RESUMO
OBJECTIVE：To optimize the extraction technology of Zhideke granules. ME THODS：The extraction technology （water extraction ，alcohol extraction ，water extraction and ethanol precipitation ）of Zhideke granules was initially screened by ammonia-induced cough experiment and xylene-induced ear swelling experiment in mice. Based on its preparation route ，the immersion time of medicinal materials containing volatile oil was investigated with water absorption as index firstly. The single factor test was adopted to investigate the amount of water added and the extraction time taking the volatile oil yield as index to optimize the extraction technology of medicinal materials containing volatile oil. Taking the contents of irisflorentin and total flavonoids as indicators ，on the basis of single factor investigation ，orthogonal test was adopted to examine the influence of three factors including the amount of water added ，extraction time and extraction frequency ，so as to optimize the water extraction technology of Zhideke granules and the validation tests were conducted. RESULTS ：The results of pharmacodynamics experiment showed that the cough latency of mice in water extract low-dose and high-dose groups （6.34，12.68 g/kg，by crude drug ）and water-extraction alcohol-precipitation extract high-dose group （12.68 g/kg，by crude drug ）were significantly longer than those inmodel group ，and the number of cough within 2 minutes was significantly reduced （P＜0.05 or P＜0.01）. Compared with model group ， the ear swelling of mice in water extract low-dose and high-dose groups （6.34，12.68 g/kg，by crude drug），ethanol extract high-dose group （12.68 g/kg，by crude drug） and water-extraction alcohol-precipitation extract hig dose group （12.68 g/kg，by crude drug ） were decreased significantly （P＜0.05 or P＜0.01）. The swelling inhibition rates were 42.26％，55.08％，33.49％，51.56％，39.57％ and 44.36％ in low-dose and high-dose groups of water extract ，alcohol extract ， water-extraction and alcohol-precipitation extract respectively ，indicating that the water extract had better antitussive and anti-inflammatory effects. The optimal extraction technology of volatile oil was adding 5-fold water ，soaking for 30 minutes，and extracting for 3 hours. The optimal water extraction technology was adding 12-fold water ，extracting for 3 times after soaked for 50 min，lasting for 1 h each time. Results of 3 times of validation tests showed that average content of irisflorentin in the extract obtained by optimal technology was 76.47 μg/g（RSD＝ 2.15％，n＝3）and the average content of total flavonoids was 92.45 mg/g（RSD＝0.48％，n＝3）. CONCLUSIONS ：The optimal extraction technology of Zhideke granules is stable and feasible.
RESUMO
OBJECTIVE:To optimize the hypoglycemic active parts from the leaves of Dimocarpus longan for type 2 diabetes mellitus mice,and to provide reference for material basis study of it. METHODS:Total extract was obtained from the leaves of D. longan with 95% ethanol percolation extraction. After confirming hypoglycemic effect of total extract,extract was suspended with water. Relevant extract was obtained by using petroleum ether,ethyl acetate and n-butanol in turn(respectively marked as ST,YT,ZT and WT). The model of type 2 diabetes mellitus mice was established by high glucose and high fat diet combined with STZ. Model mice were randomly divided into model group(distilled water),positive control group(metformin,0.1 g/kg), ST low-dose,medium-dose and high-dose groups(22.2,29.6,44 g/kg,calculated by crude drug,similarly hereinafter),YT low-dose,medium-dose and high-dose groups(10.7,16,32 g/kg),ZT low-dose,medium-dose and high-dose groups(10.5, 15.8,31.6 g/kg),WT low-dose,medium-dose and high-dose groups(18.5,24.6,37 g/kg),with 10 mice in each group. Other 10 normal mice were included in normal group(distilled water). Those groups were given relevant medicine intragastrically,once a day,for consecutive 28 d.The growth of mice and the change of body weight were observed and recorded. Fasting blood glucose(FBG)of mice was detected,and glucose tolerance test of mice was conducted. The contents of TC,TG,HDL-C and LDL-C in serum of mice were also determined. RESULTS:Compared with normal group,body weight and serum content of HDL-C in model group were decreased;FBG and the blood sugar after giving glucose for 0.5,1,2 h and area under glucose tolerance curve of mice were increased,and serum contents of TC,TG and LDL-C were alse increased,with statistical significance(P<0.01). compared with model group,body weight of mice in YT high-dose group after administration of 25 d was increased,FBG of mice in YT groups and ZT medium-dose,high-dose groups were decreased(P<0.05 or P<0.01);blood sugar of mice in YT medium-dose,high-dose groups and ZT high-dose group after giving glucose for 1 h as well as blood sugar of mice in YT high-dose group after giving glucose for 2 h were all decreased;area under glucose tolerance curve of mice and serum content of TC were decreased in YT groups and ZT high-dose group,while serum content of HDL-C was increased. Serum content of TG and LDL-C in YT high-dose group and ZT high-dose group as well as serum content of TG in YT medium-dose group were all decreased,serum content of HDL-C in ZT medium-dose group was increased,with statistical significance(P<0.05 or P<0.01). CONCLUSIONS:The ethyl acetate part and n-butanol part from the leaves of D. longan show good hypoglycemic effect on type 2 diabetes mellitus mice.