Analysis of the role of membrane transport and polyglutamation of methotrexate in gut and the Ehrlich tumor in vivo as factors in drug sensitivity and selectivity.

The synthesis and retention of methotrexate polyglutamate derivatives were evaluated in the Ehrlich ascites tumor cell and the intestinal mucosa in vivo. When tumor-bearing mice re ceived injections of methotrexate (12 mg/kg), polyglutamate derivatives accumulated within the tumor cells as long as intracellular methotrexate was above the dihydrofolate reductase-binding capacity, the metabolites themselves reaching concentrations in excess of this level. Methotrexate attained maximum ¡ntracellular concentration in tumor cells within 1 hr but rapidly declined to below the enzyme-binding capacity in less than 5 hr. Methotrexate polyglutamates, however, re mained above the binding capacity even after 24 hr. One hr after administration, the majority of tumor cell dihydrofolate reductase was associated with methotrexate; however, as polyglutamate levels increased and free methotrexate declined, these derivatives replaced bound methotrexate on the enzyme so that, within 24 hr, over 80% of the enzyme was associated with methotrexate polyglutamates. Although intestinal mucosa cells attained intracellular levels of methotrexate comparable to that of the tumor, relatively little accumulation of polygluta mates was detected; in contrast to the tumor, total antifolate levels within these cells were minimal 8 hr after injection. Thus, the large difference in the rate of decline of total cell antifolate between these tissues was due to the accumulation and reten tion of methotrexate polyglutamates in the tumor. The rate of decline of the monoglutamate was identical in intestinal and tumor cells, which also paralleled the decline in the plasma methotrexate level. The data suggest that the toxic effects of methotrexate on intestinal cells may be related primarily to the action of the monoglutamate, while toxic effects on tumor cells are related to the buildup and persistence of the polyglutamate forms. A computer model used to simulate in vivo pharmacokinetics, transport, and metabolism of methotrexate predicted that the rates of decline of free intracellular methotrexate within cells having different membrane transport parameters for this agent would be identical with one another and would indeed parallel the decline in plasma methotrexate. Hence, shortly after ad ministration of drug, intracellular and extracellular free metho trexate is always at steady state, because the rates of transport across the cell membrane are far greater than the rates of decline of plasma methotrexate. The simulations further showed that, although the membrane transport parameters do not affect the rate of decline of intracellular monoglutamate or polyglutamates in vivo, they do govern the absolute levels of ' Supported in part by NIH Grants CA16906 and CA09340. 2 To whom requests for reprints should be addressed. Received July 28, 1982; accepted December 7, 1982. intracellular drug achieved which, in turn, influences the extent of synthesis of methotrexate polyglutamates and the duration over which critical levels are sustained within the cell. Hence, differences in the decline of total cell antifolate observed among different tumor lines with different transport properties for methotrexate must be due to differences in the net rate of decline of the polyglutamate derivatives.