Novel Alternatively Spliced Exon in the Extracellular Ligand-binding Domain of the Pituitary Adenylate Cyclase-activating Polypeptide (PACAP) Type 1 Receptor (PAC1R) Selectively Increases Ligand Affinity and Alters Signal Transduction Coupling during Spermatogenesis*

The expression of the paracrine signaling hormone pituitary adenylate cyclase-activating polypeptide (PACAP) is regulated in a cyclical fashion during the 12-day spermatogenic cycle of the adult rat testis. The precise functions of PACAP in the development of germ cells are uncertain, but cycle- and stage-specific expression may augment cAMP-regulated gene expression in germ cells and associated Sertoli cells. Here we report the existence of a heretofore unrecognized exon in the extracellular domain of the PACAP type 1 receptor (PAC1R) that is alternatively spliced during the spermatogenic cycle in the rat testis. This splice variant encodes a full-length receptor with the insertion of an additional 72 base pairs encoding 24 amino acids (exon 3a) between coding exons 3 and 4. The PAC1R(3a) mRNA is preferentially detected in seminiferous tubules and is expressed at the highest levels in round spermatids and Sertoli cells. Analyses of ligand binding and signaling functions in stably transfected HEK293 cells expressing the two receptor isoforms reveals a 6-fold increase in the affinity of the PAC1R(3a) to bind PACAP-38, and alterations in its coupling to both cAMP and inositol phosphate signaling pathways relative to the wild type PAC1R.

The expression of the paracrine signaling hormone pituitary adenylate cyclase-activating polypeptide (PACAP) is regulated in a cyclical fashion during the 12-day spermatogenic cycle of the adult rat testis. The precise functions of PACAP in the development of germ cells are uncertain, but cycle-and stage-specific expression may augment cAMP-regulated gene expression in germ cells and associated Sertoli cells. Here we report the existence of a heretofore unrecognized exon in the extracellular domain of the PACAP type 1 receptor (PAC1R) that is alternatively spliced during the spermatogenic cycle in the rat testis. This splice variant encodes a full-length receptor with the insertion of an additional 72 base pairs encoding 24 amino acids (exon 3a) between coding exons 3 and 4. The PAC1R(3a) mRNA is preferentially detected in seminiferous tubules and is expressed at the highest levels in round spermatids and Sertoli cells. Analyses of ligand binding and signaling functions in stably transfected HEK293 cells expressing the two receptor isoforms reveals a 6-fold increase in the affinity of the PAC1R(3a) to bind PACAP-38, and alterations in its coupling to both cAMP and inositol phosphate signaling pathways relative to the wild type PAC1R. These findings suggest that the extracellular region between coding exons 3 and 6 of PAC1R may play an important role in the regulation of the relative ligand affinities and the relative coupling to G s (cAMP) and G q (inositol phosphates) signal transduction pathways during spermatogenesis. PACAP 1 is a paracrine signaling peptide that functions as a neurotransmitter, a neurotrophic factor, and a secretagogue for insulin, catecholamines, and pituitary hormones (reviewed in Ref. 1). PACAP is synthesized in the form of a precursor (proPACAP) and is processed into active peptides of either 27 (PACAP-27) or 38 (PACAP-38) amino acids, differing by an 11-amino acid extension at the carboxyl terminus. High levels of PACAP are produced by the germ cells of the rat testis (2)(3)(4) and are translated from a testis-specific mRNA (5) transcribed from the PACAP gene using an alternative promoter active in round spermatids (6,7). PACAP regulates gene expression in germ cells (8), increases cAMP in Sertoli cells (9), and stimulates steroidogenesis in Leydig cells (10).
Part of the diversity in PACAP actions is due to the existence of multiple receptors. Three distinct genes encode receptors capable of interacting with PACAP-27 and PACAP-38 isoforms at physiological concentrations. The PACAP-specific type 1 receptor and the PACAP receptors type 2 and 3, which also bind vasoactive intestinal peptide (VIP), and are therefore referred to as VIPR1 (VPAC1R) (11,12) and VIPR2 (VPAC2R) (13,14). The PACAP type 1 receptor (PAC1R: identified by several laboratories and reviewed in Ref. 15) has a markedly higher affinity for PACAP and does not function as a VIP receptor at physiological concentrations. All three receptors are seventransmembrane-spanning domain G protein-coupled receptors that stimulate adenylate cyclase activity. However, the PAC1R also stimulates phospholipase C (PLC), leading to the accumulation of inositol phosphates and diacyl glycerol. The VPAC1R and VPAC2R may also couple to PLC in some cells (16,17).
Multiple receptor isoforms are generated by alternate splicing of the PAC1R mRNA. Two 81-bp exons ("hip" and "hop") are alternatively spliced within the third intracellular loop and play a role in modulating inositol phosphate responsiveness (18,19). Similarly, a variant of the receptor lacking the fourth and fifth exons encoding sequence in the extracellular domain has been described in hypothalamus and pituitary (20). The loss of 21 amino acids from the extracellular ligand-binding domain results in an increased affinity for PACAP-27. Other variants of the type I PACAP receptor have also been described (21,22), as well as alternative splicing of the 5Ј-untranslated region (23). Here we describe a novel PAC1R splice variant highly expressed in the testis and consisting of the insertion of a 72-bp exon encoding 24 amino acids in the extracellular ligand-binding domain. This alternatively spliced exon sub-stantially alters PACAP-27 and PACAP-38 ligand binding affinities and the relative coupling of the receptor to the cAMPcoupled and inositol phosphate-coupled signal transduction pathways.
RT-PCR and Southern Blotting-Whole cell RNA was extracted with TriZOL reagent (Life Technologies, Inc.) in accordance with the manufacturer's specifications. RNA, in 10 l of H 2 O, was combined with 0.5 g of oligo(dT) 16 and heated to 65°C for 10 min, then cooled on ice. RT buffer, dNTPs (50 M each), dithiothreitol (5 mM), SuperScript II (Life Technologies, Inc.) enzyme (50 units), and H 2 O were added for a total volume of 40 l, and reactions were incubated at 42°C for 40 min. All PCRs were performed in 50-l reactions using 2 l of cDNA template. Reactions contained 20 pmol each of forward and reverse primers, 0.2 mM each of dNTPs, and 2.5 units of thermostable Taq polymerase (TaKaRa Biomedical Inc., Berkeley, CA). For nested PCR reactions, a second reaction with nested primers was carried out using 2 l of the first reaction as template. PACAP receptor PCR from cDNA in Figs. 1 and 3A was carried out in two rounds with primers PRF-79 and PRR-1256 followed with primers PRF-70 and PRR-552. The first round of PCR used 10-s denaturation at 94°C, 20-s annealing at 58°C, and a 2-min extension at 72°C for 20 cycles. The second round was identical except for a shorter extension time (1 min) and 30 cycles. PAC1R(3a) PCR from cDNA in Fig. 3B was carried out using the same protocol, but with a different primer pair for the first round (PRF-79 and PRR-1465).
Cloning of the full-length PAC1R and PAC1R(3a) was done with Pfu polymerase (Stratagene, La Jolla, CA) and the primer pairs PRF-79 and PRR-1465, followed by PRF-7 and PRR-1404. Second round extension time was increased to 2 min, but other parameters remained as described above. For PACAP (PCPTF-32 and PCPR-1120) and SOX-17 (SOX-17F1 and SOX-17R1): 10-s denaturation at 94°C, 20-s annealing at 58°C, and 1-min extension at 72°C for 25 cycles. For follicle-stimulating hormone receptor, FRF-11 and FRR-2196: 10-s denaturation at 94°C, 20-s annealing at 58°C, and 2-min extension at 72°C for 30 cycles. The intron between coding exons 3 and 4 was amplified from rat genomic DNA with the primers PRF-215 and PRR-285 and a 1:50 mixture of Pfu and standard Taq polymerase. A "touchdown" PCR protocol was used, with cycling parameters of 2-s denaturation at 94°C and 3-min extension at 72°C for 5 cycles, followed by 2 s at 94°C and 3 min at 67°C for 30 cycles.
For Southern hybridization, PCR products were transferred to Mag-naGraph membrane (Micro Separations Inc., Westboro, MA) by capillary transfer. Hybridization with [␥-32 P]ATP-labeled oligonucleotide probes was done in a solution of 5 ϫ SSC, 1% SDS, 10ϫ Denhardt's, and 100 g/ml denatured salmon sperm DNA for 3 h at 37°C. Blots were washed to a maximum stringency of 0.5 ϫ SSC at 52°C.
Cloning and Sequencing of DNAs-Products of RT-PCR reactions were cloned into pCR2.1 by TA cloning (Original TA cloning kit: Invitrogen, Carlsbad, CA). To facilitate ligation, products amplified with Pfu polymerase were purified and treated with standard Taq polymerase for 10 min at 72°C in the presence of dNTPs. Sequencing was carried out with reagents from a Sequenase 2.0 kit (Amersham Pharmacia Biotech). Sequencing gels were exposed to Kodak film, from which scanned images were generated on a computing densitometer (Molecular Dynamics). Nucleotide sequences were interpreted from gel images using DNA codes software (PDI Imageware Systems, New York City).
Creation of Stable Cell Lines Expressing Recombinant PACAP Receptor-encoding DNA-The PAC1R and PAC1R(3a) were subcloned into pCR 3.1 (Invitrogen). Linearized constructs were transfected into HEK293 cells, and stably expressing cells were selected by exposure to G418 (Life Technologies, Inc.) at 0.2-1.0 mg/ml. Cell lines used in experiments described in the legends to Figs. 4 and 5 were selected on the basis of a similar maximal binding of 125 I-PACAP-27 by an identical number of cells. Cells were cultured in Dulbecco's modified Eagle's medium (DMEM: Life Technologies) ϩ 10% FCS.
Measurement of Ligand Binding Affinities-HEK293 cells stably expressing the receptors were harvested with trypsin and transferred to receptor binding buffer (RBB: 138 mM NaCl, 5.6 mM KCl, 1.2 mM MgCl 2 , 0.4 mM CaCl 2 , and 10 mM HEPES, pH 7.4) with 0.1% BSA. Approximately 20,000 -50,000 cells were used per binding reaction. For details of binding reactions, see Kieffer et al. (24). 125 I-PACAP-27 (30,000 cpm) and competing unlabeled peptide were added followed by a 30-min incubation, a single wash in RBB, 0.1% BSA, and determination of bound labeled peptide in an Apex automatic ␥ counter (Micromedic Systems Inc.).
Measurement of cAMP Formation-Stable cell lines in 10-cm dishes were transfected with reporter plasmid pCRE-luc and control plasmid pRSV-␤-galactosidase (Stratagene, La Jolla, CA) 18 h prior to assay. Immediately following transfection, the cells were trypsinized and replated in 24-well plates. At the time of assay, cells were transferred into DMEM ϩ 0.1% BSA, and agonist (PACAP-27 or PACAP-38; Novabiochem, San Diego, CA) was added in a range of concentrations (10 Ϫ11 M through 5 ϫ 10 Ϫ6 M and an unstimulated control). Luciferase activity was quantified after a 6-h incubation using the Promega luciferase system in accordance with the manufacturer's instructions (Promega Corp., Madison, WI). Results shown in Fig. 4, A and B, are from single experiments, with each treatment performed in quadruplicate. Inhibitor compounds H89 and U73122 (Calbiochem) were used to verify the predominance of the cAMP pathway in pCRE-luc activation.
Measurement of Inositol Phosphate Formation-Stably expressing cell lines were plated into 24-well plates at matching densities and cultured for 24 -48 h prior to assay. In the 24 h preceding assay, the media were substituted with inositol-free DMEM ϩ 10% FCS containing 3 Ci/ml myo-[ 3 H]inositol (Amersham Pharmacia Biotech). At the time of assay, wells were washed twice in DMEM ϩ 0.1% BSA and incubated for 40 min in 0.8 ml of DMEM ϩ 0.1% BSA containing 10 mM LiCl and a predetermined concentration of agonist (PACAP-27 or PACAP-38; Novabiochem). Reactions were terminated by replacing the media with 0.5 ml of ice-cold 5% trichloroacetic acid. Wells were scraped and transferred to 1.5-ml tubes. Recovered precipitate was extracted with 0.5 ml of chloroform and neutralized with 10 l of 0.1 M NaOH. Extraction of total inositol phosphates was performed by a modified chromatographic procedure using AG1-X8 resin (Bio-Rad) prepared in 10 mM myo-inositol. 400 l of the supernatant was added to ϳ100 l of resin and gently mixed at room temperature for 20 min. The resin was washed twice with 1 ml of 60 mM sodium borate, 5 mM borax. Inositol phosphates were eluted with 200 l of 2 M ammonium formate, 0.1 M formic acid and quantified by scintillation counting. Results shown in Fig. 4, A and B, are from single experiments performed in quadruplicate.
Measurement of Intracellular Calcium-Receptor effects on intracellular calcium were investigated in the stably transfected HEK293 cells, stimulated with PACAP-27 or PACAP-38 at 100 nM. Measurements of intracellular calcium were carried out according to the methods described in Ref. 25.
Fractionation of Testis Cell Populations-Separation of testis cells was achieved by velocity sedimentation at unit gravity in a STAPUT chamber based on the method of Bellve (26). Briefly, a single testis from adult rat was dissociated to single cells by sequential digestion in collagenase (5 mg/ml: Sigma) and trypsin/DNase I (5 mg/ml and 1 g/ml: Sigma and Roche Molecular Biochemicals). Cells were centrifuged at 800 ϫ g for 2 min and then resuspended in DMEM ϩ 10% FCS and sieved through a 100-m mesh strainer. Cells were collected by centrifugation once more and resuspended in 60 ml of enriched Krebs-Ringer buffer with 0.5% BSA. This suspension was introduced into a 23-cm diameter STAPUT chamber, followed by a 2.1-liter gradient of 2-4% BSA in enriched Krebs-Ringer buffer. Separation was carried out at 4°C for 2-3 h. The apparatus was then drained and the contents collected into 120-ml fractions. Cells were concentrated by centrifugation, and early fractions not containing significant amounts of cells were discarded. Samples were taken from each remaining fraction and processed for RT-PCR as described above. Cultures of adult Sertoli cells used for RT-PCR analysis were prepared from rat testes by sequential digestion as above, followed by selection of adherent cells by plating in DMEM ϩ 10% FCS on plastic culture flasks. The cells were trypsinized and replated after 24 h in culture, and again after 72 h, to reduce germ cell contamination. The Sertoli cells were further purified by one or two rounds of trypsinization and replating, and in some cases by culture in media containing 3 g/ml cytosine arabinofuranoside to suppress fibroblast growth.
Data Analyses-The hydrophobicity profile (Fig. 2B) was generated with Protean 3.14 (DNASTAR Inc., Madison, WI). The K m and EC 50 values (Table I), and associated confidence intervals, were calculated with Prism 2.0.1 (GraphPad Software Inc., San Diego, CA).

Detection of PAC1R Variants by Nested PCR-PCR primers
were designed to detect the existence of different isoforms of PAC1R mRNA following a nested PCR protocol. Distal primers encompassing a region including all of the extracellular domain and transmembrane domains were used for initial DNA amplification, followed by proximal primers specific for the extracellular domain (Fig. 1A). PAC1R products were then identified by Southern hybridization with an end-labeled oligonucleotide probe. RT-PCR analysis of RNA from 2-mm segments of seminiferous tubules (Fig. 1B) revealed the presence of three isoforms. Sequencing of the cloned products identified the lower most band (arrow 1) to be identical to the published rat PAC1R sequence (GenBank TM accession number Z23279). The largest product (arrow 3) contained additional sequence between coding exons 4 and 5 and is likely a consequence of the inclusion of part of the 5Ј-region of the intron between exons 4 and 5 (intron slippage). The additional sequence includes two in-frame stop codons. The product corresponding to arrow 2 contained a novel region of 72 nucleotides inserted between coding exons 3 and 4 with no interruption of the translational reading frame. The predicted translation products are shown (Fig. 1C). Exon 3a is an alternatively spliced "cassetted" exon and not due to intron slippage. The nested PCR approach was further utilized to isolate full-length receptor clones. Three full-length versions of the PACAP type 1 receptor were cloned (Fig. 1C): 1) PAC1R, equivalent to GenBank TM accession number Z23279; 2) the PAC1R with the additional 24 amino acids; and 3) the PAC1R receptor "HOP" variant. The identities of all three clones were verified by sequencing. The PAC1R clone 1 was identical to the published sequence (18). In addition to the extra 72 bp of coding sequence ( Fig. 2A), the longer clone 2 had one substitution (G to A) at nucleotide 798, compared with the published sequence. This G to A substitution did not alter the conceptual translation product and is most likely the result of misincorporation of a nucleotide during the PCR procedure. Clone 3 (Fig. 1C) is an intron slippage splice variant resulting in a frameshift and the formation of a truncated protein.
To determine whether the sequence coding for the insertion of an additional 24 amino acids into the PAC1R was due to the splicing in of a separate exon or was part of an existing exon or intron incorporated by a splice slippage event, the region between coding exons 3 and 4 was amplified from rat genomic DNA. Partial sequencing of this ϳ2.5-kilobase pair fragment revealed that the inserted sequence was a separate exon (termed 3a), separated from the preceding exon 3 by 161 bp. (Fig. 2B). The additional 24 amino acids encoded by the alternatively spliced exon 3a predict a markedly altered hydrophobicity profile of the extracellular domain by introducing a short stretch of hydrophilic amino acids, followed by a predominantly hydrophobic region (Fig. 2C). Furthermore, analysis of the inserted exon 3a using PROSITE software (27, 28) identified a potential N-linked glycosylation site (NTSS), which may be a glycosylation-sensitive site for the regulation of the biologic functions of exon 3a.

PAC1R(3a) mRNA Occurs in Distinct Cell
Populations of the Testis-The tissue distribution of the novel receptor variant was assessed by RT-PCR of several rat tissues responsive to PACAP (Fig. 3). A DNA product corresponding to PAC1R was amplified from cerebellum, hypothalamus, pituitary, pancreatic islets, seminiferous tubule, and the INS-1 cell line ( Fig. 3A; 482-bp band in upper panel). A product corresponding in size to that expected for PAC1R(3a) is also strongly detected in seminiferous tubules. The identity of this product is confirmed by Southern hybridization to an oligonucleotide probe (DP1) complementary to the exon 3a sequence (lower panel). The strongest signal for the PAC1R(3a) occurs in the seminiferous tubule. Other tissues show only trace amounts of PAC1R(3a) PCR product.
To determine which cell populations within the testis express PAC1R(3a), dissociated rat testis cells were fractionated by velocity sedimentation at unit gravity. Fractions were examined by RT-PCR for products specific for certain cell types, as well as PAC1R(3a) (Fig. 3B). A PCR product corresponding to PAC1R(3a) was detected in highest amounts in fractions enriched for Sertoli cells (located by the follicle-stimulating hormone receptor PCR signal) and round spermatids (PACAP PCR signal). The product is absent in spermatocytes (fractions 4 and 5), in which a pachytene spermatocyte marker, transcription factor SOX-17, is expressed. PCR products corresponding in size to PAC1R and PAC1R(3a) were also detected in cultures of adult rat Sertoli cells substantially purified as described under "Materials and Methods" (data now shown). However, as some germ cells remain associated with Sertoli cells in these cultures, and in STAPUT fractionations, the presence of the variant receptor on adult Sertoli cells cannot be verified.
PAC1R(3a) Has an Increased Binding Affinity for PACAP-38 -Binding studies were carried out in HEK293 cells stably transfected with recombinant plasmids expressing either the PAC1R or the PAC1R(3a). The receptors were compared for their relative binding of PACAP-27 and PACAP-38 by displacement of 125 I-labeled PACAP-27 with unlabeled peptides (Fig.  4). Whereas the PAC1R and PAC1R(3a) displayed equal affinity for PACAP-27 (Fig. 4A), the PAC1R(3a) showed a significantly increased (6.1-fold) affinity for PACAP-38 over the PAC1R (Fig. 4B). The amino-terminal truncated PACAP-(6 -38) peptide, which lacks signaling activity and therefore acts as an antagonist of the PACAP receptor, displayed a 1.9-fold increase in binding affinity for PAC1R(3a) compared with the PAC1R (Fig. 4C). The K m values of the different ligand/receptor combinations are summarized in Table I. In other experiments, the sequence-related peptides VIP, PHI (peptide histidine isoleucine), and PRP did not displace 125 I-PACAP-27 from either the PAC1R-or PAC1R(3a)-expressing cell lines (data not shown).
The PAC1R(3a) Has Altered Signal Transduction Coupling Efficiencies Compared with PAC1R-The efficiency of signaling through the cAMP pathway was assayed using a cAMPresponsive reporter gene. The specificity of this method was assessed by using specific inhibitors of protein kinase A (H89) and PLC (U73122). H89 dose-dependently inhibited the accumulation of luciferase activity in both PAC1R-and PAC1R(3a)expressing cell lines treated with PACAP-27, and no inhibition was observed with U73122 (data not shown). The cAMP responsiveness of the PAC1R(3a)-expressing cell line was rightshifted with respect to the PAC1R-expressing line with both PACAP-27 and PACAP-38 (Fig. 4, A and B). This difference in ligand concentration coupling to cAMP formation between PAC1R and PAC1R(3a) was observed in a total of four independent experiments. Coupling to PLC-dependent signaling pathways was assessed by measuring the accumulation of inositol phosphates in response to PACAP-27 or PACAP-38 in the receptor-expressing stable cell lines. The PAC1R(3a)-expressing cell line clearly has a reduced sensitivity to both PACAP-27 and PACAP-38 compared with the PAC1R-expressing cell line (Fig. 4B). This observation, seen in a total of three independent experiments, suggests that the 24 amino acids inserted in the alternatively spliced PAC1R(3a) also reduces coupling to the PLC-dependent signaling pathway. The EC 50 values for cAMP and inositol phosphate generation are given in Table I. No difference was detected in the ability of PAC1R and PAC1R(3a) to elevate intracellular calcium in response to treatment of the HEK293 with PACAP38 or PACAP27. DISCUSSION Alternative splicing of the PAC1R pre-mRNA has been described previously for the 5Ј-untranslated region, the third intracellular loop, and the extracellular ligand-binding domain (18,20,23). Here we describe a new splice variant of the PAC1R resulting from the cassetting in of an exon that adds an additional 72 nucleotides in the amino-terminal extracellular domain and results in alterations of both ligand binding and signaling function. Inclusion of the additional 72-bp in-frame exon (termed exon 3a) results in the translation of an addition 24 amino acids within the amino-terminal extracellular domain of the mature receptor providing the unique receptor PAC1R(3a). Although the identity of amino acids between PACAP and VIP receptors (VPAC1R and VPAC2R) in the extracellular amino-terminal domain is 38% for PAC1R, compared with VPAC1R and 41% for PAC1R, compared with VPAC2R, the region encoded by coding exons 3a, 4, and 5 is unique to the PAC1R. We speculate that the presence of exon 3a perturbs the overall alignment of conserved residues within the extracellular domains of the three receptors. The inclusion of coding exon 3a in the receptor increases the affinity of the receptor for PACAP-38 by 6-fold, but does not affect the affinity of the receptor for PACAP-27. PAC1R(3a) also has a small increase in affinity for the truncated PACAP peptide, PACAP-(6 -38), compared with the PAC1R.
The interaction between PACAP and its receptors has been examined in experiments with mutated and truncated receptors. The amino-terminal extracellular domain of the PAC1R and a single membrane-anchoring region is sufficient to mediate specific binding of PACAP, albeit with a 20-fold reduction in  affinity compared with the full-length receptor (29). The role of the extracellular loops in stabilizing ligand binding is apparent in experiments in which ligand binding by the VPAC1R was found to be dependent on a conserved cysteine residue in the third extracellular loop (30). Three distinct regions within PACAP are believed to contribute to receptor binding. A region within the carboxyl terminus of the 1-27 peptide comprising two adjacent ␣-helices (31, 32) may provide primary ligand-receptor recognition (33)(34)(35). The integrity of the helices is important for high affinity binding (33). A similar structural arrangement is also detected in VIP (36). The first 6 amino-terminal residues of PACAP are important for both binding and signal transduction. Loss of these residues abolishes cAMP signaling for both PACAP-(6 -27) and PACAP-(6 -38) (33,37). In comparison with PACAP-27 the binding affinity of PACAP-(6 -27) is severely reduced. The affinity of PACAP-(6 -38), however, remains high rendering it an effective antagonist. The COOH-terminal extension of PACAP-38 contains a short ␣-helix and a preponderance of cationic amino acids (31). The contribution of the COOH-terminal extension to the binding of PACAP- (6 -38) suggests that it forms a third receptor interaction domain.
The coding exons 3a, 4, and 5 in the extracellular domain of the receptor may provide conformational alterations that affect the relative affinity of the receptor for PACAP-27 and PACAP-38 (Fig. 5). The length of this extracellular domain of the receptor may affect the degree to which the three regions within the PACAP ligands can cooperate. The longer domain (exons 3a, 4, and 5) may enhance cooperation between the COOH-terminal cationic region (region 3 of PACAP in Fig. 5) and the other receptor binding regions of PACAP. The absence of exons 3a, 4, and 5 may enhance the interaction of the PACAP amino-terminal region (region 1 of PACAP in Fig. 5) while weakening the effect of the COOH-terminal region, thus bringing PACAP-27 and PACAP-38 closer together in their receptor binding affinities (20). The smaller increase in affinity of PACAP-(6 -38) for binding to PAC1R(3a) compared with that of PACAP-38 suggests that the amino-terminal activation region of PACAP also plays an important role in the binding to its receptor. The increased binding affinity of PACAP to PAC1R(3a) compared with PAC1R suggests that the aminoterminal activation domain of the receptor also plays an important role in this cooperative binding of PACAP to its receptor.
PAC1R(3a) and PAC1R are coupled to both the cAMP and inositol phosphate signaling pathways. PAC1R(3a) exhibits in-creased EC 50 values of coupling to both pathways in response to both PACAP-27 and PACAP-38 compared with PAC1R. Notably the increased affinity of PACAP-38 binding to the PAC1R(3a) is proportionately reflected in the accumulation of cAMP. This circumstance indicates that the stimulation of cAMP production by the PAC1R(3a) is coupled to cAMP production over the full range of receptor occupancy, whereas in contrast, the PAC1R generates near maximum levels of cAMP at only partial receptor occupancy (Fig. 4B). For example, the PAC1R generates half-maximal levels of cAMP with 10 -20% receptor occupancy by PACAP-38, whereas the PAC1R(3a) receptor requires near 90% receptor occupancy to reach halfmaximal cAMP production and continues to generate increasing amounts of cAMP throughout the entire range of receptor occupancy. With regard to inositol phosphate production, the wild type receptor, PAC1R, begins to generate inositol phosphate at about 80% of receptor occupancy, whereas the variant PAC1R(3a) receptor requires near 100% receptor occupancy to begin to generate inositol phosphates. These observations suggest that at low concentrations of PACAP (10 -50% of receptor occupancy), the PAC1R receptor is coupled exclusively to G s and cAMP formation. At higher concentrations of PACAP the receptor recruits G q , resulting in the activation of phospholipase C and the formation of inositol phosphates. In contrast to PAC1R, the variant-spliced receptor PAC1R(3a) remains coupled predominantly to G s and cAMP formation at all concentrations of PACAP and only begins to recruit G q at very high concentrations of PACAP, equivalent or close to 100% receptor occupancy. It is worth noting that our findings support earlier studies showing that the deletion of exons 4 and 5 enhances receptor coupling to inositol phosphate production relative to cAMP formation (20).
In addition to the apparent differences in coupling to G proteins described above, the PAC1R(3a) differs from the wild type PAC1R in another interesting aspect. The range of PACAP concentrations involved in the generation of both cAMP and inositol phosphate is one to 2 orders of magnitude greater for PAC1R(3a) than it is for PAC1R. For example, for both PACAP-27 and PACAP-38 cAMP formation occurs over a concentration range of 10 Ϫ11 to 10 Ϫ5 M with PAC1R(3a) compared with a range of 10 Ϫ11 to 10 Ϫ7 M with PAC1R. This difference of an order of magnitude in the range of PACAP in the activation of signal transduction between the two receptors further emphasizes the situation that quantitative aspects of signal transduction by the PACAP receptor are altered, or modulated, by the presence of alternatively spliced exon 3a in the extracellular domain of the receptor. Although several other splice variants of the PAC1R are reported to affect the relative coupling to cAMP or inositol phosphate signaling pathways (18,19,38), these splice variations occur in the third intracellular loop, a region that could interact directly with G protein subunits. In instances cited above in which variations occur in the aminoterminal extracellular domain, the effects appear to be mediated by conformational changes in the receptor/ligand interaction.
Several lines of evidence suggest that ligand/receptor interactions can affect coupling of receptors to signaling pathways. There are reported differences in the relative potencies of PACAP-27 and PACAP-38 in stimulating second messenger synthesis. In general, PACAP-38 is reported to be a more effective stimulator of cAMP production than is PACAP-27 (19,20,33), although there are exceptions (18). In two instances it has been reported that PACAP-38 is more effective than PACAP-27 in the activation of phospholipase C (18,20). If alterations in the relative affinities of PAC1R for the two PACAP peptides, PACAP-27 and PACAP-38, indicate a change FIG. 5. Model of PACAP-38 interaction with PAC1R. PACAP-38 contains three potential receptor interaction domains, labeled 1 to 3 from the amino to the carboxyl terminus. Only two of the binding domains can engage concurrently, and only one conformation leads to receptor activation and G protein release. In PAC1R(3a), the second, weak-activating conformation may be favored over the first. Thus, the affinity of PACAP-38 for the PAC1R(3a) is increased without an accompanying increase in second messenger signaling over the entire range of receptor occupancy.
in the way regions within PACAP interact with the receptor, it is possible that these changes also affect the potency of PACAP for stimulating the activation of coupling to different G protein subunits. In at least one system, the glucagon-like peptide-1 hormone, relative activation of two different G protein subunits by one receptor has been shown to be ligand dose-dependent (39). If our experimental model reflects the true biology of the interactions of PACAP peptides with the different spliced forms of the PACAP receptor in the seminiferous tubules during the cycles of spermatogenesis, then the implications of our findings are significant because they indicate that local ambient concentration gradients of PACAP can regulate the recruitment of different signal transduction pathways during the spermatogenic cycle. Ambient concentrations of PACAP are likely to be high at certain stages of the spermatogenic cycle due the stagespecific production of PACAP by round spermatids (2-4, 7). The PAC1R and PAC1R(3a), expressed in round spermatids and Sertoli cells, will therefore experience high concentrations of autocrine and paracrine PACAP, potentially in the micromolar range. Given that PACAP levels are higher in the testis than any other tissue, it is tempting to speculate that these isopeptides are important regulators of the differentiation of germ cells. Concentration gradients established by the stagespecific expression of PACAP, and acting through the PAC1R/ PAC1R(3a), may modulate and coordinate cellular responses in germ and Sertoli cells.