DNA replication in S phase requires a balanced supply of the four deoxynucleotides, and the thymidine nucleotides are the most vulnerable part of that supply because their synthesis depends on a short, linear pathway. Folate enters the pathway in its oxidized form and must be reduced to tetrahydrofolate (THF) by dihydrofolate reductase (DHFR). THF is then converted to 5,10-methylenetetrahydrofolate, which donates the methyl group that thymidylate synthase (TS) uses to convert deoxyuridine monophosphate (dUMP) into deoxythymidine monophosphate (dTMP). The reaction consumes the folate cofactor, leaving dihydrofolate that must be re-reduced by DHFR to sustain the cycle.
Methotrexate binds DHFR with high affinity and blocks this regeneration step. The reduced-folate pool collapses, TS loses its cofactor, and dTMP production falls. 5-Fluorouracil follows a different route to the same enzyme: after intracellular conversion it becomes fluorodeoxyuridine monophosphate (FdUMP), which binds the TS active site and, together with the folate cofactor, forms a stable ternary complex that inactivates the enzyme. Both drugs therefore deplete thymidine nucleotides, but one starves the enzyme of its cofactor while the other traps the enzyme itself.
Pemetrexed and other antifolates inhibit several folate-dependent enzymes at once, including TS and DHFR, which broadens the block. The pyrimidine and purine analogs act one step further downstream or on parallel pathways: cytarabine is phosphorylated to a triphosphate that is incorporated into DNA and terminates chain elongation, and 6-mercaptopurine is converted to a fraudulent purine nucleotide that disrupts purine synthesis and incorporation. The common feature is that each drug is a substrate analog — a molecule close enough to the real substrate to be handled by the same enzymes and transporters, but different enough to jam the reaction.