Valorization of banana peel into α-amylase using one factor at a time (OFAT) assisted artificial neural network (ANN) and its partial purification, characterization, and kinetics study

Abstract

In this present study α-amylase producing bacteria were isolated from the soil near North Bengal University canteen on an amylase agar medium. Primary screening of the isolates was done by iodine test, based on the clear zone around the bacterial colonies in the media plates. Seven bacterial isolates were selected and quantitatively screened for amylase production. Among them, strain sps2 showing the highest amylase activity (215 ± 2 IU/ml) was selected for further studies. Isolated strain sps2 was identified both morphologically and biochemically and further confirmed by 16s rRNA sequencing as Streptomyces pratensis sps2 (Gene bank accession No. OP 236721). Optimization of α-amylase produced by the isolate was conducted under submerged fermentation of low-cost carbon source banana peel, using a one factor at a time (OFAT) approach. Maximum amylase activity of 821.33 ± 0.57 IU/ml was recorded after 2 days of fermentation. The artificial neural network (ANN) model was employed using the OFAT data to generate a predictive model for amylase production and the R-value of 0.94496 and the RMSE of 64.82 suggesting that the neural network has the precise capability to generate an optimization environment. Ammonium sulfate precipitation and dialysis techniques were used to partially purify the α-amylase and then characterized. The enzyme was found to have a molecular mass of 28 kDa (SDS-PAGE and zymogram) with a Km and Vmax of 2 mg/ml and 1000 μmol/min, respectively. The enzyme was found to be thermostable in the temperature range of 4–90 °C and more stable at alkaline pH with optimum pH of 7.