Determination of adenosine 3':5'-cyclic monophosphate in cerebral tissues by saturation analysis. Assessment of a method using a binding protein from ox muscle.

1. The Gilman (1970) procedure for determining cyclic AMP (adenosine 3':5'-cyclic monophosphate) by saturation analysis gave erroneous results when applied to the analysis of extracts of whole brain or preparations of membrane fragments from brain. 2. The extracts contained a non-diffusible factor, which enhanced the binding of cyclic AMP by the muscle protein fraction. 3. Extracts also contained material which inhibited binding, but net inhibition of binding was only observed when relatively concentrated extracts were analysed. 4. The error introduced by the factors modifying binding could be eliminated by incorporation of unlabelled internal standards in the unknowns. The design adopted enables a statistical estimate to be made of the standard error of a single assay. 5. The modified assay was used to determine bound cyclic AMP and adenylate cyclase activity in cerebral membrane fragments. Five preparations of synaptic membrane fragments contained less than 3.5pmol of cyclic AMP/mg of protein; a microsomal fraction from rat contained 65pmol of cyclic AMP/mg of protein.