| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Medicine, Oregon Health & Sciences University and Veterans Affairs Medical Center, Portland, Oregon 97239
ABSTRACT
In somatic cells, RHOA mediates actin dynamics through a GNA13-mediated signaling cascade involving RHO kinase (ROCK), LIM kinase (LIMK), and cofilin. RHOA can be negatively regulated by protein kinase A (PRKA), and it interacts with members of the A-kinase anchoring (AKAP) family via intermediary proteins. In spermatozoa, actin polymerization precedes the acrosome reaction, which is necessary for normal fertility. The present study was undertaken to determine whether the GNA13-mediated RHOA signaling pathway may be involved in acrosome reaction in bovine caudal sperm, and whether AKAPs may be involved in its targeting and regulation. GNA13, RHOA, ROCK2, LIMK2, and cofilin were all detected by Western blot in bovine caudal sperm. Overlay, immunoprecipitation, and subsequent mass spectrometry analysis identified several RHOA-interacting proteins, including proacrosin, angiotensin-converting enzyme, tubulin, aldolase C, and AKAP4. Using overlay and pulldown techniques, we demonstrate that phosphorylation of AKAP3 increases its interaction with the RHOA-interacting proteins PRKAR2 (the type II regulatory subunit of PRKA, formerly RII) and ropporin (ROPN1, a PRKAR2-like protein, or R2D2). Varying calcium concentrations in pulldown assays did not significantly alter binding to R2D2 proteins. These data suggest that the actin-regulating GNA13-mediated RHOA-ROCK-LIMK-cofilin pathway is present in bovine spermatozoa, that RHOA interacts with proteins involved in capacitation and the acrosome reaction, and that RHOA signaling in sperm may be targeted by AKAPs. Finally, AKAP3 binding to PRKAR2 and ROPN1 is regulated by phosphorylation in vitro.
AKAP, calcium, gamete biology, kinases, PKA, RHO, signal transduction, sperm
The RHO signaling pathway is involved in stimulating actin polymerization, a process that regulates many cellular functions, including cell division, motility, and polarity [1]. The RHO family of signaling molecules is a group of small GTP-binding proteins within the Ras-related small GTPase superfamily. RHO GTPases are present in all eukaryotic cells, where they alternate between inactive GDP-bound and active GTP-bound states. Guanine nucleotide exchange factors (GEFs) catalyze the GDP for GTP exchange [2]. Activation is negatively regulated by both guanine nucleotide dissociation inhibitors (RHO GDIs) and GTPase-activating proteins (GAPs) [1, 2]. Endogenous RHO can be inactivated via C3 exoenzyme ADP-ribosylation, and studies have demonstrated RHO involvement in actin-based cytoskeletal response to extracellular signals, including lysophosphatidic acid (LPA) [2–4]. LPA is known to signal through G-protein-coupled receptors (GPCRs) [4, 5]; specifically, LPA-activated GNA13 (formerly G
13) promotes RHO activation through GEFs [4, 6]. Activated RHO-GTP then signals RHO kinase (ROCK), resulting in the phosphorylation and activation of LIM-kinase (LIMK), which in turn phosphorylates and inactivates cofilin, an actin depolymerizer, the end result being actin polymerization [7–9] (See Supplemental Fig. 1 for diagram available online at www.biolreprod.org).
|
A-kinase anchoring proteins (AKAPs) are defined by their ability to bind one or more of the regulatory subunits (PRKAR1A, PRKAR2A, PRKAR1B, and PRKAR2B, formerly RI, RII, RIβ, and RIIβ) of cAMP-dependent protein kinase A (PRKA); these subunits interact with an amphipathic helix on the AKAP. AKAPs target the action of PRKA by acting as scaffolding proteins, spatially restricting function by simultaneously binding several related signal transduction enzymes [10–13]. Activation of PRKA usually results from the binding of cAMP to the R subunits of PRKA, which promotes dissociation and activation of the catalytic subunits, leading to a wide variety of cellular responses. However, AKAP3 can simultaneously bind PRKA and activated GNA13, causing the dissociation of the catalytic subunits of PRKA, thus activating PRKA in a cAMP-independent manner [14]. PRKA activation has been shown to negatively regulate RHOA signaling in two ways: PRKA directly phosphorylates RHOA, leading to increased interaction with RHO-GDI and translocation from the membrane to the cytosol [15–17], and PRKA phosphorylates GNA13, substantially reducing RHO activation (an effect that was blocked by the introduction of a GNA13 mutant incapable of PRKA-mediated phosphorylation) [18]. Ht31, an anchoring inhibitor peptide (AIP) that competitively disrupts PRKA binding to AKAPs, blocks PRKA phosphorylation of RHO and prevents PRKA-induced inhibition of RHOA in human SGC-7901 cells [19].
A yeast two-hybrid screen using AKAP3 as bait identified two other proteins that interact with the amphipathic helix in a manner similar to PRKAR2: AKAP-associated sperm protein (ROPN1L, formerly ASP, a novel protein) and ropporin (ROPN1), a protein involved in the RHO pathway via interaction with rhophilin [20, 21]. Due to the sequence similarity of ROPN1L and ROPN1 to the N-terminal dimerization and docking domain of RII, these proteins have been named R2D2 proteins [22]. Sequence analysis identified two other proteins that contain this R2D2 domain: sperm protein 17 (SPA17) and calcium-binding tyrosine phosphorylation-regulated protein (CABYR, formerly FSII) [23, 24]. All four of these proteins were originally identified as sperm-specific proteins [22].
In bovine spermatozoa, redistribution of actin-regulating proteins and increased actin polymerization takes place during capacitation and is suggested to have a role in the acrosome reaction (AR) [25–27]. Brener et al. [26] have shown that C3 exoenzyme inhibition of RHO blocks capacitation-induced actin polymerization and subsequent acrosome reaction. The authors suggest that cofilin inactivation via RHO signaling may be involved in this process. However, GNA13, the G protein that mediates cofilin signaling through RHOA, has not been previously demonstrated in sperm. RHOA and ROCK1 are known to be present in sperm, and have been localized to both the head and the tail of bovine spermatozoa [21, 28]. LIMK2 and cofilin are present in testes, but they have not yet been specifically identified in spermatozoa [29, 30]. It has been suggested that cofilin and LIMK signaling requires tight spatial regulation [1, 8]. As scaffolding molecules known to associate with RHOA pathway members, AKAPs are potential candidates for this targeting [14, 20, 21, 31–35]. To date, six AKAPs have been identified in spermatozoa: AKAP1, AKAP3, AKAP4, AKAP11, MAP2, and WAVE1 [13, 36–38].
Although much work has been conducted on the regulation of proteins associated with AKAPs (i.e., PRKA phosphorylation of RHOA has an inhibitory effect [15–18]), studies on regulatory modification of the AKAPs themselves have been less abundant [39]. Several recent studies illustrate regulation of AKAP/PRKAR2 interaction via phosphorylation. Luconi et al. demonstrated that tyrosine phosphorylation of AKAP3 enhances sperm motility by increasing its binding to PRKAR2B [40]. AKAP9 phosphorylation by PRKA in response to β-adrenergic receptor (β2AR) stimulation regulates the IKs channel complex in the heart [41]. Phosphorylation of AKAP12 by PRKA enhances its association with the β2AR in a signaling complex that includes PRKC [42, 43]. Finally, phosphorylation of AKAP13 by PRKA leads to 14-3-3 binding, which inhibits the RHO-GEF activity of the AKAP [31–33, 35, 44].
In addition to phosphorylation, mammalian sperm function also is mediated by calcium signaling [45]. Calcium signaling regulates sperm motility and hyperactivation, chemotaxis, and the acrosome reaction, which all are determinants of normal fertility [46]. There is some convergence of the calcium and phosphorylation pathways in sperm via soluble adenylyl cyclase, which responds to both calcium and bicarbonate to produce cAMP, resulting in activation of PRKA [45]. Beta-adrenergic receptor-mediated stimulation of cardiac calcium channels is a result of cAMP signaling, perhaps through direct PRKA phosphorylation of the channel [47].
In the present study, we aim to determine which components of the GNA13-mediated RHOA-ROCK-LIMK-cofilin signaling cascade are present in bovine spermatozoa and to identify novel RHOA-binding partners. Based on previously demonstrated AKAP3 binding of PRKA, GNA13, and ROPN1, we hypothesize that AKAP3 is a potential binding partner that would serve to target and regulate RHOA signaling in sperm. Finally, we investigate the effects of PRKA phosphorylation and/or calcium-based regulation of AKAP3 binding to the RHOA-interacting proteins PRKAR2A and ROPN1, as well as the ROPN1-related R2D2 proteins, all in an attempt to better understand the actin dynamics that may serve to regulate the acrosome reaction, a key determinant of fertility.
Chemicals, Equipment, and Sources
All chemicals and protease inhibitors were from Sigma-Aldrich (St. Louis, MO) unless otherwise noted. Anti-GST antibody also was from Sigma. Glutathione sepharose fusion protein beads, protein A-sepharose 4B beads, and HiTrap Chelating HP columns were from Amersham Biosciences (Uppsala, Sweden). S-protein agarose, S-protein horseradish peroxidase conjugate, and BL21(dE3)-pLysS competent cells were from Novagen (Madison, WI). GTP
S was from Cytoskeleton (Denver, CO). Immobilon-P PVDF membrane was from Millipore Corp. (Bedford, MA). Phospho-(Ser/Thr) PRKA substrate antibody (monoclonal), cofilin, and LIMK2 antibodies were from Cell Signaling Technologies (Beverly, MA). Production of -AKAP3 antibody (rabbit, polyclonal) was described previously in Vijayaraghavan et al. [11]. Rabbit IgG and -ROCK2 (monoclonal) antibody were from BD Pharmingen (San Diego, CA). Goat -rabbit and goat -mouse horseradish peroxidase-conjugated secondary antibodies, polyclonal -RHOA(119), polyclonal -ROCK2 (H-85), and polyclonal -GNA13 antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). Western Lightning chemiluminescence and 32P-ATP were from Perkin Elmer (Boston, MA). Modified trypsin for mass spectrometry sample preparation was from Promega. Sonication was performed using the Sonic Dismembrator 60 from Fisher Scientific (Pittsburgh, PA). PRKA (cat) was kindly supplied by the Susan Taylor lab. Bovine testes came from Carlton Farms, (Carlton, OR), and ejaculated spermatozoa (cryopreserved) were from Select Sires Inc. (Plain City, OH). Investigations involving animals were conducted in accordance with "Guide for Care and Use of Laboratory Animals."
Bovine caudal spermatozoa were obtained as described previously [12]. For RHOA and ROCK2 Western blots, 4.0 x 108 spermatozoa were lysed in 250 µl buffer 6 (5 mM Tris, pH 7.6, 0.1% Triton X-100, 250 mM sucrose, 1 mM sodium vanadate, 1 mM sodium fluoride, 5 ng/ml leupeptin, 1 mM AEBSF [4-(2-aminoethyl)benzenesulfonyl fluoride hydrochloride], 20 µg/ml aprotinin, 1 mM benzamidine, and 10 µg/ml soybean trypsin inhibitor) by triturating, then sonicating three times (10 bursts, power 10). Samples were centrifuged for 10 min at 13 000 x g, 4°C, and 5x SDS sample buffer (62.5 mM Tris-HCl, pH 6.8, 10% glycerol, 2% SDS, 5% 2-mercaptoethanol, and 0.05% bromophenol blue) was added to the supernatants. Samples were separated by 10% or 15% SDS-PAGE and transferred to Immobilon-P membrane. After 1 h blocking solution (TBS [10 mM Tris, 150 mM NaCl, pH 7.5] with 0.1% BSA and 5% nonfat milk), blots were probed with -RHOA antibody (diluted 1:250 in TTBS [TBS with 0.05% Tween-20], 1 h at room temperature), or -ROCK2 antibody (diluted 1:250 in TTBS with 2.5% BSA at 4°C overnight). Secondary antibodies were horseradish peroxidase-conjugated goat -rabbit or goat -mouse, both diluted 1:5000 in TTBS and incubated for 1 h at room temperature. For GNA13, LIMK2, and cofilin Western blots, caudal sperm were lysed as above, but in sperm lysis buffer (20 mM Tris-HCl, pH 7.6, 50 mM NaCl, 1% Triton X-100, 1% β-mercaptoethanol, 0.1 mM EDTA, 1 mM sodium vanadate, 1 mM sodium fluoride, 5 ng/ml leupeptin, 1 mM AEBSF, 20 µg/ml aprotinin, 1 mM benzamidine, and 10 µg/ml soybean trypsin inhibitor). For LIMK2 Western blots, bovine and murine (C57 Black6x129 SvEv Brd hybrid) testes were frozen and crushed to a fine powder in liquid nitrogen, then placed in tissue lysis buffer (10 mM Tris-HCl, pH 7.6, 150 mM NaCl, 1% Triton X-100, 1% deoxycholate, 0.1% SDS, 5 mM EDTA, and protease inhibitors as above), homogenized on ice in a glass homogenizer, sonicated, and separated by electrophoresis as above. Western blots were performed using -GNA13 antibody (diluted 1:1000 in TTBS and incubated overnight at 4°C), -cofilin antibody, or -LIMK2 antibody (both diluted 1:1000 in TTBS with 5% BSA and incubated overnight at 4°C). Secondary antibody in all cases was horseradish peroxidase-conjugated goat -rabbit, diluted 1:5000 in TTBS and incubated for 1 h at room temperature.
RHOA was cloned from human testis cDNA library using touchdown PCR and with annealing temperatures as follows: five cycles at 60°C, five cycles at 58°C, and 25 cycles at 56°C. The purified PCR product was cloned into pGEX 5X-1 and expressed in BL21 (DE3) pLysS cells. Protein expression was induced with 0.5 mM Isopropyl-β-D-1-thiogalactopyranoside (IPTG) at 30°C for 2 h. The cells were lysed in RHO lysis buffer (50 mM Tris-HCl, pH 7.6, 5 mM MgCl2, 100 mM NaCl, 100 µM GDP, 0.5% Nonidet P-40, 1 mM dithiothreitol [DTT], 1 mM AEBSF, and 2 µg/ml aprotinin). The sample showed induction of RHOA-GST that was checked using the antibodies for RHOA and GST.
To detect RHOA-binding proteins, Rho overlays were performed using methods modified from Zong et al. [48]. Briefly, RHOA-GST or GST was lysed in RHOA lysis buffer, sonicated three times (10 bursts, power 10), and centrifuged at 16 000 x g to remove insoluble material. The supernatants were rotated with glutathione sepharose 4B beads at 4°C overnight, then washed three times with lysis buffer. Bead-bound RHOA-GST was loaded with nucleotide by incubation with loading buffer (50 mM Tris-HCl, pH 7.6, 50 mM NaCl, 5 mM EDTA, and 1 mM DTT) plus 200 µM GTP or 200 µM GDP for 20 min at 30°C. As an additional control, bead-bound GST was incubated with loading buffer plus 200 µM GTP. To stop the reactions, MgCl2 was added to 10 mM final concentration, and free nucleotide was removed by washing beads three times in loading buffer with 10 mM MgCl2 in place of nucleotide/EDTA. To elute proteins, beads were incubated on ice for 20 min with 200 µl elution buffer (20 mM glutathione, 100 mM Tris-HCl, pH 8.0, 50 mM NaCl, 5 mM MgCl2, 10% glycerol, and 1 mM DTT). Bovine caudal spermatozoa were lysed in sperm lysis buffer as for Western blotting above, subjected to SDS-PAGE, and transferred to Immobilon-P. Membrane was blocked in 1% BSA, 50 mM Tris, pH 7.4, 150 mM NaCl, 1 mM DTT, and 0.05% Tween 20 for 2 h at room temperature. Membranes were incubated for 15 min at 4°C with eluted proteins that were diluted 1:1000 in overlay buffer (blocking buffer with 10 mM MgCl2 and 100 µM GDP), then washed 3 x 5 min with the overlay buffer. Membranes were incubated for 20 min at room temperature with milk/Tween blocking buffer (TBS with 0.1% Tween 20 and 5% nonfat milk), rinsed once in TTBS, then incubated 40 min at room temperature with HRP-conjugated -GST antibody (diluted 1:3000 in TTBS).
Immunoprecipitation and Mass Spectrometry Analysis
Bovine caudal spermatozoa (2 x 109) were lysed in buffer 6. The supernatant was precleared by adding 30 µl of 50% protein A-sepharose slurry (made in lysis buffer) and rotating for 30 min at 4°C. For RHOA immunoprecipitation (IP), the precleared supernatant was incubated for 1 h at 4°C with 1.4 µg rabbit IgG, 1.4 µg -RHOA antibody, or 1.4 µg -RHOA antibody preincubated for 15 min with 4.2 µg blocking peptide. After adding 50 µl of 50% protein-A slurry, the mixtures were rotated for 30 min at 4°C. The beads were washed four times with lysis buffer by centrifugation at 13 000 x g for 4 min at 4°C and boiled in 30 µl SDS sample buffer for 5 min. Alternatively, the proteins were eluted from the beads by incubation with 4.2 µg blocking peptide for 4 h, followed by dialysis against 10 mM ammonium bicarbonate, then boiling in SDS sample buffer. All samples were separated by 15% SDS-PAGE, proteins were detected using silver stain [49], and bands unique to the RHOA IP lanes (one lane with beads and one without [peptide elution]) were excised using a clean razor blade and cut into 1-mm pieces. Gel pieces were incubated at room temperature in 500 µl of a solution containing 50 mM ammonium bicarbonate and 50% (v/v) acetonitrile solution for 30 min to destain. For all incubations, samples were placed on an orbital shaker to facilitate mixing. Solution was removed, and wash was repeated. Samples were placed in a vacuum concentrator until the gel pieces were completely dehydrated. To dry samples, 100 µl of 10 mM DTT/100 mM ammonium bicarbonate was added and incubated at 56°C in a water bath for 45 min. DTT solution was then removed, and 100 µl of 55 mM iodoacetamide/100 mM ammonium bicarbonate was added, and samples were incubated at room temperature for 30 min in the dark. Iodoacetamide solution was removed, and 500 µl of 50 mM ammonium bicarbonate, 50% (v/v) acetonitrile solution was added and incubated for 15 min. Samples were dehydrated again in a vacuum concentrator. Enough trypsin solution (78.7 mM ammonium bicarbonate, 7.87 mM CaCl2, 0.197 mM HCl, 0.02 µg/µl modified trypsin, 0.4 M urea, and 1 mM DTT) to cover gel pieces was added (approximately 100 µl), and samples were incubated on ice for 15 min. After samples were completely rehydrated, excess trypsin solution was removed. Digestion buffer without trypsin (50 mM ammonium bicarbonate and 50 mM CaCl2) was added, and samples were incubated at 37°C for 16 h (overnight). Samples were dehydrated in a vacuum concentrator, then redissolved in 5% formic acid. Samples then were analyzed by liquid chromatography-electrospray ionization tandem mass spectrometry (LC-ESI-MS-MS) at the OHSU Proteomics Core using SEQUEST and DTASelect software to interpret output. Confidence levels were determined using Scaffold Version 01-07-00 with peptide thresholds set at 90% minimum and protein thresholds set at 99% minimum, two peptides minimum. In all cases, the protein identification probability was greater than 99%. A list of the -RHOA antibody-associated proteins was generated. Three separate experiments were performed.
Expression, Purification, and Phosphorylation of AKAP3
AKAP3(1–350)-pET30a was cloned and expressed as described previously [20]. A pellet from 500 ml induced bacterial culture was lysed in 20 ml His-binding buffer (5 mM imidazole, 0.5 M NaCl, 20 mM Tris-HCl, pH 7.9, 1 mM AEBSF, 10 mg/ml soybean trypsin inhibitor, 50 mM leupeptin, and 1.25 mg/ml aprotinin). The pellet was resuspended, then sonicated three times (15 bursts at 20 power). Lysate was centrifuged at 11 500 x g for 15 min at 4°C to remove insoluble material, and the supernatant was further clarified by passage through a 0.45 µm syringe filter. The supernatant then was applied to a HiTrap Chelating HP column. The column was prepared by washing with deionized water, charging with 80 mM NiSO4, and equilibrating with His-binding buffer. Protein was eluted by adding 60 mM to 1 M imidazole (stepwise) in His-binding buffer. Purification was confirmed by 10% SDS-PAGE, transfer to Immobilon-P, Coomassie stain, and Western blot with -AKAP3 antibody (1:2000 in TTBS, 1 h at room temperature).
Purified AKAP3(1–350) was phosphorylated by the catalytic subunit of PRKA. Reaction was prepared in 100 µl kinase buffer (50 mM MOPS, 10 mM MgCl2, 0.25 mg/ml BSA, pH 7.0) with 2.5 mM MgATP, purified AKAP3, and 6% (v/v) kinase. A nonphosphorylated control, in which water replaced the kinase, also was prepared. Tubes were incubated 1 h at 25°C, and phosphorylation was checked by performing Western blots using an -phospho-(Ser/Thr) PRKA substrate antibody, diluted 1:500 in 2.5% BSA/TTBS, incubated overnight at 4°C.
Pulldown assays were performed as described previously [12], with the following modifications. For pulldown assays in which the effects of phosphorylation were being investigated, bacterially expressed, S-tagged AKAP3(1–350)-pET30 or S-tagged pET30 was immobilized on S-protein agarose, and then phosphorylated by PRKAC as described above. Beads were washed before addition of the secondary protein (ROPN1L-pGEX-5X, CABYR-pGEX-5X, ROPN1-pGEX-5X, or SPA17-pGEX-5X; all GST tagged). Equal loading and phosphorylation of AKAP3 was checked by AKAP3 and phospho-(Ser/Thr) PRKA substrate antibody Western blots. For pulldown assays in which the effects of calcium were being investigated, after S-tagged AKAP3(1–350) was immobilized on the beads, CaCl2 over a range of 0.1 µM to 1000 µM, 1000 µM + 1 µg/ml calmodulin, or 10 mM EDTA was added in addition to the secondary protein (CABYR-pGEX-5X, GST tagged).
Purified AKAP3(1–350)-pET30a, nonphosphorylated and PRKAC phosphorylated, was separated by electrophoresis and transferred to Immobilon-P using a protocol identical to that described for Western blotting. The blots then were probed with 32P-labeled PRKAR2A (100 000 cpm/ml), and PRKAR2A-binding proteins were visualized by autoradiography as previously described [50, 51]. Equal loading and phosphorylation of AKAP3 was checked as for phosphorylation pulldowns above. All densitometric analyses were performed on a Macintosh computer using the public domain NIH Image program (developed at the US National Institutes of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/).
All error bars represent standard error of the mean. Significance was determined using a Student t-test on Microsoft Excel.
RHO Pathway Components Are Present in Sperm
To determine whether known components of the RHO signaling pathway are present in bovine caudal spermatozoa, Western blots were performed using antibodies to GNA13, RHOA, ROCK2, LIMK2, and cofilin. A band of appropriate size was detected in each blot, at approximately 42 kDa for GNA13 (Fig. 1A), 25 kDa for RHOA (Fig. 1B), 160 kDa for ROCK2 (Fig. 1C), 47 kDa for LIMK2 (Fig. 1D), and 18 kDa for cofilin (Fig. 1E). These results indicate that all of the components of the GNA13-mediated RHOA signaling pathway are present in bovine caudal spermatozoa.
RHOA Binding Proteins in Sperm
To investigate the identity of RHOA-binding proteins in sperm, we performed an overlay assay using RHOA as a probe. Bacterially expressed recombinant GST and RHOA-GST were immobilized on beads, bound to either GTP (active RHOA) or GDP (inactive RHOA), then eluted from the beads. Bovine caudal spermatozoa proteins were separated by electrophoresis on a 10% acrylamide gel, and GST (+GTP), inactive GDP-bound RHOA-GST, or activated GTP-bound RHOA-GST elutions were incubated with the previously blocked membranes. A Western blot detecting the GST tag resulted in several prominent bands in the GTP-bound RHOA overlay (Fig. 2, panel 3). These bands were either much less prominent (GST with GTP; Fig. 2, panel 1), or not present (GDP-bound RHOA; Fig. 2, panel 2) in the control overlays. These results suggest the presence of multiple RHOA-binding proteins in sperm.
|
In an attempt to identify some of the RHOA-binding proteins revealed by the overlay assay, we performed an IP assay using RHOA antibody and then identified associated proteins by mass spectrometry analysis. To ensure that the interacting proteins were specific to the RHOA antibody, only bands that were unique to the RHOA IP lane (i.e., detected neither in the IgG nor RHOA plus antigenic peptide control lanes) were excised, trypsin digested, and subjected to mass spectrometry analysis. Proteins with a protein identification probability of greater than 99% included outer dense fiber protein 2 (ODF2), - and β-tubulin, proacrosin, aldolase C, and major fibrous sheath protein (also known as AKAP4; Table 1). The data presented in Table 1 represent one of three separate experiments. Of the proteins listed, ODF2, - and β-tubulin, and AKAP4 were identified in two of the three analyses. These results serve to identify several proteins that interact with RHOA in bovine spermatozoa, either through direct interaction or intermediary proteins.
|
Binding of AKAP3 to PRKA and R2D2 Proteins Is Regulated by Phosphorylation, but Not Calcium, In Vitro
The PRKAR2-like proteins ROPN1 (a protein involved in the RHO pathway via interaction with rhophilin) and ROPN1L were identified as AKAP3-binding proteins in a yeast two-hybrid screen [20, 21]. To determine whether the binding of AKAP3 to the RHOA-interacting proteins PRKAR2, ROPN1, or the ROPN1-related R2D2 proteins is regulated, we performed a series of pulldown and overlay experiments. To ascertain the role of AKAP3 phosphorylation in binding regulation, pulldown experiments were performed with AKAP3 and the R2D2 proteins. S-tagged AKAP3 was immobilized on S-protein agarose and was either left unphosphorylated or was phosphorylated with PRKA. GST-tagged ROPN1L, CABYR, ROPN1, or SPA17 was incubated with the immobilized AKAP3 or with immobilized S-protein as a nonspecific binding control (Fig. 3, A, B, C, and D, respectively). Samples were separated by electrophoresis, transferred, and immunoblotted for the R2D2 using GST antibody. Phosphorylation and equal loading of AKAP3 were confirmed as in Figure 4 (data not shown). Densitometric quantitation of bands in lanes 2 and 3 of each panel indicate that all R2D2 proteins bind to AKAP3, although SPA17 may bind more weakly than the others (note the small difference between SPA17 binding in the S-protein pulldown compared with S-tagged AKAP3 pulldown [Fig. 3D, lanes 1 and 2]). However, only the interaction between ROPN1 and AKAP3 (Fig. 3C) is significantly increased (P = 0.01) upon PRKAC phosphorylation of the AKAP.
|
|
To further investigate whether phosphorylation of AKAP3 affects its interactions, purified AKAP3(1–350)-pET30 (the PRKAR2 binding region of the AKAP) was either phosphorylated with the catalytic subunit of PRKA or left unphosphorylated. Samples were separated by SDS-PAGE, then transferred to Immobilon-P. Membranes were subsequently blocked and subjected to PRKAR2A overlay analysis (Fig. 4, A and B), which revealed that phosphorylation of AKAP3 increases its interaction with PRKAR2A by 36% compared with the nonphosphorylated control (P = 0.03). To confirm phosphorylation and equal loading of AKAP3, Western blots were performed with a PRKA-(Ser/Thr) substrate antibody (Fig. 4C) or -AKAP3 (Fig. 4D). These results indicate that the interactions between ROPN1 and PRKAR2 with AKAP3 are regulated by phosphorylation.
As one of the R2D2 proteins (CABYR) is a calcium-binding protein, we wanted to ascertain whether calcium signaling might regulate the AKAP3/CABYR interaction. A pulldown experiment was conducted in which S-protein or S-protein-tagged AKAP3 was immobilized on S-protein agarose, then GST-tagged CABYR or ROPN1L was added in the presence of various concentrations of calcium, plus or minus calmodulin (a calcium-binding protein that mediates calcium signaling) and EDTA (a chelator that should bind any calcium present, making it unavailable to bind to CABYR). Samples were separated by electrophoresis and transferred, and Western blots for the R2D2 proteins were performed. Bands from three separate experiments were quantitated by densitometry. Results for both CABYR (Fig. 5) and ROPN1L (data not shown) indicate that although both R2D2 proteins bind to AKAP3, calcium concentration does not significantly alter interaction in vitro. These data suggest that calcium does not regulate AKAP3 interactions with R2D2 proteins.
|
Membrane remodeling due to actin polymerization is known to precede the acrosome reaction in mammalian spermatozoa [25–27]. RHOA, a small GTPase, has been shown to regulate actin cytoskeletal remodeling through a pathway involving the phosphorylation of ROCK, LIMK, and cofilin, an actin depolymerizer that is inactivated by LIMK phosphorylation [8]. Although it has been suggested that this pathway may be present in spermatozoa [26], to our knowledge, the presence of GNA13, ROCK2, LIMK2, and cofilin has not previously been demonstrated in sperm. As such, we performed Western blot analysis to determine whether each of these proteins is expressed in bovine spermatozoa (Fig. 1). Previous studies using sperm membrane extracts and Western blot analysis have suggested that GNA13 and the related protein, GNA12, are not present in human sperm [52, 53]. However, a more recent report using total sperm protein for Northern and Western blotting, along with immunohistochemistry and immunofluorescence, demonstrates that GNA12 is present in human spermatozoa but is not localized to the plasma membrane, as would be expected for a G protein [54]. In this report, using total protein from bovine caudal sperm, we detected GNA13 by Western blot for the first time. Similar to previous findings [21, 28], RHOA was detected in caudal sperm lysates. Although ROCK1 has been localized in the spermatozoa of various mammalian species [28], ROCK2 had not. Although studies overexpressing ROCK1 and ROCK2 imply that they are functionally redundant, downregulation of each kinase with short-interfering RNA has revealed unique compartmentalization of the two kinases, as well as differential regulation of myosin II activity [55]. Here, we detected ROCK2 by Western blot, suggesting that both ROCK1 and ROCK2 may have roles in regulating sperm function.
Our LIMK2 Western blot detects a single band at approximately 47 kDa, which was lower than the expected size of 70 kDa. This antibody was raised in rabbit against a synthetic peptide (KLH coupled) corresponding to carboxy-terminal residues of human LIMK2, and it is known to recognize human, mouse, rat, and monkey LIMK2. One explanation for this size discrepancy is that the antibody is recognizing the previously described testis-specific variant of LIM kinase (tLIMK2) [56]. Discovered in pachytene spermatocytes and round spermatids in mouse testes, tLIMK2 lacks the N-terminal LIM domains and part of the PDZ domain of full-length LIMK2 while retaining the entire kinase domain. Takahashi et al. [56] observed a LIMK2 antibody cross-reacting band unique to testes that runs at approximately 50–55 kDa. While this is a bit larger than our observed band, this difference could be due to species variation (see Supplemental Fig. 2 for an LIMK2 Western blot of bovine sperm, bovine testes, and mouse testes). Finally, we detected cofilin in spermatozoa.
Thus, the proteins necessary for the GNA13-mediated RHOA-ROCK-LIMK-cofilin signaling pathway, which regulates actin polymerization and membrane remodeling in other cell types [8, 9], are present in bovine spermatozoa, suggesting that these proteins are involved in the actin cytoskeletal remodeling that occurs during capacitation and the acrosome reaction [25–27].
To further investigate possible RHOA function in bovine sperm, we performed binding assays to identify proteins that may interact with RHOA in vivo. Initially, a simple overlay technique was used (Fig. 2). Several bands were detected that interacted with activated RHOA + GTP, but not inactive RHOA + GDP, demonstrating that RHOA binds to a variety of proteins in spermatozoa. A number of RHOA-associated proteins were identified by performing mass spectrometry analysis on RHOA immunoprecipitations. Proteins that are unique to the RHOA IP are presented in Table 1 and discussed in the following paragraphs.
Angiotensin-converting enzyme (ACE) catalyzes the conversion of angiotensin I to its activated form, angiotensin II, which has a wide range of effects, including potent vasoconstriction. In capacitated bovine spermatozoa, angiotensin II induces acrosomal exocytosis [57]. A testis-specific form of ACE (tACE) has been demonstrated to be necessary for normal fertilization in vivo. Although murine spermatozoa with a disruption in the tAce gene that prevents synthesis of tACE have normal viability, motility, and ability to undergo capacitation and AR, they are subfertile, perhaps due to defects in sperm transport in the oviduct and impaired binding to the zona pellucida [58].
Major Fibrous Sheath Protein (AKAP4)
AKAP4 is the most abundant protein in the fibrous sheath of spermatozoa, where it has been found to bind fibrous sheath interacting proteins 1 and 2, and AKAP3 [59]. In Akap4 gene knockout mice, subcellular distributions of PRKA, PI 3-kinase, and SPA17 (R2D2 protein) are disrupted, activity and phosphorylation of PP12 are significantly altered, and the mice are infertile due to lack of sperm motility [60].
Proacrosin is the inactive precursor to acrosin, which is a trypsinlike proteolytic enzyme located in the acrosomal matrix. Proacrosin is autoactivated during the AR, has been shown to accelerate the dispersal of acrosomal proteins, and may be involved in activation and/or deactivation of other acrosomal proteins through serine proteolytic activity. Although mouse sperm lacking acrosin protease activity are still fertile, there is a significant correlation between acrosin activity and male infertility in humans [61, 62].
Tubulins— and β—are globular proteins that form heterodimers, which polymerize into microtubules [63]. While RHO-GTPases have long been known to regulate actin dynamics, they have recently been found to be involved in microtubule dynamics in a manner that can be independent of actin. For example, RHOA promotes the formation of stabilized microtubules in fibroblasts, demonstrated by an increase in detyrosinated tubulin. How this occurs is not clear, but it appears to be mediated by mDia (a downstream RHO effector) and does not affect the actin cytoskeleton [1, 64]. Additionally, several different RHO-GEFs have been shown to bind microtubules. For example, in Xenopus development, microtubules regulate RHO through binding to XLFC (a RHO-GEF). Again, how this regulation occurs is not clear—nocodazole (a microtubule destabilizer) inhibits actin-based protrusion in these cells, but the exact mechanism is unknown. XLFC nucleotide exchange activity is required for this inhibition, and the effects of nocodazole are partially rescued by addition of dominant negative RHO or RHO kinase inhibitor [65]. In podocytes (renal glomerular visceral epithelial cells), a RHO-ROCK pathway seems to negatively regulate microtubule-based process elongation [66]. Clearly, the interactions between microtubules, actin, and RHO-GTPases are varied and complex.
ODF2, Hexokinase, and Aldolase C
Outer dense fiber protein 2 (ODF2) is a major component of the cytoskeletal structure of the sperm tail and is associated, albeit indirectly, with microtubules [67]. Hexokinase binds ATP in an actin fold-like structural motif to phosphorylate glucose, and it has been localized in the acrosome, midpiece, and tail of human spermatozoa [68, 69]. Aldolase C is both a component of the glycolytic complex associated with the fibrous sheath and an F-actin-binding enzyme [70, 71].
Thus, we have identified a number of very interesting potential RHOA-binding proteins, many of which affect capacitation and the acrosome reaction, both essential processes for fertility.
RHO is undoubtedly involved in a wide variety of signaling pathways in many different cell types. It has been suggested that the number and variety of these pathways require that RHO signaling be tightly spatially regulated [1, 8]. AKAPs are known to act as scaffolding proteins that target and coordinate many signaling molecules [20]. AKAP3 has been shown to bind activated GNA13, which is upstream of the RHOA-ROCK-LIMK-cofilin pathway. Although this interaction is suggested to provide a cAMP-independent mechanism for PRKA activation [14], other studies indicate that PRKA has a negative regulatory effect on RHOA through direct phosphorylation and inactivation of RHOA [15–17], phosphorylation and inactivation of upstream activator GNA13 [18], or some combination of the two. Additionally, AKAP3 binds ROPN1, an R2D2 protein that interacts indirectly with RHOA via binding to rhophilin [20, 21] (see Fig. 6 for simplified models of potential AKAP3/AKAP4 interactions with the RHOA pathway). We therefore hypothesized that AKAP3 is a scaffolding molecule for the RHO pathway in bovine spermatozoa. As such, we were surprised to find AKAP4 in our RHOA IPs. To our knowledge, AKAP4 has not been shown previously to bind RHOA or any RHO-associated proteins, and it was specifically shown not to directly bind SPA17 [23], which is a ROPN1-related R2D2 protein. Therefore, we did not consider it a strong candidate for targeting of the Rho pathway in sperm. While it is possible that AKAP3 simply does not associate with the RHOA pathway in sperm, there are other possible explanations for not detecting it. AKAP3 is highly insoluble [12] and may not have been extracted in our mild buffer conditions. Although AKAP4 is also quite insoluble, it is more abundant in the fibrous sheath than AKAP3, and it also has a soluble precursor form [59]. Interestingly, mature AKAP4 binds to AKAP3 [59], which allows for the possibility that there may be multiple AKAPs involved in targeting a single pathway, such as RHOA signaling.
|
Several recent studies have indicated regulation of AKAP protein interactions via PRKA phosphorylation of the AKAPs (see Introduction and [31–33, 35, 40–44]). Consequently, we wanted to address whether or not AKAP3 binding to RHO-interacting molecules, such as the regulatory subunit of PRKA and ROPN1 (along with the related R2D2 proteins), was regulated by phosphorylation. Binding to both ROPN1 and PRKAR2A significantly increased when AKAP3 was phosphorylated (Figs. 3C and 4, respectively). Although AKAP3 also bound to the other R2D2 proteins, there were no significant differences in binding when AKAP3 was phosphorylated. It would appear then that PRKA phosphorylation of AKAP3 has a regulatory effect on its association with specific proteins in vitro. We are especially excited to see a specific regulatory effect with ROPN1 binding, as ROPN1 complexes with RHOA via rhophilin, and thus this regulation has the potential to influence RHOA signaling.
Calcium signaling is another major contributor to sperm function and has been shown to be involved in the RHOA signaling pathway in other cell types [72–74]. Since the ROPN1-related R2D2 protein CABYR is a calcium-binding protein whose interaction with AKAP3 was not affected by phosphorylation (Fig. 3), we performed experiments to determine whether varying concentrations of calcium would affect AKAP3 binding. We detected no change in binding between AKAP3 and either CABYR (Fig. 5) or ROPN1L (data not shown) when calcium concentration was varied in vitro. While these results seem to indicate that calcium does not affect AKAP3 binding, signaling can be quite complex, and it is possible that our experimental model is too simple, or lacking in necessary co-factors, to detect effects of calcium. Another possibility is that calcium does not affect binding to R2D2 proteins but may have effects on other aspects of AKAP interactions. In support of this notion, Tao et al. recently demonstrated that AKAP12 translocation to the cytoplasm of transfected A431 cells is mediated by calcium/calmodulin binding to the membrane interacting sites of the AKAP [75].
In summary, we have provided evidence for the presence of the components of the actin-remodeling, GNA13-mediated RHOA-ROCK2-LIMK2-cofilin signaling pathway in bovine spermatozoa. We also have identified several potential RHOA-interacting proteins that affect the critical functions of capacitation and acrosome reaction. Specifically, we report that RHOA immunoprecipitates AKAP4, providing the first evidence of possible AKAP targeting of the RHO pathway in bovine spermatozoa. Finally, we demonstrate that AKAP3 binding to the RHOA-associated proteins ROPN1 and PRKAR2 is regulated in vitro by PRKA phosphorylation. Taken together, these results serve to enhance our knowledge of actin regulation in spermatozoa and may lead to better understanding of the actin dynamics that precede the acrosome reaction and subsequent fertilization.
ACKNOWLEDGMENTS
We would like to thank Dr. Susan Taylor for providing the PRKA catalytic subunit. We would also like to thank Dr. Robynn Schillace and Dr. Sonemany Salinthone for their helpful comments in the preparation of this manuscript, and Jenny Ruan for technical assistance.
FOOTNOTES
1Supported by the Department of Veterans Affairs Biomedical Laboratory Research and Development Service, National Institutes of Health grant HD36408 to D.W.C., the Proteomics Shared Resource, which is funded by the Oregon Opportunity, and by National Institutes of Health center grants 5P30CA069533 and 5P30EY010572. 
Correspondence: 2Correspondance: Daniel W. Carr, Veterans Affairs Medical Center, Mail Code R&D8, 3710 SW US Veterans Hospital Rd., Portland, OR 97239. FAX: 503 721 1082; e-mail: carrd{at}ohsu.edu
Received: 16 May 2007.
First decision: 8 June 2007.
Accepted: 26 September 2007.
REFERENCES
| ||||||||||||||||||
