Differential Effects of Peptide Histidine Isoleucine (PHI) and Related Peptides on Stimulation and Suppression of Neuroblastoma Cell Proliferation

The growth rate of rodent embryonic neuroblasts and human neuroblastoma cell lines is regulated in part by autocrine or paracrine actions of neuropeptides of the family that includes vasoactive intestinal peptide (VIP), peptide histidine isoleucine (PHI), and pituitary adenylate cyclase-activating peptide (PACAP). These peptides act via seven transmembrane G-protein-linked receptors coupled to cAMP elevation, phospholipase C activation, intracellular Ca2+ release, and/or of mitogen-activated protein (MAP) kinase activation. Here we investigated the action of these peptides on the mouse neuroblastoma cell line Neuro2a. PHI and VIP inhibited proliferation at concentrations as low as 10−13 m and 10−10 m, respectively. In contrast, PACAP action was biphasic, with stimulation occurring at subnanomolar doses and inhibition at higher doses. Peptide actions were studied further by measuring cAMP and ERK1/2 MAP kinase activity and by assessing3H-thymidine incorporation in conjunction with a panel of signal transduction pathways inhibitors. The data obtained indicated that the PHI-inhibitory and PACAP-stimulatory activities were mediated by corresponding changes in activity of the MAP kinase pathway and independent of protein kinase A (PKA) or protein kinase C (PKC). In contrast, the inhibitory actions of VIP and PACAP were specifically blocked by antagonists of PKA. Northern blot analysis revealed gene expression for only the PACAP-preferring (PAC1) receptor. However, binding experiments using 125I-labeled PACAP27, PHI, and VIP, demonstrated the presence of PACAP-preferring sites, bivalent VIP/PACAP sites, and PHI-binding sites that did not interact with VIP. The studies demonstrate potent regulatory actions of PACAP, PHI, and VIP on neuroblastoma cell proliferation which appear to be mediated by multiple subsets of receptors which differentially couple to MAP kinase and PKA signaling pathways.

The vasoactive intestinal peptide (VIP) 1 neuropeptide family includes VIP, pituitary adenylate cyclase-activating peptide 27 , the carboxyl-terminal extended isoform PACAP-38, peptide histidine/isoleucine (PHI), and its human homologue peptide histidine/methionine (PHM), secretin, glucagon, growth hormone releasing factor (GRF), and also analogues isolated from reptile venom (1). Expression of the VIP and PACAP genes results in the biosynthesis of peptide precursors, which also give rise to PHI/PHM and PACAP-related peptide (PRP), respectively (2,3). The neuropeptides VIP, PHI/PHM, and PACAP are commonly expressed and secreted in neuroblastoma and certain neuroendocrine tumors (4). Moreover the finding of functional VIP and PACAP receptors expressed in neuroblastoma cell lines and freshly excised tumor resections suggests that these neuropeptides may be involved in autocrine/paracrine loops (5,6). Although these peptides are well known to act in the processes of neurotransmission (1,7,8), they have been shown to regulate the proliferation rate and differentiation of numerous cell lines (9). These include those derived from neuroblastoma (10 -15), glioblastoma (16), and pheochromocytoma (17) and from a variety of epithelial tumors including breast, intestine, lung, pancreas, and prostate (9). In contrast, much less information is available concerning physiological actions of PHI and PRP (18 -21) on either neurotransmission or proliferation.
High affinity receptors for VIP-related peptides have been identified in numerous tissues from several species. At least three different receptors have been characterized in terms of sequence, peptide affinity profiles, tissue expression, and signaling pathways. These belong to the seven-transmembranespanning domain, G-protein-coupled receptor family. Two of these recognize VIP and PACAP with equal high affinity, VPAC 1 and VPAC 2 (22,23), and the third is a PACAP-preferring receptor, PAC 1 (24). 2 In addition, at least seven splice variants have been reported for the PACAP receptor (25)(26)(27)(28). Cloned and expressed forms of these variants have been reported to differ in their ability to bind PACAP analogs and in their coupling to signaling pathways. In contrast, no splice variant has yet been described for VIP receptors. VPAC 1 and VPAC 2 , and PAC 1 receptors are commonly expressed alone or in combination in neuroblastomas (6,29). The cAMP-dependent pathway mediates many of the actions of VIP and PACAP, but many recent reports indicate the involvement of nitric oxide, phospholipase C (PLC), phosphatidylinositol 3-kinase (PI3-kinase), and MAP kinase pathways (7, 8, 18, 25, 30 -32). In this study, we studied the ability of VIP, PACAP-27, PACAP-38, and PHI to modulate proliferation of the mouse neuroblastoma cell line Neuro2a. We also determined if changes in cAMP content and MAP Kinase activity were associated with these proliferative actions, and if these actions were sensitive to specific pathway inhibitors (33)(34)(35). Finally, we analyzed receptor expression by Northern analysis and by radiolabeled ligand displacement.
Proliferation Studies-The actions of VIP-related peptides (purchased from Sigma) on cell proliferation were initially investigated using a tetrazolium-derived (MTS) colorimetric assay (cell titer 96 from Promega) as described previously (17). In brief, subcultures were performed in 96-well plates seeded with 8,000 cells/well. 24 h after seeding, vehicle or peptides were added from 10ϫ stock solutions. After 48 h of peptide treatment, the MTS proliferation assay was initiated by incubation of the cells in 100 l of fresh medium for 1 h at 37°C, followed by addition of 20 l of reagent solution to each well. After incubation at 37°C for 90 min, absorbency at 490 nm was determined on a 96-well plate reader (Bio-Rad).
For 3 H-thymidine incorporation assay, cells were seeded in 24-well plates (80,000 cells/well) in 1 ml of medium and cultured for 24 h. Medium was then replaced with fresh serum-free medium. After 1.5 h at 37°C, cells were preincubated for an additional hour with vehicle or signal transduction inhibitors (H89 and GF109203x from Calbiochem, and PD98059 from New England Biolabs). Peptides were then added, and 1 h later, 1 Ci/well 3 H-thymidine was also added. Four hours later, 3 H-thymidine incorporation was determined by TCA precipitation.
cAMP Measurements-Cells (80,000/well) were cultured for 3 days in 24-well plates. Medium was replaced with serum-free medium with or without 1 M isobutylmethylxanthine (IBMX). After 15 min incubation at 37°C, peptides were added and incubated for 15 min. Cell lysis and cAMP radioimmunoassay (NEN Life Science Products) were performed as described (17).
Northern Analysis-Neuro2a cells were cultured in eight flasks (75 cm 2 ) and harvested at 90% confluency with 0.05% trypsin, 0.02% EDTA. Trypsin was neutralized with the tissue culture medium (containing 8% serum), and then cells were centrifuged and rinsed with phosphate-buffered saline. Subsequent isolation of poly(A)-selected mRNA was carried out as described (36). Receptor probes and hybridization conditions were exactly as described previously (29).
Radioiodination of PHI-Because 125 I-PHI is not commercially available, PHI was radioiodinated with the chloramine-T technique as described previously (37). Iodinated peptide was separated by reversephase HPLC (Spectraphysics) using a 5 m VYDAC C18 column (Interchrom, France). A gradient of acetonitrile, H 2 O, 0.1% trifluoroacetic acid was required for elution. Fractions containing 125 I PHI were evaporated under nitrogen and stored at -20°C.
Binding Studies-Cells were seeded at 80,000/ml/well in 24-well dishes and cultured for 3 days. Fifteen min before the initiation of the binding, cells were incubated in fresh medium at 37°C. Culture medium was removed and replaced with 270 l of cold binding medium containing 30,000 cpm of radiotracer ( 125 I-VIP and 125 I-PACAP, 2200 Ci/mmol, from NEN Life Science Products) and 30 l of specified concentrations of unlabeled peptides (for more details, see Ref. 13). Radioiodinated PHI binding experiments were performed on intact cells in suspension, using a slightly different procedure to increase the sensitivity. In brief, nonconfluent cultures were harvested by incubation for 5 min at 37°C in a phosphate buffer (0.05 M, pH 7.4) containing 0.53 mM EDTA.
Cells were counted and aliquoted into Eppendorf tubes at a concen-tration of 1.2 million cells per tube. Cells were then quickly centrifuged, pellets were reconstituted in the binding buffer containing 35,000 cpm of 125 I-PHI and specified concentrations of the unlabeled ligands, and then cells were incubated under shaking. The incubation times were determined by time course experiments conducted at 4°C and then fixed at 120, 150, and 210 min for VIP, PACAP, and PHI displacements, respectively. Incubation was stopped with a quick rinse with 2 ml of cold phosphate-buffered saline containing 0.1% bovine serum albumin, followed by cell lysis in 0.5 M NaOH solution. Extracts were transferred into 5-ml tubes, and radioactivity was counted using a ␥-counter (Wallac). MAP Kinase Assays-Cells were cultured for 24 h in 24-well plates (150,000 cells/well). After a quick rinse, cells were incubated for 12 h in the presence of culture medium containing 0.2% fetal bovine serum. Cultures were then rinsed twice with serum-free medium and incubated in this medium for 120 min. PD98059 (30 M) or vehicle was added for another hour and then peptides for 10 min at 37°C. The incubation was stopped by replacement of the stimulation buffer with 100 l of lysis buffer (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 2 mM EGTA, 2 mM dithiothreitol, 1 mM orthovanadate supplemented with 1 mM phenylmethylsulfonyl fluoride, 1 M pepstatin, and leupeptin, aprotinin, and antipain (10 g/ml each) (Sigma). ERK activity was determined by the Biotrack p42/p44 MAP kinase enzyme assay kit according to the recommendation of the manufacturer (Amersham Pharmacia Biotech). Protein was determined by the Bradford assay with Bio-Rad reagents and bovine serum albumin as a standard.

RESULTS
Cell Proliferation-Both MTS (Fig. 1) and thymidine incorporation assay (Fig. 2) demonstrated that VIP and PHI were potent inhibitors of Neuro2a cell proliferation. The MTS assay, carried out at the end of a 2-day period of drug stimulation, indicated that maximal inhibition by VIP was reached at about 10 nM, whereas PHI action reached near maximal values at concentrations in the subnanomolar range (Fig. 1). PACAP-27 and PACAP-38 induced a biphasic response: a significant stimulation at very low concentrations (0.1 and 1 pM, respectively) and then an inhibition at the highest doses of peptides. In subsequent experiments, thymidine incorporation over a 5-h period after initiation of drug treatment was used to assess the effect of peptides on DNA synthesis. This assay produced qualitatively similar results, although the effects were generally shifted to higher concentrations ( Fig. 2A).
Various inhibitors were added to characterize the possible signaling pathways mediating the proliferative actions. a positive proliferative response that could otherwise only be seen with lower concentrations of PACAP. H89 also blocked the antiproliferative action of VIP but clearly did not block the inhibitory action of PHI. Nearly identical effects on neuropeptide responses were seen when cultures were treated with another PKA inhibitor, RT-cAMP (200 M, data not shown). In contrast to the effects of PKA inhibitors, treatment with the MEK1/2 inhibitor PD98059 (30 M) (34) completely suppressed the PACAP-induced stimulation ( Fig. 2A versus 2C). Paradoxically PD98059 inhibited the antiproliferative action of PHI but appeared not to affect the antiproliferative actions of VIP or high concentrations of PACAP. The blockade of PHI action by PD98059 implies that inhibitory action of this peptide required a basal level of MAP kinase activity. DNA synthesis was in fact inhibited by PD98059 in control cultures (Fig. 2C, second column) implying that a MAP kinase pathway coupled to growth stimulation was indeed active in these cells under basal conditions. The PKC inhibitor GF109203X (10 M) (35) also strongly inhibited 3 H-thymidine incorporation in control cultures, but did not appear to interfere with the neuropeptide modulation of DNA synthesis ( Fig. 2A versus 2D).
cAMP Response-Intracellular cAMP content was specifically and dose-dependently increased by addition of PACAP-38, PACAP-27, and VIP, either in the absence (Fig. 3A) or presence (Fig. 3B) of the phosphodiesterase inhibitor IBMX (1 M). No elevation of the cAMP levels was detected when cells were stimulated by PHI alone (Fig. 3A). Indeed, only a slight increase in cAMP content was observed in the presence of IBMX at the highest concentrations of PHI.
MAP Kinase Activity-Changes in ERK1/2 MAP kinase activity were measured at the selected peptide concentrations shown in Fig. 4. For PACAP-27, the peak of activation was reached at 1 pM, whereas peak activation by PACAP-38 was shifted to about 0.1 nM. In contrast, VIP did not modify MAP kinase activity at any concentration from 1 pM to 1 M. PHI at 1 nM, on the other hand, significantly inhibited MAP kinase activity by about 25%. The MEK1/2 inhibitor PD98059 reduced ERK1/2 activity in control cells (Fig. 4, second bar), indicating that some basal activity of the MAP kinase pathway existed in these cells, perhaps from active compounds in serum (medium contained 0.2% FBS, see "Materials and Methods"). Consistent with this possibility, the PHI action was more pronounced (78% inhibition) when experiments were carried out in medium containing 2% FBS (data not shown).
Expression of VIP and PACAP Receptor Genes-Northern analysis indicated that the PACAP-preferring PAC 1 receptor gene was expressed in Neuro2a cells, whereas neither VPAC 1 nor VPAC 2 receptor gene expression could be detected under these conditions (Fig. 5, lane 1). The specificity and integrity of hybridization was tested using internal positive and negative Radioligand Binding-Binding sites for VIP and PACAP in Neuro2a cells were assessed, using commercially available radiotracers, and for PHI, using chloramine T-labeled radioiodinated peptide (see "Materials and Methods"). Preliminary studies of time-course binding demonstrated that the optimized incubation times at 4°C were 120, 150, and 210 min for 125 Ilabeled VIP, PACAP, and PHI, respectively. Preliminary saturation experiments and Scatchard analyses were also per- Displacement curves of 125 I-PHI binding by different analogues (native PHI, VIP, and PACAP isoforms) also suggested a multiple site system in Neuro2a cells. Three different PHI binding sites appeared to be present based on the capacity of these analogues to displace the binding of 125 I-PHI (Fig. 6C). Displacement by native PHI resulted in curve characterized by a Hill slope only 0.33 (IC 50 0.2 Ϯ 0.45 nM). VIP displaced only 55% of total 125 I-PHI binding with IC 50 51 Ϯ 1.65 pM (Hill slope: 0.71). PACAP-27 and PACAP-38 displaced almost completely (87%) the PHI binding. Data fit to a two-site model indicated high affinity sites (IC 50 PACAP27, 4.7 Ϯ 2.1 pM; PACAP38, 40 Ϯ 2.36; representativity: 56%) and low affinity binding sites (IC 50 PACAP-27, 61 Ϯ 5.21 nM; PACAP-38, 60 Ϯ 6.54 nM; representativity: 44%). Finally about 13% of radioiodinated PHI was displaced only by native PHI. DISCUSSION We demonstrate here that VIP, PHI, and PACAP potently regulate the proliferation of mouse neuroblastoma Neuro2a cells. VIP and PHI inhibited proliferation, PHI being considerably more potent. PACAP, on the other hand, stimulated proliferation at low concentrations and inhibited proliferation at higher concentrations. These observations led us to hypothesize that multiple receptors were present in these cells, coupling to different growth-related signaling pathways. An analysis of proliferation and intracellular responses to peptides in the presence and absence of specific signaling pathway inhibitors, and an assessment of binding sites using radiolabeled peptides strongly supported this hypothesis.
Of particular interest was the anti-proliferative action of PHI. Although PHI acts as a moderate to weak agonist on all well known forms of VIP receptors, its anti-proliferative action in the present system was more than 100-fold more potent than VIP. One explanation for this finding is that PHI acted on a unique receptor that preferred PHI to VIP. Consistent with this possibility was the fact that these peptides appeared to act by different signaling pathways. The action of VIP (but not PHI) was associated with an increase in cAMP levels and was selectively blocked by inhibitors of PKA. In contrast, PHI action was uniquely associated with ERK1/2 inhibition and was selectively blocked by a MAP kinase pathway inhibitor. Finally, binding studies using 125 I-PHI as a tracer indicated the presence of up to two classes of high affinity receptors that selectively bound PHI over VIP.
We also found the action of PACAP to be intriguing. Very low concentrations of this peptide (up to 0.1 nM) stimulated MAP kinase activity and proliferation, whereas higher PACAP concentrations, like VIP, stimulated cAMP production and inhibited proliferation. MAP kinase and PKA pathways were very likely involved in these opposing proliferative responses. Whereas the MEK1/2 inhibitor blocked the stimulation of proliferation observed at low PACAP doses (Fig. 2C), blockade of PKA resulted in stimulation of proliferation at all PACAP doses (Fig. 2B). Biphasic responses to peptides in this family have been reported in other systems. For example, VIP at 50 pM stimulated c-src activity and the proliferation of cultured retinal pigmented epithelial cells, whereas higher concentrations stimulated cAMP production and inhibited proliferation (38). In addition, biphasic action of VIP on cell growth has been reported for Lo-Vo colonic carcinoma cells (39) and T98G glioblastoma cells (17).
The Northern analysis revealed expression of only the PACAP-preferring PAC 1 receptor gene, whereas binding studies suggested the presence of multiple receptors that bind these peptides. One possible explanation is that receptors arising from alternatively spliced forms of PAC 1 differentially bind and respond to these peptides. Several forms have been characterized (25)(26)(27)(28), including one displaying a 21-amino acid deletion in the amino-terminal part of the receptor believed to be important for ligand binding. A cloned form of this latter receptor has been shown to differ slightly from the original receptor with respect to the relative binding affinities for PACAP-27 versus PACAP-38 (27). However, the ability of this variant to bind VIP, PHI, or related peptides has not been reported. We tried to detect alternative transcripts corresponding to this region of the PAC 1 receptor by reverse transcription-polymerase chain reaction (27). Only the original form of receptor was identified in these experiments (data not shown). Alternatively, a low level of VPAC 1 receptor expression (not detectable by Northern analysis) could potentially explain some of the actions of VIP and/or PHI because the rat VPAC 1 binds PHI with an affinity only 3-fold lower than VIP. However, the considerably more potent action of PHI on growth (Fig. 1) and its selective coupling to the MAP kinase pathway suggest the presence of a novel high affinity receptor that preferentially binds PHI versus VIP.
Because very high affinity PHI receptors are not well described in the literature, we radioiodinated PHI to investigate the pharmacological properties of the PHI binding sites. Displacement studies were consistent with the presence of polyvalent PACAP/VIP/PHI sites (estimated to be 40% of the total) and up to two other receptors, PACAP/PHI (50%) and possibly PHI-preferring (10%), whose molecular characteristics have not been determined. The presence of a VPAC 2 receptor seems unlikely because the VPAC 2 -selective agonist helodermin (23) was unable to displace the binding of 125 I-VIP or 125 I-PACAP (data not shown). In the context of our data suggesting the existence of novel receptors, a recent report indicating that VIP family peptides interact with "clearance receptors" for the seemingly unrelated atrial natriuretic peptide (ANP) family of neuropeptides is of interest (40). These receptors were named "clearance" because early attempts to characterize these ANP binding sites did not reveal any signaling function, suggesting that they might be involved in removal of the peptide. Interestingly, ANP interaction with these single transmembrane domain receptors in smooth muscle cells triggered inhibition of MAP kinase activation and growth inhibition (41). Because PHI also inhibits MAP kinase pathways and growth in Neuro2a cells, one could speculate that a "clearance-like" receptor could mediate the activity induced by PHI in Neuro2a cells.
In conclusion, we demonstrated here that at least three closely related peptides of the neuropeptide VIP family regulated the proliferation of the mouse neuroblastoma Neuro2a cells. These actions appeared to occur through interaction with at least three different functional receptors that were differentially coupled, positively or negatively, to the PKA and MAP kinase signaling pathways. Moreover, the combined actions of these neuropeptides on proliferation may be relevant to the specific behavior of neuroblastoma tumor since an autocrine/ paracrine action for these neuropeptides has been proposed (5,6,10,11,29).