A general folate cycle model has been completed for the MCF-7 breast cancer cell line. It consists of 21 ordinary differential equations describing the principal reactions of the folate cycle as well as those describing the glutamation, deglutamation and transport of methotrexate (MTX). Steady-state folate pools in drug-free cells have been described by adjusting Vmaxs while leaving Michaelis constants at their experimental values. By introducing observed mechanisms of inhibition and experimental inhibition constants into the model, we have found that it qualitatively reproduces folate pool dynamics during 21 hours of exposure to drug. Quantitative agreement was found if inhibition constants for MTXglu2 and MTXglu3 at thymidylate synthase (TS) and for dihydrofolate (FH2) and formyl-FH2 (FFH2) at glycinamide ribonucleotide and amidoimidazolecarboxamide ribonucleotide transformylases (GT and AT) are reduced within experimental uncertainty (i.e. 2-fold). Purine and thymidylate synthesis rates were found to also agree with experimental except that purine synthesis was somewhat overpredicted at exposure times between 10 to 21 hours. The folate cycle model has also been examined for its sensitivity to inclusion of enzyme activities that are cell cycle dependent. TS and dihydrofolate reductase (DHFR) were treated as periodic enzymes with S-phase activity 20-fold higher than in G- phase. The earlier model, where these activities had been represented as cycle-averaged quantities, was found to still represent folate dynamics provided (a) the rate constant describing FFH2 production from FH2 was reduced by 23% and (b) all quantities were properly averaged using G and S-phase population densities obtained from solution of a maturity-time representation of the cell cycle. S-phase cells were found to be more folate depleted at methylene tetrahydrofolate (MF) and formyl tetrahydrofolate (FFH4) than indicated by population average measurements. However, FH2 and FFH2 inhibitions at GT and AT still play a significant role since their presence is needed in the model in order to keep these pools elevated above zero, as required by experiment.
