Isothermal titration calorimetric studies of Saccharomyces cerevisiae myristoyl-CoA:Protein N-myristoyltransferase: Determinants of binding energy and catalytic discrimination among Acyl-CoA and peptide ligands

Saccharomyces cerevisiae myristoyl-CoA:protein N-myristoyltransferase (Nmt1p) is an essential, monomeric enzyme that catalyzes the transfer of myristate from CoA to the amino-terminal Gly residue of cellular proteins. Product inhibition studies indicate that Nmtlp has an ordered Bi Bi reaction mechanism with myristoyl-CoA binding to the apo-enzyme to form a high affinity binary complex followed by binding of peptide with subsequent release of CoA and then the myristoylpeptide product. We have used isothermal titration calorimetry to quantify the effects of varying acyl chain length and removing the 3′-phosphate group of CoA on the energetics of interaction between Nmtlp and acyl-CoA ligands. Myristoyl-CoA binds to apo-Nmt1p with an affinity of 15 nM, corresponding to a binding free energy of -10.9 kcal/mol. This free energy is composed of a large favorable enthalpy of -24 kcal/mol and a large unfavorable entropic term. This large negative ΔH° is consistent with a conformational change in the enzyme upon ligation, allowing synthesis of a functional peptide binding site. Binding of palmitoyl-CoA and lauroyl-CoA is driven by an exothermic enthalpy change which is much smaller than the corresponding parameter for myristoyl-CoA binding. The large differences in binding enthalpy and entropy (ΔΔH° and TΔΔS° = 8-9 kcal/mol) demonstrate that the "off-length" acyl-CoAs bind to Nmtlp in a significantly different energetic fashion from myristoyl-CoA, even though the enzyme does not have a great deal of specificity among these ligands in terms of binding free energy (ΔΔG° ≤ 1 kcal/mol). The effect of removing the CoA 3′-phosphate group from myristoyl-CoA is similar to the effect of a two-carbon change in acyl chain length: i.e. an enthalpy dominated reduction in binding affinity. However, kinetic studies reveal that removing the 3′-phosphate from myristoyl-CoA has little effect on Nmt1p's catalytic efficiency, indicating that the 3′-phosphate group contributes binding free energy but little catalytic destabilization. The greater ΔΔG°, with smaller ΔΔH° and ΔΔH° components, produced by removing the 3′-phosphate compared to increasing chain length suggests that it is not primarily the interactions of the 3′-phosphate which are disrupted when palmitoyl-CoA is substituted for myristoyl-CoA. No detectable interactions were noted between apo-Nmt1p and the substrate peptide, GAAPSKIV-NH₂, providing additional support for the preferred ordered reaction mechanism. In contrast, GAAPSKIV-NH₂ is able to bind with high affinity to Nmtlp saturated with either nonhydrolyzable myristoyl-CoA or palmitoyl-CoA analogs. The 3.6 μM dissociation constant for binding of GAAPSKIV-NH₂ to Nmt1p-S-(2-oxo)pentadecyl-CoA is driven by favorable enthalpy (-6.5 kcal/mol) and entropy (0.9 kcal/mol at 300 K). The fact that a high affinity peptide binding site is also formed with a nonhydrolyzable palmitoyl-CoA analog, S-(2-oxo)heptadecyl-CoA, indicates that the energy of binding of the off-length acyl-CoA is sufficient to induce the cooperative transition which allows peptide binding but not to generate a highly efficient active site. The similar affinities of Nmtlp for myristoyl-CoA, palmitoyl-CoA, and lauroyl-CoA provide little specificity for binding one acyl-CoA rather than another, suggesting that for this myristoyl-transferase to avoid catalyzing the transfer of other acyl chains or being competitively inhibited by them, there must be some form of functional segregation of cellular acyl-CoA pools.