Dioxin and the AH Receptor: Synergy of Discovery

Abstract

Why the interest in dioxins, a group of structurally related chemicals which have a common mechanism of action, a common spectrum of biological responses and are environmentally and biologically persistent? A plethora of effects have been reported in people, wildlife, and domestic animals since chloracne was first described in 1899. Cattle, horses, sheep, and chickens have all been shown to be affected during poisoning episodes with polychlorinated byphenyls (PCBs). Fish, birds, and marine mammals have shown adverse outcomes, such as loss of reproduction and immune suppression, at environmental levels. And in the laboratory, species from all vertebrate classes have been used to study the biological effects from exposure to dioxins [1]. While chloracne is diagnostic of poisoning by dioxins, it is only associated with high levels of exposure. However, industrial accidents such as in Nitro, West Virginia, in 1949, Seveso, Italy in 1976, the polybrominated biphenyl (PBB) flame retardant poisoning in Michigan in 1973, and the Binghamton office building fire in 1981, all resulted in some chloracne. In addition, other human poisonings, such as that due to PCB/polychorinated dibenzofuran (PCDF) contaminated rice oil in Japan in 1968 (“Yusho”) and Taiwan in 1979 (“Yucheng”), demonstrated a wide range of toxic effects, both on those who ingested the contaminated oil and on their children born afterwards. Intentional poisoning by 2,3,7,8-tetrachloridibenzo-p-dioxin (TCDD), the most toxic polychlorinated dibenzo-p-dioxin (PCDD) congener, occurred to five people in Vienna in 1999, and to the Ukrainian President in 2004 [2].

1. INTRODUCTION

While 2,3,7,8- TCDD is the most studied and most toxic of the dioxins, there are many other dioxin-like compounds referred to as “dioxins,” including, the laterally halogenated (≥3) dibenzo-p-dioxins, dibenzofurans, and biphenyls. In addition, the laterally halogenated naphthalanes, azo- and azoxy-benzenes can all have dioxin-like activity. None of these chemicals are ever found in isolation, but usually in complex mixtures. This led to the development of the Toxic Equivalency, or weighted potency, approach for risk assessment, in which the chlorinated PCDDs/Fs/Bs are assigned toxic equivalency factors in relationship to TCDD [3]. More recently, brominated dioxins have been assigned values as well [4]. It is important to note that while the halogenated biphenyls, naphthalenes, and azo/azoxy-benzenes were used commercially, the dioxins and furans have never been intentionally synthesized, except for laboratory purposes, but are unwanted contaminants of uncontrolled combustion and certain industrial processes.

Dioxins cause a wide range of effects, ranging from molecular to organismal. Effects are context dependent but can occur in almost every tissue during some life stage and in both sexes of animals. Biochemical effects include induction of Phase I, II, and III metabolizing enzymes and transporters, genes involved in proliferation, cytokines, growth factors and receptors, and more. Clearly adverse effects range from lethality, wasting, and gonadal and lymphoid atrophy, at high exposures, to hyperplasia, metaplasia, carcinogenesis, endocrine disruption, reproductive and developmental toxicity, including functional developmental toxicity, dermal toxicity, immunotoxicity, neurotoxicity, hepatic toxicity, cardiovascular toxicity, bone and teeth toxicity, cardiovascular toxicity, and diabetes [1]. Of greatest concern are the developmental alterations associated with prenatal exposure occurring at the high end of the background population: decreased learning and memory, altered behaviors, suppressed immune system, compromised endocrine systems, and impacts on growth [5]. The good news is that regulations work, and levels of dioxins have decreased both in the environment and in people since peaking in the 1960s through the 1980s.

How do dioxins cause this multitude of effects? All the effects of TCDD and related dioxin-like chemicals involve the AH (aryl hydrocarbon) Receptor (AHR). The AHR is a highly conserved protein throughout Vertebrata, with homologs present throughout invertebrates. In fact, it is a member of a growing family of key regulatory proteins in development and aging, hypoxia, and circadian rhythms, among others. It is necessary, but not sufficient, for all of the effects of dioxin. Its existence was first postulated by Nebert and Gelboin [6] based on their studies of the induction of aryl hydrocarbon hydroxylase (AHH) in “responsive” (C57BL/6) versus “non-responsive” (DBA/2) mice. Poland and Kende [7] showed that dioxin and related compounds could overcome the “non-responsiveness” of the DBA/2 mice, although a ~10-fold higher dose was required than in the C57BL/6 mice. Poland and coworkers demonstrated that various dioxin-like compounds had varying binding affinities to the AHR, and that there was a direct relationship between the binding affinity (Kd) and enzyme induction (AHH) as well as teratogenicity, induction of cleft palate in mice. The differential binding was further explored by Steve Safe and colleagues [8] for a wide variety of halogenated ligands, leading to the development of a relative potency ranking scheme, and eventually the toxic equivalency factors (TEFs). Working in Poland’s laboratory, Chris Bradfield was the first to clone the AHR [9], and, to everyone’s surprise, found that it was not a member of the nuclear steroid receptor superfamily, but was a basic-helix-loop-helix (bHLH) protein. This followed the isolation of the protein by Gary Perdew and Alan Poland [10]. What soon became clear is that it was a member of a family of proteins containing a per-arnt-sim (PAS) domain which act as biological sensors [11].

The classic model for AHR action involves not only the AHR, but additional proteins, XAP2, p23, and two molecules of HSP90 in the cytosol. Upon binding a ligand, XAP2 and p23 dissociate from the AHR-ligand/HSP90 complex, which then translocates into the nucleus where another bHLH protein, ARNT, binds and HSP90 is released. The ligand-activated AHR/ARNT heterodimer binds to a dioxin-response element (DRE) upstream of the promoter for genes such as CYP1A1 and initiates transcription. Within ten years of its cloning, it became clear that other proteins can be recruited to the AHR-ARNT complex prior to DNA-binding, thus modulating its transcriptional activity [12]. Thyroid hormone receptor/Retinoblastoma-interacting protein 230 (TRIP230) is an example of a co-activator, and estrogen receptor alpha (ERα) can function as a co-repressor. In fact, TCDD action can be thought of similar to that of any kind of hormone and involves at least three separate steps. The first involves recognition of its signal – the ligand binds with high specificity to its receptor, which is part of a multimeric complex. This complex falls apart, and after movement into the nucleus, transduction of the signal is effected by heterodimerizing with ARNT, forming the DNA binding species. This acts as a transcriptional enhancer binding to specific sites on the DNA leading to a response.

The complexity of AHR action has continued to increase. While much of the focus has been on the genomic mode of actions of the AHR, studies by Fumio Matsumura [13] and others have demonstrated that AHR activity can also be nongenomic. Activation of src-family kinases occurs within minutes of ligand activation, too rapidly to involve all the steps in transcriptional control. Theses rapid responses occur without nuclear translocation and without ARNT. Mobilization of calcium stores also occurs rapidly in multiple cell types. The rise in intracellular calcium leads to events which are dominated by the activation of protein kinases and phosphatases. One might think of the AHR as repressing the activity of XAP2 and p23, which upon release when a ligand activates the AHR, can induce calcium mobilization and phosphorylation/dephosphorylation.

The relatively simple model of the AHR acting to activate second messenger mechanisms in the cytosol and inducing transcription in the nucleus has been modified in the past decade. If one thinks of the function of the ligand as inducing a conformational change in the AHR, it is possible to understand how a conformational change could be induced by other mechanisms. Omeprazole, sunitinib, geinistein, and daidzein do not bind to the AHR directly, but appear to activate the AHR by changing its phosphorylation state and thus its conformation, leading to loss of XAP and p23 from the cytosolic complex, and movement of the AHR/HSP90 complex into the nucleus where it can dimerize with ARNT to initiate transcription from the DRE [14]. Cytosolic activation can also occur via NFκB, Src1, and PKA. RelB (a component of NFκB) can dimerize with the activated AHR in the nucleus and induce transcription.

In addition, it is becoming clear that the “activated” AHR in the nucleus can dimerize with proteins other than ARNT, leading to recognition of other DNA sequences. We have known for over 20 years that certain TCDD-inducible genes, such as PAI-1, did not have a consensus DRE upstream of their promoter. Recent work from Cornelius “Kees” Elferink’s lab [15] has demonstrated that ligand-activated AHR can bind to Kruppel-like-factor 6 (KLF6). This heterodimer then binds CPS1, a protein that homocitrullinates histone H1 resulting in a more transcriptionally active chromatin stricture. Recent studies have shown that regulation of the immune system, specifically of Th17 and Treg cells, require the AHR, but one is dependent on the classical DRE and other on non-canonical DNA binding pathways [16].

Interaction with the estrogen receptor (ER) can have different consequences depending on the context. For example, while TCDD acts like an anti-estrogen in breast cancer, it can lead to an increase in breast tumors following developmental exposures [17]. ER can dimerize with the activated nuclear AHR leading to degradation of the AHR. The ER homodimers, bound to ligands such as 3-methylcholanthrene or PCBs, can act on the estrogen response element (ERE) and act like AHR agonists. However, in the absence of AHR ligands, the ER can bind to ERE upstream of the DRE and the AHR/ARNT complex can act to repress ER-mediated transcription. Various co-activators and co-repressors can complex not only with AHR/ARNT but with other nuclear transcription factors. For example, p300 is also needed for ER/ER activity. Thus, there can be competition for these factors, known as squelching [18]. In addition, ARNT has many partners. It is required to respond to a hypoxia challenge by binding to HIF1α. Also, while the amounts of AHR in a cell appear to be inducible, those of ARNT remain fairly constant. Thus, under high AHR conditions, there may not be enough ARNT available to respond to hypoxia. Both AHR and ARNT also undergo circadian rhythmicity, but they are not synchronized [19], again allowing for ARNT to become limiting.

One of the first indications that endogenous AHR ligands might exist was the observation that much of the AHR was nuclear during development [20]. In the past ten years, there have been numerous reports of endogenous ligands such as kynurenine, FICZ, and ICZ, among others. Their affinity for the AHR varies, and all are rapidly metabolized [21]. Most of these are derived from tryptophan, some from photolysis in the skin. In addition, thousands of short-lived natural and synthetic ligands have been found. What is extremely important is that diverse ligands can result in diverse outcomes [22]. While many ligands activate the same battery of genes, ligand-specific effects are also observed, likely reflecting the recruitment of different coactivators or corepressors due to slight conformational differences in the ligand-bound conformation of the AHR.

What has become increasingly clear is that the 3D conformation of the activated receptor can vary subtly. While the human AHR has lower Kd for TCDD than the mouse, there are other ligands which have greater affinity for the human AHR [23]. The persistence of TCDD and other dioxin-like chemicals is part of their toxicity. Recent studies have shown inhibition of the metabolism of FICZ in vivo leads to dioxin-like effects [24].

The AHR can have apparently opposite effects depending upon the context. For example, it can function both as a tumor promoter and as a tumor suppressor in prostate cancer, depending on the timing of exposure [25]. AHR signaling in the skin can cause both proliferation and differentiation, leading Haarmann-Stemmann, Esser, and Krutmann [26] to describe this protein as “Janus-Faced.” Many of the high-dose effects of dioxin are similar to what is seen in AHR-knockout mice – in part could be related to the fact that the TCDD-activated receptor is actually degraded, and thus may no longer be available for its key role in endogenous processes: development, differentiation, proliferation. Recent studies have shown the key role of the AHR in stem cell regulation, with bidirectional talk between it and Oct4 [27]. This may be why dioxins are such problems: they hijack the AHR, disrupting its endogenous function. As proposed in 2009 [1], the physiological activator of the AHR likely induces rapid on/off signaling through the receptor. Dioxins, however, are believed to induce toxicity through the persistent activation of the AHR, thereby preventing the AHR from functioning in the maintenance of homeostasis. Persistent AHR activation may also result in increased competition for ARNT, and thus ARNT’s other partners might have reduced ability to dimerize and carry out their functions. AHR plays an evolutionarily conserved role in development and an endogenous role in cell cycle control.

2. DISCUSSION

While TCDD was first described as the cause of chick edema disease in 1957, and the AHR first postulated in the late 1960s and cloned in the 1990s, it is clear that much of the work studying the AHR was initially driven by the toxicity of dioxins. The toxicology studies laid the groundwork for the understanding of the AHR as a key regulatory protein in development and differentiation, in stem cells, and in the immune system, in cancer and in heart disease, in eye and long would it have taken to discover the AHR without TCDD? While this question is not really answerable, what is clear is that science has clearly shown that dioxins are bad skin health, etc. We can ask how for living creatures. And as this understanding has increased, studies on the toxicity of dioxins are starting to decrease, while those focusing on the key endogenous roles of the AHR continue to climb.

HIGHLIGHTS.

• TCDD and related chemicals, “dioxins,” cause a plethora of effects in all vertebrates, including humans

• The Ah receptor is necessary, but not sufficient, for all the effects of dioxin

• The Ah receptor is an essential regulatory protein in development and differentiation

• The effects of the Ah receptor are context-dependent

• Studies of dioxin have led to the increasing study of the multiple roles of the ah receptor in health and disease