Artificial intelligence-based optimization for extracellular L-glutaminase free L-asparaginase production by Streptomyces violaceoruber under solid state fermentation conditions

Abstract

Abstract The bacterial L-asparaginase is a highly effective chemotherapeutic drug and a cornerstone of treatment protocols used for treatment the acute lymphoblastic leukemia in pediatric oncology. A potential actinomycete isolate, Streptomyces sp. strain NEAE-99, produces glutaminase-free L-asparaginase was isolated from a soil sample. This potential strain was identified as S. violaceoruber strain NEAE-99. The central composite design (CCD) approach was utilized for finding the optimal values for four variables including the mixture of soybean and wheat bran in a 1:1 ratio (w/w), the concentrations of dextrose, L-asparagine, and potassium nitrate under solid state fermentation conditions. Through the use of an artificial neural network (ANN), the production of L-asparaginase by S. violaceoruber has been investigated, validated, and predicted in comparison to CCD. It was found that the optimal predicted conditions for maximum L-asparaginase production (216.19 U/gds) were 8.46 g/250 mL Erlenmeyer flask of soybean and wheat bran mixture in a 1:1 ratio (w/w), 2.2 g/L of dextrose, 18.97 g/L of L-asparagine, and 1.34 g/L of KNO3. The experimental results (207.55 U/gds) closely approximated the theoretical values (216.19 U/gds), as evidenced by the validation. This suggests that the ANN exhibited a high degree of precision and predictive capability.

Keywords

• L-asparaginase; Streptomyces; Identification
• Optimization
• Central composite design
• Artificial neural network