Tapas K. Pradhan

Discovery of a High Affinity Radioligand for the Human Orphan Receptor, Bombesin Receptor Subtype 3, Which Demonstrates That It Has a Unique Pharmacology Compared with Other Mammalian Bombesin Receptors

An orphan receptor discovered in 1993 was called bombesin receptor subtype 3 (BRS-3) because of 47–51% amino acid identity with bombesin (Bn) receptors. Its pharmacology is unknown, because no naturally occurring tissues have sufficient receptors to allow studies. We made two cell lines stably expressing the human BRS-3 (hBRS-3). hBRS-3 was overexpressed in the human non-small cell lung cancer cells, NCI-H1299, and the other was made in Balb 3T3 cells, which lack endogenous BRS-3. [d-Phe6,β-Ala11,Phe13,Nle14]Bn-(6–14) (where Nle represents norleucine) was discovered to have high potency for stimulating inositol phosphate formation in both cell lines. [125I-d-Tyr6,β-Ala11,Phe13, Nle14]Bn-(6–14) bound to both cell lines with high affinity. Neither Bn nor 14 other naturally occurring Bn peptides bound to hBRS-3 with aK d <1000 nm. Twenty-six synthetic peptides that are high affinity agonists or antagonists at other bombesin receptors had an affinity >1000 nm. Guanosine 5′-(β,γ-imido)triphosphate inhibited binding to both cells due to a change in receptor affinity. These results demonstrate hBRS-3 has a unique pharmacology. It does not interact with high affinity with any known natural agonist or high affinity antagonist of the Bn receptor family, suggesting the natural ligand is either an undiscovered member of the Bn peptide family or an unrelated peptide. The availability of these cell lines and the hBRS-3 ligand should facilitate identification of the natural ligand for BRS-3, its pharmacology, and cell biology.

Identification of a unique ligand which has high affinity for all four bombesin receptor subtypes

Four subtypes of bombesin receptors are identified (gastrin-releasing peptide receptor, neuromedin B receptor, the orphan receptor bombesin receptor subtype 3 (BB3 or BRS-3) and bombesin receptor subtype 4 (BB4)), however, only the pharmacology of the gastrin-releasing peptide receptor has been well studied. This lack of data is due in part to the absence of a general ligand. Recently we have discovered a ligand, 125I-[D-Tyr6,betaAla11,Phe13,Nle14]bombesin-(6-1 4) that binds to BRS-3 receptors. In this study we investigate its ability to interact with all four bombesin receptor subtypes. In rat pancreatic acini containing only gastrin-releasing peptide receptor and in BB4 transfected BALB cells, this ligand and 125I-[Tyr4]bombesin, the conventional gastrin-releasing peptide receptor ligand, gave similar results for receptor number, affinity for bombesin and affinity for the unlabeled ligand. In neuromedin B receptor transfected BALB cells, this ligand and 125I-[D-Tyr0]neuromedin B, the generally used neuromedin B receptor ligand, gave similar results for receptor number, neuromedin B affinity or the unlabeled ligand affinity. Lastly, in BRS-3 transfected BALB cells, only this ligand had high affinity. For all four bombesin receptors this ligand had an affinity of 1-8 nM and was equal or greater in affinity than any other specific ligands for any receptor. The unlabeled ligand is specific for gastrin-releasing peptide receptors on rat pancreatic acini and did not inhibit binding of 125I-cholecystokinin octapeptide (125I-CCK-8), 125I-vasoactive intestinal peptide (125I-VIP) or 125I-endothelin to their receptors. The unlabeled ligand was an agonist only at the gastrin-releasing peptide receptor in rat acini and did not interact with CCK(A) receptors or muscarinic M3 acetylcholine receptors to increase [3H]inositol phosphates. These results demonstrate 125I-[D-Tyr6,betaAla11,Phe13,Nle14]bombesin-(6-1 4) is a unique ligand with high affinity for all subtypes of bombesin receptors. Because of the specificity for bombesin receptors, this ligand will be a valuable addition for such pharmacological studies as screening for bombesin receptor agonists or antagonists and, in particular, for investigating BRS-3 cell biology, a receptor for which no ligand currently exists.

Rational Design of a Peptide Agonist That Interacts Selectively with the Orphan Receptor, Bombesin Receptor Subtype 3

The orphan receptor, bombesin (Bn) receptor subtype 3 (BRS-3), shares high homology with bombesin receptors (neuromedin B receptor (NMB-R) and gastrin-releasing peptide receptor (GRP-R)). This receptor is widely distributed in the central nervous system and gastrointestinal tract; target disruption leads to obesity, diabetes, and hypertension, however, its role in physiological and pathological processes remain unknown due to lack of selective ligands or identification of its natural ligand. We have recently discovered (Mantey, S. A., Weber, H. C., Sainz, E., Akeson, M., Ryan, R. R. Pradhan, T. K., Searles, R. P., Spindel, E. R., Battey, J. F., Coy, D. H., and Jensen, R. T. (1997) J. Biol. Chem. 272, 26062–26071) that [d-Tyr6,β-Ala11,Phe13,Nle14]Bn-(6–14) has high affinity for BRS-3 and using this ligand showed BRS-3 has a unique pharmacology with high affinity for no known natural Bn peptides. However, use of this ligand is limited because it has high affinity for all known Bn receptors. In the present study we have attempted to identify BRS-3 selective ligands using a strategy of rational peptide design with the substitution of conformationally restricted amino acids into the prototype ligand [d-Tyr6,β-Ala11,Phe13,Nle14]Bn-(6–14) or its d-Phe6 analogue. Each of the 22 peptides synthesized had binding affinities determined for hBRS-3, hGRPR, and hNMBR, and hBRS-3 selective ligands were tested for their ability to activate phospholipase C and increase inositol phosphates ([3H]inositol phosphate). Using this approach we have identified a number of BRS-3 selective ligands. These ligands functioned as receptor agonists and their binding affinities were reflected in their potencies for altering [3H]inositol phosphate. Two peptides with an (R)- or (S)-amino-3-phenylpropionic acid substitution for β-Ala11 in the prototype ligand had the highest selectivity for the hBRS-3 over the mammalian Bn receptors and did not interact with receptors for other gastrointestinal hormones/neurotransmitters. Molecular modeling demonstrated these two selective BRS-3 ligands had a unique conformation of the position 11 β-amino acid. This selectivity was of sufficient magnitude that these should be useful in explaining the role of hBRS-3 activation in obesity, glucose homeostasis, hypertension, and other physiological or pathological processes.

In vitro and in vivo antitumor effects of cytotoxic camptothecin-bombesin conjugates are mediated by specific interaction with cellular bombesin receptors.

Most human tumors overexpress or ectopically express peptide hormone/neurotransmitter receptors, which are being increasingly studied as a means to selectively deliver cytotoxic agents. Although a number of peptide ligand-constructs demonstrate tumor cytotoxicity, the role of specific tumoral receptor interaction in its mediation is unclear. To address this question, we synthesized camptothecin (CPT) bombesin (Bn) analogs, in which CPT was coupled via a novel carbamate linker, L2 [N-(N-methyl-amino-ethyl)-glycine carbamate], that were chemically similar but differed markedly in their potency/affinity for human Bn receptors. We then examined their ability to interact with Bn receptors and cause in vitro and in vivo tumor cytotoxicity. CPT-L2-[d-Tyr6,β-Ala11,d-Phe13,Nle14] Bn (6-14) (BA3) bound with high affinity and had high potency for all three human Bn receptor subtypes, whereas CPT-L2-[d-Tyr6,β-Ala11, d-Phe13,Nle14] Bn (6-14) [d-Phe-CPT-L2-BA3] had >1400-fold lower affinity/potency. 125I-CPT-L2-BA3 but not 125I-d-Phe-CPT-L2-BA3 was internalized by Bn receptor subtype-containing cells. CPT-L2-BA3 displayed significantly more cytotoxicity than d-Phe-CPT-L2-BA3 toward NCI-H1299 lung cancer cells in both 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide and clonogenic assays and more potently inhibited H1299 xenograft growth in nude mice. CPT-L2-BA3 was also metabolically more stable than its parent peptide and inhibited growth of a number of other tumor cell lines in vitro and in vivo. These results demonstrate that specific tumoral receptor interaction is important in mediating the ability of peptide ligand-cytotoxic constructs to cause cytotoxicity. Because many tumors overexpress Bn receptors, these results also demonstrate that CPT-L2-BA3 will be a useful agent for delivering receptor-mediated cytotoxicity to many different human tumors.

Pharmacology and selectivity of various natural and synthetic bombesin related peptide agonists for human and rat bombesin receptors differs

The mammalian bombesin (Bn)-receptor family [gastrin-releasing peptide-receptor (GRPR-receptor), neuromedin B-receptor (NMB receptor)], their natural ligands, GRP/NMB, as well as the related orphan receptor, BRS-3, are widely distributed, and frequently overexpressed by tumors. There is increased interest in agonists for this receptor family to explore their roles in physiological/pathophysiological processes, and for receptor-imaging/cytotoxicity in tumors. However, there is minimal data on human pharmacology of Bn receptor agonists and most results are based on nonhuman receptor studies, particular rodent-receptors, which with other receptors frequently differ from human-receptors. To address this issue we compared hNMB-/GRP-receptor affinities and potencies/efficacies of cell activation (assessing phospholipase C activity) for 24 putative Bn-agonists (12 natural, 12 synthetic) in four different cells with these receptors, containing native receptors or receptors expressed at physiological densities, and compared the results to native rat GRP-receptor containing cells (AR42J-cells) or rat NMB receptor cells (C6-glioblastoma cells). There were close correlations (r=0.92-99, p<0.0001) between their affinities/potencies for the two hGRP- or hNMB-receptor cells. Twelve analogs had high affinities (≤ 1 nM) for hGRP receptor with 15 selective for it (greatest=GRP, NMC), eight had high affinity/potencies for hNMB receptors and four were selective for it. Only synthetic Bn analogs containing β-alanine(11) had high affinity for hBRS-3, but also had high affinities/potencies for all GRP-/hNMB-receptor cells. There was no correlation between affinities for human GRP receptors and rat GRP receptors (r=0.131, p=0.54), but hNMB receptor results correlated with rat NMB receptor (r=0.71, p<0.0001). These results elucidate the human and rat GRP-receptor pharmacophore for agonists differs markedly, whereas they do not for NMB receptors, therefore potential GRP-receptor agonists for human studies (such as Bn receptor-imaging/cytotoxicity) must be assessed on human Bn receptors. The current study provides affinities/potencies on a large number of potential agonists that might be useful for human studies.

Activities of calcitonin gene-related peptide (CGRP) and related peptides at the CGRP receptor.

In guinea pig pancreatic acini rat calcitonin gene-related peptide (CGRP) increased amylase release 2-fold, salmon calcitonin had an efficacy of only 44% of that of CGRP and [Tyr0]CGRP(28-37) and human calcitonin had no actions. [Tyr0]CGRP(28-37), but not human calcitonin, antagonized the actions of CGRP in pancreatic acini with an IC50 of 3 microM. [Tyr0]CGRP(28-37) produced a parallel rightward shift in the dose-response curve for CGRP-stimulated amylase secretion. The inhibition was specific for CGRP and was reversible. Studies with 125I-CGRP demonstrated that CGRP, salmon calcitonin and [Tyr0]CGRP, but not human calcitonin, interacted with CGRP receptors on pancreatic acini. These results indicate that various CGRP-related peptides demonstrate different relationships between their abilities to occupy the CGRP receptor and to affect biologic activity, with CGRP itself being a full agonist, salmon calcitonin a partial agonist, [Tyr0]CGRP(28-37) a competitive antagonist, and human calcitonin having no actions.

Camptothecin-somatostatin conjugates inhibit the growth of small cell lung cancer cells

The effects of camptothecin-somatostatin (CPT-SS) conjugates were investigated on small cell lung cancer (SCLC) cells. CPT was coupled to a SS agonist (SSA), c(Cys-Phe-DTrp-Lys-Thr-Cys)Thr-NH2 using the built in nucleophile assisted-releasing group (L1) N-methyl-aminoethyl-Gly-Dser-Nle-Dtyr-Dser or (L2) aminoethyl-Gly-Dser-Nle-Dtyr-Dser. The resulting CPT-L1-SSA and CPT-L2-SSA inhibited the specific binding of [125I-Tyr11]SS to NCI-H69 cell membranes with IC50 values of 0.2 and 2.1 nM, respectively. [125I]CPT-L1-SSA was internalized by SCLC cells at 37 degrees C but not at 4 degrees C. CPT-L1-SSA and CPT-L2-SSA inhibited in a dose-dependent manner the increase in adenylylcyclase activity caused by 25 microM forskolin. In vitro, 0.3 microM CPT-L1-SSA half-maximally inhibited the clonal growth of SCLC cells and 1 microM CPT-L1-SSA strongly inhibited 3H-thymidine incorporation into DNA and trypan-blue exclusion. These results suggest that CPT conjugated peptides such as CPT-L1-SSA may prove useful for exploring the efficacy of receptor-directed cytotoxicity to inhibit the proliferation of SCLC cells.

Identification of bombesin receptor subtype-specific ligands: effect of N-methyl scanning, truncation, substitution, and evaluation of putative reported selective ligands.

Mammalian bombesin (Bn) receptors include the gastrin-releasing peptide receptor, neuromedin B receptor, and bombesin receptor subtype 3 (BRS-3). These receptors are involved in a variety of physiological/pathologic processes, including thermoregulation, secretion, motility, chemotaxis, and mitogenic effects on both normal and malignant cells. Tumors frequently overexpress these receptors, and their presence is now used for imaging and receptor-mediated cytotoxicity. For these reasons, there is an increased need to develop synthetic, selective receptor subtype-specific ligands, especially agonists for these receptors. In this study, we used a number of strategies to identify useful receptor subtype-selective ligands, including synthesizing new analogs (N-methyl-substituted constrained analogs, truncations, and substitutions) in [d-Tyr6,βAla11,Phe13,Nle14]Bn(6–14), which has high affinity for all Bn receptors and is metabolically stable, as well as completely pharmacologically characterized analogs recently reported to be selective for these receptors in [Ca2+]i assays. Affinities and potencies of each analog were determined for each human Bn receptor subtype. N-Methyl substitutions in positions 14, 12, 11, 10, 9, and 8 did not result in selective analogs, with the exception of position 11, which markedly decreased affinity/potency. N-Terminal truncations or position 12 substitutions did not increase selectivity as previously reported by others. Of the four shortened analogs of [d-Phe6,βAla11,Phe13,Nle14]Bn(6–14) reported to be potent selective BRS-3 ligands on [Ca2+]i assays, only AcPhe,Trp,Ala,His(τBzl),Nip,Gly,Arg-NH2 had moderate selectivity for hBRS-3; however, it was less selective than previously reported Apa11 analogs, demonstrating these are still the most selective BRS-3 analogs available. However, both of these analogs should be useful templates to develop more selective BRS-3 ligands.

The gastrin-releasing peptide receptor is differentially coupled to adenylate cyclase and phospholipase C in different tissues

Recent studies suggest that in some tissues GRP receptor activation can both stimulate phospholipase C and the adenylate cyclase pathway and that activation of the latter pathway may be important in mediating some of its well-described growth effects. However, other studies suggest GRP-R may not be coupled to adenylate cyclase. To investigate this possibility, in the present study we determined the coupling of the GRP receptors to each pathway in mouse, rat, and guinea pig pancreatic acini and compared it to that in mouse Swiss 3T3 cells and human SCLC cells, all of which possess well-characterized GRP receptors. Moreover, we tested the effect of PKC activation on the ability of GRP-related peptides to increase cAMP accumulation in these tissues. Changes in cAMP levels were determined with or without IBMX present, with or without forskolin, or both to amplify small increases in cAMP. In mouse, rat and guinea pig pancreatic acini, murine Swiss 3T3 cells and human SCLC cells, GRP-related peptides caused a 600%, 500%, 250%, 300% and 60% increase, respectively, in [3H]IP with 1-3 nM causing a half-maximal effect. In murine Swiss 3T3 cells, IBMX, forskolin, and IBMX plus forskolin caused a 300%, 3500% and 10500% increase in cAMP, respectively. GRP-related peptides and VIP caused an additional 70% increase in cAMP with GRP causing a half-maximal (EC50) increase in cAMP at 2.1 +/- 0.5 nM, which was not significantly different from the EC50 of 3.1 +/- 0.9 nM for increasing [3H]IP in these cells. GRP-related peptides did not stimulate increases in cAMP in mouse, rat or guinea pig pancreatic acini or in SCLC cells either alone, with IBMX or forskolin or both. However, in pancreatic acini IBMX, forskolin or both increased cAMP 3 to 8-, 10 to 500-, and 100 to 1000-fold increase and the addition of VIP caused an additional 20-, 2-, and 3-fold increase in cAMP in the different species. In mouse pancreatic acini with TPA alone or IBMX plus TPA, neither bombesin nor GRP increased cAMP. Furthermore, in mouse pancreatic acini, neither TPA nor TPA plus IBMX altered basal or VIP-stimulated increases in cAMP. In mouse Swiss 3T3 cells TPA significantly increased cAMP stimulated by Bn, GRP or VIP. These results demonstrated that GRP receptor activation in normal tissues from three different species and a human tumoral cell line do not result in adenylate cyclase activation, whereas in Swiss 3T3 cells it causes such activation. The results suggest that the difference in coupling to adenylate cyclase is likely at least partially due to a difference in coupling to an adenylate cyclase subtype whose activation is regulated by PKC. Therefore, the possible growth effects mediated by this receptor in different embryonic or tumoral cells through activation of adenylate cyclase are not likely to be an important intracellular pathway for these effects in normal tissues.

Gastric chief cells possess NK1 receptors which mediate pepsinogen secretion and are regulated by agents that increase cAMP and phospholipase C.

In order to determine whether tachykinins alter the function of chief cells and to characterize the receptors mediating the effect, we investigated the abilities of various substance P (SP)-related peptides to inhibit the binding of 125I-Bolton-Hunter labeled substance P (125I-BH-SP) and their abilities to alter cell function in dispersed chief cells from guinea pig stomach. Binding of 125I-BH-SP was saturable, reversible, time- and temperature-dependent and was inhibited by several SP-related peptides with relative potencies of SP = physalaemin (IC50:0.19 nM) > SP methyl ester (SP-ME) (IC50:3.3 nM) > eledoisin (IC50:6.1 nM) > neurokinin A (NKA) (IC50: 65 nM) > neurokinin B (NKB) (IC50:80 nM). Analyses of these binding data demonstrated that chief cells possess a high and low affinity class of binding sites. Neither 125I-NKA nor [phenylalanyl-3,4,5-3H]senktide demonstrated saturable binding to chief cells. Acid stripping experiments demonstrated rapid ligand internalization with 55% of the bound radioligand internalized by 10 min. Phospholipase C activating agents (carbachol, CCK-8), adenylate cyclase activating agents (secretin, VIP), TPA and the calcium ionophore, A23187, all inhibited the binding of 125I-BH-SP and it was due to inhibition of ligand internalization with no change in surface bound parameters. SP (0.1 microM) stimulated pepsinogen secretion but was 4-times less efficacious than CCK-8 (10 nM) or carbachol (1 mM). 10 nM SP stimulated a rapid increase in cytoplasmic free calcium concentration ([Ca2+]i) followed by a sustained elevation lasting 2 min. Single cell spectroscopy demonstrated SP (10 pM to 1 microM) did not cause calcium oscillations. The NK1 receptor antagonist, CP96,345 specifically inhibited the SP-stimulated changes in [Ca2+]i and pepsinogen secretion. The relative potencies of SP-related peptides to stimulate pepsinogen secretion and [Ca2+]i demonstrated a close agreement with their abilities to inhibit the binding of 125I-BH-SP, and comparison of the dose-response curves suggests occupation of the low affinity sites mediate changes in biologic activity. In conclusion, the present study demonstrates that chief cells possess a NK1 subtype of tachykinin receptor, occupation of the low affinity sites of this receptor cause calcium mobilization and pepsinogen secretion, and that binding to this receptor is regulated by agents that activate phospholipase C, adenylate cyclase, protein kinase C and calcium mobilization.