Optimization of the in vitro oxidative biotransformation of glimepiride as a model substrate for cytochrome p450 using factorial design

BACKGROUND AND PURPOSE OF THE STUDY: Glimepiride (GLM) was chosen as a model substrate in order to determine the kinetic parameters for in vitro metabolism via human liver micrososmes (HLM). We aimed to optimize the turnover of the substrate by the test system in relation to incubation time and HLM concentration in such a way that it was linearly dependent on time and less than 20% of the substrate was consumed which utilized the lowest amount of the HLM. Further we aimed to report K(m) and V(max) values for GLM. METHODS: Linearity of enzyme reactions in microsomal incubations was assessed by monitoring the effect of incubation time (from 5 to 60 min) and HLM concentration (from 0.2 to 0.75 mg/ml) on metabolite formation of GLM. The ideal conditions for turnover of GLM were justified using 3x3 factorial design. F value was calculated to confirm the omission of insignificant terms from the full-model to derive a reduced- model polynomial equation. The regression equation was used to develop a contour plot that showed turnover rate within the limits of this design. The optimized reaction velocity data was extrapolated to carry out the kinetic studies in vitro to generate a saturation curve for the determination of K(m) and V(max) values. RESULTS: The reaction was found to be linear with respect to both incubation time between 24 and 50 min and HLM concentration between 0.3 to 0.65 mg/ml. The K(m) and V(max) values obtained by nonlinear least squares regression method was found to be 28.9 ± 2.97 μMole and 0.559 ± 0.017 μMole respectively. Lineweaver-Burk plot was also used to estimate K(m) and V(max) which yield value of 29.411 ± 1.25 μMole and 0.571 ± 0.020 μMole/min/mg protein respectively. MAJOR CONCLUSION: The statistical approach successfully allows for the optimization of reaction time course experiments. The results obtained with linear as well as the nonlinear transformation were found to be in close agreement with each other which shows the best precision for estimates of K(m) and V(max).