Vitamin A is obtained from the diet, either as preformed vitamin A (retinyl ester, retinol, and small amount of RA) or as provitamin A (carotenoids) (Fig. 1). In the lumen of the small intestine or in the intestinal mucosa, dietary retinyl esters are hydrolyzed to retinol, through the action of retinyl ester hydrolases (REHs). Provitamin A (mainly in the form of ^-carotene) absorbed by the mucosal cells is converted to retinaldehyde through the actions of carotene-15,15'-dioxygenase, and this form is further reduced to retinol by retinaldehyde reductase. Within the enterocyte, retinol, independently of its dietary origin, is reesterified by the enzyme lecithin:retinol acyl-transferase (LRAT), and retinyl esters are packaged into chylomicrons, together with other dietary lipids. Although LRAT is considered the primary enzyme for esterification of retinoids, LRAT null mice maintain some ability to convert retinol to retinyl esters, supporting the notion that another enzyme, namely acyl-CoA:retinol acyltransferase is involved in this process (O'Byrne et al., 2005).
Once packaged into chylomicrons, the bulk of dietary retinoid is taken up by the liver, while the remaining 25% is taken up by extrahepatic tissues (Goodman, 1962; Goodman et al., 1965). Upon uptake in the liver, chylo-micron retinyl esters are once again hydrolyzed to retinol, which can either be secreted by the hepatocyte bound to retinol-binding protein (RBP) or it can be transferred to the hepatic stellate cells for storage (Vogel et al., 1999). At this time the mechanism of transfer of retinol from the hepatocytes to the stellate cells for storage is still not completely elucidated. However, when dietary vitamin A is abundant, ^80-90% of the stored retinyl esters are in the stellate cells. Both hepatocyte and stellate cells produce significant amounts of REH and LRAT, as well as the cellular retinol-binding protein type I (CRBPI). CRBPI is a chaperone protein necessary to solubilize retinol in the aqueous environment of the cell (Vogel et al., 1999).
One of the earliest observations of dioxin toxicity was that exposure to these compounds altered hepatic retinyl ester storage in a variety of mammalian model systems (Hakansson et al., 1991). In rats exposed to an acute dose of TCDD, storage of retinyl ester in the stellate cells was reduced, until the TCDD was eliminated from the liver (Hakansson and Hanberg, 1989; Hakansson et al., 1994; Thunberg et al., 1979, 1980). This is postulated to be somewhat influenced by a TCDD-induced mobilization of retinoids from hepatic and extrahepatic storage, as well as increased elimination in the urine (Brouwer et al., 1989). The inhibition of retinyl ester storage following
TCDD exposure is most likely responsible for the reduction of retinoids in the rest of the animal (Hakansson et al., 1991).
Although it is clear that TCDD and related congeners alter hepatic retinyl ester storage, the mechanism of this reduction is less clear. The reduction in storage in the hepatic stellate cell population is not a result of a reduction in the number of stellate cells, or from any toxicity of TCDD on the stellate cell population (Hanberg et al., 1996). However, exposure to TCDD does appear to reduce LRAT activity in stellate cells (Nilsson et al., 1997). Further, LRAT mRNA levels are reduced in whole liver homogenates from TCDD-exposed rats (Hoegberg et al., 2003). These data indicate that TCDD may directly alter the expression of LRAT in hepatic stellate cells, and thereby reduce LRAT-induced retinyl ester formation for storage. Interestingly, TCDD treatment results in an increase in retinyl esters in the kidney, and this is preceded by an increase in the expression of LRAT (Hoegberg et al., 2003).
To maintain solubility in an aqueous environment, retinoids are bound to retinoid-specific binding proteins. The cellular retinol-binding proteins (CRBPs) and the cellular retinoic acid-binding proteins (CRABPs) are entirely intracellular; whereas the RBP and the intracellular retinol-binding proteins (IRBP) are extracellular. CRBPI is thought to have a critical role in regulation of retinoid storage by regulating ROH esterification by LRAT in the stellate cells (Ghyselinck et al., 1999; Nilsson et al., 1997). Further, several microsomal enzymes that are involved in retinoid metabolism prefer ROH bound to CRBPI, including RDHs I, II, and III. Also, SCADs prefer retinol that is bound to CRBPI (Napoli, 1999). There are also RALDH that are more effective toward retinal in complex with CRBPI. It has been suggested that retinol bound to CRBPI is protected from metabolism from liver enzymes such as ADH and the CYP450s (Napoli, 1999). CRBPI also acts as an atRA chaperone which may result in the metabolism of the low levels of free atROH, thus preventing excessive ROH oxidation but allowing a small amount to be converted to RAL for atRA synthesis (Duester, 2000). CRBPI is also critical to delivery of retinol to newly synthesized RBP for secretion from the liver into circulation, and it is implicated in facilitating uptake of retinol-RBP complexes by the extrahepatic cells. Therefore, CRBPI is an essential intracellular transporter of retinol and forms a link between retinol mobilization, metabolism, and uptake.
Although it is compelling to postulate that TCDD and the AhR pathway may alter the expression of the CRBPs, no change in CRBPI expression in the liver of TCDD-treated rats is observed (Schmidt et al., 2003). However, data from knockout animals suggest that the CRBP proteins somehow modulate the effect of TCDD on RA storage. TCDD exposure of mice that lack CRBPI, CRABPI, and CRABPII results in complete depletion of total retinoids in the liver. However, exposure of mice that are null for only CRABPI and CRABPII maintained 60-70% of the total hepatic retinoids
(Hoegberg et al., 2005). This suggests that loss of CRBPI may account for the increased susceptibility of the triple knockout mice to TCDD-induced retinoid depletion.
As discussed above, maintenance of whole-body retinoid metabolism is a complex process involving several organ systems, with the main storage of retinoids in the liver. This system of regulation provides the body with optimal amounts of retinoids despite changes in retinoid dietary intake. It has been proposed that there exists a feedback mechanism involving the hepatic and renal retinoid pools in conjunction with the circulating retinoids and that this feedback mechanism is regulated by a yet identified set point, which may be a critical level of a specific retinoid in the liver or kidney. This mechanism would allow for maintaining retinoid homeostasis even in times of vitamin A deficiency or excess. It has been proposed that TCDD, by the significant alterations in liver retinoid levels, may alter the set point for the feedback system, thereby resulting in a cascade of mis-regulation of retinoid acid synthesis, metabolism, and storage (Nilsson and Hakansson, 2002). Although it is unclear how TCDD and the AhR pathway may alter the set point, there are several candidates suggested for this mechanism. One is the apo:holo ratio of the CRBPI, which may be involved in maintaining a balance between retinoid hydrolysis, esterification, as well as conversion of retinol to RA (Boerman and Napoli, 1991, 1996; Herr and Ong, 1992). The importance of the binding proteins in the retinoid homeostasis is supported by the finding that CRBPI null animals do not store retinyl esters properly (Ghyselinck et al., 1999). Further, RBP knockouts fail to efficiently mobilize stored hepatic retinoids (Quadro et al., 1999).
V. MOLECULAR INTERACTIONS BETWEEN THE RA AND AhR PATHWAYS
Cross talk between the AhR and RA pathway extends beyond effects on retinoid metabolism, also affecting transcriptional regulation. Both the AhR and RA pathways regulate transcription of a variety of genes that are critical for the physiological effects mediated by these pathways. Like the numerous interactions observed for these pathways in retinoid metabolism, there are also several levels of molecular interactions between these pathways, including direct inhibition, alteration of receptor availability, and competition for transcriptional coactivators.
One of the first indications that the RA and AhR pathways interact at the level of gene expression was that TCDD exposure of SCC-4 keratinocytes inhibits atRA-induced activation of transglutaminase, an enzyme critical for proper differentiation of skin. The role of the AhR pathway in mediating this inhibition is indicated by two AhR-activating compounds, methyl-cholanthrene or benzo[a]pyrene, preventing transglutaminase activation
(Rubin and Rice, 1988). TCDD inhibition of transglutaminase in the SCC-4 cells is mediated primarily at the level of transcription, and does not result from a change in mRNA stability (Krig and Rice, 2000). Interestingly, the data also indicate that TCDD does not alter binding to and activation of the RARE, as there was no effect of TCDD on an RARE-luciferase construct transfected into these cells (Krig and Rice, 2000). Therefore, the mechanism of TCDD-induced interference of atRA-induced transglutaminase expression is still unknown. TCDD also demonstrates an inhibitory action toward other atRA target genes, including RDH9 (Tijet et al., 2006), and CRABPII (Weston et al., 1995). TCDD activates expression of retinal oxidase, the enzyme that catalyzes the conversion of retinal to RA. However, cotreatment with atRA and TCDD results in the downregulation of retinal oxidase expression and activity (Yang et al., 2005). Although the majority of data indicate that TCDD/AhR inhibit RA-mediated gene expression, there is growing evidence indicating that the interaction is more complex and may be tissue and cell-type specific.
Data from AhR knockout mice support the hypothesis that the AhR pathway interferes with expression of RA pathway target genes. For example, the expression of CRBPI is higher in the livers of AhR null animals than in their wild-type counterparts (Andreola et al., 1997). Interestingly, TCDD exposure does not appear to alter CRBPI expression in mice (Hoegberg et al., 2005). In addition, atRA levels are elevated in the livers of AhR null mice in comparison to wild-type mice (Andreola et al., 1997), which is coupled to a downregulation in CYP2C39 mRNA expression in the AhR null animals (Andreola et al., 2004). These data suggest that the AhR pathway, in the absence of exogenous ligand, is inhibitory toward the basal expression of genes that encode for proteins critical for retinoid homeostasis.
Conversely, the RA pathway also has an inhibitory effect on AhRmediated transcription, and one of the most extensively studied is the effect of atRA on expression of CYP1A1. Because of the presence of an RARE in the human CYP1A1 promoter, it was originally postulated that atRA would enhance CYP1A1 expression. This was supported by findings demonstrating that the CYP1A1 RARE is able to bind nuclear proteins as well as mediate atRA-induced expression of a CYP1A1 reporter construct (Vecchini et al., 1994). However, studies in the expression of the endogenous CYP1A1 gene did not support this conclusion: neither mouse embryos nor Hepa-1c1c7 cells exposed to atRA show induction of endogenous CYP1A1 (Soprano et al., 2001). It is now accepted that atRA exposure is inhibitory to xenobiotic-induced CYP1A1 expression and activity, and that this inhibition is mediated through the RARE in the promoter (Wanner et al., 1996). In support of this conclusion, RARa null animals display an increase in hepatic CYP1A1 activity after TCDD treatment compared to wild-type mice, suggesting that RARa may play an inhibitory role in TCDD-mediated CYP1A1 gene regulation (Hoegberg et al., 2005).
In addition to the effects of these pathways on target genes containing either the XRE or RARE elements, other target genes for both pathways have been identified that may be important for mediating some of the lesions observed following exposure to TCDD and atRA. Examples of potential target genes are the enzymes that mediate matrix metabolism, including the matrix metalloproteinases (MMPs). MMPs are a family of endopeptidases that mediate the cleavage of proteins involved in tissue structure, such as type I and type IV collagen. Further, these enzymes are also involved in the regulation of proliferation and angiogenesis through the cleavage and release of growth factors and receptors from the extracellular matrix and the cell surface (reviewed in Brinckerhoff and Matrisian, 2002). It is long established that atRA exposure alters the expression of MMPs in a variety of cell types, primarily downregulating expression through interference with the AP-1-signaling pathway (Vincenti et al., 1996). Data indicate that TCDD also modulates the expression and activity of the MMPs (Murphy et al., 2004; Villano et al., 2006). Interestingly, cotreatment with TCDD and atRA in normal human keratinocytes results in an enhancement of MMP-1 expression over exposure to TCDD alone. The coactivation of atRA and TCDD was also observed for PAI-2, a regulator of matrix remodeling, indicating that atRA/TCDD coactivation is not limited to MMPs. The induction of MMP-1 by cotreatment with atRA and TCDD does not rely on transcriptional interaction between the RARs and AhR, but instead is mediated through two distinct mechanisms: TCDD-induced transcription and atRA enhancement of MMP-1 mRNA stability (Murphy et al., 2004).
Although it is clear that there are interactions between these pathways at the level of transcriptional activation, it is unclear how these interactions are accomplished. Although atRA inhibition of CYP1A1 is most likely mediated by steric interference between proteins binding to the XRE and RARE, not all genes that are coregulated have both XREs and RAREs. A potential mechanism underlying changes in target gene expression by AhR and RA pathway interaction may be through changes in receptor availability. AhR availability in the cell is mediated by transcriptional and posttransla-tional mechanisms. Further, the targeted degradation of the AhR protein is also considered an important mechanism in regulating this pathway (Pollenz, 2002). Studies using the murine AhR promoter demonstrate that treatment of a murine epidermal cell line with atRA results in reduced AhR promoter activity (FitzGerald et al., 1996). Data from human keratinocytes did not demonstrate any change in either AhR or Arnt mRNA expression following atRA exposure (Murphy et al., 2004). This difference may be a consequence of species-specific differences between the human and murine AhR promoter activity or from a difference in atRA responsiveness of the human versus the murine cell line tested. In support of this idea is the fact that the murine cell line (JB6-C1 41-5a) used in the reported studies is highly responsive to RA (FitzGerald et al., 1996).
TCDD and the AhR pathway are known to alter the expression of the RARs and RXRs, although the effect of TCDD on RAR and RXR gene expression is receptor and cell-type dependent. RARb expression is inhibited by TCDD exposure of embryonic palate mesenchymal cells (Weston et al., 1995). However, in SCC12Y cells TCDD treatment results in a decreased binding of atRA to RARa without any change in RARa gene expression (Lorick et al., 1998). In normal human keratinocytes, TCDD treatment results in an increase in RARg and RXRa mRNA levels (Murphy et al., 2004). Therefore, one way in which TCDD may alter atRA target gene expression in some cell types may be through alterations of the receptor availability.
Activation of gene expression by the AhR/Arnt- or RA-signaling pathways requires the recruitment of coactivators and general transcription factors to the promoter region (Hankinson, 2005; Wei, 2003). The recruitment of coactivators to target genes can either enable chromatin remodeling or aid in recruiting basal transcription machinery to the promoter of the target gene (Chen, 2000). Coactivators are classified based on the mechanism used to induce transcription. HAT coactivators transfer an acetyl group onto specific lysine residues of histone tails destabilizing chromatin structure while histone methyltransferases (HMTs) modify arginines (to enhance) or lysines (to repress) of histone tails with methyl groups. Phosphorylation of histone tail serine residues as well as ubiquitination also serves as potential signals for altering chromatin structure. These signals serve to facilitate access to the DNA by destabilizing local chromatin structure through his-tone modification as well as being markers to recruit other coactivator proteins to the modified sites. Corepressors function by recruiting a complex of silencing proteins to the promoter region, including histone deacetylases (HDACs) (Baniahmad, 2005). A list of coactivators known to interact with the RA and AhR pathways are shown in Table II.
The importance of nucleosomal structure on TCDD-mediated CYP1A1 transcription was demonstrated in a study by Morgan and Whitlock (1992) identifying a nucleosome structure associated with CYP1A1 mouse promoter/enhancer region that is altered following TCDD treatment. There is also a concomitant increase in protection of this area located at —40 and —60 bp of the promoter which contains the TATAAA box and an NF-1-like recognition motif. This indicates that activated AhR alters nucleosome structure to facilitate transcriptional activation and suggests that coactivator or corepressors may be involved in TCDD-mediated CYP1A1 expression. Indeed, coactivator estrogen receptor associating protein 140 (ERAP 140) and the corepressor SMRT both physically interact with AhR/Arnt transcription factor complex and are able to increase AhR/Arnt binding to an XRE (Nguyen et al., 1999). However, the exact nature of the involvement of SMRT in AhR-mediated transcription is not yet elucidated. One study indicates that SMRT acts to inhibit AhR-mediated transcription (Nguyen et al., 1999);
while data from another study demonstrates that overexpression of SMRT activates TCDD-mediated transcription in some cell types and inhibits it in others (Rushing and Denison, 2002). These data indicate that SMRT-AhR interaction has a role in mediating AhR transcriptional activation, and suggest that the interaction of AhR and SMRT is dependent on cell-type-specific signals or factors.
Recent evidence indicates that corepressors may be a link between the AhR and RA expression pathways. The SMRT corepressor is known to interact with both the AhR and RARs and modulate their transactivating function (Nguyen et al., 1999; Rushing and Denison, 2002; Widerak et al., 2005). Further, SMRT may also be involved in TCDD-mediated effects on RAR binding and transactivation through the RARE. It has been known for some time that in some cell types, TCDD is able to activate expression of RARE-driven reporter constructs (Vecchini et al., 1994; Widerak et al., 2005). However, the mechanism of this activation was unknown. Data show that TCDD activation of the RARE-CAT construct is inhibited by cotransfection with an expression vector containing the SMRT corepressor (Widerak et al., 2005). Taken together these data suggest that the involvement of the corepressor as well as the coactivator proteins may provide a molecular pathway for the transcriptional cross talk between the AhR and RA pathways.
The data presented in this chapter demonstrate both direct and indirect interactions between the AhR- and RA-signaling pathways. These interactions include changes in the availability of atRA in the liver and extrahepatic tissues by AhR-mediated regulation of atRA synthesis and metabolism, as well as on storage and transport. Further, these two pathways directly impact each others signaling pathways through alterations in receptor availability and modulation of transcriptional regulation. Although it appears that the intersection of these two pathways may be mediated by specific coactivator and corepressor proteins, the exact mechanism is yet undefined. However, it is clear that a portion of toxicity related to TCDD and related congeners is mediated through their effect on RA homeostasis and on the atRA-signaling pathway.
Was this article helpful?