A computational model of thiopurine metabolism

Abstract

A computational model of thiopurine metabolism The thiopurines, azathioprine, mercaptopurine (MP) and thioguanine are used as immunosuppressants and in the treatment of leukaemia. These drugs undergo extensive metabolism to form their cytotoxic metabolites that correlate with drug efficacy and the likelihood of side effects. Although these inexpensive drugs are effective in many patients, a deeper understanding of thiopurine metabolism would enable better individualisation of therapy resulting in increased efficacy and safety. A computational model of MP metabolism using data generated from mercaptopurine treated MOLT-4 (human T-ALL cell line) by a novel liquid chromatography mass spectrometry assay was built in CoPaSi. The model qualitatively reproduced published data about the effects of changes in thiopurine methyltransferase activity on MP metabolism. In vitro studies showed that high concentrations of allopurinol, a xanthine oxidase inhibitor, reduced the sensitivity of MOLT-4 cells to MP from 2.9 μM to 43 μM (P<0.001), whereas lower concentrations of allopurinol slightly but not significantly increased the sensitivity of MOLT-4 cells to MP. Combination of MP and allopurinol treatment of MOLT-4 cells resulted in lower concentrations of thioguanine nucleotides and methylated thioinosine monophosphate metabolites compared to MP only treatment, as determined by liquid chromatography mass spectrometry. These data were used to extend the model of MP metabolism to test hypotheses that addition of allopurinol decreased methylated thioinosine monophosphate and increased the concentration of TGNs. The computational model suggested that the mechanism by which allopurinol interacts with MP metabolism is by inhibiting hypoxanthine guanine phosphoribosyl transferase resulting in altered levels of MeTIMP and TGNs.