中文摘要：

目的：研究放射敏感与放射抵抗的鼻咽癌组织基因表达差异,寻找预测鼻咽癌放射敏感性基因指纹。方法：采用BioStarH-141s（2004）型含14112点的人类全长基因cDNA表达谱芯片,对19例不同放射敏感性鼻咽癌组织进行检测,其中放射敏感8例,放射抵抗11例,用T-test检验、聚类、LOOCV等方法进行生物信息分析。结果：放射抵抗组与放射敏感组比较,表达差异的基因P〈0.5有923条,P〈0.01有164条,P〈0.001有17条;与慢性炎症组织相比,放射敏感组70%以上病例有2倍以上表达差异的基因267条,明显高于放射抵抗组106条。这些基因包括已知的与放射敏感性密切相关的基因PRKD,DCK等;还有大部分以前没有描述与放射敏感性相关的基因,如SIRT1,RABAC1,PPP1CC,ZNF638,APPBP1,RABAC1等,它们主要与细胞周期、S期和M期调控点、细胞凋亡、细胞生长、DNA绑定,RNA绑定等功能有关;还有一些新发现的,功能尚未明确的基因,如CUTA,ZNF148,Sep15,TMTC3,TOP1等。取P〈0.01的差异表达基因98条进行聚类分析,19例患者能分成放射敏感与抵抗二类,除1例外,所有病例敏感性的分类均与临床结果相符。用LOOCV验证,准确率100%。结论：不同放射敏感性的鼻咽癌组织,基因表达存在明显差异,与炎症组织比较,放射敏感组的基因表达差异更大,可利用差异基因簇进行放射敏感性分类,提示可利用基因指纹法进行鼻咽癌放射敏感性预测。

英文摘要：

Objective： To investigate the difference in gene expression of radiation-sensitive and radiation-resistant nasopharyngeal carcinoma（NPC） tissues. Methods： Genetic fingerprinting was utilized to predict the radiation sensitivity of NPC. Nineteen NPC samples were assessed by BioStarH-141s（2004） cDNA gene expression profiling, with each chip containing 14112 spots of full length human genes. The 19 biopsies were classified into two groups： 8 samples were radiation-sensitive cases, whereas the other 11 were radiation-resistant cases. Methods of biostatistics and bioinformatics were used to analyze the differentially expressed genes. Results： Comparing gene expression profiles of the radiation-sensitive group with the resistant group, there were 164 differentially expressed genes（P〈0.01）, and 17 differentially expressed genes（P〈0.001）. In a comparison with nasopharyngeal chronic inflammatory tissue, 267 genes were found to be expressed differentially more than two-fold in at least 70% of the radiation-sensitive samples, and 106 similarly expressed genes were found in at least 70% of the radiation-resistant cases. Some of the genes were related to radiation sensitivity such as PRKD and DCK, whereas others had no reported connection with radiation sensitivity including SIRT1, RABAC1, PPP1CC and ZNF638. These genes were mostly related to cell cycle, checkpoint of S and M phases, cell death and development, or DNA or RNA binding. Also, some genes, including CUTA, ZNF148, Sep15, TMTC3 and TOP1, were recently discovered and their functions remain unknown. Using a group of 98 genes expressed differentially between the resistant and sensitive groups（P〈0.01）, 19 specimens could be classified into two groups： radiation sensitive and radiation resistant. Eighteen of them were consistent with their clinical outcome. Testified by LOOCV, the cluster accuracy was 100%. Conclusion： Gene expression of radiation sensitive NPC tissue is significantly different from that of radiation resistant tumors