Identification of multiple male reproductive tract-specific proteins that regulate sperm migration through the oviduct in mice

Abstract

CRISPR/Cas9-mediated genome editing technology enables researchers to efficiently generate and analyze genetically modified animals. We have taken advantage of this game-changing technology to uncover essential factors for fertility. In this study, we generated knockouts (KOs) of multiple male reproductive organ-specific genes and performed phenotypic screening of these null mutant mice to attempt to identify proteins essential for male fertility. We focused on making large deletions (dels) within 2 gene clusters encoding cystatin (CST) and prostate and testis expressed (PATE) proteins and individual gene mutations in 2 other gene families encoding glycerophosphodiester phosphodiesterase domain (GDPD) containing and lymphocyte antigen 6 (Ly6)/Plaur domain (LYPD) containing proteins. These gene families were chosen because many of the genes demonstrate male reproductive tract-specific expression. Although Gdpd1 and Gdpd4 mutant mice were fertile, disruptions of Cst and Pate gene clusters and Lypd4 resulted in male sterility or severe fertility defects secondary to impaired sperm migration through the oviduct. While absence of the epididymal protein families CST and PATE affect the localization of the sperm membrane protein A disintegrin and metallopeptidase domain 3 (ADAM3), the sperm acrosomal membrane protein LYPD4 regulates sperm fertilizing ability via an ADAM3-independent pathway. Thus, use of CRISPR/Cas9 technologies has allowed us to quickly rule in and rule out proteins required for male fertility and expand our list of male-specific proteins that function in sperm migration through the oviduct.

Keywords: CRISPR/Cas9; fertilization; infertility; transgenic; uterotubal junction.

Fig. 1.

Male fertility of mice lacking the region between Pate8 and Pate10. (A) Pate family genes within murine genomic locus |chromosome 9qA4|. Pate family genes (Pate1–Pate14) were shown as “1” to “14.” Genes with “Gm-” were listed as serial numbers. The inequalities showed the direction of transcription. (B) Multitissue gene expression by RT-PCR analysis. Actin β (Actb) was used as the control. Brain (Br), heart (He), kidney (Ki), liver (Li), lung (Lu), spleen (Sp), thymus (Th), ovary (Ov), uterus (Ut), testis (Te), epididymis (Epi), caput (Cap), corpus (Cor), cauda (Cau), prostate (Pr), coagulating gland (CG), seminal vesicle (SV), placenta (Pl). (C) gRNA design for del of gene cluster between Pate8 and Pate10. Arrowheads show gRNAs for Pate8 and Pate10. Primers (Fw1, Fw2, Rv1, and Rv2) were used for genotyping with PCR. (D) Genotyping with PCR in (Pate8–Pate10)^del/del mice. Primers shown in C were used for PCR. wild-type (WT). (E) DNA sequencing. The sequence of PCR amplicon was analyzed. (Pate8–Pate10)^del/del mice deleted 841-kb genomic region between Pate8 and Pate10. (F) Pregnancy rates (delivery/plug). Males were caged with two WT females. (Pate8–Pate10)^del/del males succeeded in the mating, but the pregnancy rates of these females significantly reduced (del/wt: 80% [16/20], del/del: 2.9% [1/34]).

Fig. 2.

Phenotypic analysis of (Pate8–Pate10)^del/del male mice. (A) Sperm fertilizing ability of using cumulus-intact oocytes in vitro. (Pate8–Pate10)^del/del spermatozoa could efficiently fertilize eggs (del/wt: 91.4 ± 10.5%, del/del: 84.0 ± 13.0%; mean ± SD). (B) Sperm ZP-binding assay. Spermatozoa were inseminated with cumulus-free oocytes, but (Pate8–Pate10)^del/del spermatozoa hardly bind to the ZP. (C) Average number of spermatozoa bound to the ZP. The number of spermatozoa bound to the ZP in (Pate8–Pate10)^del/del males was significantly reduced (del/wt: 25.9 ± 4.7 spermatozoa/egg, del/del: 1.5 ± 1.6 spermatozoa/egg; mean ± SD). **P < 0.01, Student’s t test. (D) Detection of ADAM3. ADAM3 was not detected in (Pate10–Pate8)^del/del spermatozoa (also see SI Appendix, Fig. S4B). GAPDH was used as the control. Sperm: Cau Epi spermatozoa. (E) Observation of sperm migration into the female reproductive tract. For sperm observation in the female reproductive tract after mating, we crossed (Pate8–Pate10)^del/del mice with a transgenic mouse line in which the acrosome and mitochondria can be visualized by EGFP and DsRed2, respectively. (Pate8–Pate10)^del/del spermatozoa could not pass through the UTJ.

Fig. 3.

Male fertility of mice lacking the region between Cstl1 and Cstdc2. (A) Male reproductive CST family within murine genomic location |chromosome 2qG3|. The direction of the arrow indicates the direction of the transcription. (B) Multitissue gene expression by RT-PCR analysis. Actb was used as an expression control. (C) gRNA design for del of gene cluster between Cstl1 and Cstdc2. Arrowheads show gRNAs for Cstl1 and Cstdc2. Primers (Fw3, Fw4, Rv3, and Rv4) were used for genotyping with PCR. (D) Genotyping with PCR in (Cstl1-Cstdc2)^del/del mice. Primers shown in C were used for PCR. (E) DNA sequencing. The sequence of PCR amplicon was analyzed. (Cstl1-Cstdc2)^del/del mice deleted 114-kb genomic region between Cstl1 and Cstdc2. (F) Pregnancy rates (%) of WT female mice mated with (Cstl1-Cstdc2) mutant male mice. The average pregnancy rate (delivery per plug) of females coupled with (Cstl1-Cstdc2)^del/wt and (Cstl1-Cstdc2)^del/del male mice were 100% (31/31) and 10.0% (4/40), respectively.

Fig. 4.

Phenotypic analysis of (Cstl1-Cstdc2)^del/del male mice. (A) In vitro fertilization rates using (Cstl1-Cstdc2) mutant spermatozoa. Average fertilization rates of (Cstl1-Cstdc2)^del/wt and (Cstl1-Cstdc2)^del/del spermatozoa were 90.4 ± 9.6% (136/148 eggs) and 83.4 ± 14.3% (222/256 eggs), respectively. (B) Sperm-ZP-binding assay. Spermatozoa were inseminated with cumulus-free oocytes, but (Cstl1-Cstdc2)^del/del spermatozoa hardly bind to the ZP. (C) Average number of ZP-binding spermatozoa in vitro. The number of ZP-binding spermatozoa in (Cstl1-Cstdc2)^del/del mice (2.2 ± 1.6 spermatozoa/egg; mean ± SD) was significantly reduced compared with that of (Cstl1-Cstdc2)^del/wt mice (21.2 ± 6.7 spermatozoa/egg). **P < 0.01, Student’s t test. (D) Immunoblot analysis of ADAM3. ADAM3 was not detectable in (Cstl1-Cstdc2)^del/del spermatozoa. IZUMO1 was used as the control. Sperm: Cau Epi spermatozoa. (E) Observation of sperm migration into the female reproductive tract. For sperm observation in the female reproductive tract after mating, we crossed (Cstl1-Cstdc2)^del/del mice with a transgenic mouse line in which the acrosome can be visualized by EGFP. (Cstl1-Cstdc2)^del/del spermatozoa could not pass through the UTJ.

Fig. 5.

Characterization of LYPD4 in mice and humans. (A) Te-specific expression of mouse Lypd4 by multitissue RT-PCR analysis. The expression of each gene was examined by RT-PCR using RNA isolated from various organs. Lypd4 was detected only in the mouse Te. The Actb gene was used as an expression control. (B) RT-PCR analysis of human LYPD4 in the human tissues. Human LYPD4 was also detected only in the Te. (C) Immunostaining of LYPD4 in Cau Epi spermatozoa. LYPD4 (red signal) localized to the sperm acrosomal membrane. IZUMO1 (green signals) is a sperm acrosome membrane protein used as a marker for the acrosome reaction. (D) Confocal microsocopic observation of LYPD4 and IZUMO1 in Cau Epi spermatozoa. Although IZUMO1 (green signal) was detected on the outer acrosomal membrane (OAM), LYPD4 (red signal) localized to the inner acrosomal membrane (IAM).

Fig. 6.

Male fertility and in vitro fertilizing ability in Lypd4 KO mice. (A) Pregnancy rates of WT female mice mated with Lypd4^+/− and Lypd4^−/− male mice. Pregnancy rate is the success rate from natural matings (delivery per plug). The average pregnancy rate of females coupled with Lypd4^+/− and Lypd4^−/− male mice were 93.3% (28/30) and 3.7% (3/82), respectively. (B) In vitro fertilization rates using WT and Lypd4 KO spermatozoa. Average fertilization rates of WT and Lypd4 KO spermatozoa were 98.0 ± 1.5% (337/344 eggs) and 96.6 ± 7.0% (308/319 eggs), respectively. (C) Observation of ZP-binding in WT and Lypd4 KO spermatozoa. Lypd4 KO spermatozoa have an impaired ZP-binding ability in vitro (Scale bars: 50 μm.) (D) Average number of ZP-binding spermatozoa in vitro. The number of ZP-binding spermatozoa in Lypd4 KO mice (1.0 ± 0.8 spermatozoa; mean ± SD) was significantly reduced compared with that of WT mice (43.2 ± 9.1 spermatozoa). **P < 0.01, Student’s t test. (E) Observation of female reproductive tract. Ut and oviducts from WT females mated with WT and Lypd4 KO males carrying fluorescent protein-tagged spermatozoa indicated the failure of Lypd4 KO spermatozoa to pass through the UTJ. Photographs were taken 4 h after coitus (Scale bars: 1 mm.)

Fig. 7.

Immunoblot analysis of Lypd4 KO and ADAM3-associated KO mouse lines. (A) Immunoblot analysis using TGC lysates collected from WT and Lypd4 KO mice. There were no differences between WT and Lypd4 KO mice. (B) Immunoblot analysis using sperm lysates collected from WT and Lypd4 KO mice. Although ADAM3 and CMTM2A reduced in Lypd4 KO mice, they remained. (C and D) Immunoblot analysis using TGC (C) and sperm (D) lysates collected from WT, Ace-t, Adam3, Cmtm2a/b, and Ly6k KO mice. LYPD4 remained in TGCs and spermatozoa from WT and these 4 genes’ KO mice. BASIGIN and SLC2A3 were used as loading controls.