Huang G M, Ng W L, Farkas J, He L, Liang H A, Gordon D, Yu J, Hood L
Department of Molecular Biotechnology, University of Washington, Seattle, Washington 98195, USA.
Genomics. 1999 Jul 15;59(2):178-86. doi: 10.1006/geno.1999.5822.
Prostate cancer is a frequently diagnosed solid tumor that is originated mostly from prostate epithelium. One of the key issues in prostate cancer research is to develop molecular markers that can effectively detect and distinguish the progression and malignancy of prostate tumors. Automated, single-pass cDNA sequencing was utilized to rapidly identify expressed genes in a number of cDNA libraries constructed from various normal and tumor prostatic tissues. These included cell lines as well as short-term epithelial culture. A total of 6604 expressed sequence tags (ESTs) were generated and searched against on-line nucleotide and protein databases. A relational database centric software system was constructed to process, store, and analyze EST data rapidly. cDNA contigs were also obtained by assembly of multiple EST sequences. Protein structural signatures were annotated using motif analysis tools including BLOCKS and an in-house-designed neural network. Cross-library comparisons revealed their unique gene expression profiles. Several differentially expressed cDNA clones were identified, and their expression patterns were confirmed by RNA dot blot and RT-PCR analyses.
前列腺癌是一种常见的实体瘤，主要起源于前列腺上皮。前列腺癌研究的关键问题之一是开发能够有效检测和区分前列腺肿瘤进展及恶性程度的分子标志物。利用自动化单通道cDNA测序技术，快速鉴定从各种正常和肿瘤前列腺组织构建的多个cDNA文库中表达的基因。这些文库包括细胞系以及短期上皮培养物。共产生了6604个表达序列标签（EST），并与在线核苷酸和蛋白质数据库进行比对。构建了一个以关系数据库为中心的软件系统，用于快速处理、存储和分析EST数据。还通过多个EST序列的组装获得了cDNA重叠群。使用包括BLOCKS和内部设计的神经网络在内的基序分析工具对蛋白质结构特征进行注释。跨文库比较揭示了它们独特的基因表达谱。鉴定出几个差异表达的cDNA克隆，并通过RNA斑点印迹和RT-PCR分析证实了它们的表达模式。
