View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Developmental Biology 274 (2004) 426–435 www.elsevier.com/locate/ydbio

Enkurin is a novel calmodulin and TRPC channel binding in sperm

Keith A. Sutton1, Melissa K. Jungnickel1, Yanli Wang, Kay Cullen, Stephen Lambert, Harvey M. Florman*

Department of Cell Biology, University of Massachusetts Medical School, Worcester, MA 01655, United States Received for publication 5 May 2004, revised 23 July 2004, accepted 26 July 2004 Available online 25 August 2004

Abstract

The TRPC cation channel family has been implicated in receptor- or phospholipase C (PLC)-mediated Ca2+ entry into animal cells. These channels are present in mammalian sperm and are assigned a role in ZP3-evoked Ca2+ influx that drives acrosome reactions. However, the mechanisms controlling channel activity and coupling Ca2+ entry through these channels to cellular responses are not well understood. A yeast two-hybrid screen was carried out to identify TRPC-interacting that would be candidate regulators or effectors. We identified a novel protein, enkurin, that is expressed at high levels in the testis and vomeronasal organ and at lower levels in selected other tissues. Enkurin interacts with several TRPC proteins (TRPC1, TRPC2, TRPC5, but not TRPC3) and colocalizes with these channels in sperm. Three protein–protein interaction domains were identified in enkurin: a C-terminal region is essential for channel interaction; an IQ motif binds the Ca2+ sensor, calmodulin, in a Ca2+-dependent manner; and a proline-rich N-terminal region contains predicted ligand sequences for SH3 domain proteins, including the SH3 domain of the p85 regulatory subunit of 1-phosphatidylinositol-3-kinase. We suggest that enkurin is an adaptor that functions to localize a Ca2+ sensitive signal transduction machinery in sperm to a Ca2+-permeable ion channel. D 2004 Elsevier Inc. All rights reserved.

Keywords: Acrosome reaction; Fertilization; Ion channel; PI3 kinase; SH3 domain; Signal transduction; Sperm; TRPC channel

Introduction 2000). In mammals, products of the seven trpc (TRPC1-7) were identified based on sequence similarities Activation of phospholipase C (PLC) generates a Ca2+ to TRP (Wes et al., 1995; Zhu et al., 1996) and have been entry into many animal cells. The transient receptor potential- implicated in a rapidly expanding set of physiological canonical (TRPC) family of cation channels have emerged as processes (Clapham, 2003; Montell et al., 2001; Putney, candidate subunits of these PLC-dependent Ca2+ influx 2004). channels (Clapham, 2003). The eponymous member of this These Ca2+ entry pathways are also present in mammalian family, the TRP channel of Drosophila melanogaster,is sperm and control the initial stages of fertilization. Sperm expressed in photoreceptor cells and functions with its interaction with the egg extracellular matrix, or zona relatives, the trpl and trpc products, to generate a pellucida (ZP), is a key early event during fertilization. In sustained ionic current in response to illumination (Minke order to penetrate the ZP and fuse with the egg, sperm must and Cook, 2002; Montell, 2003). Homologues of dipteran first undergo a secretory event or acrosome reaction (Evans TRP have been identified in animals ranging from nematodes and Florman, 2002; Primakoff and Myles, 2002). The to the vertebrates (Birnbaumer et al., 2003; Harteneck et al., initiation of an acrosome reaction following gamete adhesion is controlled by ZP3, a glycoprotein component of the zona pellucida (Bleil and Wassarman, 1983). Early events of ZP3 * Corresponding author. Department of Cell Biology, University of signal transduction in sperm include activation of the G class Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA i 01655. Fax: +1 508 856 1033. of heterotrimeric G proteins (Wilde et al., 1992), elevation of E-mail address: [email protected] (H.M. Florman). internal pH (Florman et al., 1989), the generation of a 2+ 2+ 1 These authors contributed equally to this work. transient Ca influx through a low voltage-activated Ca

0012-1606/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2004.07.031 K.A. Sutton et al. / Developmental Biology 274 (2004) 426–435 427 channel (Arnoult et al., 1996, 1999), and the stimulation of and interacted with mTRPC2 when the inserts were PLCy (Fukami et al., 2001, 2003). These upstream elements exchanged between plasmids. These positive clones were lead to a sustained Ca2+ entry into sperm that directly drives subjected to secondary rounds of screening against other secretion (O’Toole et al., 2000). TRPC2 was assigned a role TRPC sequences, as described in the Results section. in this process based on observations that an antibody directed against an extracellular domain of that channel DNA constructs and cell culture blocked a substantial portion of this ZP3-evoked Ca2+ entry and also inhibited the acrosome reaction (Jungnickel et al., The sequences for cloned m- and h-enkurin cDNAs are 2001). Other members of the TRPC channel genes are also given in Genbank AY454124 and AY454125, respectively. expressed during spermatogenesis, and in some cases channel cDNA clones of TRPC channels were kindly provided by proteins are retained in sperm (Castellano et al., 2003; Lutz Birnbaumer (National Institute of Environmental Trevino et al., 2001; Wissenbach et al., 1998). Among the Health Science; hTRPC1), David Clapham (Harvard Med- unresolved questions regarding TRPC channels in sperm, as ical School; mTRPC5), and Hong-Sheng Li (University of well as in somatic cells, are the following: What are the Massachusetts Medical School; hTRPC3). The following transducer elements that couple PLC activation to alterations oligonucleotides (5V-3V) were used as primer pairs to clone in channel conductance state, and what are the downstream sequences for secondary rounds of enkurin interaction effector elements that convert cation entry through TRPC in yeast: hTRPC1, 5V-ATGGCAGCGGCCCTGTACCCGA channels into cellular responses such as the acrosome GC, 3V-TCCCGACATCTGTCCAAACCA; hTRPC3, 5V-C reaction? ATGTTCTAACACGGGAT, 3V-AAGACATGGTCCCTGT The efficient operation of signal transduction pathways CC; and mTRPC5, 5V-ATGGCCCAGCTGTACTACAAG, depends upon the assembly of multimeric protein complexes 3V-TCCAGGGAAGCCATCGTA. (Ferrell and Cimprich, 2003; Hung and Sheng, 2002; FLAG-tagged enkurin was constructed by inserting a Pawson, 2003). The study of dipteran and mammalian Kozak-containing oligonucleotide with three FLAG epitopes channels suggests that coupling both to upstream activating upstream of the enkurin coding sequence in pEF1 (Invitro- signals and to downstream effector pathways may depend gen, Carlsbad, CA). GST and MBP fusion proteins were upon proteins that are associated with TRPC channels produced in BL21 bacteria. Expression was induced by the (Harteneck, 2003; Montell, 2003). Yet, the mechanisms by addition of 0.3 mM IPTG. Induction was for 3 h at 308C. which TRPC-associated proteins regulate channel activation Protein purification was performed following manufacturers and the downstream signal processing are not completely protocols for MBP (New England Biolabs, Beverly, MA) and understood. We approached this problem by carrying out a GST (Amersham Biosciences, Piscataway, NJ). yeast two-hybrid screen designed to identify TRPC binding HEK 293 cells (ATCC, Manassas, VA) were grown in proteins that were candidate regulators or effectors of these DMEM supplemented with 10% FCS. Transfections were channels in sperm. Here, we report the identification of performed using Effectene (Qiagen, Valencia, CA) accord- enkurin (derived from the Greek verb enkuros: to trip or to ing to the manufacturer’s protocol. Total RNA was prepared stumble upon), a conserved protein that is present in several from murine tissues by homogenization in Rnazol and regions of the mouse sperm. Based on the results presented Northern blots were performed as previously described here, it is proposed that enkurin functions as an adapter (Sutton et al., 1998). protein, tethering signal transduction proteins to TRPC Enkurin sequence analysis was performed using DNAS- channels. TAR (DNASTAR Inc., Madison, WI).

Antibodies and immunological methods Materials and methods Primary antibodies used in this study are listed in Table 1. Yeast two hybrid Peptides corresponding to TRPC1481–509 (NH2-KVVAHNK FHDFADRKDWDAFH-COOH) and to enkurin14–27 (NH2- Nucleotides 423–1967 of mTrp2-17 (AF111107) encode IPSDLKEPPQHPRY-COOH; designated IPSD peptide) an intracellular region of mTRPC2 (mouse TRPC2, hence- were synthesized and used to generate affinity-purified forth m- and h- designates mouse and human gene products, antibodies from rabbits (Research Genetics, Huntsville, respectively) from the N-terminus to the beginning of the H1 AL). The TRPC2-specific anti-RDAS was generated as domain, including the three ankyrin repeats (Vannier et al., described previously (Jungnickel et al., 2001). Antibodies 1999). This sequence was cloned into a LexA vector against TRPC3, TRPC5, and p85 were obtained from (pBTM116) and used to screen a murine testis cDNA library Alomone Laboratories (Jerusalem, IS), Santa Cruz Biotech- that had been cloned into the GAL4 activation domain nology (Santa Cruz, CA) (C-17 antibody), and Upstate (Lake plasmid pACT using the L40 yeast strain (BD Biosciences Placid, NY), respectively. AlexaFluor 546-conjugated sec- Clontech, Palo Alto, CA). Positive clones were identified as ondary antibodies were obtained from Molecular Probes those that did not activate transcription in the absence of bait (Eugene, OR). 428 K.A. Sutton et al. / Developmental Biology 274 (2004) 426–435

Table 1 excess of competing peptide. Blots were incubated with Specificities and sources of antibodies primary antibody (overnight, 48C), washed five times in Antibody Epitope Source TBST, incubated (1 h) with the horseradish peroxidase- (accession no.) (catalog number) conjugated secondary antibody (1:2000; Santa Cruz Bio- TRPC1 Residues 481–509 Our laboratory technology), washed five times with TBST, and the signal _ (NP 035773) was detected by chemiluminescence (Perkin Elmer NEN, TRPC3 Residues 822–835 Alomone Laboratories (Q9QZC1) (ACC-016) Boston, MA). TRPC5 C-terminal 17 Santa Cruz Biotechnology residue peptide (sc-18737) Calmodulin and p85 binding Enkurin Residues 14–27 Our laboratory (AY454124) FLAG-enkurin-transfected HEK293 cells were solubi- p85 Full-length p85 fusion protein Upstate (610045) lized in 50 mM Tris (pH 7.4), 150 mM NaCl, and 0.1% Triton X-100 in the presence or absence of 1 mM Ca2+. Insoluble Mouse sperm were obtained from the cauda epididymis, material was removed by sedimentation (14,000 Â g, air-dried onto glass slides, rinsed in TBS (20 mM Tris, pH 48C). Aliquots of the supernatant fraction (500 Al) were 7.6, 137 mM NaCl), fixed (4% paraformaldehyde, 10 min), incubated with agarose beads (2 h, 48C) and then with 50 Al extracted with CH3OH (À208C, 5 min), rinsed in TBS, calmodulin–agarose (5 h, 48C; Calbiochem, San Diego, CA) quenched (fresh 0.1% NaBH4/TBS, 5 min), and rinsed as that had been preequilibrated in the same buffer. In selected before. Slides were incubated with 1% bovine serum experiments, calmodulin–agarose was preincubated with 10 albumin/TBS (1 h) and then with primary antibody for AM m-enkurin174–191 (NH2-NYIQENLKKAAMKRLSDE- 1 h in a humid chamber at 378C. The specificity of labeling COOH; Prosci Inc., Poway, CA) before addition of cell was assessed in control experiments by preabsorbing extracts. Calmodulin–agarose was washed five times with primary antibody with competing peptide (10 AM). After either Ca2+-supplemented or -depleted buffer, bound proteins primary antibody treatment, the slides were incubated (1 h, were eluted in SDS sample buffer, and proteins were resolved 378C) in a humid chamber with secondary antibody (1:400 by SDS-PAGE. Free Ca2+ concentrations were estimated dilution; Molecular Probes). Slides were washed three to using MaxC (Chris Patton, Stanford University; http:// five times with PBS after each step. www.stanford.edu/~cpatton/maxc.html). Immunohistochemistry was performed on 5 Am thick SH3 Domain Array 1 (Panomics, Redwood City, CA) was sections of fixed (emersion in Bouin’s Fixative), paraffin- screened with lysates from FLAG-enkurin-transfected HEK embedded testis with an ABC kit (Vector Laboratory, 293 cells. Pulldown experiments were performed using GST- Burlingame, CA) using the enkurin-specific antibody, anti- SH3 domain fusion proteins, which had been expressed in ISPD. Sections were counterstained with dilute hematoxylin. BL21 bacteria and purified on glutathione–agarose (Amer- Samples for immunoprecipitation were extracted with sham Biosciences). Protein levels were quantified after RIPA buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 0.1% SDS, release of recombinant proteins from glutathione–agarose 1% NP-40, 0.5% DOC), insoluble components were removed with 0.5% SDS. Binding studies with FLAG-enkurin by sedimentation (100,000 Â g, 30 min), and the supernatant expressed in HEK 293 cells were performed using recombi- fraction was incubated for 2 h with preimmune IgG or protein nant protein bound to glutathione–agarose beads. In vitro A–agarose (Seize; Pierce Biotechnology, Rockford, IL). The translation was performed using the ProteinScriptII system nonbound fraction was recovered and incubated with primary (Ambion, Austin, TX) with enkurin cDNAs carrying site- antibody or protein A–agarose overnight at 48C. Immune directed mutations introduced using the QuickchangeII complexes were washed three times and bound antigen was system (Stratagene, La Jolla, CA). The following oligonu- dissociated from the antibody as described by the manufac- cleotides were used as primer pairs (5V-3V) to clone mouse turer. mTRPC5-MBP fusion proteins were incubated with p85a (GCTGCAGACATGAGTGCAGA and AAAGAATT FLAG-enkurin, immobilized on amylose resin (New England TCACCCCCT) and p85h (GCCATGGCAGGAGCC- Biolabs), isolated by sedimentation, washed, bound protein GAGGG and TGGCTCTGGCTCGACCA) into pGEX2T resolved by SDS-PAGE, and enkurin visualized with anti- (Amersham Biosciences). FLAG. Control binding reactions with MBP were carried out under identical conditions. Proteins were resolved by SDS/polyacrylamide gel Results electrophoresis (Laemmli, 1970) and transferred to Immobi- lon-P membranes (100 V, 2 h, 48C; Millipore, Bedford, MA). Cloning and characterization of enkurin Membranes were blocked for 1 h in medium TBST (TBS, 0.1% Tween 20) containing 4% (w/v) nonfat milk and probed Our previous work identified the mTRPC2 channel as a with the appropriate antibody (diluted in TBST/4% nonfat mediator of ZP3-induced Ca2+ entryinmousesperm milk) as indicated in the figure legends. In controls experi- (Jungnickel et al., 2001). In order to study the function of ments, primary antibody was first incubated with molar this protein, we carried out a yeast two-hybrid screen of a K.A. Sutton et al. / Developmental Biology 274 (2004) 426–435 429 mouse testis cDNA library using the amino terminus of Table 2 mTRPC2 as the bait. In designing this screen, we noted that Summary of TRPC and enkurin localization in mouse sperm an antibody against an extracellular domain of mTRPC2 TRPC1 TRPC2a TRPC3 TRPC5 Enkurin completely inhibited the TRPC2-dependent current in a Head heterologous expression system but only reduced the ZP3- Acrosomal ++ ++ evoked Ca2+ response of mouse sperm by 80–85% crescent i Equatorial + (Jungnickel et al., 2001). This suggested that other channels, segment in addition to mTRPC2, may also be activated by ZP3 Posterior head + during sperm interaction with the zona pellucida and Flagellum contribute a fraction of the evoked Ca2+ entry. Midpiece Mammalian sperm express several members of the TRPC Principal piece + + channel family (Castellano et al., 2003; Trevino et al., 2001) Distributions of TRPC1, TRPC3, TRPC5, and enkurin are shown in Figs. 1 2+ and 5 and summarized here. that might conduct the residual Ca influx seen following a TRPC2 distributions had been reported previously (Jungnickel et al., antibody block of mTRPC2 (Jungnickel et al., 2001). We 2001) and are provided for comparison. mapped the distribution of TRPC proteins in mouse sperm, observing the following: TRPC1 and TRPC5 in the acrosomal crescent of the anterior sperm head; and TRPC3 amino termini of TRPC1, TRPC3, and TRPC5 to obtain in the posterior head and in the flagellar principal piece (Fig. candidates that may interact with several sperm channels. 1, Table 2). Previously, we had found that TRPC2 was Positive clones in this secondary screen were analyzed for present in the anterior sperm head, including both the the degree of conservation of sequence and expression site acrosomal crescent and the equatorial segment (Jungnickel using the GenBank EST databases. We identified nine et al., 2001). Channels that participate in the acrosome candidate clones by this approach, two of which were reaction are expected to be in the head, possibly in the independent clones of the same cDNA. These independent acrosomal crescent and/or equatorial segment. clones had identical 3V ends, which included a stop codon We took advantage of these observations to narrow the and extended into a polyA region, but had different 5V ends pool of positive clones in our yeast two-hybrid screen for (Fig. 2A). A full length was assembled from EST data and further study. An initial set of mTRPC2-interacting cDNA confirmed by PCR. This clone, which we designated clones was screened in the two-hybrid system against the enkurin, interacted with TRPC1, TRPC2, and TRPC5, but

Fig. 1. TRPC channel distribution on mouse sperm. (A) The mouse sperm is organized into morphological domains, including the acrosomal crescent and equatorial segment, which overlie the acrosome; the posterior head; and the midpiece and principal piece regions of the flagellum. (B–G) Localization of TRPC1, TRPC3, and TRPC5 antibody binding sites in mouse sperm in the absence (B–D, total antibody binding) and presence of competitive antigenic peptide (E–G, nonspecific antibody binding). Typical distribution patterns were drawn from observations on groups of N50 sperm for each antibody. Results are summarized in Table 2. 430 K.A. Sutton et al. / Developmental Biology 274 (2004) 426–435

Fig. 2. Enkurin amino acid sequences. (A) Key features of mouse enkurin include the following: the positions of the two independent enkurin clones obtained from library screens; and the relative positions of the proline-rich N terminal domain, the p85 binding site, the IQ motif, and the C-terminal TRPC binding site. (B) Alignment of enkurin sequences. Protein sequences were derived from ESTs and PCR clones (human and mouse) or from assembled ESTs (chicken, zebrafish, ciona, and sea urchin). In cases where no upstream stop codon was identified, then the methionine at which translation is presumed to be initiated is shown in bold. Amino acid identities and sequence gaps are based on comparison to mouse sequences and shown as yellow blocks and dots, respectively. Percent identities to mouse enkurin, as determined from complete sequences using the Jotun Hein method (PAM250 matrix), were as follows: human, 87.9; chicken, 65.1; zebrafish, 56.7; ciona, 55.6; sea urchin, 55.6. Symbols above sequences are as follows: red dashed line (- - - -), IPSD peptide target of enkurin antibody; black dashed line (- - - -), predicted p85 SH3 ligand sequence; red double line (= = = =), predicted IQ motif; asterisk (*), DP1 proline residues at positions 83, 86, and 89 that, when mutated to alanine, greatly reduced p85 binding to m-enkurin (see Fig. 7); cross (+), DP2 proline residues at positions 21 and 22 that, when mutated to alanine, have no effect on p85 binding to m-enkurin (see text). not with TRPC3 in yeast two-hybrid assays (Fig. 3), and (44/255 residues, 17%; deduced pI, 9.58). It lacks predicted was selected for further analysis. hydrophobic or transmembrane domains and is most likely a Sequence analysis indicates that enkurin is a single-copy soluble, nonsecreted protein. The sequence contains protein gene with no homology to characterized proteins. The interaction motifs, including an amino terminal proline-rich deduced amino acid sequence of m-enkurin is lysine-rich region (m-enkurin4–95, 14% proline) that contains predicted K.A. Sutton et al. / Developmental Biology 274 (2004) 426–435 431

present in the same regions of the sperm cell (compare to Fig. 1, Table 2). We found that the cellular localization of endogenous enkurin differed from that in heterologous expression systems. There was no indication of a nuclear localization of endogenous enkurin in spermatogenic cells, sperm (Figs. 5B and C), or in the lung (data not shown). In contrast, we observed extensive nuclear localization of m-enkurin follow- ing heterologous expression in HEK-293 cells. In this regard, m-enkurin is a highly basic protein and might be expected to accumulate in acidic compartments. We are presently exploring the possibility that endogenous enkurin assembles into multiprotein complexes in male germ cells and that these associated proteins account for nonnuclear localization. Concentration of enkurin in the nucleus of HEK 293 cells and other heterologous expression systems complicated Fig. 3. Yeast interaction assays of enkurin–TRPC association. Negative attempts to assess enkurin effects on TRPC channel function. control, pGAD-STE4 + pBTM-GPA1; positive control, pGAD-STE4 + pBTM-TRPC2. Enkurin interactions with TRPC1, TRPC2, TRPC3, and Enkurin interaction with TRPC channels, calmodulin, and TRPC5 were screened as described (see Materials and methods). PI3 kinase p85 ligands for SH3 domain proteins (Obenauer et al., 2003)and We next examined three types of interactions of enkurin a candidate IQ motif (m-enkurin176–187), suggesting that with other proteins. First, we have noted that enkurin enkurin may be a calmodulin-binding protein. Orthologous associates with N-terminal regions of several (TRPC1, sequences can be found in vertebrate and invertebrate TRPC2, and TRPC5) but not all (TRPC3) TRPC proteins species (Fig. 2B). Both sequence motifs are conserved in in yeast interaction assays (Fig. 3). These interactions were h-enkurin and possibly in other orthologs (see below). confirmed in HEK 293 cells for the cases of TRPC1 and Enkurin transcripts were detected at high levels in the TRPC5. As shown in Fig. 6A, FLAG-enkurin formed an testis and VNO, which are both sites of mTRPC2 function immunoprecipitable complex with HA-hTRPC1. We also (Jungnickel et al., 2001; Leypold et al., 2002; Liman et al., found that FLAG–enkurin was bound to immobilized MBP- 1999; Stowers et al., 2002), and at lower levels in ovary, heart, lung, and brain (Figs. 4A and B). We were unable to detect enkurin expression in other tissues (Figs. 4A and B) or in several cell lines, including HEK293, HD-57, and COS cells (data not shown). To investigate the sites of protein expression in the testis, we prepared an antibody against a synthetic peptide corresponding to residues 14–27 of mouse enkurin (Table 1). This antibody specifically interacted with a 25-kDa protein in sperm extracts (Fig. 5A). This is somewhat smaller than the 29.6-kDa mass anticipated from the cDNA sequences and may reflect either processing of m-enkurin RNA or protein or proteolysis during extraction. In particular, the lysine-rich sequence of m-enkurin provides a number of potential cleavage sites for the family S1 peptidase of sperm, acrosin (EC 3.4.21.10). Enkurin immunoreactivity localized to the postmeiotic spermatid population of differentiating spermatogenic cells, where it was associated with the developing acrosome. In contrast, germ cells in the earlier spermatogonial and spermatocyte stages were not labeled (Fig. 5B). Enkurin is also retained in sperm (Fig. 5A), where it is localized Fig. 4. Enkurin RNA expression in mouse and human tissues. (A) RNA specifically in the acrosomal crescent and the principle piece loaded on each lane: VNO, 3 Ag; testis, 5 Ag; other tissues, 10 Ag. (B) Human expression studies used a Clontech human MTN blot in which of the flagellum (Fig. 5C). Antibody binding in all regions RNA loads were normalized to give equivalent h-actin levels Abbreviations was specific and was greatly reduced by coincubation with used: (A) VNO, vomeronasal organ; (B) PBL, peripheral blood leukocytes; the antigenic peptide. Enkurin and TRPC channels are then sm. intestine, small intestine. 432 K.A. Sutton et al. / Developmental Biology 274 (2004) 426–435

Fig. 5. Enkurin protein expression in mouse testis and sperm. Enkurin was detected using an antibody against residues 14–27 (Table 1). (A) Enkurin migrates as an approximately 25 kDa protein in mouse sperm extracts. Labeling was reduced by N95% following incubation of antibody with 10 AM antigenic peptide and so was considered specific. The top (ori) and dye front (df ) of this gel are also shown. (B) Enkurin localizes in or near the developing acrosome in mouse spermatids. Insets: Antibody labels the acrosomal region of condensing spermatids (inset B1) and also of early stages of acrosome formation in spermatids (inset B2). In contrast, enkurin could not be detected in spermatogonia or spermatocytes (small and large arrows, respectively). (C) Distribution of enkurin in mouse sperm. Specific enkurin labeling is observed in the acrosomal crescent and flagellar principal piece.

TRPC5 but not to control matrixes containing immobilized in HEK 293 cells (Fig. 7A). These interactions were specific MBP. as enkurin did not bind GST alone (Fig. 7A). Second, inspection of the deduced amino acid sequence of Binding was also observed between in vitro translated m-enkurin had revealed a candidate calmodulin-binding IQ enkurin and p85a and -h fusion proteins, demonstrating that motif (m-enkurin176–187; Fig. 3). The ability of m-enkurin to these interactions were direct. Binding was greatly reduced bind calmodulin was tested by incubating lysates of FLAG- when enkurin prolines at residues 83, 86, and 89 were m-enkurin-transfected HEK293 cells with calmodulin–agar- converted to alanine (Fig. 7B; DP1). These residues form ose. As shown in Fig. 6B, m-enkurin was bound to calmodulin the predicted ligand for the p85 SH3 domain (Obenauer et in a Ca2+-dependent manner. A peptide corresponding to al., 2003). In contrast, mutation of enkurin prolines 21 and m-enkurin174–191 ([NH2]-NYIQENLKKAAMKRLSDE- 22 to alanines failed to produce observable effects on p85 [COOH], predicted IQ sequence italicized) inhibited FLAG- binding: enkurin-DP2 mutants (proline 21/22 converted to m-enkurin/calmodulin–agarose association in a concentra- alanine) bound p85 as well as wild-type enkurin (data not tion-dependent fashion (Fig. 6C), suggesting that the IQ shown), and the concerted mutation of all 5 prolines did not sequence represents the CaM binding site. further reduce binding below levels observed for DP Finally, examination of the deduced sequence of m- and h- mutants alone (data not shown). enkurin reveals a proline-rich amino terminus that contains predicted ligand sequences for the SH3 domain of the p85 regulatory subunit of 1-phosphatidylinositol-3-kinase and Discussion other SH3 domain proteins (Obenauer et al., 2003). SH3 domain array screens also demonstrated that p85a and h bind In this study, we characterize enkurin, a novel TRPC to enkurin in vitro (data not shown). To confirm the binding protein. This protein was initially isolated due to an specificity of array data, fusion proteins were generated interaction with TRPC2, and consistent with this, it is consisting of GST coupled to the SH3 domains of p85a or synthesized at high levels in known sites of TRPC2 p85h. We observed that p85-GST formed immunoprecipit- expression and function, such as the testis and VNO able complexes with FLAG-enkurin that had been expressed (Jungnickel et al., 2001; Leypold et al., 2002; Liman et K.A. Sutton et al. / Developmental Biology 274 (2004) 426–435 433

al., 1999); and binding studies in yeast and in mammalian cells demonstrate that enkurin binds several TRPC proteins in addition to TRPC2, including TRPC1 and TRPC5 but not TRPC3 (Figs. 3 and 6). Enkurin then is a TRPC channel binding protein with functions that are not restricted to TRPC2. This conclusion is consistent with the presence of enkurin in the anterior sperm head, where three TRPC binding partners are also located, but not with the situation in the flagellar principal piece. In contrast, flagellar enkurin may reflect interactions with other TRPC proteins there or with additional and yet uncharacterized binding partners. Enkurin has an apparent modular structure. An N- terminal approximately 100 residue domain is proline-rich and contains predicted binding sites for SH3 domain proteins (Obenauer et al., 2003). In vitro binding studies indicate strong interactions with the p85 regulatory subunit of PI3 kinase and weaker interactions with several other SH3 domains. Binding to p85 was confirmed and the likely interaction site identified by mutagenesis experiments as prolines 83/86/89 within sequence PKKPAVP (mEnkurin83– 89). This sequence is conserved in h-enkurin. In addition, the other enkurin orthologs described here all contain putative p85 SH3 ligand sites (Obenauer et al., 2003). We are presently exploring the functional significance of enkurin’s interaction with p85 and other SH3 domain proteins. In contrast, a C-terminal domain (mEnkurin160–255) contains the TRPC2 binding site. Enkurin then has the anticipated characteristics of a scaffold protein that tethers cargo, such as SH3 domain proteins, to a subset of TRPC channels. Another feature of the primary sequence of enkurin is the presence in the C terminus of an IQ motif that binds Fig. 6. Enkurin binds to TRPC proteins and to calmodulin. (A) Enkurin calmodulin in a Ca2+-dependent fashion. Calmodulin regu- forms precipitable complexes with hTRPC1 and with mTRPC5. Left gel: Lysates of HEK293 cells transfected with FLAG-enkurin were incubated with HA-hTRPC1, precipitated with anti-HA, and probed with either anti- HA or anti-FLAG. Right gel: Lysates of HEK293 cells transfected with FLAG-enkurin were incubated with agarose-MBP-mTRPC5, sedimented, and bound FLAG-enkurin identified with anti-FLAG antibody. Controls included lysates incubated with agarose-MBP. (B) Enkurin binds calm- odulin in a Ca2+-dependent manner. Lysates were prepared from FLAG- enkurin-transfected HEK293 cells and incubated with calmodulin–agarose in buffer containing either 1 mM or approximately 10 nM Ca2+,as indicated. (C) Calmodulin binding to enkurin is inhibited by a peptide containing the putative IQ motif of enkurin (NH2-DNYIQENLK- KAAMKRLSDE-COOH). Extracts of FLAG-enkurin-transfected HEK293 cells were incubated with calmodulin–agarose in the absence (no addition, buffer addition) or presence (7.5 nM, 75 AM) of IQ peptide and calmodulin-bound enkurin visualized by chemiluminescence. HEK293 cell extract starting material (input) is shown by silver staining of total protein. Arrows indicate the predicted migration of HA-TRPC1 or HA- TRPC5 (A) and of FLAG-enkurin (A–C). Fig. 7. p85 binds to m-enkurin. (A) GST pulldown of FLAG-enkurin. Fusion proteins consisting of GST coupled to the SH3 domains of p85a or al., 1999; Menco et al., 2001; Stowers et al., 2002). The p85h were incubated with extracts of FLAG-enkurin-transfected HEK293 evidence that it also interacts with a number of other TRPC cells. Blots were probed with anti-FLAG. (B) Binding of wild-type and channels can be summarized as follows: enkurin is mutant enkurin to p85 SH3 domains. 35S-methionine-labeled FLAG– expressed in other mouse tissues, such as the lung (Fig. enkurin and a mutant enkurin (enkurin DP1), in which prolines 83, 86, and 89 were mutated to alanine (see Fig. 2), were produced by in vitro 4), where TRPC2 has not yet been detected (Tiruppathi et translation and used to assess binding to p85a-GST and p85h-GST. Control al., 2002); it is highly conserved in human (Figs. 2 and 4), incubations used template-free in vitro translation products. Data represent where TRPC2 is a pseudogene (Wes et al., 1995; Vannier et autoradiographic visualization of bound proteins. 434 K.A. Sutton et al. / Developmental Biology 274 (2004) 426–435 lation may be another conserved feature of enkurin: The IQ However, the observation that TRPC1 and TRPC5 proteins sequence is completely conserved in human (Fig. 2) and colocalize with TRPC2 in the acrosomal crescent region of predicted sites can be detected in other vertebrate orthologs sperm (Fig. 2), and that a portion of ZP3-evoked Ca2+ (http://calcium.uhnres.utoronto.ca/ctdb/pub_pages/resources response is resistant to the inhibition of TRPC2 channels /index.htm; chicken, residues 97–115, IQSKKNFIKTNAVA (Jungnickel et al., 2001), provides a plausible explanation. AITGL; zebrafish, residues 216–234, MKQLSGEERFKIL- According to this suggestion, ZP3 may activate TRPC2 as QALKKN) but not in the invertebrate sequences examined. well as other TRPC channels, and these other channel IQ motifs often bind apocalmodulin in a Ca2+-independ- proteins, notably TRPC1 and TRPC5, remain functional and ent manner. Yet, a Ca2+ requirement for the binding of may account for male fertility in the absence of functional calmodulin to IQ domains has previously been demon- TRPC2. In this regard, enkurin is conserved in human strated in the cases of IQ-GAP2, Ras GRF1, and Myr4 sperm (Fig. 2), where it may interact with TRPC1 and myosin (Bahler and Rhoads, 2002; Jurado et al., 1999). The TRPC5 in the acrosomal crescent (Castellano et al., 2003), role of calmodulin in enkurin function is intriguing as it and this suggests a conserved signal transduction pathway. places is places a Ca2+ sensor on a protein (enkurin) that is This situation differs from the case of the murine bound to a Ca2+-conducting ion channel. It may be vomeronasal organ, where TRPC2 is the only member of speculated that enkurin/cargo recruitment to TRPC channels the TRPC family expressed in the pheromone-responsive or release from those channel binding sites is controlled epithelium and may account for the phenotypic consequence locally by channel opening and the subsequent influx of of ablation of this gene (Leypold et al., 2002; Lucas et al., Ca2+. Thus, local activation of signaling molecules, which 2003; Stowers et al., 2002). may include but not be restricted to PI3 kinase, can be linked to TRPC channel conductance. It may be relevant to Acknowledgments note then that PI3 kinase is localized in both the sperm head and flagellum (Luconi et al., 2004; Nagdas et al., 2002) and We are grateful to Robert O’Connell for demonstration of has been implicated in the regulation of motility and the the art of VNO dissection; to Lutz Birnbaumer, David acrosome reaction (Fisher et al., 1998; Luconi et al., 2004; Clapham, and Hong-Sheng Li for TRPC cDNA clones; and Nagdas et al., 2002). TRPC channels exhibit both a similar to the NIH for support of this project (HD40310 and cellular localization and may also regulate similar physio- HD46948, to HMF). logical events in sperm (Castellano et al., 2003; Jungnickel et al., 2001; Trevino et al., 2001). We are presently investigating the hypothesis that enkurin provides a novel References link between TRPC channel and PI3 kinase signaling pathways. Arnoult, C., Cardullo, R.A., Lemos, J.R., Florman, H.M., 1996. Activation 2+ TRPC channels or related channels within the TRP of mouse sperm T-type Ca channels by adhesion to the egg zona pellucida. Proc. Natl. Acad. Sci. 93, 13004–13009. superfamily regulate sperm function in a wide range of Arnoult, C., Kazam, I.G., Visconti, P.E., Kopf, G.S., Villaz, M., Florman, animals, including mammals (Jungnickel et al., 2001), H.M., 1999. Control of the low voltage-activated calcium channel of Caenorhabditis elegans (Xu and Sternberg, 2003), Droso- mouse sperm by egg ZP3 and by membrane hyperpolarization during phila (Gao et al., 2003; Watnick et al., 2003), and sea urchin capacitation. Proc. Natl. Acad. Sci. 96, 6757–6762. (Mengerink et al., 2000, 2002; Neill et al., 2004; Vacquier Bahler, M., Rhoads, A., 2002. Calmodulin signaling via the IQ motif. FEBS Lett. 513, 107–113. and Moy, 2000). Localization of TRPC proteins to both the Birnbaumer, L., Yidirim, E., Abramowitz, J., 2003. A comparison of the head and flagellum is consistent with the participation of genes coding for canonical TRP channels and their M, Vand P relatives. 2+ these channels in such Ca -dependent processes as sperm Cell Calcium 33, 419–432. capacitation (Fraser, 1987; Visconti et al., 1995), the Bleil, J.D., Wassarman, P.M., 1983. Sperm-egg interactions in the mouse: acrosome reaction (Dan, 1954; Yanagimachi and Usui, sequence of events and induction of the acrosome reaction by a zona pellucida glycoprotein. Dev. Biol. 95, 317–324. 1974), and the control of flagellar motility (Ishijima and Castellano, L.E., Trevino, C.L., Rodriguez, D., Serrano, C.J., Pacheco, J., Hamaguchi, 1993). In particular, the inhibitory effects of a Tsutsumi, V., Felix, R., Darszon, A., 2003. Transient receptor potential TRPC2-specific function-blocking antibody have led to the (TRPC) channels in human sperm: expression, cellular localization and assignment to this channel of a role in mediating ZP3- involvement in the regulation of flagellar motility. FEBS Lett. 541, dependent Ca2+ influx during the initiation of mouse sperm 69–74. Clapham, D.E., 2003. TRP channels as cellular sensors. Nature 426, acrosome reactions (Jungnickel et al., 2001). This con- 517–524. clusion appears at variance with subsequent observations Dan, J.C., 1954. Studies on the acrosome. III. Effect of calcium deficiency. that TRPC2 has become a pseudogene in some primates, Biol. Bull. 107, 335–349. including humans (Vannier et al., 1999; Wes et al., 1995), Evans, J.P., Florman, H.M., 2002. The state of the union: the cell biology of old world monkeys, and apes (Liman and Innan, 2003), and fertilization. Nat. Cell Biol. 4 Suppl, s57–s63. Ferrell, J.E., Cimprich, K.A., 2003. Enforced proximity in the function of a possibly in bovine (Wissenbach et al., 1998), as well as with famous scaffold. Mol. Cell 11, 289–291. the persistent fertility of mice bearing targeted deletions of Fisher, H.M., Brewis, I.A., Barratt, C.L., Cooke, I.D., Moore, H.D., 1998. the trpc2 gene (Leypold et al., 2002; Stowers et al., 2002). Phosphoinositide 3-kinase is involved in the induction of the human K.A. Sutton et al. / Developmental Biology 274 (2004) 426–435 435

sperm acrosome reaction downstream of tyrosine phosphorylation. Mol. Montell, C., 2001. Physiology, phylogeny, and functions of the TRP Hum. Reprod. 4, 849–855. superfamily of cation channels. Sci. STKE 90 (RE1. 7-10-2001). Florman, H.M., Tombes, R.M., First, N.L., Babcock, D.F., 1989. An Nagdas, S.K., Winfrey, V.P., Olson, G.E., 2002. Identification of ras and its adhesion-associated agonist from the zona pellucida activates G protein- downstream signaling elements and their potential role in hamster promoted elevations of internal Ca and pH that mediate mammalian sperm motility. Biol. Reprod. 67, 1058–1066. sperm acrosomal exocytosis. Dev. Biol. 135, 133–146. Neill, A.T., Moy, G.W., Vacquier, V.D., 2004. Polycystin-2 associates with Fraser, L.R., 1987. Minimum and maximum extracellular Ca requirements the polycystin-1 homolog, suREJ3, and localizes to the acrosomal during mouse sperm capacitation and fertilization in vitro. J. Reprod. region of sea urchin spermatozoa. Mol. Reprod. Dev. 67, 472–477. Fertil. 81, 77–89. Obenauer, J.C., Cantley, L.C., Yaffe, M.B., 2003. Scansite 2.0: proteome- Fukami, K., Nakao, K., Inoue, T., Kataoka, Y., Kurokawa, M., Fissore, wide prediction of cell signaling interactions using short sequence R.A., Nakamura, K., Motoya, K., Mikoshiba, K., Yoshida, N., motifs. Nucleic Acids Res. 31, 3635–3641. Takenawa, T., 2001. Requirement of phospholipase Cy4 for the zona O’Toole, C.M.B., Arnoult, C., Darszon, A., Steinhardt, R.A., Florman, pellucida-induced acrosome reaction. Science 292, 920–923. H.M., 2000. Ca2+ entry through store-operated channels in mouse Fukami, K., Yoshida, M., Inoue, T., Kurokawa, M., Fissore, R.A., Yoshida, sperm is initiated by egg ZP3 and drives the acrosome reaction. Mol. N., Mikoshiba, K., Takenawa, T., 2003. Phospholipase Cy4 is required Biol. Cell 11, 1571–1584. for Ca2+ mobilization essential for acrosome reaction in sperm. J. Cell Pawson, T., 2003. Organization of cell-regulatory systems through Biol. 161, 79–88. modular-protein-interaction domains. Philos. Trans. Ser. A Math. Phys. Gao, Z., Ruden, D.M., Lu, X., 2003. PKD2 cation channel is required Eng. Sci. 361, 1251–1262. for directional sperm movement and male fertility. Curr. Biol. 13, Primakoff, P., Myles, D.G., 2002. Penetration, adhesion, and fusion in 2175–2178. mammalian sperm-egg interaction. Science 296, 2183–2185. Harteneck, C., 2003. Proteins modulating TRP channel function. Cell Putney, J.W., 2004. The enigmatic TRPCs: multifunctional cation channels. Calcium 33, 303–310. Trends Cell Biol. 14, 282–286. Harteneck, C., Plant, T.D., Schultz, G., 2000. From worm to man: three Stowers, L., Holy, T.E., Meister, M., Dulac, C., Koentges, G., 2002. Loss of subfamilies of TRP channels. Trends Neurosci. 23, 159–166. sex discrimination and male-male aggression in mice deficient for Hung, A.Y., Sheng, M., 2002. PDZ domains: structural modules for protein TRP2. Science 295, 1493–1500. complex assembly. J. Biol. Chem. 277, 5699–5702. Sutton, K.A., Vu, M.N., Wilkinson, M.F., 1998. Distal V beta promoters Ishijima, S., Hamaguchi, Y., 1993. Calcium ion regulation of chirality of transcribe novel T-cell receptor-beta transcripts in early development. beating flagellum of reactivated sea urchin spermatozoa. Biophys. J. 65, Immunology 93, 213–220. 1445–1448. Tiruppathi, C., Freichel, M., Vogel, S.M., Paria, B.C., Mehta, D., Flockerzi, Jungnickel, M.K., Marrero, H., Birnbaumer, L., Lemos, J.R., Florman, V., Malik, A.B., 2002. Impairment of store-operated Ca2+ entry in H.M., 2001. Trp2 regulates entry of Ca2+ into mouse sperm triggered by TRPC4(À/À) mice interferes with increase in lung microvascular egg ZP3. Nat. Cell Biol. 3, 499–502. permeability5. Circ. Res. 91, 70–76. Jurado, L.A., Chockalingam, P.S., Jarrett, H.W., 1999. Apocalmodulin. Trevino, C.L., Serrano, C.J., Beltran, C., Felix, R., Darszon, A., 2001. Physiol. Rev. 79, 661–682. Identification of mouse trp homologs and lipid rafts from spermatogenic Laemmli, U.K., 1970. Cleavage of structural proteins during assembly of cells and sperm. FEBS Lett. 509, 119–125. the head of bacteriophage T4. Nature 227, 680–685. Vacquier, V.D., Moy, G.M., 2000. The fucose sulfate polymer of egg jelly Leypold, B.G., Yu, C.R., Leinders-Zufall, T., Kim, M.M., Zufall, F., Axel, binds to sperm REJ and is the inducer of the sea urchin sperm acrosome R., 2002. Altered sexual and social behaviors in trp2 mutant mice. Proc. reaction. Dev. Biol. 192, 125–135. Natl. Acad. Sci. U. S. A. 99, 6376–6381. Vannier, B., Peyton, M., Boulay, G., Brown, D., Qin, N., Jiang, M., Zhu, Liman, E.R., Innan, H., 2003. Relaxed selective pressure on an essential X., Birnbaumer, L., 1999. Mouse trp2, the homologue of the human component of pheromone transduction in primate evolution. Proc. Natl. trpc2 pseudogene, encodes mTrp2, a store depletion-activated capaci- Acad. Sci. U. S. A. 100, 3328–3332. tative Ca2+ channel. Proc. Natl. Acad. Sci. 96, 2060–2064. Liman, E.R., Corey, D.P., Dulac, C., 1999. TRP2: a candidate transduction Visconti, P.E., Bailey, J.L., Moore, G.D., Pan, D., Olds-Clarke, P., Kopf, channel for mammalian pheromone sensory signaling. Proc. Natl. Acad. G.S., 1995. Capacitation of mouse spermatozoa. I. Correlation between Sci. 96, 5791–5796. the capacitation state and protein tyrosine phosphorylation. Develop- Lucas, P., Ukhanov, K., Leinders-Zufall, T., Zufall, F., 2003. A ment 121, 1129–1137. diacylglycerol-gated cation channel in vomeronasal neuron dendrites Watnick, T.J., Jin, Y., Matunis, E., Kernan, M.J., Montell, C., 2003. A is impaired in TRPC2 mutant mice: mechanism of pheromone trans- flagellar polycystin-2 homolog required for male fertility in Drosophila. duction. Neuron 40, 551–561. Curr. Biol. 13, 2179–2184. Luconi, M., Carloni, V., Marra, F., Ferruzzi, P., Forti, G., Baldi, E., 2004. Wes, P.D., Chevesich, J., Jeromin, A., Rosenberg, C., Stetten, G., Montell, Increased phosphorylation of AKAP by inhibition of phosphatidylinositol C., 1995. TRPC1, a human homolog of a Drosophila store-operated 3-kinase enhances human sperm motility through tail recruitment of channel. Proc. Natl. Acad. Sci. 92, 9652–9656. protein kinase A. J. Cell Sci. 117, 1235–1246. Wilde, M.W., Ward, C.R., Kopf, G.S., 1992. Activation of a G protein in Menco, B.P., Carr, V.M., Ezeh, P.I., Liman, E.R., Yankova, M.P., 2001. mouse sperm by the zona pellucida, an egg-associated extracellular Ultrastructural localization of G-proteins and the channel protein TRP2 matrix. Mol. Reprod. Dev. 31, 297–306. to microvilli of rat vomeronasal receptor cells. J. Comp. Neurol. 438, Wissenbach, U., Schroth, G., Phillipp, S., Flockerzi, V., 1998. Structure and 468–489. mRNA expression of a bovine trp homologue related to mammalian Mengerink, K.J., Moy, G.W., Vacquier, V.D., 2000. suREJ proteins: new trp2 transcripts. FEBS Lett. 429, 61–66. signalling molecules in sea urchin spermatozoa. Zygote 8 (Suppl 1), Xu, X.Z., Sternberg, P.W., 2003. A C. elegans sperm TRP protein required S28–S30. for sperm-egg interactions during fertilization. Cell 114, 285–297. Mengerink, K.J., Moy, G.M., Vacquier, V.D., 2002. suREJ3, a polycystin-1 Yanagimachi, R., Usui, N., 1974. Calcium dependence of the acrosome protein, is cleaved at the GPS domain and localizes to the acrosomal reaction and activation of guinea pig spermatozoa. Exp. Cell Res. 89, region of sea urchin sperm. J. Biol. Chem. 277, 943–948. 161–174. Minke, B., Cook, B., 2002. TRP channel proteins and signal transduction. Zhu, X., Jiang, M., Peyton, M., Boulay, G., Hurst, R., Stephani, E., Physiol. Rev. 82, 429–472. Birnbaumer, L., 1996. Trp, a novel mammalian gene family Montell, C., 2003. The venerable inveterate invertebrate TRP channels. Cell essential for agonist-activated capacitative Ca2+ entry. Cell 85, Calcium 33, 409–417. 661–671.