Expressional Subpopulation of Cancers Determined by G64, a Co-regulated Module

초록

Studies of cancer heterogeneity have received considerable attention recently because the presence or absence of resistant sub-clones may determine whether or not certain therapeutic treatments are effective. Previously, we have reported G64, a co-regulated gene module composed of 64 different genes, can differentiate tumor intra- or inter-subpopulations in lung adenocarcinomas (LADC). Here, we investigated whether the G64 module genes were also expressed distinctively in different subpopulations of other cancers. RNA sequencing–based transcriptome data derived from 22 cancers except LADC, were downloaded from The Cancer Genome Atlas (TCGA). Interestingly, the 22 cancers also expressed the G64 genes in the correlated manner as observed previously in the LADC study. Considering that gene expression levels were continuous among different tumor samples, tumor subpopulations were investigated using extreme expressional ranges of the G64, i.e., tumor subpopulation with the lowest 15% of G64 expression, tumor subpopulation with the highest 15% of G64 expression and tumor subpopulation with intermediate expression. In each 22 cancer, we examined whether patients’ survivals were differentiated among the three different subgroups, and found that the G64 could differentiate tumor subpopulations in six other cancers including sarcoma, two kidneys, brain, liver, and esophageal cancers.

키워드