Molecular cloning of a novel variant of the pituitary adenylate cyclase-activating polypeptide (PACAP) receptor that stimulates calcium influx by activation of L-type calcium channels.

Pituitary adenylate cyclase-activating polypeptide (PACAP) is a novel neuropeptide that produces its biological effects by interacting with G protein-coupled receptors. Molecular cloning of the PACAP receptor revealed the existence of five splice variant receptor forms differing in the third intracellular loop region, with four variants activating both adenylyl cyclase and phosphoinositide phospholipase C and one variant activating only adenylyl cyclase (Spengler, D., Waeber, C., Pantaloni, C., Holsboer, F., Bockaert, J., Seeburg, P. H., and Journot, L. (1993) Nature365, 170-175). Here, we report cloning of a novel PACAP receptor variant, designated PACAPR TM4 (transmembrane domain IV), that differs from the previously cloned short form of the PACAP receptor (PACAPR) primarily by discrete sequences located in transmembrane domains II and IV. Reverse transcriptase-polymerase chain reaction and primer extension analyses demonstrated tissue-specific differential expression of mRNAs encoding PACAPR TM4 and splice variant forms of the PACAP receptor. PACAPR TM4 and PACAPR possess identical intracellular domains, implicated as primary determinants of G protein recognition by rhodopsin-like receptors. However, unlike the PACAPR, PACAPR TM4 does not activate either adenylyl cyclase or phosphoinositide phospholipase C in response to PACAP in either transient or stable expression systems. However, PACAP stimulates increases in [Ca2+]i in cells expressing PACAPR TM4 by activating L-type Ca2+ channels, a response not elicited by stimulation with vasoactive intestinal polypeptide. The signaling phenotype of PACAPR TM4 is characteristic of the PACAP receptor involved in regulation of insulin secretion from pancreatic β islets, a tissue expressing transcripts for PACAPR TM4 but not for PACAPR or its longer splice variant forms. These findings are consistent with a role of PACAPR TM4 in the physiological control of insulin release by PACAP in β-islet cells. The finding that PACAPR TM4 has a unique signaling phenotype, although it possesses intracellular domains identical to those of the PACAPR, suggests that receptor-G protein recognition by rhodopsin-like receptors can be determined by sequences other than those located in intracellular receptor domains.

Pituitary adenylate cyclase-activating polypeptide (PACAP) 1 was first discovered in 1989 as a novel hypothalamic hormone that stimulates increases in cAMP in rat anterior pituitary cells (1). Structural studies revealed the peptide to be a Cterminal amidated 38-amino acid peptide (PACAP- 38), and subsequent studies resulted in identification of the other molecular form of PACAP (PACAP-27) that lacks the last 11 amino acids of the longer form (2). These two bioactive forms of PACAP are derived from a single gene and are formed by proteolytic processing of the encoded PACAP precursor protein (3). PACAP exhibits high sequence homology to VIP, indicating that it is a member of the VIP/secretin/glucagon/growth hormone-releasing factor family of neuropeptides. PACAP-containing nerve fibers are present throughout the central and peripheral nervous system, and the two molecular forms of PACAP exhibit a broad distribution and range of tissue concentrations (3)(4)(5). PACAP possesses, in addition to its role as a hypophysiotropic hormone, a diverse array of biological activities consistent with its suggested role as a neurotransmitter, neuromodulator, and vasoregulator (3,6).
The biological effects of PACAP are mediated through PACAP binding to at least two types of high affinity receptors, the Type I PACAP-preferring receptor and the Type II receptor that does not distinguish between PACAP and VIP and which is also known as the VIP receptor (VIPR) (3,7,8). Molecular cloning of receptors with the pharmacological properties of each of these receptors has revealed that they are members of the G protein-coupled receptor superfamily (9 -15). The recombinant VIPR binds VIP and PACAP with equal affinity and activates adenylyl cyclase (9). The recombinant PACAP receptor (i.e. Type I) binds PACAP with an affinity 1000 times higher than that for VIP (14,15). Spengler et al. (10) identified five splice variants of the rat PACAP receptor differing only in their predicted third intracellular domains, a region implicated in coupling of a variety of receptors to G proteins (16 -22). The short splice variant form of the receptor was designated PACAPR with the longer splice variant forms designated as hip, hop, or hip-hop isoforms of the PACAPR. Four of these splice variants exhibited the multifunctional signaling characteristic of PACAP receptors found in various cells (23)(24)(25), activating both adenylyl cyclase and phosphoinositide phospholipase C (PI-PLC), while one variant activated only adenylyl cyclase. Sequence analysis of the PACAPR shows that it is a member of the new family of G protein-coupled receptors first identified by cloning of secretin, calcitonin, and parathyroid hormone receptors and which now includes receptors for VIP, glucagon, glucagon-like peptide, and growth hormone-releasing factor (9, 26 -31).
The PACAPR is expressed in the central nervous system, pituitary, adrenal medulla, and testicular germ cells (10), while the VIPR is predominant in the lung, liver, and gastrointestinal tract (9). Recently, a third PACAPR subtype was cloned, designated PACAPR-3, which has a tissue distribution distinct from that of the PACAPR and VIPR (32). This new subtype binds PACAP and VIP with equal affinity, like the VIPR, and activates both adenylyl cyclase and phospholipase C. PACAPR-3 is expressed in lung, brain, stomach, colon, and pancreatic ␤-islet cells. Yada et al. (33) recently showed that PACAP is a physiologically occurring neuropeptide in the pancreas and that PACAP is by far the most potent insulin secretagogue known. Expression of PACAPR-3 in ␤-islet cells implicated it in the insulin secretagogue actions of PACAP. However, PACAP stimulates insulin release in ␤-islet cells by interacting with PACAP receptors exhibiting the pharmacological selectivity of Type I PACAPRs and by a mechanism involving increases in [Ca 2ϩ ] i mediated exclusively by activation of L-type calcium channels (33). None of the cloned PACAP receptors exhibit the signaling specificity characteristics of the receptor present in pancreatic ␤-islet cells.
Here, we report isolation of a novel PACAPR variant that differs from the cloned PACAPR primarily in predicted transmembrane domain IV (TM4) and, hence, is designated PACAPR TM4. PACAPR TM4 exhibits unique signaling properties, although it has intracellular domains identical to those of the PACAPR. Unlike all previously cloned PACAP receptors, PACAPR TM4 does not stimulate adenylyl cyclase or phospholipase C. Instead, it stimulates Ca 2ϩ influx through an L-type Ca 2ϩ channel with the pharmacological selectivity of the Type I PACAPR, a signaling phenotype characteristic of the PACAP receptor expressed in ␤-islet cells. The mRNA encoding PACAPR TM4 is expressed in a tissue-specific pattern unique from that of splice variant forms of the PACAPR, and ␤-islet cells express transcripts for PACAPR TM4 but not for PACAPR or its longer splice variant forms. These results suggest that PACAPR TM4 represents the receptor mediating the insulinreleasing actions of PACAP in ␤-islets. The present results suggest further that discrete sequences within transmembrane domains of a rhodopsin-like receptor can determine the specificity of receptor signaling.

EXPERIMENTAL PROCEDURES
Materials-PACAP-27, PACAP-38, and VIP were purchased from Bachem. Fura-2 AM was obtained from Molecular Probes, and nifedipine and lanthanum were from Sigma. myo-[2-3 H]Inositol (16.5 Ci/mmol) and adenosine 5Ј-[␥ 32 P]triphosphate (6000 Ci/mmol) were purchased from Amersham. LipofectAMINE, OptiMEM, and inositol-free Dulbecco's modified Eagle's medium were obtained from Life Technologies, Inc. AmpliTaq was from Perkin Elmer, and pcDNAI vector was from Invitrogen. Cell culture medium and serum were provided by the Diabetes Endocrinology Research Center (the University of Iowa). Oligonucleotide primers and other molecular biological reagents were obtained from the University of Iowa DNA Core. BHK-21 cells (ATCC) were a gift from Dr. Jeffrey Pessin (the University of Iowa).
Isolation of PACAPR TM4 cDNA-A novel PACAPR variant was amplified from oligo(dT)-primed rat cerebellum cDNA using a PCRbased strategy we described previously (34). The PCR strategy used would amplify any variants of the PACAPR sharing homology to the PACAPR at regions complementary to the primers used in PCR. Briefly, a 776-bp 5Ј-cDNA fragment (nucleotides (Ϫ)9 -767) and a 713-bp 3Ј-cDNA fragment (nucleotides 710-1423) of the PACAPR cDNA were amplified by PCR using 5Ј-single-stranded ligation of cDNA end and 3Ј-rapid amplification of cDNA end methods, respectively. Recombinant PCR, i.e. using the overlapping cDNA fragments as template with forward and reverse primers corresponding to nucleotides (Ϫ5)-22 and 1375-1406 of the PACAPR, respectively, resulted in amplification of the expected full-length 1.4-kilobase cDNA product. The amplified cDNAs were cloned into eukaryotic expression vector pcDNAI, and sequence analysis of selected clones revealed the presence of a unique clone, differing from the PACAPR cDNA that we isolated using this approach (34). Double-stranded sequencing of the cloned PACAPR variant was performed by automated fluorescent dideoxynucleotide sequencing by the University of Iowa DNA Core Facility. We designated this new clone as PACAPR TM4 due to its divergence in sequence from the PACAPR primarily in transmembrane domain IV.
RT-PCR Analysis of Tissue Distribution of mRNAs Encoding PACAPR and PACAPR TM4 -Oligo(dT)-primed cDNAs were synthesized from total or poly(A) ϩ RNA using avian myeloblastosis virus reverse transcriptase (Promega) or Superscript H Ϫ reverse transcriptase (Life Technologies, Inc.). The resulting cDNAs were used in PCR with forward and reverse primers corresponding to amino acids 23-29 (DCIFKKE) and 383-388 (FELGLG) of the PACAPR ( Primer Extension Analysis to Identify mRNAs Encoding PACAPR and PACAPR TM4 in Tissues-Oligo(dT)-primed cDNAs were synthesized from poly(A) ϩ RNA (1 g) of rat cerebellum, lung, and pancreatic ␤-islet cells. One-tenth of the cDNA synthesis mixtures were PCR amplified using forward and reverse primers corresponding to amino acids 154 -163 (ALYTVGYSTS) and 307-317 (VIKGPVVGSIM) of the PACAPR (Fig. 1), respectively. These primers are complementary to sequences conserved between the PACAPR and PACAPR TM4 cDNAs and span a region that encompasses the divergent sequences encoding the transmembrane IV region of these two receptor forms. The resulting PCR products were purified (Wizard PCR Preps, Promega), and approximately 50 ng of the DNAs were used as templates for extension of an end-labeled 23-nucleotide primer (0.3 pmol) using Klenow large fragment DNA polymerase I in the presence of 1 mM dideoxy-CTP and 0.1 mM each of dATP, dTTP, and dGTP. Authentic cloned PACAPR and PACAPR TM4 cDNAs were used as positive controls. The extension reaction was performed at 37°C for 10 min, and extension products were separated on a DNA sequencing gel. Extension products were visualized by autoradiography.
Transient Transfections and Measurements of Inositol Phosphate and cAMP Accumulation-BHK cells were plated in 24-well culture dishes at a density of 5 ϫ 10 4 cells/well and allowed to grow for 24 h prior to transfection. Cells were transfected with PACAPR or PACAPR TM4 plasmid DNA (0.5 g/well) using LipofectAMINE (5 l/g DNA) for 16 h according to the manufacturer's protocol. Experiments were performed 48 h following termination of transfection. For measurement of inositol phosphates (IPs), cells were labeled overnight with myo-[ 3 H]inositol (2 Ci/ml) in inositol-free Dulbecco's modified Eagle's medium containing 10% dialyzed fetal bovine serum (Sigma) 30 h following transfection. Incubations with vehicle or agonists were performed for 20 min at 37°C in Earle's balanced salt solution containing 10 mM LiCl for measurements of IPs and for 5 min (also 1 min for experiments shown in Fig. 5C) in Dulbecco's modified Eagle's medium containing 1% bovine serum albumin and 0.5 mM isobutylmethylxanthine for measurements of cAMP. Total IPs were extracted with choloroform:methanol:water (1:1: 0.5) and then separated on Dowex AG1-X8 columns as described by Brown et al. (35). Total IPs were eluted with 1 M ammonium formate, 0.1 M formic acid. IP accumulation is expressed as dpm of IPs/10 5 dpm in the lipid fraction. Cyclic AMP was measured with a commercially available kit (Diagnostic Products).
Isolation of PACAPR TM4 Stable Transfectants and Measurement of [Ca 2ϩ ] i -CHO cells were transfected with PACAPR TM4 plasmid DNA by lipofection as described above. Clonal cell lines expressing PACAPR TM4 were isolated by selection in medium supplemented with 1 mg/ml G418 (Life Technologies, Inc.) followed by replating selected cells at a low density in this selection medium and isolation of colonies using cloning cylinders. Four colonies were expanded and used for measurements of PACAP-mediated increases in [Ca 2ϩ ] i . All four cell lines re-sponded to PACAP with similar increases in [Ca 2ϩ ] i . We selected one of these cell lines to characterize the functional activity of PACAPR TM4. These stable PACAPR TM4 transfectants, designated CHO TM4, expressed 19,200 Ϯ 1,600 125 I-labeled PACAP-27 binding sites/cell with a K D of 18 Ϯ 3.1 nM. The level of expression of PACAPR TM4 in CHO TM4 cells (33 fmol/10 6 cells) is nearly identical to that found in a gonadotrope cell line (35 fmol/10 6 cells) where PACAP stimulates adenylyl cyclase and PI-PLC and increases in [Ca 2ϩ ] i (25). Moreover, this physiologically relevant receptor level is equivalent to levels of receptors exhibiting dual signaling in the group III family of G protein-coupled receptors (32,36,37). Effects of PACAP on IPs and cAMP in control CHO cells and CHO TM4 were performed essentially as described above for transient transfectants. [Ca 2ϩ ] i was measured in individual cells using fluorescent and digital image analysis of Fura-2-loaded cells as described previously (38). Fig. 1 shows the alignment of the deduced amino acid sequence of cDNAs encoding the PACAPR, PACAPR TM4, and VIPR (i.e. the "Type II PACAP receptor"). PACAPR TM4 differs primarily from the PACAPR in predicted transmembrane domain IV where there is a substitution and deletion of two amino acids each leading to a CVTV to SAXX conversion, resulting from replacement of the sequence GTGTAACAGTG (nucleotides 836 -846) of the PACAPR with CAGCA in the PACAPR TM4 cDNA. PACAPR TM4 differs from PACAPR by two other substitutions; a D136N substitution (resulting from replacement of GAT with AAT, nucleotides 406 -408 of the PACAPR cDNA) in the amino-terminal region, which is conserved in the VIPR, and a N190D substitution (resulting from replacement of AAC with GAC, nucleotides 568 -570 of the PACAPR cDNA) in transmembrane domain II. Interestingly, each of these amino acid sequence differences between PACAPR TM4 and PACAPR occur at sites where there is a similar divergence in the VIPR. These sites of divergence are contiguous with sequences conserved between all three receptors. PACAPR TM4 does not differ from the PACAPR in any intracellular regions including intracellular domains 2 and 3 and the C-terminal cytoplasmic domain, regions identified as primary determinants of receptor-G protein recognition in G protein-coupled receptors (16 -22).

RESULTS AND DISCUSSION
The divergence in sequence of the cDNA encoding PACAPR TM4 enabled us to design primers to examine expression of its mRNA in tissues by RT-PCR. Alternative splicing of the PACAP receptor primary transcript has been shown to produce short and long forms of the PACAP receptor that differ in their predicted third intracellular loop (10). Therefore, RT-PCR was used to compare expression of mRNA encoding PACAPR TM4 with that of short and long splice forms of the receptor. As shown in Fig. 2, PCR amplification using primers specific for the splice variant forms of the PACAP receptor (lanes a) produced two major bands corresponding to the size expected for the short form (PACAPR) and hip-or hop-cassette forms of the PACAPR. Amplification using primers specific for PACAPR TM4 (lanes b) produced a single band corresponding to the size expected for this variant. Control experiments using no template DNA or using authentic cloned PACAPR or PACAPR TM4 cDNAs as templates demonstrated the specificity of the primer pairs used in this RT-PCR (see Fig. 2 legend).
Tissue-specific differential expression of mRNAs encoding splice variant forms of the PACAP receptor and PACAPR TM4 was observed. The ratio of amplified products corresponding to the short and hip-or hop-cassette forms of the PACAP receptor varied in different tissues, as reported by Spengler et al. (10). In addition, exclusive amplification of the short and of the hipor hop-cassette forms of the receptor occurred in vascular smooth muscle and liver, respectively, providing clear evidence

FIG. 2. Tissue distribution of mRNAs encoding PACAPR TM4
and that of short and long splice variant forms of the PACAP receptor as determined by RT-PCR. cDNAs from various rat tissues or cells were subjected to PCR using primer pairs specific for (a) short and long splice variant forms of the PACAP receptor cDNA or (b) PACAPR TM4 cDNA as described under "Experimental Procedures." Simultaneously, as negative controls, PCR was performed using each primer pair with no template cDNA ((Ϫ) template), using the primer pair specific for splice variant forms of the PACAP receptor with 10 ng of the cloned PACAPR TM4 cDNA as template (PACAPR TM4) and using the primer pair specific for PACAPR TM4 with 10 ng of a cloned PACAPR cDNA as template (PACAPR*). PACAPR* is designated with an asterisk because it differs from PACAPR by three nucleotides that are located in regions not amplified in this PCR. The small and large PCR fragments on the gel in lanes labeled a are the size expected for PACAPR and its hip and/or hop isoform, respectively, and the product in lanes labeled b correspond to that of PACAPR TM4. of tissue-specific processing of splice variant forms of the PACAP receptor. The results shown in Fig. 2 demonstrate that transcripts encoding PACAPR TM4 also are expressed in a tissue-specific manner and with unique patterns of expression relative to that of mRNAs encoding splice variant forms of the PACAP receptor. Coexpression of mRNAs for PACAPR TM4 and splice variant forms of the PACAP receptor was observed in cerebellum, cerebral cortex, brainstem, vas deferens, and lung. In contrast, transcripts encoding PACAPR TM4 were not detected in spinal cord, heart, liver, kidney, and vascular smooth muscle where transcripts encoding splice variant forms of the PACAP receptor were clearly expressed. The transcript encoding PACAPR TM4 was the only PACAP receptor mRNA detected in ␤-islet cells.
A primer extension assay was developed to assess relative proportions of mRNAs encoding PACAPR and PACAPR TM4 in selected tissues. Rat cerebellum, lung, and pancreatic ␤-islet cell cDNAs were subjected to PCR with primers common to both PACAPR and PACAPR TM4 to amplify DNA segments encompassing the divergent regions existing between these forms of the receptor. The relative contribution of these forms of the PACAPR in the amplified product is expected to be proportional to their respective mRNA levels in these tissues. The amplified products were subjected to primer extension according to the scheme shown in Fig. 3, which will distinguish transcripts encoding PACAPR (the hip, hop, and hip-hop forms of the PACAPR share the targeted sequence of the PACAPR and, therefore, would not be detected individually in this assay) and PACAPR TM4. Fig. 3 shows that this assay reliably de-

FIG. 3. Primer extension analysis to identify transcripts encoding PACAPR and PACAPR TM4 in tissues.
Primer extension analysis was performed on cDNA generated by RT-PCR using oligonucleotide primers to regions conserved between the PACAPR and PACAPR TM4 as described under "Experimental Procedures." The amplified DNA encompassed the divergent regions existing between these two receptor forms. The upper panel shows the sequence of PACAPR and PACAPR TM4 cDNAs targeted for primer extension analysis and the extension primer sequence. The extension reaction was performed with dideoxy-CTP in the deoxynucleotide mixture to terminate the extension reaction at the cytosines (underlined) resulting in formation of 35-and 29-bp products (arrows) for PACAPR and PACAPR TM4 cDNAs, respectively. Authentic cloned PACAPR and PACAPR TM4 cDNAs were used as positive controls. The lower panel shows the resulting autoradiogram of the extension products separated on a DNA sequencing gel. tected the expected extension products from each of the authentic receptor cDNAs and that the major extension product present in cerebellum and lung is the size expected for the PACAPR cDNA, while the major product in ␤-islets corresponds to that of the PACAPR TM4 cDNA. Although RT-PCR analysis demonstrated amplification of the transcript encoding PACAPR TM4 in cerebellum and lung (Fig. 2), these primer extension results suggest they represent minor transcripts relative to that of mRNA encoding PACAPR in these tissues. The opposite appears to be true in ␤-islet cells, in agreement with the exclusive amplification of mRNA encoding PACAPR TM4 by RT-PCR in these cells.
Like other members of the group III family of G proteincoupled receptors, PACAP receptors exhibit multifunctional signaling activity stimulating both adenylyl cyclase and phosphoinositide-specific phospholipase C. Spengler et al. (10) reported minor differences in signaling by splice variants of the PACAP receptor that differ in their predicted third intracellular loop region. All splice variants stimulated adenylyl cyclase in response to PACAP-27 or PACAP-38, although the hip and hip-hop isoforms exhibited 6 -17-fold higher EC 50 values for this effect, and all variants except the PACAPR-hip isoform also stimulated PI-PLC. Therefore, we examined effects of PACAP on cAMP and inositol phosphate production in BHK cells transiently expressing PACAPR and PACAPR TM4. Control experiments summarized in Fig. 4A demonstrated that PACAP-27 stimulated increases in cAMP and inositol phos-phate accumulation in PACAPR transfectants with EC 50 values of approximately 1 and 5 nM, respectively, while VIP was at least 100-fold less potent. In contrast, 100 nM PACAP-27 had no significant effects on cAMP or inositol phosphate accumulation in PACAPR TM4 transfectants, although it stimulated 20 -40-fold increases in these second messengers in PACAPR transfectants in parallel incubations (Fig. 4B). PACAP-38 was also inactive in stimulating increases in either cAMP or inositol phosphate accumulation in PACAPR TM4 transfectants, although it was equipotent to PACAP-27 in producing these responses in PACAPR transfectants (not shown). The absence of effects of PACAP on adenylyl cyclase and PI-PLC in PACAPR TM4 transfectants suggested that this receptor is incapable of activating these signaling pathways and raised the intriguing possibility that it has different signal-transducing capabilities.
Because PACAPR TM4 is the primary, if not exclusive, Type I PACAP receptor expressed in ␤-islet cells (Figs. 2 and 3), we investigated whether this variant possesses the unique signaling activity of the PACAP receptor in ␤-islets. In these cells, PACAP stimulates increases in [Ca 2ϩ ] i , and hence insulin secretion, by a mechanism involving influx of Ca 2ϩ through dihydropyridine-sensitive L-type Ca 2ϩ channels rather than by mobilization of intracellular Ca 2ϩ stores through activation of PI-PLC (33). The molecular structure of this unique PACAP receptor is unknown. Because PACAP has no effects on PI-PLC in BHK cells transiently transfected with PACAPR TM4 (Fig.   FIG. 5. PACAP Fig. 6A). Although our studies in BHK cells transiently transfected with PACAPR TM4 demonstrated the inability of the receptor to mediate activation of PI-PLC, it seemed essential to investigate whether the observed Ca 2ϩ response in CHO TM4 cells was similarly dissociated from PACAPR TM4-mediated PI-PLC activation. Therefore, we examined effects of PACAP on PI-PLC in CHO TM4 and CHO cells by measurements of inositol phosphate accumulation. As shown in Fig. 5B, PACAP-27 (1 M) had no effects on inositol phosphate accumulation in untransfected CHO cells or CHO TM4 cells, whereas stimulation of endogenous purinergic receptors with ATP (10 M) produced expected increases in inositol phosphates in each cell line. These findings demonstrate that PACAPR TM4 mediates increases in [Ca 2ϩ ] i by a mechanism independent of phospholipase C-catalyzed hydrolysis of inositol phospholipids. In contrast, PACAP stimulates increases in [Ca 2ϩ ] i in CHO cells stably expressing the PACAPR by inositol trisphosphate-mediated intracellular calcium mobilization (39). Thus, the functional differences between PACAPR and PACAPR TM4 appear to reflect intrinsic differences in their abilities to recognize G proteins that activate distinct signaling systems. Fig. 5C shows that CHO TM4 cells also did not respond to PACAP-27 (100 nM) with increases in intracellular cAMP. These results confirm our findings in BHK cell transfectants and demonstrate that PACAPR TM4 has a signaling phenotype unique from other cloned PACAP receptors.
Additional experiments were performed to determine how PACAP increases [Ca 2ϩ ] i in CHO TM4 cells. Fig. 6A illustrates the time course of effects of PACAP in CHO cells and CHO TM4 cells. As shown, 5 nM PACAP-27 produced a rapid, transient monophasic increase in [Ca 2ϩ ] i in CHO TM4 cells but had no significant effects in CHO cells at a 200-fold higher concentration. Fig. 6B shows that VIP did not elicit increases in [Ca 2ϩ ] i in CHO TM4 cells, indicating that PACAPR TM4 displays the pharmacological selectivity of the Type I PACAP receptor in mediating this response. In contrast to results obtained in Ca 2ϩ -containing medium, PACAP did not stimulate increases in [Ca 2ϩ ] i when CHO TM4 cells were incubated in Ca 2ϩ -free medium or in Ca 2ϩ -containing medium supplemented with the Ca 2ϩ channel blocker lanthanum (20 M) or the L-type Ca 2ϩ channel antagonist nifedipine (1 M) (Fig. 6C). These results indicate that PACAPR TM4 stimulates increases in [Ca 2ϩ ] i by stimulating Ca 2ϩ influx via a dihydropyridine-sensitive L-type Ca 2ϩ channel. Thus, PACAPR TM4 displays the signaling phenotype of the PACAP receptor in ␤-islet cells (33), a tissue expressing mRNA encoding PACAPR TM4 but not for splice variant forms of the PACAPR (Figs. 2 and 3). Of considerable interest is the mechanism by which PACAPR TM4 stimulates L-type Ca 2ϩ channels. These channels are known to be activated by the stimulatory G protein G s and via phosphorylation by protein kinase A (40). Our results demonstrate that PACAPR TM4 does not stimulate adenylyl cyclase in either transient or stable expression systems, indicating that the receptor does not couple to G s . Moreover, the absence of increases in cAMP in cells expressing PACAPR TM4 was not a result of influx of Ca 2ϩ because cells expressing this receptor did not increase cAMP in response to PACAP in incubation medium containing 1 M nifedipine (not shown). Therefore, the stimulation of the L type-Ca 2ϩ channel by PACAPR TM4 appears to occur by a mechanism that does not involve coupling to G s and subsequent increases in cAMP.
These results show that PACAPR TM4 has unique signal transduction properties compared with those of PACAPR. This finding was unexpected because both receptors possess identical intracellular domains, implicated as primary determinants of coupling of rhodopsin-like receptors to G proteins and, hence, of receptor signaling specificity. The involvement of intracellular domains of rhodopsin-like receptors in signaling specificity can be rationalized on a conceptual basis because G proteins are intracellularly located, and, therefore, receptors that regulate them presumably do so via their intracellular sequences. This has been borne out by a variety of studies using various approaches to study the role of these intracellular domains in signaling that have ascribed the specificity of G protein coupling to sequences within the third intracellular loop but also to sequences in the second intracellular loop and C-terminal cytoplasmic tail (16 -22). However, the present results suggest that regions other than these domains are responsible for determining the specificity of receptor-G protein coupling of PACAPR TM4, which differs from PACAPR only by discrete sequences located in transmembrane domains II and IV and the extracellular domain of the receptor. It is unclear from the present studies whether these sequence differences produce differences in receptor signaling as a result of indirect conformational changes in G protein-coupling domains of the receptor or by direct physical interactions of membrane-spanning domains of the receptor with G proteins, known to be associated closely with the plasma membrane (41,42).
In summary, the present study documents the existence of a novel PACAP receptor variant that has a signaling phenotype distinct from that of all previously cloned PACAP receptors. This receptor, designated PACAPR TM4, does not mediate activation of adenylyl cyclase or PI-PLC but stimulates increases in [Ca 2ϩ ] i by stimulating Ca 2ϩ influx via a dihydropyridine-sensitive L-type Ca 2ϩ channel. Our results provide evidence that mRNA encoding PACAPR TM4 is expressed in a tissue-specific fashion unique from that of the splice variant forms of the PACAPR. The primary, or exclusive, expression of transcripts encoding PACAPR TM4 in ␤-islet cells, along with the fact that this receptor has a signaling phenotype like that of the PACAP receptor in ␤-islet cells mediating insulin secretion, suggests that PACAPR TM4 represents the receptor involved in physiological control of insulin release in ␤-islet cells by PACAP.