Effects of food contaminant mixtures on putative targets of endocrine disrupters exposure : an integrated toxicogenomic-biomarkers approach

Abstract

The Endocrine Disruptors (EDs) are a class of chemicals that may interfere with the endocrine system that plays a crucial role in maintaining the physiological homeostasis of the human body as well as in regulating body growth, metabolism, reproduction and behavior (Faroon et al., 2001; Schell and Gallo, 2010). The EDs can exert their effects through a number of different mechanisms in particular interacting with nuclear (NRs) and orphan receptors, enzymatic pathways involved in steroid biosynthesis and detoxification metabolism (Diamanti-Kandarakis, 2009). Polychlorinated biphenyls (PCBs) form a group of fat soluble and environmentally persistent EDs able to bioaccumulate in lipid fraction of animal tissue leading to biomagnification in the food chain. PCBs intake via food of animal origin represents the main route of human exposure (EFSA, 2010). PCBs can be divided into two groups according to their biochemical and toxicological properties: Dioxin-like (DL) PCBs, mainly interacting with the Aryl Hydrocarbon Receptor (AhR) and Non Dioxin-like (NDL) PCBs whose mechanisms of action are not fully clarified. Actually, organisms are exposed to several congeners at the same time and, accordingly, the observed toxicological effects, such as on liver, thyroid, immune function, reproduction and behavior as well as carcinogenicity (EFSA, 2005), are exerted by mixtures. Therefore, new researches are required to study EDs mixtures, in particular on possible additive/synergistic effects at “real-life” dose levels. In this study, 21 environmentally relevant PCB congeners were divided in three groups according to the proposed classification of Wolff et al. (1997) and Negri et al. (2003), based on structural and modes of action similarities: Mix1 (PCB 44, 49, 52, 101, 174, 177, 187, 201), potentially estrogenic NDL congeners, Mix2 (PCB 77, 81, 105, 114, 118, 126, 169) featuring DLPCB and Mix3 (PCB 99, 153, 180, 183, 196, 203), NDL-PCBs highly persistent in the organism and phenobarbital-like inducers of Cytochrome P450 (CYP450). The aim of the present project was to evaluate potential effects on liver and on adipose tissue as target organs of PCBs, elicited by the mixtures, at realistic human exposure levels, as derived from human exposure data through the evaluation of biomarkers response. Two in vitro models have been selected as representative of liver tissue and corresponding to two different life stages: infant (HuH6) and pubertal (HepG2) cells. To evaluate synergic effect, biomarkers of effect were used as indicators to measure PCB-induced variation in cellular or biochemical processes (NRs, NRF2, CYP450 enzymes and markers of oxidative stress). Moreover gene expression profiles of PCBs treated HepG2 cell line were evaluated by microarray analysis. Moreover, PCBs bioaccumulation and permanence in animal fat could affect preadipocyte programming and differentiation and adipocytes metabolism. Therefore, differentiation of mouse (3T3-L1) and human preadipocytes (SGBS cells) was evaluated to determine the effect of PCBs exposure. Moreover gene expression of biomarkers involved in adipocytes differentiation and metabolism were analyzed in 3T3-L1. Previous reports classified Mix1 as potentially antiandrogenic congeners (Wolff et al., 1995), on the contrary, we observed in HepG2 cell line, that Mix1 significantly up-regulated only AR suggesting a potential androgenic effect. In HepG2 cells, we also observed an uncoupling in ROS-CYP1A1 regulation and in HuH6, GSH/GSSG ratio was significantly higher. As for Gene Ontology enrichment analysis, Mix1-modulated gene list was uniquely significant for terms related to Nuclear Transport and significantly affected only one pathway, the Intestinal Immune Network for IgA production, implying a stimulation of the immune response. Overall, these molecular processes may be considered specific markers of Mix1 effect in hepatic cells. As regards effects on adipocytes, Mix1 increased SGBS differentiation at 3x concentration and in 3T3-L1 affected only C/EBP involved in adipose tissue development. Mix2, featuring DL-PCB congeners, exerted the up-regulation of AR and AhR in HepG2 cells. Moreover, Mix2 increased ROS levels as well as CYP1A1 mRNA expression both mediated by AhR (Kopf et al., 2010). However Mix2 significantly decreased the CYP1A1 enzyme activity, indicating a different effect possibly due to the low concentrations used. In HuH6, Mix2 induced a down regulation of CAR mRNA expression that may explain the reduction of CYP1A1 and CYP3A4 activity. Therefore, CAR gene expression and CYP1A1 and CYP3A4 activities may be assumed as target of DL-PCBs disruption on this hepatic cell line. Microarray results evidenced Mix2 affected genes involved in RNA processing and splicing, protein localization and catabolism and macromolecules degradation. Noteworthy, main and characteristic effects of Mix2 relied a) on enrichment of genes related several cancer pathways as p53, activator of genes involved also in cell cycle arrest (Lanni and Jacks, 1998), apoptosis and in communication in adjacent cells, and b) on cell cycle progression, inducing of G1 phase and inhibiting S phase. So we may hypothesize that a sum of contrasting mode of action exerted by PCB congeners in Mix2 occurred. As regards effects on adipocytes, Mix2 significantly induced SGBS differentiation only at 3x concentration, probably due to the mixture effect and/or the lower concentration levels used in this study. Mix3 treatment affected the wider number of NRs expression (ER , ER , AR, PXR, RAR , THR ) as well as NRF2, in particular, in HuH6. To our knowledge, it is the first study that showed AR mRNA expression due to PCB congeners included in Mix3, whereas the observed ER and ER induction confirmed the estrogenic effect of PCB 99 and 153 (Warner et al, 2012).The decrease in the CYP3A4 enzyme activity in HuH6 cells was in agreement with NRF2 gene expression down-regulation (Itoh et al, 1997). Microarray analysis showed that Mix3 modulated the higher number of genes (5979 genes) with 1501 shared with Mix2-affected genes. Apart some GO enriched terms related to cell organization, Mix3 did not share other effects with Mix1, the other NDL mixture. One of the most interesting evidence is the enrichment of genes involved in the Wnt signaling pathway by only Mix3, suggesting an unbalance toward pluripotency promotion (Takemarua and Moona, 2000). Mix3 also affected some cancer pathways as well as the adherens junction pathway as Mix2 but modulating a different panel of involved genes. Moreover Mix3 exerted the higher magnitude of effect on pre-adipocytes differentiation in both murine and human cell lines. Gene expression evaluation analysis outlined that Mix3 down regulated Hes1, a DNA binding protein whose expression blocks adipogenesis (Ross et al., 2006). Overall results confirmed that the adopted grouping of PCBs in three mixtures served to highlight different modes of action as well as biomarkers responses, also among mixtures featuring NDL congeners. Moreover, the two hepatic cell lines and the two pre-adipocytes cell lines appear to be differently reactive to PCBs, indicating the need to use different in vitro models and a panel of biomarkers in order to characterize the EDs effects. These represented new results since there are no studies on PCB mixtures summarizing effects on oxidative stress, metabolism, nuclear receptors gene expression and adipocytes differentiation responses following treatment at human real exposure concentrations. As prompted by EFSA (2005), it is necessary to provide evidences concerning comprehensive toxicological end-points of NDL effects. Therefore, these data may serve as a basis for developing relative toxicological factors for the NDL congeners risk assessment.