Multi-objective optimization of lignocellulolytic enzyme cocktail production from Pseudolagarobasidium acaciicola TDW-48 by artificial neural network-genetic algorithm (ANN-GA) strategy and its application in lignocellulose waste bioconversion
Thi Thu Huong Luong,Supattra Poeaim
DOI: https://doi.org/10.1007/s13399-024-05802-z
IF: 4.05
2024-06-13
Biomass Conversion and Biorefinery
Abstract:A massive amount of lignocellulose waste is generated annually, causing many environmental concerns. The bioconversion of these wastes into value-added products by the lignocellulolytic enzymes (LCE) is one of the effective and environmental approaches. However, the use of LCE has not been extended due to high costs. This study aimed to enhance the yield of crude LCE cocktail production from Pseudolagarobasidium acaciicola TDW-48 by optimizing cultural conditions using statistical tools. Firstly, the effect of cultural factors on LCE production was identified through the Plackett-Buman design. Then, the artificial neural network-genetic algorithm (ANN-GA) strategy was applied to optimize the significant factors. The result shows that the production of carboxymethyl cellulase (CMCase), xylanase, and laccase responded differently to cultural conditions. Among these, five factors (incubation time, water content, medium pH, glucose, and CuSO 4 concentration) were identified to have significant effects on enzyme activities. The ANN-GA optimization with a neuron network architecture (5-23-3) successfully modeled the crude LCE cocktail production, where the R -value achieved 0.98369 for the total dataset. A set of optimum conditions was proposed with an incubation time of 8 days, 72.6% water content, medium pH at 2.97, 0.5% glucose, and 0.53 g/L of CuSO 4 . With the above conditions, P. acaciicola TDW-48 could produce 23.97 U/g of CMCase, 26.02 U/g of xylanase, and 139.11 U/g of laccase, which enhanced 14.4%, 8.7%, and 405% activity, respectively, compared with non-optimization. In addition, the P. acaciicola TDW-48's crude LCE cocktail performed a high bioconversion efficiency on lignocellulose waste, the reducing sugar yield achieved 327.29 mg/g on rice straw, 308.02 mg/g on rice husk, and 312.29 mg/g on corn stover after 8-h incubation. These results provided a highly effective approach for LCE production with multi-objective optimization based on an artificial intelligence platform and supported the reuse of lignocellulose waste toward the eco-friendly strategy.
energy & fuels,engineering, chemical