【正文】 n). MTT assay showed that SGC7901 pGPU/GFP/Neo/livin2 transfectants were more sensitive to cisplatin and 5fluorouracil than negative control and parent cells (Figs. 7A, 6B). The number of apoptotic cells induced by cisplatin and 5fluorouracil increased to about 168。C3fold in pGPU/GFP/Neo/livin2 transfectants pared with their control cells (P 。 Fig. 7C). Furthermore, stable transfectants underwent spontaneous apoptosis more readily without proapoptotic stimuli than the control cells (P 。 Fig. 7C).4. Discussion In this study, we show that livin, a new member of the IAP family, was found to be not expressed in any of the NOT cancerous gastric tissues, and expressed only in a proportion of gastric cancer patients (%), and also show that suppressing livin expression or function causes spontaneous apoptosis and inhibition of SGC 7901 cells growth and make cells more susceptible to proapoptotic stimuli. It was thought that livin has two isoforms, a and b. Although both isoforms are involved in blocking apoptosis induced by TNFa and antiCD95 in vitro, they show some different antiapoptotic properties. livin b seems to be more effective than livin a in blocking apoptosis induced by DNA damaging agents[13]. Some study on tissue distribution of livin has recently shown that elevated levels of both livin isoforms a and b have been detected in heart, placenta, lung, spleen and ovary, while livin balone has been detected specifically in fetal tissues and dult kidneyand livin a alone has been detected in brain, skeletal muscle and peripheral blood lymphocytes [1114]. Furthermore, while livin expression was detected in a variety of cancerous cell lines and some tumor tissues [1418] and antilivin antibody was recognized in sera of gastric cancer and lung cancer patients [19,20], no data were available concerning the expression of livin isoforms in gastric tumor tissues. Our study for the first time demonstrates that livin isoforms a and b were almost both expressed in a proportion of gastric cancer tissues (%) and livin expression correlate with some of the known prognostic variables, such as grade and lymphonode metastasis. Data from the literature have demonstrated that both livin isoforms are involved in blocking apoptosis and may give cells with livin overexpression a strong resistance to chemotherapyinduced apoptosis. Gastric cancer in general is highly resistant to chemoradiotherapy and moderately resistant to apoptosis [21]. These result suggested that overexpression of livin may effect the responsibility of chemotherapy on some gastric cancer patients and prognosis of patients. The specific interference with factors contributing to the apoptosis resistance of tumor cells may provide a novel basis for the development of rational intervention strategies in cancer therapy [22,23]. Since the expression of livin could contribute to the apoptosisresistant phenotype of cancer cells and its specific expression in tumors could make livin an interesting therapeutic target for tumorspecific intervention strategies, we chose the livin gene as a molecular target. The shRNA technology representiong an extremely powerful tool to inhibit endogenous gene expression [24,25] be made to inhibit livin gene and attempt to correct the apoptosis deficiency of gastric tumor cells. The efficacy of shRNAs to silence expression of a tageted gene is different, relation with the halflife and abundance of the gene product as well as with accessibility of target mRNA [2427]. In this study, we observed that silivin1 was regularly more strongly silence the livin gene than silivin2. Our study results also shown that silencing livin gene expression may strongly increase apoptotic response of SGC7901 cells in the presence or absence of several proapoptotic agents and inhibit the cells growth, which indicate that the interference with livin leads to a sensitization to proapoptotic stimuli. The similar result on hela cell was reported by CrnkovicMertens [18].In summary, our results showed that inhibition of livin expression and function resulted in spontaneous apoptosis and inhibitor cell growth enhanced sensitivity to cytotoxic drugs in vitro. Because of the preferential expression of livin in gastric cancer but not in normal tissues, these data suggest that targeting the livin pathway alone or with cytotoxic drugs may be useful in the treatment of gastric cancer. Despite their therapeutic potential, major technical hurdles still have to be overe, in order to apply shRNAs as drugs. Under therapeutic aspects, will have to meet the general challenges of gene therapy approaches, such as efficient delivery into the target cells or the circumvention of immune responses. Notably, recent in vivo studies showed that shRNAs could be directly applied to organs of postnatal mice by highpressure injection into the tail vein, leading to the specific inhibition of target genes [2830]. These data show that a direct application of active shRNAs via the bloodstream is principally feasible.