A Combined ULBP2 and SEMA5A Expression Signature as a Prognostic and Predictive Biomarker for Colon Cancer

Abstract

Background: Prognostic biomarkers for cancer have the power to change the course of disease if they add value beyond known prognostic factors, if they can help shape treatment protocols, and if they are reliable. The aim of this study was to identify such biomarkers for colon cancer and to understand the molecular mechanisms leading to prognostic stratifications based on these biomarkers. Methods and Findings: We used an in house R based script (SSAT) for the in silico discovery of stage-independent prognostic biomarkers using two cohorts, GSE17536 and GSE17537, that include 177 and 55 colon cancer patients, respectively. This identified 2 genes, ULBP2 and SEMA5A, which when used jointly, could distinguish patients with distinct prognosis. We validated our findings using a third cohort of 48 patients ex vivo. We find that in all cohorts, a combined ULBP2/SEMA5A classification (SU-GIB) can stratify distinct prognostic sub-groups with hazard ratios that range from 2.4 to 4.5 (p≤0.01) when overall- or cancer-specific survival is used as an end-measure, independent of confounding prognostic parameters. In addition, our preliminary analyses suggest SU-GIB is comparable to Oncotype DX colon(®) in predicting recurrence in two different cohorts (HR: 1.5-2; p≤0.02). SU-GIB has potential as a companion diagnostic for several drugs including the PI3K/mTOR inhibitor BEZ235, which are suitable for the treatment of patients within the bad prognosis group. We show that tumors from patients with worse prognosis have low EGFR autophosphorylation rates, but high caspase 7 activity, and show upregulation of pro-inflammatory cytokines that relate to a relatively mesenchymal phenotype. Conclusions: We describe two novel genes that can be used to prognosticate colon cancer and suggest approaches by which such tumors can be treated. We also describe molecular characteristics of tumors stratified by the SU-GIB signature.

Keywords: Biomarker.; Colon Cancer; Prognosis.

Figure 1

SEMA5A/ULBP2 (SU) gene signature for colon cancer prognostication. Kaplan-Meier graphs based on the SU signature for GSE17536 (A), GSE17537 (B) and the Ankara cohort (C), and their respective log-rank p values are shown. Survival times are in months. A combined score based on ULBP2 and SEMA5A expression that separates patients into good (SEMA5A high, ULBP low), bad (SEMA5A low, ULBP high) and intermediate (patients with both SEMA5A and ULBP2 high or both low) groups results in better stratification of colon cancer patients. ULBP2 and SEMA5A cut-off values were 4 and 6, respectively for in silico analyses. For the Ankara cohort, cut-off values correspond to the median expression value for both genes.

Figure 2

SU-GIB can predict recurrence-free survival especially for microsatellite stable patients in GSE39582. Kaplan-Meier graphs of microsatellite instable (MSI, A) and stable (MSS, B) patients stratified independently using SU-GIB, and log-rank p values are shown. Cut-off values for both genes were the most significant within the 25^th and 75^th interquartile range.

Figure 3

SU-GIB can predict recurrence-free survival for micosatellite stable patients with stage 2 or 3 disease in GSE39582. Kaplan-Meier graphs and log-rank p values are shown for stage 2 (A, B) and 3 (C, D) patients with microsatellite instable (A, C) and stable (B, D) tumors. Cut-off values for both genes were the most significant within the 25^th and 75^th interquartile range.

Figure 4

TCGA based proteome analysis of colon cancer tumor tissue reveals increased EGFR phosphorylation and decreased Caspase 7 cleavage in the good prognosis group. RNA seq. and proteome data for 132 colon cancer primary tumor tissues downloaded from “cancergenome.nih.gov” via the TCGA data portal classified according to the SU signature (bad survivors: 40, good survivors: 37, intermediate survivors: 55) revealed increased EGFR 1068 phosphorylation among patients with better prognosis (p<0.001, 1-way Anova) (left). The same analysis showed increased Caspase 7 cleavage in patients within the worse survival group (p<0.0001) (right). Caspase 7 cleavage was directly correlated with ULBP2 and inversely with SEMA5A expression (p<0.001 for both genes by Pearson's r). The mean and standard deviation for each group are indicated.

Figure 5

TCGA based RNA seq. analysis of colon cancer tumor tissue reveals increased inflammatory cytokine gene expression in the bad prognosis group. TCGA colon tumor samples were stratified into GIB groups using median expression values as cut-offs for ULBP2 and SEMA5A. RPKM values of IL6, IL1B, TGFB1 are plotted for “good”, “intermediate” and “bad” groups. T-test p values between “good” and “bad” groups. *p < 0.05, **p <0.01, ***p <0.0001. RPKM: reads per kilobase per million mapped reads. The median and inter-quartile ranges for each group are indicated.

Figure 6

Chemosensitivity profiles of colon cancer cell lines corresponding to SU-G, -I and -B phenotypes. In silico analysis of colon cancer cell lines as classified by the SU signature into G, I or B phenotypes for NVP-BEZ235 cytotoxicity based on CGP data. Cut-off values used to generate a SU-GIB stratification were either the SSAT generated (SEMA5A) or the median expression values (ULBP2).