The Distal Appendage Protein CEP164 Is Essential for Efferent Duct Multiciliogenesis and Male Fertility

Cilia are evolutionarily conserved microtubule-based structures that perform diverse biological functions. Cilia are assembled on basal bodies and anchored to the plasma membrane via distal appendages. Multiciliated cells (MCCs) are a specialized cell type with hundreds of motile multicilia, lining the brain ventricles, airways, and reproductive tracts to propel fluids/substances across the epithelial surface. In the male reproductive tract, MCCs in efferent ducts (EDs) move in a whip-like motion to stir the luminal contents and prevent sperm agglutination. Previously, we demonstrated that the essential distal appendage protein CEP164 recruits Chibby1 (Cby1), a small coiled-coil-containing protein, to basal bodies to facilitate basal body docking and ciliogenesis. Mice lacking CEP164 in MCCs (FoxJ1-Cre;CEP164fl/fl) show a significant loss of multicilia in the trachea, oviduct, and ependyma. In addition, we observed male sterility, however, the precise role of CEP164 in male fertility remained unknown. Here, we report that the seminiferous tubules and rete testis of FoxJ1-Cre;CEP164fl/fl mice exhibit substantial dilation, indicative of dysfunctional multicilia in the EDs. Consistent with these findings, multicilia were hardly detectable in the EDs of FoxJ1-Cre;CEP164fl/fl mice although FoxJ1-positive immature cells were present. Sperm aggregation and agglutination were commonly noticeable in the lumen of the seminiferous tubules and EDs of FoxJ1-Cre;CEP164fl/fl mice. In FoxJ1-Cre;CEP164fl/fl mice, the apical localization of Cby1 and the transition zone marker NPHP1 was severely diminished, suggesting basal body docking defects. TEM analysis of EDs further confirmed basal body accumulation in the cytoplasm of MCCs. Collectively, we conclude that deletion of CEP164 in the MCCs of EDs causes basal body docking defects and loss of multicilia, leading to sperm agglutination, obstruction of EDs, and male infertility. Our study therefore unravels an essential role of the distal appendage protein CEP164 in male fertility. Author Summary Multicilia are tinny hair-like microtubule-based structures that beat in a whip-like pattern to generate a fluid flow on the apical cell surface. Multiciliated cells are essential for the proper function of major organs such as brain, airway, and reproductive tracts. In the male reproductive system, multiciliated cells are present in the efferent ducts, which are small tubules that connect the testis to the epididymis. However, the importance of multiciliated cells in male fertility remains poorly understood. Here, we investigated the role of the critical ciliary protein CEP164 in male fertility using a mouse model lacking CEP164 in multiciliated cells. Male mice are infertile with reduced sperm counts. We demonstrate that, in the absence of CEP164, multiciliated cells are present in the efferent ducts but fail to extend multicilia due to basal body docking defects. Consistent with this, the recruitment of key ciliary proteins is perturbed. As a result, these mice show sperm agglutination, obstruction of sperm transport, and degeneration of germ cells in the testis, leading to infertility. Our study therefore reveals essential roles of CEP164 in the formation of multicilia in the efferent ducts and male fertility.


93
Spermatogenesis is a highly complex and finely tuned process in the testis that begins 94 with a diploid stem cell and ends with four haploid germ cells after two successive 95 8 airway MCC differentiation. More recently, we reported that mice lacking CEP164 in 162 MCCs (FoxJ1-Cre;CEP164 fl/fl ) show a significant loss of multicilia in the trachea, 163 oviduct, and ependyma, and about 20% die due to profound hydrocephalus. In tracheal 164 MCCs, the basal body recruitment of Cby1 and ciBAR proteins was severely diminished 165 in the absence of CEP164 [17]. Furthermore, we found that FoxJ1-Cre;CEP164 fl/fl males 166 are completely infertile. However, the precise role of CEP164 in the male reproductive 167 system has remained unexplored. 168

169
Here, we report the characterization of male infertility in FoxJ1-Cre;CEP164 fl/fl mice. We 170 found that, in these mice, the seminiferous tubules and rete testis show substantial 171 dilation, reminiscent of dysfunctional multicilia in EDs. In addition, sperm accumulation 172 and agglutination were commonly observed in the seminiferous tubules and EDs, We have previously reported that FoxJ1-Cre-mediated deletion of CEP164 leads to fully 186 penetrant male sterility [17]. To begin to understand the function of CEP164 in male 187 fertility, we first examined testis weights and sperm counts. We found no significant 188 difference in testis weights between control CEP164 fl/fl and FoxJ1-Cre;CEP164 fl/fl mice 189 ( Fig 1B). However, there was a dramatic decrease in the number of sperm from the 190 cauda epididymis in FoxJ1-Cre;CEP164 fl/fl mice (Fig 1C), yet these sperm were motile. 191 Interestingly, some of the sperm from FoxJ1-Cre;CEP164 fl/fl mice displayed 192 morphological defects such as detached heads and gaps between the midpiece and the 193 principal piece (Fig 1D). Histological analyses of testes from FoxJ1-Cre;CEP164 fl/fl mice revealed marked luminal 198 dilation of seminiferous tubules with thinning of the epithelium (Fig 2A). Elongated 199 flagella were visible in the lumen of seminiferous tubules in FoxJ1-Cre;CEP164 fl/fl mice 200 (bottom panels). In the control CEP164 fl/fl cauda epididymis, tubules were filled with 201 spermatozoa (Fig 2B). In stark contrast, many cauda epididymal tubules in FoxJ1-202 Cre;CEP164 fl/fl mice contained only few sperm with abnormal fluid secretions (i and iii). 203 Commonly, we also observed a few tubules containing agglutinated spermatozoa 204 (asterisks and ii). 205

206
The thinning of the seminiferous epithelium is a hallmark of dysfunctional multicilia in 207 the efferent ducts (EDs) [33]. It is typically associated with dilation of the rete testis, 208 which connects seminiferous tubules to EDs (Fig 1A). Histological assessment of the 209 rete testis revealed substantial dilation in FoxJ1-Cre;CEP164 fl/fl mice compared to that 210 of control CEP164 fl/fl mice (Fig 3A, arrows). Intriguingly, clogged seminiferous tubules 211 were frequently visible near the rete testis in intact testes from FoxJ1-Cre;CEP164 fl/fl 212 mice (Fig 3B, arrowheads). Cross sections of the clogged seminiferous tubules 213 confirmed extensive sperm agglutination (Fig 3C, arrowheads). Sperm agglutination 214 was detected in all 6 mice that we examined. Taken together, FoxJ1-Cre;CEP164 fl/fl 215 mice exhibit significant dilation of the seminiferous tubules and rete testis and sperm 216 agglutination, most likely due to defective multicilia in EDs. and sperm are concentrated as they progress from the rete testis into the caput 226 epididymis ( Fig 1A) [5,9]. Multiciliated cells (MCCs) are present in the ED epithelium 227 and have been shown to play a critical role in preventing the agglutination of sperm by 228 stirring the luminal fluids [11]. In control CEP164 fl/fl mice, CEP164 localized to the apical 229 surface of the ED epithelium ( Fig 4A). No such apical CEP164 signals were detectable 230 in FoxJ1-Cre;CEP164 fl/fl EDs, confirming efficient Cre-mediated recombination and 231 antibody specificity. We also consistently noticed dilation of EDs to varying degrees. 232 Intriguingly, costaining with the lectin peanut agglutinin (PNA), which binds to the outer 233 acrosomal membrane of sperm [37], showed robust accumulation of sperm in the EDs 234 of FoxJ1-Cre;CEP164 fl/fl mice, although no sperm were typically present in those of 235 control CEP164 fl/fl mice due to rapid transport of sperm through EDs (Figs 4A and S1). Our data so far point to the possibility that defective multicilia in the EDs of FoxJ1-250 Cre;CEP164 fl/fl mice lead to sperm agglutination, profound dilation of the rete testis and 251 seminiferous tubules, and ultimately infertility. To directly test this, we performed 252 histological analyses of EDs. As expected, the EDs of control CEP164 fl/fl mice showed 253 abundant multicilia with a length of 10 to 15 m, extending into the lumen ( Fig 5A). In 254 contrast, multicilia were hardly detectable in the EDs of FoxJ1-Cre;CEP164 fl/fl mice. We 255 also frequently observed the agglutination of sperm in the ED lumem of FoxJ1- apical enrichment in secretory cells (Fig. 6C, arrow). These cells were clearly present in 273 the EDs of FoxJ1-Cre;CEP164 fl/fl mice, and, as predicted, the A-tub-positive structures 274 coincided with AQP-1-positive secretory cells (Fig 6C, arrowheads). 275 Previously, we demonstrated that CEP164 physically interacts with and recruits Cby1 to 277 the distal appendages of nascent basal bodies to facilitate the efficient formation of 278 ciliary vesicles during the early stages of airway MCC differentiation, and both remain at 279 the base of mature cilia [16,17,43]. We observed Cby1 at the apical surface of MCCs, 280 basal to A-tub staining, in control CEP164 fl/fl EDs. However, apical Cby1 staining was 281 lost in FoxJ1-Cre;CEP164 fl/fl EDs (Fig 6D). Similarly, the apical localization of NPHP1, a 282 maker for the ciliary transition zone that acts as a diffusion barrier at the ciliary base [44, 283 45], was also altered in FoxJ1-Cre;CEP164 fl/fl EDs ( Fig S3). These results are 284 consistent with basal body docking defects. 285 286

CEP164 is critical for basal body docking in multiciliated cells in efferent ducts 287
CEP164 plays pivotal roles in recruitment of small preciliary vesicles to distal 288 appendages to assemble ciliary vesicles, thereby promoting basal body docking [16][17][18]. 289 In airway MCCs, loss of CEP164 is associated with defects in basal body docking and 290 subsequent cilium elongation [17]. To investigate if this is also the case for MCCs in the 291 EDs of FoxJ1-Cre;CEP164 fl/fl mice, we performed transmission electron microscopy 292 (TEM) analyses. In control CEP164 fl/fl EDs, we observed basal bodies docked to the 293 apical cell surface with multicilia extending into the lumen (Fig 7, arrowheads). In FoxJ1-294 Cre;CEP164 fl/fl EDs, however, many basal bodies were found undocked deep into the 295 cytoplasm (Fig 7, arrowheads). These results underscore the key role of CEP164 in 296 basal body docking.

299
MCCs are found in a restricted set of tissues, such as the airway, oviduct, spinal cord, 300 and brain ventricles and play crucial roles in fluid propulsion across the epithelial surface. 301 A growing body of evidence suggests that MCCs in EDs play a major role in male fertility 302 [11,33,34]. It has recently been shown that these multicilia act as agitators that generate 303 strong turbulence to maintain immotile spermatozoa in suspension in the ED lumen [11]. 304 Upon loss of these multicilia, aggregation of sperm leads to ED occlusions, thereby 305 preventing the transport of sperm into the epididymis. 306

307
We found that loss of CEP164 in MCCs leads to male infertility due to a nearly complete 308 ablation of multicilia in the ED epithelium. Consequently, FoxJ1-Cre;CEP164 fl/fl mice 309 consistently develop sperm agglutinations in the seminiferous tubules, EDs, and cauda 310 epididymis (Figs 2, 3, 5, and S1). In particular, large stretches of seminiferous tubules 311 near the rete testis were visibly clogged ( Fig 2B). Ultimately, this results in dilation of the 312 rete testis and seminiferous tubules, followed by testicular degeneration. Thus, it indicates 313 that dysfunctional multicilia in EDs leads to sperm agglutinations and obstruction of sperm 314 passage in different segments of the male reproductive tract. 315

316
The impact of CEP164 deletion on MCC differentiation in EDs is prominent as multicilia 317 are barely detectable (Figs 5-7). We demonstrated that the recruitment of basal body 318 proteins to the apical surface is abolished, and many basal bodies are undocked in the 319 EDs of FoxJ1-Cre;CEP164 fl/fl mice. Our findings therefore confirm the key role of CEP164 320 in basal body docking during MCC differentiation in vivo. 321 A recent single-cell RNA sequencing analysis suggests that CEP164 is ubiquitously 323 expressed throughout mouse spermatogenesis (Fig S2) [40]. Our IF staining 324 demonstrated that CEP164 intensely localizes to the base of elongating flagella in 325 spermatids (Fig 4B), suggesting that CEP164 is involved in flagellogenesis. This is in 326 contrast to the fly CEP164 homolog (CG9170), which is not expressed in the testis 327 (Flybase). A recent study has shown that CEP164 also localizes to the capitulum and 328 striated columns of the human sperm neck [46], although its functional significance is 329 unknown. Thus, it is likely that CEP164 plays multiple roles during spermatogenesis. 330

331
Although CEP164 is still present at the base of sperm flagella in FoxJ1-Cre;CEP164 fl/fl 332 mice at comparative levels to control (Fig 4B), some spermatozoa collected from the 333 cauda epididymis show morphological defects such as detached heads and a gap 334 between the midpiece and principal piece of the sperm tail ( Fig 1D). One possible 335 explanation is that fluid retention and back pressure, caused by ED obstructions, may 336 contribute to the sperm morphological defects. Alternatively, it could be attributable to 337 changes in the luminal microenvironment of the epididymis as testicular sperm undergo 338 complex maturation processes as they travel through the epididymis [47][48][49].

Immunofluorescence staining 382
For IF staining of PFA-fixed paraffin embedded tissues, tissue sections were 383 deparaffinized and rehydrated, and then antigen retrieval was performed using sodium 384 citrate buffer (10 mM sodium citrate, 0.05% Tween 20, pH 6.0) in an autoclave for 17.5 385 minutes. The specimens were then subjected to extraction using 0.5% Triton X-100 in 386 PBS for 5 minutes, followed by blocking for 1 hour at room temperature using 5% goat 387 serum in antibody dilution solution [2% bovine serum albumin (BSA) in PBS]. Primary 388 antibodies were applied overnight at 4°C, and then secondary antibodies were applied 389 for 1 hour at room temperature in antibody dilution solution. Finally, DAPI was applied 390 for 3 minutes at room temperature, and the specimens were mounted with Fluoromount-391 G (Southern Biotech). For primary antibodies used, see supplemental table S1. The were snap-frozen using 2-methylbutane and liquid nitrogen, and 10-m thick sections 399 were cut. The tissue was allowed to air dry for 30 minutes and then fixed with 4% PFA 400 at room temperature, followed by acetone at -20°C for 10 minutes each. Subsequent 401 steps were performed as described above. 402 403

Generation and purification of NPHP1 antibody 404
A rabbit anti-NPHP1 antibody was generated at Open Biosystems, Inc against His-405 tagged human NPHP1 N-terminal region (aa1-153) containing coiled-coil and SH3 406 domains. The antigen was expressed and purified from E. coli. The antibody was 407 purified by antigen affinity chromatography using a