An X-linked Gene Encodes a Major Human Sperm Fibrous Sheath Protein, hAKAP82

Mammalian sperm motility is regulated by a cascade of cAMP-dependent protein phosphorylation events mediated by protein kinase A. A-kinaseanchor proteins (AKAPs) direct protein kinase A activity by tethering the enzyme near its physiological substrates. We have characterized a major human sperm fibrous sheath AKAP, hAKAP82, and its precursor, pro-hAKAP82, the homologues of the mouse fibrous sheath proteins mAKAP82 and pro-mAKAP82. The cDNA sequence of pro-hAKAP82 was highly homologous to the mouse sequence, and the functional domains of the pro-hAKAP82 protein, the protein kinase A binding, and the pro-hAKAP82/hAKAP82 cleavage sites were identical to those of the mouse protein. The genomic organization of mousepro-AKAP82 was determined. Alternative splicing occurred in both the mouse and human pro-AKAP82 genes that resulted in at least two distinct transcripts and possibly two different proteins. Compared with pro-mAKAP82, considerably less pro-hAKAP82 was processed to hAKAP82 in human sperm. Although pro-mAKAP82 localizes only to the proximal portion of the principal piece of the flagellum, pro-hAKAP82 localized to the entire length of the principal piece. The pro-hAKAP82 gene mapped to human chromosome Xp11.2, indicating that defects in this gene are maternally inherited. These studies suggest several roles for hAKAP82 in sperm motility, including the regulation of signal transduction pathways.

Eukaryotic flagellar assembly and the regulation of flagellar motility are complex processes. For example, in the blue-green alga Chlamydomonas reinhartii, more than 25 loci have been characterized that affect flagellar assembly, and over 52 loci have been identified that result in altered motility (1). In contrast to Chlamydomonas, only a few of the molecular defects underlying naturally occurring cases of human flagellar immo-tility have been characterized (2)(3)(4)(5). It is likely that the molecular basis of human sperm motility is more complicated than that of Chlamydomonas because, in addition to the axoneme, mammalian sperm have several flagellar accessory structures including a fibrous sheath (FS). 1 The FS is a structure found exclusively in the principal piece of the flagellum where it surrounds the outer dense fibers and axoneme. It is believed to play a structural role in sperm motility by restricting the plane of bending of the flagellum (6).
Although many of the target proteins have not been identified, it is well accepted that protein phosphorylation/dephosphorylation events are involved in the initiation and maintenance of mammalian sperm flagellar motility and that phosphorylation occurs via a cAMP-dependent pathway (7)(8)(9)(10). In mammalian sperm, the major downstream target of cAMP is protein kinase-A (PK-A), thus making it likely that this enzyme is involved in the regulation of sperm motility (11). Although typically soluble in somatic cells, PK-A also can be found tethered to subcellular organelles via binding of its regulatory (RII) subunit to A-kinase anchor proteins (AKAPs) (12)(13)(14). In response to cAMP binding to the RII subunit of PK-A, the catalytic subunit of the kinase is released and becomes free to catalyze phosphorylation of its substrates. By tethering PK-A close to certain substrates, AKAPs may play critical roles in determining the specificity of PK-A action (15)(16)(17).
We have cloned and characterized a cDNA encoding mouse AKAP82 (mAKAP82), the major protein of the sperm FS and a member of the AKAP family (11,18). Mouse AKAP82 is synthesized in the cell body of condensing spermatids as a M r 97,000 precursor (pro-mAKAP82, GenBank accession number U07423) (18). This precursor polypeptide is transported down the flagellum to the principal piece where it is processed by the proteolytic cleavage of the amino-terminal 179 amino acids to produce mAKAP82 and the free 179 amino acid pro domain (19). Coincident with or following cleavage, mAKAP82 is assembled into the FS. Mouse AKAP82 could tether PK-A close to the axoneme and other components of the flagellum that are involved in sperm motility, thus regulating the action of PK-A by directing its activity to specific motility-related targets. With the characterization of mAKAP82 there is now experimental evidence for a functional role of the FS as a mediator of PK-A activity.
Previous work from our laboratories has shown that the human homologue of mAKAP82, hAKAP82, localizes to the FS (20). Additionally, both hAKAP82 and its predicted precursor protein, pro-hAKAP82, bind the RII subunit of PK-A and are the major polypeptides of a limited subset of proteins that become tyrosine-phosphorylated in a time-dependent manner after incubation in a medium supporting capacitation. In human sperm, capacitation is associated with several cellular changes, such as the tyrosine phosphorylation of specific proteins (20 -22), that must take place before fertilization can occur. Capacitation also is associated with alterations in sperm motility patterns characteristic of sperm hyperactivation (23).
Because the evidence to date is supportive of both structural and functional roles for pro-mAKAP82 and mAKAP82 in mouse sperm motility, we hypothesize that pro-hAKAP82 and hAKAP82 play central roles in human sperm motility. The hypothesis is based on the observation that these proteins are phosphoproteins that are likely to be involved in the regulation of phosphorylation of other proteins in the sperm flagellum; such phosphorylation events are known to be critical for regulating motility (7-10, 18, 20). Support for this hypothesis requires further characterization of hAKAP82 and its precursor. In this paper, we report that pro-hAKAP82 and hAKAP82 were the human homologues of pro-mAKAP82 and mAKAP82. We provide evidence that two alternative transcripts of the gene were made during both human and mouse spermatogenesis because of the use of different donor/acceptor splice junctions. Additionally, both pro-hAKAP82 and hAKAP82 localized specifically to the entire length of the FS of ejaculated sperm. Finally, we show that the pro-hAKAP82 gene mapped to human Xp11.2, a finding that has significant implications for germ cell development and male infertility.

Isolation of Human and Mouse cDNA and Genomic Clones Encoding
Pro-AKAP82-To isolate a cDNA clone corresponding to pro-hAKAP82, a random-primed human testis cDNA library in a gt11 vector (CLON-TECH, Palo Alto, CA) was screened by filter hybridization with a radiolabeled 1.9-kilobase pair cDNA fragment representing the 5Ј end of the pro-mAKAP82 cDNA (24). After multiple rounds of screening, a 1.4-kilobase pair partial pro-hAKAP82 cDNA clone homologous to the 5Ј-UTR and the first 1222 bp of coding sequence of pro-mAKAP82 was isolated and subcloned into the EcoRI site of pGEM-3™ (Promega, Madison, WI).
A human testis Expressed Sequence Tag (EST) clone containing an approximately 750-bp insert, which was over 90% homologous to bases 1948 -2520 of the pro-mAKAP82 cDNA coding region and the pro-mAKAP82 3Ј-UTR, was identified by performing a BLAST (basic local alignment search tool (25)) search on the GenBank EST data base with the pro-mAKAP82 cDNA sequence. The human EST clone was obtained from the I.M.A.G.E. consortium (LLNL) (I.M.A.G.E. consortium clone 726916, Research Genetics, Inc. Huntsville, AL) (26,27).
To map the chromosomal location of the pro-hAKAP82 gene by fluorescence in situ hybridization (FISH), a genomic clone containing pro-hAKAP82 was isolated from a genomic human bacterial artificial chromosome (BAC) library (in vector pBeloBAC 11) by Research Genetics using a PCR-based technique. PCR primers were designed based on the cDNA sequence for pro-hAKAP82. Sequence analysis of the BAC clone showed that it contained pro-hAKAP82.
A genomic clone containing the 5Ј-flanking region of pro-mAKAP82 was isolated from a mouse genomic library in GEM™-11 (Promega) using filter hybridization with a radiolabeled PCR fragment from the mouse cDNA as a probe (24). A second genomic clone containing the full-length genomic sequence of pro-mAKAP82 together with its 5Ј-and 3Ј-flanking regions was isolated from a mouse P1 embryonic stem cell library using a PCR-based technique and primers derived from the pro-mAKAP82 cDNA sequence (Genome Systems Inc., St. Louis, MO).
Generation and Analysis of PCR Products-A 1.6-kilobase pair PCR product corresponding to the region of pro-hAKAP82 that was not included in the human cDNA or EST clones (homologous to bases 437 to 2092 of the pro-mAKAP82 cDNA) was amplified by PCR from a human testis cDNA library using primers based on the pro-hAKAP82 cDNA sequence (corresponding to bases 437 to 457 of the pro-hAKAP82 cDNA coding region) and the pro-hAKAP82 EST sequence (corresponding to bases 2097 to 2077 of the pro-hAKAP82 cDNA coding region).
To assay for the presence of an alternative splice variant of pro-hAKAP82, a DNA fragment was amplified by PCR from a human testis cDNA library using primers corresponding to a region in the 5Ј-UTR of mouse Fsc1 (GenBank accession number U10341 (28)) and to a region in the 5Ј end of the coding region of the pro-hAKAP82 cDNA. PCR products were purified with the Wizard™ PCR preps kit (Promega) before sequencing.
DNA Sequencing and Computer Analysis-All sequencing was done with the AmpliTaq®, FS dye terminator cycle sequencing kit chemistry or the BigDye™ terminator cycle sequencing kit chemistry and the appropriate primers using a 373A DNA sequencer (PE Applied Biosystems, Foster City, CA). Ambiguities were resolved by sequencing the opposite strand. DNA and protein sequence analyses were performed using the MacVector™ (Kodak Scientific Imaging Systems, New Haven, CT) and Sequencher™ (Gene Codes Corp., Ann Arbor, MI) software programs.
Preparation of Sperm and Sperm Proteins-Samples of human semen were obtained by masturbation from normal, healthy donors with good sperm motility (total sperm motility greater than 75%, progressive motility greater than 60%). Ejaculate volume, percentage of total and progressively motile sperm, and sperm concentration were determined for each ejaculate. Sperm were washed 3ϫ in PBS. Before final centrifugation, the volume, sperm concentration, and total sperm numbers again were determined for each sample. After the final wash, sperm pellets were dissolved in SDS sample buffer containing 40 mM dithiothreitol and boiled for 5 min. The amount of protein in each sample was determined by the Amido Black procedure (29).
Immunological Reagents-An antiserum against the predicted pro domain of pro-hAKAP82 (anti-hpro) was prepared as follows. A peptide corresponding to residues 131-145 (NH 2 -VGDTEGDYHRASSEN-COOH, the hpro peptide) of the pro-hAKAP82 protein was synthesized with a cysteine added to the amino terminus (Quality Control Biochemicals, Hopkinton, MA). The peptide was conjugated to keyhole limpet hemocyanin via the cysteine and then used for immunization. The antiserum was characterized by immunoblotting using protein extracts from ejaculated human sperm. A portion of the antiserum was affinitypurified, eluted, neutralized, and dialyzed in PBS.
Immunoblot Analysis of Sperm Proteins-Proteins from ejaculated human sperm were separated under reducing conditions by SDS-polyacrylamide gel electrophoresis on a 10% (w/v) gel and electrophoretically transferred to nitrocellulose membranes. Equal amounts of protein were analyzed in each lane. The blots were blocked, probed with anti-hpro (1:2000 (v/v)), processed, and developed using an ECL kit (Amersham Pharmacia Biotech) as described previously (19) before being exposed to Reflection™ film (NEN Life Science Products). For the preabsorption experiments, the antiserum was incubated with the hpro peptide (1 mg/ml) in PBS containing 0.1% (v/v) Tween 20 and 3% (w/v) bovine serum albumin for 1 h at room temperature and then used to probe the immunoblots as described above.
Indirect Immunofluorescence of Sperm-One-ml aliquots of sperm diluted in PBS to 3 ϫ 10 6 cells/ml were permeabilized in 0.1% (v/v) Triton X-100 for 15 min and then washed once in PBS. Pellets were resuspended in 1 ml of PBS, and the cells were transferred onto coverslips. After settling, cells were fixed in 4% (w/v) paraformaldehyde, incubated in Ϫ20°C methanol, washed, and blocked in normal goat serum as described previously (19). Sperm then were incubated in anti-hpro diluted 1:10 (v/v) in 10% goat serum overnight at 4°C and washed in PBS. Sperm were incubated for 1 h at 37°C in the secondary antibody (fluorescein isothiocyanate-conjugated goat anti-rabbit IgG, Jackson Immunoresearch Laboratories, Inc., West Grove, PA), diluted 1:50 (v/v) in 10% goat serum, and again washed in PBS before being mounted on slides with mounting media (Fluoromount-G, Southern Biotechnology Associates Inc., Birmingham. AL). Slides were viewed with a Zeiss Photomicroscope III equipped with epifluorescence. Photographs were taken with Kodak T-Max film, 3200 ASA. Paired sample and control photographs were exposed for equal amounts of time.
Chromosomal Mapping of Pro-hAKAP82-Localization of the pro-hAKAP82 gene to a single chromosome band was accomplished using FISH mapping. The entire human genomic BAC clone containing an insert of approximately 100 kilobase pairs, including pro-hAKAP82, was labeled with digoxigenin-11-dUTP (Boehringer Mannheim) by nick translation and hybridized to metaphase chromosome spreads prepared from peripheral blood lymphocytes from a normal man. Labeled probe (300 ng) was incubated overnight at 37°C with 3 g of human cot-1 DNA (Amersham Pharmacia Biotech) in Hybrisol VII (Oncor Inc., Gaithersburg, MD). The probe then was denatured at 72°C for 5 min and pre-annealed for 30 min at 37°C. Slides were dehydrated and denatured before hybridizing overnight at 37°C with the probe in a humidified chamber. Slides were washed in 50% formamide,1ϫ SSC (0.15 M NaCl and 0.015 M sodium citrate) for 10 min and then twice in 2ϫ SSC for 4 min each. All washes were done at 40°C with gentle shaking. Slides were transferred to a solution of 1ϫ phosphate-buffered detergent (Oncor). Detection was performed with rhodamine-labeled anti-digoxigenin antibody, and chromosomes were counterstained with diamidinophenylindole. Metaphase chromosome spreads were visualized using a Zeiss universal microscope with a Photometrics™ cooled-CCD camera and Quips™ Imaging Software (Vysis™ Inc., Downers Grove, IL). Twenty metaphase spreads with signals on both chromatids at the same band position were used to determine chromosomal location.

RESULTS
Pro-hAKAP82 Is the Human Homologue of pro-mAKAP82-We determined the full-length sequence of the pro-hAKAP82 cDNA by aligning and sequencing the pro-hAKAP82 cDNA clone, PCR product, and EST clone (see "Experimental Procedures"; GenBank accession number AF072756). The composite cDNA sequence contained an initiator methionine with an in-frame, upstream stop codon. The coding region of pro-hAKAP82 was 2535 bases long and contained one open reading frame, which predicted a protein of 845 amino acids and concluded with an in-frame stop codon. Although no consensus polyadenylation signals (AATAAA) were present upstream of the putative poly(A) tract, two less well conserved potential polyadenylation signals (AATACA) were present at bases 2807 and 2827. The predicted molecular weights for the various forms of the human protein were 93,500 for pro-hAKAP82 and, assuming that the cleavage site in pro-hAKAP82 is the same as in pro-mAKAP82, 73,272 for hAKAP82 and 20,244 for the pro domain (hpro).
Two cDNA sequences, pro-mAKAP82 and Fsc1, have been reported for the major mouse FS protein (18,28). The predicted amino acid sequence of pro-hAKAP82 was highly homologous to both of these proteins. Specifically, at least 79% of the amino acids were identical, and 91% were conserved between the mouse (both pro-mAKAP82 and Fsc1) and human sequences. Critical functional domains that have been defined previously within the mouse protein; specifically, the RII binding region (FYVNRLSSLVIQMA) and the pro-mAKAP82/mAKAP82 cleavage site (KNTNNNQSPS) (11,18) were identical in the human homologue.
Both the Mouse and Human Genes for pro-AKAP82 Are Alternatively Spliced-The protein coding regions and 3Ј-UTRs of the pro-mAKAP82 and Fsc1 cDNAs are essentially identical; however, the 5Ј-UTRs of the two sequences share no homology. Furthermore, the Fsc1 5Ј-UTR contains an alternative inframe start codon (27 bases upstream of the start codon for pro-mAKAP82) that could result in a protein containing an additional 9 amino acids at its amino terminus. These observations suggested that the two transcripts arise from alternative splicing of the same gene. To determine whether this was the case, genomic clones of pro-mAKAP82 and its 5Ј-and 3Јflanking regions were isolated and sequenced. The 5Ј-UTRs of both pro-mAKAP82 and Fsc1 were present in a single genomic clone and were separated from each other by 585 bp (Figs. 1, A  and B), indicating that the two cDNA clones (pro-mAKAP82 and Fsc1) resulted from alternative splicing of the pro-mAKAP82 gene. The coding region of pro-mAKAP82 contained 5 exons and 4 introns with consensus splice donor/acceptor sites present at most exon/intron boundaries (Table I; Gen-Bank accession numbers AF087516 and AF087517). An over-view of the structure of pro-mAKAP82 is shown in Fig. 1A The 5Ј-UTR of the pro-hAKAP82 cDNA was highly homologous to the 5Ј-UTR of pro-mAKAP82 but shared no homology with the 5Ј-UTR of Fsc1. To determine whether the pro-hAKAP82 gene, like the mouse homologue, was alternatively spliced, we used a PCR-based approach to search for a human cDNA sequence homologous to the Fsc1 5Ј-UTR. Using primers corresponding to regions in the 5Ј-UTR of Fsc1 and the 5Ј end of the coding region of the pro-hAKAP82 cDNA, an approximately 400-bp product (5Ј-UTR f ) was amplified from a human testis cDNA library. The sequence of 5Ј-UTR f was highly homologous to the 5Ј-UTR/5Ј end of the coding region of Fsc1 (67% identical bases, Fig. 1B). The 5Ј-UTR f sequence, like the sequence of Fsc1, contained an alternative in-frame start codon 27 bp upstream of the start codon for pro-hAKAP82, which could result in a protein containing an additional 9 amino acids at its amino terminus compared with pro-hAKAP82. Five of these 9 deduced amino acids were identical to the 9 predicted additional amino acids of Fsc1. This finding is strong support for the concept that, like the mouse, there are at least two alternative splice variants of the pro-hAKAP82 gene, one with a 5Ј-UTR homologous to pro-mAKAP82 and one with a 5Ј-UTR homologous to Fsc1.
Pro-hAKAP82 and the Free pro Domain Localize to the Entire Length of the Principal Piece of Human Sperm-A polyclonal antibody generated against the mature mAKAP82 protein recognizes two bands in ejaculated human sperm; one at M r 82,000 (hAKAP82) and another at M r 97,000 (20). These findings suggest that, like the mouse, hAKAP82 is formed by proteolytic cleavage of a higher M r precursor, pro-hAKAP82. An antibody (anti-hpro) raised against a peptide sequence in the predicted processed (hpro) region of pro-hAKAP82 (a region presumed to be absent from mature hAKAP82) was used to probe immunoblots of ejaculated human sperm protein. Antihpro identified a polypeptide at M r 97,000, which is pro-hAKAP82 ( Fig. 2A). In addition, a M r 18,000 protein was detected. This protein approximates the size predicted for the hpro domain of pro-hAKAP82 (M r 20,244) and indicated that some of this fragment persisted in mature sperm. As expected, because the hpro peptide has been removed from the mature protein, no band was recognized at M r 82,000 (hAKAP82). Preabsorption of anti-hpro with the hpro peptide abolished the immunoreactivity, demonstrating that the antiserum reacted specifically with the hpro sequence. The sizes of hAKAP82 (20), pro-hAKAP82, and hpro are consistent with the hypothesis that the pro-hAKAP82/hAKAP82 cleavage site is similar, if not identical, to that of the mouse.
Immunoreactivity was seen along the entire length of the principal piece when human sperm were probed with anti-hpro (Fig. 2, B and C). All sperm labeled in a similar fashion. This result is in contrast to the findings in mature cauda epididymal mouse sperm in which pro-mAKAP82 and mouse pro are found only in the proximal portion of the principal piece (19). Antihpro was specific for the FS as it reacted exclusively with the principal piece of the flagellum. The antiserum reacted specifically with the hpro sequence as immunoreactivity was not seen in samples in which anti-hpro was preabsorbed with the hpro peptide. Additionally, no staining was seen with the preimmune serum.
The pro-hAKAP82 Gene Maps to Xp11.2-To map the chromosomal location of the pro-hAKAP82 gene and to determine whether the gene is a candidate for any previously mapped human genetic diseases involving male infertility, we analyzed chromosomes from a normal man by FISH using a digoxigenin-11-dUTP-labeled BAC genomic clone of pro-hAKAP82. Results showed that pro-hAKAP82 mapped to Xp11.2, adjacent to the  1. A, schematic diagram of the structure of the pro-mAKAP82 gene and its flanking regions. Plain lines denote the locations and relative sizes of the introns. The dashed line represents a region that has not yet been sequenced. Numbered boxes denote the locations and relative sizes of the exons. Exons 2 through 6 (black boxes) are common to both the Fsc1 and pro-mAKAP82 cDNAs. The 5Ј-UTRs of the cDNAs encoding Fsc1 (exon 1Ј, shaded box) and pro-mAKAP82 (exon 1Љ, white box) are indicated. The relative positions of the potential Fsc1 start codon and the pro-mAKAP82 start codon are shown. The region demarcated by the brackets is detailed in B. B, both pro-mAKAP82 and pro-hAKAP82 are alternatively spliced. The relevant sequence of the 5Ј-flanking region of the mouse genomic clone is listed on the top line as genomic. Aligned beneath it are the 5Ј-UTRs of Fsc1, pro-mAKAP82 (designated mAKAP82) and pro-hAKAP82 (designated hAKAP82), and the sequence of the human PCR product (5ЈUTR f ). Shaded regions mark identical bases. The sequence of the 5Ј-UTR of Fsc1 was separated from the sequence of the 5Ј-UTR of pro-mAKAP82 by a 585-bp intron (sequence not shown). Note that the Fsc1/5ЈUTR f sequences both contained an alternative in-frame, upstream ATG (boxed). The numbers beside the Fsc1, pro-mAKAP82, and pro-hAKAP82 cDNA sequences correspond to the bases of the respective sequences as they appear in the GenBank. hAKAP82 in Human Sperm centromere (Fig. 3). This region of the human X chromosome is syntenic to the most proximal end of the mouse X chromosome, the location of the pro-mAKAP82 gene (30). DISCUSSION The high homology of the amino acid sequences of pro-hAKAP82, pro-mAKAP82/Fsc1, and a 75-kDa rat FS protein (76% identical to and 89% conserved with pro-hAKAP82 at the amino acid level) indicates that the structure and function of AKAP82 is highly conserved in sperm of a number of mammalian species. Of particular interest is the finding that the amino acid sequence of the RII binding site of pro-mAKAP82 (11) is identical in pro-hAKAP82 (14 of 14 identical amino acids) and is highly conserved in the 75-kDa rat FS protein (13 of 14 identical amino acids). Furthermore, when human sperm proteins were probed with radiolabeled RII, prominent bands were identified at M r 97,000 and M r 82,000 that corresponded to pro-hAKAP82 and hAKAP82, respectively (20). The binding was eliminated by preincubation of the RII subunit with a synthetic RII-binding peptide (11) corresponding to the sequence of the pro-mAKAP82 RII binding site but not by preincubation of the RII subunit with a peptide containing a scrambled version of the sequence of the RII-binding peptide (data not shown). These results confirmed that pro-hAKAP82 and hAKAP82 are AKAPs and that the RII binding site was conserved between the mouse and human proteins. The ability of pro-hAKAP82 and hAKAP82 to bind the RII subunit of PK-A together with the importance of protein phosphorylation events for sperm motility makes it likely that hAKAP82 anchors PK-A to the sperm FS and directs the signal transduction pathways that control human sperm motility.
Although the ability to bind the RII subunit of PK-A has historically been used to identify and define AKAPs, recent evidence indicates that AKAPs function as scaffolding proteins for a variety of signal transducing molecules (31-35). Thus, multifunctional kinases and phosphatases can be targeted to FIG. 3. A, FISH of normal human metaphase chromosomes with a digoxigenin-labeled genomic BAC clone containing pro-hAKAP82 detected with rhodamine-labeled anti-digoxigenin (red dots). The red dots localized the gene to that part of the X chromosome adjacent to the centromere. Note that only one chromosome was labeled as the cell was from a normal male. B, banding pattern of the same metaphase spread shows that the red markers localized to Xp11.2. C, schematic diagram of the human X chromosome showing the location of band 11.2 adjacent to the centromere. An enlargement of the labeled X chromosome is provided beside the schematic for clearer identification of the banding pattern. FIG. 2. A, immunoblot analysis of proteins from ejaculated human sperm with anti-hpro. Immunoblots were probed with anti-hpro, anti-hpro preabsorbed with the peptide against which it was generated, or the preimmune serum. Anti-hpro immunoreacted with pro-hAKAP82 (M r 97,000) and hpro (M r 18,000). As expected, anti-hpro did not cross-react with hAKAP82, because hAKAP82 does not contain the pro domain. Immunoreactivity was abolished by preincubation of the antiserum with the peptide against which it was generated and was not present in the preimmune serum. B, immunofluorescence and (C) the corresponding phase contrast images of ejaculated human sperm labeled with anti-hpro. Punctate immunoreactivity was detected throughout the entire length of the principal piece but not in the sperm head or in the midpiece or endpiece of the sperm flagellum. Arrows denote the junction of the principal piece and the midpiece. Labeling of human ejaculated sperm incubated with anti-hpro that was preabsorbed with the peptide against which it was generated showed no immunoreactivity (data not shown). subcellular locations via anchoring to a common AKAP. The highly polarized structure of the flagellum together with a paucity of cytoplasm in sperm lends itself to the idea of a scaffolding protein serving to sequester typically soluble proteins. In the future, it will be important to determine whether AKAP82 functions as a scaffold for other proteins involved in sperm motility.
In addition to the ability of pro-hAKAP82 and hAKAP82 to bind RII, the processing of pro-hAKAP82 into hAKAP82 represents another possible mechanism through which these proteins may be involved in the regulation of sperm motility. Compared with mouse sperm (18), human sperm contained a relatively large amount of pro-hAKAP82 in addition to hAKAP82 (20). Also, although pro-mAKAP82 is located only in the proximal principal piece of mouse cauda epididymal sperm, reactivity to the anti-hpro antibody persisted throughout the entire length of the principal piece of mature human sperm. These data indicated that the processing of pro-hAKAP82 to hAKAP82 is different from, and possibly less efficient than, the processing of pro-mAKAP82 to mAKAP82 (19). One hypothesis is that a decrease in processing of pro-hAKAP82 to hAKAP82 might be associated with sperm with poor motility. However, anti-hpro appeared to immunoreact similarly against pro-hAKAP82 in all sperm, suggesting that pro-hAKAP82 is not associated preferentially with subpopulations of immotile or poorly motile sperm in normal human ejaculates. Immunofluorescence experiments in this report were performed on ejaculates obtained from normal, fertile donors, and all samples had very high percentages of motile sperm. Thus, differences in the efficiency of processing may be found in patients with more dramatic reductions in sperm motility.
Taken together, the data on mAKAP82 and hAKAP82 suggest several potential ways in which pro-hAKAP82 and hAKAP82 could be associated with sperm motility. First, these proteins, by their ability to bind the RII subunit of PK-A, may function to direct signal transduction pathways by channeling the activity of the kinase. Other studies have suggested that the interaction of RII with sperm AKAPs may itself be important for the regulation of sperm motility independent of the activity of the catalytic subunit of PK-A (36). Second, the processing of pro-hAKAP82 into hAKAP82 may be associated with alterations in flagellar motility. Third, the capacitationdependent tyrosine phosphorylation of pro-hAKAP82 and hAKAP82 (20) may regulate an undefined function important to sperm motility. Fourth, pro-hAKAP82/hAKAP82 may also be important to the cytoskeletal integrity of the FS. And finally, the persistence of the free pro domain in both mature mouse and human sperm suggests that the pro domain itself may have some as yet unknown function in the FS. Future studies will focus on the potential association of alterations in processing or phosphorylation of pro-hAKAP82 and hAKAP82 with changes in human sperm motility.
Alternative splicing of the pro-AKAP82 gene resulted in the generation of two transcripts in both the mouse (pro-mAKAP82 and Fsc1) and the human (pro-hAKAP82 and the transcript represented by 5Ј-UTR f ). The Fsc1 and 5Ј-UTR f sequences predict proteins that contain 9 more amino acids at their amino termini than do the pro-mAKAP82 and pro-hAKAP82 proteins. Recently, a human testis-specific cDNA (termed hi) encoding a predicted protein sequence that is 92% identical to and 95% conserved with that of pro-hAKAP82 was reported (37). The 5Ј-UTR of hi is 97% identical to the sequence of 5Ј-UTR f , indicating that the hi sequence encodes the alternative splice variant of pro-hAKAP82. Surprisingly, and with no direct experimental evidence, the predicted hi protein is described as a sperm surface glycoprotein.
Alternative splicing has been reported in other AKAP genes including the Drosophila AKAP, DAKAP550, and another human/mouse male germ cell AKAP, S-AKAP84 (38,39). In mouse AKAP-KL, alternative splicing together with the use of alternative translation initiation codons, results in six different protein isoforms (40). The expression of these isoforms is differentially regulated and results in different isoforms predominating in different tissue types. These data suggest that the various isoforms of AKAP-KL may play different physiologic roles. At this time, we do not know the significance of the two alternative splice products of pro-hAKAP82 or if both of the two alternative forms of the protein even exist. However, conservation of the alternative splice event between mouse and human lends support to the hypothesis that each of the two pro-AKAP82 alternative splice products are functionally significant.
We previously showed that the gene for mouse pro-AKAP82 maps to the proximal end of the X chromosome (30). In the current study, the human gene for pro-AKAP82 mapped to the region of the human X chromosome (Xp11.2) that is syntenic to the proximal X chromosome of the mouse. In the mouse, pro-mAKAP82 is expressed post-meiotically in late spermiogenesis (18,28). During this haploid phase, the X and Y chromosomes are generally considered inactive. This means that, in the mouse, at least some regions of the X chromosome are actively transcribed even in late spermiogenesis. In addition, because the cells are haploid at the time that the transcript first appears, yet the protein is present in all cells, mRNA and/or protein for mAKAP82 must be shared between conjoined spermatids via their intercellular bridges. Our data suggest that a similar event may occur in humans.
Although several genes and genetic diseases have been mapped to Xp11.2, the reported phenotypes are not suggestive of a defect in a sperm FS protein. Thus, clinically significant mutations occurring in the hAKAP82 gene have not yet been mapped or genetically characterized. Because pro-hAKAP82 is on the X chromosome, defects in the gene in males would necessarily be maternally inherited. If men with mutations in pro-hAKAP82 were sterile, then the defect would be self-limiting and thus would be rare. Candidates for a defect in this gene are men with dysplasia of the fibrous sheath. Affected individuals have more than 95% immotile sperm. The sperm tails are short and thick, and the FS is disorganized (41). It will be important to determine whether or not hAKAP82 is involved in this or any other form of male factor infertility associated with reduced sperm motility.