The substrate binding sites of adenylate kinase (AK) proposed by X-ray crystallographic studies [Pai, E. F., Sachsenheimer, W., Schirmer, R. H., & Schulz, G. E. (1977) J. Mol. Biol. 114, 37-45, and subsequent revisions] were evaluated by site-specific mutagenesis in conjunction with structural analysis by NMR. The residues examined in this report include two near an adenosine site (threonine-39 and arginine-44) and two in the phosphate binding region (arginine-128 and arginine-149). The results and conclusions are summarized as follows: (a) Although Thr-39 is very close to an adenine site [Egner, U., Tomasselli, A. G., & Schulz, G. E. (1987) J. Mol. Biol. 195, 649-658], it is nonessential either structurally or functionally. (b) The R44M mutant enzyme showed significant increases in the Michaelis and dissociation constants of adenosine 5'-monophosphate (AMP) (36- and 22-fold, respectively) while all other kinetic parameters were relatively unperturbed. The proton NMR property of this mutant was unchanged in the free enzyme and only slightly perturbed in the binary complexes with AMP and with MgATP (adenosine 5'-triphosphate), and in the ternary complex with MgAP5A [P1,P5-bis(5'-adenosyl) pentaphosphate]. These results indicate that Arg-44 interacts specifically with AMP starting at the binary complex, and suggest that the MgATP site proposed by Pai et al. (1977) is likely to be the AMP site. (c) The kinetic parameters of R149M were dramatically perturbed: kcat decreased by a factor of 1540, Km increased to 130-fold, and kcat/Km decreased by a factor of 2 X 10(5).(ABSTRACT TRUNCATED AT 250 WORDS)