ON THE NATURE OF MACROMOLECULAR CODING IN NEURONAL MEMORY OSCAR HECHTER and IAN D. K. HALKERSTON* This paper is an attempt to contribute to the class ofideas concerning a possible molecular basis for the memory engram, one of the major aspects ofthe mindproblem. Insuggestingplausible reactions at the molecular level dealing with "memory" and "recall," we are fully conscious of the many gaps in our formulation and aware that the approach employed has in effect bypassed all ofthe really "difficult" conceptual questions inherent in the problem [?, 2]. Our purpose is not to present an all-inclusive theory ofthe memory engram but to suggest possible future work which may contribute to the development ofa "molecular neurology." Thepresent ideas developed as a consequence ofattending a convocation ofthe Neurosciences Research Program entitled "Mind, Memory, Molecules ," where F. O. Schmitt and his associates dramatically presented the possibility that a revolutionary advance in understanding ofthe mind and its mechanisms—with all this implies in terms ofhuman intercommunication —might result ifthe powerful concepts and techniques of molecular biology could be appropriately coupled with the classical disciplines of brainstudy. Itwas emphasized that theproblem ofmemoryin mind might profitably be examined from the point of view of information transfer and macromolecular coding, as had been done in genetics, where mecha- * Worcester Foundation for Experimental Biology, Shrewsbury, Massachusetts. The views expressed here were developed in discussionwith various individuals at our own and other institutions. We particularly thank the following: Francis O. Schmitt, for inviting one ofus to participate in several sessions ofthe Neuroscience Research Program and therebyintroducing us to the scope and problems ofmolecular neurology; Irvin lsenberg, who helped to clarify our mechanistic thinking at the molecular level and suggested that the S-S bonds in proteins might serve as electron acceptors in charge-transferreactions; Hudson Hoagland and Gregory Pincus, for important assistanceand support at a variety of levels; and, perhaps most important, Emil Conason, who, starting many years ago, told us repeatedly, even though we did not listen at the time, that brain function is expressed in immunological terms. I83 nisms ofheredity and protein synthesis regarded just a few years ago as "virtually impossible to solve" had been elucidated by the interaction of molecular biology with classical genetics. Heredity involves a type ofbiological memory which passes information from one generation to another in a nucleic acid code; the information is transferred from one polynucleotide to another, eventually to be "translated" into a polypeptide code. The immunologically specific interactions of antibodies with antigens represent yet another type of biological "memory" carried in a polypeptide code. Accordingly, the question was raised whether memory in mind may not likewise involve a chemical code carried in macromolecules. If the mind-code could be "cracked" as the genetic code has been broken, the new possibilities which arise seem limitless. This challenging presentation, coming at a time when we were writing a review about hormone action from the point of view of information transfer, struck a responsive chord. The problem ofmemory in mind, like the problem ofhormone action, involves a series ofvectorial informationtransfer reactions whereby information received at the boundaries ofthe cell is translated into intracellular information which modifies cell function. To be sure, the information from the environment is processed differently in these two situations, but we wondered whether certain informationtransfer reactions might not be similar in these two examples ofbiological action. We had been considering the possibility that there might be a "code" inhormone action—a group ofhormones using an essentially similar series ofinformation-transferreactions—primarilybecause oftheubiquitous biochemical effects of the nucleotide 3',5'-cyclic adenosine monophosphate (3',5'-AMP). Discovered in the course ofstudies designed to elucidate the mechanism ofepinephrine action, 3',5'-AMP has now been shown to bean intermediary in the action ofpeptide hormones ofsuch diverse structure as glucagon, ACTH, and perhaps vasopressin, as well as the biogenic amine serotonin [3]. The mechanism of action of these biogenic: amines and peptide hormones at a cellular level involves hormonal regulation of3',5'-AMP generation; this is achieved, in part at least, by way of activation ofa membrane-sited enzyme (adenyl cyclase) involved in the formation of3',5'-AMP from ATP. Once formed, the nucleotide activates cell function in diffèrent ways depending...