Abstract
Asthenoteratospermia is a common cause of male infertility. Recent studies have revealed that CFAP65 mutations lead to severe asthenoteratospermia due to acrosome hypoplasia and flagellum malformations. However, the molecular mechanism underlying CFAP65-associated sperm malformation is largely unclear. Here, we initially examined the role of CFAP65 during spermiogenesis using Cfap65 knockout (Cfap65−/−) mice. The results showed that Cfap65−/− male mice exhibited severe asthenoteratospermia characterized by morphologically defective sperm heads and flagella. In Cfap65−/− mouse testes, hyper-constricted sperm heads were apparent in step 9 spermatids accompanied by abnormal manchette development, and acrosome biogenesis was abnormal in the maturation phase. Moreover, subsequent flagellar elongation was also severely affected and characterized by disrupted assembly of the mitochondrial sheath (MS) in Cfap65−/− male mice. Furthermore, the proteomic analysis revealed that the proteostatic system during acrosome formation, manchette organization and MS assembly was disrupted when CFAP65 was lost. Importantly, endogenous immunoprecipitation and immunostaining experiments revealed that CFAP65 may form a cytoplasmic protein network comprising MNS1, RSPH1, TPPP2, ZPBP1 and SPACA1. Overall, these findings provide insights into the complex molecular mechanisms of spermiogenesis by uncovering the essential roles of CFAP65 during sperm head shaping, acrosome biogenesis and MS assembly.
Introduction
Spermiogenesis is the final phase of sperm development that involves complex and highly ordered spermatid differentiation. During this process, round haploid germ cells undergo remarkable nuclear and cytoskeletal modifications to transform into the highly polarized spermatozoa with the capacity for fertility (1). Three key events are included: development of the acrosome covering the apical pole of the nucleus; assembly of the flagellum as a motility apparatus; and nuclear condensation and remodeling into a species-specific shape (2,3). Defects in genes involved in these processes contribute to flagella immobility and acrosome deformity, resulting in male infertility frequently seen in asthenoteratospermia, including multiple morphological abnormalities of the sperm flagellum (MMAF) and globozoospermia (4–7). Understanding the molecular causes of asthenoteratospermia requires addressing how complex molecular machines get composed during spermiogenesis.
Previous studies have uncovered that CFAP65 (OMIM: 614270) mutations cause severe asthenoteratospermia in humans manifesting acrosome hypoplasia and flagellum malformations (8–10), but the underlying mechanisms of CFAP65 and its interactome networks involving spermiogenesis are still unknown. Unlike other identified asthenoteratospermia-related genes (11), CFAP65 has been shown to localize at both the acrosomal region and the flagellar mid-piece of spermatozoa in humans (8). Hence, further investigation of the functional role of CFAP65 during spermiogenesis could offer new insights into sperm development by linking up at least two essential events (the sperm acrosome formation and flagellar assembly) via one single gene.
A mouse model can be used to address the function of one gene during spermiogenesis. In mice, the process of spermiogenesis has been divided into 16 identified steps with four acrosome phases (12,13). During acrosome biogenesis, the Golgi sorts and traffics proacrosomal vesicles to the nuclear surface (Golgi phase, steps 1–3), followed by a fusing process of these proacrosomal granules to form a large acrosomal granule near the nuclear surface, with the acrosome gradually spreading over the nuclear membrane (Cap phase, steps 4–7). Subsequently, the acrosome undergoes condensation, attaches itself to the inner acrosomal membrane (IAM) (acrosome/elongation phase, steps 8–12) and then migrates to the final sites, whereas the nucleus goes through a few morphological changes (maturation phase, steps 13–16) (14). The attachment of the developing acrosome to the nuclear surface seems to be mediated by the perinuclear theca, which is a cytoskeletal-based structure with unclear components between the acrosomal and nuclear membranes (2,15). The manchette assembles at step 8 of the spermiogenesis involved in spermatid head shaping and acrosome formation and serves as a protein transport platform for intramanchette transport (IMT), which is essential for the assembly of the mitochondrial sheath (MS) and outer dense fibers (ODFs) until its disassembly at step 14 (1,16). The IMT has also been suggested to communicate with the intraflagellar transport (IFT), which is required in the sperm flagellar development (16,17). Overall, it is still unclear which processes CFAP65 may be involved in as the spermiogenesis proceeds.
In this study, we report novel findings of sperm phenotypes when CFAP65 is deficient and identify new partners of CFAP65 during spermiogenesis. The deletion of CFAP65 causes a cascade of development defects in both sperm head and flagella and leads to oligoasthenoteratospermia in mice. Our data highlight the multifaceted role of CFAP65 based on phenotypic and molecular evidence, which provides a new perspective on the complex regulatory networks during spermiogenesis.
Results
Cfap65−/− male mice exhibited severe oligoasthenoteratospermia
Our previous study revealed that CFAP65 mutations cause severe asthenoteratospermia in humans because of acrosome hypoplasia and flagellum malformations (8). CFAP65 was showed high expression in testis and cilia-related tissues (Supplementary Material, Fig. S1). To further evaluate the role of CFAP65 during spermiogenesis in vivo, we generated a Cfap65-knockout mouse model using clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) technology. Single-guide RNAs (sgRNAs) were designed to target exons 4 and 11 of Cfap65 (Fig. 1A and Supplementary Material, Fig. S2A). Knockout genotype identification and efficiency were confirmed using polymerase chain reaction (PCR) combined with Sanger sequencing (Supplementary Material, Fig. S2B), western blotting (Fig. 1B) and immunofluorescence (IF) (Supplementary Material, Fig. S3A and B). Specifically, using a customized C-terminal antibody, CFAP65 was shown to be expressed in the cytoplasm of the elongating spermatids and the flagella of the spermatozoa in Cfap65+/+ mice but was absent in Cfap65−/− mice (Supplementary Material, Fig. S3A and B). The body weight and testis weight of 2-month-old Cfap65−/− mice was significantly lower than those of Cfap65+/− and Cfap65+/+ mice (Fig. 1C), and the Cfap65−/− mice also displayed developmental delay (data not shown). Mating tests confirmed that Cfap65−/− males were able to mate normally, but failed to produce any pups, whereas the Cfap65+/− males generated comparable numbers of pups to the Cfap65+/+ males (n = 6, Supplementary Material, Fig. S4A). We then focused on the fertility-associated phenotypes of Cfap65−/− mice.
Cfap65 knockout leads to severe oligoasthenoteratospermia in male mice. (A) Schematic representation of the recombination events targeting Cfap65 exon 4–11 using CRISPR/Cas9 technology. (B) Western blotting showed the absence of CFAP65 in the testes of Cfap65−/− mice at 2 months of age. (C) Cfap65+/+, Cfap65+/−, Cfap65−/− mice at 2 months of age. Testes of Cfap65−/− mice were smaller than those of Cfap65+/+, Cfap65+/− mice. The body weight and testis weight were highly reduced in Cfap65−/− mice; *P < 0.05. (D) H&E staining of seminiferous tubules from Cfap65+/+ and Cfap65−/− mice at 2 months of age. (E) H&E staining of caudal epididymis from 2-month-old Cfap65+/+ and Cfap65−/− mice. (F) Sperm counts and percentages of motile sperm in 2-month-old Cfap65+/+ and Cfap65−/− mice; ***P < 0.001. (G) H&E staining of epididymis spermatozoa from Cfap65+/+ and Cfap65−/− mice. The aberrant sperm in Cfap65−/− mice showed malformed flagella, hyper-constricted head shapes or redundant cytoplasm. Scale bars: 10 μm.
Histological analysis of the testes showed that the Cfap65−/− mice had no apparent abnormalities in proliferating spermatogonia, and there were spermatocytes and round spermatids in the seminiferous tubules in the Cfap65−/− mice (Supplementary Material, Fig. S4C). However, there was a substantial decline in elongating spermatids reaching the seminiferous tubule lumen compared with the Cfap65+/+ mice (Fig. 1D). Importantly, we observed elevated pyknotic germ cells with dark staining characteristic of apoptosis in stage XI–XII, and these cells seemingly belonged to spermatocytes (Supplementary Material, Fig. S4C and Fig. 2B). Correspondingly, terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay also showed a marked increase of the apoptotic germ cells close to the basal aspect of the tubule in Cfap65−/− testes compared with that in Cfap65+/+ mice (Supplementary Material, Fig. S4D), which may account for the oligozoospermia phenotype. However, it is still yet to be determined which apoptotic pathway was triggered by the absence of CFAP65 deficiency. The number of spermatozoa in the cauda epididymis from Cfap65−/− mice was also significantly reduced and appeared 100% immotile in contrast with those in Cfap65+/− and Cfap65+/+ mice (Fig. 1E and F). In comparison with Cfap65+/+ mice sperm, the sperm of Cfap65−/− mice demonstrated severe morphological defects (Fig. 1G) including hyper-constricted head shapes and multiple flagellum deformities (shorten, bent, curved or absent sperm flagella) (Supplementary Material, Table S1). In addition, there were large amounts of beheaded sperm with separated heads scattered in the Cfap65−/− smears, indicating a fragile neck region in these spermatozoa (Supplementary Material, Fig. S4B). Taken together, as CFAP65 deficiency led to increased spermatocyte apoptosis and a reduced number of haploid germ cells entering spermiogenesis along with multiple abnormalities in sperm head shaping and tail formation, we suggest that CFAP65 may be required both in the function of spermatocytes and the subsequent haploid germ cell differentiation.
Malformed sperm head and manchette in Cfap65−/− male mice. (A) SEM of epididymis spermatozoa from Cfap65+/+ and Cfap65−/− mice. Misshapen heads (asterisk) and abnormal flagella were apparent in Cfap65−/− mice. Scale bars: 5 μm. Histogram indicating the sperm head morphology in Cfap65+/+ and Cfap65−/− mice as shown on the right (three mice of each genotype were evaluated). Data represent mean ± SD; **P < 0.01. (B) The Periodic acid–Schiff (PAS) staining of seminiferous tubules at stage IX and XII in Cfap65+/+ and Cfap65−/− mice. Compare with the Cfap65+/+ mice, abnormal shaping of sperm heads in Cfap65−/− mice were shown to have swollen anterior and tapered posterior. Arrowheads indicate the pyknotic spermatocytes. Scale bars: 10 μm. (C) TEM of testes from Cfap65+/+ and Cfap65−/− mice showing spermatids at steps 8–15. Reduced electrodense of the cytoskeleton components (asterisk) in the PAR at steps 8–10 were showed in Cfap65−/− mice. The equatorial zone was shown in the enlarged insets. The manchette microtubules at steps 11–13 seemed abnormally elongated and malformed in Cfap65−/− mice. The distance from the perinuclear ring to the nuclear cauda was indicated by double-headed arrows. Misshapen sperm heads with abnormal acrosomes (arrowhead) at steps 14–15 in Cfap65−/− mice were also shown. Scale bars: 2 μm.
Cfap65−/− male mice exhibit abnormal sperm head shaping and deficiency in manchette organization during spermiogenesis
Compared with Cfap65+/+ mice, Cfap65−/− mice had increased head deformities in sperm (11.4 ± 4.6 versus 65.7 ± 4.5%) (Fig. 2A), which was also observed in patients with CFAP65 mutations (8). To investigate the possible cause of formation of abnormal sperm heads when CFAP65 defects occurred, we traced the process of sperm nuclear shaping during spermiogenesis. Importantly, we observed a distinct abnormality of Cfap65−/− sperm head from the Periodic acid–Schiff (PAS) staining of the testis, characterized by rounded ends of the anterior head and tapered ends of the posterior head at step 9 (stage IX) (Fig. 2B). As the nucleus continued elongating, compared with the falciform spermatozoa with a hooked anterior and a relatively flat posterior in Cfap65+/+ mice, the most common nuclear shape abnormality in Cfap65−/− mice was a bulged shape of the apical head and a narrow cylindrical shape of caudal part (Fig. 2B), reminiscent of the phenotype in the azh/azh mouse model with manchette defects (18).
The manchette is a transient microtubule- and F-actin-based structure formed during spermiogenesis (steps 8–14 in mice), which is also the most influential factor in shaping the sperm heads (16). The testis transmission electron microscopic (TEM) results showed an ectopic manchette with respect to the edge of the acrosome in steps 8–10 spermatids of Cfap65−/− mice compared with Cfap65+/+ mice. The equatorial zone (including the marginal ring and circumferential groove) and the post-acrosomal region (PAR) in steps 8–9 spermatids of the Cfap65−/− mice showed lower electron density (Fig. 2C), implying a deficiency of cytoskeleton components, such as F-actin, calmodulin or α-spectrin, at these areas. We also observed the abnormal manchette organization in Cfap65−/− mice specifically characterized by elongated or misshapen manchette microtubules in steps 11–13 spermatids (Fig. 2C and Supplementary Material, Fig. S5A and B), suggesting there might be a defect in the distal migration of manchette. The posterior part of nucleus starting from the perinuclear ring appeared lengthened and malformed in Cfap65−/− mice (Fig. 2C), consistent with the narrow cylindrical shape of the caudal nucleus observed in the PAS staining (Fig. 2B). This may be due to the continuing constriction of the sperm nucleus during its elongation by the detained manchette microtubules. All these findings indicated an error in the normal organization of manchette and/or related cytoskeleton structures during spermiogenesis in Cfap65−/− mice, which may be related to the sperm head shaping.
Abnormal acrosome biogenesis in Cfap65−/− male mice
During the later steps of spermiogenesis (steps 14–15), the defects in head shapes and acrosomal morphologies were manifest, and the aberrant acrosomes were separated from the nuclei (Fig. 2C). The epididymis sperm of Cfap65−/− mice also displayed either absent or morphologically defective acrosomes compared with normal controls. Most of the acrosomes showed less condensed contents with barely recognizable inner and outer acrosomal membranes (IAMs and OAMs) (Supplementary Material, Fig. S6A). A previous study showed CFAP65 staining (using an N-terminal antibody) in the acrosome region and mid-piece of human spermatozoa (8). Combined with the presence of abnormal acrosomes in Cfap65−/− mice, we were prompted to investigate the acrosome biogenesis during spermatogenesis. In Cfap65+/+ mice, all four acrosome biogenesis phases, including the Golgi, cap, elongation and maturation phases, could be identified by their standard characteristics through TEM of testes (13). However, in Cfap65−/− mouse testes, the acrosome attached anomalously along the perinuclear region in the maturation phase (Fig. 3A), and there were irregularly shaped swellings with several vacuole-like contents in the perinuclear theca. These results suggest that the malformed acrosomes in Cfap65−/− mice might result from the failure of acrosome condensation and its attachment to IAM during the later stages of the acrosome biogenesis. We also detected the acrosomal structures before the completion of the first spermatogenic wave using Cfap65+/+ and Cfap65−/− mice from postnatal day 21, in which the acrosomal formation was supposed to progress into the elongation phase (Supplementary Material, Fig. S6B). However, there seemed to be a delay in the spreading of acrosomal granules on the nuclear surface in Cfap65−/− mice, with fewer round spermatids showing a flattened acrosomal structure with a very thin layer on the nucleus. Taken together, these results indicate that acrosome biogenesis was affected in Cfap65−/− mice.
Analysis of acrosome biogenesis by TEM and PNA staining in Cfap65+/+ and Cfap65−/− male mice. (A) TEM showed seemingly similar acrosome biogenesis in the testes of Cfap65+/+ and Cfap65−/− mice during the Golgi, cap and elongation phases. The maturation phase showed well-formed, condensed acrosomes in Cfap65+/+ mice, but irregularly shaped acrosomes with uncondensed or vacuole-like contents (asterisk) in the perinuclear theca of Cfap65−/− mice. (B) Acrosome and proacrosomal vesicle staining using PNA-FITC. The acrosomes during the maturation phase are shown with an abnormally expanding staining pattern (red arrow) in Cfap65−/− mice compared with Cfap65+/+ mice. Moreover, a part of the PNA staining in the elongation phase also began to exhibit deformities along with malformed sperm heads (arrow) in Cfap65−/− mice. Scale bars: 20 μm. (C) The PNA staining of epididymis spermatozoa was absent from Cfap65−/− mice compared with Cfap65+/+ mice. Scale bars: 10 μm. The histogram on the right indicates the statistics for PNA staining in the sperm from Cfap65+/+ and Cfap65−/− mice (n = 200).
To further examine how the Cfap65 defect affects acrosome biogenesis, the acrosome was labeled with fluorescein isothiocyanate (FITC)-conjugated peanut agglutinin (PNA), which could bind to the OAMs. The results show that a single acrosomal vesicle was attached to one end of the nucleus in spermatids of Cfap65+/+ mice and then gradually spread along the nuclear surface, indicating that proacrosomal vesicles derived from the Golgi apparatus were fused (Fig. 3B). In contrast, in the maturation phase of Cfap65−/− mice, poorly condensed PNA-positive structures with a more expanded staining pattern were observed in the perinuclear region (Fig. 3B), which was similar to the observations from TEM analysis. Moreover, the PNA staining was absent or aberrant in most Cfap65−/− spermatozoa (Fig. 3C), suggesting that a large amount of the loose acrosome structures may detach from the nucleus during epididymal transport and ejaculation. These results indicate that Cfap65 may be involved in maintaining the cap-like structure to form the compact moon-shaped structure covering the nucleus. This putative role of CFAP65 is also consistent with its function in sperm head shaping and manchette organization, manifesting as a scaffold protein to maintain the connection between acrosome, nuclear lamina and manchette.
The sperm of Cfap65−/− male mice showed misassembled MS
We also used TEM to analyze the flagellum ultrastructure in Cfap65-deficient sperm. The cross-sections of sperm flagella of Cfap65−/− mice showed disorganized axoneme and peri-axoneme structures, such as absence or disorganization of outer doublet microtubules and ODFs (Fig. 4A). Interestingly, the central pair complex (CPC) was frequently absent in the mid-piece and principal piece, but was mostly formed in the end-piece (Fig. 4A). Partitioned MS in the mid-piece was observed in Cfap65−/− mice, and the mitochondrial contents seemed abnormally vacuolated or tenuous (Fig. 4A), which was also corresponding to frequent absences of TOMM20 (an MS marker) signal in the flagella of Cfap65−/− mice (Supplementary Material, Fig. S7A). Through testis TEM, we also observed redundant recruitment of mitochondria that could not be correctly assembled in the Cfap65−/− testes (Fig. 4B), which may relate to an unstable formation of MS and lead to the detachment of the MS after spermiation. Overall, these results indicate that a lack of Cfap65 may also lead to various defects in sperm flagellum assembly.
Defects in sperm flagellum ultrastructure and MS assembly in Cfap65−/− male mice. (A) TEM of the cross-sections of sperm flagella showed abnormally arranged mitochondria with vacuolated or tenuous contents (asterisk) in the mid-piece, disorganized/absent outer dense fibers (ODFs) and axonemal components, such as central pairs, outer doublet microtubules (DMTs) and radial spokes (arrow). Scale bars: 100 nm. (B) TEM analysis of MS biogenesis during spermiogenesis in Cfap65+/+ and Cfap65−/− mice. In Cfap65−/− mice, the mitochondria around the ODFs were accumulated and could not be correctly assembled. The axonemal structure was also absent in Cfap65−/− mice (asterisk). Scale bars: 1 μm.
Expressional pattern of proteins for acrosome formation and flagellum assembly in Cfap65−/− mice
To better elucidate the molecular basis for these phenotypes caused by Cfap65 deficiency, liquid chromatograph-mass spectrometer/mass spectrometer-based proteomic analysis was performed to assess protein expression profiles in adult Cfap65−/− mouse testes. In the proteomic analysis of Cfap65+/+ and Cfap65−/− mouse testes, a total of 6545 proteins were quantified, of which 59 proteins were significantly upregulated and 123 proteins were significantly downregulated (Fig. 5A and Supplementary Material, Table S2), and a good consistency among mice was seen (Supplementary Material, Fig. S8). According to the literature, the expression profile for the acrosome biogenesis and flagellar assembly changes significantly, wherein a considerable number of proteins required for the acrosomal formation or head shaping, such as ENKUR, CRISP2, SPACA1, ZPBP, CCIN and CAPZA3; for flagellar structure maintenance, such as FSIP2, ODF2, AKAP4 and RSPH1; and for centriole functions, such as CFAP53, SPATC1 and NME7, were significantly reduced (Fig. 5B). In addition, most calcium pathway proteins, such as SLIT2, EFHC2 and EFHC1; IMT or IFT process proteins, such as SMRP1, MNS1, IFT20 and TRAF3IP1; phosphorylation function proteins, such as TSSK4, ROPN1 and GAPDHS; and cation transport-related proteins, such as AQP7, ATP1A4, ATP1B3, but not ATP1B2, were also significantly reduced (Fig. 5B). Several mitochondrial function-related proteins, such as NSUN4 and TPPP2 were reduced, whereas others, such as GTPBP10, MRPI47 and PDK1, were abnormally elevated. In addition, testis-enriched ubiquitinated proteins, such as UBQLN3 and OTUB2, were downregulated, whereas others, such as RNF181, RNF2 and USP39, were upregulated (Fig. 5B). Gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG) pathway analysis showed that several crucial functions related to sperm motility, cilium assembly and fertilization were enriched (Supplementary Material, Fig. S9).
Changes of protein expression during CFAP65 deficiency in mouse testes. (A) Proteins with altered expression from the proteomics analysis in the testes of Cfap65−/− mice shown by volcano plots. A total of 59 significantly upregulated (red) and 123 significantly downregulated (green) proteins were found (ratio >1.2). (B) Function-enrichment analysis of the differentially expressed proteins in the testes of Cfap65−/− mice compared with Cfap65+/+ mice at 2 months of age. (C) Western blotting validated the expressions of TPPP2, RSPH1, SLIT2, EFHC1, ENKUR, ZPBP, SPACA1, MNS1, DYNLT1 and IFT20 in testes from Cfap65+/+ and Cfap65−/− mice.
Based on the proteomic results and phenotypic characteristics in Cfap65−/− mice, we selected several differentially expressed proteins related to acrosome formation and flagellum assembly for verification. Western blotting subsequently confirmed the decreased expression of 10 proteins in the Cfap65−/− mice compared with the Cfap65+/+ mice, including three acrosome-related proteins, namely ZPBP, SPACA1 and ENKUR, two calcium pathway proteins, SLIT2 and EFHC1, one mitochondrial protein, TPPP2, one axoneme radial-spoke (RS) protein, RSPH1, and three IMT- or IFT-related proteins, MNS1, DYNLT1 and IFT20 (Fig. 5C). Taken together, these results strongly suggested that a proteostatic system within sperm acrosome formation and flagellar assembly or maintenance were disrupted when CFAP65 was absent.
Proteostasis network of CFAP65 predicted by STRING and confirmed by endogenic immunoprecipitation. (A) The interaction network of CFAP65 with other proteins from different functional groups analyzed using the STRING software. The network was centered by CFAP65 with the predicted primarily interactive proteins distributed (green color). Other groups of proteins characterized were also shown, and their expression levels between Cfap65+/+ and Cfap65−/− mice are differentiated by different colors. Blue, downregulated; red, upregulated; gray, unchanged. (B) Endogenous CFAP65 immunoprecipitation using a customized CFAP65 C-terminal antibody (amino acids 1401–1635) showed its binding of MNS1, RSPH1, TPPP2, ZPBP, SPACA1, but not DYNLT1 and SLIT2 from Cfap65+/+ mouse testes at postnatal day 60. IgG was used as a control. Protein molecular weights displayed in KDa.
CFAP65 may interact with key proteins participating in spermiogenesis
Spermiogenesis is a strictly regulated process, in which every single gene is equipped with an elaborate network. We analyzed the interaction network of CFAP65 using the STRING software. When centered by CFAP65, key biological events, such as acrosomal formation, flagellar assembly, IMT or mitochondrial function, were connected by multiple predicted proteins differentially expressed/unchanged in Cfap65−/− mice through proteomic analysis (Fig. 6A). We speculated that CFAP65 could form a successive set of complexes with several of these proteins to fulfill its roles during acrosome formation and flagellar assembly. To examine the interactions between CFAP65 and the related proteins, we immunoprecipitated endogenous CFAP65 from postnatal Cfap65+/+ testes (D60). Finally, we found endogenous interactions of CFAP65 with several proteins, including MNS1, RSPH1, TPPP2, ZPBP1 and SPACA1 (Fig. 6B) but not SLIT2 and DYNLT1, which might explain the phenotypes observed in Cfap65−/− mice.
Immunodetection of SPACA1, ZPBP, TPPP2 and SLIT2 in the seminiferous tubules of Cfap65+/+ and Cfap65−/− mouse testes during different stages. (A) Compared with Cfap65+/+ testes, a reduced and irregular expression of SPACA1 at steps 5–6 (stages V–VI) and steps 11–12 (stages XI–XII) spermatids in Cfap65−/− testes were shown. (B) Compared with Cfap65+/+ testes, a reduced expression of ZPBP at steps 5–6 (stages V–VI) and steps 11–12 (stages XI–XII) spermatids in Cfap65−/− testes were shown. (C) Compared with Cfap65+/+ testes, a reduced expression of TPPP2 in sperm MS at steps 15–16 (stages VI–VIII) and in the spermatid cytoplasm at steps 11–12 (stages XI–XII) in Cfap65−/− testes was shown. (D) SLIT2 staining of seminiferous tubules from Cfap65+/+ and Cfap65−/− mouse testes. A failure of SLIT2 assembly in the sperm flagella at step 15 (stages VI–VII) spermatids and an abnormal expressional pattern at steps 11–12 (stages XI–XII) spermatids in Cfap65−/− testes is shown. Scale bars: 5 μm.
To investigate the underlying link between CFAP65 and the related proteins, IF was performed for Cfap65+/+ and Cfap65−/− testes samples. SPACA1, a protein proposed to play a role in the acrosome–acroplaxome interaction (19), showed an abnormal localization in steps 11–12 spermatids (Fig. 7A) and was absent in the mature spermatozoa (Supplementary Material, Fig. S7B) in Cfap65−/− mice. ZPBP, an acrosomal protein essential for the integrity of acrosome matrix (14), exhibited a reduced expression from the round spermatids and failed to locate in the acrosomal region of the elongated spermatids (Fig. 7B). These results suggest that loss of CFAP65–SPACA1 and CFAP65–ZPBP interactions may underlie the pathogenic effect of acrosome in Cfap65−/− mouse and CFAP65 mutant patients. Moreover, the mitochondrial protein TPPP2 also displayed interactions with CFAP65 endogenously (20). The immunostaining of TPPP2 showed a highly reduced expression in the cytoplasm in steps 11–12 spermatids and a lack of MS-like structure in step 15 spermatids in Cfap65−/− mouse testes, which could explain the defects in mitochondrial contents and MS assembly (Fig. 7C). In addition, although the expression of SLIT2 and DYNLT1 were reduced in Cfap65−/− mice, and a defect in SLIT2 recruitment and migration during CFAP65 deficiency was observed (Fig. 7D), there were no interactions between these two proteins and CFAP65, suggesting other mechanisms may exist for the involvement of CFAP65 in the calcium pathway and IFT process. Taken together, perturbations to the proteostasis network associated with CFAP65 may cause a series of defects in sperm head and flagella, leading to severe asthenoteratospermia.
Discussion
In this study, we characterize the defective sperm phenotypes in Cfap65−/− mice and try to understand the mechanism behind CFAP65 dysfunction by proteomic analysis and endogenous immunoprecipitation (IP). With our findings, we highlight the crucial role of CFAP65 during sperm head shaping, acrosome formation and flagellar assembly and demonstrate that CFAP65 may form a cytoplasmic protein network comprising MNS1, RSPH1, TPPP2, ZPBP1 and SPACA1 (Fig. 8) to fulfill its functions.
Working hypothesis for the role of CFAP65 in acrosome biogenesis, head shaping and flagellar assembly. Schematic illustration of the mechanisms that are possibly dependent on the presence of the CFAP65. Multiple proteins may participate in the acrosome biogenesis (ZPBP and SPACA1), head shaping (MNS1) and flagellar assembly (TPPP2 and RSPH1) by forming functional complexes with CFAP65. Loss of CFAP65 leads to asthenoteratospermia characterized by an abnormal acrosome, misshapen head and malformed MS along with highly reduced motility.
The critical events, i.e. sperm acrosome formation along with head shaping, flagellar assembly, and nuclear condensation and modification, during spermiogenesis occur in an overlapping manner and are often co-dependent, wherein hundreds of proteins are attached to their dedicated working networks to execute distinct steps (2,13,16). For example, the lack of acrosome formation could initiate a cascade of incorrectly regulated spermiogenic steps and disturb sperm head elongation and/or flagellum formation (21).
In both humans with CFAP65 mutations (8) and Cfap65−/− mice, we found acrosomal abnormality and morphologic defects in both sperm head and flagella, although there seemed to be several species-related differences between human and mouse when considering the embodied forms and severity degrees. Importantly, these defects may be initiated by one or interfering with each other during CFAP65 absence, which has caused us great difficulties in investigating the role of CFAP65 during spermiogenesis. However, different CFAP65 splicing isoforms are translated during spermiogenesis, which has been supported by recent single-cell sequencing of human, macaque and mouse testes (22). So, these different phenotypic defects may be caused by different isoforms incapable of their functions during acrosome biogenesis, sperm head shaping and/or flagellar assembly. To some extent the localization of CFAP65 may provide some hints: using the N-terminal antibody, CFAP65 staining was at the acrosomal region (especially the equatorial segment) and mid-piece flagella of human spermatozoa (8), whereas using a C-terminal antibody, the staining was showed at the whole sperm flagella (9). In this study, we generated a C-terminal antibody specifically binding the mouse CFAP65 amino acids 1401–1635 and found a localization of CFAP65 in the whole flagella of mouse spermatozoa. We can have a bold speculation that a shorter form of CFAP65 with a truncated C-terminus may function in the head shaping, acrosome formation and MS assembly, whereas a longer form of CFAP65 is essential for the assembly and maintenance of sperm flagella. However, further biochemical studies are required to define the specific functions of different CFAP65 isoforms.
The nucleus changes its shape at the acrosome phase during acrosome biogenesis (23). Deficiency in proteins related to the acroframosome–acroplaxome–manchette axis (24) could cause head deformity and acrosomal defect (25). In our study, the manchette is malformed when CFAP65 is absent, causing a special nucleus shape, i.e. bulges at the anterior ends and tapers or has a column shape at the posterior, with loosened acrosomal structure, which have been rarely reported in previous studies with a genetic basis. The equatorial area is where the acrosome, nuclear lamina and manchette connect closely, which is important for both nuclear shaping and acrosome attachment. Combining the reduced cytoskeleton context at the equatorial area, we proposed that CFAP65 may function in the manchette (responsible for the nucleus shaping and acrosome elongation) and the equatorial zone (responsible for transverse acrosome extension, anchor and concentration). Moreover, the endogenous IP showed a presence of CFAP65 in the manchette-related complex containing MNS1 and the acrosome anchoring-related complex containing SPACA1 and ZPBP (19,26,27). Expression and localization of these proteins were changed during CFAP65 absence. These results also suggest that correct localization of CFAP65 is essential for the recruitment and/or transport of these proteins. Since CFAP65 is a protein with a transmembrane domain and has a strong expression at the equatorial zone, it possibly serves as a scaffold protein on the nuclear surface related to both the acrosome and manchette during spermiogenesis. Interestingly, several cytoskeleton-related proteins, such as CCIN and CAPZA3, were also downregulated in CFAP65-deficiency mice in our study (28,29). Whether the abnormalities in acrosome and head shape were caused by cytoskeletal dysregulation because of a CFAP65 defect will be an important next step toward understanding the role of CFAP65 during spermiogenesis.
In mice, the nuclei begin to have an asymmetrical shape and develop dorsal and ventral surfaces in step 9 (stage IX) (30). During this stage, over 50% of the dorsal surface becomes covered by the acrosome, whereas on the ventral surface of the spermatid, the acrosome covers only the apex of the head. Where there is no acrosome covering, the manchette forms (31). Therefore, it has been proposed that the nuclear membrane of spermatids might have recognition sites with unique features, either of a molecular or conformational nature, which would serve to distinguish the beginning of the manchette microtubule organization (31), which is important for the nuclear shaping and acrosome expansion. In CFAP65-deficient mice, the ectopic manchette was identified. We speculate that CFAP65 might play a role in the recognition sites, and a defect in CFAP65 leading to a delay in manchette organization triggers numerous other morphological abnormalities of spermatozoa.
The malformed MS in Cfap65−/− male mice is also corresponding to a short mid-piece previously observed in patients with CFAP65 mutations (8), indicating a failure of packaging mitochondria into the flagellar mid-piece around ODFs and the axoneme. However, through TEM of the testis, we can infer that the recruitment of mitochondria to the mid-piece could be normal, but the attachment to the flagellum may be cut off. In addition, the failed assembly of MS may trigger a quality-control response to delete these disrupted mitochondria as we see many mitochondria with abnormal morphology and contents. TPPP2 might be a working partner of CFAP65 during MS assembly. In Tppp2−/− mice, the sperm exhibited increased irregular mitochondria lacking lamellar cristae, abnormal expression of electron transfer chain molecules, low ATP levels and decreased mitochondrial membrane potential, leading to male subfertility with a significantly decreased sperm count and motility (20). However, we are still unaware of how CFAP65 functions in the mitochondria or MS and whether it acts as a structural component in the sperm mitochondria or as a scaffold protein responsible for mitochondrial anchoring.
Recent proteomic analysis of the CPC in Chlamydomonas has revealed numerous new candidate CPC proteins, wherein FAP65, the homolog of mammalian CFAP65, is predicted in the C2 complex (32). In our study, we found a localization of CFAP65 in the whole sperm flagella and the existence of RSPH1 in the endogenous CFAP65 interactome. RSPH1 is a RS-head protein, which has been found to be important for the correct building of CPC and RS proteins in cilia (33). It is known that motile cilia have a similar axoneme structure with sperm flagella (9 + 2), and we have also found phenotypic defects related to the motile cilia in Cfap65−/− mice (data not shown). Correspondingly, the CPC of sperm flagella in both Cfap65−/− mice and patients with CFAP65 mutations (8) were frequently absent. These findings strongly suggest a possible CPC structural role of CFAP65 in mature spermatozoa.
In summary, loss of CFAP65 leads to severe malformations in both sperm head and flagella, and the structural deficits and abnormally expressed proteins could account for the asthenoteratospermia phenotype in Cfap65−/− male mice. However, how exactly CFAP65 regulates nuclear envelope–acrosome–manchette synchrony and what is the role of CFAP65 in sperm mitochondria and the CPC are yet to be determined. In addition, although the intracytoplasmic sperm injection treatment has achieved great success in most asthenoteratospermia patients, individuals with CFAP65 mutations have shown unsatisfactory outcomes (8). Therefore, further investigations of CFAP65 could also elucidate the mechanisms that lead to inferior embryo development and address the requirements of patients with CFAP65 mutations.
Taken together, our results demonstrated the severe phenotypes caused by CFAP65 deletion during spermiogenesis and proposed multiple emerging roles for CFAP65 using a Cfap65 knockout mouse model. CFAP65 may specifically function in the sperm head shaping and acrosome formation by being involved in the manchette organization from the early steps of elongated spermatids and participate in the assembly of the sperm MS and axoneme in the late steps. Overall, these findings can provide insights into the molecular mechanisms involved in sperm head shaping, acrosome biogenesis and flagellar MS assembly.
Materials and Methods
Generation of Cfap65 knockout mice
Cfap65 knockout mice (C57BL/6 N) were generated by CRISPR/Cas9 genome editing as described in our previous study (34). Briefly, sgRNAs (sgRNA-1: AATTTTTGGTGCGCGACATTAGG, sgRNA-2: CTGGCCACAAGGATGACCCGTGG) in plasmids against exon 4 and 11 of Cfap65 (NM_001039495.1) were designed and constructed, respectively. Cas9 messenger RNA and sgRNAs were transcribed by T7 RNA polymerase in vitro, and then mixed and co-microinjected into the fertilized oocytes of C57BL/6 mice. Offspring were genotyped by PCR using tail genomic DNA via Ex Taq DNA polymerase (Bio-Rad, Hercules, CA, USA) with the specific primers listed in Supplementary Material, Table S3 and identified by Sanger sequencing. Tail genomic DNA was extracted using a TIANamp Genomic DNA Kit (TianGen Biotech, Beijing, China). Cfap65 expression in the testicular tissue from Cfap65+/+ (wild-type), Cfap65+/− (heterozygotes) and Cfap65−/− (homozygotes) mice was validated by reverse transcription (RT)-PCR followed by Sanger sequencing. All animal procedures were approved by the Institutional Animal Care and Use Committee of Central South University (Changsha, China) and carried out according to the standard protocols.
Phenotypic analysis of Cfap65 knockout mice
The different genotypes of 2-month-old mice were used for fertility testing. We mated Cfap65+/+, Cfap65+/− and Cfap65−/− male mice with Cfap65+/+ female mice and the coital plugs of the females were checked. The number of pups per litter was recorded, and the average litter size of each male mouse was measured. Sperm samples from different genotypes of male mice were collected from the cauda epididymis. Briefly, the cauda epididymis was dissected and sperm was squeezed out from the cauda epididymis and incubated for 30 min at 37°C under 5% CO2. Sperm counts were determined using a hemocytometer under a light microscope (Olympus). The mobility of sperm was assessed according to a previously described method (35). For sperm morphology analysis, the sperm were spread onto pre-coated slides, dried at room temperature, fixed in 4% paraformaldehyde and stained with hematoxylin and eosin (H&E).
Scanning electron microscopy (SEM)
Sperm samples from Cfap65+/+ and Cfap65−/− male mice were fixed in 2.5% phosphate-buffered glutaraldehyde at 4°C for 2 h and then deposited on poly-L-lysine-coated coverslips. The coverslips were washed in distilled water, dehydrated via an ascending gradient of 50, 70, 80, 9 and 100% cold ethanol, and dried at critical point using a Quorum K850 Critical Point Dryer (East Sussex, UK). Specimens were then attached to specimen holders and coated with gold particles using an ion sputter coater (Q150RS Rotary-Pumped, Quorum Technologies, East Sussex, UK) before being viewed with an S-3400 N SEM (Hitachi, Tokyo, Japan).
TEM
Sperm samples from Cfap65+/+ and Cfap65−/− male mice were treated as previously described (34). Briefly, samples were fixed with glutaraldehyde (Sigma-Aldrich, St. Louis, MO, USA) and osmium tetroxide, OsO4 and sucrose, and dehydrated using graded ethanol. Subsequently, the samples were embedded in Epon812, dodecenylsuccinic anhydride, methylnadic anhydride and dimethylaminomethyl phenol. Ultrathin (70–90 nm) sections were contrasted with uranyl acetate and lead citrate. An HT7700 Hitachi electron microscope (Hitachi) and a MegaView III digital camera (Munster, Germany) were used for image capturing.
Histological analysis
For histological analysis, testicular tissues from different genotypes of 2-month-old male mice were fixed in 4% paraformaldehyde or Bouin’s solution (Sigma-Aldrich), embedded in paraffin, sectioned, processed and used for subsequent experiments. For H&E staining, semen smears were dehydrated with graded ethanol, stained with H&E, dehydrated again with graded ethanol and processed with dimethylbenzene for 5 min twice. For PAS staining, the testes of 2-month-old Cfap65+/+ and Cfap65−/− male mice were fixed overnight in Bouin’s solution, embedded in paraffin, sectioned, and then stained with hematoxylin and PAS reagent to visualize the acrosome.
IF and immunohistochemistry (IHC) analyses
IF analysis was performed as previously described (36). The slides were incubated with primary antibodies for 2 h at 37°C. The details of all antibodies are listed (Supplementary Material, Table S4). The slides were then incubated with secondary antibodies for 1.5 h at 37°C. For evidence of the acrosome, the samples were treated with FITC-conjugated PNA and incubated for 1.5 h at 37°C. Finally, all slides were stained using 2-(4-amidinophenyl)-1H-indole-6-carboxamidine for 5 min. An Olympus IX51 fluorescence microscope (Olympus) and VideoTesT-FISH 2.0 software (VideoTesT Ltd., Petersburg, Russia) were utilized for photographing fluorescence signals. IHC analysis was performed using the Bouin’s fixed sections. For antigen retrieval, section was boiled for 15 min in sodium citrate buffer; 3% H2O2 was used to eliminate internal peroxidase activity. After blocking with 5% bull serum albumin (BSA) and incubating primary antibody at 4°C overnight, a conventional non-biotin system was used to detect the bound antibody according to manufacturer’s instructions (PV9000, ZSGB-BIO, Beijing, China). Finally, the sections were visualized by staining with 3,3′-diaminobenzidine, and the nuclei were counterstained with hematoxylin. Images were taken using an Olympus CX21 microscope (Olympus).
TUNEL assay
TUNEL assay was performed to detect apoptotic germ cells from Cfap65+/+ and Cfap65−/− male mice using the DeadEnd™ Colorimetric TUNEL System (Promega, Madison, WI, USA) according to the manufacturer’s instructions. Digital images were photographed using an Olympus IX51 fluorescence microscope (Olympus).
Protein extraction and quantitative proteomic analysis
Testicular tissue from 2-month-old Cfap65+/+ (n = 3) and Cfap65−/− (n = 3) male mice were used for proteomic analysis, which was carried out in the Jingjie PTM BioLab Co., Ltd. (Hangzhou, China). Briefly, testicular tissues were homogenized with Mammalian Protein Extraction Reagent (M-PER, Pierce Biotechnology, Rockford, IL, USA) and supplemented with a Halt Protease Inhibitor Cocktail. Then, 100 μg of protein from each sample were digested using trypsin, and the peptides were labeled using the 6-plex TMT Kit (90068, Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s protocol. The peptide mixtures were fractionated by high pH reverse-phase high performance liquid chromatography (HPLC) into 18 fractions using an Agilent 300 Extend C18 column. The peptide fractions were subsequently analyzed using a Q ExactiveTM hybrid quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific). The Mascot search engine (v.2.3.0) was used to process the mass spectrum results. Trypsin/p was designated as a cleavage enzyme, allowing for no more than two missing cleavages. The mass error was set at 0.02 Da for fragment ions and 10 ppm for precursor ions. The fixed and variable modifications were designated to be the carbamidomethylation of Cys and the oxidation of Met, respectively. TMT 6-plex in Mascot was used for protein quantification. The peptide ion score and P-value were set at R 20 and <0.05 respectively, whereas the false discovery rate was adjusted to <1%. Only proteins identified with at least two unique peptides were accepted. Proteins with quantitative ratios above 1.20 or below 0.83 were deemed to be significantly differentially expressed (P < 0.05).
Bioinformatics analysis of differentially expressed proteins
The GO annotation of the selected differentially expressed proteins was derived from the UniProt-GOA database. The KEGG database was used to annotate protein pathways. The Cytoscape software and the STRING database (version 10.0) were used to analyze the protein–protein interaction network. After comparing the database numbers or protein sequences of the differential proteins obtained in different comparison groups using the STRING (v.11.0) protein network interaction database, the differential protein interaction relationship was extracted according to confidence score >0.7 (high confidence). Then, R package ‘networkD3’ was used to visualize the differential protein interaction network.
Western blotting
Proteins were homogenized and blotted onto a polyvinylidene difluoride membrane, blocked with 5% skim milk for 2 h at room temperature, and then incubated overnight at 4°C with the antibodies listed in Supplementary Material, Table S4. The membranes were then incubated with secondary antibodies (goat anti-mouse IgG or goat anti-rabbit IgG, MultiSciences Biotech Co., Ltd., Hangzhou, China) for 1 h at room temperature, and the blots were revealed using the ECL Western blotting kit (Pierce Biotechnology) according to the manufacturer’s instructions.
Endogenous IP and immunoblots
Endogenous IP was performed using protein extracts from Cfap65+/+ mouse testes (postnatal day 60) lysed under the M-PER (Pierce Biotechnology) supplemented with protease inhibitor cocktail (Thermo Fisher Scientific) according to the manufacturer’s instructions. A customized CFAP65 antibody (Abclonal, China) specifically binding mouse amino acids 1401–1635 was incubated with protein A/G magnetic beads (Pierce Biotechnology) according to the manufacturer’s instructions. Subsequent immunoblotting for MNS1, RSPH1, TPPP2, ZPBP, SPACA1, DYNLT1 and SLIT2 were performed to detect the interactions. The antibodies used were listed in Supplementary Material, Table S4.
Statistical analysis
Statistical analysis was performed by Student’s t-test or one-way analysis of variance using IBM SPSS statistics version 19.0 and GraphPad PRISM version 5.01. Data are presented as the mean ± standard error of the mean. Differences were considered significant when the P-value <0.05 (*), 0.01 (**) or 0.001(***).
Acknowledgements
The authors would like to thank all families and individuals participated in this study. The authors thank the Center of Cryo-electron Microscopy at Zhejiang University for technical support. The authors also thank Dr Xiaoyin Wu and Dr Junpu Wang at the Electron Microscope Laboratory of Central South University for technical assistance.
Conflict of Interest statement. None declared.
Funding
The National Key Research & Developmental Program of China (2018YFC1004900 to Y.Q.T.); the National Natural Science Foundation of China (81771645, 81971447 to Y.Q.T.); the Key Grant of Prevention and Treatment of Birth Defect from Hunan Province (2019SK1012 to Y.Q.T.).
References
Author notes
Weili Wang, Shixong Tian contributed equally to this work.
