Cdc42 activity is essential for the interplay between cAMP/PKA pathway and CatSper function

Sperm acquire the ability to fertilize in a process called capacitation and undergo hyperactivation, a change in the motility pattern, which depends on Ca2+ transport by CatSper channels. CatSper is essential for fertilization and it is subjected to a complex regulation that is not fully understood. Here, we report that similar to CatSper, Cdc42 distribution in the principal piece is confined to four linear domains and this localization is disrupted in CatSper1-null sperm. Cdc42 inhibition impaired CatSper activity and other Ca2+-dependent downstream events resulting in a severe compromise of the sperm fertilizing potential. We also demonstrate that Cdc42 is essential for CatSper function by modulating cAMP production by sAC, providing a new regulatory mechanism for the stimulation of CatSper by the cAMP/PKA-dependent pathway. These results reveal a broad mechanistic insight into the regulation of Ca2+ in mammalian sperm, a matter of critical importance in male infertility as well as in contraception.


Results 134
Cdc42 localizes in four longitudinal lines along the flagellum resembling CatSper 135 distribution pattern. 136 The presence of Cdc42 protein in mouse sperm was analyzed by 137 immunoblotting using a specific antibody against this small GTPase. As shown in Fig.  138 1A, a single band of the expected size for Cdc42 (21.3 kDa) was observed in these 139 cells. Immunofluorescence studies revealed the localization of Cdc42 mainly along the 140 flagellum as well as in the acrosome of mouse sperm (Fig. 1B). 141 To investigate the spatial distribution of Cdc42, 3D STORM was used. This type 142 of microscopy allows 3D reconstruction with ~20 nm resolution (Huang et al., 2008;143 Rust et al., 2006), making it possible to study protein localization within the sperm 144 flagellum, which is less than 1 μm in diameter (Chung et al., 2014;Gervasi et al., 145 2018). The identification of specific structures within the flagellum was performed by 146 analyzing the localization of well-known flagellar proteins. For the axoneme, the fibrous 147 sheath and the plasma membrane, β-tubulin, A kinase anchoring protein 4 (AKAP4), 148 and the facilitative glucose transporter 3 (GLUT3) were used as previously described 149 (Chung et al., 2014) (Suppl. Fig. 1 and 2). Taking advantage of the cylindrical 150 symmetry of the flagellum, the molecule localizations were converted into cylindrical 151 coordinates (r, Ɵ, z') where z' represents the flagellum axis. In cartesian coordinates (x, 152 y, z), z is the direction normal to the coverslip, so that in this coordinate system the 153 sperm lies on the x,y plane as represented in Suppl. Fig. 1A. As expected, the spatial 154 distribution of β-tubulin was axonemal and localized in the center of the flagellum, while 155 AKAP4 (localized in the fibrous sheath) and GLUT3 (present throughout the flagellar 156 plasma membrane) showed a continuous rim distribution in cross-section (Suppl. Fig.  157 1B, right panel). The radial distributions of β-tubulin, AKAP4 and GLUT3 in the principal 158 piece showed that β-tubulin peaked at 23 ± 2 nm (mean ± standard error) from the 159 center of the flagellum while AKAP4 and GLUT3 peaked at 146 ± 1 nm and 269 ± 1 160 nm, respectively (Suppl. Fig. 1C). 161 Mouse sperm exhibited a Cdc42 localization confined to four columns along the 162 principal piece (Fig. 1C and Suppl. Video 1), which resembles the localization of 163 CatSper in the same region (Chung et al., 2014). On cross-sections of the flagellum, 164 Cdc42 appeared as four tight puncta ( Fig. 1C right panel). Similar to Cdc42, P-CaMKII 165 showed the CatSper-like distribution in the principal piece  3C-D), as previously described (Chung et al., 2014). The radial distributions of Cdc42 167 and P-CaMKII peaked at 238 ± 16 nm and 235 ± 2 nm respectively ( Fig. 1D and F). 168 These values are similar to those observed for GLUT3 (Suppl. Fig. 1B-C), indicating 169 that both proteins localized close to the plasma membrane. A graphical representation of Cdc42 localization in terms of the azimuth angle Ɵ vs. axial distance z' along the Ca 2+ brought in from CatSper channels is essential for hyperactivation and 245 successful fertilization (Ren et al., 2001). Given the decrease in [Ca 2+ ]i observed as a 246 result of Cdc42 inhibition, hyperactivated motility and in vitro fertilization was assessed 247 in presence of MLS-57151. For this purpose, motility patterns were analyzed by 248 computer-assisted semen analysis (CASA) in those samples exposed to Cdc42 249 inhibitor or control capacitating medium. Cdc42 inhibition by MLS-573151 produced a 250 significant decrease in the percentage of hyperactivated cells when compared with 251 controls (Fig. 4E). This decrease was also observed in the kinetic parameters 252 evaluated, in particular, in those that are used to identify the hyperactivated population 253 such as VCL, LIN and ALH (Suppl. To determine whether Cdc42 activity is required for CatSper function, we used 265 a fluorescence method as previously described (Ernesto et al., 2015). This assay is 266 based on the fact that removing external Ca 2+ and Mg 2+ after adding EGTA allows 267 CatSper to efficiently conduct monovalent cations (Kirichok et al., 2006), where a 268 sudden influx of Na + depolarizes the cells (Espinosa and Darszon, 1995;Torres-Flores 269 et al., 2011). The magnitude of this depolarization mainly depends on the extent of 270 CatSper opening, and is inhibited by CatSper channels blockers (Ernesto et al., 2015;271 Torres-Flores et al., 2011). Sperm membrane potential (Em) was measured with a 272 fluorescent cyanine dye DiSC3(5). Because this cationic dye accumulates on cells with 273 hyperpolarized membranes, an increase in DiSC3 (5)  CatSper channels are strongly activated by an intracellular alkalinization 280 (Kirichok et al., 2006;Lishko et al., 2010). To evaluate if Cdc42 alters the sperm intracellular pH (pHi), this parameter was analyzed in live cells by flow cytometry, 282 loading sperm with the pHi sensitive probe BCECF AM. Cells were incubated for 90 283 min under non-capacitating and capacitating conditions in the presence or absence of 284 Cdc42 inhibitor. The intracellular alkalinization-associated with capacitation was not 285 modified by the presence of MLS-573151 (Fig. 5C). 286 To better characterize the Cdc42-dependent CatSper gating, whole-cell voltage-287 clamp recordings of CatSper currents (ICatSper) were performed. ICatSper from cauda 288 epididymal sperm were analyzed using a typical voltage-ramp protocol (from -80 to +80 289 mV during 750 ms with a holding potential of 0 mV) and Na + out/Cs + in as the main 290 conducting ions in the absence of divalent cations (divalent cation free condition, DVF). 5F-G). All together, these results suggest that Cdc42 activity is necessary for CatSper 297 function, by a mechanism different from pHi regulation. 298 There is evidence that CatSper channels are promiscuous and may be inhibited 299 by different compounds in a non-specific manner (Barratt and Publicover, 2012;300 Brenker et al., 2012). To validate our results, we explored the specificity of the Cdc42 301 inhibitor used in previous experiments. To this aim, we used a condition where Cdc42 302 activity is compromised (absence of GTP) and took advantage of the fact that mouse 303 CatSper is strongly activated by intracellular alkalinization (promoted by NH4Cl 304 addition), as demonstrated by an increase in ICatSper (Kirichok et al., 2006). The 305 mechanism for pH-dependent activation is not yet fully understood but it has been 306 reported that the protein EFCAB9 together with the highly rich histidine CatSper amino 307 terminus domain are responsible for sensing changes in pHi (Hwang et al., 2019;Ren 308 et al., 2001). Our experiment is based on the premise that if CatSper is non-specifically 309 blocked by the action of Cdc42 inhibitor, alkalinization would not produce a robust 310 increase in CatSper activity. 311 In order to analyze whether Cdc42 modulation on CatSper disturbs pHi 312 activation, non-capacitated sperm loaded with the Em sensitive probe DiSC3(5) were 313 first exposed to Cdc42 inhibitor and then to NH4Cl. CatSper opening was evidenced by 314 the magnitude of Em depolarization resulting from Ca 2+ influx. MLS-573151 pre-315 incubation did not alter CatSper sensitivity to pH alkalinization ( Fig. 6A-B). Em depolarization under these experimental conditions. These results support that this 323 inhibitor is not producing a non-specific blockade of CatSper channels and may be 324 safely used to investigate the role of Cdc42 in mouse sperm. 325 In addition to MLS-573151, other two specific pharmacological inhibitors were 326 also used to study Cdc42. Secramine A blocks membrane recruitment of Cdc42 by 327 impeding the release of GDI from Cdc42 (Cdc42/GDI complex), which sequesters the 328 inactive GDP-bound Cdc42 in the cytosol and inhibits Cdc42 activation (GTP loading) 329 (Pelish et al., 2006). CASIN targets Cdc42 by specifically blocking GEF activity on the 330 Cdc42/GDI complex (referred to in (Peterson et al., 2006) as pirl1-related compound 331 2). On one hand, Secramine A partially impairs alkalinization-dependent CatSper 332 activation, observed by a decrease in Em depolarization after NH4Cl addition in 333 comparison with the control condition (without Cdc42 inhibitor) (Suppl. Fig. 5A-B). To investigate if Cdc42 inhibition alters the capacitation-induced PKA activation, 349 PKA-dependent phosphorylation was evaluated in the presence or absence of Cdc42 350 inhibitor. Incubation with MLS-573151 led to a decrease in the phosphorylation of PKA 351 substrates (pPKAs) and tyrosine residues (pY) (Fig. 7A). These data indicate that 352 active Cdc42 is crucial for proper PKA activity. To analyze whether the requirement of 353 Cdc42 is upstream PKA activation, sperm were incubated with Cdc42 inhibitor in the 354 presence or absence of a membrane permeable analog of cAMP (8-Br-cAMP) and a phosphodiesterase inhibitor (IBMX). As a result, addition of 8-Br-cAMP/IBMX bypassed 356 the inhibition of pPKAs and pY caused by 20 µM MLS-573151 ( Fig. 7B) suggesting that 357 Cdc42 activity is required upstream PKA activation. Furthermore, the decreased in 358 [Ca 2+ ]i provoked by MLS-573151 was also bypassed by 8-Br-cAMP/IBMX ( Fig. 7C and  359 Suppl. Fig. 6A). 360 Since our results indicate that Cdc42 is acting upstream PKA activity, we 361 analyzed whether active Cdc42 is necessary for cAMP production by sAC. Sperm 362 cAMP levels were measured in the presence or absence of 20 µM MLS-573151 and as 363 a result, a significant decreased in cAMP concentration was observed after incubation 364 with the Cdc42 inhibitor (Fig. 7D). All together, these results suggest that Cdc42 activity 365 is required for cAMP production by sAC, and consequently for the upregulation of 366

CatSper through PKA. 367
We also confirmed that the other Cdc42 inhibitors, that were discarded in 368 previous experiments (Suppl. Fig. 5), display non-specific inhibition on CatSper 369

function. Both Secramine A and CASIN provoke a decrease in phosphorylation 370
patterns of PKAs and pY (Suppl. Fig. 6B) that could be restored by 8-Br-cAMP/IBMX 371 (Suppl. Fig. 6C), but contrary to MLS-573151, sustained [Ca 2+ ]i inhibition was not 372 recovered even with the addition of membrane permeable analog of cAMP and IBMX 373 (Suppl. Fig. 6D-G). This reinforces that these drugs are not suited to study Cdc42 374 function in mouse sperm. 375 Finally, to further support our approach, the CatSper-dependent Ca 2+ increase 376 was investigated in the presence of PKA and CatSper inhibitors, where the addition of 377 8-Br-cAMP/IBMX should not be able to restore Ca 2+ levels. In one case, because 378 CatSper is stimulated downstream PKA activation and hence, direct inhibition of the 379 kinase is not bypassed by cAMP analogs. In the other case, the direct blockade of the 380 channel with CatSper inhibitors also impedes its cAMP-dependent activation. In 381 agreement with our interpretation of the results presented above, the decrease in the 382 In this report, we obtain similar results in terms of CatSper function and PKA activity. 428 However, we are proposing a new regulatory mechanism that connects CatSper function and the cAMP/PKA pathway. We found that Cdc42 is important for cAMP 430 production by sAC, which in turn results in activation of PKA. We hypothesize that the 431 localization of Cdc42 in the CatSper signaling complex is essential for the tight 432 interplay between the increase in the [Ca 2+ ]i and the cAMP/PKA pathway. Although a 433 cross talk between cAMP/PKA-dependent pathways and Ca 2+ clearly plays a key role 434 in sperm capacitation, the connection between these signaling events is incompletely 435 understood and our results contribute to unravel this complex regulatory mechanism. 436 Intracellular alkalinization has been univocally demonstrated to alter CatSper 437 function endogenously (Kirichok et al., 2006). In this regard, it was recently reported 438 the asymmetrically distribution of Hv1 channel in the human sperm flagellum, which is linked since a connection between Cdc42 and H + efflux through NHE1 has been 452 reported in somatic cells. A bistable positive feedback regulation between Cdc42 and 453 NHE1 activities has been previously proposed (Frantz et al., 2007). In summary, in this study we present novel findings indicating that Cdc42 plays 464 a central role in mammalian sperm capacitation. Figure 8 summarizes the results of 465 this work in a proposed model. In the female reproductive tract (or when sperm are exposed to conditions that support capacitation), HCO3stimulation of sAC resulting in 467 cAMP production requires the activity of Cdc42. This protein is localized in the 468 signaling complex organized by CatSper together with other proteins (not shown in this 469 simplified model). The increase in cAMP levels promotes the activation of PKA which in 470 turn, phosphorylates a subset of proteins that lead to phosphorylation in tyrosine 471 residues of others. In addition, the stimulation of the cAMP/PKA pathway leads to the 472 activation of CatSper (either directly or through other intermediates). As a result, a 473 sustained Ca 2+ influx promotes the development of hyperactivation. Considering that 474 CatSper is sperm specific and has a fundamental role in hyperactivation, this channel 475 has emerged as a potential therapeutic target in male infertility as well as in the 476 development of contraceptive strategies. Thus, understanding the modulation of 477 CatSper activity and all the players involved in its regulation is of great importance. 478 479

Sperm capacitation 534
Animals were euthanized and cauda epididymal mouse sperm were collected. 535 Both cauda epididymis were placed in 1 ml of non-capacitating modified TYH medium 536 (NC: without BSA and NaHCO3) in the presence or absence of Ca 2+ as described for 537 each experiment. After 15 min of incubation at 37°C (swim-out), epididymis were 538 removed, and sperm were resuspended to a final maximum concentration of 1x10 7 539 cells/ml on 100 µl of the appropriate medium, depending on the experiment performed.

Three-Dimensional Stochastic Optical Reconstruction Microscopy (3D STORM) 550
After swim-out, cells were washed twice with NC medium by centrifugation (5 551 min at 400 x g) and finally resuspended in NC medium. Sperm were seeded in poly-L-552 lysine coated coverslips (Corning #1.5), air-dried for 10 min, fixed with 4% fresh 553 paraformaldehyde in PBS for 10 min at room temperature, and followed by three 554 washes with PBS (5 min at 400 x g). Cells were then permeabilized with 0.5% Triton X-555 100 in PBS for 5 min at room temperature and washed three times with PBS (5 min at 556 400 x g). Samples were blocked with 3% BSA/PBS for 1 h at room temperature and 557 then incubated overnight at 4°C with primary antibody in a humidified chamber: anti-558 inhibitor (20 µM MLS-573151) at a concentration between 0.65 and 0.95 mg protein/ml 616 were added into each well. Plain lysis buffer and a standard of constitutively active 617 purified GTP-bound Cdc42 protein were added to duplicate wells as a blank and a 618 positive control respectively. After binding, anti-Cdc42 primary antibody was added to 619 each well followed by secondary antibody labeled with HRP, which was developed by 620 adding HRP reagent. Each well was read at OD 492 nm on a 96-well plate 621 spectrophotometer. Lysis buffer background was subtracted and results were 622 normalized to protein concentration. 623

624
Membrane potential assay in cell populations 625 The procedure was done as previously described (Demarco et al., 2003). 626 Sperm obtained from swim-out were diluted in NC or CAP medium and loaded with 1 627 μM of the membrane-potential-sensitive dye DiSC3(5) for 5 min. Mitochondrial 628 membrane potential was dissipated incubating sperm for 2 min with 500 nM CCCP. 629 After this, 1 ml of the sperm suspension was transferred to a stirred cuvette at 37°C 630 and fluorescence monitored with an Ocean Optics USB4000 spectrofluorometer 631 On one hand, CatSper activity was assessed measuring Em in TYH medium 643 after adding EGTA (3.5 mM final, pH adjusted with NaOH to ~10 so that media pH 644 does not change upon H + release in exchange for Ca 2+ ), which allows Na + influx 645 through CatSper, depolarizing non-capacitated sperm. The magnitude of the 646 depolarization caused by Na + influx correlates with the extent of CatSper opening. 647 Briefly, after sperm treatment (capacitation in presence or absence of Cdc42 648 inhibitors), cells were collected by centrifugation (400 x g, 3 min) and concentration 649 adjusted to 2x10 6 sperm/ml with NC medium with the addition of the corresponding 650 inhibitor or not. Then, sperm were loaded with 1 μM of DiSC3(5). Results were represented as ΔFluorescence, the difference between DiSC3(5) fluorescence after 3.5 652 mM EGTA addition and before (resting Em) (Ernesto et al., 2015). In these 653 experiments, calibration curves were performed to confirm that sperm were viable. Em 654 values were not calculated from these calibrations. 655 On the other hand, non-capacitated sperm were loaded with 1 μM of DiSC3(5) 656 and then exposed to 10 mM NH4Cl, with prior addition of Cdc42 inhibitor or not. 657 Alkalinization allows CatSper activation, which results in Ca 2+ influx depolarizing the 658 cell. Again, the magnitude of the depolarization correlates with the extent of CatSper 659 opening. Results were represented as ΔEm, the difference between Em after 10 mM 660 NH4Cl addition and before (Em with or without Cdc42 inhibitor). were evoked employing a conventional voltage-ramp protocol from -80 mV to +80 mV, with duration of 750 ms from a holding potential of 0 mV. In all cases, the addition of 689 Cdc42 inhibitors was recorded until reaching a stable-state effect (3-5 min). 690 691

Extraction of sperm proteins and western blotting 692
After incubation in the appropriate medium, sperm were washed by 693 centrifugation (5 min at 400 x g), resuspended in sample buffer without reducing agents 694 (62.5 mM Tris-HCl pH 6.8, 2% SDS, 10% glycerol) and boiled for 5 min. After 695 centrifugation for 5 min at 13,000 x g, 5% β-mercaptoethanol and 0.0005% 696 bromophenol blue was added to the supernatants then boiled again for 5 min. Protein antibodies was sufficient. Antibodies were diluted in 2% nonfat dry milk in T-PBS as 707 follows: 1:500 for anti-Cdc42 (Santa Cruz sc-8401); 1:750 for anti-Cdc42 (Santa Cruz 708 sc-87); 1:3000 for anti-pPKAs, anti-pY and anti-β-tubulin T4026. The corresponding 709 secondary antibodies were incubated for 1 h at room temperature, diluted in 2% nonfat 710 dry milk in T-PBS as follows: 1:3000 for HRP anti-rabbit and 1:3000 for HRP anti-711 mouse. In all cases the reactive bands were visualized using a chemiluminescence 712 detection solution consisting of 100 mM Tris-HCl buffer pH 8, 205 μM coumaric acid, 713 1.3 mM luminol, 0.01% H2O2 and were exposed for different time periods to CL-714 XPosure film (Thermo Scientific). In all experiments, molecular masses were expressed 715 in kiloDaltons (kDa). ImageJ 1.48k (National Institute of Health, USA) was used for 716 analysis of the western blot images following the specifications of ImageJ User Guide, 717 IJ 1.46r. The optical density of all bands were quantified and normalized, first to the β-718 tubulin or Actin band and then to the capacitating condition of each experiment. 719 720 Immunofluorescence 721 To perform immunofluorescence on mouse sperm, a previously described 722 method was used (Gervasi et al., 2018). Briefly, mouse sperm were washed twice by 723 centrifugation for 5 min at 400 x g, resuspended in PBS containing 4% room temperature. After washing twice for 5 min in PBS sperm were placed onto glass 726 slides. Sperm were air-dried and then permeabilized with 0.5% Triton X-100 in PBS for 727 5 min. The slides were washed twice for 5 min in T-PBS and blocked in 3% BSA in 728 PBS for 1 h at room temperature, then were incubated with primary antibody (