[Cluster analysis and identify significance of novel genes related to glioma in molecule classification]

To screen differentially expressed genes in the development of human glioma and establish molecular classification of glioma preliminary based on gene expression using cDNA microarray.

Brain specimens were obtained from 18 patients with glioma, 10 males and 8 females, aged 14 approximately 62 with an average age of 44.4. The total RNAs of these glioma specimens and 2 specimens of donated brain of normal adults were extracted. BioStarH140S microarray (including 8347 old genes and 5592 novel genes) were adopted and hybridized with probes which were prepared from the total RNAs. Differentially expressed genes between the normal tissues and glioma tissues were assayed after scanning cDNA microarray with ScanArray 4000. Northern hybridization, and in situ hybridization (ISH) were used to identify the functions of novel genes. Those differentially expressed genes were studied with Hierarchical method and molecular classification of glioma was preliminarily carried out.

Among the 13 939 target genes, there were 1200 (8.61%) differentially expressed genes and 395 (2.83%) novel genes. 348 genes were up-regulated and 852 genes were down-regulated in glioma. The results of bioinformatical analysis, Northern hybridization and ISH revealed that those novel genes were highly associated with glioma. There were multiple genes which were relevance to classification by Hierarchical method, such as MAP gene, cytoskeleton and matrix motility genes, etc. Molecular classification of glioma with Hierarchical cluster was in accordance with pathology and revealed internal essence in tumorigenesis and development.

Multiple genes play important roles in development of glioma. cDNA microarray technology is a powerful technique in screening for differentially expressed genes between two different kinds of tissues.