Characterization of novel splice variants of the PAC1 receptor in human neuroblastoma cells: consequences for signaling by VIP and PACAP

Expression of VPAC and PAC1 receptor isoforms was determined in six neuroblastoma cell lines as well as in human embryonic and adult brain using reverse transcriptase PCR and quantitative PCR. PAC1 receptor splice variants missing a 21 amino acid sequence in the amino terminal domain were found to be the major receptor variants in the neuroblastoma cell lines and also were highly expressed in embryonic brain compared to adult brain. In four of the neuroblastoma cell lines, VIP and PACAP stimulated cyclic AMP production with different potencies and levels of maximal stimulation. High potency and greatest maximal stimulation of cyclic AMP for each peptide were recorded in SH-SY5Y cells, indicating the presence of high affinity VIP and PACAP receptors. Further characterization of specific VPAC and PAC1 receptor isoforms was carried out in the SH-SY5Y cell line, where along with known PAC1 receptor splice variants and the VPAC2 receptor, a number of novel PAC1 receptor splice variants were identified. The comparatively low level expression of the VPAC2 receptor along with the poor responsiveness of SH-SY5Y cells to the VPAC2 receptor-specific agonist Ro 25-1553 indicated that this receptor did not contribute significantly to the observed VIP responses. When the individual PAC1 receptor isoforms were expressed in COS 7 cells, the ability of VIP to activate cyclic AMP production was increased more than 50-fold at the majority of the PAC1 receptor variants lacking the 21 amino acid amino terminal domain sequence compared to those with the complete domain. Smaller changes were seen in the potency of PACAP-38. Similar trends were seen with inositol phosphate responses, where in each case agonist potencies were lower than for cyclic AMP production. The results of this study show that the combination of different amino terminal and intracellular loop 3 splicing variants in the PAC1 receptor dictates the ability of agonists, particularly VIP, to activate signaling pathways. VIP has considerably greater potency at most PAC1 receptors with the short amino terminal domain, and these therefore may mediate physiological effects of both VIP and PACAP. Furthermore, there may be a phenotypic switch in the expression of different PAC1 receptor amino terminal splice variants between embryonic and mature nervous system, indicating that regulation of this event may have an important role in VIP/PACAP function, particularly in the developing nervous system.

Specific interaction between the hop1 intracellular loop 3 domain of the human PAC(1) receptor and ARF

The PAC(1), VPAC(1) and VPAC(2) receptors are members of the secretin (Group II) family of G protein-coupled receptors. All members of this family activate adenylate cyclase and several have also been shown to activate phospholipase C. We have recently reported that the rat VPAC(1), VPAC(2) and PAC(1) receptors activate phospholipase D and that distinct pathways are utilised by two intracellular loop 3 splice variants of PAC(1), one of which is ARF-dependent. Phospholipase D activation by the hop1, but not the null (short), form of the PAC(1) receptor is sensitive to brefeldin A, an inhibitor of GTP exchange at ARF. We have expressed the null and hop1 intracellular loop 3 domains of the human PAC(1) receptor in bacteria as GST-fusion proteins and used them as peptide affinity matrices to determine whether a functional interaction exists between these domains and ARF. Using this GST pull-down assay, we have shown binding of the small G protein ARF6 to the hop1 but not the null domain of this receptor.

Additional signals from VPAC/PAC family receptors

The receptors for the neuropeptides vasoactive intestinal polypeptide and pituitary adenylate cyclase-activating polypeptide are strong activators of adenylate cyclase, but recent evidence suggests that they can elicit a number of additional intracellular signals. Some of these are likely to be downstream of the conventional adenylate cyclase pathway, but it is now clear that others reflect novel primary coupling events of the receptors.

ADP-Ribosylation Factor-Dependent Phospholipase D Activation by VPAC Receptors and a PAC1 Receptor Splice Variant

The VPAC(1) and VPAC(2) receptors for vasoactive intestinal polypeptide and the PAC(1) receptor for pituitary adenylate cyclase-activating polypeptide are members of a subfamily of G protein-coupled receptors (GPCRs). We recently reported that phospholipase D (PLD) activation by members of the rhodopsin group of GPCRs occurs by at least two routes, one of which seems to involve the small G protein ADP-ribosylation factor (ARF) and its physical association with GPCRs. Here we report that rat VPAC and PAC(1) receptors can also stimulate PLD (albeit less potently than adenylate cyclase) in transfected cells and also in cells where they are natively expressed. PLD responses of the VPAC receptors and the hop1 spice variant of the PAC(1) receptor but not its null form are sensitive to brefeldin A (BFA), an inhibitor of GTP exchange at ARF. The presence of the hop1 cassette in the rat PAC(1) receptor facilitates PLD activation in the absence of marked changes in ligand binding, receptor internalization, and adenylate cyclase activation, with some reduction in phospholipase C activation. Both VPAC(2) and PAC(1-hop1) (but not PAC(1-null)) receptors were shown to associate with immunoprecipitates directed against native or epitope-tagged ARF. A chimeric construct of the VPAC(2) receptor body with intracellular loop 3 (i3) of the PAC(1-null) receptor mediated BFA-insensitive activation of PLD, whereas the response of the corresponding PAC(1-hop1) construct was BFA-sensitive. Motifs in i3 of the PAC(1-hop1) receptor may act as critical determinants of coupling to ARF-dependent PLD activation by contributing to the GPCR:ARF interface.

Mechanisms of phospholipase C activation by the vasoactive intestinal polypeptide/pituitary adenylate cyclase-activating polypeptide type 2 receptor

The vasoactive intestinal polypeptide/pituitary adenylate cyclase-activating polypeptide type 2 (VPAC(2)) receptor was shown to induce both [(3)H]inositol phosphate ([(3)H]InsP)and cAMP production in transfected COS7 cells and in GH(3) cells where it is natively expressed. Neither cholera toxin nor forskolin could elicit an equivalent [(3)H]InsP response, suggesting independent coupling of the two pathways. The VPAC(2) receptor-mediated [(3)H]InsP response was partially inhibited by pertussis toxin (Ptx) and by the G beta gamma-sequestering C-terminal fragment of GRK2 (GRK2-ct) in COS7 and GH(3) cells, whereas responses of control receptors were unaffected. Blockers of receptor-activated Ca(2+) influx pathways (Co(2+) and SKF 96365) also partially inhibited VPAC(2) receptor-mediated [(3)H]InsP responses. This inhibition was not present in the component of the response remaining after Ptx treatment. A range of blockers of voltage-sensitive Ca(2+) channels were ineffective, consistent with the reported lack of these channels in COS7 cells. The data suggest that the VPAC(2) receptor may couple to phospholipase C through both Ptx-insensitive and Ptx-sensitive G proteins (G(q/11) and G(i/o), respectively) to generate [(3)H]InsP. In addition to G beta gamma, G(i/o) activation appears to require receptor-activated Ca(2+) entry. This is consistent with the possibility that not only G alpha(q/11)-responsive and G beta gamma-responsive isoforms of phospholipase C but also Ca(2+)-responsive forms may contribute to the overall [(3)H]InsP response.

Gonadotropin-releasing hormone receptor activation of extracellular signal-regulated kinase and tyrosine kinases in transfected GH3 cells and in alphaT3-1 cells

GH3 cells were stably transfected with the wild-type murine GnRH receptor and a clonal cell line selected on the basis of inositol phosphate production and PRL/GH release in response to GnRH. This cell line (wt28) was characterized by [125I]GnRH analog binding, [3H]inositol phosphate response to GnRH, and hormone secretion. We examined the activation of the mitogen-activated protein kinase isoforms, extracellular signal-regulated kinase 1/2 (ERK1/2) and tyrosine kinases in wt28 cells and alphaT3-1 cells (which express a native GnRH) using specific phospho-ERK1/2 and phosphotyrosine antibodies. Concentration-response and time-course data revealed that a sustained ERK1/2 response was seen only in aT3-1 cells. Furthermore, GnRH-induced tyrosine phosphorylation was detectable in alphaT3-1 cells, but not in wt28 cells. Activators for several different signaling pathways revealed distinct differences between the cell types. Protein kinase C activation by phorbol 12,13-dibutyrate was very effective in alphaT3-1 cells at phosphorylation of both ERK1/2 and tyrosine, whereas raising cAMP levels using forskolin also strongly increased wt28 cell ERK1/2 phosphorylation. Only the tyrosine phosphatase inhibitor pervanadate increased tyrosine phosphorylation in wt28 cells. The lack of sustained ERK1/2 phosphorylation in wt28 cells could be the result of minimal tyrosine kinase activation by GnRH compounded by a different pathway profile for ERK1/2 activation. When pervanadate and GnRH were combined, ERK1/2 phosphorylation was synergistic and sustained in wt28 cells, whereas the response was additive in alphaT3-1 cells. In sum, the intracellular pathways leading to ERK1/2 and tyrosine phosphorylation in alphaT3-1 and wt28 cells are distinct; thus, activating GnRH receptors in each of the two cell types leads to different sequelae of events regarding ERK1/2 activation.

Effect on gastric and duodenal mucosal prostaglandins of repeated intake of therapeutic doses of naproxen and etodolac in rheumatoid arthritis

The synthesis of gastric and duodenal mucosal prostaglandin E2, prostaglandin I2, and thromboxane B2 during a 60 minute incubation of biopsy specimens, the degree of endoscopic and histological damage, and the anti-inflammatory response were all studied after a four week, double blind study of therapeutic doses of two non-steroidal anti-inflammatory drugs, naproxen and etodolac, received by 27 patients with active rheumatoid arthritis (13 receiving naproxen, 14 etodolac). Prostaglandin values after treatment did not differ from the baseline levels when all the patients were analysed as one group. Subgroup analysis showed that naproxen suppressed gastric prostaglandin E2 from a median of 29 to 9 ng/mg protein, duodenal prostaglandin E2 from 34 to 11 ng/mg, and duodenal prostaglandin I2 from 62 to 15 ng/mg protein. No overall suppression occurred with etodolac. Also, on the second assessment patients receiving naproxen had lower gastric and duodenal prostaglandin E2 and prostaglandin I2, but higher values of duodenal thromboxane B2, than patients receiving etodolac. Both drugs had comparable anti-arthritic activity and caused microscopic gastritis in similar proportions of patients. No correlation was detected between prostaglandin values and the mucosal damage which developed in seven patients receiving naproxen (54%) and three receiving etodolac (21%). These findings indicate that, unlike naproxen, etodolac does not seem to affect gastric or duodenal prostaglandin synthesis; other mechanisms of injury need to be considered.