Objective To compare the sub-cellular proteome of isoniazid ( INH)-resistant Mycobacterium tuberculosis (MTB) with that of sensitive strains for identifying of unique proteins of these strains and discussing their preliminary application in clinical diagnosis. MethodsProteins of cell wall and membrane of 5 INH-resistant strains and 5 INH-sensitive strains were extracted by density gradient centrifugation.The extracts were subsequently analyzed using weak cation exchange (WCX) liquid chromatography ( LC )followed reverse phase (RP) liquid chromatography to compare the sub-cellular protein patterns. A total of 1280 fractions were collected and identified by matrix assisted laser desorption ionization-time of flight-tandem mass spectrometry (MALDI-TOF MS/MS). The Database for Annotation, Visualization and Integrated Discovery (DAVID) was used for cell component and biological process analysis. Normalized Spectral Abundance Factors (NASF) was used for semi-quantity of protein expression. 5 proteins significantly up-regulated in INH-resistant strains and 2 proteins significantly up-regulated in INH-sensitive strains were selected for ELISA analysis with autologous sera respectively. Results A total of 347 proteins were identified. Cell component analysis showed that 58％ proteins were cells well or membrane proteins. Biological process analysis showed that 31％ proteins involved in carboxylic/monocarboxylic acid biosynthetic and metabolic process, 26％ and 15％ proteins involved in organic acid or fatty acid biosynthetic and metabolic process,while 28％ proteins involved in lipid biosynthetic , metabolic, transport and localization process. O-succinylbenzoate synthase, monooxygenase, hypothetical protein Rv2255c, nicotinate-nucleotide--dimethylbenzimidazole phosphoribosyltransferase and membrane phosphatidate cytidylyltransferase cdsA were up-regulated in INH-resistant strains and fractions contained these proteins could elicit specific antibody response with autologous sera. The A450 was higher than that with INH-sensitive sera. The differences between the INH-resistant sera and the INH-sensitive sera were significant ( t = 0.028, 0.044, 0.066, 0.064, 0.083, all P＜0.01 ). Chain A of Rv2002 Gene Product and Chain A of Crystal Structure Of Rv2632c were up-regulated in INH-sensitive strains and fractions contained these proteins could elicit specific antibody response with autologous sera. The A450 was higher than that with INH-resistant sera. The differences between the INH-sensitive sera and the INH-resistant sera were significant (t=0.053, 0.073, both P＜0.05). ConclusionThe combination of density gradient centrifugation and 2D-LC MS/MS technology is useful in enrichment and identification of differential expressed proteins between INH-resistant and INH-sensitive strains at sub-cellular level. It is useful in finding antigens associated with INH-resistant MTB infection, which may prove useful for further study in the mechanism of INH resistant, as well as interaction between MTB and host.

ObjectiveTo explore and establish the liver injury risk prediction model of indirect toxicity of Chinese medicinals under the condition of compound formulas， and provide new ideas and methods for the study of evaluation of liver injury of Chinese medicinals based on indirect toxicity. MethodsTaking Buguzhi （Fructus Psoraleae） pre-parations as model drug， the combined Chinese medicinals with Buguzhi （Fructus Psoraleae） of high frequency are screened out， and their components and action targets were obtained through TCMSP， TCMIP and PharmMapper databases. The association strength value and risk value of Chinese medicinals that acted on the nuclear factor κB （NF-κB） pathway were analyzed. For those having greater values than the median association strength value and risk value were regarded as indirect Chinese medicinals of liver injury risk. In this way， a prediction model of liver injury risk of Chinese medicinals was constructed based on immune activation-related indirect liver injury process （taking NF-κB pathway as an example）. And verification of the prediction model was performed using Heshouwu （Radix Polygoni Multiflori） preparations. ResultsThe prediction model of liver injury risk based on important immunoactivated pathway （taking NF-κB pathway as an example） found that Yinyanghuo （Herba Epimedii） （association strength value = 0.18， risk value = 0.25） was a Chinese medicinal with potential risk of indirect liver injury within Buguzhi （Fructus Psoraleae） prepartions， which may increase the risk of liver injury by positively regulating Bruton's tyrosine kinase （Btk） and protein kinase C theta （PKCθ） on NF-κB pathway. Further verification of prediction model by Heshouwu （Radix Polygoni Multiflori） preparations showed that Buguzhi （Fructus Psoraleae） （association strength value = 0.25， risk value = 0.33） and Tusizi （Semen Cuscutae） （Semen Cuscutae， association strength value = 0.34， risk value = 0.33） may increase the liver injury risk of Heshouzu. ConclusionThe liver injury risk prediction model of indirect toxicity of Chinese medicinals has been constructed in this study， providing metho-dological reference for the identification of Chinese medicinals of indirect liver injury risk under the condition of compound formulas.