The 2.3-A crystal structure of recombinant human dihydrofolate reductase (EC 1.5.1.3, DHFR) has been solved as a binary complex with folate (a poor substrate at neutral pH) and also as a binary complex with an inhibitor, 5-deazafolate. The inhibitor appears to be protonated at N8 on binding, whereas folate is not. Rotation of the peptide plane joining I7 and V8 from its position in the folate complex permits hydrogen bonding of 5-deazafolate's protonated N8 to the backbone carbonyl of I7, thus contributing to the enzyme's greater affinity for 5-deazafolate than for folate. In this respect it is likely that bound 5-deazafolate furnishes a model for 7,8-dihydrofolate binding and, in addition, resembles the transition state for folate reduction. A hypothetical transition-state model for folate reduction, generated by superposition of the DHFR binary complexes human.5-deazafolate and chicken liver.NADPH, reveals a 1-A overlap of the binding sites for folate's pteridine ring and the dihydronicotinamide ring of NADPH. It is proposed that this binding-site overlap accelerates the reduction of both folate and 7,8-dihydrofolate by simultaneously binding substrate and cofactor with a sub van der Waals separation that is optimal for hydride transfer.

No abstract

A procaryotic high-level expression vector for human dihydrofolate reductase has been constructed and the protein characterized as a first step toward structure-function studies of this enzyme. A vector bearing the tac promoter, four synthetic oligodeoxynucleotides, and a restriction fragment from the dihydrofolate reductase cDNA were ligated in a manner which optimized the transcriptional and translational frequency of the enzyme mRNA. The reductase, comprising ca. 17% of the total soluble protein in the host bacteria, was purified to apparent homogeneity as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and characterized by amino acid composition, partial amino acid sequence, and steady-state kinetic analysis. This expression vector has been used as a template for double-stranded plasmid DNA site-specific mutagenesis. Functional studies on a Cys-6----Ser-6 mutant enzyme support the contention that Cys-6 is obligatory for organomercurial activation of human dihydrofolate reductase. The Ser-6 mutant enzyme was not activated to any extent following a 24-h incubation with p-(hydroxymercuri)benzoate and nicotinamide adenine dinucleotide phosphate (reduced) (NADPH), whereas the kcat for Cys-6 reductase increased 2-fold under identical conditions. The specific activities of the Cys-6 and Ser-6 enzymes were virtually identical as determined by methotrexate titration as were the Km values for both dihydrofolate and NADPH. The Ser-6 mutant showed a decreased temperature stability and was more sensitive to inactivation by alpha-chymotrypsin when compared to the wild-type enzyme. These results suggest that the Ser-6 mutant reductase is conformationally altered relative to the Cys-6 native enzyme.

No abstract

No abstract