Introduction
Infertility is becoming a global emerging health problem; approximately 15% of couples suffer from fertility-related problems, and as many as 50% of cases can be attributed to male factors [
1,
2]. Three-quarters of these factors can be accounted for by aberrant spermatogenesis, a process divided into three phases: mitosis, meiosis, and spermiogenesis [
3–
6]. These special and complex processes demand spermatid-specific genes to execute their intricate roles, and an estimated 600–1000 germ cell-unique genes participate in spermiogenesis [
7]. Disruption of the protein homeostasis required during spermatogenesis can lead to male infertility [
8,
9]. Thus, identifying the underlying pivotal genetic regulators of spermiogenesis may improve the diagnostic rate and clinical treatments for male infertility. Understanding the etiological factors of infertility and systematically identifying associated candidates is important.
Mainly regulated by the ubiquitin–proteasome system, proteolysis in germ cell meiosis and spermiogenesis is essential for virtually all steps of mammalian male gamete formation. A growing number of studies have verified that E3 ubiquitin ligase plays diverse and critical roles during spermatogenesis [
10–
13]. The Cullin-RING finger ligases (CRLs), which consist of one of seven Cullin protein scaffolds that interact with specific adaptors and various substrate receptors, are the largest family of ubiquitin ligases in eukaryotes [
14]. Previous studies have shown that CRL4 plays important roles in cell cycle progression, DNA damage response, DNA replication regulation, histone modification, hematopoiesis, and organismal development, as well as in pathogenesis of neuronal diseases and in tumorigenesis [
15–
19]. CRL4 can regulate the survival of both male and female germ cells [
20]. The difference between CRL4 and other CRLs is that CRL4 employs WD40-like repeat-containing protein DDB1 (DNA damage binding protein 1) as linker protein. CRL4 substrate receptors that give substrate specificity to the appropriate CLR4 complex for ubiquitination are called DCAFs (DDB1-associated and Cullin-4-associated factors) [
21,
22].
Several CUL4-DDB1-DCAF components related to spermiogenesis have been studied in gene-targeted mouse models. One of the neddylation members of the Cullins, CUL4A, and its homolog, CUL4B, play distinct and vital roles during spermatogenesis [
10,
23–
25]. CUL4A is specifically expressed during the meiosis stage of spermatocytes, and
Cul4a-deficient male mice are infertile because of the defective meiotic progression of their gametes [
10,
25]. CUL4B is expressed in spermatogonia but predominantly in post-meiotic spermatids; male infertility resulting from
Cul4b disruption has been attributed to abnormal post-meiotic sperm development [
23,
24].
Dcaf17 knockout (KO) mice successfully represent the male infertility phenotype due to abnormal sperm development [
26]. Here, we investigated the role of another CRL4 substrate adapter, DCAF8, on male fertility in mice.
DCAF8, also known as WDR42A, functions as a substrate recognition receptor for CUL4-DDB1 E3 ubiquitin ligase [
22,
27,
28]. Histone H3 can be uniquely polyubiquitinated by CRL4
DCAF8 ubiquitin ligase to modify the fetal liver gene signature during liver maturation [
29]. CRL4
DCAF8 regulates ubiquitination, degradation, and stabilization of MLF1 and MLF2, which play critical roles in hematopoiesis, leukemia, and various cancers [
30]. The DCAF8 p.R317C mutation causes hereditary motor and sensory neuropathy with giant axons and mild cardiomyopathy [
31]. The
Dcaf8 mRNA level was significantly upregulated in
Miwi knockdown GC-2 cells, indicating that
Dcaf8 has a potential role in histone degradation during the pachytene stage of spermatogenesis [
32,
33]. However, the specific function of
Dcaf8 in spermatogenesis remains unknown.
In this study, we generated Dcaf8 KO mice. We found that DCAF8 was predominantly expressed in adult mouse testis in the cytoplasm of spermatocytes but not in their nuclei. More than 40% of spermatids in Dcaf8 KO male mice presented bent head malformations, and their fertilizing capability was impaired dramatically. These findings showed that Dcaf8 plays an important role in spermatogenesis.
Materials and methods
Mice and genotyping
Dcaf8 KO mice were generated to introduce a myriad of mutations to exon 6 of the Dcaf8 gene using CRISPR-Cas9 gene editing approaches. One founder mouse was found to harbor insertion of a fragment of 115 base pair nucleotides, which resulted in a frameshift in its coding sequence. Heterozygous mice were intercrossed to generate offspring with three different genotypes. Genomic DNA was isolated from mice toes using a genomic purification kit (Promega). Genotype validations were performed by PCR. Table 1 lists the primers used. The PCR conditions were as follows: 94 °C for 2 min, 35 cycles of 94 °C for 30 s, 60 °C for 45 s, 72 °C for 30 s, and a 10 min extension at 72 °C at the end of the run. PCR products were separated on 1.5% agarose gels; the wildtype (WT) and mutant alleles were 538 and 653 bp, respectively.
Analyses of mRNA expression
To detect the mRNA levels of
Dcaf8 in different mice tissues, we extracted total RNA using Trizol Reagent (Thermo Fisher) and quantified it using a Thermo Fisher Nano-Drop 2000c spectrophotometer. We took 500 ng of total RNA for reverse transcription using the Prime-Script RT Reagent kit (Takara). Real-time PCR was performed using cDNA products as templates from different tissues. Table 1 lists the primers used.
Dcaf8 expression levels were normalized to
Gapdh by the 2
−ΔCt method and reported as fold changes [
34].
Extraction of nuclear and cytoplasmic protein
After removing the tunica albuginea, the testes were digested to single cells using collagenase I (Thermo Fisher) and trypsin (Basal Media). The nuclear and cytoplasmic proteins were extracted according to a Nature Protocol publication [
35]. In brief, the cells were lysed with hypotonic lysis buffer (10 mmol/L Tris HCl pH 8.0, 1 mmol/L KCl, 1.5 mmol/L MgCl
2, 1 mmol/L DTT, protease inhibitors, and ddH
2O) for 30 min, and the supernatant was collected as the cytoplasmic extract after centrifugation. The pellet was resuspended in NP-40 buffer (50 mmol/L Tris-HCl pH 8.0, 400 mmol/L NaCl, 5 mmol/L EDTA pH 8.0, 1% IGEPAL, 0.2% SDS, and ddH
2O) as the nuclear extract.
Western blot
The testes and cauda epididymis of mice were homogenized in liquid nitrogen and then lysed in 2× SDS lysis buffer (1 mol/L Tris buffer (pH 6.8), DTT, 10% SDS, glycerin, and ddH2O). After three alternate 30 min cooling and boiling periods, we harvested the protein samples. On the basis of the molecular weight of the proteins, we subjected different concentrations of SDS-PAGE gels to electrophoresis and transferred the separated proteins to polyvinylidene fluoride membranes. We blocked the membranes with 5% nonfat milk for 1 h at room temperature and then incubated them with the special primary antibodies overnight at 4 °C. To visualize the protein bands, we incubated the membranes with labeled secondary antibody at room temperature for 1 h. Finally, proteins were detected with an ECL Plus kit. The primary antibodies used were anti-DCAF8 (Abcam), anti-Histone H3 (Abcam), and anti-GAPDH (Proteintech).
Male fertility test
Sexually mature male mice of two genotypes in WT groups (n = 4) and Dcaf8 KO groups (n = 16) were mated with WT females at a 1:1 sex ratio for four continuous days, and the females were changed every day. Vaginal plugs were checked in the next morning every day. We counted the numbers of total females, plugged females, litters, and offspring in each group to calculate the frequencies of copulatory plug (FCP) and of conception (FC).
Sperm count
To evaluate epididymal sperm counts, the cauda epididymis from 10-week-old mice were dissected, placed into 1 mL of Hanks’ balanced salt solution (HBSS), and maintained at 37 °C in a 5% CO2 environment. Each epididymis was minced, allowing the sperm to swim out. After 20 min of incubation, we diluted the sperm suspensions for quantitative assessments.
Histology
To explore the progress of spermatogenesis, testes and epididymis were isolated from sexually mature 10-week-old mice with different genotypes. We directly fixed tissues in 4% paraformaldehyde overnight at 4 °C. Fixed tissues were rinsed with PBS and dehydrated in a series of 70%–100% ethanol solutions and embedded in paraffin. Sections (5 μm) were prepared on glass slides and stained with HE. The stained testis and epididymis sections were observed using an optical microscope to determine the different germ cell types during varying stages and the amounts of spermatozoa.
Immunofluorescence and immunohistochemical analyses
After treating the samples as described, we prepared slides for immunostaining of testis and epididymis. We hydrated and boiled the slides for 10 min in 0.1 mol/L sodium citrate using a microwave oven and cooled them in an ice box for antigen unmasking. We washed the slides three times for 5 min in PBS and blocked them in 10% goat serum for 1 h at room temperature. After blocking, the samples were triple-washed with PBS and incubated with anti-DCAF8 antibody (Abcam) overnight at 4 °C. Subsequently, the slides were triple-washed again in PBS and incubated with Alexa Fluor 555-conjugated goat anti-rabbit secondary antibody (CST) at room temperature for 1 h in a dark room. After a last PBS wash, we used Hoechst 33342 to dye the nuclei and Alexa Fluor 488-conjugated lectin peanut agglutinin (PNA) to label sperm acrosomes. For sperm immunofluorescence analyses, we dissected the cauda epididymis, and the sperm that swam out at 37 °C in HBSS were smeared onto slides. We entrusted the immunohistochemical staining to the Biossci Biotechnology Company. We observed all the samples at room temperature using a microscope.
Computer-assisted semen analysis (CASA)
Parameters of sperm motility were measured by CASA using an integrated visual optical system software (Hamilton-Thorne IVOSII). In brief, we dissected two sides of the cauda epididymis from each mouse (10 weeks old) and cut them once to allow sperm dispersion. We collected the sperm samples into 1.5 mL tubes containing 500 μL of modified Krebs–Ringer’s bicarbonate solution (TYH medium) at 37 °C [
36]. After 5 min of incubation, sperm were added to a counting chamber for analysis, and we examined five fields for each cauda epididymis. The instrument settings of CASA were as follows: temperature, 37 °C; minimum cell size, five pixels; minimum contrast, 50; minimum static contrast, 25; low VAP cutoff, 20.0; low VSL cutoff, 30.0; threshold straightness, 50%; static head size, 0.3–1.95; static head intensity, 0.5–1.3; and magnification, 0.89.
Transmission electron microscopy (TEM) and scanning electron microscopy (SEM)
TEM and SEM analyses were performed at the Laboratory of Electron Microscopy, Shanghai Jiao Tong University School of Medicine, following standard protocols [
37]. In brief, testes were cut roughly into 1 mm cubes, and sperm isolated from cauda epididymis of both WT and
Dcaf8 KO mice were fixed in glutaraldehyde and osmium tetroxide (OsO
4). The specimens were then dehydrated in sucrose and embedded in epoxy resin. Ultrathin sections were prepared and picked up on nickel grids to be stained with uranyl acetate and methylcellulose. Finally, sections were observed under an electron microscope.
In vitro fertilization (IVF) assays
Female mice were injected with human chorionic gonadotropin, and the oocytes were collected after 14 h treatments and placed in human tubal fluid (HTF) medium covered with mineral oil. Sperm were freshly isolated from cauda epididymis of WT and Dcaf8 KO mice and incubated in HTF medium at 37 °C under a 5% CO2 environment for 1 h to allow capacitation. Oocytes were separated into two averaged pools and fertilized with sperm from either WT or Dcaf8 KO mice individually. The percentages of 2-cell, 4-cell, and 8-cell embryos examined 24, 48, and 72 h later, respectively, were used as IVF rates.
Statistical analysis
All data from sperm counts, motility and IVF assays were analyzed using the GraphPad prism 7 software and presented as mean±standard error of the mean. All the experiments were repeated at least three times. For each experiment, we used 3–7 animals. We performed two-tailed unpaired Student’s t-tests and Chi-square test to compare the data and considered P<0.05 as statistically significant.
Results
Expression patterns of Dcaf8 in adult mice
We detected the tissue expression profiles of Dcaf8 by RT-qPCR. We extracted RNA from 12 different tissues (brain, cauda epididymis, heart, liver, lung, ovary, kidney, spleen, adrenal gland, thymus, testis, and uterus). Fig. 1A shows the fold changes of Dcaf8 mRNA level normalized to that of Gapdh. Among all the detected tissues, the testis exhibited the highest Dcaf8 expression, suggesting that Dcaf8 could be involved in spermatogenesis. The expression of DCAF8 in tissues was further verified by Western blot. The highest expression of DCAF8 protein was shown in the liver, while it ranked second in testicular expression (Fig. 1B). Using publicly available protein data from different species, we found that the WD40 repeat region of Dcaf8 was highly conserved among human, monkey, mouse, rat, pig, and cattle DNA, further indicating that it may play an important biological function (Supplementary Table).
Generation of Dcaf8 KO mice
The mouse Dcaf8 gene includes 14 exons spanning approximately 7 kb. To investigate the function of Dcaf8 in vivo, Dcaf8 KO mice were generated by inserting 115 bp nucleotides into exon 6 of Dcaf8, resulting in frameshifts of its coding sequence. The strategy was shown in Fig. 1C. The KO effect was verified by PCR analysis (Fig. 1D) and Western blots of testis and cauda epididymis tissues (Fig. 1E and 1F). Homozygous mice could be obtained by intercrossing heterozygous mice and were viable with no noticeable abnormalities. The total body weight was comparable with their WT littermates (Fig. 3A).
Male Dcaf8 KO mice showed impaired fertility
To determine whether Dcaf8 is associated with the reproduction process, we carried out breeding experiments where WT and Dcaf8 KO mice were allowed to mate continuously with C57BL/6J females at a 1:1 ratio for 4 days. We found no differences in mating behavior between WT and Dcaf8 KO mice by checking their ability to produce vaginal plugs in the female mice. The FCPs were 68.75% and 51.56% in WT and Dcaf8 KO mice, respectively (Table 2). However, only 3 out of 33 WT females plugged by 16 Dcaf8 KO males became pregnant and produced only 15 pups. In parallel, the 11 WT females plugged by 4 WT males produced 64 offspring pups in 9 litters (Table 2). Together, these data suggested that DCAF8 proteins were needed for male fertility, and the impaired fertility phenotype was not associated with abnormal sexual behavior.
Localization of DCAF8 in testis
As Dcaf8 is expressed highly in testis, we performed immunofluorescence and immunohistochemistry experiments to investigate its distribution. We found that the DCAF8 protein was expressed in the cytoplasm of spermatocytes and almost absent in elongated sperm and mature sperm of adult mice. In addition, we confirmed the Dcaf8 KO effect (Fig. 2A–2D). To further verify the distribution pattern, we extracted cytoplasmic and nuclear proteins after testis digestion into single cells and found that DCAF8 showed cytoplasmic expression but was absent from nuclei (Fig. 2E). Such a unique expression pattern may indicate an essential physiologic role of Dcaf8 during spermiogenesis in male mice.
Sperm counts and motility of Dcaf8 KO male mice
To explore the mechanisms resulting in the impaired fertility of Dcaf8-deficient mice, we focused on the phenotypes of Dcaf8 KO mice. The weights of testis and epididymis from sexually mature WT and Dcaf8 KO mice were similar (Fig. 3B). We found no significant quantity differences in sperm production from the cauda epididymis between WT and Dcaf8 KO mice (Fig. 3C and 3D). All these observations suggested that sperm production was unlikely to be associated with the impaired fertility of Dcaf8 KO males. Moreover, we assessed the sperm motility activation after releasing sperm from the cauda epididymis by CASA. The movement characteristics including progressive motility, path velocity (VAP), progressive velocity (VSL), and track speed (VCL) were significantly lower in Dcaf8 KO mice than in WT ones (Fig. 3E and 3F). However, total motility values were similar between WT and Dcaf8 KO sperm (Fig. 3F).
Defective sperm morphology of Dcaf8 KO mice
To examine sperm morphology of Dcaf8 KO mice, we prepared sperm smears and stained them with Diff-Quick stain. The sperm from Dcaf8 KO male mice showed higher percentages of bent head malformation than those of WT mice (data not shown). To understand the sperm defects in detail, we observed the sperm ultrastructure by SEM and TEM. Almost all the WT sperm were normal with a hook-shaped head, midpiece, and tail structures (Fig. 4A), whereas up to 40% of Dcaf8 KO mice sperm exhibited bent head malformation (Fig. 4B), and the typical sperm head can be seen in Fig. 4C and 4D. To further analyze Dcaf8 KO sperm defects, we stained WT and mutant sperm with PNA (green) and DAPI (blue), labeling the acrosomes and nuclei, respectively. No differences were observed between WT and Dcaf8 KO sperm, and they all showed typical crescent-like acrosomes (green) covering the anterior portion of the head, even in the abnormal sperm of Dcaf8 KO mice (Fig. 4E and 4G). Immunofluorescence showed that the number of spermatogenic cells were not influenced by Dcaf8 KO (Fig. 4F). As mentioned above, the sperm bent head malformation may partially explain the impaired fertility of Dcaf8 KO male mice.
Dcaf8 KO males showed severely reduced in vitro fertilization rates
We performed IVF assays to assess whether the fertility of Dcaf8 KO sperm in vitro is disrupted or not. The two-cell stage embryos (24 h after sperm mixed with oocytes) of WT sperm appeared at a rate of approximately 60%, while those of Dcaf8 KO sperm were dramatically fewer at a rate of approximately 8% (Fig. 5A). Furthermore, the 4-cell and 8-cell division stages were significantly better in Dcaf8 WT mice than in the KO mice at 48 and 72 h, respectively (Fig. 5B and 5C). These results may explain the low in vivo fertilization rate of Dcaf8 KO mice shown in Table 2. These findings suggested that the Dcaf8 KO male mice’s inability to reproduce could be explained by their low in vivo migration ability and in vitro sperm–egg binding ability.
Discussion
Mammalian male fertilization is a complex event for which success depends on multiple steps, including the generation of spermatozoa carrying intact genomes and capability of fusion with the egg. Various factors have been reported to be essential for fertilization in studies targeting mice genes [
38,
39]. However, many specific genes with crucial roles during fertilization still need to be uncovered to provide guidance on diagnosis and treatments for male patients.
Many E3 ubiquitin ligases are essential for spermatogenesis regulation from spermatogonia to mature sperm stages [
40,
41]. Many of the genes involved are expressed at high rates and/or uniquely in the testis, including
Cul4a, Cul4b, and
Dcaf17 [
10,
24,
26,
42]. Studies have shown that deletion of
Ddb1, which is associated with both
Cul4a and
Cul4b, leads to defects in oocyte survival, oocyte meiotic maturation, and post-meiotic oocyte maintenance [
43–
45]. These findings suggested that the CRL4 E3 ubiquitin complex is indispensable during germ cell development. In this study, we revealed that DCAF8, a previously uncharacterized E3 ubiquitin ligase substrate adapter, plays an important role during male fertilization, and the fertility of male mice deficient in
Dcaf8 is significantly impaired.
Interestingly, the
Dcaf8 sequence is highly conserved among several species. On the basis of the predominant expression of
Dcaf8 mRNA levels in testis shown in our study,
Dcaf8 may play an essential testicular function role. Mice with
Dcaf8 deficiency displayed normal embryonic development without clear anatomical abnormalities. We verified the impaired fertility of
Dcaf8 male KO mice using mating tests as shown in Table 2. Our results showed that the frequency of conception after mating with WT male mice was 81.82%, while that after mating with
Dcaf8 KO mice decreased to 9.09%. However, the FCP were similar between these two groups, implying that the
Dcaf8-deficient male mice displayed normal sexuality. Studies have revealed that
Cul4a KO mice testis was twofold smaller in size and weight compared with WT testis [
10]. However,
Dcaf8 KO mice had no effects on the size or weight of their testis in our study. The normal mating behavior of our
Dcaf8 KO mice and their apparently normal anatomical structures suggested that the impaired infertility was not related to aberrant reproductive physiology, sex organ development, or sexual behavior.
To achieve successful fertilization, enough sperm need to be deposited into the female reproductive tract; sperm with normal morphology, vigorous motility, and sound fertilizing capabilities are necessary [
46,
47]. To uncover the mechanisms leading to impaired fertility of the
Dcaf8 KO male mice, we assessed all the variables mentioned above. However, the amount of sperm produced by the
Dcaf8 KO mice was similar to that of their WT littermates. Moreover, CASA analysis revealed similar total motilities between the two genotypes, while the
Dcaf8 KO mice showed just slightly lower progressive motility, VAP, VSL, and VCL than the WT mice. We observed aberrant sperm morphology under SEM and TEM with almost half of the spermatids exhibiting bent head malformation. This specific malformation may have adverse effects on the fertilization rate. In addition, the reduced fertilization rate was verified by IVF experiments. To a considerable degree, the impaired fertility phenotype in
Dcaf8 KO male mice was caused by the sperm’s defective morphology and attenuated motility. The subfertility phenotype observed in
Dcaf8 KO male mice demonstrated vital functions of DCAF8 proteins during spermatogenesis. DCAF8 is a member of DCAF family proteins; it serves as a substrate receptor that provides substrate specificity to ubiquitin for CRL4 E3 ligase complexes. Spermatogenesis is affected in
Cul4A,
Cul4B, Dcaf17, and
Dcaf8 KO mice [
10,
24,
26], and
Dcaf1 and
Dcaf13 KO mice show vital functions in maintaining mammalian oogenesis [
44,
48]. One can reasonably assume that the CRL4 E3 ligase complexes may be involved in the regulation of the reproductive system via degrading certain proteins through ubiquitination, and the insights observed may also be applicable to humans. However, these assumptions need to be explored by further investigation.
In the study, we used Dcaf8-deficient mice to demonstrate the significant role of Dcaf8 in male fertility. Even though DCAF8 is exclusively expressed in spermatocytes and is nearly vanished in mature sperm, it was found to participate in sperm motility and morphogenetics. Further dissecting this functional network will enhance our understanding of this E3 ligase complex component during spermiogenesis.
PDF
(1778KB)
