Harnessing Bacillus inaquosorum AGSP2 for Enhancing ω-Transaminase Production Through Classical and AI-Supported Statistical Design

ω-transaminases are PLP-dependent enzymes able to catalyze the synthesis of various chiral amines, an important building block in the pharmaceutical industry. Here we describe the isolation and optimization of a wild-type strain of Bacillus species ( Bacillus inaquosorum AGSP2) isolated from an industrially polluted site of Amlakhadi, Gujarat, India. The isolate AGSP2 is a chiral amine producer and can synthesize a wide variety of compounds (Aldehydes, ketones, amino acids, and amines). The optimization was performed using the OFAT method (One Factor at a Time), followed by RSM-CCD (Response Surface Methodology-Central Composite Design), with further validation by an AI (Artificial Intelligence) tool, SVM (Support Vector Machine). The media optimized by statistical means were designated as Modified Luria Bertani (MLB) medium, which contains fructose, NaCl, yeast extract, and peptone supplemented with α-MBA. An overall 2.8-fold increase in transaminase production was observed with an enzyme activity of 6121.88 ± 42 U/ml. The other optimized parameters were temperature, pH, agitation speed, and inoculum size. AGSP2 is an (S) - selective ω-transaminase producer and has synthesised acetophenone using (S) -α-Methylbenzylamine with a 64.35% conversion and 51% of enantiomeric excess.