Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Vasoactive intestinal peptide (VIP) is a 28-amino acid peptide widely distributed in both the central nervous system and peripheral tissues where it functions as a neurotransmitter and neuromodulator (Dockray, 1994). VIP is involved as a neural regulator of pancreatic and intestinal secretion, gastrointestinal motility, and blood flow, and also functions in a similar capacity in the cardiovascular, respiratory, and urogenital tracts (Dockray, 1994). VIP also has important growth effects (Muller et al., 1995), effecting the growth of cancers (prostate, lung, pancreatic, breast) (Moody, 1996; Jiang et al., 1997; Moody et al., 1998) as well as normal tissues (Gozes and Brenneman, 2000). Recent reports propose a role of VIP in a number of clinically important areas, including asthma (Bolin et al., 1995); treatment of impotence (Sandhu et al., 1999); pain transmission (Dickinson and Fleetwood-Walker, 1999); pathogenesis of vascular headaches (Edvinsson, 2000); having a neuroprotective effect that could be used to treat Alzheimer's disease (Gozes and Brenneman, 2000) or Parkinson's disease (Offen et al., 2000); and having potent anti-inflammatory effects that could be useful in the treatment of septic shock, Crohn's disease, and rheumatoid arthritis (Gomariz et al., 2001; Said, 1996). In addition VIP receptors are highly expressed in a number of cancers, including those of the breast and gastrointestinal tract, and it has been shown these can be used to localize these tumors by imaging methods and, therefore, their presence may be important in various antitumor treatments (Virgolini, 1997).

VIP is structurally similar to pituitary adenylate cyclase-activating peptides, secretin, peptide histidine isoleucine, glucagon, glucagon-like peptides, gastric inhibitory polypeptide, and growth hormone-releasing factor (Dockray, 1994). Two receptors that share 50% homology, VPAC₁ and VPAC₂, mediate the actions of VIP (Ulrich et al., 1998). Both receptors are members of the G protein-coupled, heptahelical superfamily (Ulrich et al., 1998) but these differ in their distribution (Waschek et al., 1995; Reubi et al., 2000) and pharmacology (Harmar et al., 1998).

The true role of VIP in the above-mentioned processes is poorly understood for a number of reasons. These include a lack of potent antagonists, lack of simplified analogs that are resistant to degradation and could be used in vivo, and a lack of understanding of the VIP pharmacophore for its receptor to use to design simplified analogs. Although there are numerous studies of the pharmacology of VIP and related peptides (O'Donnell et al., 1991; Jensen, 1994; Bolin et al., 1995; Nicole et al., 2000), these studies provide limited information on the VIP pharmacophore. The principal reason is that except for one study (Nicole et al., 2000), all of these studies were performed before the recognition that two subtypes of VIP receptors mediate VIP's actions and are frequently present in the same tissue (Usdin et al., 1994; Waschek et al., 1995; Ulrich et al., 1998; Ito et al., 2000). Furthermore, many studies were performed using receptor preparations from guinea pigs and rodents. In some studies (Robberecht et al., 1988; O'Donnell et al., 1991; Bolin et al., 1995), but not others (Leroux et al., 1994; Ito et al., 2000), the VIP pharmacophore in these species is reported to be similar to human receptors. Even in studies on VIP receptor pharmacology performed since the VIP receptor structures have been identified, by using transfected VIP receptors, it has not been established the pharmacology revealed by the transfected cell systems used faithfully reflects that seen with the native VIP receptor. Studies demonstrate that the entire molecule is required for high-affinity interaction with VIP receptors (Jensen, 1994). Furthermore, in vivo VIP has a half-life of less than 1 min (Domschke et al., 1978). Developing simplified, stable VIP analogs with high affinity for VIP receptors would be useful as prototypes to develop more stable analogs, selective receptor analogs that interact with one of the two G protein-coupled receptors mediating VIP's actions (Harmar et al., 1998), or for in vivo physiological and pathological studies. However, to eventually develop such analogs it is essential to understand VIP's pharmacophore in humans and different species commonly used in experimental studies.

Therefore, the goal of the present study was to determine the VIP pharmacophore for the VPAC₁ receptor, which is the predominant subtype in cancers and widely distributed in normal tissues (Reubi et al., 2000) by assessing its interaction with the human VPAC₁ and the VPAC₁ from different species commonly used in laboratory studies. To accomplish this it was necessary to develop cell systems expressing the VPAC₁-R whose pharmacology reflected that of the native receptor. We identified human, rat, and guinea pig cells containing only or >90% native VPAC₁-R and prepared different cell lines that stably expressed VPAC₁-Rs and determined the VIP pharmacophore in these VPAC₁-R-containing cells by performing alanine and D-amino acid scans of VIP. Our results demonstrated that there were significant species differences in the VIP pharmacophore for the VPAC₁ receptor and that the type of cell used for the stable expression of the VPAC₁-R has a significant effect on the VIP pharmacophore. Furthermore, our results identify at least nine amino acids in VIP that we could replace with alanines to develop a simplified analog that was metabolically stable and retained high affinity for the VPAC₁ in human, rat, and guinea pig.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. Male guinea pigs (100-150 g) were obtained from the Charles River Laboratories, Inc. (Wilmington, DE). Male Sprague-Dawley rats (80-100 g) were obtained from Taconic Farms (Germantown, NY). CHO cells, PANC1 human pancreatic cancer cells, H508 human colon cancer cells, T98G human glioblastoma cells, and Sup T₁ human lymphoblastoma cells were obtained from American Type Culture Collection (Rockville, MD). T47D human breast cancer cells were a gift from Dr. Terry W. Moody (Cell and Cancer Biology Department, Medicine Branch, National Cancer Institute, Rockville, MD). Porcine VIP was purchased from Bachem Biosciences (King of Prussia, PA); purified collagenase (type CLSPA) from Worthington Biochemicals (Freehold, NJ); basal Eagle's medium amino acid mixture, basal Eagle's medium vitamin solution, fetal bovine serum, and LipofectAMINE transfection reagent from Invitrogen (Carlsbad, CA); geneticin (G418 sulfate) from Mediatech (Herndon, VA); pcDNA 3.1(+) and pcDNA 3.1() from Invitrogen; bacitracin, soybean trypsin inhibitor, and 1,3-dimethylxanthine (theophylline) from Sigma Chemical (St. Louis, MO); bovine serum albumin (BSA) fraction V from ICN Biomedicals (Aurora, OH); ¹²⁵I-VIP (2200 Ci/mmol) from PerkinElmer Life Sciences (Boston, MA); Na¹²⁵I (2200 Ci/mmol) from Amersham Biosciences, Inc. (Piscataway, NJ); 1,3,4,6-tetrachloro 3-, 6-diphonylglycouril (IODO-GEN) from Pierce Chemical (Rockford, IL); and Phadebas amylase test reagent from Pharmacia Diagnostics (Piscataway, NJ). The standard incubation solution contained 24.5 mM HEPES, pH 7.45, 98 mM NaCl, 6 mM KCl, 2 mM KH₂PO₄, 5 mM sodium pyruvate, 5 mM sodium fumarate, 5 mM sodium glutamate, 2 mM glutamine, 11.5 mM glucose, 0.5 mM CaCl₂, 1 mM MgCl₂, 1% (w/v) BSA, 0.2% (w/v) soybean trypsin inhibitor, 1% (v/v) amino acid mixture, and 1% (v/v) essential vitamin mixture.

Preparation of Peptides. Native VIP has an alanine in position 4 and 18, so the remaining 26 amino acids were replaced sequentially. Therefore, 26 VIP analogs with a single alanine substitution in VIP numbered from the amino terminal amino acid position 1 to position 28 in the carboxyl terminus of VIP were synthesized. In addition, 28 VIP analogs with a single D-amino acid substitution in VIP numbered from the amino terminal amino acid to position 28 in the carboxyl terminus of VIP were synthesized. Synthesis was performed using standard solid phase methods. Homogeneity of the peptides was assessed by thin layer chromatography and analytical reverse phase HPLC, and purity was at least 97% for each peptide. Using the same methods, one VIP analog with multiple alanine substitutions ([Ala^{2,8,9,11,19,24,25,27,28}]VIP) was also synthesized and purified. [Lys¹⁵, Arg¹⁶, Leu²⁷]VIP(1-7)GRF(8-27) (a VIP analog selective for VPAC₁-R) (Harmar et al., 1998; Ito et al., 2000) and Ro 25-1553 (a cyclic VIP analog selective for VPAC₂-R) (O'Donnell et al., 1994; Harmar et al., 1998; Ito et al., 2000) were also synthesized using a similar procedure.

Transfection of CHO Cells with Human VPAC₁-R (hVPAC₁-R) and Rat VPAC₁-R (rVPAC₁-R), and PANC1 Cells with rVPAC₁-R and Selection of Stable Transfectants. Construction of the rVPAC₁-R, hVPAC₁-R expression vector was described previously (Ito et al., 2000, 2001). Construction of hVPAC₁-R or rVPAC₁-R stably transfected CHO or PANC1 cells (hVIP₁-R or rVIP₁-R/CHO or PANC1 cells) was described previously (Ito et al., 2001). To prepare stably transfected CHO and PANC1 cells containing hVPAC₁-R or rVPAC₁-R, after transfection with lipofectAMINE, individual colonies were isolated and expanded, and cloned cells were screened for VIP-R expression by receptor binding of ¹²⁵I-VIP. For rat and human VIP-Rs in each cell type at least four clones were isolated and binding of ¹²⁵I-VIP assessed in more detail by dose-inhibition curves for VIP and related peptides. For each cell type and species the affinities of the different clones were similar for VIP and the other peptides tested. The clone showing the highest binding was selected for additional studies.

Cell Culture. The hVIP₁-R/CHO and rVIP₁-R/CHO cells were grown in HAM medium supplemented with 10% (v/v) fetal bovine serum, 1% (v/v) antibiotics, and 300 µg/ml G418. The rVIP₁-R/PANC1 cells were grown in DMEM supplemented with 10% (v/v) fetal bovine serum, 1% (v/v) antibiotics, and 300 µg/ml G418. The hVIP₁-R/PANC1 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) fetal bovine serum, 1% (v/v) antibiotics, and 200 µg/ml G418. T47D breast cancer cells were grown in RPMI-1640 medium supplemented with 10% (v/v) fetal bovine serum, 1% (v/v) antibiotics, and 1.4 µM bovine insulin. H508 human colon cancer cells and Sup T₁ human lymphoblastoma cells were grown in RPMI-1640 medium supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) antibiotics. T98G human glioblastoma cells were grown in minimum essential medium supplemented with 10% (v/v) fetal bovine serum, 1 mM sodium pyruvate, and 1% (v/v) antibiotics. Cultures were maintained in incubators at 37°C in an atmosphere of 5% CO₂ and 95% air.

Southern Blot Analysis. Total RNA from human VPAC₁-R-transfected PANC1 cells and CHO cells, T47D cells, H508 cells, T98G cells, and Sup T₁ cells was used to synthesize first-strand cDNA. The first-strand cDNA was synthesized using 1.0 µg of total RNA with the First-Strand cDNA synthesis kit (Invitrogen). The probe for the hVIP₁-R (a 327-bp fragment) was generated from human lung total RNA. The gene-specific PCR primers used were as follows: 5'-GAAGGCTGGACGCACCTGGA-3' (sense) and 5'-TTACAGCCACCGAGCCTGGA-3' (antisense). The probe for hVIP₂-R (a 150-bp fragment) was generated from Sup T₁ human lymphoblastoma cells total RNA. The gene-specific PCR primers used were as follows: 5'-CCACCTGAACCTGTTCCTGT-3' (sense) and 5'-TACTGCAGGAAGAAGACCAGGCT-3' (antisense). The PCR reaction was carried out using the following conditions: an initial step of 3 min at 95°C; 30 cycles of 1 min of denaturation at 95°C, 1 min of annealing at 60°C, and 1 min of extension at 72°C with 5 min of final extension at 72°C. The DNA sequences of the PCR products were verified by sequencing both strands. For amplification from the first-strand cDNAs, the same gene-specific primers for hVIP₁-R and hVIP₂-R used to make probes were used as described above. The PCR products were electrophoretically separated in 1.2% (w/v) SeaKem GTG agarose gels (FMC BioProducts, Rockland, ME) and transferred to nitrocellulose. Nitrocellulose filters of Southern transfers were hybridized at room temperature with the ³²P-labeled hVIP₁-R probe (327-bp fragment) and hVIP₂-R probe (150-bp fragment). Labeling of the probes was performed using Random Primers DNA labeling system (Invitrogen). Hybridization and washing procedure was carried out as described previously (Ito et al., 2000).

Preparation of Dispersed Pancreatic Acini. Dispersed acini from guinea pig and rat were prepared as described previously (Ito et al., 2000). Unless specified otherwise dispersed acini from the pancreas of one animal were suspended in 100 ml of standard incubation solution. All incubations were at 37°C. The incubation solution was equilibrated with 100% O₂, and incubations for measurement of amylase release were performed with 100% O₂ as the gas phase.

Assessment of Amylase Release from Guinea Pig Pancreatic Acini. Amylase release was measured using the procedure published previously (Ito et al., 2000). Amylase activity was determined using the Phadebas reagent, and results were expressed as the percentage of the total cellular amylase released into the extracellular medium during the incubation. For all peptides the EC₅₀ was calculated, which was the concentration of the peptide that gave half-maximal stimulation of a maximally effective concentration of VIP (10 nM). The EC₅₀ was calculated using the curve-fitting program Kaleidagraph (Synergy Software, Reading, PA).

Binding Studies. Binding of ¹²⁵I-VIP to pancreatic acini was performed as described previously (Zhou et al., 1989; Fishbein et al., 1994; Ito et al., 2000). Acini were incubated with nonradiolabeled peptides for 45 min at 37°C with 75 pM ¹²⁵I-VIP without (total binding) or with 1 µM VIP (nonsaturable binding). Samples (100 µl) of cell suspension were centrifuged through silicon oil (density = 1.05) in microcentrifuge tubes to separate bound from unbound ligand. Radioactivity was determined by a Packard auto-gamma counter (Packard Instrument Co., Meriden, CT). Nonsaturable binding for ¹²⁵I-VIP was less than 3% of total binding.

Preparations and Degradation Study of ¹²⁵I- [Ala^{2,8,9,11,19,24,25,27,28}]VIP. ¹²⁵I-[Ala^{2,8,9,11,19,24,25,27,28}]VIP at a specific activity of 2200 Ci/mmol was prepared by a modification of the methods described previously (Zhou et al., 1989). ¹²⁵I-[Ala^{2,8,9,11,19,24,25,27,28}]VIP or ¹²⁵I-VIP corresponding to 500,000 cpm (250 pM) was incubated in 1 ml of standard incubation solution in the absence of bacitracin, with or without hVIP₁-R/PANC cells (0.3 × 10⁶ cells/ml) at 37°C for 7.5 min. After incubation, aliquots were centrifuged and distribution of the radioactivity in the supernatants was analyzed by HPLC. Each supernatant, including ¹²⁵I-VIP corresponding to 220,000 cpm or ¹²⁵I-[Ala^{2,8,9,11,19,24,25,27,28}]VIP corresponding to 180,000 cpm, was injected onto an HPLC with a Vydac C18 column. Fractions (1 ml) were collected and radioactivity determined.

Statistical Analysis. The results are mean ± S.E.M. of three or more experiments. IC₅₀ and EC₅₀ were calculated using the curve-fitting program Kaleidagraph. Binding curves for VIP₁-R/PANC1 cells and T47D cells were fitted using a least-squares curve fitting program, LIGAND (P. J. Munson; NIH, Bethesda, MD), to calculate dissociation constant (K_d) and binding capacities (B_max). Statistical comparisons were made using the Student's t test.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Expression of Subtypes of VIP Receptors in Human Tumor Cell Lines. To access the pharmacology of VIP at the VPAC₁ receptor we used two different approaches. Because VIP can interact with VPAC₂-R also and at high concentrations with the PACAP receptor (Harmar et al., 1998) and because many cell types contain more than one subtype of VIP/PACAP receptor (Waschek et al., 1995; Reubi et al., 2000), we first prepared cell lines with VPAC₁-R stably transfected. Second, to confirm that the pharmacology in the transfected cell lines resembled the native receptor, we used cells that contained only or >90% native VPAC₁-R. To obtain a native hVPAC₁-R in sufficient numbers to determine pharmacology we screened a number of human cell lines reported to contain hVIP receptors (Fig. 1). The expression of human VIP receptors is reported in T47D human breast cancer cells (Waschek et al., 1995), NCI-H508 human colon cancer cells (Frucht et al., 1992), T98G human glioblastoma cells (Vertongen et al., 1996), and Sup T₁ human lymphoblastoma cells (Robberecht et al., 1996). RT-PCR with Southern blotting with probes specific for hVPAC₁-R or hVPAC₂-R was performed to determine whether mRNA for these receptors was present in these cells (Fig. 1). T47D cells expressed hVPAC₁-R mRNA only (Fig. 1, lanes 3 and 13). H508 cells showed expression of both hVPAC₁-R and hVPAC₂-R mRNA (Fig. 1, lanes 2 and 12). T98G cells and Sup T₁ cells expressed hVPAC₂-R mRNA only (Fig. 1, lanes 1, 6, 7, 14, and 15).

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Southern blot analysis of the RT-PCR products from total RNA of several cell lines hybridized with a specific probe for the hVIP1-R or VPAC2 receptor (hVIP2-R). Left, RT-PCR products from total RNA of several cell lines by using hVIP1-R- or hVIP2-R-specific primers were hybridized with a specific hVIP1-R probe as described under Experimental Procedures. Position of the expected VIP1-R PCR product (327 bp) is indicated by the arrow. Right, RT-PCR products from total RNA of several cell lines by using hVIP1-R- or hVIP2-R-specific primers were hybridized with a specific hVIP2-R probe. Position of the expected VIP2-R PCR product (150 bp) is indicated by the arrow. hVIP1-R/PANC1, PANC1 (human pancreatic cancer cells) stably transfected with hVIP1-R; T47D, T47D human breast cancer cells; H508, H508 human colon cancer cells; Sup T1, Sup T1 human lymphoblastoma cells; T98G, T98G human glioblastoma cells.

Affinity of Single Alanine-Substituted Analogs of VIP for Human, Rat, or Guinea Pig VPAC₁ Receptor. To better understand the pharmacophore of VIP for the VPAC₁ receptor, we performed an alanine scan by the substitution of each amino acid one at a time with alanine. We analyzed both the human VPAC₁-R (by using hVPAC₁-R transfected CHO and PANC1 cells, and T47D cells containing native hVPAC₁-R) as well as rVPAC₁-R (by using rat pancreatic acini containing native rVPAC₁-R and rVPAC₁-R-transfected PANC1 cells), as well as guinea pig pancreatic acini possessing native VPAC₁-R. A recent study has demonstrated rat and guinea pig acini possess primarily VPAC₁-R (>90%) (Ito et al., 2000). hVPAC₁-R transfected into CHO cells was used because this cell line has been extensively used to characterize the cell biology and pharmacology of VIP receptors in previous studies (Gaudin et al., 1996). hVPAC₁-R-transfected PANC1 cells were used because the PANC1 cells we used did not possess natively VPAC₁-R, confirmed by RT-PCR with Southern blotting and binding studies (data not shown). However, these cells resemble a number of human adenocarcinoma cell lines recently reported to contain native hVPAC₁-R and in one study were found to contain low levels of native hVPAC₁-R (Jiang et al., 1997). A dose-inhibition curve by VIP of ¹²⁵I-VIP binding on VPAC₁-R-transfected PANC1 cells showed that VIP caused half-maximal inhibition at 1.7 ± 0.1 nM (Table 1). Analysis of the VIP dose-inhibition curve by a curve-fitting program demonstrated it was significantly better (p = 0.018) fit by a two-binding site model (n = 3), consistent with the Hill coefficient of 0.76. The high-affinity site had a K_d = 0.16 ± 0.04 nM and the low-affinity site had a K_d = 54.2 ± 28.2 nM. Receptor densities of high-affinity site were 2.4 ± 0.4 pmol/mg protein or 196 ± 37 fmol/10⁶ cells, whereas those of low-affinity sites were 36.5 ± 19.9 pmol/mg protein or 3045 ± 1664 fmol/10⁶ cells. In hVIP₁-R/CHO cells, analysis of the VIP dose-inhibition curve demonstrated it was also significantly better (p < 0.0001) fit by a two-binding site model (n = 4), consistent with the Hill coefficient of 0.66. The high-affinity site had a K_d = 0.34 ± 0.06 nM and the low-affinity site had a K_d = 68.9 ± 34.8 nM. Receptor densities of the high-affinity site were 406 ± 51 fmol/10⁶ cells, whereas for low-affinity sites they were 3685 ± 1289 fmol/10⁶ cells.

View larger version (51K): [in this window] [in a new window]   Fig. 2.   Relative affinities of various single alanine-substituted VIP analogs for native or transfected human (top), rat or guinea pig (bottom) VPAC1 receptors compared with native VIP. Top, hVPAC1-R-transfected CHO and PANC1 cells and T47D breast cancer cells, which natively possess hVPAC1-R, were incubated for 60 min at room temperature with 75 pM 125I-VIP alone or with various concentrations (0.01 nM-1 µM) of the indicated alanine-substituted VIP analogs or native VIP. By using the curve-fitting program Kaleidagraph, the IC50 for all peptides was calculated, which was the concentration that gave half-maximal inhibition of the saturable binding seen with 75 pM 125I-VIP alone. Results are shown with VIP analogs with a single alanine substitution from position 1 on the left to position 28 of VIP on the right. Native VIP has an alanine in position 4 and 18. Data are expressed as the ratio of the affinity of a given peptide divided by the affinity (IC50) of native VIP. The affinity of native VIP for human VPAC1-transfected cells or T47D breast cancer cells was as follows: hVIP1-R/CHO cells, 0.2 ± 0.01 nM; hVIP1-R/PANC1 cells, 1.7 ± 0.1 nM; and T47D cells, 1.6 ± 0.1 nM. Bottom, rat VPAC1 receptor-transfected PANC1 cells were incubated for 60 min at room temperature with 75 pM 125I-VIP alone or with various concentrations (0.01 nM-1 µM) of the indicated alanine-substituted VIP analogs or native VIP. Rat or guinea pig acini, which possess primarily VPAC1 (i.e., 90%) (Jensen, 1994; Ito et al., 2000), was incubated for 45 min at 37°C under similar conditions described above. The affinity (IC50) of native VIP for rat VPAC1-R-transfected PANC1 cells, or rat or guinea pig acini was as follows: rVIP1-R/PANC1 cells, 0.54 ± 0.03 nM; rat acini, 2.4 ± 0.2 nM; and guinea pig acini, 2.5 ± 0.1 nM. Less or more potency indicates the amount of decrease or the increase in affinity a given alanine substitution caused compared with native VIP, respectively. Data are the means from at least three experiments, and in each experiment each point was determined in duplicate.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Comparison of the IC50 of single alanine-substituted analogs for native human VPAC1 receptors (T47D cancer cells) to human VPAC1-transfected PANC1 cells (top) or CHO cells (bottom). IC50 values are shown in nanomolar concentration from Table 1 and Fig. 2. Each dot represents data from a single alanine-substituted analog. The data from the 25 substitutions with IC50 values more than 500 nM were excluded. The best fit and correlation coefficient (r) were calculated by least-squares analysis. The correlation coefficient between native human VPAC1-R and hVIP1-R/PANC1 cells (top) (r = 0.97) is significantly better (<0.01) than obtained between native VPAC1-R and hVIP1-R/CHO cells (r = 0.50) (bottom).

Affinity of Single D-Amino Acid-Substituted Analogs of VIP for Human, Rat, or Guinea Pig VPAC₁ Receptor. To investigate the importance of the orientation of the amino acid backbone substitution at each position of VIP for determining affinity at the human VPAC₁-R, similar studies to those performed with alanine-substituted analogs were performed with each amino acid of VIP replaced one at a time with its D-isomer (Fig. 4; Table 3). As seen in alanine scanning, the VIP pharmacophore demonstrated by comparing the relative affinity of each D-amino acid-substituted analog to VIP was generally close among different VPAC₁-R cell types (Fig. 4). The native hVPAC₁-R on T47D cells had almost identical affinities for the 28 different analogs to hVPAC₁-R-transfected PANC1 cells (Table 3). In contrast, with D-amino acid substitution in most positions, VPAC₁-R transfected into CHO cells had a higher affinity, varying from 1.3- to 18.6-fold with a mean of 6.5-fold for the identical D-amino acid-substituted analog compared with the results from the other two cell lines (Fig. 4; Table 3). Substitution of D-His¹, D-Gln¹⁶, and D-Leu²⁷ had the least effect, causing only a 2- to 3-fold decrease in affinity compared with VIP. The greatest decrease occurred with substitution of D-Phe⁶ (>80-fold), D-Thr⁷ (>130-fold), D-Asp⁸ (>80-fold), D-Thr¹¹ (>50-fold), D-Lys²¹ (>60-fold), and D-Tyr²² (>50-fold), demonstrating the orientation of these amino acids' side chains were particularly important for high-affinity receptor interaction.

View larger version (73K): [in this window] [in a new window]   Fig. 4.   Relative potencies of D-amino acid-substituted VIP analogs for native or transfected human (top), rat, or guinea pig (bottom) VPAC1 receptors compared with native VIP. The experimental conditions were the same as described in the legend to Fig. 2. The IC50 for each peptide was calculated as described in the legend to Fig. 2, and the data are expressed as the ratio of the affinity of a given peptide to that of native VIP as described in the legend to Fig. 2. Data are the means from at least three experiments and in each experiment each point was determined in duplicate.

Affinity and Biological Activity of Single Alanine-Substituted Analogs of VIP in Guinea Pig Acini. To compare the effect of alanine substitutions in VIP on biological activity with its effect on the affinity for the hVPAC₁-R, we used guinea pig pancreatic acini, because studies (Ito et al., 2000) demonstrate that 90% of the amylase release through the VIP receptor in this species is mediated by VPAC₁-R. At a concentration of 1 µM, all of the 26 single alanine-substituted analogs had full agonist activity, causing a 4-fold increase in enzyme secretion, which was equal in efficacy to a maximally effective concentration of VIP (10 nM) (data not shown). Alanine substitution for His¹, Asp³, Phe⁶, Thr⁷, Tyr¹⁰, Arg¹², Arg¹⁴, and Leu²³ produced a marked decrease of greater than 20-fold in affinity for either binding affinity or potency for stimulating amylase release (Fig. 5; Table 2), especially Asp³, Phe⁶, and Leu²³. In contrast, substitutions for Thr¹¹, Lys¹⁵, Asn²⁴, or Leu²⁷ increased the potency for stimulating enzyme secretion (Fig. 5; Table 2). There was a very close correlation (r = 0.997; p < 0.0001) between the ability of each alanine-substituted analog to stimulate enzyme secretion from pancreatic acini and inhibit ¹²⁵I-VIP binding to VPAC₁-R on pancreatic acini. These results demonstrate that the replacement of the various amino acid side chains of VIP by a methyl group did not result in a dissociation between binding affinity and potency for receptor activation, demonstrating that for all of the alanine-substituted analogs, binding affinity and receptor activation are closely coupled.

View larger version (45K): [in this window] [in a new window]   Fig. 5.   Comparison of the relative affinity of single alanine-substituted VIP analogs (top) or D-amino acid-substituted VIP analogs (bottom) from binding studies with their potency for stimulating amylase release in guinea pig acini compared with native VIP. The relative affinity from binding studies was calculated as described in the legend to Fig. 3. For measuring potency for stimulating amylase secretion, dispersed acini from one guinea pig pancreas were suspended in standard incubation solution containing 5 mM theophylline at 37°C for 30 min alone or with various concentrations (0.00001 nM-1 µM) of the VIP analogs or native VIP. For all peptides, the EC50 for stimulating amylase release was calculated, which was the concentration of the peptide that gave half-maximal stimulation of the secretion caused by a maximally effective concentration of VIP (10 nM). Data are expressed as the affinity ratio (binding, IC50) or potency ratio (amylase release, EC50) of a given peptide to that obtained with native VIP. Less or more potency indicates the amount of decrease or the increase in affinity compared with native VIP, respectively. The IC50 of native VIP in binding studies was 2.5 ± 0.1 nM and the EC50 for stimulating amylase release was 0.050 ± 0.005 nM. The control and maximal (10 nM VIP) stimulated values of amylase release were 3.7 ± 0.4 and 13.7 ± 0.7% of the total cellular amylase release into the intracellular medium during the incubation. Results are means from at least three experiments and in each experiment each point was determined in duplicate.

Affinity and Biological Activity of the D-Amino Acid-Substituted Analog of VIP in Guinea Pig Pancreatic Acini. Each of the 28 D-amino acid VIP analogs was an agonist for stimulating amylase release (Fig. 5; Table 4). If a sufficiently high concentration was used, all but [D-Phe⁶]VIP was a full agonist. [D-Phe⁶]VIP was a partial agonist causing 50 ± 3% of the maximal secretion caused by a maximally effective concentration of VIP. There was a close correlation (r = 0.85; p < 0.0001) between the binding affinity of each D-amino acid-substituted analog and its potency for stimulating amylase release (Fig. 5; Table 4). The stereochemistry of Asn⁹, Gln¹⁶, Ala¹⁸, Lys²⁰, and Leu²⁷ had either no or a minimal (<3-fold change) effect on biological potency. In contrast, the stereochemistry of Phe⁶, Thr⁷, Asp⁸, and Lys²¹ was as essential to high potency for secretion as it was for binding affinity (Fig. 5; Table 4).

Affinity, Biological Activity, and Stability of a Simplified VIP Analog with Multiple Alanine Substitution. In an attempt to identify a simplified human VPAC₁-R ligand, we used the alanine and D-amino acid scan data from our study to identify amino acids where the replacement of a given amino acid had either no or minimal effect on receptor affinity to design a simplified analog. The analog, [Ala^{2,8,9,11,19,24,25,27,28}]VIP, had an equal or better affinity than VIP for human (IC₅₀ = 1.6 ± 0.1 versus 1.6 ± 0.2 nM) and rat VPAC₁-R (IC₅₀ = 0.78 ± 0.01 versus 2.4 ± 0.2 nM), respectively (Fig. 6, top) and also had the same potency for stimulating amylase release from guinea pig acini (EC₅₀ = 0.048 ± 0.004 versus 0.033 ± 0.004 nM) (Fig. 6, top). Furthermore, ¹²⁵I-[Ala^{2,8,9,11,19,24,25,27,28}] showed much greater stability than ¹²⁵I-VIP. Specifically, >80% of ¹²⁵I-VIP was degraded during an incubation with hVIP₁-R/PANC1 cells and 0% of ¹²⁵I-[Ala^{2,8,9,11,19,24,25,27,28}] (Fig. 6, bottom) was degraded.

View larger version (31K): [in this window] [in a new window]   Fig. 6.   .: Affinity and stability of the simplified multialaninated VIP analog [Ala2,8,9,11,19,24,25,27,28]VIP. Top left, to determine the abilities of VIP and [Ala2,8,9,11,19,24,25,27,28]VIP to inhibit binding of 125I-VIP to T47D human breast cancer cells (1.2 × 106 cells/ml) and rat pancreatic acini, which possesses native VPAC1 receptors, each cell type was incubated for 60 min (human cells) or 45 min (rat acini) with 75 pM 125I-VIP alone or with the indicated concentrations of unlabeled peptides. Results are expressed as the percentage of the saturable binding of 125I-VIP observed in absence of competing peptide. In each experiment, each value was determined in duplicate and results given are means ± S.E.M. from at least three separate experiments. Top right, abilities of VIP and [Ala2,8,9,11,19,24,25,27,28]VIP to stimulate amylase release from guinea pig pancreatic acini. Acini from guinea pig were incubated in standard incubation solution with 5 mM theophylline for 30 min at 37°C. Results are the percentage of the stimulation caused by a maximally effective concentration of VIP (10 nM). The control and 10 nM stimulated values were 7.2 ± 0.3 and 27 ± 11% of the total cellular amylase in the acini before the incubation that was released into the extracellular medium during the incubation period. In each experiment, each value was determined in duplicate and results given are means ± S.E.M. from at least three separate experiments. Bottom, degradation of 125I-VIP or 125I-[Ala2,8,9,11,19,24,25,27,28]VIP by hVIP1-R/PANC1 cells. Shown are the HPLC elution profiles of supernatants after incubation of 125I-VIP (left) or 125I-[Ala2,8,9,11,19,24,25,27,28]VIP (right) (500,000 cpm/ml) with or without hVIP1-R/PANC cells (0.3 × 106/ml) in standard incubation solution for 7.5 min at 37°C. After incubation supernatants containing 200,000 cpm were analyzed by HPLC. Fractions (1 ml) were collected each minute, and radioactivity was determined in each fraction. Arrows indicate the point of elution of intact 125I-VIP (left) or 125I-[Ala2,8,9,11,19,24,25,27,28]VIP (right).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

A number of our results show there are important differences in the VIP pharmacophore for the human VPAC₁ compared with that in rat and guinea pig. Furthermore, there are significant differences in the VIP pharmacophore for the human VPAC₁ receptor expressed in CHO cells and PANC1 cells that need to be considered and that were not reported previously. First, with alanine or D-amino acid substitutions in VIP there were a number of important differences in their effect on affinity for the rat or guinea pig VPAC₁ compared with the human VPAC₁ that need to be recognized. Second, the VIP pharmacophore assessed by transfection into CHO cells, which has been extensively used in assessing the pharmacology of the VIP receptor, does not accurately reflect the true pharmacology of the native receptor. These results demonstrate that to accurately assess the VIP pharmacophore for the human or rat VPAC₁, it is essential that either cells containing native VPAC₁ receptors be used or VPAC₁ transfected into cells such as PANC1 be used, wherein the pharmacology closely reflects that seen in the native receptor.

Alanine scanning demonstrated the side chain substitutions that are most important for determining the overall pharmacophore of VIP for the human VPAC₁ are Phe⁶ Leu²³ > Asp³, Arg¹⁴ > His¹, Arg¹² (Fig. 7). With the rat and guinea pig VPAC₁ receptor, Thr⁷ is also important. The side chains of eight amino acids (Val⁵, Thr⁷, Tyr¹⁰, Lys¹⁵, Gln¹⁶, Lys²⁰, Tyr²², Ile²⁶) are of intermediate importance in determining the VIP pharmacophore for the native human VPAC₁ (Fig. 7). In contrast, with the guinea pig and rat VPAC₁ receptor, alanine substitution of Lys¹⁵, Gln¹⁶, Lys²⁰ caused little or no change. These results have both similarities and differences from previous studies. In one (O'Donnell et al., 1991) of the two studies (O'Donnell et al., 1991; Nicole et al., 2000) assessing the VIP pharmacophore at human VIP receptors in which binding to human lung membranes was assessed, it was concluded using alanine scanning that the most important amino acids in determining high-affinity interaction were Thr⁷, Tyr¹⁰, and Tyr²² in addition to Asp³, Phe⁶, Arg¹⁴, and Leu²³, which were also found to be important in our study. The difference from our results is likely explained by the fact the lung contains both VPAC₁ and VPAC₂ receptors (Usdin et al., 1994; Busto et al., 2000). This conclusion is supported by recent studies that demonstrate Thr⁷, Tyr¹⁰, and Tyr²² are much more important in determining the affinity of VIP for the VPAC₂ receptor than the VPAC₁ receptor (Nicole et al., 2000). Our results are in agreement with a recent study (Nicole et al., 2000) with human VPAC₁-R-transfected CHO cells, which concluded from alanine scanning that His¹, Phe⁶, Arg¹², Arg¹⁴, and Leu²³ were very important in determining the VIP pharmacophore for the human VPAC₁ receptor. However, our results on the relative importance of these different amino acids as well as those having an intermediate or no effect on affinity differ significantly from Nicole et al. (2000). That this difference was likely due to the different cell expression systems used was supported by the results in our study. We found that the VIP pharmacophore for either rat or human VPAC₁ receptors when expressed in CHO cells, but not PANC1 cells, differed significantly from that determined using cells possessing native VPAC₁ receptors. Our results using T47D cells and hVPAC₁-transfected PANC1 cells demonstrated Ser², Asp⁸, Asn⁹, Thr¹¹, Val¹⁹, Asn²⁴, Ser²⁵, Leu²⁷, and Asn²⁸ substitution with alanine caused a 2-fold decrease in affinity in either of these cells and these amino acid substitutions caused little or no loss of agonist potency. When these nine amino acids were alanine-substituted in VIP to form a simplified VIP analog, there was no loss of potency for VPAC₁ in any species (Fig. 6). [Ala^{2,8,9,11,19,24,25,27,28}]VIP also had a greater metabolic stability, demonstrating no degradation by VPAC₁-R alone, thereby suggesting it could be useful in vivo.

View larger version (144K): [in this window] [in a new window]   Fig. 7.   VIP pharmacophore for the human VPAC1 receptor. VIP amino acid residues 1 to 28 are shown as two -bends (residues 2-5, 7-10) and a single -helix (residues 11-26) as proposed from conformational, CD, and NMR studies (Fournier et al., 1988; Fry et al., 1989). The most important amino acids revealed by alanine scanning (>50-fold decrease in affinity with alanine substitution) on binding or biological activity are shown in yellow. Amino acid residues having intermediate importance (alanine substitution causes a 4- to 22-fold decrease) are shown in blue. The 12 amino acid residues whose backbone substitutions were not essential for high affinity or potency (substitution causes <4-fold decrease) at the VPAC1 are shown in green. **, indicates the nine amino acid residues that can be substituted by alanine with no loss of affinity either for the native or stably expressed human VPAC1 receptor and with no loss of potency of enzyme secretion from guinea pig acini.

The present study for the first time provides insight into the VIP pharmacophore for a human VIP receptor by using D-amino acid scanning. This approach can provide information on the importance of the orientation of a given amino acid side chain (Bolin et al., 1995; Coy, 2000). Such a strategy has been successfully used to investigate the pharmacophore and design potent agonists of growth hormone-releasing factor, which shares a close homology to VIP (Coy, 2000). This approach demonstrated that, except for substitution of D-His¹, D-Gln¹⁶, and D-Leu²⁷, alteration of the orientation of the side chain of any of the other 25 amino acids in VIP had a significant effect on receptor affinity in all species. The largest effects were seen with substitution of D-amino acids in the NH₂-terminal amino acids Phe⁶, Thr⁷, Asp⁸, and in the COOH-terminal amino acids Val¹⁹, Lys²¹, Tyr²², and Ile²⁶. Circular dichroism and nuclear magnetic resonance data obtained in organic solvents indicate VIP has a helical conformation, especially involving residues 15-28 and some studies suggest the N terminus has either one or two type-III -bends at residues 2-5 and 7-10 (Fournier et al., 1988; Musso et al., 1988; Fry et al., 1989; Wray et al., 1998) (Fig. 7). In one study in 50% methanol/water (Fry et al., 1989), the side chains of Asp³, Phe⁶, Thr⁷ were found to be clustered near the NH₂ terminus and a second cluster of residues involving Ala¹⁸, Val¹⁹, Tyr²², Ile²⁶ occur near the COOH terminus. It was proposed (Fry et al., 1989) that hydrogen bonding between Thr⁷ and Asp³ and Asp⁸ and Lys¹² may be important in stabilizing this helical structure. Based on these conformation studies, it was suggested VIP may have two separate regions that interact with the VIP receptor (Fry et al., 1989). Our data support this conclusion by both alanine scanning and D-amino acid substitutions. We could find a clustering of NH₂-terminal amino acids (residues 3, 6-8) and COOH-terminal amino acids (residues 19, 21, 23, 26) whose side chain conformation and backbone integrity are particularly important to the VIP pharmacophore.

A decrease in affinity with alanine or D-amino acid substitution could be due to either altered binding where the structural change was important in ligand-VPAC₁ receptor interaction or to the amino acid change altering the global structure of VIP, perhaps by altering its secondary structure. Our results suggest that differences in the affinities of VIP analogs with alanine substitutions for Leu¹³, Lys¹⁵, Gln¹⁶, Met¹⁷, Lys²⁰, Lys²¹, Tyr²², Asn²⁴, Ser²⁵, Ile²⁶, and Asn²⁸ are not due to global conformation changes in VIP, but are likely due to differences in the importance of these amino acids' backbone substitution for receptor-ligand interaction in the different species. This speculation is supported by the fact that each of these single alanine-VIP analogs had unaltered receptor affinity in at least one assay system. With alanine substitutions for His¹, Asp³, Val⁵, Phe⁶, Thr⁷, Tyr¹⁰, Arg¹², Arg¹⁴, and Leu²³, which had decreased receptor affinity in all systems examined, either mechanism discussed above could account for the decrease in affinity.

To determine the effect of amino acid substitutions in VIP on the VIP pharmacophore as assessed by its ability to activate the VIP receptor, we determined their ability to stimulate enzyme secretion from guinea pig pancreatic acini, which natively possess VPAC₁ receptors (Ito et al., 2000). These cells possess no secretin or PACAP receptors (PAC₁) coupled to enzyme secretion (Jensen, 1994) that VIP could also interact with to stimulate secretion, and >90% of the secretion caused by activation of VIP receptors on these cells is due to VPAC₁ receptors (Ito et al., 2000). In general, the VIP analog potencies for stimulating secretion correlated closely (r > 0.85; p < 0.0001) with their relative binding affinities. All alanine- and D-amino acid-substituted analogs were full agonists at the VPAC₁ receptor except for [D-Phe⁶]VIP, which was a partial agonist. In previous studies [D-Phe⁶]VIP or related analogs are reported to function as VPAC₁ receptor antagonists (Jensen, 1994; Usdin et al., 1994; Gaudin et al., 1996), as a partial agonist (Fishbein et al., 1994), or as a full agonist (Bolin et al., 1995). In our study, [D-Phe⁶]VIP functioned as a partial agonist of the VPAC₁ receptor with 50% of the efficacy of VIP. However, because of its low affinity (IC₅₀ > 0.1 µM) this analog is unlikely to be a good template for design of more potent antagonists. Our results are consistent with studies of the ability of these substituted VIP analogs to stimulate cyclic AMP accumulation in hVPAC₁-transfected CHO cells (Nicole et al., 2000) or to cause VIP receptor activation, resulting in the relaxation of guinea pig tracheal smooth muscle (O'Donnell et al., 1991; Bolin et al., 1995). These results demonstrate that a strategy of substitution of D-amino acids in VIP is unlikely to yield potent antagonists.

In conclusion, we studied the VIP pharmacophore for the VPAC₁ receptor in guinea pig, rat, and human by using both transfected cell systems and cells bearing native receptors. Our results demonstrated that there were significant species differences in the VIP pharmacophore for the VPAC₁ receptor in these different species. Our studies also demonstrate that the type of cell used for the stable expression of the VPAC₁-R could have a significant effect on the VIP pharmacophore, a factor that has not been considered in a number of previous studies on VIP receptor pharmacology. Last, our results allowed us to synthesize a simplified VIP analog containing 11 alanines that had equal affinity to VIP for the VPAC₁-R in all species but was more metabolically stable than VIP. This analog should be useful as a prototype for design of selective VPAC₁ receptor agonists/antagonists that could be useful in vivo, either therapeutically or to elucidate VIP's role in biological processes.