|Home | About | Journals | Submit | Contact Us | Français|
Ligand-dependent activation of the aryl hydrocarbon receptor (AhR) pathway leads to a diverse array of biological and toxicological effects. The best-studied ligands for the AhR include polycyclic and halogenated aromatic hydrocarbons, the most potent of which is 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). However, as new AhR ligands are identified and characterized, their structural and physiochemical diversity continues to expand. Our identification of AhR agonists in crude extracts from diverse materials raises questions as to the magnitude and extent of human exposure to AhR ligands through normal daily activities. We have found that solvent extracts of newspapers from countries around the world stimulate the AhR signaling pathway. AhR agonist activity was observed for dimethyl sulfoxide (DMSO), ethanol, and water extracts of printed newspaper, unprinted virgin paper, and black printing ink, where activation of luciferase reporter gene expression was transient, suggesting that the AhR active chemical(s) was metabolically labile. DMSO and ethanol extracts also stimulated AhR transformation and DNA binding, and also competed with [3H]TCDD for binding to the AhR. In addition, DMSO extracts of printed newspaper induced cytochrome P450 1A associated 7-ethoxyresorufin-O-deethylase activity in zebrafish embryos in vivo. Although the responsible bioactive chemical(s) remain to be identified, our results demonstrate that newspapers and printing ink contain relatively potent metabolically labile agonists of the AhR. Given the large amount of recycling and reprocessing of newspapers throughout the world, release of these easily extractable AhR agonists into the environment should be examined and their potential effects on aquatic organisms assessed.
The aryl hydrocarbon receptor (AhR) is a ligand-dependent, basic helix-loop-helix, Per-Arnt-Sim-containing transcription factor that mediates a diverse array of biological and toxicological effects in a variety of species. Mechanistically, the unliganded AhR exists in the cytosol as an inactive multiprotein complex consisting of the AhR, two molecules of the chaperone protein Hsp90 (Perdew, 1988), the X-associated protein (Meyer et al., 1998), and the cochaperone p23 (Kazlauskas et al., 1999). Ligand binding stimulates nuclear translocation of the AhR protein complex (Ikuta et al., 1998; Pollenz et al., 1994), wherein the AhR is released from the protein complex and dimerizes with the AhR nuclear translocator (ARNT) protein (Reyes et al., 1992). The resulting ligand:AhR:ARNT complex binds to its specific DNA recognition sequence, the dioxin response element (DRE) (Denison et al., 1988), leading to transcriptional activation of downstream genes. Although the best-studied AhR-responsive genes are enzymes involved in drug and chemical metabolism, microarray studies have identified a large number of AhR-responsive gene products (Hanlon et al., 2005; Vezina et al., 2004). It has been proposed that the toxic and biological effects of some AhR ligands result from their ability to persistently activate AhR-dependent gene expression, although the responsible genes and exact mechanistic events remain to be identified.
Halogenated aromatic hydrocarbons (HAHs), such as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD, dioxin) and related polychlorinated dibenzo-p-dioxins, dibenzofurans and biphenyls, represent the best characterized and highest affinity ligands for the AhR (Bandiera et al., 1983, 1984; Mason et al., 1986; Safe, 1990). The resistance of these HAH ligands to metabolism results in their ability to activate AhR-dependent gene expression persistently and to produce AhR-dependent biological and toxicological effects (Safe, 1990). In contrast, polycyclic aromatic hydrocarbons (PAHs) such as 3-methylcholanthrene, β-naphthoflavone (βNF), indolo[3,2-b]carbazole, and others are also relatively high affinity AhR ligands (Burbach et al., 1992; Wei et al., 1998), but their metabolic lability results in transient activation of AhR-dependent gene expression but not toxicity. In contrast to the many high affinity, planar, and hydrophobic HAH and PAH ligands, which have common physiochemical characteristics, recent studies have revealed that the AhR can bind to and be activated by chemicals with dramatically different structural and physiochemical characteristics (reviewed in Denison and Nagy, 2003; Denison et al., 1999). The ability of the AhR to stimulate expression of numerous gene products responsible for the metabolism and degradation of structurally diverse chemicals, including most AhR ligands, supports one of its proposed roles as an internal environmental sensor for such compounds. Although activation of the AhR signal transduction pathway clearly plays a role in enhancing the metabolism and elimination of numerous chemicals, the structural diversity of chemicals, which can bind to and activate the AhR, is only now being elucidated. The application of AhR-based in vitro and cell screening bioassays has resulted in the identification of numerous novel and structurally diverse AhR agonists and antagonists. In addition, the presence of AhR ligands (agonists and antagonists) in crude polar and nonpolar extracts of foods and natural products has been reported (Amakura et al., 2003; Jeuken et al., 2003; Seidel et al., 2000), supporting the widespread distribution of naturally occurring AhR ligands. In preliminary studies, we demonstrated the presence of relatively potent AhR agonists in crude extracts of commercial and consumer products with extracts of common printed newspapers containing relatively high activity (Seidel et al., 2000). Here we have extended these initial observations and further characterized the AhR agonist activity of extracts of common newspapers obtained from countries throughout the world.
[3H]TCDD (18 Ci/mmol), unlabeled TCDD, and 2,3,7,8-tetrachlorodibenzofuran (TCDF) were generously provided by Dr Steven Safe (Texas A&M University). Dimethyl sulfoxide (DMSO) was purchased from Sigma (St Louis, MO) and ethanol (ETOH) from GoldShield Chemical (Hayward, CA). [γ32P]ATP (adenosine triphosphate) (~6000 Ci/mmol), from Amersham Biosciences (Piscataway, NJ), poly[d(I·C)] from CalBiochem (La Jolla, CA), hydroxyapatite (BiogelHTP (HAP)) from BioRad (Hercules, CA), and other reagents from Fisher Scientific (Chicago, IL).
Newspapers used for screening analysis were obtained directly from vendors in the indicated countries. Samples of virgin (unprinted) paper, printed paper, and the black ink (U.S. Ink, Carlstadt, NJ) used for printing of the same newspaper were obtained from a local newspaper printing house and the paper itself contained either 40% or 100% recycled pulp. For extraction, virgin paper, newspapers (only black ink newsprint containing normal sized text font was used) or printing inks were incubated for 24 h at room temperature in DMSO, ETOH, or distilled water at a ratio of 1 g paper per 9 ml of solvent or 1 g ink per 1 ml of solvent. The solvent fraction from each extraction was collected after centrifugation (10 min at 12,000 × g) and stored in a Teflon-capped glass vial at room temperature in the dark.
Male Hartley guinea pigs (400 g) were purchased from Charles River Laboratories (Wilmington, MA) and maintained in a 12-h light:12-h dark cycle with free access to food and water. Hepatic cytosol was prepared in HEDG buffer (25mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, pH 7.5, 1mM ethylenediaminetetraacetic acid, 1mM dithiotreitol, 10% [vol/vol] glycerol) as previously described (Denison et al., 2002). Cytosolic protein concentrations were determined by dye binding (Bradford, 1976) and samples stored at −80 °C until use.
Complementary synthetic oligonucleotides containing the DRE3 AhR DNA binding site (5′-GATCTGGCTCTTCTCACGCAACTCCG-3′ and 5′-GATCCGGAGTTGCGTGAGAAGAGCCA-3′ were prepared, reannealed, and end-labeled with [32P]ATP as described (Denison et al., 2002). Guinea pig hepatic cytosol (8 mg/ml in HEDG) was incubated for 2 h in a room temperature water bath with DMSO (2% final concentration), TCDD (20nM final concentration in DMSO), or the indicated solvent or extract at a 50-fold dilution. An aliquot of the reaction was mixed with poly[dI·dC] and [32P]-DRE (100,000 cpm), and AhR:DRE complexes were resolved by gel retardation analysis (Denison et al., 2002), visualized by autoradiography and the relative amount of AhR:[32P]-DRE complex formation was quantified by phosphorimager analysis (Molecular Dynamics, Sunnyvale, CA) of the dried gels.
Aliquots of guinea pig hepatic cytosol (2 mg/ml) were incubated with 2nM [3H]TCDD in the presence of DMSO (1%), TCDF (200nM), or the indicated solvent or sample extract for 2 h in a room temperature water bath. [3H]TCDD binding in aliquots of the incubation (200 μl) was determined by HAP binding as previously described (Denison et al., 2002). The total amount of [3H]TCDD specific binding was obtained by subtracting the nonspecific binding ([3H]TCDD and TCDF) from the total binding ([3H]TCDD). The ability of a chemical(s) in a sample extract to bind to the AhR was indicated by its ability to competitively reduce [3H]TCDD specific binding.
Recombinant guinea pig intestinal adenocarcinoma (G16L1.1c8) and mouse hepatoma (H1L1.1c2 and H1L6.1c2) cells were grown at 37°C and 5% CO2 in medium (α-MEM; Gibco/Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum; Atlanta Biologicals, Lawrenceville, GA) as described (Garrison et al., 1996; Han et al., 2004). Each of these cell lines contains a stably transfected AhR-responsive luciferase reporter plasmid that responds to AhR agonists in a dose-, time-, and chemical-specific manner (Garrison et al., 1996; Han et al., 2004). The optimal induction time for TCDD in the G16L1.1c8 and H1L1.1c2 cells is 4 h, whereas in the H1L6.1c2 cells, it is 24 h. For analysis, 75,000 cells were added to each well of a clear bottom 96-well white microplate (Corning, Lowell, MA), allowed to attach overnight in 100 μl of medium, followed by replacement with medium with 100 μl of media containing DMSO (1%), TCDD (1nM), solvent (1%), or sample extract (10 μl/ml) and incubation for the indicated time. Studies determining the relative potency of the extracts used 0.1% DMSO. Cells were then rinsed twice with phosphate-buffered saline, lysed with 50 μl of luciferase cell lysis buffer (Promega, Madison, WI) for 20 min with shaking and luciferase activity measured using an Anthos Lucy 2 microplate luminometer with automated injection of 50 μl of Promega stabilized luciferase substrate.
Newly fertilized AB* zebrafish embryos were collected in egg-water (1× Danieu water; Nasevicius and Ekker, 2000). Embryos were washed and maintained at 28°C on a 14-h light, 10-h dark light cycle for the duration of the experiments. The zebrafish cytochrome P450 1A morpholino (zfcyp1a-MO), previously used to block successfully the initiation of translation of CYP1A messenger RNA in vivo (Carney et al., 2004) (5′-TGGATACTTTCCAGTTCTCAGCTCT-3′) was obtained from Gene Tools (Philomath, OR) and was fluorescein tagged to allow monitoring of injection success. Morpholino was diluted to a working concentration of 0.15mM in sterile 1× Danieau’s solution (58mM NaCl, 0.7mM KCl, 0.4mM MgSO4, 0.6mM Ca(NO3), 5mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, pH 7.6) for injection, stored in the dark at 4°C, and heated for 5 min at 65°C prior to use (Nasevicius and Ekker, 2000). Healthy embryos at the one-to four-cell stage were injected with zfcyp1a-MO using a Narishige IM300 Microinjector (Tokyo, Japan). At 24 h, embryos were assessed for fluorescence to evaluate injection success and distribution of the morpholino throughout the embryo tissue. Only undamaged embryos exhibiting strong, uniform fluorescence at 24 h were subsequently used in dosing experiments.
Zebrafish embryos (four to five embryos per vial) were exposed for 72 h (from 24 through 96 h of development) in a static dosing regimen using DMSO, βNF in DMSO (1 μg/l) or DMSO newspaper extract added to egg-water to a final dilution of 1:5000 (0.03%). For determination of ethoxyresorufin-O-deethylase (EROD) activity, all dosing solutions contained 7-ethoxyresorufin (21 μg/l). At 96 h of development, zebrafish larvae were transferred to clean egg-water for imaging and quantification of in vivo CYP1A EROD activity using a modified in ovo EROD method (Nacci et al., 2005; Wassenberg and Di Giulio, 2004). Zebrafish larvae were anesthetized with MS-222 and immobilized in 3% methylcellulose. The accumulated fluorescent product of CYP1A EROD metabolism, (i.e., resorufin), was visualized in the left-lateral gastrointestinal tract of the zebrafish by fluorescent microscopy (rhodamine red filter set; Axioskop; Zeiss, Thornwood, NY), quantified digitally by IPLab software (Scanalytics, Inc., Fairfax, VA), and expressed as a percentage of that in control embryos. EROD data were analyzed by two-way analysis of variance (ANOVA) using Statview for Windows (version 5.0.1; SAS Institute, Cary, NC). When ANOVA yielded significance ( p < 0.05), Fisher’s protected least-significant differences was used as a post hoc test.
Activation of the AhR, and the AhR signal transduction pathway, by structurally diverse chemicals as well as extracts from a variety of natural and synthetic materials has been reported by numerous laboratories (reviewed in Denison and Nagy, 2003; Denison et al., 1999). In preliminary studies we observed the ability of DMSO extracts of several U.S. newspapers to stimulate transformation and DNA binding of guinea pig hepatic cytosolic AhR, these extracts were relatively weak activators of AhR-dependent gene expression in mouse hepatoma cells (Seidel et al., 2000). Here we have extended this analysis and have also determined whether AhR ligands were present in newspapers from throughout the world, and whether these ligands are derived from the paper itself and/or the printing ink.
To determine whether AhR ligands/agonists are present in newspapers from throughout the world, we obtained newspapers directly from vendors in the indicated countries and prepared simple DMSO extracts using only black ink newsprint containing normal sized text font. The ability of each DMSO extract to stimulate transformation and DNA binding of guinea pig hepatic cytosolic AhR in vitro was examined by gel retardation analysis (Fig. 1). Although DMSO extracts from most newspapers from around the world stimulated AhR:ARNT:DRE complex formation in vitro, indicating the presence of AhR agonists, the amount of complex/agonist activity was varied (Fig. 1). Because the ability of a chemical or extract to stimulate binding of the AhR to the DRE does not guarantee that it will induce gene expression (Seidel et al., 2000), we then examined the ability of a subset of these newspaper extracts (the lower panel samples in Fig. 1) to induce AhR-dependent gene expression. Guinea pig adenocarcinoma (G16L1.1c8) cells containing a stably transfected AhR-responsive luciferase reporter gene were incubated with DMSO, TCDD (1nM), or the indicated newspaper DMSO extract at a 100-fold dilution for 4 h, and luciferase activity determined (Fig. 2). Not only did each extract induce AhR-dependent reporter gene activity to a level comparable to that produced by a maximal inducing concentration of TCDD, but their efficacy as inducers was comparable to their ability to stimulate AhR transformation and DNA binding in vitro (compare results in Fig. 2 to that of the lower panel in Fig. 1). Interestingly, the DMSO extract of the Australian newspaper induced AhR-dependent reporter gene activity to a level significantly higher than that of a maximal inducing concentration of TCDD. The reason for this “superinduction” is currently unclear, but not surprising, because we have observed this “superinduction” phenomenon previously in studies with solvent extracts of sediment and soils (data not shown). Taken together, these results demonstrate that DMSO extracts of newspapers obtained throughout the world contain relatively high AhR agonist(s) activity.
To identify the source of the active compounds in the DMSO newspaper extracts, we obtained two different newspapers printed at a local company that varied in the percentage of recycled pulp (40% and 100%) they contain and also collected the precursor materials for these newspapers (i.e., unprinted [virgin] paper and the black ink used for printing these newspapers). DMSO extracts prepared from these materials were added to G16L1.1c8 cells at a 1:100-fold dilution, incubated for 4 h and luciferase activity determined (Fig. 3A). The relative activity of the DMSO extract of each virgin paper was comparable to the activity of DMSO extracts of the respective printed papers (50–70% of that of TCDD), suggesting that the paper was the primary source of the AhR agonists. However, significant activity was still obtained using the black ink DMSO extract, indicating that ink also contains AhR agonists. The amount of ink per unit paper, and hence its actual contribution to the overall activity of the DMSO extracts of the printed papers in these experiments, if any, is unknown. Interestingly, DMSO extracts prepared from virgin paper that we obtained and tested in the year 2000 demonstrated little AhR agonist activity in AhR-based cell bioassays (data not shown). The fact that virgin paper that was extracted and analyzed at that time contained significantly less recycled material could suggest that the AhR active chemicals may have come from the recycled materials in the paper, however, differences in paper processing between these samples could be responsible for the agonist activity. The identity and source of the responsible AhR active chemical(s) in the newspaper extracts reported here remains to be determined.
Given the significant differences in structural and physiochemical properties of AhR ligands/agonists (Denison and Nagy, 2003; Jeuken et al., 2003; Khim et al., 2001; Nagy et al., 2002a; Nishiumi et al., 2005), and because DMSO would somewhat preferentially extract nonpolar chemicals, we also examined whether printed newspapers also contained AhR ligands/agonists with more polar characteristics. Accordingly, each of these materials was also incubated (extracted) overnight with 95% ETOH or distilled water as was done for DMSO, and extracts tested for their ability to induce AhR-dependent reporter gene expression in G16L1.1c8 cells at a 1:100-fold dilution (Fig. 3A). Interestingly, ETOH extracts of the papers and ink induced luciferase activity to levels comparable to that induced by DMSO extracts of the same materials. In contrast, although all of the water extracts induced luciferase activity to a level significantly greater than background (p < 0.05), the magnitude of induction was relatively low (less than 25% of maximal induction by TCDD).
To determine whether printed newspaper, virgin paper, and ink extracts could also activate the AhR in species other than guinea pig, we examined their ability to induce AhR-dependent luciferase reporter gene expression in stably transfected mouse hepatoma (H1L1.1c2) cells. The induction profile for extracts in this cell line (Fig. 3B) was comparable with that obtained in the guinea pig cells (Fig. 3A), except that the black ink extract induced luciferase gene expression in mouse treated cells to a level 50–100% higher than that induced by TCDD, suggestive of a species-specific cooperative or synergistic induction event(s) in these cells.
However, although DMSO and ethanol could extract polar and nonpolar ligands from newspapers, it is possible that the newspapers contain a high concentration of a single ligand that is poorly soluble in water but more soluble in DMSO or ethanol. If the latter possibility is true, it would suggest that extensive “washing” of the newspaper with a solvent like hexane could remove these nonpolar AhR agonists and a subsequent DMSO extract of the hexane-washed newspaper extraction should contain little AhR agonist activity. Accordingly, newspaper was extracted three times with a relatively large volume of hexane (one part newspaper to 100 parts hexane) and the resulting newspaper and hexane extract dried; a standard newspaper DMSO extract was prepared (one part newspaper to nine parts DMSO) as a control. The dried newspaper was subsequently extracted with DMSO (one part hexane-washed newspaper to nine parts DMSO) and the DMSO extract collected. In addition, the residue of the dried hexane extract was resuspended in DMSO (one part original newspaper extracted to nine parts of DMSO). Mouse hepatoma (H1L6.1c2) cells were incubated for 24 h with each of these extracts and luciferase activity determined and compared with that of TCDD. Although extensive hexane washing of the newspaper significantly reduced the amount of DMSO-extractable AhR agonist activity, the DMSO extract still induced AhR-responsive luciferase activity in H1L6.1c2 cells to 10% of that of the control newspaper extract (Fig. 4). Interestingly, the total amount of luciferase activity induced by newspaper directly extracted with DMSO was comparable to the sum of the activity of the same amount of newspaper extracted with hexane and the activity of the DMSO-extracted/hexane-washed newspaper. Considering the large volume of hexane used to extract the newspaper, these results suggest the presence of multiple AhR agonists in newspaper with different physiochemical characteristics and solvent extractability.
The above results support the presence of various chemicals with distinct physiochemical properties in newspapers and newspaper ink extracts that can activate the AhR signaling pathway. In order to estimate the relative inducing potency of newspaper extracts, G16L1.1c8 cells were exposed to increasing concentrations of DMSO or ETOH extract of newspaper or TCDD and luciferase gene expression results compared (Fig. 5). The TCDD dose response curve was modeled using a four parameter Hill equation and the calculated EC50 of luciferase induction by TCDD was 198pM (i.e. 6.5 pg of TCDD in the 100 μl incubation). The volume of original newspaper DMSO or ETOH extract needed to induce luciferase reporter gene activity to a level comparable to the EC50 value of TCDD was determined from analysis of the dilution response data in Figure 5B (lower panel) and found to be 0.06 and 0.066 μl, respectively. Back calculation from the newspaper extract concentration curve indicated that the EC50 for the DMSO and ETOH extracts contained 108 and 98 pg of TCDD inducing equivalents (TCDD-IEQs) per μl of extract, which equates to 970 and 880 ng of TCDD-IEQs per gram of newspaper, respectively. These results indicate that printed newspaper contains relatively high levels of and/or potent AhR agonists.
Although the identity of the inducing chemicals in these extracts are unknown, examination of the magnitude of the induction response in cells exposed for 4 and 24 h can provide some insight into their persistence or metabolic stability. In these experiments, mouse hepatoma (H1L6.1c2) cells containing a more stable luciferase reporter gene (Garrison, et al., 1996; Han, et al., 2004) were used to assess the induction response at 4 and 24 h by extracts of newspaper 1. These results (Fig. 6) revealed that DMSO and ethanol extracts of virgin paper and ink and the ethanol extract of printed newspaper induced significantly more reporter gene activity at 4 h as compared with 24 h; induction by the DMSO extract of printed newspapers at 24 h was slightly lower than that observed at 4 h, but not significantly different. The decrease in the overall induction response presumably results from the metabolism of the inducing and enhancing chemicals by basal and/or induced enzymes in these cells leading to a decreased magnitude of induction over the 24-h time period. These and other cell lines have previously been used to demonstrate the transient nature of the induction response by a variety of metabolically labile chemicals such as PAHs (Han et al., 2004; Machala et al., 2001a,b; Nagy et al., 2002b).
Although the above results demonstrate the presence AhR agonists in newspapers that can activate AhR- and DRE-dependent gene expression, these assays do not confirm the ability of these chemicals to directly bind to and/or activate the AhR. Accordingly, the ability of DMSO, ETOH, and water extracts of virgin and printed papers to stimulate guinea pig AhR transformation and DNA binding in vitro was determined by gel retardation analysis (Fig. 7). Similar to the gene induction results, DMSO and ETOH extract of virgin paper, newspapers, and ink stimulated AhR transformation and DNA binding, however, little or no induction or AhR:DRE complex formation was observed with the water extracts. Although it is clear that black ink contains AhR agonists and it can contribute to the overall agonist activity of newspaper extracts (only black ink printed pages were analyzed), many different colored inks are also used in a typical newspaper. Accordingly, it was of interest to determine whether these inks also contained AhR agonists. Gel retardation analysis of AhR activation by DMSO, ethanol and water extracts of white, purple, blue, yellow, and red inks were carried out (Fig. 8). Interestingly, although the ethanol extracts of all inks stimulated AhR transformation and DNA binding, the DMSO extracts of purple and red inks were inactive in this analysis (all other DMSO extracts were active); water extracts of all inks were weakly active or inactive. These results suggest the presence of differentially extractable agonists in red and purple inks as compared with the other colored inks. Alternatively, it is possible that DMSO was able to extract an AhR antagonist(s) from these inks (purple ink would almost certainly contain red ink as one of its components). As expected, DMSO and ethanol extracts of the AhR transformation/DNA binding active colored ink extracts induced AhR-dependent luciferase activity in mouse hepatoma (H1L1.1c2) cells (data not shown). Overall, the above results suggest that the overall AhR agonist activity of newspaper extracts can derive from agonists present in the paper and printing inks.
To confirm that the AhR agonist activity of the DMSO and ethanol newspaper extracts was due to direct binding of the activating chemicals to the AhR ligand binding site, we examined the ability of the extracts to inhibit [3H]TCDD binding to the AhR. Although DMSO and ETOH extracts of both printed newspapers inhibited [3H]TCDD specific binding, extracts of the virgin paper were less effective, inhibiting between 30% and 60% of [3H]TCDD specific binding (Fig. 9). Together, the above results are consistent with the hypothesis that induction of AhR-dependent gene expression by chemicals present in DMSO and ETOH extracts of newspaper occurs via direct binding to and activation of the AhR.
The above experiments confirm the ability of newspaper and printing ink extracts to bind to and activate the AhR in vitro, and to stimulate AhR-dependent gene expression in intact cells in culture. However, the significance of these observations to animals in vivo remains to be determined. Zebrafish embryos have previously been used to demonstrate the ability of PAHs and metabolically labile chemicals to stimulate AhR-dependent gene induction (CYP1A and EROD) in vivo (Billiard et al., 2006; Wassenberg and Di Giulio, 2004). Accordingly, we examined the ability of a DMSO extract of printed newspaper to induce CYP1A-dependent EROD activity in zebrafish embryos after a 72-h exposure (Fig. 10). A 1:5000 final dilution of the DMSO newspaper extract in egg-water resulted in a twofold increase in EROD activity as compared with DMSO control; βNF induced EROD activity by sevenfold. Injection of zebrafish embryos with a CYP1A-morpholino, previously shown to block induction of EROD activity by βNF, a well characterized AhR agonist (Billiard et al., 2006; Carney et al., 2004), blocked the DMSO extract induced increase in EROD activity (Fig. 10). The lack of effect of the CYP1A-morpholino on basal EROD activity suggests this activity is due to other metabolic enzymes. In contrast, although the ETOH extract could stimulate AhR-dependent gene induction in cells in culture, it failed to induce EROD activity above background in exposed zebrafish at any concentration tested (data not shown), suggesting that the responsible chemicals are very metabolically labile in fish in vivo and/or eliminated relatively rapidly. Overall, these results confirm that AhR active compounds present in the DMSO newspaper extract are both soluble and stable enough to be absorbed into the fish embryos and stimulate AhR-dependent gene expression in vivo before being metabolized or degraded.
The AhR has been well characterized with regard to its role in mediating the biochemical and toxic effects of a variety of environmental contaminants such as dioxins and related dioxin-like chemicals, primarily HAHs. However, recent studies from our laboratory and others have demonstrated the wide structural and physiochemical diversity of AhR ligands/agonists as well as the presence of AhR ligands in a variety of food and commercial consumer product extracts (Amakura et al., 2003; Denison et al., 1999; Jeuken et al., 2003; Seidel et al., 2000). In our previous study, we found that DMSO extracts of several newspapers could stimulate AhR transformation and DNA binding (Seidel et al., 2000). Here, we have extended this work and have demonstrated the presence of DMSO, ETOH, and water soluble compounds that can be extracted from newspapers and that can bind to and activate the AhR and the AhR signal transduction pathway. Use of these extracts may allow investigation into the events and/or signaling processes involved in the synergistic induction response.
The ability of DMSO extracts of newspapers that we obtained in various countries around the world to stimulate both AhR transformation and DNA binding as well as AhR-dependent reporter gene expression, indicates that the source of the AhR agonists in these samples is not unique to a particular paper or ink product, or printing company from one country, but is likely due to common constituents. In addition, the active compounds in these extracts are very efficacious activators of the AhR signaling pathway as demonstrated by their ability to induce AhR reporter gene expression in the bioassay at levels up to 90% of a maximally inducing concentration of TCDD. Interestingly, the DMSO extract from an Australian paper (Fig. 2) and DMSO and ETOH ink extracts (Figs. (Figs.3B3B and and5)5) induced AhR-dependent reporter gene expression to greater than 150% of that of TCDD. This was not completely surprising given that we have identified a number of pure chemicals and crude extracts that can induce AhR-dependent gene expression to a level greater than that produced by a maximally inducing concentration of TCDD (unpublished observations). Although there may be multiple mechanisms for this enhancement, one possibility is enhanced cross-talk between the AhR and another signaling pathway that is stimulated by chemicals present in the extracts leading to a synergistic or cooperative increase in gene expression. For example, activation of protein kinase C has been shown to synergistically increase the magnitude of AhR-dependent gene induction by TCDD both in cells in vitro and in rats in vivo (Chen and Tukey, 1996; Long et al., 1999). Additionally, the ability of dexamethasone to potentiate AhR-dependent signaling in a glucocorticoid-dependent manner has also been reported (Celander et al., 1997; Hoogenboom et al., 1999). Accordingly, it is not unexpected that the complex mixture of chemicals in a crude extract could enhance (or inhibit) AhR-dependent gene expression by their concurrent effect on multiple cell signaling pathways.
The results obtained here do not clearly indicate whether the AhR agonists in these newspaper sample extracts are entirely derived from the paper or whether the printing ink can contribute to the induction response. DMSO extracts of virgin paper, printed paper, and black ink were all able to induce reporter gene expression significantly in both guinea pig and mouse luciferase reporter cell lines indicating that the response can occur in cell lines from at least two different species. However, induction by the black ink extract was approximately twice that of all of the paper extracts and this suggests that ink could be a major contributing source of AhR ligands in printed newspapers. Accordingly, one would then expect the printed paper to have greater activity than virgin paper and this was not observed in our experiments. One possible explanation for this discrepancy is that the amount of ink present per unit paper may be so small that its AhR ligand contribution might be negligible in comparison to the amount of AhR ligands extracted from the larger quantity of paper itself. It is also possible that the virgin paper tested here was derived in some part from recycled paper (either 40% or 100%) and therefore may contain residual ink-derived chemicals that can be extracted.
Although the exact source and identity of AhR activators present in these newspaper extracts remain to be determined, we can draw some conclusions regarding their physiochemical properties. Not only did the ETOH extracts of the virgin paper, printed paper, and black ink stimulate transformation and DNA binding of the AhR, but both ETOH and water extracts significantly induced AhR-dependent reporter gene expression in cells from two species, guinea pig and mouse, with the ETOH extracts being far more active. ETOH and water extracts of recycled paper fibers have been previously shown to induce AhR receptor dependent reporter gene expression in rat cells (Binderup et al., 2002). Taken together, these results are consistent with the existence of water soluble AhR agonists in newspaper and paper products. However, it is possible that the newspaper could contain relatively high concentrations of an AhR agonist that is poorly soluble in water and the limited activity observed with water extracts is due to the limited solubility of the active chemical in water. Although extensive extraction of newspaper with hexane extracted/removed about 90% of the AhR agonist activity from the newspaper, it did not entirely remove the activity. These results not only suggest that the remaining AhR active chemical(s) has somewhat more polar characteristics, but also that newspapers contain more than one AhR agonist and argue against the active compounds simply being PAHs derived from carbon black, a common ingredient in black printing inks. The decrease in AhR-dependent reporter gene expression typically observed with the DMSO extracts at later time points of incubation indicates that the AhR active compounds are metabolically labile and are therefore not TCDD-like HAHs, which are metabolically stable and persistent activators of AhR-dependent gene expression. The flavonoids, a group of polyphenolic phytochemicals including more than 5000 chemicals, are one class of compounds that is likely to be present in these extracts. This conclusion derives from the fact that most newspapers are printed with soy-based inks and a variety of several soy flavonoids, such as genistein, quercetin, and kaempferol have been previously reported to interact with the AhR pathway (Ciolino et al., 1999; Han et al., 2003; Zhang et al., 2003). Bioassay-based fractionation studies are currently underway to identify the responsible AhR agonists in these extracts.
Although the results of our in vitro and intact cell experiments demonstrate the presence of AhR agonists in newspaper extracts, the biochemical and toxicological relevance of these observations to intact animals is unclear. One area of concern, particularly in the European Union, is the potential risk associated with the use of recycled fibers in food packaging materials reported to contain extractable estrogen receptor and AhR agonists (Binderup et al., 2002; Vinggaard et al., 2000). Here, we found that zebrafish embryos exposed in vivo to a crude DMSO extract of newspaper had significantly elevated EROD activity after 72 h of exposure. The ability of a CYP1A-morpholino to inhibit the induction response clearly indicated that this was a CYP1A-dependent activity and previous studies have shown that induction of CYP1A in zebrafish is AhR-dependent (Billiard et al., 2006; Wassenberg and Di Giulio, 2004). These results demonstrate both the bioavailability of AhR agonists present in the sample extract and their ability to activate the AhR pathway in vivo. Based upon what is currently understood about dioxin-like compounds and AhR-mediated toxicity, one of the defining characteristics of compounds that produce the suite of TCDD-like toxic effects is their metabolic stability. The significant decrease in reporter gene expression by these extracts at 24 h compared with 4 h indicates that the AhR ligands in the crude newspaper and black ink extracts are metabolically labile and rapidly metabolized and unlikely to produce AhR-dependent toxicity. However, at least in fish, there is some evidence that chronic exposure to relatively high doses of a metabolically labile PAH AhR agonist (i.e., βNF) can be as effective at producing AhR-dependent toxicity as a single exposure to those that are metabolically persistent (Grady et al., 1992). In addition, PAHs, which are normally rapidly metabolized and eliminated have been shown to produce AhR-dependent toxic effects when the PAH metabolizing enzyme (CYP1A) is either chemically inhibited or knocked down via morpholino in fish embryos (Billiard et al., 2006; Wassenberg and Di Giulio, 2004). The resulting increase in persistence of the PAHs in this experimental system leads to more persistent activation of the AhR and thus some level of AhR-dependent toxicity. One can also envision that a crude extract of a complex mixture of chemicals could contain both CYP1A inhibitors and AhR agonists and that exposure to such an extract on a chronic basis could lead to AhR-dependent toxicity. Conversely, chronic induction of CYP1A and other metabolizing enzymes has the potential to produce adverse health effects through their ability to metabolically activate other compounds (i.e., PAHs) to which animals may be exposed.
Although direct exposure of the general public to these AhR active chemicals is expected to be minimal through the handling and reading of newspapers, environment exposure to higher levels of these newspaper-derived chemicals is a more likely scenario. A potential site for aquatic exposure to the AhR ligands present in newspapers and printing inks could occur downstream of a recycling plant. Newspaper recycling incorporates a variety of repulping and deinking processes which involve aqueous mixes that can contain a number of additives such as hydrogen peroxide, acidic and alkaline surfactants, alcohols, fatty acids, and vegetable oil products (Toda, 2004). The amount of newspapers recovered in the United States has steadily increased since 1990 with close to 10 million tons recovered in 2005. This represents almost 70% of newsprint produced in the country (Paper Industry Association Council, 2006). Based on our results, we calculate that there are 970 and 880 ng of TCDD-IEQs per gram of printed newspaper for the DMSO and ETOH extract, respectively. Given that a double page of a typical newspaper weighs ~16 g, the estimated extractable TCDD-IEQs is 15.5 and 14 μg per double newspaper page for the DMSO and ETOH extracts, respectively. With greater than 14 million tons of newspapers printed per year in the United States alone, this could equates to greater than 11,000 kg of polar, extractable TCDD-IEQs present in newspapers printed each year in the United States alone. For our fish exposure experiments, a 1:5000 dilution of a DMSO newspaper extract, or 108 pg of TCDD-IEQs per 5 ml of water, was sufficient to significantly induce AhR-dependent CYP1A EROD activity. How this would compare to newspaper recycling effluent remains to be determined because the presence and levels of these chemicals in wastewater streams are unknown. However, in the European Union, paper recycling plants generate an estimated 8–16 m3 of wastewater per ton of pulp. In general, this wastewater is either sent to a municipal sewage plant or is treated to remove particulates and to decrease oxygen demand prior to being released into surface waters (Toda, 2004). Future work directly testing samples of paper recycling wastewater using these AhR-based bioassays and fish exposure assays will allow both bioassay-directed fractionation approaches to identify the responsible chemicals and can provide information as to whether compounds that can activate the AhR pathway are extracted and released into the environment during newspaper re-pulping and deinking processes.
National Institutes of Environmental Health Sciences (ES012498) to M.S.D., (ES007685) to M.S.D., (ES007031) to R.T.D.; Superfund Basic Research Grants (ES004699) M.S.D., (ES010356) to R.T.D.; and the California Agricultural Experiment Station.