Differential effects of dithiothreitol and iodoacetamide on corticotropin-releasing factor (CRF) activity of bovine hypothalamic CRFs and vasopressin.

Various fractions were tested in vivo for corticotropin-releasing factor (CRF) activity after Sephadex G-100 fractionation of 0.1-N HCl extracts of bovine hypophyseal stalk or cerebral cortex. Female rats pretreated with chlorpromazine, morphine, and Nembutal were used for CRF assay. CRF-A (void volume fractions; big CRF), CRF-B (Kav = 0.583), and CRF-C (salt volume fractions) of bovine hypophyseal stalk and lysine or arginine vasopressin all induced clear-cut stimulation of ACTH and corticosterone in the assay rat, whereas they were ineffective in acutely hypophysectomized rats. Control fractions purified from bovine cerebral cortex had no CRF activity. Treatment of arginine and lysine vasopressin and CRF-C with dithiothreitol and iodoacetamide completely abolished their CRF activity, whereas the CRF activities of CRF-A and CRF-B were unaltered by these treatments. Treatment with iodoacetamide alone had no effect on the CRF activity of any of these substances. Fractionation of either CRF-C or arginine vasopressin on Sephadex G-15 yielded a CRF-active peak at a Kav of 0.35. We conclude that 1) three different forms of CRF exist in bovine hypophyseal stalk; 2) CRF-A and CRF-B are unrelated to vasopressin and require neither a disulfide bond(s) nor a sulfhydryl group(s) for their CRF activity; 3) reduction of the disulfide bond of vasopressin destroys both CRF and antidiuretic activities; 4) CRF-C requires an intact disulfide bond(s) for its CRF activity and is likely to be either vasopressin itself or a substance closely related to vasopressin; and 5) CRF-B is likely to be the physiologically important form of bovine CRF.