Moving Toward Integrating Gene Expression Profiling Into High-Throughput Testing: A Gene Expression Biomarker Accurately Predicts Estrogen Receptor α Modulation in a Microarray Compendium

Abstract

Microarray profiling of chemical-induced effects is being increasingly used in medium- and high-throughput formats. Computational methods are described here to identify molecular targets from whole-genome microarray data using as an example the estrogen receptor α (ERα), often modulated by potential endocrine disrupting chemicals. ERα biomarker genes were identified by their consistent expression after exposure to 7 structurally diverse ERα agonists and 3 ERα antagonists in ERα-positive MCF-7 cells. Most of the biomarker genes were shown to be directly regulated by ERα as determined by ESR1 gene knockdown using siRNA as well as through chromatin immunoprecipitation coupled with DNA sequencing analysis of ERα-DNA interactions. The biomarker was evaluated as a predictive tool using the fold-change rank-based Running Fisher algorithm by comparison to annotated gene expression datasets from experiments using MCF-7 cells, including those evaluating the transcriptional effects of hormones and chemicals. Using 141 comparisons from chemical- and hormone-treated cells, the biomarker gave a balanced accuracy for prediction of ERα activation or suppression of 94% and 93%, respectively. The biomarker was able to correctly classify 18 out of 21 (86%) ER reference chemicals including "very weak" agonists. Importantly, the biomarker predictions accurately replicated predictions based on 18 in vitro high-throughput screening assays that queried different steps in ERα signaling. For 114 chemicals, the balanced accuracies were 95% and 98% for activation or suppression, respectively. These results demonstrate that the ERα gene expression biomarker can accurately identify ERα modulators in large collections of microarray data derived from MCF-7 cells.

Using a gene expression biomarker to identify DNA damage‐inducing agents in microarray profiles

“…However, the high level of accuracy demonstrated the robustness of the computational procedures despite the fact that the biosets were derived from heterogeneous experiments with various exposure conditions carried out in multiple labs. These results are consistent with our past experience in accurately identifying chemical modulators of transcription factors in the mouse liver (Oshida et al, , , ) and MCF‐7 cells (Ryan et al, ). Our approach has a number of advantages compared to other methods used to identify DDI chemicals.…”

“…Thus, in the nCounter data, the cumulative contribution of all genes to the biomarker correlations and resultant P values increased the −Log( P value)s compared to the microarray data. In our earlier studies, we found that removal of groups of genes from the ER biomarker consistently led to lower significance across bioset comparisons (Ryan et al, ). Our conclusion from these studies is that our computational approach is useful to distinguish between DDI and non‐DDI chemicals independent of the platform used to generate the gene expression data.…”

“…High‐throughput transcriptomic (HTTr) technologies have the potential to accurately identify toxic chemicals in in vitro screens of environmental chemicals. Our group has previously determined that a biomarker approach can be used to identify ER modulators in a large compendium of microarray profiles derived from chemically treated human cell lines (Ryan et al, ). In this study, we used similar computational methods to determine if our approach can also identify chemicals that cause DNA damage using the previously characterized TGx‐DDI biomarker (Li et al, ; Yauk et al, ).…”

BCScreen: A gene panel to test for breast carcinogenesis in chemical safety screening

“…For example, some genes overlap between BCScreen and PAM50 [54], the estrogenicity panel [42], and the Tox21 S1500+ panel [31]. This overlap indicates that while the other panels cover some of the biological space in BCScreen, our approach produced a unique gene set.…”

“…Studies comparing gene expression signatures of known breast carcinogens with putative non-carcinogens and with carcinogens that don't target the breast can begin to evaluate whether BCScreen can classify chemicals. A similar approach was reported by Ryan et al [42], who developed a consensus gene expression signature for estrogen action, although our expectation is that breast carcinogens act by diverse biological pathways and we designed BCScreen to capture all of them. Another important area of work is to compare gene expression responses among different breast cell and tissue models, including standard cancer cell lines used in high throughput testing and realistic normal human breast tissue models.…”