Modeling and optimization of uricase production from a novel Pseudomonas mosselii using response surface methodology and artificial neural network

Uricase is potential therapeutic and diagnostic enzyme having wide applications in pharmaceutical sector, but production yield and cost of production are challenging. Hence, isolation of potent uricase-producing sources and optimizing its production are essential for industrial-scale uricase production. In this study, a novel extracellular uricase-producing bacterium was isolated using uric acid plate assay. The strain was identified by 16S rRNA sequencing and it was found to be Pseudomonas mosselii with similarity of 99%. Further, multi-level statistical optimizations were conducted to enhance the uricase production. Firstly, Plackett-Burman design (PBD) was adapted for screening of different medium components and found that uric acid, peptone, and sucrose are highly significant medium components which affect the uricase production. Further, central composite design (CCD) and artificial neural network (ANN) combined genetic algorithm (GA) were used to optimize the uricase production. The use of CCD prior to ANN improved the predictive capability of ANN and GA model. Under the optimum conditions obtained using ANN-GA, the uricase production was enhanced and found to be 59.91 U/mL ± 0.32 which is 2.02 folds higher compared to unoptimized medium. The results of the current study enunciate that the newly isolated Pseudomonas mosselii is a potential bacterium for enhanced production of uricase and optimization using multilevel optimization approaches gave better prediction of optimum process conditions for enhanced uricase production.