Citation:

Rooney, J., B. Chorley, S. Hiemstra, S. Wink, X. Wang, D. Bell, B. van de Water, AND Jon Corton. Mining a human transcriptome database for chemical modulators of Nrf2 (PLOS ONE). PLOS ONE . Public Library of Science, San Francisco, CA, 15(9):e0239367, (2020). https://doi.org/10.1371/journal.pone.0239367

Impact/Purpose:

Gene expression profiling represents a robust complementary approach to HTS screening and has the potential to cover some of the biological space missed by HTS assays. The Library of Integrated Network-Based Cellular Signatures (LINCS) and the Connectivity Map (CMAP) projects have both made significant contributions to the field of large-scale gene expression profiling, using network-based approaches to link chemical exposure to gene expression and disease. The CMAP project screened ~1300 small molecules in 3 cell lines with whole transcriptome expression data [7, 8]. One of the major hurdles to these early gene expression profiling efforts was the inherently low throughput of microarray technologies. However, the field has seen great advancements in throughput in recent years. The aforementioned LINCS database uses the L1000 gene expression technology that measures the expression of 1000 genes, and through computational inference can assess transcriptional changes in nearly 80% of the genome. The LINCS project has generated over 1 million profiles using the L1000 technology from a large battery of human cell lines perturbed with chemicals and gene probes [9]. Furthermore, new RNA sequencing (RNA-seq) technologies, such as the TempO-Seq platform show great promise in their ability to measure expression changes in both smaller, targeted gene sets, and the whole transcriptome in a high-throughput manner [10]. Lastly, there is a wealth of microarray- and RNA-Seq-derived gene expression data currently available in multiple public repositories that can be used to develop procedures for predicting the molecular targets of chemicals. Nuclear factor erythroid-2 related factor 2 (Nrf2) is a key transcription factor important in cellular responses to oxidative stress and xenobiotics. Under normal conditions, Nrf2 is bound in the cytoplasm by Kelch-like ECH-associated protein 1 (Keap1), which results in ubiquitination and targeting of Nrf2 for proteasomal degradation. When activated, Nrf2 dissociates from Keap1, translocates to the nucleus and binds to genomic antioxidant response elements (AREs) as a heterodimer with Maf proteins, MafF, MafG, or MafK [26]. Nrf2 binding to AREs promotes the transcription of a diverse battery of genes involved in the antioxidant response and detoxification, including many genes related to the cytoprotective processes of phase 2 metabolism [27]. Nrf2 is intricately linked with carcinogenesis, as Nrf2-nullizygous mice are more susceptible to many chemical carcinogens, yet paradoxically Nrf2 and its target genes are also upregulated in many cancers [28, 29]. The activation status of Nrf2 is also linked to hepatocyte steatosis [30], a condition in which there is an accumulation of triglycerides. The ability to readily identify chemicals that modulate Nrf2 in microarray studies could help to build predictive models for cancer or steatosis. In the present study, we developed computational methods for constructing and testing a gene expression biomarker that can predict Nrf2 activation or suppression in human cells. We used this biomarker, coupled with an annotated database of gene expression profiling experiments, to perform an in silico screen for chemical perturbations that lead to Nrf2 modulation. We validate our findings using an ARE-linked reporter system in HepG2 cells.

Description:

Nuclear factor erythroid-2 related factor 2 (Nrf2) is a transcription factor critical for protecting cells from chemically-induced oxidative stress. We developed computational procedures to identify chemical modulators of Nrf2 in a large database of human microarray data. A gene expression biomarker was built from statistically-filtered gene lists derived from microarray experiments in primary human hepatocytes and cancer cell lines exposed to Nrf2-activating chemicals (oltipraz, sulforaphane, CDDO-Im) or in which the Nrf2 suppressor Keap1 was knocked down by siRNA. Directionally consistent biomarker genes were further filtered for those dependent on Nrf2 using a microarray dataset from cells after Nrf2 siRNA knockdown. The resulting 143-gene biomarker was evaluated as a predictive tool using the correlation-based Running Fisher algorithm. By comparing the biomarker to 59 gene expression data sets from cells treated with known Nrf2-activating chemicals, the biomarker gave a balanced accuracy of 93%. The biomarker was comprised of many well-known Nrf2 target genes (AKR1B10, AKR1C1, NQO1, TXNRD1, SRXN1, GCLC, GCLM), 69% of which were found to be bound directly by Nrf2 using ChIP-Seq. Nrf2 activity was assessed across ~9840 microarray comparisons from ~1460 studies examining the effects of ~2260 chemicals in human cell lines. A total of 260 and 43 chemicals were found to activate or suppress Nrf2, respectively, most of which have not been previously reported to modulate Nrf2 activity. Using a Nrf2-responsive reporter gene in HepG2 cells, we confirmed the activity of chemicals predicted using the biomarker. Our biomarker-based approach accurately identified Nrf2 activators. The biomarker will be useful for future gene expression screening studies.

URLs/Downloads:

Record Details: