OBJECTIVE：To establish and optimize a extraction method of Fructus Gleditsiae Abnormallis ，and to analyze and identify chemical components of the extract simultaneously. METHODS ：Fructus Gleditsiae Abnormallis was extracted with CO 2 supercritical fluid extraction （SFE）method. Based on single factor tests ，using extraction yield as index ，extraction temperature ， extraction pressure and extraction time as investigation factors ，SFE technology was optimized with orthogonal test ，and validation test was performed. Chemical components in the extract were identified by GC-MS. Relative percentage of each component was calculated with area normalization method. RESULTS ：The optimal SFE extraction technology of Fructus Gleditsiae Abnormallis was extraction temperature of 60 ℃，extraction pressure of 300 MPa and pression time of 15 min. Average extraction of 3 times of validation tests was 1.73％（RSD＝1.78％，n＝3）. The 48 components in the extracts of Fructus Gleditsiae Abnormallis were identified，which accounted for 98.31％ of the total amount of the extracts. The extracts of Fructus Gleditisae Abnormalis mainly included organic acids ，accounting for 36.99％，followed by alkaloids ，accounting for 12.59％ in total. Main components were palmitic acid （16.62％），oleic acid （14.12％），N-aminotetrahydropyrrole（9.79％），2，6-dimethyloctane-1，7-dien-3-ol（5.95％）， tetrahydropyran（3.83％），vanillin（3.39％），etc. CONCLUSIONS ：SFE method of Fructus Gleditisae Abnormalis is established successfully，and the extract is mainly organic acids.

OBJECTIVE： To investigate the effects of different doses of total alkaloids from Aconitum racemulosum （ARTA） on serum inflammation factors and FOS protein expression in synovial tissue of joint in collagen-induced arthritis （CIA） model rats， and to investigate its potential mechanism of anti-rheumatoid arthritis （RA）. METHODS： Male SD rats were randomly divided into blank group， model group， positive group （Compound dexamethasone acetate ointment， 0.2 g/kg）， ARTA low-dose， medium-dose and high-dose groups （56.26， 112.50， 225.00 mg/kg， by the weight of ARTA in the extract）， with 10 rats in each group. Except for blank group， other groups were given subcutaneous injection of Bovine collagen Ⅱ emulsified with incomplete Freund’s adjuvant into the left foot to establish CIA model； the left foot were smeared with relevant medicine from the day of modeling. Blank group and model group were smeared with constant volume of 65％ ethanol， 3 times a day， for consecutive 28 days. On the 7th， 14th， 21st and 28th day of administration， the thickness of left hind toe was measured with vernier caliper， and the degree of foot swelling was calculated. The serum contents of IL-1β， IL-6 and TNF-α in rats were measured by ELISA after last administration. The expression of FOS protein in synovial tissue was determined by immunohistochemical method [expressed by HIS]. The comprehensive score was conculated by entropy weight method. Effects of each dosage on above indexes of CIA model rats were evaluated with the comprehensive score. RESULTS： Compared with blank group， the degree of foot swelling， serum content of inflammatory factors and HIS value were increased significantly in model group （P＜0.05）. Compared with model group， the degree of foot swelling in each administration group， serum contents of IL-1β， IL-6 and TNF-α， HIS in positive group and ARTA high-dose group， serum contents of IL-6 and TNF-α in ARTA medium-dose group as well as serum content of TNF-α in ARTA low-dose group were decreased significantly（P＜0.05）. Comprehensive score of above indicators were 0.37（positive group）， 0.31（ARTA high-dose group）， 0.23（ARTA medium-dose group） and 0.09（ARTA low-dose group）. CONCLUSIONS： ARTA can improve CIA model rats， and the effect tends to increase with the increase of dose. Above effect may be associated with reducing serum content of inflammatory factors and inhibiting the expression of FOS protein in synovial tissue.