Cyclic adenosine monophosphate, CA++, and membranes.

Since the initial discovery of cyclic 3',5'-adenosine monophosphate (3',5'AM{P) by Rall, Sutherland, and Berthet' and its role in glycogenolysis in the liver,2 evidence has accumulated indicating that this nucleotide is an intermediate in the action of many peptide hormones.3-9 More recently it has also been implicated in the release of insulin from the pancreas,10 the release of amylase from the salivary gland,"1 and the action of estrogen on the uterus.'2 As this evidence has accumulated, no satisfactory theory has developed to explain the cellular or molecular basis of these apparently diverse effects of 3',5'AMP. Initially, it was thought that in the case of both epinephrine-induced glycolysis2 and adrenocorticotropin (ACTH)-induced steroid release from the adrenal cortex3 the effect of cyclic AMP was confined to an activation of phosphorylase. However, as this nucleotide was identified as an intermediate in the epinephrine-induced lipolysis in adipose tissue5 and the vasopressin-induced change in water permeability in the toad bladder6 it became clear that an effect upon the single enzyme, phosphorylase, was insufficient to account for all of the nucleotide's effects. Even in the case of ACTH action, it soon became apparent that cyclic AMP must have effects upon adrenal cell metabolism other than an activation of phosphorylase.4 Nonetheless, the notion persisted that its major effects were upon enzymes, particularly when it was discovered that it activated phosphofructokinase in certain tissues.4 Thus considerable credence was given to the notion that cyclic AiIP acted as a modifier of the activities of one or more enzymes in a particular cell.13 The specificity of its particular effects upon a given cell was explained as being due to the uniqueness of each particular cell type in regard to its enzymatic composition. If this were the case, it would mean that the search for the role of cyclic AMP in each particular cell type would be a separate problem in biochemical investigation. The alternative would be that cyclic AMP has a different function, but that investigative attention as to its site of action had been misdirected. An alternative mechanism was first proposed when we noted an apparent relationship between 3',5'-AMP and Ca++ in the action of vasopressin on the toad bladder and several other systems. 14' 15 The possible importance of this relationship was further emphasized by the discovery that adenyl cyclase, the enzyme responsible for the conversion of adenosine-5'-triphosphate (ATP) to 3',5'AMP, and phosphodiesterase, the enzyme involved in converting 3',5'-AMP to AMP, were present in high concentrations in the brain,'6' 17 and appeared to be localized primarily in the synaptosome portion of brain homogenates.'8 This discovery was followed by the reports that cyclic AMP also appeared to be an intermediate in the secretion of amylase by the parotid gland" and in the secretion of insulin by the A cells of the pancreatic islets.'0 These apparently diverse systems have several striking similarities. A consideration of these common