IDENTIFICATION OF SITES FOR FEEDBACK-REGULATION OF GLUTAMINE 5-PHOSPHORIBOSYLPYROPHOSPHATE AMIDOTRANSFERASE BY NUCLEOTIDES AND RELATIONSHIPTO RESIDUES IMPORTANT FOR CATALYSIS

Glutamine phosphoribosylpyrophosphate amido-transferase, the key regul atory enzyme for de novo purine nucleotide synthesis, is subject to fe edback regulation by adenine and guanine nucleotides. Affinity labelin g with 5'-p-fluorosulfonylbenzoyladenosine (FSBA) and 8-azidoadenosine 5'-monophosphate (N3-AMP) was used to identify purine nucleotide site s for feedback control of the Escherichia coli amidotransferase. FSBA inactivated the amidotransferase with saturation kinetics. Specificity for inactivation was shown by the covalent attachment of 2.0-2.4 eq o f [H-3] sulfobenzoyladenosine (SBA) per subunit and protection by GMP and AMP against inactivation and incorporation of [H-3]SBA. Six chymot ryptic peptides modified with [H-3]SBA were isolated and identified by differential labeling followed by high performance liquid chromatogra phy and radioactivity. Mass spectrometry and Edman degradation analysi s were used to identify 5 residues that were covalently modified by [H -3]SBA: Tyr74, Tyr258, Lys326, Tyr329, and Tyr465. Tyr258 was also mod ified by N3-AMP. Mutant enzymes K326Q and Y329A had activity similar t o that of the wild type enzyme. However, both mutants exhibited decrea sed sensitivity to inhibition by GMP and decreased binding of GMP but were inhibited by AMP. Mutant enzymes Y74A and Y258F were normally fee dback-inhibited but were defective in glutamine amide transfer and syn thase functions, respectively. Therefore Tyr74 and Tyr258 are importan t for activity and modification by FSBA and N3-AMP accounts for enzyme inactivation. These results localize residues important for catalysis in close proximity to a site for nucleotide binding. Two additional m utant enzymes, G331I and N351A, were constructed which were refractory to inhibition by GMP with little change in inhibition by AMP. A repla cement of Tyr465 indicates that this residue is not essential for cata lysis or feedback inhibition. Overall, these results are interpreted i n terms of a two-nucleotide site model with Lys326, Tyr329, Gly331, an d Asn351 defining a site required for inhibition by GMP. A second nucl eotide site not affinity labeled by analogs is very close to or overla ps with the catalytic site.