At the time of its discovery in 2005, full-length GIV protein was found to be most highly expressed in two organs: testis and brain (Figure 1A). Immunohistochemical studies curated by the Human Protein Atlas further confirm that GIV is most highly expressed in the testis (Figure 1—figure supplement 2A). Single-cell sequencing (Figure 1B; Figure 1—figure supplement 2B–E) and immunohistochemistry (IHC; Figure 1C) studies on human testis pinpoint sperm as the major cell type in the testis that expresses GIV mRNA and protein. We confirmed by confocal immunofluorescence on mouse testis that GIV is indeed expressed in the spermatozoa and localizes predominantly to the acrosomal cap, as determined by colocalization with the mouse acrosomal matrix protein, sp56 (Kim et al., 2001) (tGIV; Figure 1D). As expected, a tyrosine phosphorylated pool of GIV (pYGIV), however, localized mostly to the plasma membrane (PM) (Figure 1E). Both antibodies detected the endogenous GIV protein in testicular lysates at the expected size of ~220 kDa (Figure 1F). We also noted that GIV consistently and predominantly localizes to the acrosome as it matures from a rudimentary vesicle into a vesicular cap during sperm maturation (Figure 1G). GIV’s localization to the acrosome, which is derived from the Golgi apparatus (Khawar et al., 2019), is in keeping with GIV’s predominant localization to the Golgi and Golgi-associated transport vesicles in diverse cell types (Le-Niculescu et al., 2005; Lo et al., 2015). Taken together, we conclude that GIV is highly expressed in sperm and may be important for sperm functions.

Previously, in a publicly available patent (WO2017024311A1), the GIV gene (CCDC88A) was identified as one among a panel of genes whose altered expression due to DNA methylation may help diagnose male fertility status and/or the quality of the embryo (Carrell, 2016). We asked if the abundance of GIV transcripts in testis or sperm may be altered in infertile men. To this end, we curated all publicly available transcriptomic datasets from the NCBI GEO portal and analyzed them for differences in the abundance of CCDC88A transcripts across the annotated (in)fertility phenotypes (Figure 2A). CCDC88A transcripts were significantly and consistently downregulated in infertile men across all independent datasets analyzed (Figure 2B–E), regardless of whether the samples used for transcriptomic studies were testis or sperm. In Klinefelter’s syndrome (KS), the most common sex chromosomal disorder in humans that causes primary infertility, reduced CCDC88A expression was seen only after puberty and not in pre-pubertal boys (Figure 2E). This finding is in keeping with our observation that GIV is most prominently expressed in spermatocytes (Figure 1) and that spermatocytes are depleted in KS patients only at the onset of puberty (Wikström et al., 2004). Finally, in a study that segregated subfertile from fertile men using commonly used clinical parameters for semen quality, we found that sperm motility, but not concentration or morphology, was the key parameter (Figure 2F); when reduced sperm motility was used as a metric of infertility, semen from those subfertile men displayed reduced levels of GIV transcript. These results indicate that reduced GIV expression in testis and sperm is associated with clinically determined male infertility. Given the heterogeneous nature of the datasets (i.e., diagnosed cause of infertility, ranging from genetic syndromes with developmental or hormonal defects to post-chemotherapy to idiopathic), reduced GIV expression could be considered as a shared common molecular phenotype among infertile men.

Mature sperm, by virtue of being transcriptionally and translationally inactive, rely entirely upon rapid post-translational modifications to regulate all pre-zygotic processes. Because GIV is a multimodular signal transducer that straddles both tyrosine-based and G protein signaling pathways (Ghosh, 2016; Ghosh, 2015), we sought to investigate how GIV’s functions are altered during sperm capacitation. Because PI3K-Akt signals downstream of TKs is a critical pathway for actin remodeling in the sperm flagellum and for hypermotility (Tan and Thomas, 2014; Breitbart et al., 2005; Etkovitz et al., 2007; Roa-Espitia et al., 2016), and GIV serves as a point of convergence for multi-TK-dependent PI3K signaling (Lin et al., 2011; Midde et al., 2018), we first asked if GIV is indeed tyrosine phosphorylated in human and mouse sperm during capacitation. Using the sperm swim-up assay, we first confirmed that in human ejaculates low-motile sperm have just as much total GIV as their highly motile counterparts, but by contrast, tyrosine phosphorylated GIV was significantly elevated in the latter (compare tGIV and pYGIV, lanes 1–2 in immunoblots; Figure 3A). As a positive control, we simultaneously analyzed the same samples by dual-color immunoblotting with an antibody that detects pan-tyrosine phosphoproteins. As expected (Ecroyd et al., 2003; Ficarro et al., 2003; Arcelay et al., 2008; Yunes et al., 2003), the highly motile sperms have higher tyrosine phosphorylation (pan-pY; Figure 3A). pYGIV and pan-pY signals co-migrated in the SDS page gel, indicating that GIV is one of the tyrosine phosphorylated proteins in high-motile sperms. The pan-pY and pYGIV signals were found to further increase in capacitated sperms, maximally by 4 hr, without any change in total GIV (lanes 3–4; Figure 3A). Such phosphorylation was dependent on the activity of PKA because pretreatment of sperm with the PKA inhibitor H89 virtually abolished both pan pY and pYGIV (Figure 3B); these findings are in keeping with the fact that PKA activity is essential for tyrosine phosphorylation cascades during capacitation (Luconi et al., 2005; Lamirande and Gagnon, 2004). Immunofluorescence studies on human sperm confirmed that pan-pY and pYGIV signals colocalized in the mid-piece and tails of high-motile sperm (Figure 3C) where they were significantly induced upon capacitation (Figure 3D). Findings in human sperm were mirrored in murine sperm (Figure 3E,F), with some notable differences in temporal-spatial dynamics. For example, pY/pYGIV of murine sperms are induced more rapidly and transient. During murine sperm capacitation, pYGIV is induced in 30 min and then reduced in 120 min (Figure 3F) and was not as restricted to the sperm tail and mid-piece as in humans (compare sperm head regions in Figure 3D and F). Although full-length GIV (~250 kDa expected size) could be detected in murine sperm (Figure 3F), we often detected numerous breakdown products, presumably proteolytic in nature, in both murine and human sperm lysates (Figure 3A–B and F). Regardless of the size of the breakdown products, total tGIV, pYGIV, and pan-pY co-migrated in the gels at the same size, suggesting that GIV may be one of the major phosphotyrosine proteins in capacitating sperm. We conclude that GIV is a major phosphotyrosine substrate in sperm tail during capacitation and that its phosphoactivation requires upstream activation of PKA. Our findings suggest that this PKA→TK→pYGIV axis may enhance PI3K-Akt signals and sperm motility. Because the sperm Ca²⁺ channel, Catsper, exerts both spatial and temporal control over tyrosine phosphorylation as sperm acquire the capacity to fertilize Chung et al., 2014, and there is some evidence that H89 may directly inhibit Catsper (Wang et al., 2020), the contributions of a possible alternative Ca²⁺→TK→pYGIV pathway towards sperm motility cannot be ruled out.

Next, we asked how the G protein modulatory function of GIV is regulated during capacitation. The evolutionarily conserved C-terminal GEM motif in GIV that enables it to both activate Gi (Garcia-Marcos et al., 2009) and inhibit Gs (Gupta et al., 2016) is phosphoregulated by two Ser/Thr kinases, cyclin-dependent-like kinase 5 (CDK5) (López-Sánchez et al., 2013) and protein kinase C Ɵ (PKCƟ) (López-Sánchez et al., 2013) (summarized in Figure 4A). Phosphorylation at Ser(S)1,674 induces GIV’s ability to activate Gi by ~2.5- to 3.0-fold, whereas phosphorylation at S1689 inhibits GIV’s ability to activate Gi; neither phosphoevent impacts GIV’s ability to bind and inhibit Gs. By activating the inhibitory Gi and inhibiting the stimulatory Gs proteins, GIV overall inhibits mACs and suppresses production of cellular cAMP (Getz et al., 2019). Because post-translational protein modification is the predominant way mature sperm rapidly respond to environmental cues, we used two previously validated phosphosite-specific antibodies (López-Sánchez et al., 2013; Bhandari et al., 2015) that detect pS1674-GIV and pS1689-GIV. We found that in mouse (Figure 4B, left) and human (Figure 4C, left) sperm, HCO3^--induced capacitation induced the levels of phosphorylation at the activation site pS1674 in the sperm tails of both species, with two notable inter-species differences: (i) in murine sperm, the acrosomal cap showed phosphorylation at baseline with no further increase upon capacitation; and (ii) in human sperm, the mid-piece region showed phosphorylation at baseline with no further increase upon capacitation. Unlike the activating pS1674 site, distribution/intensity of phosphorylation at the inhibitory pS1689 site was observed at baseline in the head, mid-piece, and tail of the murine sperm (Figure 4B, right), and the mid-piece and tail in human sperm (Figure 4C, right) and did not change during capacitation.

We next repeated the studies with the sequential addition of HCO3^- (for 2 hr) followed by two other stimuli that are commonly used to trigger the AR, the calcium ionophore, A23186 (Tateno et al., 2013), and the reproductive hormone, progesterone (López-Torres and Chirinos, 2017). To monitor the phosphomodifications in GIV and their temporal relationship with acrosome exocytosis, we co-stained the sperm with peanut agglutinin (PNA) and Ser/Tyr-GIV. PNA binds specifically to galactose residues on the outer acrosomal membrane, and its disappearance is a widely accepted method of monitoring acrosome exocytosis (Mortimer et al., 1987; Kallajoki et al., 1986). pYGIV was induced predominantly in the mid-piece and tail during capacitation (cap 2 hr; Figure 4D) as seen before (Figure 3F) but also in the sperm head in the presence of A23187 and progesterone (Figure 4D). The localization of pYGIV in sperm head was seen only when the acrosomes were intact and lost in those where the acrosome was shed (compare AC-intact vs. -shed; Figure 4D). Similarly, phosphorylation at the activation site pS1674 was detected in sperm heads in the presence of A23187 and progesterone, but exclusively when the acrosomes remained intact (compare AC-intact vs. -shed; Figure 4E).

Taken together, the predominant findings can be summarized as follows (see legend of Figure 4F): GIV-GEM is inactive at baseline and activated upon capacitation. It remains active in both head and tail regions of capacitated sperm until the moment the acrosome is shed. Capacitation is also associated with robust tyrosine phosphorylation of GIV in the sperm tail and mid-piece throughout the process of acrosomal reaction (AR).

To determine if GIV is required for male fertility, we next co-housed female mice with conditional GIV knockout male mice (henceforth referred to as GIV-cKO; generated using tamoxifen in Ccdc88a^fl/fl-Ubc^Cre-Ert2 mice) or control littermates (WT; Ccdc88a^fl/fl mice) (see Materials and methods; see legend of Figure 5A) and analyzed diverse readouts. GIV knockdown was confirmed by genotyping tail tips (Figure 5B) and assessing GIV mRNA (Figure 5C) and protein (Figure 5D) in the testis. We noted a significant reduction of cumulative probability of pregnancy (100% vs. 55% rate for WT and KO groups, respectively, within 40 days after co-housing; Figure 5E) and average litter size (Figure 5F) in GIV-cKO mice. Surprisingly, both WT and GIV-cKO mice had similar sperm counts (Figure 5G), testes sizes, and weights (Figure 5—figure supplement 1). We confirmed by IHC that GIV was predominantly expressed in sperm in the testis of WT mice and that it was effectively depleted in GIV-cKO mice (Figure 5H). RNA-seq of the testis followed by unsupervised clustering showed that GIV-cKO testis differentially expressed only a handful of transcripts compared to WT testis (Figure 5I). The predominantly upregulated genes mapped to the ‘aberrant activation of PI3K/Akt signaling’ pathway (Figure 5J). This was largely attributable to Esr1 (highlighted in red; Figure 5I); polymorphisms of this gene are known to predispose to male fertility (Ge et al., 2014; Galan et al., 2005), and its induction represents a negative feedback event, resulting in the setting of inhibition of PI3K signaling (Bosch et al., 2015). The predominantly downregulated genes mapped to the IL12 pathway (Figure 5K), which is consistent with prior studies in men showing that IL12 may be important for male fertility and that its dysregulation may reflect infertility (Naz and Evans, 1998; Naz et al., 1998). Notably, both pathways reflect changes that are largely contributed by non-sperm cells in the testis; Esr1 is expressed exclusively in the Leydig cells in mouse testis (Zhou et al., 2002; Kotula-Balak et al., 2005) and IL12 is largely expressed by endothelial cells, peritubular cells, and macrophages (Terayama et al., 2014).

These findings demonstrate that GIV is required for male fertility and suggest that the role of GIV and its various phosphomodifications we observe in sperm is largely post-transcriptional and post-translational in nature.

Next, we assessed the role of GIV during sperm capacitation using a previously validated approach, that is, exogenous addition of cell-permeable His-tagged GIV-derived ~210 aa long peptides (Ma et al., 2015); these peptides either have an intact functional GEM motif (WT peptides) or, as negative control, a well-characterized F1685A (FA) mutant of the same motif ,which lacks such activity (Garcia-Marcos et al., 2009; Kalogriopoulos et al., 2019; Figure 6A; top). By anti-His staining followed by flow cytometry, we confirmed that TAT-His-GIV peptides were indeed taken up as we could detect uptake only when staining was conducted under permeabilized conditions (Figure 6A, bottom). Peptide uptake was efficient, varying within the range of ~80–90% (Figure 6A, bottom). Immunofluorescence studies confirmed that uptake was seen in all segments of the sperm (Figure 6B). The peptides were detected and functional (i.e., retained their ability to bind Gαi) at 1 and 6 hr post-uptake, as determined using lysates of peptide-transduced sperm as source of GIV in pulldown assays with recombinant GDP-loaded GST-tagged G protein, Gαi3 (Figure 6C).

Next, we analyzed phosphoproteins in TAT-GIV-transduced sperm undergoing in vitro capacitation by immunoblotting. Although PKA activation (Figure 6D) and pan-Y or pYGIV phosphorylation (Figure 6E) were relatively similar between WT and FA-transduced sperm, phosphorylation of Akt differed; TAT-GIV-WT induced phosphorylation of Akt much more robustly than TAT-GIV-FA (Figure 6D). This finding is consistent with the established role of GIV-GEM in the activation of the PI3K→Akt pathway via the activation of Gi and the release of ‘free’ Gβγ15. Because Akt phosphorylation has been implicated in sperm hypermotility and survival during capacitation (Quan and Liu, 2016; Pujianto et al., 2010), we performed computer-assisted sperm analysis (CASA) and MTT assays, respectively (Figure 6F). Consistent with the patterns of Akt phosphorylation, WT, but not FA peptide-transduced sperm showed greater overall motility as well as hypermotility (Figure 6G) and greater viability (Figure 6H,I).

These findings indicate that GIV’s GEM function may be dispensable for the PKA→TK→tyrosine phosphorylation pathway, but is required for Akt activation, sperm motility, and survival during capacitation (Figure 6J).

Prior studies have underscored the importance of mACs and their role in the regulation of cAMP and acrosome exocytosis in sperm (summarized in Figure 1—figure supplement 1A). mACs are localized most abundantly in the head (Figure 1—figure supplement 1A), and their activation by Gs or inhibition by Gi is known to finetune cAMP surge in that location, a function that is conserved in numerous species (Spehr et al., 2004). Thus, mACs and sACs regulate cAMP surges in the sperm head and tail, respectively, in a spatiotemporally segregated and independent manner (Figure 1—figure supplement 1B). Upon approaching the zona pellucida of an egg, a timely surge in cAMP in sperm head is required for the downstream activation of effectors PKA (Romarowski et al., 2015) and the exchange proteins directly activated by cAMP (EPAC) (Sosa et al., 2016; Mata-Martínez et al., 2021), which in turn coordinate the activation of several small GTPases (Branham et al., 2009; Pelletán et al., 2015; Bustos et al., 2015; Bustos et al., 2012) of the Ras superfamily. These GTPases enable rapid cytoskeletal remodeling and membrane trafficking events that culminate in acrosome exocytosis. As an activator of Gi and an inhibitor of Gs (Gupta et al., 2016) using the same conserved GEM motif (Figure 7A), GIV is known to tonically and robustly suppresses cAMP (Getz et al., 2019; Getz et al., 2020), and by that token, it is expected to inhibit the cAMP surge. Because GIV-GEM was activated upon capacitation and remained active until the acrosome was shed (Figure 4F), we hypothesized that GIV’s GEM function may be required for the prevention of a premature cAMP surge in the sperm head, and hence, premature acrosome exocytosis. We first confirmed that cAMP is modulated by a variety of stimuli targeting Gi- (adenosine) and Gs-coupled (progesterone) GPCRs (Figure 7—figure supplement 1A), consistent with what has been observed before (Wang et al., 2020; Parinaud and Milhet, 1996). When the same studies were carried out on TAT-GIV peptide-transduced sperm, the expected degree of cAMP induction were observed once again (Figure 7B), but TAT-GIV-WT, but not the GEM-deficient FA mutant peptides could significantly suppress the degree of cAMP surge across all stimuli tested (Figure 7C, Figure 7—figure supplement 1B).

Next we assessed the effect of GIV-GEM on acrosome exocytosis under the same conditions, that is, capacitation followed by AR, as we did in (Figure 4D,E), using a highly sensitive immunofluorescence-based assay that monitors the progressive exposure during AR of the inner acrosomal membrane protein, CD46 (Carver-Ward et al., 1997; Frolikova et al., 2016) (a.k.a. membrane cofactor protein [MCP]; Figure 7D). At 15 min after exposure to 10 µM A23186 or 100 µM progesterone, the TAT-GIV-WT, but not TAT-GIV-FA-transduced sperm had more intact acrosomes (Figure 7E) and fewer completely reacted acrosomes (Figure 7F). These results indicate that AR in response to both A23186 and progesterone was suppressed by TAT-GIV-WT, but not the GEM-deficient FA mutant. Instead, the FA mutant peptide had a higher proportion of sperm that completed AR.

These findings demonstrate that GIV is sufficient to inhibit cAMP surge and AR, and that these functions require a functional GEM module. Taken together with the temporal nature of the GIV-GEM activity (see Figure 4F), our findings also suggest that GIV-GEM may inhibit premature cAMP surge and acrosome shedding. Because these premature events may compromise fertilization only in the in vivo setting where sperm is required to remain in capacitated state while maintaining intact acrosomes for prolonged periods of time within the female reproductive tract before encountering the egg, we hypothesized that GIV’s function may be bypassed in the setting of in vitro fertilization (IVF; Figure 7G). We found this indeed to be the case because TAT-GIV-WT peptide-transduced sperm successfully fertilized the eggs in vitro to a similar extent as PBS control (Figure 7H,I). The GEM-deficient FA mutant-transduced sperm, which had higher surges in cAMP (Figure 7C) and a higher proportion of completely reacted acrosomes (Figure 7F), showed an approximately twofold increase in fertility.

Taken together, these findings indicate that GIV-GEM inhibits cAMP surge and AR to primarily prevent both events from occurring prematurely in vivo until in the presence of an egg for successful fertilization.

The major discovery we report here is a role of GPCR-independent (hence, non-canonical) G protein signaling in the sperm that is mediated by GIV/Girdin. Expressed most abundantly in the testis, and primarily in sperm, GIV is required for male fertility, and low GIV transcripts in men were invariably associated with infertility. We show that GIV is rapidly phosphomodulated on key tyrosine and serine residues in a manner segregated in space and time in various segments of the sperm (head, mid-piece, and tail) during capacitation and acrosomal reaction. These specific phosphomodifications, which are known to regulate GIV’s interactions with other key proteins (PI3K, Gαi/s proteins, etc.) and its functions as an effector of multiple TKs, as a cytoskeletal remodeler, and as a signal transducer, regulate key sperm phenotypes in at least two sperm compartments (summarized in Figure 8). First, in the sperm head, GIV’s GEM activity is induced upon capacitation. Once activated, GIV modulates both Gαi/s via the same GEM motif to suppress premature cAMP surges downstream of ligand-activated Gi/Gs-coupled GPCRs. Consequently, GIV-GEM inhibits premature acrosome shedding. Because both premature AR or failure to do so are important causes of male infertility (Liu et al., 2006), deciphering the signaling events that precisely regulate the timing of acrosome exocytosis has remained one of the most challenging and unresolved questions concerning mammalian reproductive biology (Buffone et al., 2014). Despite emerging evidence in the last decade that has challenged the long-held paradigms in the field, and mechanistic insights into sperm-extrinsic factors responsible for premature AR (Sánchez-Cárdenas et al., 2021; Balestrini et al., 2021; Harper et al., 2004), the identity of sperm-intrinsic pathways/processes/proteins that inhibit premature acrosome exocytosis was unknown. Our conclusion that GIV-GEM serves as a ‘brake’ for cAMP surge and prevents AR is consistent with the fact that the PDE-inhibitor sildenafil citrate (Viagra) increases cAMP to cause premature acrosomal reaction (Glenn et al., 2007). It is noteworthy that although canonical G protein signaling that is triggered by ligand-activated GPCRs has been implicated in the activation/inhibition of mACs and cAMP signaling in the sperm head (Adeoya-Osiguwa et al., 2006; Schaefer et al., 1998; Flegel et al., 2016), the role of non-canonical G protein we report here was never recognized previously. Because GIV is most highly expressed in sperm, the cAMP-regulatory role of GIV-GEM we define here implies that it may fulfill a major role in the regulation of cAMP in the sperm head. pYGIV was also detected in the sperm head, but its role in AR was not studied here. Because pYGIV activates class 1 PI3K, it is possible that the pYGIV→PI3K axis at that location could also influence rapid lipid phosphorylations that are also known to regulate AR (Cohen et al., 2016).

Second, in the sperm tail, GIV is an effector within the sAC→cAMP→PKA→multi-TK axis that gets robustly phosphorylated on Y1764; this site is known to directly bind and activate class 1 PI3Ks, which ultimately enhance Akt signals. GIV’s GEM activity is also activated in the tails of capacitated sperms and enhances Akt signals, presumably via the previously defined GIV→Gi→‘free’ Gβγ→class 1 PI3K axis. These two mechanisms of Akt signaling have previously been shown to act as an ‘AND’ gate to maximally enhance Akt signaling in diverse cell types to increase cell survival and motility (Lin et al., 2011; Lopez-Sanchez et al., 2015) Furthermore, both GIV transcripts and its phosphoactivation by TKs (pYGIV) were reduced in sperms with lower motility. Because the global trend of progressive reduction in the number of motile and viable sperm in the ejaculate has been associated with a concomitant increase in the rates of infertility (Dcunha et al., 2020), our findings in the case of GIV add to the growing number of proteins that enrich the signal-ome of healthy sperm. For example, as an intrinsically disordered protein (IDP) and a multi-modular scaffold that generates crosstalk between diverse signaling pathways, GIV appears to be in a prominent position to orchestrate rapid cooperativity between these pathways and processes in the otherwise transcriptionally and translationally silent sperm cell.

In conclusion, our results provide evidence that GIV may perform different roles in the distinct spatial compartments of capacitating sperms. This study not only sheds light on defective GIV-signaling as potential ‘marker’ of male infertility, but also reveals that inhibitors of GIV-dependent signaling will inhibit fertility by reducing sperm motility and viability and by promoting premature AR. The latter is a promising strategy for the development of a male contraceptive ‘pill’ specifically targeting sperm.

Human sperm were collected from volunteers via masturbation, liquefied at room temperature for 30 min, and subsequently washed with TYH media before being exposed to non-capacitating (NC) TYH media or capacitating conditions (C) TYH plus 5 mg/ml of BSA and 15 mm NaHCO₃ for 4 hr at 37°C and 5% CO₂, as described previously (Munné and Estop, 1993). The study proposal was approved by Institutional Review Board of University of California, San Diego (UCSD human subjects IRB protocol #16027, Gagneux). All samples were deidentified before use in studies. A written informed consent was obtained before participating in study. Consent to publish aggregate data with subject’s anonymity was obtained. The study design and the use of human study participants were conducted in accordance to the criteria set by the Declaration of Helsinki.

Ubc^Cre-Ert2/+ + Ccdc88a^fl/fl and Ccdc88a^fl/fl mice were generously provided by Dr. Masahide Takahashi (Nagoya University Graduate School of Medicine, Nagoya, Japan). Males Ubc^Cre-Ert2/+ × Ccdc88a^fl/fl were bred to females Ccdc88a^fl/fl to generate Ubc^Cre-Ert2/+ × Ccdc88a^fl/fl (experimental group), and Ubc^{Cre-Ert2/Cre-Ert2} × Ccdc88a^fl/fl (control group) mice. Genotyping was performed by PCR, and only male mice were used in this study. Wildtype female C57BL/6 mice were purchased from The Jackson Laboratory (stock number: 000664; Bar Harbor, ME). All mice were housed in standard cages in an Association for Accreditation and Assessment of Laboratory Animal Care-approved animal facility at the University of California San Diego School of Medicine. This study was approved by the UCSD Institutional Animal Care and Use Committee (protocol #S17223; Ghosh), which serves to ensure that all federal guidelines concerning animal experimentation are met.

For all biochemical, immunofluorescence, peptide transduction, and functional studies, the source of sperms was C57BL/6J mice, which were bred and housed under another protocol, which was also approved by the UCSD Institutional Animal Care and Use Committee (protocol #S16223; Gagneux).

All reagents were of analytical grade and obtained from Sigma-Aldrich (St. Louis, MO) unless otherwise stated.

The affinity-purified anti-pS1689-GIV and pS1674-GIV were generated commercially in collaboration with 21st Century Biochemicals (Marlboro, MA) and validated previously (López-Sánchez et al., 2013; Bhandari et al., 2015.) Rabbit anti-GIV CT (T-13) was obtained from Santa Cruz Biotechnology; and a previously validated custom-raised anti-p-GIV (pY1764) was from Spring Bioscience (Pleasanton, CA, USA) (Midde et al., 2015; Lin et al., 2011; Midde et al., 2018; Dunkel et al., 2016). Mouse mAbs against pTyr was from BD Biosciences; mouse anti-His, anti-a tubulin and anti-actin were obtained from Sigma; rabbit monoclonal anti-phospho-(p)Akt (Thr308) and anti-total (t)Akt were from Cell Signaling Technology (Beverly, MA). Rabbit anti-GIV-coiled-coil (CC) was obtained from EMD Millipore (Carlsbad, CA). Other commercially obtained antibodies used in this work were mouse anti-sp56 (MA1-10866) was purchased from Thermo Fisher Scientific (Waltham, MA), mouse anti-human hexokinase 1/2 monoclonal antibody (catalog # MAB8179; R&D Systems, Minneapolis, MN), and rabbit anti-phospho-PKA substrate (RRXS*/T*) (100G7E) mAb #9624 from Cell Signaling Technology.

Goat anti-rabbit and goat anti-mouse Alexa Fluor 680 or IRDye 800F (ab0)2 used for immunoblotting were from LI-COR Biosciences (Lincoln, NE). Goat anti-rabbit Alexa Fluor 488 and goat anti-mouse Alexa Fluor 594 for immunofluorescence were purchased from Life Technologies.

Mouse testes were fixed in zinc paraformaldehyde to prepare FFPE tissue blocks. Tissue sections of 4 μm thickness were cut and placed on glass slides coated with poly-L-lysine, followed by deparaffinization and hydration. Heat-induced epitope retrieval was performed using sodium citrate buffer (pH 6) in a pressure cooker. Tissue sections were incubated with 3% hydrogen peroxidase for 10 min to block endogenous peroxidase activity, followed by incubation with primary antibody overnight in a humidified chamber at 4°C. Antibodies used for immunostaining were SP173 rabbit monoclonal, anti-GIV antibody. Immunostaining was visualized with a labeled streptavidin–biotin using 3,3′-diaminobenzidine as a chromogen and counterstained with hematoxylin.

Mouse sperm suspension were obtained from cauda epididymis of mature male (9 weeks old) placed in 1 ml NC buffer prewarmed at 38.1°C for 25 min in siliconized/low-adhesion microfuge tubes. The caudae epididymis was cut to let the spermatozoa swim out. The microfuge tube was agitated on an orbital shaker for 10 min to facilitate the swim out of the sperm. The tubes were then placed upright on a bench top, and the epididymal tissues were allowed to settle for 10 min. The sperm suspension was then removed from the top, and an aliquot was taken to ensure purity and for counting (∼99%).

Human ejaculates were collected from volunteers via masturbation after 1 week of abstinence under UCSD human subject protocol (UCSD human subjects IRB protocol #16027, Gagneux). After liquefaction at room temperature, 1 ml ejaculate was transferred into the bottom of 1 ml prewarmed NC buffer and incubated for additional 1 hr to swim up procedure. Highly motile sperm mobilized to the upper layer was collected for the experiment.

Freshly obtained human and mouse sperms were segregated into low-motile and high-motile populations using ‘swim-up’ technique. Subsequently highly motile sperms were capacitated in TYH buffer containing 5 mg/ml BSA and 15 mM NaHCO₃ at 37°C under 5% CO₂ for the indicated time mentioned in figure legends. Sperms were both lysed in reducing sample buffer for immunoblotting and fixed in 3% paraformaldehyde for immunofluorescence staining. Acrosomal reaction in sperm was triggered by incubating capacitated sperm either with Ca2+ ionophore or progesterone for the indicated time at 37°C in 5% CO₂. Sperm were then fixed and co-stained for peanut agglutinin (PNA-488; green, an acrosomal marker) and either pYGIV, or pSerGIV and DAPI.

Sperms were fixed with 3% paraformaldehyde in PBS for 25 min at room temperature, treated with 0.1 M glycine for 10 min, and subsequently blocked/permeabilized with PBS containing 1% BSA and 0.1% Triton X-100 for 20 min at room temperature. Primary and secondary antibodies were incubated for 1 hr at room temperature in PBS containing 1% BSA and 0.1% Triton X-100. Dilutions of antibodies used were as follows: GIV (1:500); phospho-GIV (Tyr1764; 1:500); phospho-pan-Tyr (1:500); α-tubulin (1:500); phosphor-GIV (Ser1674; 1:500); phosphor-GIV (Ser1689; 1:500); Peanut agglutinin (PNA) (1:500); His (1:500) and DAPI (1:2000). Secondary Alexa conjugated antibodies were used at 1:500 dilutions.

In the case of frozen sections of mouse testes, the protocol used was as follows: cryosections were washed three times with PBS, followed by 0.15% glycine for 10 min at room temperature and incubated for 20 min in blocking buffer (1% BSA in PBS), then 2 hr in primary antibodies and 45 min in secondary antibodies. Dilutions of antibodies used were as follows: GIV (1:500); phospho-GIV (Tyr1764; 1:250); ZP3R (1:500); DAPI (1:1000). Secondary Alexa conjugated antibodies were used at 1:250 dilutions. Sperms and sections were imaged on a Leica SPE confocal microscope using a 63× oil objective using 488, 561, 633, and 405 laser lines for excitation. The settings were optimized, and the final images scanned with line-averaging of 3. All images were processed using ImageJ software (NIH) and assembled for presentation using Photoshop and Illustrator software (Adobe).

Protein samples were separated by SDS/PAGE and transferred to PVDF membranes (Millipore). Membranes were blocked with PBS supplemented with 5% nonfat milk (or with 5% BSA when probing for phosphorylated proteins) before incubation with primary antibodies. Infrared imaging with two-color detection and quantification were performed using a LI-COR Odyssey imaging system. Primary antibodies were diluted as follows: anti-His 1:1000; anti-GIV (tGIV) 1:500; anti-phospho-Tyr-1764-GIV (pYGIV) 1:500; anti-phospho-Tyr (pan pY) 1: 500; anti-phospho-PKA 1: 500; anti-hexokinase 1:1000; anti-phospho-Akt (Thr308) 1:500; anti-Akt 1:500; anti-β-tubulin 1:1000. All Odyssey images were processed using ImageJ software (NIH) and assembled for presentation using Photoshop and Illustrator software (Adobe).

Cloning of TAT-GIV-CT-WT and TAT-GIV-CT-FA mutant has been described (Ma et al., 2015). TAT-constructs were expressed using BL21(DE3)-pLysS (Invitrogen) and Terrific Broth (BioPioneer) supplemented with additives as per auto-induction protocols outlined by Studier, 2005. Briefly, cultures of bacteria were grown at 300 rpm at 37°C for 5 hr, then at 25°C overnight. Cells were lysed in 10 ml of lysis buffer containing 20 mM Tris, 10 mM imidazole, 400 mM NaCl, 1% (vol:vol) sarkosyl, 1% (vol:vol) Triton X-100, 2 mM DTT, 2 mM Na3oV4 and protease inhibitor mixture (Roche Diagnostics), pH 7.4, sonicated (3 × 30 s), cleared at 12,000× g for 20 min at 4°C and affinity-purified on Ni-NTA agarose resin (Qiagen) (4 hr at 4°C). Proteins were eluted in elution buffer containing 20 mM Tris, 300 mM imidazole, 400 mM NaCl, pH 7.4, dialyzed overnight against TBS containing 400 mM NaCl and stored at −80°C.

TAT transduction in sperms was performed by incubating them with 400–800 nM TAT-GIV-CT peptides for 30 min. Efficiency uptake was measured by flow cytometry. Different batches of TAT-GIV-CT protein preparations were used for optimization of equal uptake of recombinant TAT-GIV-CT WT and FA peptides. Optimization steps included timing of transduction, wash step, and concentrations of each peptide used to ensure that WT and FA peptides are equal in sperm. The recombinant protein that showed the most efficient uptake was subsequently used in four different mouse sperm samples to document consistent uptake test by western blotting, by FACS and immunofluorescence, and finally, to confirm that GIV peptides retain functionality (G protein binding) upon uptake.

Freshly obtained sperms were segregated into low-motile and high-motile populations using ‘swim-up’ technique and highly motile sperms were subsequently capacitated in TYH buffer containing 5 mg/ml BSA and 15 mM NaHCO₃ along with TAT-GIV-CT peptides at 37°C under 5% CO₂ for 3 hr. The sperm motility and progressive motility were measured on CASA on a Hamilton Thorne IVSO-CASA (Berns Laboratory, UC San Diego).

Mouse sperms at a density of 2 × 10⁷ cells/ml (6 × 10⁶ cells in total) were first peptide transduced with TAT-GIV-CT for 30 min, washed gently with PBS three times to remove excess peptides before their use in cAMP assays. Peptide-transduced sperms were pre-incubated with 0.5 mM isobutyl methyl xanthine (IBMX) prior to exposure to various chemicals at the following final concentrations: 25 mM NaHCO_3, 0.1 mM adenosine, or 100 µM. After mixing with the respective stimulus, the samples were incubated for 30 min at 37°C, followed by the addition of 0.25 M HCl (final concentration) to quench the biochemical reactions. After incubation for 30 min at room temperature, cell debris was sedimented by centrifugation at 3000 g for 5 min at room temperature. The cAMP concentration in the supernatant was determined by a competitive enzyme immunoassay according to the product manual (catalog # ADI-900-066, Enzo Life Sciences).

4–5-week-old mice experimental or control mice received an intraperitoneal (i.p.) tamoxifen injection 1 mg/100 µl/mouse/day (Millipore Sigma, St. Louis, MO) solubilized in 100% corn oil for five consecutive days. Ccdc88a gene knockout (GIV-knockout) was confirmed using DNA qPCR as described in our previous study (Swanson et al., 2020). Mice were then housed for 3 weeks with random females to promote mating with the intent to discharge sperms in which tamoxifen has not yet induced Cre expression.

After 3 weeks, each male mouse was housed with three 7–8-week-old C57BL/6 fertile females for 3 months. The frequency of successful live births and the litter size was recorded. After 3 months, experimental and control males were sacrificed, and testicles and epididymis were collected for IHC, immunoblotting, and mRNA analysis.

Publicly available microarray (GSE4797 Feig et al., 2007, E-TABM-234 Feig, 2008, GSE6872 Platts et al., 2007, GSE26881 Pacheco et al., 2011) and RNASeq dataset (GSE103905 Winge et al., 2018) were downloaded from the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus website (GEO) and ArrayExpress. The data was processed using the Hegemon data analysis framework repositories for male infertility (Sahoo, 2012; Sahoo et al., 2008; Sahoo et al., 2010). Microarray datasets (GSE4797; GSE6872; GSE26881; E-TMAB-234) were normalized using robust multi-array average (RMA). RNA-sequencing dataset (GSE103905) was normalized using transcripts per millions (TPM) normalization method; for downstream analyses, log2(TPM) if TPM >1 and (TPM – 1) if TPM <1 is used. Distribution of gene expression values is illustrated using boxplots and mean as circle with 95% confidence intervals as arrows. Numbers on top indicate the p values, which were derived from Welch’s t-test. All semen samples mentioned in the above datasets were classified based on WHO (2010) guidelines for semen parameters (Cooper et al., 2010).

Total RNA from mouse testicles were isolated using the Direct-zol RNA Miniprep Kit (Zymo Research, Irvine, CA) following the manufacturer’s instructions. RNA concentration was measured using the NanoDrop One (ThermoScientific, Waltham, MA). The RNA Integrity Number (RIN) was assessed using the 4200 TapeStation system and the TapeStation RNA ScreenTape & Reagents (Agilent Technologies, Santa Clara, CA).

Total RNA samples were submitted to the IGM Genomics Center (University of California San Diego) for library preparation and sequencing. mRNA stranded sequencing libraries were generated with the TruSeq Stranded mRNA Sample Prep Kit with TruSeq Unique Dual Indexes (Illumina, San Diego, CA). Resulting libraries were multiplexed and sequenced with 100 bp paired-end reads (PE100) to a depth of approximately 30 million reads per sample on an Illumina NovaSeq 6000. Samples were demultiplexed using bcl2fastq v2.20 Conversion Software (Illumina).

To determine which genes were differentially expressed in WT and GIV-knockout mice testes, transcript-level abundance of paired-end RNA-seq data was estimated by Salmon (1.1.0) using the mouse transcriptome from Genecode (vM24). Tximport (1.14.2) was used to aggregate transcript-level quantification to the gene level. The resulting gene counts were used as an input to DESeq2 Bioconductor package. Differentially expressed genes below a Benjamini-Hochberg (BH)-adjusted p-value of 0.05 were considered significant. Also, all genes differentially expressed were included in REACTOME pathway enrichment analysis. Statistically significant pathways of upregulated and downregulated DEGs are listed in the table and bar plots of upregulated and downregulated enriched pathways. In Gene Set Enrichment Analysis (GSEA) analysis, some fertility-related gene sets from Molecular Signatures Database (MSigDB) were tested. Genes from these gene sets were used to rank order the samples and test for GIV-knockout versus WT phenotype classification using the area under the curve (AUC) receiver operating characteristics (ROC) curve and displayed such classification using violin plots.

Statistical significance between datasets with three or more experimental groups was determined using one-way analysis of variance (ANOVA) followed by Tukey’s test for multiple comparisons. Unpaired t-test is used to test the statistical difference between two experimental groups. For all tests, a p-value>0.05 is considered as significant. All experiments were repeated at least three times. All statistical analyses were performed using GraphPad Prism 9.