CARF regulates the alternative splicing and piwi/piRNA complexes during mouse spermatogenesis through PABPC1

Abstract

ADP-ribosylation factor collaborator (CARF), which is also known as CDKN2AIP, was first recognized as an ADP-ribosylation factor-interacting protein that participates in the activation of the ARF-p53-p21 (WAF1) signaling pathway under different conditions, such as oxidative and oncogenic stresses. The activation of this pathway often leads to cell growth arrest and apoptosis as well as senescence. Previous studies revealed that CARF, an RNA-binding protein, is critical for maintaining stem cell pluripotency and somatic differentiation. Nevertheless, its involvement in spermatogenesis has not been well examined. In this study, we show that male mice deficient in Carf expression present impaired spermatogenesis and fertility. IP-MS and RNA-seq analyses reveal that CARF/ Carf interacts with multiple key splicing factors, such as PABPC1, and directly targets 356 different types of mRNAs in spermatocytes. Carf-associated mRNAs display aberrant splicing patterns when Carf expression is deficient. In addition, our results demonstrate that PIWIL1 expression and localization are altered in the Carf ^{-/ -} mouse model through the downregulation of PABPC1, which further affects the ratio of pachytene-piRNA. Our study suggests that CARF is critical for regulating alternative splicing in mammalian spermatogenesis and determining infertility in male mice.

Keywords: CARF; PABPC1; alternative splicing; male infertility; spermatogenesis.

Figure 1

CARF is highly expressed in spermatocytes and spermatids of the testis (A) NCBI database analysis of the expression profile of Carf mRNA in mouse tissues. (B) qPCR analyses of Carf mRNA levels in multiple organs of mice. Data are presented as the mean ± SEM, n = 3. (C) Immunofluorescence staining analysis of CARF (green) in testis sections. γH2AX (red) was used as a marker for spermatocytes. The nuclei were stained with DAPI (blue). Spc indicates spermatocytes, Ser indicates Sertoli cells, Ley indicates Leydig cells, and Rspd indicates round sperm. Scale bar: 50 μm. (D) Immunofluorescence staining analysis of CARF (red) and PNA (green) in testis sections. The nuclei were stained with DAPI (blue), Scale bar: 50 μm. (E,F) The expression pattern of CARF in the mouse germline atlas was analyzed via a single-cell sequencing database (http://malehealthatlas.cn/ ).

Figure 2

CARF is essential for germ cell development (A) Expression of Carf mRNAs in developing testes at postnatal day 8 (P8), P10, P12, P16, P28 and P35 were analyzed via qPCR. Data are presented as the mean ± SEM, n = 3. (B) Expression of the CARF protein in developing testes at postnatal day 8 (P8), P10, P12, P16, P28 and P35 were analyzed via western blot analysis. (C) Schematic of the generation of Carf–/–-deficient mice with the CRISPR-Cas9 genome editing system. Four small guide RNAs (sgRNAs) were designed to generate Carf-knockout mice. The DNA fragments covering exon 1 to exon 3 were deleted. (D) Representative image of PCR genotyping via the F1, R1 and R2 primers. Heterozygotes (+/–), wild-type (+/+) and knockout (–/–) alleles generate PCR products of 538 bp and 772 bp, 538 bp, and 772 bp, respectively. H2O was used as a negative control for the PCR. M, 1000 bp marker. (E) Litter size produced by heterozygotes (n = 5/group). Data are presented as the mean ± SEM. (F) Western blot analysis of CARF protein expression in testis extracts from wild-type and Carf-knockout testes at P56. GAPDH served as a loading control. (G) The mating strategy of heterozygous mice was studied to analyze the litter size of female and male mice after mating. (H) Gross morphology of testes from wild-type and Carf –/– mice at the age of 20 weeks. Panoramic images of testicular tissue were scanned with ImageScope software. (I) Testis weight/body weight ratio of 20-week-old wild-type and Carf-knockout male mice. The Carf-knockout male mice presented significantly reduced testis weight (42% of wild-type) (P < 0.001, n = 6). (J) Testicular histology of wild-type and Carf–/– mice at the age of 20 weeks. The number of germ cells in Carf –/– male mice was significantly lower than that in wild-type male mice. The scale bar of the picture in the left panel is 100 μm, whereas the scale bar of the picture in the right panel is 50 μm. (K) Sperm counts of wild-type and Carf –/– mice.Data are presented as the mean ± SEM, n = 6, ***P<0.001.

Figure 3

CARF interacts with PABPC1 to participate in RNA alternative splicing (A) Abnormal alternative splicing patterns, including skipped exons, alternative 5′ splice site (A5SS), alternative 3′ splice site (A3SS), mutually exclusive exons (MXE) and retained intron (RI) caused by Carf defects. (B) Abnormal variable splicing of the functional gene Hira, which is related to germ cell development caused by Carf defects. It belongs to the alternative 5′ splice site (A5SS) exception mode. (C) Abnormal variable splicing of the functional gene Surf1, which is related to germ cell development caused by Carf defects. It is an alternative 3′ splice site (A3SS). (D) Abnormal variable splicing of the functional gene Usf2, which is related to germ cell development caused by Carf defects. It belongs to the mutually exclusive exons (MXE). (E) Abnormal variable splicing of the functional gene Lrmp, which is related to germ cell development caused by Carf defects. It belongs to the retained intron (RI) family. (F) Abnormal variable splicing of the functional gene Pkd2l1, which is related to germ cell development caused by Carf defects. It belongs to the retained intron (RI) family. (G) Co-IP analysis of the interaction between CARF and PABPC1. PABPC1 expression was detected in the IP products of CARF, and IgG was used as a control. GAPDH served as a loading control. (H) Co-IP analysis of the interaction between CARF and PABPC1. CARF expression was detected in the IP products of PABPC1, and IgG was used as a control. GAPDH served as a loading control. (I) Immunostaining of PABPC1 in wild-type and Carf–/– testis sections (PABPC1: green); DAPI was used to stain the dye the nuclei, scale bar: 50 μm. (J) Western blot analysis of PABPC1 protein levels in testes from wild-type and Carf–/– mice. GAPDH served as a loading control. (K) Immunohistochemical analysis of the expression of PABPC1 in testes from wild-type and Carf–/– testis sections. Scale bar: 50 μm. (L) Quantitative results of (K). n = 3, Data are presented as the mean ± SEM. ***P < 0.001.

Figure 4

Carf defects inhibits the expression of PIWIL1 and piRNA maturation (A) Immunostaining of PIWIL1 in wild-type and Carf–/– testis sections (PIWIL1: green; γH2AX: red); γH2AX is used as a marker of DNA damage. Scale bar 50 μm. (B) PCR analysis of Piwil1 mRNA levels in testes from wild-type and Carf–/– mice. GAPDH served as a loading control. Data are presented as mean ± SEM. ***P < 0.001. (C) Immunohistochemical analysis of the expression of PIWIL1 in testes from wild-type and Carf–/– testis sections. Scale bar: 50 μm. (D) Quantitative results of (C). n = 3, Data are presented as the mean ± SEM. *P < 0.05. (E) Transcript and piRNA abundances in wild-type and Carf–/– testes are shown for illustrative examples from 16.5 dpp mice. (F) Analysis of the length distribution of piRNAs in wild-type and Carf–/– testes from P16.5 dpp mice. (G) The first nucleotides of the piRNAs, showing a strong U bias in wild-type and Carf–/– testes from P16.5 mice. (H) The tenth nucleotides of the piRNAs, showing a strong U bias in wild-type and Carf–/– testes from P16.5 mice.