RESUMO
This study investigated the impact of Candida tropicalis NITCSK13 on sugarcane bagasse (SCB) consolidated bioprocessing (CSB) using various parameters, such as pH, steam explosion (STEX) pretreatment, and temperature (at two different temperatures, cellulose hydrolysis and ethanol fermentation). The backpropagation neural network (BPNN) method simulated the optimal CSB conditions, achieving a maximum ethanol yield of 44 ± 0.32 g/L (0.443 g of ethanol/g of SCB) from STEX pretreated SCB within 48 h at 55 °C for cellulose hydrolysis and 33 °C for ethanol fermentation and pH 3.5. The simulated conditions were experimentally validated and showed an R2 value of 0.998 and absolute average deviation (AAD) of 1.23%. The strain NITCSK13 also exhibited a high ethanol tolerance of 16% (v/v). The interactions between the inhibitors, cellobiose, furfural, and thermocellulase were assessed through molecular docking. The results revealed a maximum inhibitory constant of 3.7 mM for furfural against the endoglucanase (EnG) of Humicola insolens (2ENG) at 50 °C. Acremonium chrysogenum endoglucanase (5M2D) exhibited a maximum of 88.7 µM for cellobiose at 50 °C. The SWISS homology model of EnG from Candida viswanathii exhibited inhibitory effects similar to those of EnG from Thermoascus and Thermotoga, indicating that the moderately thermophilic yeast Candida sp. cellulase may be capable of efficiently tolerating inhibitors and could be a promising candidate for consolidated bioprocessing of cellulosic ethanol.
RESUMO
There are numerous elements of daily life where plastic is employed, yet it is uncertain exactly when it will deteriorate. Poly-(3-hydroxybutyrate) (PHB), a biodegradable polymer, is viewed as a possible substitute for synthetic plastics made from petroleum. With Pseudomonas putida SS9, the current study sought to enhance operational conditions and nutritional factors to enhance PHB production. To maximize the impacts of operational factors, a combination of response surface modeling (RSM) and artificial neural networks (ANN) has been applied. PHB content was used as the response while the interaction effects of the factors were examined. The optimal parameters for PHB synthesis were further tested in a lab scale fermentor. Under optimal conditions, 13.83 g/L of C, 0.57 g/L of N, 0.59 g/L of P, the maximal productivity of PHB obtained with Pseudomonas putida SS9 is 12.89 g/L after 84 h. A mean square value of 15.7 with P < 0.0001 were obtained from the ANOVA results of quadratic polynomial model using RSM. The same construct was employed in MATLAB software to train a feed-forward ANN using the back-propagation approach, generating 12.88 g/L. The data indicated that a properly trained ANN model outperforms the RSM model in prediction. Furthermore, employing dairy waste (cheese whey) as a low-cost feedstock resulted in an equally proportionate PHB yield of 12.02 g/L. Therefore, cheese whey appeared to be a viable alternative carbon source over optimized synthetic media.