Optimization of fermentation conditions and enzymatic properties of extracellular lipase produced by Serratia marcescens derived from food waste.

A high lipase-producing strain, C41802, was isolated from a food-waste environment and identified as Serratia marcescens based on morphological traits, physiological and biochemical assays, and phylogenetic analysis. To enhance lipase production, fermentation conditions were optimized using a genetic algorithm–artificial neural network (GA–ANN) model. Under the optimized conditions, the enzymatic properties of the crude extracellular lipase were characterized. The GA–ANN model outperformed response surface methodology (RSM–CCD) in predicting optimal fermentation parameters and maximum enzyme yield, achieving a relative error of 0.67% compared with 2.54% for RSM–CCD. Optimization yielded the following conditions: olive oil 10 g/L, glucose 8 g/L, beef extract 3.1 g/L, K₂HPO₄ 2 g/L, inoculum size 3.5%, cultivation time 26.2 h, temperature 30 °C, and initial pH 5. Under these conditions, lipase production reached 613.39 U/mL, representing a 6.45-fold increase over the unoptimized baseline. The crude extracellular lipase exhibited an optimal temperature of 70 °C and an optimal pH of 7, retaining 52.56% of its activity after 12 h of incubation at 50 °C. These properties underscore the industrial potential of the lipase produced by S. marcescens C41802.