Using site-specific mutagenesis, we have probed the structural and functional roles of lysine-21 and lysine-27 of adenylate kinase (AK) from chicken muscle expressed in Escherichia coli. The two residues were chosen since according to the nuclear magnetic resonance (NMR) model [Mildvan, A. S., & Fry, D. C. (1987) Adv. Enzymol. 58, 241–313], they are located near the α- and the γ-phosphates, respectively, of adenosine 5′-triphosphate (ATP) in the AK-MgATP complex. In addition, a lysine residue (Lys-21 in the case of AK) along with a glycine-rich loop is considered “essential” in the catalysis of kinases and other nucleotide binding proteins. The Lys-27 to methionine (K27M) mutant showed only slight increases in kcat and Km, but a substantial increase (1.8 kcal/mol) in the free energy of unfolding, relative to the WT AK. For proper interpretation of the steady-state kinetic data, viscosity-dependent kinetics was used to show that the chemical step is partially rate-limiting in the catalysis of AK. Computer modeling suggested that the folded form of K27M could gain stability (relative to the wild type) via hydrophobic interactions of Met-27 with Val-179 and Phe-183 and/or formation of a charge-transfer complex between Met-27 and Phe-183. The latter was supported by an upfield shift of the methyl protons of Met-27 in 1H NMR. Other than this, the 1H NMR spectrum of K27M is very similar to that of WT, suggesting little perturbation in the global or even local conformations. These results suggest that Lys-27 is nonessential in either catalysis (including substrate binding) or structure, and that a gain in conformational stability does not require a global conformational change. Substitution of Lys-21 by methionine, on the other hand, resulted in a dramatic decrease in kcat corresponding to as much as 7 kcal/mol of transition-state binding energy, while causing little change in the conformational stability. However, there are substantial, and possibly global, conformational changes in K21M, as evidenced by dramatic differences between the 1H NMR spectra of K21M and WT. Examination of the 2.1-Å crystal structure of porcine muscle AK [Dreusicke, D., Karplus, P. A., & Schulz, G. E. (1988) J. Mol. Biol. 199, 359–371] suggested that the amino group of Lys-21 is within H-bonding distance (2.84 Å) of the carbonyl oxygen of Gly-15. Such a hydrogen bonding could stabilize the glycine-rich loop. Since the loop is a key feature of the tertiary structure, its disruption or destabilization in K21M could result in global conformational changes (but does not necessarily change the overall conformational stability). These results suggest that Lys-21 plays a key structural role, and that whether it is catalytically essential remains to be established. In addition, there seems to be no direct correlation between changes in conformation and changes in conformational stability in site-specific mutants. The results also strongly echo the warning [Wilde, J. A., Bolton, P. H., Dell’Acqua, M., Hibler, D. W., Pourmotabbed, T., & Gerlt, J. A. (1988) Biochemistry 27, 4127–4132] against straightforward interpretation of kinetic data without considering potential structural changes in site-specific mutants. Furthermore, we have addressed a few other issues related to the catalytic capability of AK, and the possible involvement of His-36 in substrate-induced conformational changes of AK. © 1990, American Chemical Society. All rights reserved.