Experimental evidence for a 5-exocyclic methylene-dUMP intermediate in the thymidylatesynthase reaction was recently obtained by demonstrating that tryptophan 82 mutants of the
Lactobacilluscasei enzyme produced 5-(2-hydroxyethyl)thiomethyl-dUMP (HETM-dUMP) (Barret, J. E., Maltby, D.A., Santi, D. V., and Schultz, P. G. (1998)
J. Am. Chem. Soc.
120, 449-450). The unusual product wasproposed to emanate from trapping of the intermediate with
-mercaptoethanol in competition with hydridetransfer from H
4folate to form dTMP. Using mutants of the C-terminal residue of thymidylate synthase,we found that the ratio of HETM-dUMP to dTMP varies as a function of CH
2H
4folate concentration.This observation seemed inconsistent with the conclusion that both products arose from a commonintermediate in which CH
2H
4folate was already bound to the enzyme. The enigma was resolved by akinetic model that allowed for differential partitioning of the intermediate formed on each of the twosubunits of the homodimeric enzyme in forming the two different products. With three C-terminal mutantsof
L. casei TS, HETM-dUMP formation was consistent with a model in which product formation occursupon occupancy of the first completely bound subunit, the rate of which is unaffected by occupancy ofthe second subunit. With one analogous
E. coli TS mutant, HETM-dUMP formation occurred uponoccupancy of the first subunit, but was inhibited when both subunits were occupied. With all mutants,dTMP formation occurs from occupied forms of both subunits at different rates; here, binding of cofactorto the first subunit decreased affinity for the second, but the reaction occurred faster in the enzyme formwith both subunits bound to dUMP and CH
2H
4folate. The model resolves the apparent enigma of thecofactor-dependent product distribution and supports the conclusion that the exocyclic methyleneintermediate is common to both HETM-dUMP and dTMP formation.