Two Functional Forms of ACRBP/sp32 Are Produced by Pre-mRNA Alternative Splicing in the Mouse¹

Abstract

ACRBP/sp32 is a binding protein specific for the precursor (pro-ACR) and intermediate forms of sperm serine protease ACR. In this study, we examined the expression pattern, localization, and possible role of mouse ACRBP in spermatogenic cells and epididymal sperm. Unlike other mammalian ACRBPs, two forms of Acrbp mRNA—wild-type Acrbp-W and variant Acrbp-V5 mRNAs—were generated by alternative splicing of Acrbp in the mouse. ACRBP-W was synthesized in pachytene spermatocytes and haploid spermatids and immediately processed into a mature protein, ACRBP-C, by removal of the N-terminal half. The intron 5-retaining splice variant mRNA produced a predominant form of ACRBP, ACRBP-V5, that was present in pachytene spermatocytes and round spermatids, but was absent in elongating spermatids. ACRBP-W and ACRBP-V5 were both colocalized with pro-ACR in the acrosomal granules of early round spermatids, whereas the sperm acrosome contained only ACRBP-C. Glutathione S-transferase pull-down assays revealed that ACRBP-V5 and ACRBP-C possess a different domain capable of binding each of two segments in the C-terminal region of pro-ACR. Moreover, autoactivation of pro-ACR was remarkably accelerated by the presence of ACRBP-C. These results suggest that ACRBP-V5 and ACRBP-C may function in the transport/packaging of pro-ACR into acrosomal granules during spermiogenesis and in the promotion of ACR release from the acrosome during acrosomal exocytosis, respectively.

Introduction

Spermiogenesis is a specialized process of differentiation of postmeiotic germ cells, spermatids, that involves extensive morphological remodeling of spermatids, including formation of two main sperm-specific structures, the acrosome and the tail [1–3]. The acrosome overlying the anterior part of sperm head originates from the Golgi apparatus in haploid spermatids; Golgi-derived small vesicles containing a variety of hydrolytic enzymes are transported into the acrosomal sac and fuse with each other to form a large acrosomal granule associated with the nuclear envelope. As spermiogenesis proceeds, the acrosome is enlarged and spread over the nucleus. In fertilization, following sperm binding to the zona pellucida—an extracellular matrix surrounding the oocyte—the sperm is thought to undergo the acrosome reaction, a fusion event between the outer acrosomal and plasma membranes [1, 4]. As a consequence of the exocytotic event, acrosomal components are released and interact with the zona pellucida to enable motile sperm to penetrate the oocyte coat. The acrosome reaction also generates a new status of the sperm head to achieve sperm/oocyte fusion. Indeed, only acrosome-reacted sperm are capable of fusing with the oocyte plasma membrane [1].

An acrosomal protein, ACRBP/sp32, is a binding protein specific for precursor and intermediate forms of serine protease ACR/acrosin [5, 6]. Porcine ACRBP is initially synthesized as a 61-kDa precursor protein in spermatogenic cells, and the 32-kDa mature ACRBP is posttranslationally produced by removal of the N-terminal half of the precursor during spermatogenesis and/or epididymal maturation of sperm. Because ACRBP is colocalized with the ACR precursor, termed pro-ACR, in the sperm acrosome, ACRBP has been postulated to participate in the release of ACR from the acrosomal matrix [6, 7]. In the human, pig, and mouse, ACRBP is known to be tyrosine phosphorylated during sperm capacitation [8–10], leading to the possibility that the partitioning of acrosomal components in the soluble, matrix, and acrosomal membrane compartments of the acrosome may be regulated by the phosphorylation status of ACRBP [9]. Moreover, ACRBP has been identified as a member of the cancer/testis antigen family; ACRBP is normally expressed only in the testis, but it is also expressed in a wide range of different tumor types, including bladder, breast, liver, and lung carcinomas [11–13]. Despite the importance of ACRBP in male gamete and cancer cells, the function of ACRBP has been poorly understood thus far.

In this study, to facilitate further studies of ACRBP, we have characterized mouse ACRBP and found that in addition to wild-type Acrbp (Acrbp-W) mRNA (GenBank accession number NM_016845), an alternatively spliced variant (Acrbp-V5; GenBank accession number NM_001127340) of Acrbp-W mRNA, which partly retains the fifth intron, is expressed in mouse spermatogenic cells, unlike porcine, guinea pig, and human ACRBP mRNAs. The mature form of ACRBP (ACRBP-C) is posttranslationally produced from ACRBP-W during spermatogenesis, whereas epididymal sperm contain no full-length ACRBP-V5. On the basis of these data, differential functions of ACRBP-W, ACRBP-V5, and ACRBP-C in acrosome biogenesis and acrosomal exocytosis are discussed.

Materials and Methods

All animal experiments were performed ethically, and experimentation was in accordance with the Guide for the Care and Use of Laboratory Animals at University of Tsukuba.

Isolation of cDNA and Genomic Clones

Screening of a mouse testis cDNA library in λgt11 [14] and a mouse 129/SvJ genomic DNA library in Charon 28 was carried out by the plaque hybridization method [15] using a cDNA fragment encoding porcine or mouse ACRBP, as described previously [16]. Positive clones were plaque purified, and the cDNA inserts were introduced into pUC19 at appropriate restriction sites for sequence analysis.

Northern Blot Analysis

Total cellular RNAs were extracted from various tissues of ICR mice (SLC, Shizuoka, Japan) using Isogen (Nippon Gene, Toyama, Japan), as described previously [17]. The RNA samples were glyoxylated, separated by electrophoresis on 1.2% agarose gels, and transferred onto Hybond-N⁺ nylon membranes (GE Healthcare, Piscataway, NJ). Hybridization was carried out as described previously [17].

Plasmids

DNA fragments encoding ACRBP-W, ACRBP-V5, and pro-ACR were amplified by PCR using the following sets of oligonucleotide primers: 5′-AAGAATTCGAGGAATCTCCAGCCTCCAC-3′ and 5′-AACTCGAGTCATCCAAACTGGCCTATAGC-3′ for Acrbp-W; 5′-AAGAATTCGAGGAATCTCCAGCCTCCAC-3′ and 5′-TACTCGAGTCACAATTTCCTGTACCTGCC-3′ for Acrbp-V5; and 5′-ATGAATTCGGCCCTAACGCCTTGCA-3′ and 5′-GTCTCGAGTCAGGAATGGAGG AAGGGC-3′ for pro-Acr. The amplified fragments were introduced into pET-32a (Invitrogen, Carlsbad, CA), pGEX-4T-1 (GE Healthcare), and pMAL-cRI (New Englamd Biolabs, Ipswich, England) to produce His-tagged, glutathione S-transferase (GST)-fusion, and maltose-binding protein (MBP)-fusion proteins, respectively.

Antibodies

A DNA fragment encoding an N-terminally truncated ACRBP-W mutant containing the 228-residue sequence at positions 313–540 (Supplemental Fig. S1A; all Supplemental Data are available online at www.biolreprod.org) was introduced into pET-32a, and the His-tagged recombinant protein was produced in Escherichia coli BL21 (DE3), as described previously [18]. A seven-residue synthetic peptide, Cys-Thr-Gly-Arg-Tyr-Arg-Lys-Leu, corresponding to the C-terminal sequence of ACRBP-V5, was coupled to maleimide-activated keyhole limpet hemocyanin, according to the manufacturer's protocol (Thermo Scientific, Rockford, IL). The recombinant protein and peptide/hemocyanin conjugate were used as antigens to raise antibodies against ACRBP-C and ACRBP-V5, respectively. The antigenic proteins were emulsified by sonication with Freund complete or incomplete adjuvant (Difco Laboratories, Detroit, MI) and were injected intradermally into female New Zealand white rabbits (SLC). The ACRBP-C and ACRBP-V5 antibodies were purified by fractionation with ammonium sulfate (0%–40% saturation), followed by immunoaffinity chromatography on a Sepharose 4B column that had been conjugated with a GST-fused ACRBP-C mutant protein carrying the C-terminal 228-residue sequence (Supplemental Fig. S1A) or an MBP-fused ACRBP-V5 mutant protein carrying the C-terminal 44-residue sequence (Supplemental Fig. S1B), as described previously [19, 20]. Affinity-purified rabbit antibody against mouse pro-ACR was prepared as described previously [21]. Affinity-purified rabbit anti-His-Tag, mouse monoclonal anti-ACTB (β-actin) clone AC-15 (ascites fluid), and affinity-purified goat anti-PRSS21 (ESP-1/testisin/TESP5) antibodies were purchased from Medical and Biological Laboratories (Nagoya, Japan), Sigma-Aldrich (St. Louis, MO), and Santa Cruz Biotechnology (Santa Cruz, CA), respectively. Horseradish peroxidase-conjugated antibodies against mouse, goat, or rabbit immunoglobulin G (IgG; H + L) were purchased from Jackson Immunoresearch Laboratories (West Grove, PA). Alexa Fluor 488-conjugated antibody against rabbit IgG was purchased from Molecular Probes (Eugene, OR).

Preparation of Protein Extracts

Pachytene spermatocytes, round spermatids, and elongating spermatids were purified from seminiferous tubules of ICR mice (3–5 mo old), as described previously [17]. Testicular tissues were homogenized at 4°C in a lysis buffer consisting of 20 mM Tris-HCl, pH 7.4; 0.15 M NaCl; 1% Triton X-100; 0.5 mM dithiothreitol, 0.5 mM phenylmethanesulfonyl fluoride; 1 μg/ml pepstatin A; 1 μg/ml leupeptin; and 75 U/ml aprotinin, using a Potter-Elvehjem glass homogenizer (Iuchi Co., Osaka, Japan) fitted with a Teflon pestle at a constant speed of 1000 rpm (10 strokes). Spermatogenic cells and epididymal sperm were lysed in the same lysis buffer by pipetting at 4°C. The protein extracts were centrifuged at 11 000 × g for 10 min at 4°C. The supernatant solution was used as a source of protein extracts. Protein concentration was determined using a Coomassie protein assay reagent kit (Thermo Scientific).

Immunoblot Analysis

Proteins were separated by SDS-PAGE and transferred onto Immobilon-P polyvinylidene difluoride membranes (Millipore, Bedford, MA). The blots were blocked with 4% skim milk in 20 mM Tris-HCl, pH 7.5, containing 0.1% Tween-20 and 0.15 M NaCl at room temperature for 1 h, incubated with primary antibodies at room temperature for 1 h, and then treated with secondary antibodies conjugated with horseradish peroxidase at room temperature for 1 h. The immunoreactive proteins were visualized by using an enhanced chemiluminescence kit (GE Healthcare), followed by exposure to X-ray films (Fuji Photo Film C., Tokyo, Japan). The dilution ratios of primary antibodies used were as follows: 1:200, 1:1000, 1:100, 1:1000, 1:200, and 1:20 000 for anti-ACRBP-C (original protein concentration, 0.08 mg/ml), anti-ACRBP-V5 (0.1 mg/ml), anti-pro-ACR (0.1 mg/ml), anti-PRSS21 (0.2 mg/ml), anti-His-Tag (0.8 mg/ml), and anti-ACTB (14 mg/ml) antibodies, respectively.

Immunostaining Analysis

Fresh cauda epididymal sperm of ICR mice were dispersed in a 0.2-ml drop of TYH medium [22] free of bovine serum albumin. Spermatogenic cells and epididymal sperm were transferred into a 1.5-ml microtube, washed with PBS by centrifugation at 3000 rpm for 5 min, fixed in 4% paraformaldehyde/PBS solution, pH 7.2, on ice for 15 min, washed with cold PBS, and then treated with 0.1% Triton X-100 in PBS at room temperature for 15 min. The fixed cells were blocked with 3% normal goat serum in PBS containing 0.05% Tween-20 on ice for 30 min, washed with the same buffer, incubated with anti-ACRBP-C, anti-ACRBP-V5, or anti-pro-ACR antibodies for 60 min, washed, and then reacted with Alexa Fluor 488-conjugated anti-rabbit IgG antibodies for 60 min. After washing with PBS, sperm cells were incubated with Alexa Fluor 568-conjugated peanut lectin PNA (3 μg/ml) and Hoechst 33342 (2.5 μg/ml) for 30 min, washed with PBS, mounted, and then observed under an IX-71 fluorescence microscope (Olympus, Tokyo, Japan), as described previously [23]. For testicular sections, mouse tissues were snap frozen and embedded in a TissueTek O.C.T. compound (Sakura Finetechnical, Tokyo, Japan). Sections (10 μm) were prepared in a Leica CM3000 cryostat (Wetzlar, Germany). After blocking with 3% normal goat serum in PBS containing 0.05% Tween-20, these samples were treated with primary antibody, reacted with Alexa Flour 488-conjugated anti-rabbit IgG antibody, incubated with Alexa Fluor 568-labeled PNA and Hoechst 33342, and viewed under the fluorescence microscope, as described above.

GST Pull-Down Assay

Pull-down assays were carried out as described previously [24]. Briefly, GST-fused recombinant proteins were mixed with His-tagged proteins in 20 mM Tris-HCl, pH 7.4, containing 0.3 M NaCl, 0.5% Nonidet P-40, 0.5 mM dithiothreitol, 0.5 mM phenylmethanesulfonyl fluoride, 1 μg/ml pepstatin A, 1 μg/ml leupeptin, and 75 U/ml aprotinin. The mixture was incubated with glutathione agarose beads that had been treated with 4% bovine serum albumin at 4°C for 1 h. After washing thoroughly with the above buffer, proteins on the agarose beads were boiled with the Laemmli buffer and analyzed by immunoblotting, as described previously [24].

Results

We carried out screening of a mouse testis cDNA library using a DNA fragment encoding porcine ACRBP as a probe. Nine positive clones were identified and divided into two groups by the lengths of cDNA inserts. One group contained approximately 1.8-kb cDNA inserts encoding ACRBP-W, whereas almost 1.3-kb inserts that coded for ACRBP-V5 were found in another group. The combined nucleotide sequences of cDNA clones encoding ACRBP-W and ACRBP-V5 were 1841 and 1253 nucleotides in length, respectively (Supplemental Fig. S1). The ACRBP-V5 sequence in the 5′-untranslated and protein-coding regions was identical to the corresponding sequence of ACRBP-W, although the 3′-end sequences carrying a part of the protein-coding region and the entire 3′-untranslated region were totally different from each other. ACRBP-W and ACRBP-V5 contain 540 and 316 amino acid residues, respectively (Fig. 1A). Mouse ACRBP-W shares high degrees of sequence identities (74% and 72%) with the porcine and guinea pig proteins, respectively. On the basis of the sequence homology [6, 11], mouse ACRBP-W is probably converted by proteolytic processing to a mature protein (ACRBP-C) of 268 residues during spermiogenesis or epididymal sperm maturation (Fig. 1A). The sequence of ACRBP-V5 is different from that of ACRBP-W only in the C-terminal region; ACRBP-V5 possesses a five-residue sequence, Arg-Tyr-Arg-Lys-Leu, divergent from the C-terminal 229-residue sequence of ACRBP-W (Fig. 1A and Supplemental Fig. S1).

Fig. 1

Alternative splicing of Acrbp mRNA. A) Schematic representation of porcine ACRBP and mouse ACRBP-W, ACRBP-V5, and ACRBP-C. The common and alternative regions among these three mouse proteins are shown by black and gray boxes, respectively. Hatched and open boxes indicate the EQ-rich region and the C-terminal five-residue sequence of ACRBP-V5, respectively. B) Exon/intron organization of Acrbp. Acrbp is approximately 14.0 kb in length and consists of 10 exons interrupted by 9 introns. The exons common to Acrbp-W and Acrbp-V5 mRNAs, exons specific for Acrbp-V5 mRNA, and exons specific for Acrbp-W mRNA are indicated by black, open, and gray boxes, respectively. C) Nucleotide and amino acid sequences between the fifth and sixth exons in Acrbp. The intron sequence specific for Acrbp-V5 mRNA is shown by an open box. Putative polyadenylation signal is underlined.

Database search and sequencing analysis of genomic clones coding for mouse Acrbp indicated that Acrbp is approximately 14.0 kb in length and consists of 10 exons interrupted by 9 introns (Fig. 1B). Importantly, the cDNA sequence of ACRBP-V5 at nucleotides 1023 to 1253 (Supplemental Fig. S1) completely matched the 5′-end, 231-nucleotide sequence of intron 5 in Acrbp (Fig. 1C). These data suggest that Acrbp-V5 mRNA is an alternatively spliced variant of Acrbp mRNA that partly retains intron 5.

To examine the expression pattern of Acrbp, we carried out Northern blot analysis of total RNAs from various tissues and spermatogenic cells using DNA fragments specific for both ACRBP-W and ACRBP-V5 (P1), ACRBP-W (P2), and ACRBP-V5 (P3) as probes (Fig. 2A). The P1 probe was hybridized to two mRNAs with the sizes of 2.4 and 1.7 kb and corresponding to Acrbp-W and Acrbp-V5 mRNAs, respectively, only in the testis (Fig. 2B). These two mRNA signals were first detectable in the testicular tissues at 18 days after birth (Fig. 2C). Densitometric analysis indicated that the level of Acrbp-V5 mRNA is approximately 3-fold higher than that of Acrbp-W mRNA. When Acrbp expression in spermatogenic cells was examined using the P2 and P3 probes, Acrbp-W and Acrbp-V5 mRNAs were most abundantly present in round spermatids (Fig. 2D). The sizes of these two mRNAs in round spermatids were approximately 200 nucleotides longer than those in pachytene spermatocytes. The increase in the mRNA size is due to the elongated chain length of poly(A) tail, as described previously [17]. Since only a single 2.2-kb ACRBP transcript is found in the testis of porcine and guinea pig [6], mouse Acrbp is distinguished from the porcine and guinea pig genes by the presence of the alternatively spliced variant.

Fig. 2

Expression of ACRBP-W and ACRBP-V5. A) Schematic representation of Acrbp-W and Acrbp-V5 mRNAs. The regions common to Acrbp-W and Acrbp-V5 mRNAs, specific for Acrbp-V5 mRNA, and specific for Acrbp-W mRNA are indicated by black, open, and gray boxes, respectively. Northern blot analysis was performed using DNA fragments P1, P2, and P3 as probes. B) Distribution of Acrbp-W and Acrbp-V5 mRNAs in mouse tissues. Total cellular RNAs of mouse testis (T), brain (B), submaxillary gland (Su), lung (Lu), heart (H), liver (Li), spleen (S), kidney (K), ovary (O), and uterus (U) were analyzed by Northern blotting. C) Expression of Acrbp-W and Acrbp-V5 mRNAs during testicular development. Total cellular RNAs of mouse testicular tissues were examined. D) Expression of Acrbp-W and Acrbp-V5 mRNAs in spermatogenic cells. Total cellular RNAs of pachytene spermatocytes (P), round spermatids (R), and elongating spermatids (E) were examined. E) Immunoblot analysis. Protein extracts of spermatogenic cells were separated by SDS-PAGE, blotted, and probed using anti-ACRBP-C, anti-ACRBP-V5, and anti-ACTB antibodies. F) Presence of ACRBP-W/ACRBP-C and ACRBP-V5 in ACR-deficient (ACR KO) and PRSS21-deficient (PRSS21 KO) mice. Protein extracts of testicular tissues and epididymal sperm were analyzed by immunoblotting.

To ascertain whether ACRBP-W and ACRBP-V5 are indeed present in spermatogenic cells and epididymal sperm, we prepared affinity-purified antibodies against ACRBP-C and ACRBP-V5. The specificity of the antibodies was validated by immunoblot analysis using recombinant GST-fused proteins (Supplemental Fig. S2). ACRBP-W was found as a 60/55-kDa doublet in all spermatogenic cells examined (Fig. 2E). A 30-kDa mature protein (ACRBP-C) was produced by the processing of ACRBP-W, and the amount of ACRBP-C increased as spermatogenesis proceeded. ACRBP-V5 was also detected as a 48/43-kDa doublet in pachytene spermatocytes and round spermatids (Fig. 2E). Only a negligibly low level of ACRBP-V5 was observed in elongating spermatids. Moreover, epididymal sperm of wild-type, ACR-deficient, and PRSS21-deficient mice exclusively contained ACRBP-C, although a 28-kDa protein probably corresponding to an additionally processed form was abundantly present in ACR-deficient sperm (Fig. 2F). No sperm protein was immunoreactive with anti-ACRBP-V5 antibody. These results suggest that ACRBP-V5 and ACRBP-C may play distinct roles mainly in spermatogenesis and fertilization, respectively, and that other protease(s) except for ACR and PRSS21 may be responsible for conversion of ACRBP-W into ACRBP-C during spermatogenesis. It is not clear at present whether the 60/55-kDa and 48/43-kDa doublets are produced as ACRBP-W and ACRBP-V5, respectively, by additional alternative splicing of pre-mRNA, proteolytic processing, or posttranslational modification, including glycosylation and phosphorylation.

We next examined the localization of ACRBP-W (ACRBP-C) and ACRBP-V5 in spermatogenic cells and epididymal sperm (Fig. 3 and Supplemental Fig. S3). Cells were costained with anti-ACRBP-C or anti-ACRBP-V5 antibody, Hoechst 33342, and Alexa Fluor 568-conjugated PNA that binds to the outer acrosomal membrane [25]. As described above, acrosomal components derived from the Golgi apparatus form a spherical acrosomal granule in the acrosomal sac of early round spermatids [2]. Immunostaining analysis indicated that both ACRBP-W and pro-ACR are highly enriched in acrosomal granules of round spermatids, and they spread over the apical end of the nucleus in elongating spermatids. ACRBP-V5 was colocalized with ACRBP-W and pro-ACR in the acrosomal granules of round spermatids, whereas no signal of ACRBP-V5 was found at a detectable level in elongating spermatids (Fig. 3, A and B). As expected, ACRBP-C and pro-ACR were colocalized in the acrosome of epididymal sperm (Fig. 3C). The localization of ACRBP-C was essentially similar between wild-type and ACR-deficient sperm, suggesting that pro-ACR may be dispensable, at least in part, for the precise localization of ACRBP-C in the sperm acrosome.

Fig. 3

Subcellular localization of ACRBP-W and ACRBP-V5. A and B) Immunostaining analysis of spermatogenic cells. Round spermatids (A) and elongating spermatids (B) at steps 5 and 14, respectively, were immunostained using anti-ACRBP-C, anti-ACRBP-V5, and anti-pro-ACR antibodies (green). The acrosome and nucleus were also stained with Alexa Fluor 568-conjugated peanut lectin PNA (red) and Hoechst 33342 (blue), respectively. C) Localization of ACRBP-W/ACRBP-C in epididymal sperm. Wild-type and ACR-deficient (ACR KO) sperm were immunostained. Bars = 2 μm.

As described above, ACRBP-C is capable of binding the precursor (pro-ACR) and intermediate forms of ACR; ACRBP-C fails to interact with mature ACR [6]. We thus postulate that the C-terminal domain of pro-ACR, termed DIII and consisting of three segments (SI, SII, and SIII; see Fig. 4A and Supplemental Fig. S4), may be involved in binding to ACRBP-C, because mature ACR lacks DIII, which is removed during pro-ACR maturation [14]. To ascertain this possibility, we carried out GST pull-down assays using His-tagged recombinant proteins of pro-ACR, Del-0 (wild-type), Del-1, Del-2, and Del-3, containing different regions of DIII (Fig. 4A and Supplemental Fig. S4). GST-fused ACRBP-W was bound to Del-0, Del-1, and Del-2 (Fig. 4B), whereas only Del-0 exhibited the ability to bind GST-ACRBP-V5 (Fig. 4C). These data suggest that the SIII of pro-ACR may be required for binding to ACRBP-V5. Because the ACRBP-V5 sequence largely overlaps with the N-terminal sequence of ACRBP-W (Fig. 1A and Supplemental Fig. S1), it is reasonable to consider that the N-terminal half of ACRBP-W also participates in the binding interaction with SIII. Moreover, the C-terminal half of ACRBP-W corresponding to ACRBP-C may be involved in binding to the precursor and intermediate forms of ACR, mainly through the interaction with pro-ACR SI.

Fig. 4

Binding of pro-ACR to ACRBP-W and ACRBP-V5. A) Schematic representation of the C-terminal domain, DIII, of pro-ACR. DIII consists of three segments—SI, SII, and SIII—that are removed during pro-ACR maturation; hence, the mature ACR contains no DIII [14]. B and C) GST pull-down assays. His-tagged recombinant proteins of pro-ACR were subjected to pull-down assays using GST-fused proteins of ACRBP-W (B) or ACRBP-V5 (C). Proteins bound to gel beads were analyzed by immunoblotting using anti-His-Tag and anti-GST antibodies. Asterisks indicate partially extended recombinant proteins in Del-0 and Del-1.

To identify the region of ACRBP necessary for binding to His-tagged pro-ACR Del-0, 12 GST recombinant proteins of ACRBP were prepared and analyzed by pull-down assays (Fig. 5A). Consistent with the above data (Fig. 4), ACRBP-W, ACRBP-V5, and ACRBP-C were all associated with Del-0 (Fig. 5B). Of three ACRBP-W mutants, Mut-1, Mut-2, and Mut-3, the latter two proteins were capable of binding Del-0, suggesting that ACRBP-W may contain at least two different domains required for binding to pro-ACR. When the other six mutant proteins, Muts 4 through 9, were examined, only Mut-7 and Mut-9 failed to bind Del-0. Thus, the N-terminal domains (B1 and B2) of ACRBP-V5 and ACRBP-C at residues 26–105 and 316–424, respectively, function as the binding sites for pro-ACR.

Fig. 5

ACRBP contains two domains required for binding to pro-ACR. A) Schematic representation of ACRBP-W, ACRBP-V5, ACRBP-C, and nine ACRBP mutants (Mut-1 through Mut-9). Two binding domains, B1 and B2, and an EQ-rich region of ACRBP are indicated by gray and hatched boxes, respectively. B) GST pull-down assays. His-tagged recombinant protein of pro-ACR Del-0 (Fig. 4) was subjected to pull-down assays using GST-fused ACRBP proteins. Proteins bound to gel beads were analyzed by immunoblotting using anti-His-Tag antibody.

We previously reported that porcine ACRBP-C accelerates autoactivation of pro-ACR in vitro [5, 6]. ACRBP-C also affects the maturation pathway of pro-ACR; although 55- and 53-kDa pro-ACRs are converted by autoactivation into 35-kDa ACR via a 43-kDa intermediate form of ACR, a 49-kDa intermediate, instead of the 43-kDa intermediate, is accumulated with increasing concentrations of exogenously added ACRBP-C during pro-ACR maturation [5, 6]. As shown in Figure 6, in addition to the 43-kDa intermediate, the 49-kDa form was produced from pro-ACR in the presence of recombinant ACRBP-C or ACRBP-W. Also, pro-ACR autoactivation was obviously accelerated by addition of either one of these two recombinant proteins. These data suggest that mouse ACRBP-C indeed affects autoactivation of pro-ACR in vitro.

Fig. 6

Maturation of pro-ACR in the presence of ACRBP-C and ACRBP-W. Porcine pro-ACR (a mixture of 55- and 53-kDa proteins; 2 μg) was autoactivated in a mixture (80 μl) containing 0.1 M Tris-HCl, pH 8.5, and GST-fused recombinant ACRBP-C (2 μg) or ACRBP-W (3 μg) at room temperature. Aliquots (12 μl each) were taken from the incubation mixture 3, 5, 15, 30, and 60 min after addition of pro-ACR, immediately mixed with a Laemmli SDS buffer (6 μl) to terminate pro-ACR autoactivation, and then subjected to immunoblot analysis using anti-porcine ACR antibody [14].

Discussion

This study describes the expression pattern, localization, and possible role of ACRBP in spermatogenic cells and sperm of the mouse. Two forms of Acrbp mRNA, Acrbp-W and Acrbp-V5 mRNAs, are produced by alternative splicing of a single mouse Acrbp (Figs. 1 and 2). Similar to porcine and guinea pig ACRBPs [6], mouse ACRBP-W is synthesized in pachytene spermatocytes and postmeiotic, haploid spermatids, and it is immediately processed into the mature protein ACRBP-C during spermatogenesis (Fig. 2). An intron 5-retaining splice variant of mouse Acrbp mRNA is also translated into ACRBP-V5 that is present in pachytene spermatocytes and round spermatids but absent in elongating spermatids, probably because of proteolytic degradation (Figs. 1 and 2). The expression of Acrbp-V5 mRNA is approximately 3-fold higher than that of Acrbp-W mRNA in spermatogenic cells (Fig. 2). Thus, two forms of ACRBP—ACRBP-V5 and ACRBP-C—are posttranscriptionally and posttranslationally produced in the mouse, although porcine and guinea pig ACRBPs function as a single protein [6].

Both ACRBP-W and ACRBP-V5 are colocalized with pro-ACR in the acrosomal granules of early round spermatids, whereas only ACRBP-W/ACRBP-C spreads over the acrosome in elongating spermatids and epididymal sperm (Fig. 3). ACRBP-W is mostly converted into ACRBP-C in elongating spermatids that contain no ACRBP-V5 (Figs. 2 and 3). If ACRBP-V5 is proteolytically degraded in elongating spermatids, the N-terminal half of ACRBP-W may also be proteolyzed after conversion into ACRBP-C, because the sequence of ACRBP-V5 is highly identical to that of the N-terminal half of ACRBP-W (Fig. 1 and Supplemental Fig. S1). Moreover, Acrbp-V5 mRNA is the predominant form of Acrbp mRNA in spermatogenic cells (Fig. 2). Thus, ACRBP-V5 may be mainly implicated in the transport/packaging of pro-ACR into acrosomal granules in early round spermatids. ACRBP-W may also participate partly in the pro-ACR transport as an auxiliary protein.

Mice lacking proprotein convertase subtilisin/kexin type 4, PCSK4 (PC4/SPC5) [26], have been previously reported to exhibit severe male subfertility [27]. In addition to abnormal sperm function [27, 28], the loss of PCSK4 resulted in the failure to convert ACRBP-W into ACRBP-C, despite the absence of consensus sequences for the enzymatic cleavage site [29]. Importantly, PCSK4-deficient testicular and epididymal sperm displayed aberrant shape of the acrosome; acrosomal components, including unprocessed ACRBP-W, were insufficiently spread over the acrosomal matrix, but rather were localized as acrosomal granule-like punctate structures near the apical region of the acrosome [29]. Because our data indicate that the processing of ACRBP-W into ACRBP-C rapidly occurs in postmeiotic spermatogenic cells (Fig. 2), we speculate that ACRBP-C may play an important role in the spread of acrosomal components, including pro-ACR, through the protein-protein interactions.

Our present data demonstrate that two sperm serine proteases, ACR and PRSS21, as well as PCSK4 [29], do not act as a processing enzyme toward ACRBP-W (Fig. 2). Given the phenotype of PCSK4-deficient mice [27], the ACRBP-W processing is probably catalyzed by an unknown acrosomal protease that is converted into the enzymatically active form by PCSK4. Unexpectedly, ACRBP-V5, which is produced by alternative splicing of Acrbp mRNA (Fig. 1), contains a very strong consensus sequence, Arg-Tyr-Arg-Lys at residues 346–349, for PCSK4 cleavage (Supplemental Fig. S1B). It is thus most likely that ACRBP-V5 is proteolytically processed by PCSK4 in the acrosomal granules of early round spermatids after the transport of pro-ACR. This probability seems to be supported by the previous finding that PCSK4 is present in acrosomal granules of round spermatids, in the acrosomal ridges of elongated spermatids, and on the plasma membrane overlying the acrosome of sperm [28]. Indeed, the signals immunoreactive with anti-ACRBP-V5 antibody are found only in the acrosomal granules (Fig. 3).

Mouse ACRBP possesses two binding sites, B1 and B2, for the C-terminal extension of pro-ACR (Figs. 4 and 5). B1, which is located in the N-terminal region of ACRBP-W and ACRBP-V5, is capable of binding the C-terminal segment SIII of pro-ACR. The binding property suggests that ACRBP-W and ACRBP-V5 associate with only a newly synthesized, enzymatically inactive pro-ACR through the B1-SIII interaction. This possibility is compatible with the role of ACRBP-V5 in the transport of pro-ACR into acrosomal granules, as described above. Another binding site, B2, near the N-terminal region of ACRBP-C binds Del-0, Del-1, and Del-2 mutant proteins of pro-ACR (Figs. 4 and 5), suggesting that ACRBP-C B2 is required for SI of pro-ACR as a minimum size of the binding site. Although SII is apparently dispensable for binding to ACRBP-C, mouse pro-ACR contains a 15-residue sequence in the regions of SI and SII that is highly conserved among the pro-ACRs of other mammalian species (Supplemental Fig. S4). Thus, at present it is conceivable that the 15-residue sequence of pro-ACR functions as a binding site for ACRBP-C. Consistent with our previous results [5, 6], a 49-kDa intermediate form of ACR is produced from pro-ACR by the addition of ACRBP-C during pro-ACR maturation (Fig. 6). The formation of the 49-kDa ACR intermediate may be explained by the binding of ACRBP-C to the 15-residue segment of pro-ACR; the ACRBP-C binding may block the cleavage site for conversion of the 49-kDa intermediate to the 43-kDa form by ACR itself.

Collectively, we have revealed that in the mouse, spermatogenic cells express an intron 5-retaining mRNA coding for ACRBP-V5 that may play an important role in the transport/packaging of pro-ACR into acrosomal granules during spermiogenesis. Although full-length ACRBP-W probably acts as an auxiliary protein on the pro-ACR transport, the major role of ACRBP-W is to exist as a precursor protein of ACRBP-C that serves to promote ACR release from the acrosomal matrix during the acrosome reaction. Thus, unlike porcine and guinea pig ACRBPs [6], alternative splicing of Acrbp may result in the functional differentiation of mouse ACRBP in spermiogenesis and fertilization.

Acknowledgment

We thank Drs. Chong Zhou and Woojin Kang for ACR-deficient and PRSS21-deficient mice, and Dr. Keiji Tanimoto for a mouse genomic DNA library.

Author notes

¹

Supported in part by grants from the Japan Society for the Promotion of Science (JSPS) and the Ministry of Education, Culture, Sports, Science and Technology in Japan (MEXT) to Y.K. and T.B.

2

These authors contributed equally to this work.