Abstract 2235: GAL-1 function as an alternative PI3K pathway activator depends on PTEN and EGFR status in CHO cells

Abstract

Abstract The PI3K signaling pathway, one of the most potent prosurvival pathways, has been associated with adverse outcome in glioma. The frequency of signaling alterations in the RTK/PI3K/RAS/PTEN axis in glioblastoma (GBM) is 88% and leads to therapeutic resistance and poor prognosis. Previous results in our lab have shown that Galectin-1 (GAL-1) gene expression is prognostic in five glioma microarray datasets and in four out of five GBM subpopulations. Our lab also has data suggesting an interaction between GAL-1 and PI3K leading to increased PI3K activity. The purpose of this study was to examine GAL-1's influence on survival in GBM patients in regard to their PI3K status. Therefore an exploratory subgroup analysis of the TCGA GBM dataset was conducted and findings were validated in preliminary cell culture model. The TCGA dataset was normalized using previously described methods. Survival analysis was performed using BRB tools. Survival distributions within the dataset were examined among two predefined subgroups: Patients with both EGFR mRNA expression below the median and PTEN expression above the median formed the “low PI3K” group and all other patients formed the comparison “high PI3K” group. The Kaplan Meier curves generated using graphpad prism were stratified by GAL-1 (median split) expression. Stable constructs of CHO cells transfected with an LGALS1 expression vector and/or with PTEN shRNA were created and transiently transfected with an EGFR expression vector. Western blots and clonogenic survival assay were performed. The group of patients with “low PI3K” activity had better outcome than the patients with “high PI3K”. This result suggests that patients with lack of PI3K activating EGFR over expression or PTEN loss have improved survival outcome. When survival of the “low PI3K” group is stratified for GAL-1 expression we found patients with high levels of GAL-1 to have significantly worse outcomes when compared to patients with low GAL-1 levels within this group. Analyzing survival for the patients with “high PI3K” activity, we did not observe any difference in prognosis when stratifying for GAL-1 expression. These results led to the hypothesis that high GAL-1 results in reduced patient survival through increased PI3K signaling. We hypothesize that the CHO cells without EGFR, reexpressed PTEN and GAL-1 knockdown have the lowest expression of pAKT. Furthermore we hypothesize that these cells are also the most radiosensitive cells in a clonogenic survival assay. These results indicate that GAL-1 could be an important prognostic marker and that over expression of GAL-1 is functionally similar to loss of PTEN or upregulation of EGFR. They also indicate that GAL-1 might be an alternative mediator of PI3K signaling. These results will be validated in a panel of glioma cell lines. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2235. doi:1538-7445.AM2012-2235