Abstract: To optimize the extraction process of Glehnia littoralis polysaccharides (GLP) and improve both extraction yield and bioactivity, ultrasound-assisted enzymatic extraction was employed. The effects of different enzyme preparations on extraction efficiency, polysaccharide composition, in vivo anti-inflammatory activity, and in vitro gastric mucosal protective activity in gastric mucosal cells were systematically evaluated. Subsequently, single-factor experiments combined with Box-Behnken response surface methodology were used to optimize the extraction parameters. The GLP was further purified using the Sevag method, and the contents of polysaccharides and proteins were determined. The results showed that pectinase extract exerted significant anti-inflammatory activity in the CuSO₄-induced zebrafish inflammation model, and obvious protective effects on human gastric mucosal GES-1 cells injured by MNNG and ethanol. In particular, its protective efficacy against ethanol-induced GES-1 cell damage was superior to other enzyme extracts. The optimal extraction condition for GLP was pectinase dosage of 0.44%, extraction time of 39 min, extraction temperature of 50 ℃, and solid-to-liquid ratio of 1∶30（g∶mL）. Under this condition, the polysaccharide extraction yield reached 37.17%. After purification, the polysaccharide content of GLP was 60.8%, with a residual protein content of 0.243%. In conclusion, ultrasound-assisted pectinase extraction enabled efficient preparation of GLP, and the resulting polysaccharides exhibited pronounced anti-inflammatory and gastric mucosal protective activities, thereby providing a reliable basis for the development of GLP in the functional food industry.

Key words: Glehnia littoralis polysaccharides, ultrasound-assisted enzymatic method, response surface methodology, extraction process, MNNG-induced injury model