Gene expression profiling in Caco-2 human colon cells exposed to TCDD, benzo[a]pyrene, and natural Ah receptor agonists from cruciferous vegetables and citrus fruits

Abstract

Cruciferous vegetables and citrus fruits are reported to possess health-beneficial properties, but also have been shown to contain natural aryl hydrocarbon receptor (AhR) agonists (NAhRAs). Binding to the AhR is widely assumed to activate the main pathway by which dioxins, like 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exert their toxicity. To establish whether or not activation of the AhR pathway by NAhRAs and dioxin-like substances results in similar cellular responses, gene expression profiles induced in Caco-2 cells were studied using microarray analysis. Cells were exposed to indolo[3,2-b]carbazole (ICZ), an acid reaction product from cruciferous vegetables, and to extracts of citrus pulp and grapefruit juice. Gene expression profiles induced by these NAhRAs were compared to those of the xenobiotic AhR agonists TCDD and benzo[a]pyrene (B[a]P). Over 20 genes were found more than 1.5 times up- or down-regulated by TCDD, and the expression of most of these genes was modulated in the same direction and to a similar extent by B[a]P and the NAhRAs. Results were confirmed by RT-PCR, and many of these genes may be involved in dioxin-related toxic effects. In conclusion, this in vitro study showed similar effects induced by NAhRAs, TCDD and B[a]P at the transcriptome level in a human intestinal cell line.

Introduction

The aryl hydrocarbon receptor (AhR) is a transcription factor which is activated by ligand binding, and is expressed in most human tissue types. 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) is the prototype AhR ligand and the most potent xenobiotic AhR agonist known. Upon ligand binding, the AhR-ligand complex is translocated into the nucleus of the cell and functions as a transcription factor binding to certain enhancer elements, called dioxin-responsive elements (DREs). This results in altered expression levels of AhR-responsive genes, among them the genes encoding for the biotransformation enzymes CYP1A1, CYP1A2, CYP1B1, UDPGT1A6, and GSTP1 (Hankinson, 1995, Denison and Nagy, 2003). Most if not all toxic effects of TCDD and other dioxin-like compounds are thought to be mediated by activation of AhR-regulated gene expression, e.g. chloracne, tumor promotion, immunosuppression and teratogenesis (Fernandez-Salguero et al., 1996, Mimura and Fujii-Kuriyama, 2003, Steenland et al., 2004).

Besides environmental contaminants like certain dioxins, polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocarbons (PAHs), also AhR agonists from plant and other biotic origins are known (Denison and Nagy, 2003). Cruciferous vegetables like Brussels sprouts, cabbage and broccoli contain e.g. AhR agonist precursors in the form of glucosinolates, which are converted via indole-3-carbinol (I3C) into I3C condensation products during food preparation, consumption and subsequent incubation in the acid environment of the stomach (Bjeldanes et al., 1991, Broadbent and Broadbent, 1998). Examples of these compounds are diindolylmethane (DIM) and indolo[3,2-b]carbazole (ICZ), of which the latter is a very potent AhR agonist (Kleman et al., 1994, Bergman et al., 2001). Also many other food items appear to contain high levels of AhR-activity, as judged by the DR CALUX^® bioassay (Jeuken et al., 2003). Recently, it was reported that hexane extracts of citrus products show strong AhR agonist activity (Hoogenboom, 2002). It is known that citrus fruits contain furocoumarins, like bergamottin, and polymethoxyflavones, like tangeretin, which were reported to be AhR agonists (Canivenc-Lavier et al., 1996, Baumgart et al., 2005).

Considering that both NAhRAs and persistent xenobiotic AhR agonists trigger the same gene regulatory pathway, it may be expected that the gene expression patterns induced by both groups of AhR agonists will be similar to a certain extent. Several studies describe AhR-mediated adverse effects of NAhRAs, especially for the I3C metabolite ICZ, such as inhibition of gap-junctional intercellular communication (GJIC), which is considered to play a role in tumor promotion (Herrmann et al., 2002), increased COX-2 expression leading to enhanced prostaglandin E2 production and colon cancer progression (Sherratt et al., 2003), and increased oxidative DNA damage (Bjeldanes et al., 1991, Chen et al., 1995, Park et al., 1996, Pohjanvirta et al., 2002). However, because NAhRAs were found to induce beneficial gene expression effects, such as phase II biotransformation and antioxidant enzyme induction (Chen and Yang, 2002), inhibition of phase I-mediated procarcinogen activation (Wen et al., 2002, Sarkar and Li, 2004), and antiestrogenic effects in human breast cancer cells (Liu et al., 1994), they are actually considered to be natural cancer-chemopreventive agents. Furthermore, in an in vivo study in TCDD-sensitive rats relatively high doses of ICZ did not produce any of the acute toxicity effects of TCDD, although moderate CYP1A1 induction in the liver was observed (Pohjanvirta et al., 2002).

It is generally assumed that the apparent discrepancy between the toxicity of TCDD and compounds like ICZ is caused by their great difference in persistence in biological systems (Chen et al., 1995, Pohjanvirta et al., 2002). Pohjanvirta et al. estimated that the elimination half-life of ICZ in rats is less than a day, possibly in the order of a few hours, whereas the half-life of TCDD in rats is estimated at 21 days (Pohjanvirta et al., 1990), and in humans even 6–7 years (Maruyama and Aoki, 2006). It cannot be excluded that the large difference in biostability between NAhRAs and persistent xenobiotic AhR agonists may result in differences between the gene sets modulated by both groups, and that selective AhR modulators as described by Safe et al. (1999) do exist within the large group of NAhRAs. None of these possible differences have been comprehensively studied thus far. This issue is even more relevant regarding the increasing popularity of dietary supplements with large amounts of pure I3C or DIM, especially if consumed on a regular basis (Zeligs, 2002). As a first step on examining this issue, the objective of the present study was to determine the gene expression profiles in the human intestinal epithelial cell-line Caco-2 after exposure to ICZ or extracts from citrus fruit products, and compare these with the profiles obtained after exposure to well known xenobiotic AhR agonists. These included the persistent dioxin TCDD as well as the readily biodegradable compound benzo[a]pyrene (B[a]P). The Caco-2 cell line was chosen since it is generally used as a model system to study the effect and absorption of compounds in the small intestine. By using microarrays, the expression of thousands of genes can be examined simultaneously. The more the transcriptome profiles look alike, the more likely it is that the compounds under investigation may exert similar biologic, pharmacologic or toxic effects (Aardema and MacGregor, 2002). This may also lead to the identification of a set of gene expression biomarkers for TCDD- or NAhRA-like effects. This study will focus on effects on the transcriptome level, but follow-up studies with e.g. human volunteers will need to take into account other possibly important factors, such as proteomics, absorption and metabolism.