Objective To investigate the effect of insulin and gliclazide therapy on endoplasmic reticulum (ER) stress and insulin sensitivity in the liver of type 2 diabetic rats.Methods A high fat diet plus a low-dose of streptozotocin was implemented to create a type 2 diabetic rats which were randomly divided into diabetes mellitus (DM) group,insulin treatment (INS) group and gliclazide treatment (GT)group; and healthy rats were as normal control group.Diabetic rats in INS and GT groups were given neutral protamine hagedorn (NPH) insulin and gliclazide respectively for 3 weeks.Protein expression levels of immunoglobulin binding protein (Bip),spliced X-box binding protein 1 (XBP-ls),phosphorylated c-Jun on serine 73 (p-c-Jun),phosphorylated insulin receptor substrate 1 on serine 307 (p-IRS-1),and glucose-6-phosphatase (G6Pase) in liver homogenate were detected by Western blotting.Results Compared with the normal rats,Bip and XBP-Is in the DM group were up-regulated (0.28 ±0.07 vs 0.90 ±0.10 for Bip;0.41 ± 0.07 vs 0.95 ±0.07 for XBP-1 s; both P ＜ 0.01 ) ; p-c-Jun (0.59 ± 0.18 vs 1.94 ± 0.03 ),p-IRS-1( 1.73 ± 0.18 vs 5.32 ± 0.22) and G6Pase (0.11 ± 0.01 vs 0.45 ± 0.01 ) were increased ( all P values ＜0.01 ).In the INS group,all of aforementioned changes were reversed (0.90 ± 0.10 vs 0.25 ± 0.04 for Bip; 0.95 ±0.07 vs 0.47 ±0.01 for XBP-1s; 1.94 ± 0.03 vs 0.50 ±0.10 for p-c-Jun; 5.32 ± 0.22 vs 1.59 ±0.32 for p-IRS-1 ; 0.45 ±0.01 vs 0.15 ±0.02 for G6Pase,all P values ＜0.01 ).In the GT group,all of aforementioned changes were also attenuated ( 0.90 ± 0.10 vs 0.53 ± 00.02 for Bip ; 0.95 ± 0.07 vs 0.78±0.02 for XBP-1s; 1.94 ±0.03 vs 1.33 ±0.11 for p-c-Jun; 5.32 ±0.22 vs 3.13 ±0.02 for p-IRS-1; 0.45 ± 0.01 vs 0.25 ± 0.01 for G6Pase,all P values ＜ 0.05).Furthermore,all of aforementioned protein levels were down-regulated more obviously in the INS group comparing to the GT group ( all P values ＜ 0.01 ).Conclusions Both insulin and gliclazide therapy could relieve ER stress and e-Jun N-terminal kinase activity and improved insulin sensitivity.The effect of insulin on Bip,XBP-1s,p-c-Jun,p-IRS-1 and G6Pase protein expressions is more obvious than that of glilcazide,which indicates besides lowering glucose,insulin might have protective effects of anti-inflammation,anti-oxidative stress or stimulation of lipid redistribution.

To explore the application of protein fingerprint technique and differential diagnosis in bacteriological negative pulmonary tuberculosis and pneumonia ,60 patients with bacteriological negative pulmonary tuberculosis ,60 patients with pneumonia ,and 60 healthy volunteers were selected from known clinical cases .Surface strengthening laser desorption ioniza-tion time of flight mass spectrometry (SELDI ToF Ms) and protein chip technology were applied to detect serum proteins ,and analyze their protein peaks by Ciphergen protein chip 3 .1 .1 software .Comparison of the serum protein fingerprinting data from the pool of 180 patients and healthy volunteers showed significant difference in 5 protein peaks (1 028 .49 ,4 796 .56 ,7 564 .77 , 8 048 .02 ,and 11 526 .75 m/z) identified between pulmonary tuberculosis and pneumonia (P<0 .01) .The total effective rate of the 5 protein peaks as a diagnosis model for differential diagnosis of bacteriological negative pulmonary tuberculosis and pneumonia was 84 .2% (101/120) ,the specificity was 82 .5% (52/63) ,the sensitivity was 85 .9% (49/57) ,the positive pre-dictive value was 86 .7% (52/60) ,and the negative predictive value was 81 .7% (49/60) .The total effective rate of the diagno-sis model for differential diagnosis of bacteriological negative pulmonary tuberculosis ,pneumonia and healthy volunteers was 89 .4% (161/180) .The specificity was 100% (60/60) ,the sensitivity was 84 .2% (101/120) ,the positive predictive value was 100% (101/101) ,and the negative predictive value was 75 .9% (60/79) .Protein fingerprinting technology is advanta-geous of being a simple method ,quick detection ,and requires less amount of sample .It is an effective means to screening the tuberculosis specific markers .We found the good diagnosis model through the detection of serum protein by protein fingerprint-ing technology .