Genetic Algorithm-Back Propagation Neural Network Model- and Response Surface Methodology-Based Optimization of Polysaccharide Extraction from Cinnamomum cassia Presl, Isolation, Purification and Bioactivities

Abstract

Ultrasonic-assisted enzymatic extraction (UAEE) was utilized to obtain the polysaccharides from the bark of Cinnamomum cassia Presl. (C. cassia). Taking the yield of the crude polysaccharides from C. cassia (CCCP) as the assessment indicator, response surface methodology (RSM) and a genetic algorithm-back propagation (GA-BP) artificial neural network model were employed to forecast and contrast the optimal parameters for UAEE. The outcomes demonstrated that the GA-BP model, which was superior in prediction accuracy and optimization capabilities to the RSM and BP models, identified the following conditions as optimal for the UAEE of CCCP: cellulase was employed, the temperature for enzymatic hydrolysis was 50.0 °C, the pH value was 5.248, the addition of enzyme was 3%, and the ultrasonic time was 70.153 min. Under these parameters, the yield of CCCP was significantly increased to 28.35%. Then, UAEE-extracted CCCP under optimal conditions was further separated and purified using a DEAE-52 column and SephadexG-100 column, yielding five purified polysaccharides from C. cassia (CCPs). All of these five fractions were acidic polysaccharides with safety at 3 mg/mL. The CCPs did not significantly affect the viability of HaCaT cells affected by UVB exposure. The CCPs demonstrated differential inhibition of nitric oxide production in RAW264.7 cells stimulated by lipopolysaccharide.