Acidified Hot Water Extraction of Adenosine, Cordycepin, and Polysaccharides from the Chinese Caterpillar Mushroom, Ophiocordyceps sinensis CS1197 (Ascomycetes): Application of an Artificial Neural Network and Evaluation of Antioxidant and Antibacterial Activities.

The effects of process variables (temperature, time, and pH) on the extraction of adenosine, cordycepin, and polysaccharides from Ophiocordyceps sinensis CS1197 were studied using response surface methodology (RSM) and an artificial neural network (ANN). The ANN model resulted in low root mean square errors (0.022, 0.079, 0.018) and high R2 values (0.995, 0.934, 0.997) for adenosine, cordycepin, and polysaccharide yields, respectively, which implied good agreement between the predicted and actual data. An overall desirability of 0.86 suggested optimal extraction conditions (temperature, 70°C; time, 1 hour; and pH ~4) for adenosine (0.205%) and cordycepin (0.246%) yields. For polysaccharide yield (6.34%), an overall desirability of 0.93 suggested extraction conditions: temperature, 87°C; time, 3.4 hours, and pH ~4. The water extract exhibited better antioxidant activity. The antibacterial activity of the water extract was assayed against Bacillus subtilis, Staphylococcus aureus, and Escherichia coli, and larger amounts of the extract (30 and 50 mg) exhibited antibacterial activity (<35%). The predictive ability of an ANN is superior to RSM and resulted in the best agreement between experimental and predicted values.