CCDC189 affects sperm flagellum formation by interacting with CABCOCO1

ABSTRACT Multiple morphological abnormalities of the sperm flagella (MMAF) are one of the major causes of male infertility and are characterized by multiple defects. In this study, we found that the coiled-coil domain-containing 189 (Ccdc189) gene was predominantly expressed in mouse testes and that inactivation of the Ccdc189 gene caused male infertility. Histological studies revealed that most sperm from Ccdc189-deficient mice carried coiled, curved or short flagella, which are typical MMAF phenotypes. Immunoelectron microscopy showed that the CCDC189 protein was located at the radial spoke of the first peripheral microtubule doublet in the sperm axoneme. A CCDC189-interacting protein, CABCOCO1 (ciliary-associated calcium-binding coiled-coil protein 1), was discovered via co-immunoprecipitation and mass spectrometry, and inactivation of Cabcoco1 caused malformation of sperm flagella, which was consistent with findings obtained with Ccdc189-deficient mice. Further studies revealed that inactivation of CCDC189 caused downregulation of CABCOCO1 protein expression and that both CCDC189 and CABCOCO1 interacted with the radial-spoke-specific protein RSPH1 and intraflagellar transport proteins. This study demonstrated that Ccdc189 is a radial-spoke-associated protein and is involved in sperm flagellum formation through its interactions with CABCOCO1 and intraflagellar transport proteins.


INTRODUCTION
Infertility is a multifactorial pathological condition affecting ∼10%-15% of couples worldwide, and nearly half of these infertility cases involve male infertility [ 1 ]. Male infertility often manifests as decreased sperm count (oligozoospermia), reduced sperm count (asthenozoospermia), a high proportion of morphologically defective sperm (teratozoospermia) or a combination of these abnormalities [ 2 -4 ]. An estimated 30%-50% of male infertility is caused by genetic alterations. Therefore, considerable effort has been made to identify and characterize the genes required for male germ cell development [ 5 ].
Spermatogenesis is a complex developmental process that typically consists of three stages: mitosis of spermatogonia stem cells, meiosis of spermatocytes, and spermiogenesis. Spermiogenesis is the final stage of spermatogenesis, where round spermatids terminally differentiate into spermatozoa. This process includes acrosome and flagellum for-mation, nuclear condensation and cytoplasmic exclusion [ 6 , 7 ]. Impairment of any of these processes causes defects in sperm morphogenesis. The flagella of normal sperm are composed of nine peripheral microtubule doublets (MTDs) surrounding a pair of singlet microtubules called the central pair (CP). The peripheral doublets are connected to each other by nexin-dynein regulatory complexes (NDRCs) and to the central pair by radial spokes (RSs) [ 8 ].
Multiple morphological abnormalities of the sperm flagella (MMAF) constitute one of the most severe forms of sperm defects critical for male infertility and are characterized by numerous non-motile spermatozoa in ejaculated semen that exhibit flagellar abnormalities, including short, coiled, absent and irregular flagella [ 9 ]. MMAF is caused mainly by genetic mutations, and ∼40 MMAF-associated genes have been reported in the past 10 years. The main pathogenic gene families include the DNAH protein family [ 10 -16 ], coiled-coil domaincontaining (CCDC) protein family [ 17 ], cilia-and flagella-associated protein (CFAP) family and other protein-encoding genes associated with flagellum biogenesis and morphogenesis [ 5 , 9 , 18 -27 ]. Mutations or deletions of these genes usually cause MMAF in human patients and animal models. Although MMAF is not a rare autosomal-recessive inherited disorder in humans, the genetic etiology of a large proportion of MMAF cases remains elusive [ 2 , 28 ], and the recently identified MMAFassociated genes explain only 30%-60% of the MMAF cases in different cohorts [ 9 ]. Therefore, it remains necessary to improve the ability to distinguish single candidate genes in human MMAF samples, and this study may provide information to guide ICSI (intracytoplasmic sperm injection) treatment for MMAF patients.
CCDC proteins are involved in a wide range of physiological and pathological processes. An increasing number of CCDC proteins have been shown to be involved in flagellar morphogenesis. The genes in this family, such as Ccdc9, Ccdc11, Ccdc33, Ccdc39, Ccdc42, Ccdc47, Ccdc63 and Ccdc172, are also involved in spermatogenesis [ 17 , 29 -32 ]. Mutations in these genes cause defects in sperm flagellum biogenesis and male infertility. It has been reported that the CCDC189 protein is located in the flagellum of rat sperm [ 33 ], and mutation of the Ccdc189 gene is associated with male infertility in Nordic Red dairy cattle [ 34 ]. These results suggest that Ccdc189 is probably involved in the biogenesis of sperm flagellum, however, the physiological functions of this gene have not been investigated to date.
To investigate the function of Ccdc189 in sperm development, a Ccdc189 -knockout mouse model was generated in this study. We found that Ccdc189 −/ − males were infertile. The morphology of the sperm was abnormal as they presented with coiled, curved or short flagella. Further investigations revealed that the CCDC189 protein was located at the radial spoke of the first microtubule doublet and interacted with ciliary-associated calcium-binding coiled-coil protein 1 (CAB-COCO1). We also found that the knockout of Cabcoco1 caused defects in sperm flagellum formation, which was consistent with the findings with Ccdc189 −/ − mice. This study demonstrated that Ccdc189 is a radial-spoke-associated protein and plays important roles in sperm flagellum formation by interacting with CABCOCO1.

Ccdc189-knockout male mice are sterile
The expression of Ccdc189 in mice was measured by reverse transcription quantitative real-time PCR (RT −PCR) and western blotting. As shown in Sup-plementary Fig. 1A and B, the mRNA of Ccdc189 was predominantly expressed in testes, and it was first detected in testes in mice that were 3 weeks old. CCDC189 protein was detected only in testes, first in 3-week-old mice, and the expression level was higher in adult mice (Supplementary Fig. 1C and D). To investigate the physiological functions of Ccdc189 in mouse germ cell development, we generated a Ccdc189 −/ − mouse model using clustered regularly interspaced short palindromic repeats (CRISPR) −Cas9 technology. As shown in Supplementary Fig. 2A, a truncated mutation was generated by deleting exons from 4 to 8. Using genomic polymerase chain reaction (PCR), successful knockdown of Ccdc189 was confirmed (Supplementary Fig. 2B). The results of western blotting showed that the CCDC189 protein was completely absent in the testes of the Ccdc189 −/ − mice (Supplementary Fig. 2C), indicating that the Ccdc189 gene was inactivated in the Ccdc189 −/ − mice. Adult Ccdc189 −/ − mice were grossly normal, and no obvious developmental abnormalities were observed. The size of the testes from the Ccdc189 −/ − mice was significantly decreased compared to that in the control mice, and the testis/body weight ratio was also significantly reduced in the Ccdc189 −/ − mice ( Fig. 1 A, B and I-K). The results of a fertility test showed that Ccdc189 −/ − males were completely infertile ( Fig. 1 L).
To explore the reasons for sterility in the Ccdc189 −/ − mice, the histology of the testes and cauda epididymis in control and Ccdc189 −/ − mice was analyzed by hematoxylin-eosin (H&E) staining (  Fig. 3I-P). A large number of mature spermatozoa were observed in the epididymis of the control mice, whereas very few sperm heads were observed in the Ccdc189 −/ − mice ( Fig. 1 E and F). Immunofluorescence of the flagell a marker AQP3 protein showed fiber-like sperm tail structures in the seminiferous tubule lumen of the control mice. In contrast, only a punctate AQP3 signal was detected in the Ccdc189 −/ − mice ( Fig. 1 G and H). The results of computer-assisted sperm analysis (CASA) showed that the number and motility of the sperm from the Ccdc189 −/ − mice were dramatically reduced compared to those in control mice ( Fig. 1 M and N). These results indicated that in the early stage of germ cell development, Ccdc189 −/ − mice were not affected and that the Ccdc189 gene was most likely involved in sperm morphogenesis.

Ccdc189 is essential for sperm flagellar formation
We further analyzed the process of spermiogenesis in Ccdc189 -deficient mice by periodic-acid-Schiff (PAS) and hematoxylin staining. As shown in Supplementary Fig. 4A, stages 1-12 of the seminiferous epithelial cycle were observed in both the control and Ccdc189 −/ − mice, and no difference was noted between the control and Ccdc189 −/ − mice. Immunofluorescence staining for α-tubulin showed that the manchette in the Ccdc189 −/ − mice in Steps 11-12 was longer than that in the control mice. This implied that deletion of Ccdc189 caused abnormal manchette formation. The abnormal manchette structure may have exerted an effect on flagellum biogenesis. Therefore, we further examined the morphology of mature sperm from control and Ccdc189 −/ − mice. As shown in Fig. 2 A and B, sperm with normal morphology were observed in the control males. In contrast, most of the sperm in the Ccdc189 −/ − mice were malformed and presented with short, coiled flagella and abnormal heads, which was consistent with the sperm morphology observed in MMAF patients. The immunostaining results also showed that the morphology of α-tubulin-positive flagella was abnormal and that PNA-positive acrosomes were abnormal in the sperm from the Ccdc189 −/ − mice. We also examined the location of the IFT88 protein in control and Ccdc189 -deficient sperm. As shown in Fig. 2 C-F, IFT88 was located in the sperm flagell a of control sperm (C) and co-localized with ace-tubulin during sperm elongation. In Ccdc189 -deficient sperm (D-F), IFT88 protein was diffused in germ cells and not co-localized with ace-tubulin. To analyze the morphological defects in Ccdc189 −/ − sperm, we performed an ultrastructural analysis of control and mutant sperm using scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Crescent-shaped sperm heads with wellorganized mitochondrial sheaths were observed in the control mice via SEM (Fig. 2 G). Sperm with a round or missing head presented with a typical hook-shaped appearance ( To explore whether Ccdc189 −/ − mice present with symptoms in addition to infertility, we examined airway epithelial cells in the tracheas of Ccdc189 −/ − mice by H&E staining and immunohistochemistry ( Supplementary Fig. 5A and B). The number of ciliated cells and the cilial morphology were not significantly changed in the Ccdc189 −/ − mice compared to the control mice. Then, we further analyzed serum starvation induced primary cilial formation using mouse embryonic fibroblast cells (MEFs) from control and Ccdc189 −/ − mice. As shown in Supplementary Fig. 5C, cilia were stained with an anti-acetylated tubulin antibody, a marker of primary cilia. We found no obvious differences in the number of primary cilia in the Ccdc189 −/ − mouse fibroblast cells compared with the control cells ( Supplementary Fig. 5D). We subsequently determined the cilial length of the Ccdc189 −/ − mouse fibroblast cells and control cells, and no significant difference was detected in the length of Ccdc189 −/ − mouse cilia compared with the control mouse cilia (Supplementary Fig. 5E).

The CCDC189 protein localizes to the radial spoke of the first microtubule doublet of sperm flagella
To determine the subcellular location of the CCDC189 protein in sperm, we used the CRISPR/Cas9 system to generate an EGFP-tagged Ccdc189 knockin mouse model ( Supplementary   Fig. 6A). The mouse model was verified by PCR and western blot analysis (Supplementary Fig. 6B and C). To determine whether the functions of the CCDC189 protein are affected by fusion with the EGFP tag, we first assessed the fertility of homozygous Ccdc189 -EGFP mice and found that homozygous Ccdc189 -EGFP male mice were fertile ( Supplementary Fig. 6D-F). The process of spermatogenesis was normal, and a large number of mature sperm were observed in the epididymis of the homozygous Ccdc189 -EGFP mice, and a Ccdc189 -EGFP tagged allele could rescue the fertility defect of a Ccdc189 knockout allele in the Ccdc189 −/ EGFP mice, supporting the idea that EGFP-tagged protein is similar to wild-type Ccdc189 in function (data not shown). These results indicated that the EGFP tag did not affect the functions of the CCDC189 protein. The immunostaining results showed GFPpositive fiber-like sperm flagell a in Ccdc189-GFP mice, but not in control mice. GFP co-localized with AQP3, ODF2 and α-tubulin in testes, indicating that the Ccdc189 protein was mainly located in sperm flagell a (Fig. 3 A-H).
To further determine the subcellular location of CCDC189 in sperm flagell a, immunostaining was performed to detect GFP, AQP3, AKAP4, ODF2 and ace-tubulin in sperm, and high-resolution images were obtained (Fig. 3 I-P). The sperm tail membrane was labeled with AQP3 ( Fig. 3 K and L), and the fibrous sheath of sperm tails was labeled with AKAP4 ( Fig. 3 I and J). We found that the GFP signal was located inside the AQP3-positive tail membrane ( Fig. 3 K and L) and AKAP4-positive fibrous sheath (Fig. 3 I and J). We also used ODF2 to label the outer dense fiber (Fig. 3 M and N) and acetubulin to label the microtubule structure ( Fig. 3 O and P). The results showed that the scattered GFP signal co-localized with the ace-tubulin-positive microtubule structure (Fig. 3 O and P). These results indicated that Ccdc189 was localized to the vicinity of the microtubule structure inside the flagellum of a spermatozoon.
To further determine the subcellular location of the Ccdc189 protein at the ultrastructural level, immunoelectron microscopy was performed. As shown in Fig. 3 , immunogold-labeled particles were detected between the microtubules of the middle piece ( Fig. 3 Q), principal piece (Fig. 3 R and S) and tail end (Fig. 3 T), as prepared by longitudinal sectioning. The golden particles were all located on one side of the central axis, indicating the polarized location of CCDC189 in the axoneme of the flagellum . To determine whether CCDC189 displays a regular polarized localization pattern in the flagellum, we further examined the location of golden particles in cross sections of sperm tails. As shown in Fig. 3 U-X, the golden particles were located at the radial spoke of the first peripheral microtubule doublet. This result confirmed the polarized location of CCDC189 in sperm flagell a.

CCDC189 interacts with CABCOCO1 in the flagellum of spermatids
To further explore the molecular mechanism of CCDC189 in sperm flagellum formation, CCDC189-interacting proteins in Ccdc189 -EGFP mouse testes were identified by coimmunoprecipitation (co-IP) and mass spectrometry analyses, as shown in Fig. 4 A and Supplementary Table 3. A total of 672 proteins in the testes of Ccdc189 -EGFP mice were pulled down by an anti-GFP antibody, and 144 proteins were significantly enriched in Ccdc189 -EGFP mice compared to wildtype control mice. We found that CCDC189 was the most highly enriched protein, in accordance with its protein-matching score. The 10 most highly enriched candidate CCDC189-interacting proteins are listed in Fig. 4 A. The candidate proteins were verified via co-IP experiments. In vitro studies showed that the CABCOCO1 and CCDC189 proteins in HEK293 cells were mutual ly pul led down by overexpressed FLAG-tagged Cabcoco1 and EGFP-tagged Ccdc189 (Fig. 4 B, top panel). The CABCOCO1 protein in the testes of Ccdc189 -EGFP mice was also pulled down by an anti-GFP antibody (Fig. 4 B, bottom panel). These results indicated that CABCOCO1 interacted with the CCDC189 protein in testes. The location of CABCOCO1 in sperm flagell a w as ev aluated vi a immunofluores cence assay ( Fig. 4 C and D). High-resolution images showed that CCDC189 (green) was partially co-localized with CABCOCO1 (red) (Fig. 4 E and F).

Inactivation of the Cabcoco1 gene causes aberrant sperm flagellogenesis
To identify the physiological functions of the Cabcoco1 gene in spermatogenesis, we generated Cabcoco1 -knockout mice using CRISPR −Cas9 technology. A nul l al lele mouse model was generated by deleting all exons of the Cabcoco1 gene ( Supplementary Fig. 7A). The identity of the mouse model was verified by PCR ( Supplementary  Fig. 7B), western blotting ( Supplementary Fig. 7C) and immunofluorescence staining (Fig. 4 C and D). No obvious developmental abnormalities were observed in adult Cabcoco1 −/ − mice (Fig. 5 A; Supplementary Fig. 7D), and the testes size of the Cabcoco1 −/ − mice was comparable to that of the control mice (Fig. 5 B and C; Supplementary  Fig. 7E). However, the sperm count and motility (Fig. 5 D and E) were markedly reduced in the Cabcoco1 −/ − mice. No obvious histological defects were observed in the testes of the Cabcoco1 −/ − mice (Fig. 5 F and G). However, a large number of abnormal spermatozoa were observed in the epididymis of the Cabcoco1 −/ − mice (Fig. 5 H and  I). AQP3-positive sperm flagell a were observed in the epididymis and testes of the control mice (Fig. 5 J; Supplementary Fig. 7F). In contrast, very few fiber-like AQP3 signals were observed in the epididymis and testes of the Cabcoco1 −/ − mice, and most of the AQP3 signal was punctate-like, which was consistent with that in the Ccdc189 −/ − mice (Fig. 5 K; Supplementary Fig. 7G). The ODF2 and ace-tubulin signals were markedly reduced in the Cabcoco1 −/ − samples compared to the control samples ( Supplementary Fig. 7H-K). The morphology of sperm from the caudal epididymis was further examined by H&E staining and immunofluorescence assay (Fig. 5 L and M). A large proportion of abnormal sperm, including those with a missing tail, short tail or coiled tail, were observed in the Cabcoco1 −/ − mice, similar to the abnormalities observed in the Ccdc189 −/ − mice (Fig. 5 N). These results indicated that Cabcoco1 plays an important role in regulating the morphogenesis of sperm flagella.

The interaction between CABCOCO1 and CCDC189 is mediated by the CCDC189 coiled-coil domain
To test whether the protein level of CABCOCO1 is affected in Ccdc189 −/ − mice, the expression of CABCOCO1 was measured by western blotting. We found that the expression of CABCOCO1 was markedly reduced in Ccdc189 −/ − mice (Fig. 6 A and  B). We also found that the protein level of CAB-COCO1 was significantly reduced in the testes of Ccdc189 + / − mice (Fig. 6 C and D). These results indicated that CCDC189 is required for the integrity maintenance of the CABCOCO1 protein and the inactivation of CCDC189 caused degradation of CAB-COCO1. To further determine which domain is critical for the interaction between CCDC189 and CABCOCO1, a series of CCDC189 fragments were generated as indicated in Fig. 6 E, and co-IP assays were performed by overexpressing the truncated proteins in HEK293 cells. As shown in Fig. 6 F, we found that Ccdc189 D3 (with the coiled-coil domain deleted) could not be pulled down by CABCOCO1, suggesting that the coiled-coil domain of CCDC189 is required for the interaction between these two proteins.
We also ex amined w hether CCDC189 and CABCOCO1 interacted with other radial-spokeassociated proteins by performing co-IP experiments. As shown in Supplementary Fig. 8A and B, radial spoke head 1 homolog (RSPH1) was pulled down by both CABCOCO1 and CCDC189, indicating that CCDC189/CABCOCO1 complexes are involved in stabilizing the axoneme structure through their interaction with RSPH1. In addition to CABCOCO1, several intraflagellar transport (IFT)-related proteins, including IFT20 and IFT88, were also pulled down by CCDC189 as determined by IP/MS analysis. IFT20 and IFT88 are two major components of the intraflagellar transport machinery and form a module that is important for the flagellar assembly and maintenance [ 35 , 36 ]. To verify these results, a co-IP assay was performed. We found that the IFT88 protein in the testes of Ccdc189 -EGFP mice was pulled down by an anti-GFP antibody (Supplementary Fig. 8C). We also found that both CCDC189 and CABCOCO1 interacted with IFT20 ( Supplementary Fig. 8D). We also observed that IFT88 protein was diffused in germ cells and not co-localized with ace-tubulin in Ccdc189-deficient sperm (Fig. 2 C-F ). These results suggested that CCDC189 and CABCOCO1 are probably involved in intraflagellar transport during flagellum formation, as they interact with IFT-related proteins. Considering these results, we speculated that CCDC189 is probably required for intraflagellar transportation. When Ccdc189 is inactivated, intraflagellar transport is negatively affected and the flagellar axoneme cannot assemble properly.

DISCUSSION
Previous studies have demonstrated that CCDC family members play important roles in sperm flagellum formation, and mutations in these genes lead to MMAF in human patients and animal models [ 17 ]. Mutations in CCDC9 and CCDC151 are detected in patients with severe asthenozoospermia [ 37 -39 ]. Biallelic mutations of CCDC39 and CCDC40 have been found in patients with 'radial spoke defects' in primary ciliary dyskinesia. Knockout of Ccdc9, Ccdc38, Ccdc42, Ccdc62, Ccdc87 and Ccdc136 results in male infertility in mouse models because of defects in sperm flagell a [ 31 , 39 -43 ]. Ccdc189 also belongs to the CCDC family and has not been previously reported to be associated with MMAF in human patients [ 17 ]. In this study, we found that the inactivation of Ccdc189 in a mouse model caused male infertility with malformation of sperm flagella, indicating that the CCDC189 protein also plays a critical role in sperm flagellum formation.
Sperm motility is driven by motile cytoskeletal elements in the tail called axonemes. The axoneme is composed of 9 + 2 microtubules, where a central pair of microtubules is surrounded by nine peripheral MTDs in a specific order [ 44 ]. The MTD that is perpendicular to the central pair of microtubules is defined as the first MTD, and MTDs 2-9 are named in clockwise order [ 45 ]. Radial spokes are T-shaped structures that link peripheral microtubule doublets to the central pair of microtubules. The spoke stalk binds to the A-tubule of each MTD, and the spoke head faces towards the center of the axoneme [ 46 , 47 ]. Radial spokes are prominent structural features of 9 + 2 axonemes and are crucial for flagellar beating. Each spoke consists of a thin stalk and a bulbous head [ 48 ]. The spoke heads contain structural proteins that specialize in having contact with the central apparatus, while the stalk contains regulatory elements for modulating flagellar beating. Previous studies have found that mutations in RSPH1 and CFAP61 cause disintegration of the sperm flagell ar structure and primary ciliary dyskinesia [ 23 , 49 ]. In this study, we found that the CCDC189 protein was specifically located at the radial spoke. Most interestingly, the CCDC189 protein was located only at the radial spoke of the first MTD. The MTDs of sperm tail axonemes are defined from 1 to 9 in clockwise order. Whether these nine MTDs are functionally different is unknown. It has been reported that mutation of DNAH17 leads to the destabilization of MTDs 4-7 in both human patients and mouse models [ 50 ]. Frequent loss of MTD 7 has also been observed in Ttll9-deficient mice [ 51 ]. VDAC3-deficient mice are infertile with abnormal axonemes and the missing doublets correspond to the last four doublets [ 52 ]. In this study, we found that the CCDC189 protein was specifically located at the radial spoke of the first MTD by immunoelectron microscopy. Taken together, these results suggest that the MTDs of sperm axonemes are most likely functionally different and that the composition of different radial spokes is probably also different. However, the exact functions of different MTDs need to be further investigated.  CABCOCO1 contains a CLAMP domain in the center and a coiled-coil domain in the C-terminal region. A previous study found that CABCOCO1 is located at the centrosome of spermatocytes and round spermatids. In mature sperm, CABCOCO1 is located in the tail flagell a [ 53 ]. However, the functions of Cabcoco1 in sperm development have not been investigated previously. In this study, we found that CABCOCO1 interacted with CCDC189. Importantly, Cabcoco1 −/ − mice also exhibited abnormal sperm flagella, which was consistent with the findings in Ccdc189 −/ − mice. However, the phenotypes observed in Ccdc189 −/ − mice were much more severe than those in Cabcoco1 −/ − mice, suggesting that Ccdc189 probably plays a more important role in flagellum formation. We also found that the protein level of CABCOCO1 was significantly reduced in Ccdc189 −/ − mice, suggesting that CCDC189 is required for integrity maintenance of the CABCOCO1 protein. RSPH1 is a major component of the radial spoke, and inactivation of this gene results in disintegration of the sperm flagellar structure and male infertility [ 49 ]. In this study, we found that both CCDC189 and CABCOCO1 interacted with RSPH1. These results confirmed that CCDC189 and CABCOCO1 are located at the radial spoke.
Two transport mechanisms are involved in sperm flagello genesis: intramanchette transport (IMT) and IFT [ 54 -56 ]. IFT is critical for normal sperm flagello genesis, and IFT20 and IFT88 are two major components of the intraflagellar transport machinery [ 35 , 36 ].
In the present study, we found that IFT20 and IFT88 interacted with CCDC189 and that the IFT88 protein was scattered and not co-localized with ace-tubulin in Ccdc189-deficient sperm. These results suggest that CCDC189 may play an important role in flagellogenesis through its facilitation of intraflagellar transportation. However, the exact function of CCDC189 in sperm flagell ar morp hogenesis needs further investigation.
In summary, our study demonstrated that Ccdc189 is a radial-spoke-associated protein and plays an essential role in sperm flagella formation (Fig. 7 ). Ccdc189 -knockout mice displayed typical MMAF with aberrant sperm flagella. We discovered a new flagellar protein, CABCOCO1, which interacts with CCDC189 and RSPH1. Inactivation of Cabcoco1 also led to flawed sperm flagell a formation in mice. The similar symptoms imply that CCDC189 and CABCOCO1 are essential for male mouse spermatogenesis and fertility and may have similar functions in human spermatogenesis. Our research also revealed that different radial spokes in the MTDs of sperm axonemes are likely functionally diverse and composed differently. Our findings provide insights into the molecular cause of male infertility, and offer potential targets for developing new therapeutic strategies.