Application of a Theory of Enzyme Specificity to Protein Synthesis.

The suggestion of Lipmann' that the energy required to drive this reaction to the right came from adenosine triphosphate has been supported by the extensive work2 which has been discussed by the previous speakers on this symposium. While the detailed mechanism of this energy transfer has not been fully elucidated, the ATP apparently forms an amino acid anhydride. The fact that an anhydride of relatively high-energy content is the method of driving the reaction thermodynamically fits in with organic practice in which acyl anhydrides are the usual reagents for acylations. In addition to thermodynamic activation, there is a needed "kinetic activation." For example, the acyl phosphates and acyl adenylates acylate amines non-enzymatically at appreciable rates in aqueous solution,3 but this reaction is certainly not rapid enough or selective enough to account for protein synthesis. The reaction of these acyl anhydrides to form peptides must be catalyzed, and it is this catalysis which is the subject of this paper. An analysis of why this linking of amino acids presents such formidable difficulties is revealing. The difficulty appears to be caused by a combination of requirements, each of which, taken singly, is rather easily satisfied. The first requirement is that an individual position in the protein be occupied by one and only one amino acid. This by itself is not a difficult condition to satisfy, since enzymatic reactions of equally high specificity are well known. The second requirement is that a highmolecular-weight polymer be produced. Again, this, by itself, is not a unique condition, since the enzymatic formation of high-molecular-weight molecules from a given monomer is also familiar, e.g., carbohydrate polymerization catalyzed by phosphorylases. Finally, the macromolecule is formed from many different monomer units, but this requirement is also easily achieved if the units are randomly arranged, e.g., polynucleotide formation by polynucleotide phosphorylase.4 However, the combined requirements, i.e., the synthesis of a macromolecule from many individual monomers to give a single specified sequence, is a problem of different magnitude. A mechanism involving an individual enzyme for each bond would be acceptable from the specificity point of view, but it would require an inconceivably large number of enzymes to form all the proteins of the cell. Moreover, there are other observed conditions, e.g., the necessity of feeding all the amino acids