P5-01-10: A Potential New Marker for, and Facilitator of, Hormone Independence.

Abstract

Abstract Triple negative breast cancers (TNBC) are generally more aggressive and have a worse prognosis than cancers that express estrogen (ER), progesterone (PGR) and/or human epidermal growth factor receptor 2 (HER2). Recent data suggests that another hormone receptor, that for androgen (AR) may be relevant in terms of further stratifying TNBC, with quadruple negative breast cancer (QNBC) having a worse prognosis than TNBC. In addition, TNBC vary with respect to prognosis and sensitivity to specific treatments, and foreknowledge of treatment outcomes has been the goal of numerous studies designed to generate predictive gene signatures. Development of these signatures involves the use of algorithms that can potentially exclude important biomarkers that may appear to be too weakly expressed to make the cut-off. We report here that the lipid metabolic enzyme, long-chain fatty acyl-CoA synthetase 4 (ACSL4) falls into this category of a heretofore-unrecognized biomarker for aggressive breast cancer. We have previously reported that ACSL4 expression is negatively correlated with the expression of ER and AR in both breast cancer cell lines and tumor samples. We now report a positive correlation between ACSL4 mRNA expression and QNBC status. In two separate studies involving a total of 71 separate breast cancer cell lines, values for ACSL4 expression derived from microarray data indicated increased expression in QNBC cell lines, p=3.75E-08 and p=4.59E-08, when compared with cell lines expressing one or more receptor biomarkers. An analysis of data derived from 178 tumor samples indicated a similar correlation, p=2.2E-03. ACSL4 expression was also correlated with molecular subtype. When breast cancer cell lines were classified by subtype, 24 of 26 luminal cell lines were negative for ACSL4 expression, 8 of 8 claudin-low cell lines were positive for ACSL4 expression, as were 12 of 17 basal-like cell lines. Three basal-like cell lines expressed both ACSL4 and HER2. Upon further examination, these lines were demonstrated to express relatively low levels of other markers that have been suggested to correlate with HER2 sensitivity (ESR1, XBP1, FASN, ERBB3), as well as to express relatively high levels of markers correlated with HER2 resistance (CD44, CAV2). In order to determine the relationship between ACSL4 expression and estrogen independence, ACSL4-negative MCF-7 cells were inducibly transfected with ACSL4 cDNA. Induction of ACSL4 resulted in abrogation of the proliferation response to estrogen, accompanied by a reduction in expression of ER, AR and PGR. In addition, induced cells were less sensitive to treatment with tamoxifen, etoposide and triacsin C. Information derived from RAF-1-transfected MCF-7 microarray data suggest that ACSL4 expression can be regulated by RAF-1. To further explore this notion, QNBC cells, MDA-MB-231, which express high levels of ACSL4, were treated with siRNA for RAF-1 and ACSL4 levels monitored by RT-PCR and immunoblot. The fall in levels of expression of RAF-1 was accompanied by a similar decrease in expression of ACSL4. Reduction of ACSL4, however, had no effect on RAF-1 expression. In summary, these data demonstrate a possible role for ACSL4 as a biomarker for prognosis and treatment outcome, as well as a facilitator of hormone resistance. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P5-01-10.