Two novel odorant receptor families expressed in spermatids undergo 5'-splicing.

We report the identification of two novel families of odorant receptor (OdR)-like proteins, termed spermatid chemoreceptors (SCRs), in rat spermatids of the testis. The full-length genomic clones encode seven transmembrane domain receptors that share 35-40% identity with certain OdRs and are among the most divergent members of the OdR superfamily based on phylogenetic analysis. RNase protection assays and in situ hybridization studies confirmed the expression of SCRs in spermatids, the post-meiotic, differentiating cell population in the testis. SCR transcripts were undetectable in the prepubertal testis but were readily identified in spermatids of sexually maturing and mature testis. Rapid amplification of cDNA end-polymerase chain reaction and genomic clone sequencing led to the discovery that SCRs are spliced upstream of their presumptive starting methionines. 5'-Splicing of OdRs may regulate the expression of functional chemoreceptors.

We report the identification of two novel families of odorant receptor (OdR)-like proteins, termed spermatid chemoreceptors (SCRs), in rat spermatids of the testis. The full-length genomic clones encode seven transmembrane domain receptors that share 35-40% identity with certain OdRs and are among the most divergent members of the OdR superfamily based on phylogenetic analysis. RNase protection assays and in situ hybridization studies confirmed the expression of SCRs in spermatids, the post-meiotic, differentiating cell population in the testis. SCR transcripts were undetectable in the prepubertal testis but were readily identified in spermatids of sexually maturing and mature testis. Rapid amplification of cDNA end-polymerase chain reaction and genomic clone sequencing led to the discovery that SCRs are spliced upstream of their presumptive starting methionines. 5-Splicing of OdRs may regulate the expression of functional chemoreceptors.
In an effort to understand the molecular basis of olfaction, Buck and Axel (1) cloned a family of putative odorant receptors (OdRs) 1 predicted to contain up to 1000 distinct gene products (2). The number and diversity of putative receptors is consistent with the ability of mammals to differentiate thousands of odors (2). Some of these receptors have been discretely localized to olfactory neurons by in situ hybridization (3)(4)(5)(6). Despite the compelling circumstantial evidence that this family of receptors functions in mammalian odor discrimination, definitive data documenting ligand-receptor binding and functional receptor expression are lacking. The possibility that these orphan receptors may serve a more general function in chemodetection comes from the discovery that OdR-like proteins are expressed in the testis (7). Subsequently, we (8) and others (9) demonstrated the localization of putative OdRs in spermatids of the testis and on the midpiece of mature spermatozoa. Moreover, mammalian sperm tails possess proteins involved in olfactory signal transduction including G-protein receptor kinase 3 (8), ␤-arrestin2 (8), inositol 1,4,5-trisphosphate receptors (10), adenylate cyclase (11), and cyclic nucleotide-gated channels (12). 2 In this report, RNase protection assays likewise detect OdR transcripts in the spleen. These findings suggest that sperm tails, olfactory cilia, and immune cells may share common mechanisms to detect exogenous chemical signals.
The hypothesis that OdRs function in mammalian sperm to regulate motility in response to exogenous signals derives from the existence of sperm-egg chemotaxis in invertebrates. The small peptides, speract and resact, are secreted by sea urchin eggs and attract spermatozoa in a species-specific manner by stimulating sperm motility and respiration (13)(14)(15). Recent evidence indicates that sperm-egg chemotaxis may likewise occur during mammalian fertilization (16,17).
We were interested in clarifying the role of OdR expression in the testis. During the meiotic phase of spermatogenesis, gene promoters may become exposed as a result of altered chromatin structure, potentially leading to spurious transcription of a host of gene products lacking physiologic function. To avoid this confounding issue in studying whole testis, we restricted our investigation of OdR expression to round spermatids, the post-meiotic, differentiating cell population in the testis. We report the identification of two novel families of OdR-like proteins, termed spermatid chemoreceptors (SCRs), in spermatids of rat testis. In situ hybridization studies illustrate the selective localization of SCRs to spermatids in the testis and demonstrate that SCR expression is developmentally regulated, with transcripts first detectable during sexual maturation. RACE-PCR and sequencing of genomic clones led to the discovery that SCRs undergo 5Ј-splicing, which may regulate receptor expression during spermatogenesis. The 5Ј-splicing of OdR transcripts may influence the stability of OdR mRNAs and the expression of functional OdR proteins in olfactory, reproductive, immunologic, and experimental systems.

EXPERIMENTAL PROCEDURES
RT-PCR of Putative SCRs-Total RNA was prepared from testis, whole brain, and olfactory tissue using TRIzol reagant (Life Technologies, Inc.) and Oligotex poly(A) ϩ selection (Qiagen) according to manufacturer protocols. Round spermatids were purified from the testis by elutriation (18); the cell pellet was homogenized in guanidinium isothiocyanate and RNA pelleted through a cesium chloride gradient according to established procedures (19). RNA was treated with 10 units of RNase-free DNase (Life Technologies, Inc.) per g of RNA. Five g of RNA from each source was reverse-transcribed using oligo(dT) or ran-dom primers. PCR was performed on a 1/20 aliquot of the RT mixture for 35 cycles at 94°C for 1 min, 50°C for 1 min, and 72°C for 1 min using TAQ DNA polymerase (Life Technologies, Inc.).
Degenerate primers to the conserved LHTPMYLF (OdR 1) and STCASH (OdR 2) sequences of originally cloned OdRs (1) were employed. Buck and Axel clone F2 was used as a positive control for RT-PCR with OdR 1 ϩ 2; negative control PCR reactions were run with mock reverse-transcribed RNA that was not treated with reverse transcriptase. The spermatid PCR product was TA-subcloned into pCRII (Invitrogen) and 14 clones (A-N) were sequenced. The testis PCR product was likewise subcloned and 7 clones were sequenced.
Genomic Cloning of SCRs-Two sets of PCR primers were designed for pSCRs D and G to generate probes corresponding to the two halves of each PCR product. PCR primer sequences were as follows: D.1 (EIG-YTC, 5Ј-GAGATTGGCTATACTTGC); D.2 (GMGCTV, 3Ј-TACTGTGC-ATCCCATTCC); D.3 (GLGQTN, 5Ј-GGTCTGGGACAGACCAAT); D.4 (GRHKAL, 3Ј-AAGAGCTTTGTGGCGCCC); G.1 (MFSSVE, 5Ј-ATGTT-CTCCTCTGTGGAG); G.2 (LGFLFM, 3Ј-CATGAAGAGGAACCCCAA); G.3 (ASPVVM, 5Ј-GCCAGTCCAGTTGTGATG); and G.4 (SARERQ, 3Ј-CTGTCGCTCCCTGGCTGA). Each half of the corresponding pSCR was amplified by PCR, gel-purified using QIAEX reagants (Life Technologies, Inc.), and labeled using a nick translation kit (Boehringer Mannheim). The probes were used to screen 2.5 ϫ 10 6 phage plaques of an EMBL3 rat genomic library (CLONTECH) according to established procedures (20). To identify SCR D clones, duplicate lifts were taken, and one lift was screened with the D.1-D.2 probe and the other with the D.3-D.4 probe. Fifty-two co-hybridizing plaques were identified, and 10 plaques were purified by two additional rounds of screening. The same technique was used to identify SCR G clones. For the SCR G screen, 221 co-hybridizing plaques were identified, and 10 plaques were purified by two additional rounds of screening. Purified phage were used to infect K803 cells (CLONTECH), and phage DNA were isolated using the lambda prep kit (Promega). Phage DNA was double strand sequenced directly using the fluorescent terminator method of cycle sequencing on an Applied Biosystems 373a automated DNA sequencer at the DNA Analysis Facility of The Johns Hopkins University (21,22). Oligonucleotides used for sequencing were synthesized on an ABI 394 synthesizer following ABI protocols. DNA sequence data was analyzed using Sequencher software (Gene Codes). SCR amino acid alignments were conducted using the MacVector program (Oxford Molecular Group). SCR hydropathy data were obtained by Kyte-Doolittle analysis (23) using MacVector. A putative phylogenetic tree relating SCRs to previously cloned full-length OdR sequences was constructed using the unweighted pair group method with arithmetic mean analysis (24, 25) of the Geneworks program (Oxford Molecular Group).
RNase Protection Assays-Riboprobes for RNase protection assays corresponding to D.1-D.4 and G.1-G.4 were in vitro transcribed off of pCRII-pSCR D and pCRII-pSCR G using T7 RNA polymerase (MAXIscript kit, Ambion) and [␣-32 P]UTP (NEN Life Science Products). Riboprobes were run on 6% polyacrylamide, 8 M urea precast TBE gels (Novex) and were gel-purified. RNase protection assays were performed using the RPA II kit (Ambion) according to the manufacturer's protocol.
To determine riboprobe sizes on non-denaturing polyacrylamide gels for comparison with protected fragments, the sense transcripts for pSCR D and pSCR G were in vitro transcribed using SP6 RNA polymerase. Antisense riboprobes were incubated for 24 h at 45°C with sense RNA, 10 g of cesium chloride purified total RNA from pachytene spermatocytes, round spermatids, and olfactory epithelium, and 0.5 g of poly(A) ϩ RNA (CLONTECH) from adult testis, lung, brain, liver, and spleen. Ten g of yeast RNA was used as a control. The integrity of total and poly(A) ϩ RNA employed was evaluated by running RPAs in parallel using a rat ␤-actin antisense riboprobe generated from the pTRI-␤actin-125-rat template (Ambion). RNase-treated samples were precipitated and run on non-denaturing 6% polyacrylamide TBE gels (Novex). The gels were transferred to Whatman filter paper and exposed to Kodak X-OMat film or evaluated by phosphorimaging.
In Situ Hybridization-In situ hybridization using digoxigenin-labeled probes was conducted using full-length SCRs D-2, D-8, and G-14 coding sequence and the novel 5Ј-SCR D RACE and SCR G RACE sequence. Fresh-frozen 20-mm cryostat sections of rat testis (2-, 6-, and 10-week-old rats as indicated) were cut onto Superfrost Plus slides (Fisher), allowed to air dry 1-3 h, and postfixed for 5 min in 4% paraformaldehyde/phosphate-buffered saline. When expression patterns in the testis were compared between probes, serial sections were utilized. The slides were washed 3 ϫ for 3 min in TBS, treated 10 min in 0.25% acetic anhydride, 0.1 M triethanolamine, pH 8.0, washed 3 ϫ for 3 min in TBS, and prehybridized for 2 h at room temperature in hybridization buffer (50% formamide, 5ϫ SSC, 5ϫ Denhardt's solution, 500 mg/ml sonicated herring sperm DNA, 250 mg/ml MRE tRNA). The tissue was incubated with 0.1 ml of buffer containing 40 ng of cRNA probe under a siliconized coverslip at 65°C overnight. After removal of the coverslips in 5ϫ SSC at 65°C, sections were washed as follows: 2 ϫ for 1 h in 0.2ϫ SSC, 1 ϫ for 5 min in 0.2ϫ SSC, and 1 ϫ for 5 min in TBS. Following a 1-h block at room temperature in 4% normal goat serum/TBS, the slides were incubated overnight at 4°C in 1:5000 dilution of anti-DIG Fab fragment (Boehringer Mannheim) in 4% normal goat serum/TBS. Sections were then washed 3 ϫ for 5 min in TBS and 1 ϫ for 5 min in AP buffer (0.1 M Tris-Cl, pH 9.5, 0.1 M NaCl, 50 mM MgCl 2 ). The color signal was developed in AP buffer containing 3.375 mg/ml NBT (Boehringer Mannheim), 3.5 mg/ml 5-bromo-4-chloro-3indolyl phosphate (Boehringer Mannheim), and 0.24 mg/ml levamisole. The color reaction was carried out in the dark for 24 -72 h at room temperature and then terminated with a TE wash. The developed slides were coverslipped with Aquapolymount (Polysciences). Control sections were hybridized with identical quantities of sense cRNA, and no signal was observed. The in situ hybridization protocol has been utilized to distinguish the localizations of transcripts sharing as much as 85-90% nucleic acid identity (6,26).
Rapid Amplification of cDNA Ends (RACE)-PCR-5Ј-RACE-PCR was conducted using the Life Technologies, Inc. 5Ј-RACE kit. One g of adult testis poly(A) ϩ RNA (CLONTECH) was reverse-transcribed (Superscript II RT, Life Technologies, Inc.) using 25 ng of the SCR D specific 3Ј-primer D.2. First strand cDNA was purified and oligo(dC)tailed according to the manufacturer's protocol. dC-tailed cDNA was amplified using the anchor primer and primer D.6 (PGECFL, 3Ј-GAG-GAAGCATTCTCCAGG), a nested SCR D-specific 3Ј-primer. Products resulting from PCR reactions that contained dC-tailed cDNA, but not present in reactions using mock dC-tailed control cDNA, were gelpurified (QIAEX, Qiagen) and reamplified by PCR with the adapter primer and primer D.6. To confirm the presence of SCR D sequence in the RACE products, gel-purified bands were checked by PCR with D.5 (M(T/S)VNCS, 5Ј-ATGA(C/G)TGTCAACTGTTCT), a 5Ј-primer to the presumptive amino terminus of SCR D and D.6. If the appropriate 93-base pair D.5-D.6 product could be generated, the RACE product was subcloned into pCRII vector (Invitrogen) and sequenced with sp6 and T7 primers. The identical procedure was followed using the SCR G-specific 3Ј-primer G.2 to generate first strand cDNA, a nested SCR G-specific primer G.6 (YLASLMG, 3Ј-GCCCATGAGGGAGGCCAAG-TA) for RACE-PCR with the anchor primer, and a 5Ј-primer to the presumptive amino terminus of SCR G, G.5 (MEWLM, 5Ј-ATGGAAT-GGCT(G/A)ATG), used with G.6 for PCR analysis of RACE products.
To investigate the origin of the novel sequence identified upstream of the starting methionines of SCRs, genomic clones containing SCRs G-14, G-15, and D-9 were sequenced in the 5Ј direction. PCR using primers designed against SCR D RACE sequence (5Ј-CGTGATGGGAT-CAGAAAT) and upstream SCR D-9 genomic sequence (3Ј-AGGATG-TACCAAGATCAATGT) was also performed on SCR D-9 phage DNA to identify intervening sequence.
To determine if SCR transcripts in spleen and olfactory tissue likewise undergo splicing, 1 g of spleen poly(A) ϩ RNA and olfactory epithelium total RNA were reverse-transcribed as described above using primer G.2 to generate first strand cDNA. PCR was then conducted using a 5Ј-primer to the SCR G RACE sequence (5Ј-ACCACTCAA-GATTTGCAC) and the 3Ј-G.6 primer; the products generated were TA-cloned and sequenced.

Two Novel Spermatid Chemoreceptor (SCR) Families Are Identified in Purified
Round Spermatids-In a search for OdR-like proteins that may serve functional roles in reproduction, we reasoned that the post-meiotic and actively differentiating cells of the testis would be most likely to express chemoreceptors physiologically relevant to mature spermatozoa. Accordingly, we purified round spermatids from rat testis (Fig. 1a) to isolate total RNA from this select population of cells, and we conducted RT-PCR experiments using degenerate primers (OdR 1 and 2) designed against highly conserved regions of the OdR family.
A robust product was generated using cDNA derived from the round spermatid total RNA (Fig. 1b). The 520-bp product corresponded in size to the band generated from PCR using the control template F2, an originally cloned Buck and Axel OdR. Whereas a 520-bp product was also obtained from testis cDNA, no product was observed using whole brain cDNA. No products were generated when PCR was conducted using negative control samples from each tissue that were mock reverse-transcribed in the absence of reverse transcriptase. The spermatid PCR product was TA-subcloned, and subsequent EcoRI digestion of 14 clones revealed two recurrent restriction patterns, with the exception of clone B which did not contain an insert of the appropriate size (Fig. 1c). DNA sequencing of 13 clones identified two distinct and unique sequences (Fig. 1d) that aligned with members of the OdR family. All inserts were either identical to the sequence of clone D, which contained an internal EcoRI site, or matched the sequence of clone G. In contrast, sequencing of seven subcloned PCR products from the testis yielded seven different partial receptor sequences (data not shown), indicating that a multitude of receptor transcripts are generated in whole testis.
To obtain full-length SCRs, we screened a rat genomic library using probes generated from the partial SCR clones (Fig. 1d). Three distinct clones containing the pSCR D sequence were isolated and encode 321 amino acid proteins sharing Ͼ87% amino acid identity (Fig. 2a). Three distinct clones containing the pSCR G sequence were also identified and encode 327 amino acid pro-teins sharing Ͼ86% amino acid identity (Fig. 2b). The highly homologous clones within each family were determined to be individual gene products based on their unique 3Ј-genomic DNA sequence (Fig. 2, a and b). Hydropathy plots generated by Kyte-Doolittle analysis (23) predict that the receptors contain seven hydrophobic transmembrane domains (TMDs) (Fig. 2, a and b). Multiple sequence alignment with previously cloned OdRs reveals that SCRs share sequence characteristics of the superfamily, with conserved motifs in TMDs 2, 6, and 7 and in the extracellular loop between TMDs 3 and 4.
To conceptualize the relationship of SCRs to other full-length OdR proteins, we constructed a putative phylogenetic tree using the Geneworks program (Fig. 3). Based on the program's unweighted pair group method with arithmetic mean analysis (24,25), the SCR G family is the most evolutionary divergent of all members of the OdR superfamily. The clustering of OdRs does not appear to correlate with tissue source or species of origin. For example, the SCR D family is only 35% identical to SCR G receptors and is more closely related to select mouse OdRs (OR3 and OR37) than to chemoreceptors expressed in dog and human testis (DTMT and HGMPO7I, -E, and -J).
FIG. 1. RT-PCR amplification of OdR transcripts from round spermatids. a, to identify putative OdR transcripts in the select population of post-meiotic differentiating cells of the testis, round spermatids were purified from rat testis by elutriation. One of the spermatids pictured (arrow) has begun to elongate. b, PCR was conducted with cDNA from rat testis, round spermatids, and brain using degenerate primers designed against conserved motifs of the OdR family. By using OdR1 and OdR2 primers, discrete products were obtained in testis and spermatids that correspond in size to a band derived from known OdR clone F2. No products were detected in brain or in control samples reverse-transcribed in the absence of reverse transcriptase. c, spermatid products TA subcloned into pCRII and digested with EcoRI revealed two classes of PCR products. Clones A, C, D, E, F, H, I, J, K, and L contained an insert of 520 bp corresponding to the OdR1 ϩ OdR2 PCR product. Clones G, M, and N contained an internal EcoRI site and therefore produced two bands after digestion whose sum is 520 bp. Clone B did not contain an insert of the appropriate size. d, DNA sequencing of clones D and G revealed two novel and distinct partial spermatid chemoreceptors (pSCRs). The primers used for PCR (OdR1 and OdR2) and for generating probes in subsequent experiments (D.1-D.4 and G.1-G.4) are indicated adjacent to the corresponding sequence. Amino acid identities between the two pSCRs are marked with asterisks. Tissue Localization of SCR D and G Transcripts-RNase protection assays were used to confirm the presence of SCR D and G transcripts in spermatids of the testis. Full-length protection of SCR D and G riboprobes was observed with total RNA from round spermatids but not with RNA purified from spermatocytes in the pachytene stage of meiosis (Fig. 4). Fulllength RNase protection was likewise achieved using total RNA from olfactory tissue, with multiple partially protected products generated by the SCR G riboprobe. Hybridization of antisense probes with the corresponding sense cRNAs served as markers for full-length protection. The integrity of template RNA was confirmed using a 125-bp ␤-actin riboprobe. In experiments utilizing poly(A ϩ ) RNA, we observed full-length protection of SCR riboprobes in testis, with a less robust but detectable product in spleen. No RNase protection occurred with the yeast control RNA or with poly(A ϩ ) RNA from brain, lung, or liver.
In situ hybridization was used to localize SCRs D and G in rat testis (Fig. 5). The expression of SCRs D and G is developmentally regulated, with the absence of detectable RNA message at age 2 weeks, when the seminiferous epithelium contains spermatogonia and spermatocytes but no spermatids (Fig. 5, a and b). SCR D and G transcripts were detected in select seminiferous tubule cross-sections of 6-week-old sexually maturing rats; by this stage of puberty, round spermatids are being produced by the second meiotic division (Fig. 5, c-f). Prominent hybridization with SCR D and G riboprobes was observed in 10-week-old rat testis (Fig. 5, g and h). The specific labeling seen with the antisense riboprobes was absent when sections were incubated with the corresponding sense riboprobes (Fig. 5, i and j). In rat, the cellular composition of seminiferous tubule epithelium varies along the length of the tubule according to a 14-stage growth cycle (27). Thus, different cross-sectional segments contain distinct groupings of spermatogonia, spermatocytes, and spermatids at a characteristic stage of rat spermatogenesis. High power images of in situ hybridization experiments with SCR D and G riboprobes demonstrated the absence of RNA message in spermatogonia and spermatocytes located peripherally to spermatids in the seminiferous tubule cross-sections (Fig. 5, e and f). In contrast, the more adluminal round and elongating spermatids showed robust hybridization with SCR riboprobes (Fig. 5, e and f). The SCR mRNA pattern is consistent with the localization of OdR protein to elongating spermatids in rat testis using antibodies generated against conserved OdR peptide sequences (8) and in dog testis using fusion protein antibodies directed against the DTMT receptor (9).
5Ј-Splicing of SCRs D and G-Several observations prompted us to pursue a comparative analysis of SCR cDNA and 5Ј-genomic sequence. 1) The splice acceptor sequence (28), TTTCTTTCAGG, occurred immediately upstream of the presumptive starting methionine in the genomic sequence of SCRs D-2, -8, and -9. 2) In situ hybridization studies comparing the localization of SCR D and G transcripts in serial sections of sexually mature testis revealed overlapping but also non-overlapping mRNA patterns among seminiferous tubule cross-sections (data not shown). This finding, coupled with the observed developmental expression of SCRs, suggested that distinct transcriptional and/or translational controls may govern the production of SCRs.
3) The introduction of constructs containing OdR coding sequences into transgenic mice has failed to produce detectable expression of OdR proteins. 3 We reasoned that the functional expression of OdR proteins may require 5Ј-translated or -untranslated spliced sequence for message stability, membrane targeting, and/or receptor function.
RACE-PCR experiments were conducted by employing SCR Dand SCR G-specific 3Ј-primers to generate first strand cDNAs from testis poly(A) ϩ RNA. RACE-PCR using the anchor primer and nested SCR 3Ј-primers yielded products for both SCRs D and G. SCR D RACE products contained sequence upstream of the SCR D-8 presumptive starting methionine that was distinct from the genomic clone, providing evidence of 5Ј-splicing (Fig. 6a). Interestingly, the novel sequence encodes an upstream open reading frame of 20 amino acids starting with a methionine at nt Ϫ60 which is preceded by a stop codon at nt Ϫ75. Theoretically, the RACE cDNA sequence could encode a 5Ј-MREPIPQSRHVL-KEEGFCLE amino-terminal extension, which is hydrophilic by Kyte-Doolittle analysis (23). Sequences of the SCR G RACE prod-3 R. Reed, personal communication.
FIG. 4. SCR RNase protection assays. Antisense riboprobes corresponding to pSCR D (D.1-D.4) and pSCR G (G.1-G.4) were employed in RNase protection assays using 10 g of total RNA from pachytene spermatocytes, round spermatids, and olfactory tissue and 0.5 g of poly(A) ϩ RNA from testis, lung, brain, liver, and spleen. Hybridization of the antisense riboprobe with the corresponding sense cRNA served as a marker for full-length protection. Full-length protection of both probes was observed with total RNA from round spermatids and olfactory tissue and with poly(A ϩ ) RNA from testis and spleen. Prominent smaller bands representing partial riboprobe detection were generated in assays using the pSCR G probe and olfactory tissue total RNA. No RNase protection occurred with total RNA from pachytene spermatocytes or with lung, brain, or liver poly(A ϩ ) RNA. The integrity of total and poly(A) ϩ RNA was confirmed using a rat ␤-actin antisense riboprobe. 10 g of yeast RNA served as a negative control for the experiment.
ucts are identical to the genomic SCR G-14 clone up to nt Ϫ73, at which point novel upstream sequence was identified (Fig. 6b). At position Ϫ90, a stop codon occurs, and no alternative methionine is encoded downstream. As could also be the case for SCR D, the novel SCR G RACE sequence likely represents 5Ј-untranslated sequence.
To confirm that the clones obtained by RACE-PCR represent 5Ј-splice products, SCR genomic clones were restriction digested with EcoRI/BamHI or HindIII/EcoRV, and Southern analysis was performed using probes corresponding to the novel 5Ј-sequence. The SCR G RACE probe, for example, recognized bands in SCR G-14 and G-15 digests, indicating that the novel 5Ј-RACE sequence was contained in the phage insert (data not shown). Sequencing of the SCR G-14 genomic clone identified the identical 140-bp SCR G RACE sequence 1475 bp upstream of the starting methionine (Fig. 7). In the SCR G-15 clone, sequence 88% identical to SCR G RACE was found 3503 bp upstream of the starting methionine (Fig. 7). In both cases, a splice acceptor site precedes the starting methionine by 73 bp. 5Ј-Sequencing and PCR experiments also led to identification of homologous SCR D RACE sequence in the upstream genomic DNA of SCR clone D-9 (data not shown). In contrast to SCR G receptors, the SCR D splice acceptor sequence occurs immediately adjacent to the presumptive starting methionine.
To determine if splicing of SCR transcripts likewise occurs in spleen and olfactory epithelium, PCR experiments were conducted using the 5Ј-SCR G RACE and 3Ј-nested G.6 primers on first strand cDNAs generated from both tissues with primer G.2. The identically spliced SCR G-14 product originally obtained from testis was detected in both spleen and olfactory epithelium, indicating that 5Ј-splicing of this SCR is a common phenomenon in these tissues.
To localize the spliced RNA sequences in rat testis, we conducted in situ hybridization studies using probes corresponding to the novel 5Ј-RACE sequences (Fig. 8b and data not shown). The 5Ј-spliced sequences were detected in adluminal spermatids, with robust hybridization occurring in seminiferous tubule profiles throughout the testis. Whereas in situ hybridization with a full-length SCR D-8 probe-labeled spermatids of select tubule profiles (Fig. 8a), a serial section incubated with a probe to the novel 5Ј-SCR D RACE sequence displayed a strong signal in many more profiles (Fig. 8b). This finding indicates that the 5Ј-spliced sequence is present in multiple mRNAs, in addition to the subset of SCR D-8 mRNAs localized by the FIG . 5. SCR D and G in situ hybrid-ization. Digoxigenin-labeled antisense riboprobes corresponding to full-length D-2 and G-14 receptors identified chemoreceptor transcripts in spermatids of sexually maturing and mature testis only. a and b, low power images of fresh-frozen testis from a 2-week-old rat are negative for SCR transcripts. c and d, SCR mRNAs are detected in select seminiferous tubule cross-sections in the testis of a sexually maturing 6-week-old rat. e and f, higher power images demonstrate the absence of RNA message in spermatogonia and spermatocytes located peripherally to spermatids in the seminiferous tubule cross-section (region demarcated by black arrowheads). The more adluminal round and elongating spermatids show robust hybridization with SCR riboprobes (white arrows). g and h, sexually mature testis from a 10-week-old rat contains high levels of SCR mRNAs in spermatids. i and j, the specific labeling seen with the antisense probes is completely absent when sections are incubated with the corresponding sense probes. Black bar (a-d and g-j) ϭ 150 m; white bar (e) ϭ 75 m; checkered bar (f) ϭ 60 m. specific probe. This conclusion is consistent with our identification of conserved RACE sequence in the upstream genomic DNA of distinct members of a given SCR family. Although a search of the National Center for Biotechnology Information data bases did not identify the RACE sequences in other known cDNAs, our results do not exclude the possibility that the spliced sequences may also be present in other testis transcripts unrelated to SCRs. DISCUSSION In the present study, we have identified two families of putative OdRs in spermatids of rat testis, termed spermatid chemoreceptors (SCRs) D and G. The two classes of chemoreceptors are only 35% identical to each other, suggesting that they recognize disparate ligands and thereby subserve distinct functions. Within each family, the individual chemoreceptors are highly homologous, displaying at least 86% amino acid identity. The existence of nearly identical receptors may have resulted from gene duplication and may provide a biological back-up mechanism that ensures detection of signals important to the reproductive process. In contrast to the exclusive expression of one or at most a small number of OdRs by individual olfactory neurons (3,6,29), SCR D and G transcripts are abundant and widely detected among spermatids of the rat testis.
Whereas the identical pair of degenerate primers repeatedly identifies only two classes of chemoreceptors in rat spermatid cDNA, a much wider variety of partial OdR-like products are generated from whole testis cDNA. The physiologic relevance of OdRs in the testis has been questioned, because changes in chromatin structure during meiosis may generate spurious genetic transcripts. The identification of many partial receptor transcripts in whole testis compared with spermatids may lend support to this hypothesis or represent the requirement for multiple receptors to serve eclectic functions in spermatogenic cells. Nevertheless, the identification of SCRs D and G in the post-meiotic differentiating cells of the testis by RT-PCR, RNase protection assays, and in situ hybridization implies a physiologically relevant role for SCRs in mature spermatozoa. We previously used antibodies to conserved OdR peptide sequence to identify 64-kDa proteins in the membrane fractions of rat, hamster, and human spermatozoa (8). The molecular weights observed are roughly double the ϳ35-kDa mobility predicted from SCR amino acid sequence. Whereas glycosyla-tion can account for some of the molecular weight disparities (30), others have detected multimeric forms of OdRs by immunoblot analysis (31). The latter observation may account for the approximate molecular weight doubling we observe in testis tissue and is believed to arise from the inability to completely disrupt inter-receptor hydrophobic interactions by denaturing gel electrophoresis. The localization of OdR protein to spermatids of the testis and to the midpiece of mature spermatozoa provides further support for the physiological relevance of such chemoreceptors in male reproductive function (8,9). The localization of desensitization proteins (8) and inositol 1,4,5trisphosphate receptor (10) to the tail midpiece, the respiratory center of mature sperm, provides additional intriguing support for the existence of a signaling mechanism to regulate mammalian sperm motility. Sengupta et al. (32) recently established an explicit functional role for Odr-10, a Caenorhabditis elegans olfactory receptor, in recognizing the molecule diacetyl, which results in worm chemotaxis toward the odorant. By analogy, liberation of a signaling molecule by the egg or female reproductive tract may activate sperm OdRs, modulating motility and directing locomotion toward the fertilization site.
In addition to olfactory tissue, mammalian OdR-like transcripts have now been identified in testis (7,8), taste tissue (33), and developing heart (34). In this report, we also find OdR-like transcripts by RNase protection assays in the adult spleen. Preliminary in situ hybridization experiments show a selective localization of SCR transcripts to the periarterial lymphatic sheath region of the spleen, where B-and T-lymphocytes reside (data not shown). Thus, in addition to potential roles in odorant detection, olfactory neuron axonal guidance (35), sperm-egg signaling, taste sensation, and cardiac morphogenesis, OdR-like proteins may also function as immune cell chemoreceptors. By using RT-PCR, RNase protection assays, and/or in situ hybridization, we and others (33,34,36) have found identical or at least nearly identical receptor transcripts in multiple select tissues. In these cases, it is unclear whether the OdR receptors serve common chemosensory roles, are generated by leaky receptor transcription, or undergo post-translational modification to endow tissue-specific functions.
Our phylogenetic analysis demonstrates that SCRs D and G are no more homologous to testis OdRs of different species than they are to olfactory tissue-derived OdRs from the same or FIG. 6. RACE-PCR provides evidence for 5-splicing of SCRs. a, alignment of 5Ј-flanking sequence obtained from genomic clone SCR D-8 and SCR D RACE-PCR demonstrates sequence identity beginning at the presumptive starting ATG (numbered ϩ1 and marked by a black arrowhead) but completely different sequences immediately upstream. The novel 5Ј-sequence obtained by RACE-PCR (underlined) has a potential translation initiation ATG at nt Ϫ60 (gray arrowhead) which is preceded by a stop codon at nt Ϫ75 (asterisk) and is in frame with the downstream ATG (ϩ1, black arrowhead). b, alignment of genomic SCR G-14 with SCR G RACE sequence reveals novel 5Ј-sequence (underlined) beginning at nt Ϫ73. A stop codon appears at nt Ϫ90 (asterisk) of the RACE product and is in frame with the presumptive starting ATG (numbered ϩ1 and marked by a black arrowhead). The sequence data are available from GenBank TM (AF034896, AF034897, and AF034901). different species. Vanderhaeghen et al. (36) have recently completed a thorough analysis of partial olfactory receptor transcripts identified by PCR from human, rat, dog, and mouse whole testis cDNA. Across the different species evaluated, the testicular clones show no specific sequence characteristics compared with OdRs identified in olfactory tissue; testicular receptors within a given species likewise do not cluster. However, the investigators do find highly homologous and possibly orthologous receptors in two species, with three of the receptor couples both detected in testis by RPA. Theoretically, the diversity of testicular receptors between species may enable species-specific positive and negative regulation of sperm-egg signaling, whereas orthologous receptors may serve roles common to spermatogenesis and sperm function across species.
The location of splice acceptor sites upstream of the presumptive starting methionine in genomic SCR clones led to the discovery that SCRs undergo 5Ј-splicing. Novel 5Ј-SCR D and G sequences were identified by RACE-PCR and localized in corresponding SCR genomic clones by upstream sequencing. Excluding a 2-kilobase pair gap in SCR G-14, the intron sequences of SCRs G-14 and G-15 share regions of sequence homology and are likely to contain conserved elements involved in transcriptional control of SCR G receptors. Such regulation, for example, may account for the observed expression of SCRs in the sexually maturing and mature testis but not in the prepubertal testis. Furthermore, differences in transcriptional control may explain the instances where SCR D and G spermatid mRNA patterns are non-overlapping (data not shown).  7. Identification of conserved SCR G RACE sequence in two distinct SCR G genomic clones. 5Ј-Genomic sequencing of SCRs G-14 and G-15 identified the SCR G RACE sequence (represented as black boxes labeled RACE G in the schematic and underlined in the sequence alignment below) at variable distances upstream of the respective starting methionines (arrowheads). In the SCR G-14 cDNA, the RACE sequence adjoins 73 nt upstream of the starting ATG (represented as a continuation of the black box in the schematic and marked with an asterisk in the alignment). The splicing of SCR G-14 is demonstrated by the arrow in the schematic. Excluding a 2-kilobase pair gap in the genomic sequence of SCR G-14, alignment of SCRs G-14 and G-15 reveals stretches of homologous intron sequence (represented as wide-striped boxes in the schematic). The sequence data are available from GenBank TM (AF034901 and AF034902). Scale bar ϭ 0.5 kilobase pairs.