old, and then mated with normal female to breed F1 generation. Similarly, a subset of the males in F1 generation were continued to be treated with HFD for 10 weeks, and mated with normal female to breed F2 generation. Male mice of F1 and F2 generations were exposed to CdCl₂ for 10 weeks, and named NC, NCD, HFD1D or HFD2D group respectively. All mice were euthanized at 15 weeks of age. a Testicular weight. n = 10 mice, DOF = 39, F = 8.12, P = 0.0003. b, c Testicular DDX4 protein expression was detected by immunoblotting. n = 6 mice, DOF = 23, F = 25.05, P < 0.0001. d Testicular Izumo3, Acrv1, Smc3, C-kit and Plzf mRNA levels were detected by RT-qPCR. n = 6 mice, DOF = 23, F = 10.68 and P = 0.0002 for Plzf; F = 16.53 and P < 0.0001 for C-kit; F = 25.24 and P < 0.0001 for Smc3; F = 31.55 and P < 0.0001 for Acrv1; F = 29.16 and P < 0.0001 for Izumo3. e, f Testicular PLZF, C-KIT and SYCP3 protein expression were detected by immunoblotting. n = 6 mice, DOF = 23, F = 7.15 and P = 0.002 for PLZF; F = 27.78 and P < 0.0001 for C-KIT; F = 24.44 and P < 0.0001 for SYCP3. g The expression of testicular SYCP3 was measured via immunohistochemistry. h SYCP3-positive cells were counted. n = 4 mice, DOF = 15, F = 39.87, P < 0.0001. n.s. not significant. *P < 0.05; **P < 0.01. In regard to Fig. 2a, c, d, f, h, statistical significance was evaluated by two-sided one-way ANOVA with post hoc LSD tests. Data are presented as mean ± SEM. Source data are provided with this paper.

Fig. 3. Multigenerational paternal HFD progressively aggravates environmental stress-inhibited testicular retinoic acid synthesis in offspring.

F0 generation male mice were fed NC or HFD from 5 weeks to 15 weeks old, and then mated with normal mice to breed F1 generation. Similarly, a subset of the males in F1 generation were continued to be treated with HFD for 10 weeks, and mated with normal female to breed F2 generation. Male mice of F1 and F2 generations were exposed to CdCl₂ for 10 weeks, and named NC, NCD, HFD1D or HFD2D group respectively. All mice were euthanized at 15 weeks of age. a, b GO enrichment analysis of differentially expressed mRNAs in mouse testes between HFD1D and HFD2D groups. c Testicular retinoic acid level was detected by ELISA. n = 4 mice, DOF = 15, F = 41.13, P < 0.0001. d, e Testicular RARα and STRA8 protein expressions were measured by immunoblotting. n = 6 mice, DOF = 23, F = 43.08 and P < 0.0001 for RARα; F = 46.65 and P < 0.0001 for STRA8. f Testicular Aldh1a1, Aldh1a2, Aldh1a3, Rdh10, Cyp26a1, Cyp26b1 and Cyp26c1 mRNA levels were tested by RT-qPCR. n = 6 mice, DOF = 23, F = 32.26 and P < 0.0001 for Aldh1a1; F = 27.78 and P < 0.0001 for Aldh1a2; F = 25.82 and P < 0.0001 for Aldh1a3; F = 11.75 and P = 0.0001 for Rdh10; F = 0.18 and P = 0.9063 for Cyp26a1; F = 0.78 and P = 0.5217 for Cyp26b1; F = 3.20 and P = 0.0459 for Cyp26c1. g–i Testicular ALDH1A1 and ALDH1A2 proteins were measured using immunoblotting. n = 6 mice, DOF = 23, F = 61.35 and P < 0.0001 for ALDH1A1; F = 20.01 and P < 0.0001 for ALDH1A2. n.s. not significant. *P < 0.05; **P < 0.01. In regard to Fig. 3c, e, f, h, i, statistical significance was evaluated by two-sided one-way ANOVA with post hoc LSD tests. Data are presented as mean ± SEM. Source data are provided with this paper.

Fig. 4. Multigenerational paternal HFD progressively exacerbates environmental stress-downregulated testicular WT1 expression in offspring.

F0 generation male mice were fed NC or HFD from 5 weeks to 15 weeks old, and then mated with normal female to breed F1 generation. Similarly, a subset of the males in F1 generation were continued to be treated with HFD for 10 weeks, and mated with normal female to breed F2 generation. Male mice of F1 and F2 generations were exposed to CdCl₂ for 10 weeks, and named NC, NCD, HFD1D or HFD2D group respectively. All mice were euthanized at 15 weeks of age. a The upstream transcription factors of Aldh1a1, Aldh1a2 and Aldh1a3 were predicted from ChEA3 (https://maayanlab.cloud/chea3/). b Testicular Wt1 mRNA level was tested using RT-qPCR. n = 6 mice, DOF = 23, F = 33.75, P < 0.0001. c, d Testicular WT1 protein expression was measured by immunoblotting. n = 6 mice, DOF = 23, F = 21.59, P < 0.0001. e, f The number of testicular WT1⁺ positive cells per tubule were counted by immunohistochemistry. n = 6 mice, DOF = 23, F = 23.66, P < 0.0001. *P < 0.05; **P < 0.01. In regard to Fig. 4b, d, e, statistical significance was evaluated by two-sided one-way ANOVA with post hoc LSD tests. Data are presented as mean ± SEM. Source data are provided with this paper.

Fig. 5. Paternal HFD aggravates environmental stress-impaired testicular spermatogenesis via inhibiting WT1-mediated retinoic acid synthesis in offspring.

F0 generation male mice were fed NC or HFD from 5 weeks to 15 weeks old, and mated with normal female to breed F1 generation. At postnatal day 35, Wt1 was overexpressed in F1 mice by local testicular injection of adeno-associated virus 9 (AAV9). After injection of AAV9-Wt1 or vehicle, mice were exposed to CdCl₂ for 10 weeks, and named NCD, NCD + WT1, HFD1D or HFD1D + WT1 group, respectively. All mice were euthanized at 15 weeks of age. a Experimental design flowchart. b, c Epididymal sperm counts were measured. n = 9 mice, DOF = 35, F = 6.36, P = 0.0017. d–f Testicular RARα and STRA8 protein expressions were measured by immunoblotting. n = 6 mice, DOF = 23, F = 29.78 and P < 0.0001 for RARα; F = 33.45 and P < 0.0001 for STRA8. g Testicular retinoic acid level was detected by ELISA. n = 6 mice, DOF = 23, F = 13.88, P < 0.0001. h–j Testicular ALDH1A1 and WT1 protein levels were measured by immunoblotting. n = 6 mice, DOF = 23, F = 30.92 and P < 0.0001 for ALDH1A1; F = 48.60 and P < 0.0001 for WT1. k Testicular WT1 expression was measured by immunohistochemistry. n = 4 mice. *P < 0.05; **P < 0.01 vs NCD. ^#P < 0.05; ^##P < 0.01 vs HFD1D. In regard to Fig. 5c, e-g, i, j, statistical significance was evaluated by two-sided one-way ANOVA with post hoc LSD tests. Data are presented as mean ± SEM. Source data are provided with this paper.

Fig. 6. Multigenerational paternal HFD progressively exacerbates environmental stress-elevated testicular Wt1 m6A level in offspring.

F0 generation male mice were fed NC or HFD from 5 weeks to 15 weeks old, and then mated with normal female to breed F1 generation. Similarly, a subset of the males in F1 generation were continued to be treated with HFD for 10 weeks, and mated with normal female to breed F2 generation. Male mice of F1 and F2 generations were exposed to CdCl₂ for 10 weeks, and named NC, NCD, HFD1D or HFD2D group respectively. All mice were euthanized at 15 weeks of age. a Testicular total RNA m6A level was measured. n = 4 mice, DOF = 15, F = 23.52, P < 0.0001. b, c Testicular METTL3 protein expression was detected by immunoblotting. n = 6 mice, DOF = 23, F = 117.00, P < 0.0001. d Testicular Ythdc1, Ythdc2, Ythdf1, Ythdf2, Ythdf3, Igf2bp1, Igf2bp2 and Igf2bp3 mRNA levels were tested by RT-qPCR. n = 6 mice, DOF = 23, F = 1.14 and P = 0.3565 for Ythdc1; F = 4.49 and P = 0.0145 for Ythdc2; F = 18.81 and P < 0.0001 for Ythdf1; F = 29.71 and P < 0.0001 for Ythdf2; F = 2.15 and P = 0.1259 for Ythdf3; F = 28.11 and P < 0.0001 for Igf2bp1; F = 1.46 and P = 0.2566 for Igf2bp2; F = 3.71 and P = 0.02858 for Igf2bp3. e, f Testicular YTHDF1, YTHDF2 and IGF2BP1 protein expressions were detected by immunoblotting. n = 6 mice, DOF = 23, F = 26.49 and P < 0.0001 for YTHDF1; F = 22.44 and P <0.0001 for YTHDF2; F = 23.94 and P < 0.0001 for IGF2BP1. g The four m6A modification sites in Wt1 mRNA were predicted by online tool SRAMP. h Wt1 site1 m6A level was measured by MeRIP-qPCR. n = 6 mice, DOF = 23, F = 34.76, P < 0.0001. n.s. not significant. *P < 0.05; **P < 0.01. In regard to Fig. 6a, c, d, f, h, statistical significance was evaluated by two-sided one-way ANOVA with post hoc LSD tests. Data are presented as mean ± SEM. Source data are provided with this paper.

Fig. 7. Environmental stress decreases Wt1 stability in an m6A-dependent manner in Sertoli cells.

a, b Treatment of TM4 cells with CdCl₂ (20 μM) for 0, 6, 12 or 24 h. a Cellular Wt1 mRNA expression was detected by RT-qPCR. n = 3 biologically independent samples, DOF = 11, F = 17.24, P = 0.0007. b Wt1 site1 m6A level was measured using MeRIP-qPCR. n = 3 biologically independent samples, DOF = 11, F = 15.54, P = 0.0011. c Schematic representation of m6A motif and mutation position in CDS within Wt1 mRNA were presented. d The luciferase activities of the mutant and wild-type Wt1-CDS were measured after Cd exposure by luciferase reporter assay. n = 3 biologically independent samples, DOF = 11, F = 8.79, P = 0.0065. e–h Treatment of TM4 cells with CdCl₂ for 24 h after Mettl3 siR. e Wt1 site1 m6A level was detected. n = 3 biologically independent samples, DOF = 11, F = 33.43, P < 0.0001. f Wt1 mRNA expression was measured. n = 3 biologically independent samples, DOF = 11, P = 0.0006, F = 18.46. g, h WT1 and METTL3 protein levels were detected via immunoblotting. n = 3 biologically independent samples, DOF = 11, F = 15.21 and P = 0.0011 for WT1; F = 78.05 and P < 0.0001 for METTL3. i Wt1 mRNA level was measured after treating with ActD for 3 or 6 h using RT-qPCR. n = 3 biologically independent samples, F = 8.00, P = 0.0001. j–l Treatment of TM4 cells with CdCl₂ for 24 h in the within or without of Igf2bp1 siR. WT1 and IGF2BP1 protein levels were detected via immunoblotting. n = 3 biologically independent samples, DOF = 11, F = 22.13 and P = 0.0003 for WT1; F = 10.07 and P = 0.0043 for IGF2BP1. m Wt1 mRNA level was detected after treating with ActD for 3 or 6 h using RT-qPCR. n = 3 biologically independent samples, F = 5.04, P = 0.0051. n The binding ability of IGF2BP1 and Wt1 was evaluated after Cd exposure by RIP assays. n = 3 biologically independent samples, DOF = 11, F = 75.36, P < 0.0001. *P < 0.05; **P < 0.01 vs NC. ^#P < 0.05; ^##P < 0.01 vs Cd. In regard to Fig. 7a, b, d–f, h, k–l, n, statistical significance was evaluated by two-sided one-way ANOVA with post hoc LSD tests. In regard to Fig. 7i, m, statistical significance was evaluated by two-sided two-way ANOVA. Data are presented as mean ± SEM. Source data are provided with this paper.

Fig. 8. Multigenerational paternal HFD progressively increases m6A level and downregulates Wt1 mRNA expression in paternal sperm.

F0 generation male mice were fed NC or HFD from 5 weeks to 15 weeks old, and then mated with normal female to breed F1 generation. Similarly, partial F1 generation male mice continued to be fed HFD for 10 weeks. Sperm were collected for MeRIP-microarray and RNA-microarray. a The total RNA m6A level of sperm was measured. n = 4 mice, DOF = 11, F = 20.97, P = 0.0004. b Sperm Mettl3, Mettl14, Wtap, Alkbh5 and Fto mRNA levels were tested using RT-PCR. n = 4 mice, DOF = 11, F = 31.58 and P < 0.0001 for Mettl3; F = 1.77 and P = 0.2242 for Mettl14; F = 0.10 and P = 0.9066 for Wtap; F = 0.01 and P = 0.9926 for Alkbh5; F = 2.13 and P = 0.1750 for Fto. c, d GO and KEGG enrichment analysis of hyper-methylated mRNAs in sperm. e The distribution of differentially expressed mRNAs in sperm were presented. f Hyper-methylated mRNAs and down-regulated mRNAs in sperm were intersected. g Heatmap of overlapping mRNAs in sperm was depicted. h Sperm Wt1 mRNA expression was tested using RT-qPCR. n = 4 mice, DOF = 11, F = 21.19, P = 0.0004. *P < 0.05. In regard to Fig. 8a, b, h, statistical significance was evaluated by two-sided one-way ANOVA with post hoc LSD tests. Data are presented as mean ± SEM. Source data are provided with this paper.

Fig. 9. Paternal HFD enhances testicular Wt1 downregulation and spermatogenesis disorder in offspring via increasing METTL3-mediated paternal sperm m6A level.

F0 generation male mice were fed NC or HFD from 5 weeks to 15 weeks old, and treated with STM2457 once a week from 10 weeks to 15 weeks old. F1 generation male mice were exposed to CdCl₂ from 5 weeks to 10 weeks old, and named NCD, NCD + STM, HFD1D or HFD1D + STM group, respectively. a Experimental design flowchart. b, c Epididymal sperm counts were measured. n = 10 mice for NCD group; n = 9 mice for NCD + STM group; n = 11 mice for HFD1D group; n = 8 mice for HFD1D + STM group; DOF = 37, F = 6.10, P = 0.0020. d, e Testicular WT1 protein expression was measured by immunoblotting. n = 6 mice, DOF = 23, F = 8.94, P = 0.0006. f The mRNA level of testicular Wt1 was detected. n = 6 mice, DOF = 23, F = 17.81, P < 0.0001. g Testicular Wt1 site1 m6A level was measured by MeRIP-qPCR. n = 6 mice, DOF = 23, F = 10.20, P = 0.0003. h, i The protein level of testicular METTL3 was measured. n = 6 mice, DOF = 23, F = 27.16, P < 0.0001. j Sperm total RNA m6A level was tested. n = 6 mice, DOF = 23, F = 14.33, P < 0.0001. k Sperm Mettl3 and Wt1 mRNA level were detected by RT-qPCR. n = 6 mice, DOF = 24, F = 15.51 and P < 0.0001 for Mettl3; F = 14.90 and P < 0.0001 for Wt1. *P < 0.05; **P < 0.01 vs NCD. ^#P < 0.05; ^##P < 0.01 vs HFD1D. In regard to Fig. 9c, e–g, i–k, statistical significance was evaluated by two-sided one-way ANOVA with post hoc LSD tests. Data are presented as mean ± SEM. Source data are provided with this paper.

Fig. 10. Elevated sperm m6A level and decreased sperm concentration is observed in donors who were overweight or obese.

After removal of smoking or alcohol drinking donors, 30 pairs of cases with overweight/obesity and corresponding controls were obtained by matching age. a BMI. n = 30 human sperm, t = −11.20, P < 0.0001. Two-tailed t test was used to analyze the differences. b Sperm concentration. n = 30 human sperm, t = 10.50, P < 0.0001. Two-tailed t test was used to analyze the differences. c Sperm m6A level. n = 30 human sperm, t = −2.74, P = 0.0083. Two-tailed t test was used to analyze the differences. d The association between sperm concentration and BMI. e The association between sperm m6A level and BMI. f The association between sperm m6A level and sperm concentration. g Graphical abstract. **P < 0.01. Source data are provided with this paper.