The transformylase enzymes of de novo purine bi osy n t h esis

Formyl transfer reactions play a key role in the construction of the purine heterocycle during de now purine biosynthesis. Formylation is catalyzed early in the pathway by the purN glycinamide ribonucleotide transformylase (GAR Transformylase, EC 2.1.2.2) in a tetrahydrofolate-dependent manner and also by the purT GAR transformylase in a tetrahydrofolate-independent manner in bacteria. Late in the pathway, 5-aminoimidazole-4-carboxamide ribonucleotide transformylase (AICAR Transformylase, EC 2.1.2.3) catalyzes the second and final formylation involved in purine nucleotide biosynthesis. This article summarizes the salient properties and mechanistic knowledge on the transformylases with special emphasis on the mechanism of the purN GAR transformylase as explored by mutagenesis studies. Introduction The de novo purine biosynthetic pathway produces purine nucleotides that are essential for many processes in the cell. Purines serve as building blocks in DNA and RNA synthesis, are utilized as an energy source for chemical reactions (ATP), are used in cellular redox reactions (NADH, NADPH, FAD, etc.), and also play key roles in regulatory functions (CAMP, ZTP, etc.). Virtually all organisms studied to date utilize this pathway to synthesize purines, with the exception of parasitic protozoa (1) which must scavenge purines from their environment. The overall de novo purine biosynthetic pathway consists of ten enzymatic reactions which serve to convert 5-phosphoribosyl-1 -pyrophosphate to inosine monophosphate, which can then be converted to adenosine monophosphate and guanine monophosphate. These reactions are invariant in all organisms synthesizing purines, although the organization and regulation may differ (2). Generally, prokaryotes tend to use smaller single function enzymes, while higher eukaryotic organisms place increased reliance on larger multifunctional enzymes in this pathway (2). Because cancer cells grow rapidly and require large amounts of purines to maintain such growth, the de novo purine biosynthetic pathway has attracted considerable attention as a target for cancer chemotherapy (3). Some of the most successful antiproliferative drugs developed to date have been folate antimetabolites. Two of the enzymes in this pathway require a reduced folate, and are thus natural targets for screening novel antifolates. These enzymes, glycinamide ribonucleotide transformylase (GAR Transformylase, EC 2.1.2.2) and 5-aminoimidazole-4-carboxamide ribonucleotide transformy lase (AICAR Transformy lase, EC 2.1.2.3) catalyze the third and ninth reactions of this pathway. Both of these enzymes are involved in formyl transfer reactions, and both use 10-formyl tetrahydrofolate as a cofactor. These two enzymes are products of the purN and purH genes in Escherichia coli. Recently, a second glycinamide ribonucleotide transformylase enzyme was isolated and characterized from E. coli (4, 5). This enzyme is the product of the purT gene and does not utilize a folate cofactor. The purN and purT enzymes both catalyze the formylation of P-glycinamide ribonucleotide (GAR) to produce formyl p-glycinamide ribonucleotide (fGAR), however, they do so using different cofactors and different mechanisms. The reactions catalyzed by the three transformylases of de novo purine biosynthesis are shown in Fig. 1. The E. coli purN GAR transformylase is a monomeric protein of 212 amino acids with a molecular weight of 23,200. The homologous enzyme in humans is a much larger trifunctional polypeptide encoding GAR synthase (EC 6.3.4.13) and aminoimidazole ribonucleotide synthetase (EC 6.3.3.1) activities in addition to a GAR transformylase activity (6). Sequence homology and mutagenesis of catalytic residues suggests that there is a substantial degree of mechanistic similarity between the human and E. coli enzymes (6-9). High resolution x-ray crystal structures of the E. coli purN GAR Transformylase have been reported in the absence of ligands (lo), in a ternary complex with substrate GAR and a folate inhibitor (1 l), and in a binary complex with a multisubstrate adduct inhibitor bound (12). These results have shown that the enzyme structure is composed of two domains. The amino terminal domain (residues 1 to 101) contains a