C-terminal amidation of PACAP-38 and PACAP-27 is dispensable for biological activity at the PAC1 receptor

Abstract

PACAP-27 and PACAP-38 are the exclusive physiological ligands for the mammalian PAC1 receptor. The role of C-terminal amidation of these ligands at that receptor was examined in neuroendocrine cells expressing the PAC1 receptor endogenously and in non-neuroendocrine cells in which the human and rat PAC1 receptors were expressed from stable single-copy genes driven by the CMV promoter, providing stoichiometrically appropriate levels of this Gs-coupled GPCR in order to examine the potency and intrinsic activity of PACAP ligands and their des-amidated congeners. We found that replacement of the C-terminal glycine residues of PACAP-27 and -38 with a free acid; or extension of either peptide with the two to three amino acids normally found at these positions in PACAP processing intermediates in vivo following endoproteolytic cleavage and after exoproteolytic trimming and glycine-directed amidated, were equivalent in potency to the fully processed peptides in a variety of cell-based assays. These included real-time monitoring of cyclic AMP generation in both NS-1 neuroendocrine cells and non-neuroendocrine HEK293 cells; PKA-dependent gene activation in HEK293 cells; and neuritogenesis and cell growth arrest in NS-1 cells. The specific implications for the role of amidation in arming of secretin-related neuropeptides for biological function, and the general implications for neuropeptide-based delivery in the context of gene therapy, are discussed.

1. Introduction

G protein-coupled receptors (GPCRs) activated by polypeptide ligands are ubiquitous throughout the GPCR superfamily. They are represented within all but one of the five classes of GPCRs (rhodopsin, secretin, adhesion, and frizzled), and within all of the four sub-groups of the largest of these classes, the rhodopsin-like GPCRs [42]. The peptide and polypeptide ligands for these receptors perform critical physiological roles as hormones, chemokines and neurotransmitters, and these ligand-receptor dyads are therefore the targets of intense basic and translational medical research.

A highly idiosyncratic feature of the biological activity of peptides, especially neuropeptides, at their receptors is the requirement for post-translational modification. Peptide ligands are typically synthesized from so-called prohormone precursors by proteolytic cleavage, glycosylation, N-terminal glutamate cyclization, and C-terminal amidation within secretory vesicles prior to hormone/neurotransmitter secretion; and additional sulfation, phosphorylation, acetylation, and lipidation prior to receptor engagement. Which of these modifications are required for receptor recognition and engagement, and which are required for ligand synthesis, trafficking, stability, or other ancillary functions, is known for some, but not all, of the mammalian peptide GPCR ligands. Among the best-studied neuropeptides, for example, the endorphins do not require C-terminal amidation for biological activity [25]. On the other hand, C-terminal amidation is necessary for the biological activity of substance P [17], neurokinin A [33], neuropeptide Y [46], as well as vasopressin and oxytocin [38].

Within the PACAP/PAC1-containing 15-member secretin receptor family, the role of C-terminal amidation in conferring receptor recognition and intrinsic activity has been examined and found to be critical for biological activity of corticotropin-releasing hormone [44] and parathyroid hormone [34], yet dispensable for the actions of growth-hormone-releasing hormone [47], glucagon-like peptide 1 [43], secretin [19], and vasoactive intestinal peptide [18].

Curiously, the effects of des-amidation on biological receptor engagement have yet to be examined for PACAP. Since PACAP is a key modulator of neurodegeneration, migraine, stress-related affective disorders, and tissue resistance to ischemia, and is specifically transported both to and from the CNS via specific transporters at the blood-brain-barrier [5], close attention to the structural requirements for its biological actions, and the potential therapeutic actions of peptide-based agonists and antagonists, is of specific relevance to translational efforts centered on this neuropeptide and its receptors. In particular, the lack of non-peptide inhibitors for PAC1 and related receptors has pushed the search for therapeutic approaches into the realm of peptide and peptoid bioequivalents and even gene therapy. Thus, for the PACAP/PAC1 dyad, considerations of peptide structure-activity relationships have assumed paramount importance, and resolution of the question of the need for C-terminal amidation in the biological activity of PACAP is necessary to move forward in this field.

2. Materials and Methods

2.1. Materials

Peptides were purchased from AnaSpec (Fremont, CA) and were prepared as 20 µM stocks in media. Forskolin was purchased from Tocris (Ellisville, MO) and was prepared as a 50 mM stock in DMSO. Neuroscreen-1 (NS-1) cells were from Cellomics (Pittsburg, PA), HEK293 cells were from the American Type Culture Collection (Manassas, VA), HEK293 CRE-luciferase cells were from Promega (Madison, WI), and HEK293T cells used to produce retroviral vectors were obtained from Cell Genesys Inc. (Foster City, CA). Solutions for cell culture were purchased from Invitrogen (Carlsbad, CA) unless otherwise specified.

2.2. Cell culture

NS-1 cells were grown in Roswell Park Memorial Institute (RPMI) 1640 media supplemented with 10% horse serum (HyClone, Piscataway, NJ), 5% heat-inactivated fetal bovine serum (Atlanta Biologicals, Flowery Branch, GA), 2 mM L-glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin. All vessels used for NS-1 cell culture were coated with a solution containing 50 µg/ml collagen type I from rat tail in 20 mM glacial acetic acid. After one hour, the collagen solution was removed, plates were rinsed with PBS, dried, and used immediately or stored at 4° C. HEK293 cells were grown in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin. Cell cultures were maintained at 37° C in a humidified incubator containing 5% CO₂. Cells were used between passages three and 25 for the experiments reported here and routinely tested negative for mycoplasma.

2.3. Molecular biology and retroviral transduction

Constructs used in these studies were introduced by retroviral transduction. pPRIG-rPAC1hop is a bicistronic γ-retroviral expression vector we constructed previously [22]. A human PAC1hop receptor with a C-terminal V5 epitope tag was encoded in the pLx304 lentiviral vector (GeneCopoeia, Rockville, MD), which confers resistance to blasticidin. An untagged hPAC1hop construct encoded in the Lv215 lentiviral vector (GeneCopoeia), containing an IRES eGFP and puromycin resistance, was used for comparison. For phenotypic assays, eGFP expression was introduced using the psi-LVRH1GP lentiviral vector (GeneCopoeia), which also confers puromycin resistance.

A luminescent cAMP biosensor fragment was PCR amplified from the pGloSensor-22F plasmid (Promega) using the following primers: 5'-ACCGGGACCGATCCAGCCTCCGCGGCCCCAATGCCTGGCGCAGTAGGC-3' (sense) and 5'-TGGAGACTAAATAAAATCTTTTATTTTATCGTTAAACCCCTTCTGGAGTGATC-3' (antisense) with Fusion high-fidelity DNA polymerase (New England Biolabs, Ipswich, MA) following the manufacturer’s protocol. Using a Sequence and Ligation Independent Cloning (SLIC) method [26], this amplicon was then assembled into a γ-retroviral vector (pLHCX) with flanking HindIII and ClaI sites to construct the pLHC-CBS retroviral vector.

To generate vector particles, HEK293T cells, cultured in 10 cm dishes, were co-transfected with endotoxin-free preparations of three plasmids: pVSV-G env, pMLV gag/pol, and the genome encoding the construct of interest, using a calcium phosphate method (Profection, Promega), according to the instructions provided by the manufacturer. Vector-containing supernatants were collected 48 hours post-transfection, filtered, titered, and used immediately or stored at −80 °C. For transductions, NS-1 or HEK293 cells were cultured in 24-well plates and exposed overnight to vector particles that were diluted to a multiplicity of infection less than one. The following day, cells were spilt 1:2 into media containing the appropriate selection antibiotics for the generation of stable cell lines. Antibiotic concentrations used were 200 µg/ml hygromycin-B (Invitrogen), 1 µg/ml puromycin (Invivogen, San Diego, CA), and 1 µg/ml blasticidin S (Invivogen).

2.4. Cyclic AMP biosensor (CBS) assays

Cyclic AMP biosensor measurements were conducted based on an established protocol [11] with the following modifications. CBS-expressing NS-1 cells were seeded in 96-well assay plates with opaque walls (Costar, catalog # CLS3603), coated with collagen type I, at a density of 1×10⁴ cells per well in CO₂ independent media (Invitrogen, catalog # 18045-088) supplemented with 10% horse serum, 5% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin. CBS-expressing HEK293 cells were seeded in at a density of 1×10⁴ cells per well in CO₂ independent media supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin. Following incubation overnight, media were aspirated and media containing 1.5 mg/ml of D-luciferin potassium salt substrate (Gold Biotechnology, St. Louis, MO) was added (100 µl per well). Cells were equilibrated in media containing this substrate at room temperature for two hours. Background luminescence was then measured using a multi-label plate reader (Victor3, Perkin Elmer). Following background readings, test compounds were added. Real-time luminescent readings of treated cells were obtained every 2 minutes for 20 minutes following agonist addition. Measurements taken 15 minutes after the addition of agonists were used for end-point assays.

2.5. Cyclic AMP response element (CRE) luciferase assays

HEK293 stably expressing a cyclic AMP response element (CRE)-driven luciferase reporter gene were transduced with retroviral vectors encoding PAC1 receptors as described previously [16]. Cells were seeded in 96-well plates at a density of 1×10⁴ cells per well in DMEM containing 1% FBS, 2 mM L-glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin overnight. The following day, cells were treated with agonists. After treatment for four hours, luciferase activity was determined by the addition 100 µl per well of Bright Glo luciferase assay substrate (Promega). Two minutes after substrate addition, luminescence was measured in a Victor3 multilabel reader (Perkin Elmer).

2.6. Reverse transcriptase polymerase chain reaction (RT-PCR)

NS-1 cells were grown in 24-well plates overnight and RNA was extracted the following day using the RNAqueous kit (Ambion, Grand Island, NY) following the manufacturer’s protocol. Samples were then digested with DNAse I (Invitrogen), pelleted, and resuspended in nuclease-free water. First strand cDNA synthesis was carried out on 1 µg of RNA from each sample using SuperScript II Reverse Transcriptase (Invitrogen) following the manufacturer’s instructions. Rat PAC1 receptor isoforms were amplified from 1 µl cDNA samples using AmpliTaq Gold reagents (Thermo Fisher) and the following primers: PAC1 forward, targeting transmembrane domain five, sequence: 5'-GGCCCCGTGGTTGGCTCTATAATGG-3', PAC1 reverse, targeting transmembrane domain six, sequence: 5'-GAGAGAAGGCGAATACTGTG-3', PAC1hop reverse, targeting the hop cassette, sequence: 5'-AGAGTAATGGTGGATAGTTCTGACA-3' and PAC1hip reverse, targeting the hip cassette, sequence: 5'-TGGGGACTCTCAGTCTTAAA-3'. PCR conditions were as follows: 95° C for 5 minutes, followed by 35 cycles of denaturation (95° C for 15 seconds), annealing (55° C for 30 seconds), and extension (72° C for one minute). There was a final extension (72° C for 7 minutes) at the end of the reaction. Samples were loaded on gels (1.8% agarose in Tris-acetate-EDTA buffer) stained with 100 ng/ml ethidium bromide. Gels were illuminated with UV light and photographed using a digital camera. Amplicon sizes were calculated from molecular weight ladders using AlphaView software (Protein Simple, San Jose, CA).

2.7. High-content analysis (HCA) assays

NS-1 cells that stably express eGFP were seeded in collagen type I-coated 96-well plates (TPP) in complete growth medium at a density 1 × 10³ cells per well. Two hours post-seeding, plates were photographed using an inverted fluorescence microscope outfitted with motorized objectives, shutters, filters, stage, automated focusing and focus correction (TiE Eclipse, Nikon). One phase-contrast and fluorescent image was captured in the center of each well of the plate during each acquisition, which required approximately three minutes per plate. Following each acquisition, plates were returned to an incubator. Eighteen hours after plating, cells were treated with peptides as indicated. Images were then acquired every 24 h for five days. Cell number in each field was determined by Nikon NIS-Elements HC Software. Cell growth data were normalized to values observed in each well two hours post-plating and are expressed as a percent of cell growth over time. Neuritogenesis was measured manually using NIS-Elements by tracing all neurites in each field captured 48 hours following addition of PACAP. Neurite outgrowth data were normalized as a ratio of total neurite length measured per field to number of cells per field.

2.8. Statistical analysis

Dose-response data were fit to curves using four-parameter logistic regressions. When appropriate, means EC₅₀ values from multiple groups were compared by one-way ANOVA followed by Bonferroni-corrected t-tests. In experiments where a receptor antagonist was used, agonist efficacy (intrinsic activity) was determined by factorial ANOVA followed by Holm-Šídák-corrected comparisons at each concentration of agonist in the absence or presence of antagonist. Curve fitting and statistical testing were done using Sigma Plot (Systat, San Jose, CA). Statistical significance was determined by P < 0.05.

3. Results

PACAP, the endogenous agonist of the PAC1 receptor, exists in vivo as either a 27 or 38 residue peptide (PACAP-27 or -38), depending on extent of processing of its prohormone precursor (Figure 1). To confirm that these two forms of PACAP are similar in both potency and intrinsic activity at the PAC1 receptor in a cell-based assay, as determined previously in cell-free receptor binding assays [7], we examined PACAP-induced cyclic AMP elevation in neuroendocrine cell line NS-1 that natively expresses PAC1 receptors but do not express the highly homologous VPAC1 or VPAC2 receptors [15]. For real-time measurements of cyclic AMP production, NS-1 cells were transduced with a retroviral vector encoding a cAMP biosensor (CBS) that catalyzes the oxidization of luciferin to yield a luminescent signal that is proportional to cAMP levels. CBS-expressing NS-1 cells were treated with varying concentrations of PACAP-27 or -38 and cAMP-induced luminescence was measured every two minutes following agonist addition for a total of 23 minutes. As seen in Figure 2, PACAP-27 and -38 promoted cAMP elevation with indistinguishable potencies and efficacies. Since PACAP exerted a maximal, steady-state, effect on cAMP levels 15 minutes following agonist addition, this time point was used for end-point analyses in further experiments using this cell line.

Figure 1. PACAP processing.

PACAP-38 with a c-terminal Gly Arg Arg extension is derived from prepro-PACAP by prohormone convertases (PC). The two Arg residues are cleaved by carboxypeptidase E/H (CPE), and the exposed Gly is used by peptidylglycine alpha-amidating monooxygenase (PAM) to amidate a Lys reside to yield PACAP-38. A series of Gly Lys Arg residues in PACAP-38 allows for further processing by PC, CPE, and PAM, to yield amidated PACAP-27.

Figure 2. Kinetic determinations of PACAP-induced cyclic AMP elevation in NS-1 cells. NS-1 cells stably expressing a luminescent cyclic AMP biosensor (CBS) were treated with varying concentrations of PACAP27 or PACAP38.

Cyclic AMP levels were determined in living cells at the indicated time points. Data points are means from three replicates with error bars corresponding to the SEM. Curves were fit to data using 4-parameter logistic regressions. (B) Data were replotted from individual time-course panels depicting 30 nM PACAP-27 or -38 induced cyclic AMP responses over the course of the assay.

To further validate this method as suitable for detailed pharmacological studies of the PAC1 receptor, CBS-expressing NS-1 cells were used to confirm the ability of PACAP(6–38) to competitively inhibit cAMP elevation caused by PACAP-27 and -38. As seen in Figure 3, both PACAP-27 and -38 were potent stimulators of cAMP elevation (EC₅₀ values of 310 + 40 pM and 360 + 110 pM, respectively). Exposure to 100 nM PACAP(6–38) effectively decreased the potency of both PACAP-27 and -38 by approximately five-fold: the EC₅₀ values in the presence of PACAP(6–38) were 1.86 + 0.36 nM for PACAP-27 1.43 + 0.27 nM for PACAP-38. On the other hand, PACAP(6–38) did not significantly reduce the maximal efficacy of either PACAP-27 or -38, suggesting that it is indeed a competitive PAC1 receptor antagonist as has been reported previously [37].

Figure 3. Functional evaluation of PACAP(6–38), a PAC1 antagonist.

CBS-expressing NS-1 cells were pretreated with 100 nM PACAP(6–38) for 10 minutes followed by the addition of PACAP27 (P27) or PACAP38 (P38) for 15 minutes. In the absence of PACAP(6–38), EC₅₀ values obtained for PACAP27 and PACAP38 were 0.31 + 0.04 nM and 0.36 + 0.11 nM, respectively. Pretreatment with 100 nM PACAP(6–38) caused an approximately 5-fold reduction in the potencies PACAP27 (EC₅₀ = 1.86 + 0.36 nM) and PACAP38 (EC₅₀ = 1.43 + 0.27 nM). Statistical analysis of inhibition by PACAP(6–38) was determined by 2-way ANOVA followed by Holm-Šídák post hoc comparisons for each concentration of PACAP-27 or -38 in the absence or presence or PACAP(6–38). Significant inhibition of PACAP27-induced responses is annotated as ##P<0.01, ###P<0.001. Inhibition of PACAP38 is indicated by *P<0.05, **P<0.01, ***P<0.001; NS, not statistically significant. Data points are means from three experiments performed in triplicate with error bars corresponding to the SEM. Data points are means from three experiments performed in triplicate with error bars corresponding to the SEM.

To determine whether C-terminal amidation is required for the biological activity of PACAP, CBS-expressing NS-1 cells were treated with PACAP38-amide (PACAP-38), PACAP38-free acid (desamido-PACAP38), or PACAP38 congeners with C-terminial extensions of one (Gly) or three (GlyArgArg) residues (PACAP38-G or PACAP38-GRR). The efficacy and potency of PACAP-38 did not vary significantly from its three congeners (Figure 4A). We next compared the effects of PACAP27-amide (PACAP-27) on cAMP elevation to the effects of desamido-PACAP27 (PACAP27-free acid), and PACAP27 with a C-terminal Gly extension (PACAP27-G). As seen in Figure 4B, these three analogs had similar potencies: PACAP27 promoted cAMP elevation with an EC₅₀ value of 0.82 + 0.25 nM. PACAP27-G was slightly more potent than PACAP27 (0.37 + 0.09 nM), while PACAP27-free acid was slightly less potent (EC₅₀ = 1.63 + 0.26 nM). The pooled EC₅₀ values calculated for each compound were compared by a one-way analysis of variance and were not significantly different (P > 0.05). Each PACAP analog tested promoted cyclic AMP elevation that followed an identical kinetic response profile that was elicited by its respective amidated congener. Taken together, these data support the notion that C-terminal amidation of PACAP is dispensable for its activity at the rat PAC1 receptor.

Figure 4. Activity of PACAP analogs at the rat PAC1 receptor.

CBS-expressing NS-1 cells were treated for 15 minutes with varying concentrations of (A) PACAP38 (P38), PACAP38-free acid (P38-fa), or PACAP38 with C-terminal extensions of Gly (P38-G) or Gly Arg Arg (P38-GRR). (B) NS-1 cells were treated with PACAP27 (P27), PACAP27-free acid (P27-fa), or PACAP27 with a C-terminal Gly extension (P27-G). Data points are means from three experiments performed in triplicate with error bars corresponding to the SEM. Within each experiment, data were normalized to the maximal response obtained.

Several splice variants of the PAC1 receptor have been identified in human and rat tissue. Of particular importance to the pharmacological and signaling properties of the receptor, a splice donor site located within the third intracellular loop allows for the inclusion of two exons (termed hip and hop) that can be included in this region individually, together (termed hiphop), or not at all, thus allowing for expression, in a given cell, of the PAC1 IC3 variants PAC1hip, PAC1hop, PAC1hiphop, and PAC1null (where neither the hip nor hop cassettes are present) [36, 41]. Using RT-PCR to determine the relative expression level of the transcripts encoding the PAC1 variants hip, hop, hiphop, and null in NS-1 cells, we found that PAC1hop is the predominantly expressed isoform (Figure 5). To determine unambiguously that the predominant PAC1 isoform possesses the properties observed by signaling through the ensemble of receptors present in NS-1 cells, we introduced a retroviral rPAC1hop construct into HEK293 that stably express a cAMP-response element-driven luciferase reporter gene (CRE-luciferase). As seen in Figure 6A, PACAP38 was equipotent to its unamidated and extended congeners at rPAC1hop receptors. We next introduced lentiviral constructs encoding hPAC1hop receptors with or without a C-terminal V5 epitope tag into CRE-luciferase-expressing HEK293 cells. As seen in Figure 6B, the agonist profile of PACAP-38 was similar at rPAC1hop, hPAC1hop, and hPAC1hop-v5. The parental CRE-luciferase-expressing HEK293 cell line fails to yield a luminescent signal in response to PACAP treatment, indicating the absence of functional expression of native PAC1, VPAC1, or VPAC2 receptors (A.C.E. and L.E.E., unpublished observations).

Figure 5. PAC1hop is the dominant PAC1 receptor isoform expressed in NS-1 cells.

RNA from NS-1 cells was reverse transcribed and cDNAs encoding PAC1 receptor isoforms were amplified by PCR. PAC1 primers targeted all isoforms – lower MW band corresponds to PAC1null and higher MW band corresponds to transcripts containing the hip or hop cassette. Amplification with primers specific to the hop or hip cassettes (lanes 3 and 4) suggest that hop is most abundantly expressed PAC1 receptor isoform in NS-1 cells.

Figure 6. Analysis of PAC1hop pharmacology and potential species differences in transduced HEK293 cells.

HEK293 cells stably expressing a luciferase reporter gene (luc2P) under the control of a minimal herpes simplex virus thymidine kinase promoter with five tandem cyclic AMP response elements (CRE-luciferase) were transduced with rat PAC1hop receptors encoded in the pPRIG retroviral vector followed by cloning by limiting dilution. (A) rPAC1hop-expressing cells were treated with PACAP38 and its analogs for four hours followed by determining luciferase activity. (B) Comparison of HEK293 CRE-luc cell lines expressing rPAC1hop, hPAC1hop, or hPAC1hop with a C-terminal v5 epitope tag (hPAC1hop-v5). Cells were treated with varying concentrations of PACAP38 for four hours followed by determination of CRE-driven luciferase activity. Data were normalized to responses observed following treatment with 25 µM forskolin, which caused an average 258-fold increase in luminescence across all assays performed.

To determine whether C-terminal amidation is required for the biological activity of PACAP-38 at the human PAC1 receptor, HEK293 stably co-expressing CRE-luciferase and hPAC1hop were treated with varying concentrations of PACAP38-amide (PACAP38), desamido-PACAP38 (PACAP38-free acid), and PACAP38 with either Gly or GlyArgArg C-terminal extensions. As seen in Figure 7A, none of PACAP38 congeners significantly differed from PACAP38-amide, or one another, in terms of potency or efficacy. Similarly, the effects of PACAP27-amide, PACAP27-free acid, and PACAP27-Gly on hPAC1hop-dependent CRE-luciferase did not differ significantly (Figure 7B), although PACAP27-amide (EC₅₀ = 32.97 + 1.08 nM) was somewhat less potent than either the free acid (EC₅₀ = 8.9 + 1.4 nM) or the glycine-extended compound (EC₅₀ = 9.3 + 0.57 nM) at the human PAC1hop receptor in this assay.

Figure 7. Activity of PACAP analogs at hPAC1hop receptors.

HEK293 cells stably expressing CRE-luciferase were transduced with human PAC1hop receptors encoded in the pLx304 lentiviral vector followed by antibiotic selection. (A) hPAC1hop-expressing cells were treated with PACAP38 and its analogs desamido-PACAP38 (PACAP38-free acid), PACAP38 with a C-terminal Gly-extension (PACAP38-G), or PACAP38 with a C-terminal Gly Arg Arg extension (PACAP38-GRR) for four hours followed by determining luciferase activity. (B) Cells were treated with PACAP27, desamido-PACAP27 (PACAP27-free acid), or PACAP27 with a C-terminal Gly extension (PACAP27-G). Data were normalized basal values obtained in untreated controls.

PACAP-induced cyclic AMP elevation causes neuroendocrine cells, such as NS-1 cells, to differentiate in a process that includes the extension of neurites and growth arrest [9, 15]. We wished to determine whether α-amidation of PACAP-38 affected its potency or efficacy to cause these long-term phenotypic changes. As seen in Figure 8A, 48 hours of treatment with PACAP38 or desamido-PACAP38 resulted in dose-dependent neurite elongation. The two congeners did not vary significantly in their potencies to elicit neuritogenesis: EC₅₀ values are 19.0 + 7.5 nM for PACAP38 and 10.6 + 1.2 nM for desamido-PACAP38. Likewise, both congeners were quite similar in their intrinsic activities: E_MAX values are 65.2 µm/cell for PACAP38 and 69.5 µm/cell for desamido-PACAP38. Automated fluorescence microscopy (Nikon TiE) was used to monitor cell growth in order to measure PACAP-dependent growth arrest. We first validated a software-based cell counting procedure (Nikon NIS Elements HCA) by comparing cell counts obtained by manually counting the same 288 fields (Figure 8B) as were analyzed by the software. The results from the two methods correlated well (Pearson’s r = 0.98, P < 0.000001), so automated cell counting was used for growth arrest assays. To determine whether α-amidation influenced growth arrest, NS-1 cells were seeded in 96 well plates and imaged two hours later to obtain a correction value for cell number per well. Varying concentrations of PACAP38 or desamido-PACAP38 were applied 24 hours after plating and cells were imaged daily for five days. As seen in Figures 8C and 8D, both congeners of PACAP38 caused growth arrest of a similar magnitude at a similar apparent potency.

Figure 8. Comparison of PACAP38-amide and PACAP38-free acid to promote neuroendocrine cell differentiation.

NS-1 cells were seeded in 96-well plates, grown overnight and treated with PACAP38-amide (P38-A) or PACAP38-free acid (P38-fa) at the indicated concentrations. (A) Following 48 hours of exposure to peptides, cells were imaged by automated microscopy. For each field, cells were manually counted and neurite length was measured. Data are expressed as mean neurite length (µm) per cell, n = 8 per condition. (B) To compare the concordance between manual and a software-enabled cell counting protocol, three 96-well plates of NS-1 cells treated with varying concentrations of PACAP were imaged at various time points. Cell number in each field was determined by both methods and results were correlated, Pearson’s r = 0.98, P<0.000001. To monitor growth arrest, NS-1 cells were treated with either PACAP38-amide (C) or PACAP38-free acid (D). Cells were treated with varying concentrations of PACAP (8 replicate wells per condition) 24 hours after plating. Cells were imaged by automated fluorescence microscopy over the course of five days at the indicated times. Data are expressed as the number of cells counted in each well two hours after plating. Data points represent means and error bars correspond to the SEM.

4. Discussion

The neurotransmitter PACAP is present in both a fully processed 27 amino acid form, and a partially processed 38 amino acid form. Both peptides are C-terminally amidated. PACAP-38 is the predominant form found in the mammalian brain, generally representing more than 90% of total PACAP content throughout the hypothalamus, where its content is highest, and in the remainder of the brain, and the gut, testis, adrenal, and other peripheral tissues [3, 20, 21].

Several groups have synthesized analogs of PACAP and reported on their biological activities. Mutational analyses of PACAP to determine the residues conferring its status as the only high-affinity ligand for the PAC1 receptor have been carried out [39]. Additionally, metabolically stabilized PACAP analogs have been developed and tested in vivo for neuroprotection from ischemia following middle cerebral artery occusion [8] and cytoprotection of pancreatic islets following streptozotocin treatment [32]. These studies have been valuable in determining the minimal epitope(s) required for biological activity. However, as in each case PACAP analogs tested have been C-terminally amidated, the question of whether the latter feature is required for high-affinity interaction at the PAC1 receptor has remained open.

The work reported here was initiated to fill this gap in our current knowledge of the structural requirements for PACAP’s action at the PAC1 receptor. Using whole-cell signaling assays for engagement of Gs-coupled signaling through the PAC1 receptor in the NS-1 neuroendocrine cell line, we have determined that both des-amidated and C-terminally extended PACAP-27- and PACAP-38-based peptides are fully active, in comparison to the amidated endogenous peptides, for dose-dependent cAMP elevation, as well as downstream signaling for cAMP-dependent gene activation. Since these neuroendocrine cells express multiple isoforms of the PAC1 receptor, including both the major PAC1hop [23, 28, 48] as well as the PAC1null and PAC1hip variants of the receptor [35, 41], we furthermore expressed both the human and rat PAC1hop receptors in the HEK293 non-neuroendocrine cell line. Here we obtained similar results, indicating that des-amidation and C-terminal extension are well-tolerated at this receptor when expressed in isolation from its multiple variants, as well. Since PAC1hop is likely to mediate the bulk of PAC1-dependent signaling in both central and peripheral nervous systems and in neuroendocrine cells of the adrenal medulla, endocrine pancreas, and other tissues, these results suggest that engagement of PAC1 receptors in vivo can be fully mimicked by nonamidated congeners of PACAP-27 and PACAP-38.

These results provide needed information about the requirements for C-terminal amidation among family B secretin-like receptors. Thus, at this writing, evidence exists for an absolute requirement for peptide amidation for activation of the CRF1 receptor [44], while there is apparently no requirement for amidation for biological activity at the PAC1 receptor, the GHRH receptor [47], the GLP-1 receptor [43], the secretin receptor [19], or the VPAC1/2 receptors [18]. These results raise the obvious question of why the amidation pathway is conserved among family B neuropeptides regardless of its efficacy in ligand-receptor coupling. Several possibilities exist, including a role in peptide stabilization from enzymatic destruction following hormonal or synaptic release, and structural requirements for peptide transport across the blood-brain barrier. Peripherally administered PACAP-27 and -38 both access the brain, although the two peptides appear to cross the blood-barrier through different mechanisms. Due to its lipophilicity, PACAP-27 enters the brain through transmembrane diffusion, while the less lipid-soluble and larger PACAP-38 is taken up by a carrier-mediated action, peptide transport system (PTS)-6. [5]. In contrast, PTS-6 is responsible for the efflux of both PACAP-38 and -27 from brain into plasma [4]. Efflux of PACAP-27, but not PACAP-38, is mediated by a component of PTS-6, an enzyme called β-F1 ATPase, isolated by the Banks laboratory [12]. Whether or not C-terminal amidation facilitates PACAP transport to and from the CNS is an open question, which might be usefully examined, in view of our findings that amidation is apparently not required for biological action of PACAP at the PAC1 receptor.

A translational issue that is re-cast by our data is the potential mode of gene therapeutic delivery of PACAP, or closely related peptide analogs, in vivo, in the context of neuroprotection [24], tumor suppression [30], treatment of stress-related affective disorders [29], PTSD [10], and migraine [1]. Therapeutic application of neuropeptides for these and other disorders is highly dependent on the need for amidation, since the latter can occur only in secretory granule-releasing tissues [13, 14], and cannot be programmed within DNA-based expression vectors if the latter are expressed in non-neuroendocrine tissues. The neuropeptide NPY, for example, has been administered in CNS as an alternative to treatment with non-peptide anti-seizure agents [45], since CNS-permeant Y2R/Y5R non-peptide agonists do not yet exist [6]. However, delivery is limited to vector injection into central neurons [31], where proNPY can be appropriately processed to fully active (amidated) NPY. PACAP now joins the list of neuropeptides such as catestatin and secretoneurin which can be engineered for expression from either neuroendocrine or non-neuroendocrine cells, upon arming with an N-terminal sequence enabling constitutive secretion [2, 40]. Given the paucity of non-peptide-based ligands with selectivity for the PAC1 receptor, the ability to administer peptide-based PACAP agonists and antagonists in a gene therapeutic mode, without concern for prohormone processing, should usefully accelerate such efforts [27].

Highlights.

• In vivo PACAP is a C-terminally α-amidated neuropeptide of 27 or 38 residues

• PACAP doesn’t require α-amidation to activate the rat or human PAC1 receptor

• Amidation of PACAP is also dispensable for neuroendocrine cell differentiation

Abbreviations

cAMP

cyclic adenosine 3',5'-monophosphate

CPE

carboxypeptidase E/H

CRE

cyclic AMP response element

GPCR

G protein-coupled receptor

NPY

neuropeptide Y

NS-1

Neuroscreen-1

PACAP

pituitary adenylate cyclase-activating polypeptide

PAC1

PACAP type 1 receptor

PAM

peptidylglycine alpha-amidating monooxygenase

PC

prohormone convertase

PKA

protein kinase A

PTS-6

peptide transport system 6

VIP

vasoactive intestinal polypeptide

VPAC1/2

vasoactive intestinal polypeptide receptors 1 and 2