Optimization of 1-deoxynojirimycin extraction from mulberry leaves by using response surface methodology.

Mulberry 1-deoxynojirimycin (DNJ, a potent alpha-glycosidase inhibitor) has therapeutic potency against diabetes mellitus. However, the amount of DNJ in mulberry leaves is low (about 0.1%), and a more effective extraction method is needed. Ultrasound-assisted extraction (UAE) was applied in this study for mulberry DNJ extraction, and five factors, the percentage of ethanol in the extraction solvent (x(1)), ratio of the extraction solvent to mulberry sample (x(2)), ultrasonic power (x(3)), extraction temperature (x(4)) and extraction time (x(5)), were investigated by fractional factorial 2((5-1)) design (FFD) to obtain the optimum extraction efficiency (DNJ yield, Y(1)) and extraction productivity (total yield, Y(2)). The results showed that x(2), x(3) and x(5) had significant impact on Y(1) and Y(2), and were further optimized by response surface methodology (RSM). Under the optimized conditions (x(2), x(3) and x(5) of 7 ml/g, 180 W and 260 s, respectively), DNJ-enriched powder (0.8%) was produced with high extraction efficiency (98%) and productivity (20%), enabling this product to be used for nutraceutical purposes.