Abstract
The applicability of the statistical tools coupled with artificial intelligence techniques was tested to optimize the critical medium components for the production of extracellular cholesterol oxidase (COD; an enzyme of commercial interest) from Streptomyces rimosus MTCC 10792. The initial medium component screening was performed using Placket-Burman design with yeast extract, dextrose, starch and ammonium carbonate as significant factors. Response surface methodology (RSM) was attempted to develop a statistical model with a significant coefficient of determination (R-2 = 0.89847), followed by model optimization using Genetic Algorithm (GA). RSM-GA based optimization approach predicted that the combination of yeast extract, dextrose, starch and ammonium carbonate at concentrations 0.99, 0.8, 0.1, and 0.05 g/100 ml respectively, has resulted in 3.6 folds increase in COD production (5.41 U/ml) in comparison with the un-optimized medium (1.5 U/ml). COD was purified 10.34 folds having specific activity of 12.37 U/mg with molecular mass of 54 kDa. The enzyme was stable at pH 7.0 and 40 degrees C temperature. The apparent Michaelis constant (K-m) and V-max values of COD were 0.043 mM and 2.21 mu mol/min/mg, respectively. This is the first communication reporting RSM-GA based medium optimization, purification and characterization of COD by S. rimosus isolated from the forest soil of eastern India.