Isabell Bastian

The biomarker development field within molecular medicine remains limited by the methods that are available for building predictive models. We developed an efficient method for conservatively estimating confidence intervals for the cross validation-derived prediction errors of biomarker models. This new method was investigated for its ability to improve the capacity of our previously developed method, StaVarSel, for selecting stable biomarkers. Compared with the standard cross validation method, StaVarSel markedly improved the estimated generalisable predictive capacity of serum miRNA biomarkers for the detection of disease states that are at increased risk of progressing to oesophageal adenocarcinoma. The incorporation of our new method for conservatively estimating confidence intervals into StaVarSel resulted in the selection of less complex models with increased stability and improved or similar predictive capacities. The methods developed in this study have the potential to improve progress from biomarker discovery to biomarker driven translational research.

Oesophageal adenocarcinoma is a rapidly increasing problem in which treatment options are limited. Previous studies have shown that oesophageal adenocarcinoma cells and tissues express oestrogen receptors (ERs) and show growth suppression and apoptosis in response to ER modulator agents such as tamoxifen. ERs are known to be expressed in a number of isoforms that act together to regulate cell growth and cell death. In this study, we used western blotting to profile the expression of ERα and ERβ isoforms, and expression of the oncologically related molecules p53, HER2, and EGFR, in a panel of oesophageal adenocarcinoma cell lines. The cytotoxicity of tamoxifen in the cell lines was determined with Annexin V-FITC flow cytometry, and correlations between cytotoxicity and receptor expression were assessed using Spearman’s rank-order correlation. Oesophageal adenocarcinoma cell lines showed varying cytotoxicity in response to tamoxifen. The ER species ERα90, ERα50, and ERα46, as well as p53, were positively associated with a cytotoxic response. Conversely, ERα74, ERα70, and ERβ54 were associated with a lack of cytotoxic response. The ER species detected in oesophageal adenocarcinoma cells may work together to confer sensitivity to ER modulators in this disease, which could open up a new avenue for therapy in selected patients.

TP53 gene mutations occur in 70% of oesophageal adenocarcinomas (OACs). Given the central role of p53 in controlling cellular response to therapy we investigated the role of mutant (mut-) p53 and SLC7A11 in a CRISPR-mediated JH-EsoAd1 TP53 knockout model. Response to 2 Gy irradiation, cisplatin, 5-FU, 4-hydroxytamoxifen, and endoxifen was assessed, followed by a TaqMan OpenArray qPCR screening for differences in miRNA expression. Knockout of mut-p53 resulted in increased chemo- and radioresistance (2 Gy survival fraction: 38% vs. 56%, p < 0.0001) and in altered miRNA expression levels. Target mRNA pathways analyses indicated several potential mechanisms of treatment resistance. SLC7A11 knockdown restored radiosensitivity (2 Gy SF: 46% vs. 73%; p = 0.0239), possibly via enhanced sensitivity to oxidative stress. Pathway analysis of the mRNA targets of differentially expressed miRNAs indicated potential involvement in several pathways associated with apoptosis, ribosomes, and p53 signaling pathways. The data suggest that mut-p53 in JH-EsoAd1, despite being classified as non-functional, has some function related to radio- and chemoresistance. The results also highlight the important role of SLC7A11 in cancer metabolism and redox balance and the influence of p53 on these processes. Inhibition of the SLC7A11-glutathione axis may represent a promising approach to overcome resistance associated with mut-p53.