The structure/function relationship of salmon calcitonin (sCT) analogues was investigated in heterologous calcitonin receptor (CTR) expression systems. sCT analogues with progressive amino-terminal truncations intermediate of sCT-(1-32) to sCT-(8-32) were examined for their ability to act as agonists, antagonists, or inverse agonists. Two CTR cell clones, B8-H10 and G12-E12, which express approximately 5 million and 25,000 C1b receptors/cell, respectively, were used for this study. The B8-H10 clone has an approximately 80-fold increase in basal levels of intracellular cAMP due to constitutive activation of the overexpressed receptor. In whole-cell competition binding studies, sCT-(1-32) was more potent than any of its amino-terminally truncated analogues in competition for 125I-sCT binding. In cAMP accumulation studies, sCT-(1-32) and modified analogues sCT-(2-32) and sCT-(3-32) had agonist activities. SDZ-216-710, with an amino-terminal truncation of four amino acids, behaved as a partial agonist/antagonist, whereas amino-terminal truncations of six or seven amino acid residues produced a 16-fold reduction in basal cAMP levels and attenuated the response to the agonist sCT-(1-32) in the constitutively active CTR system. This inverse agonist effect was insensitive to pertussis toxin inhibition. In contrast, the inverse agonist activity of these peptides was not observed in the nonconstitutively active CTR system, in which sCT analogues with amino-terminal truncations of four or more amino acids behaved as neutral competitive antagonists. These results suggest that the inverse agonist activity is mediated by stabilization of the inactive state of the receptor, which does not couple to G protein, and attenuates basal signaling initiated by ligand-independent activation of the effector adenylyl cyclase.

The translated calcitonin receptor (CTR) complementary DNA sequences contain potential N-linked glycosylation sites within the extracellular N-terminus. We investigated the relative molecular mass (M(r)) and degree of N-linked glycosylation of five cloned CTRs (pig, rat C1a, rat C1b, human I1-ve, and human I1+ve), together with the pig hypothalamic CTR, to analyze the potential contribution of carbohydrate moieties to the molecular identity of these receptors. Receptors were cross-linked to 125I-salmon CT with the homobifunctional reagent bis(sulfosuccinimidyl) suberate. Autoradiographic analysis of the cross-linked receptors, following SDS-PAGE, revealed apparent M(r)S, ranging between 70,000 and 80,000 for the rat, human, and pig hypothalamic receptors. However, the cloned, expressed pig CTR was much smaller (approximately 58,000). The lower M(r) of the cloned pig CTR appeared to be due to absence of N-terminal residues, but this did not impact on ligand-receptor specificity when compared with the hypothalamic pig CTR. Cleavage under nondenaturing conditions of N-linked sugars from the CTRs using endoglycosidase F (Endo F), increased the electrophoretic mobility of all receptors, except the pig CTRs, by approximately 10 kDa. Under denaturing conditions, electrophoretic mobilities increased by approximately 30 kDa for the rat C1a, rat C1b, and humanI1-ve (expressed in human embryonic kidney-293 cells) CTRs and by approximately 20 kDa for the cloned pig, pig hypothalamic, and human CTR isoforms (expressed in baby hamster kidney cells). Competition binding studies using glycosylated and partially deglycosylated (nondenaturing conditions) receptor preparations demonstrated no significant differences in binding affinity or specificity. Thus the CTRs are N-linked glycoproteins whose degree of glycosylation is both cell-type and species dependent.

Accumulating evidence supports the existence of nonthyroidal calcitonin (CT)-like peptides, more similar to fish CTs, which may act as endogenous regulators of CT receptors in brain and other tissues. In this study, we have carried out large-scale extractions from SpragueDawley rat brain diencephalon and pituitary, and purified a novel, biologically active, CT-like peptide from pituitary. Monitoring of the calcitonin-like activity of the peptides from rat brain and pituitary required different detection systems. While the brain CT cross-reacted with C-terminally directed salmon CT-specific antisera, the pituitary CT did not. However, the pituitary CT was biologically active, exhibiting specific interaction with CT receptors to activate adenylate cyclase. Conventional chromatographic techniques were employed to purify the CT-like peptides. Although the brain CT was not purified to homogeneity, size exclusion chromatography revealed the presence of multiple molecular weight forms of immunoreactive CT. Of these, only the lowest molecular weight form was biologically active. Purification from the pituitary resulted in the isolation of a biologically active peptide with a mass of 3267 Da. This mass differs from the mass of both salmon and thyroid-derived rat CT. Initial amino acid sequencing of the pituitary CT indicated that it was Nterminally blocked. Following aminopeptidase digestion, a unique six amino acid sequence, EKSQSP, was identified. Elucidation of the amino acid composition provided supporting evidence that the peptide was novel and was consistent with a full length peptide of approximately 30 amino acids. These data support the existence of novel, nonthyroidal, CTs which are potential regulators of CT receptormediated functions. (Endocrinology 139: [982][983][984][985][986][987][988][989][990][991][992] 1998)

Recently, a high affinity amylin binding site was identified in the mouse alpha-TSH thyrotroph cell line. In this study, we have characterized binding sites for 125I-salmon calcitonin (125I-sCT), 125I-rat alpha-calcitonin gene-related peptide (125I-CGRP), and 125I-rat amylin in alpha-TSH cells. Using 125I-CGRP or 125I-rat amylin, equilibrium was rapidly reached, and binding was fully reversible. Competition binding revealed the relative potency of peptides was sCT>amylin, CGRP>>rCT, which is similar to the specificity profile of amylin receptors characterized in rat brain. Furthermore, specific binding of 125I-rat amylin and 125I-CGRP to membrane preparations was reduced by 52% and 39%, respectively, in the presence of 20 microM GTP-gamma-s, indicating a requirement of G protein coupling for high affinity binding. In contrast, 125I-sCT binding reached equilibrium more slowly, was essentially irreversible, and was unaltered by GTP-gamma-s. Competition binding studies using 125I-sCT as radioligand demonstrated only weak interaction by CGRP or amylin, consistent with other described CT receptors. Assessment of ligand-induced cAMP accumulation and intracellular calcium signaling revealed a relative specificity profile of sCT>rCT with little or no second messenger signaling stimulated by amylin or CGRP, consistent with a C1-CT receptor phenotype. RT-PCR amplification of messenger RNA indicated that the predominant isoform was the C1a CT receptor. In cross-linking studies, 125I-rat amylin and 125I-CGRP specifically labeled a major band of relative molecular mass (Mr) approximately 80K, being approximately 10 kDa higher than the major 125I-sCT binding protein. Full deglycosylation of N-linked carbohydrates with endoglycosidase F reduced the Mr of each of the labeled proteins to approximately 50K. Cross-linked amylin or CT receptors were immunoprecipitated with C-terminally directed antimouse or antirat CT receptor antibodies but were not immunoprecipitated with nonimmune sera or antihuman CT receptor antibodies. The current data demonstrate expression of two biochemically distinct receptor phenotypes in mouse alpha-TSH cells, a CT receptor phenotype and an amylin receptor phenotype that have highly similar protein backbones.