ToxSci Advance Access originally published online on May 16, 2007
Toxicological Sciences 2007 98(2):375-394; doi:10.1093/toxsci/kfm122
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Polychlorinated Biphenyls–Modulated Tumorigenesis in Sprague–Dawley Rats: Correlation with Mixed Function Oxidase Activities and Superoxide (O2•–) Formation Potentials and Implied Mode of Action
,1
* General Electric Company, Fairfield, Connecticut 06431
General Electric Company, Global Research Center, One Research Circle, Niskayuna, New York 12309
1 To whom the correspondence should be addressed at 1479 Dean St., Niskayuna, NY 12309-5235. Fax: (518) 862-2702. E-mail john.brown{at}ge.com.
Received December 19, 2006; accepted May 8, 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Parallel, chronic (24 months) multidose bioassays of the PCB (polychlorinated biphenyls) Aroclors 1016, 1242, 1254, and 1260 in male and female Sprague–Dawley rats showed sex/Aroclor-dependent increases in hepatic tumors and decreases in extrahepatic tumors. To elucidate the PCB mode of action (MOA) involved, levels of a number of hypothesized mediators were measured in liver specimens collected at 3, 6, 12, 18, and 24 months and screened for correlation with late life hepatotumorigenesis (HT; mostly adenomas). Consistently correlated with HT were (1) tissue accumulations of
PCBs (correlated in both sexes) and of dioxin equivalents (toxic equivalency [TEQ]; correlated in females only); (2) net activities of six groups of mixed function oxidases (MFOs), some PCB-induced, some PCB-repressed, as determined by differential metabolism of PCB congeners; (3) activities of deproteinated, reoxidized hepatic cytosols as catalysts for superoxide (O2•–) production, such activity having the chemical characteristics of redox-cycling quinones (RCQs), e.g., those derived from the glutathionylated estrogen catechols that were identified in the female rat livers; and (4) increased expression of the indicator of cell proliferation, proliferating cell nuclear antigen. The new findings, along with other recently reported relationships, were indicative of a MOA consisting of (1) PCB/TEQ accumulation in rat tissues; (2) PCB/TEQ repression of constitutive MFOs; (3) PCB/TEQ induction of other MFOs; (4) MFO-mediated formation of RCQs; (5) RCQ-mediated formation of O2•–; (6) O2•– dismutation to H2O2; and (7) H2O2-mediated mitotic signaling, resulting in the proliferation of spontaneously or otherwise initiated cells to form hepatic tumors, as in tumor promotion. Key Words: Tumorigenesis; cancer; rat; PCB; PH50; quinone; RO3; superoxide.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
The Mayes et al. (1998)
chronic toxicity and carcinogenicity bioassay of PCB (polychlorinated biphenyls) Aroclors 1016, 1242, 1254, and 1260 in Sprague–Dawley (S-D) rats is the most extensive of the PCB cancer bioassays conducted to date (e.g., Kimbrough et al., 1975; National Cancer Institute [NCI], 1977; Norback and Weltman, 1986; Schaeffer et al., 1984). Together, these studies showed that PCBs increased the incidence of liver tumors in three of the four rat strains used, while decreasing tumors at other sites. The increases in tumorigenesis reported in these studies occurred despite absence of positive findings in in vitro mutagenicity tests (Agency for Toxic Substances and Disease Registry [ATSDR], 2000; U.S. Environmental Protection Agency [EPA], 1997; Safe, 1989, 1994; Silberhorn et al., 1990; World Health Organization [WHO], 2003). With respect to DNA binding, a presumed indicator of tumor initiation, some investigators have reported the formation of DNA adducts with quinones derived from individual nonpersistent PCB congeners in vitro (Arif et al., 2003; Srinivasan et al., 2001); protein, though not DNA, adducts with such quinones in short-term tests in vivo also using individual PCB congeners (Lin et al., 2000; Pereg et al., 2001); and hepatic mutations in rats treated with a monochlorobiphenyl (Lehmann et al., 2006). However, there is no evidence of PCB–DNA interaction products in Aroclor-dosed animals (Nath et al., 1991; Schilderman et al., 2000; Whysner et al., 1998). The results of all studies conducted to date on Aroclors and individual PCB congeners, however, are consistent with hepatotumorigenesis (HT) promotional activity, for both low-ortho ("dioxin-like") and multi-ortho ("phenobarbital [PB]–like") PCB congeners in two-stage bioassays (ATSDR, 2000; Buchmann et al., 1991; Glauert et al., 2001, Shirai, 1997). Moreover, rat liver is an organ with a high incidence of spontaneous initiation (Pitot et al., 1989, 1996; Schulte-Hermann et al., 1981, 1989; Ward, 1983) so that agents with promotional activity alone can produce tumors.
There have been many published hypotheses concerning possible modes of action (MOA) by which PCBs or related xenobiotics might contribute to the development of rat liver tumors. Suggested candidates for such events, as reviewed in detail by Glauert et al. (2001), have included induction of mixed function oxidase (MFO) activity, a characteristic response to persistent lipophile accumulation, leading to new metabolic processes and products (Nebert, 1994; Nebert et al., 2000); high dose cytotoxicity, leading to hepatic tissue necrosis and consequent mitotic stimulation and regenerative proliferation (Ames and Gold, 1990); production of "oxidative stress," "reactive oxygen species" (ROS), or reactive oxygenated metabolites (ROM) by any of a dozen possible routes, followed by oxidative attack on DNA or proteins (Bolton et al., 2000; Cerutti, 1985; Halliwell and Gutteridge, 1999; Hassoun et al., 2001; Kensler and Trush, 1984; Klaunig et al., 1995, 1998; Liehr and Roy, 1990; Nebert et al., 2000; Stohs, 1990); inhibition of gap junction intercellular communication (GJIC) (Bager et al., 1997; Rhee et al., 2003; Robertson and Hansen, 2001; Upham et al., 1997; Whysner and Wang, 2001); and possibly direct stimulation of mitosis or inhibition of apoptosis (Glauert et al., 2001). Plausibly relevant activities of all these suggested proposed contributors to tumor promotion have been demonstrated in short term or in vitro tests. However, none had ever been measured in tissues taken during the course of a chronic bioassay to determine whether they exhibited the proper timing and dose–response relationships for relevance to late life tumor development.
The Mayes et al. (1998) bioassay provided an opportunity to test many of the hypothesized intermediates for correlation with actual tumor development. This study showed that the incidence of HT depended upon Aroclor type, dose, sex, and sex-Aroclor combination in a complex pattern. Additional animals had been included to allow intermediate sacrifices at three time points for the males (6, 12, and 18 months) and five for the females (3, 6, 9, 12, and 18 months). The resulting preserved tissue specimens permitted measurements of PCB and toxic equivalency (TEQ) accumulation, representative MFO levels and activities, PCB congener ratios indicative of other MFO activities, selected Phase 2 enzymes, trace metal levels, indicators of cytotoxicity, and indicators of ROS production, including redox-cycling activity, to supplement the previously reported measurements of animal weight gains, organ weights, serum chemistry, hematology, tissue pathology (Brunner et al., 1997; Mayes et al., 1998), and tissue immunohistopathology (Whysner and Wang, 2001). Each of the newer or older measurements was assessed for covariance with tissue PCB composition (to identify the Aroclor components responsible for the responses) and with the incidence of HT (to identify the earlier responses predictive of, or correlated with, late life HT). Consequently, any biochemical change of interest could be tested for consistency with tumorigenesis over variations in Aroclor accumulation, TEQ, dose, sex, age, and target organ.
In using a study of the effects of various Aroclors on S-D rats to elucidate the events involved in tumorigenesis by these complex mixtures, certain facts pertaining to both the Aroclors and the S-D rat strain should be noted. First, all the Aroclors are complex mixtures of many individual PCB congeners, each with a different chlorination pattern, susceptibility to metabolism, TEQ, and other enzyme-inducing properties. However, because these mixtures are readily resolved into their components by high-resolution gas chromatography (Frame et al., 1996a,b) the tissue levels and fates of the individual congeners can be readily tracked and quantified. Second, the Aroclors, especially Aroclor 1254, are known for their ability to induce rat liver MFOs, including various isozymes of P450 cytochrome (CYP) families 1A, 1B, 2A, 2B, and 3A (Bandiera, 2001), most of which have at least some activity as aromatic hydroxylases (Lee et al. 2003). With respect to their CYP-inducing properties, the individual PCB congeners fall into two broad overlapping classes (Parkinson et al., 1983). Certain non-ortho– or mono-ortho–substituted congeners share with dioxin (e.g., 2,3,7,8-tetrachlorodibenzo-p-dioxin, TCDD) the ability to activate the aromatic hydrocarbon receptor (AhR) and thereby induce CYPs 1A1, 1A2, 1B1, 2S1, and a number of dioxin-like toxic responses. This activity in rats can be quantified using toxic equivalency factors (TEF) for individual congeners and summed using the dioxin TEQ for a PCB mixture (Van den Berg et al., 2006). Alternatively, almost all mono-ortho– or multi-ortho–substituted congeners share with PB the ability to induce expression of CYPs of families 2 and 3 as well as other responses (Hansen, 1999). Also induced by both PB and the ortho-substituted PCBs is the cytochrome P450 cofactor, P450 reductase, an NADPH (nicotinamide adenine dinucleotide phosphate, reduced)–reducible flavoprotein (Parkinson et al., 1980). This enzyme can also serve as an electron source for quinone-mediated redox cycling (Iyanagi, 1990). However, it has not been known whether the dioxin-like or the PB-like activities of PCBs are responsible for HT. Third, most PCB congeners are metabolized by either PCB-induced or constitutive MFOs, albeit at widely varying rates (Brown, 1994), giving mostly polychlorinated biphenols, with some further conversion to catechols (Hansen, 1999). The possibility that catechols may be oxidized to quinones that may contribute to HT has already been noted. Fourth, the gas chromatograms of rat tissues following chronic dosing with various PCB Aroclors show marked alterations in PCB congener retention patterns that reflect both the types and extent of MFO-mediated PCB metabolism. Since PCB metabolism results entirely from the aromatic hydroxylase activity of the various MFOs involved, the patterns of such alterations can not only indicate the types of MFO activity present, but also provide a measure of the total aromatic hydroxylase activity arising from both identified and unidentified MFOs.
The S-D strain of rats are outbred, relatively large (M 
850 g; F 500 g), and widely used in tumorigenicity bioassays because of their susceptibility to chemical carcinogens. The undosed controls showed substantial incidences of pituitary and mammary tumors, and lesser incidences of prostate, pancreatic, and adrenal tumors, all of which could be repressed by PCB administration (Mayes et al., 1999; supplemental Tables S-4, S-5). The S-D females exhibit the endocrine peculiarity of not undergoing a normal midlife menopause, but instead entering a state of constant estrus and estrogen production (Wetzel et al., 1994; WHO, 1999). In addition, estrogen metabolism (via the aromatic hydoxylase activity of CYP1 isozymes) in the S-D rat yields largely a mixture of glutathionylated estrogen catechols, readily oxidized to glutathionylated estrogen quinones, which are potent redox-cyclers, i.e., catalysts for superoxide (O2•–) production (Butterworth et al., 1998). Possibly related to the above are the findings that Aroclor 1254, which is a strong inducer of CYP1 MFOs in rats, is a potent hepatotumorigen in S-D females, but not in S-D males (Mayes et al., 1998). However, in the Fisher 344 rat Aroclor 1254 produced even less HT in females than in males, and was deemed noncarcinogenic in that strain (National Cancer Institute [NCI], 1978). These findings indicate that the details of the tumorigenic response to PCBs can vary with rodent strain and sex, as well as with target organ.
Finally, it should be noted that some measurements correlated with HT (i.e., PCBs, TEQ, and O2•– formation) have been previously described in conference reports (Brown et al., 1996, 2001; Fish et al., 1997). However, none of these correlations were judged sufficient to define all essential steps in a MOA. The intervening development of techniques for extracting MFO data from tissue residues of selected PCB congeners, the discovery of additional correlations, and importantly, external progress in elucidating key signal transduction pathways (Agrawal and Shapiro, 2000; Finkel, 2003; Ohtake et al., 2007; Rhee et al., 2003; Waxman and O'Connor, 2006) have now provided the necessary bases for describing a biologically plausible MOA for PCB-induced tumorigenesis. This MOA includes key events of the general types already hypothesized by Nebert et al. (2000) for cell cycle control, i.e., induction of multiple MFOs, formation of ROM, and production of ROS.
Although any MOA for cancer development is of general interest, the present report will be limited to descriptions of measurements made, correlations found, and the implied MOA in Aroclor-dosed and undosed S-D rats. Discussion of the broader implications of this novel MOA will be deferred.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Aroclors Used
The sources and chemical compositions of the Aroclor specimens used in the bioassay are described in the original report (Mayes et al., 1998
). However, a change in the WHO international consensus TEF values (Van den Berg et al., 2006) required a revision of the calculated TEQ values to 0.08, 4.8, 21.1, and 0.5 ppm for Aroclors 1016, 1242, 1254, and 1260, respectively. The accumulation potentials of these Aroclors (ratio of mean midlife rat PCB/lipid to PCB/feed) were 0.13, 0.17, 1.63, and 2.98, respectively. Thus, all four possible combinations (low–low, high–low, high–high, and low–high, respectively) of these two characteristics were represented by the Aroclors used in the bioassay.
Tissue Specimens
Necropsy samples of liver, brain, mammary adipose, and perirenal adipose tissues were removed at 3, 6, 9, 12, 18, and 24 months from up to six animals per sex per test group and quickly frozen at –70°C for future analysis as previously described (Brunner et al., 1997; Mayes et al., 1998). Liver samples intended for separate determinations of soluble versus membrane-bound analyte levels or activities were further processed to produce cytosolic and microsomal fractions by conventional centrifugal procedures (Guengerich, 1994). To produce the MFO-induced microsomes used for validating the CYP assay procedures and for the redox-cycling assays, two male and two female S-D rats were each dosed with either PB or beta-naphthoflavone for 3 days by Battelle Laboratories (Columbus, OH), then sacrificed on the fourth day and the livers removed, divided into, ca. lg. portions, and shipped to us for use. Microsome stock suspensions were prepared as above from individual liver specimens as needed, and kept frozen when not in use.
Analytical Procedures
Polychlorinated biphenyls.
For PCB analysis, thawed tissue samples from three animals from each Aroclor/dose/sex/time were weighed, homogenized, and Soxhlet-extracted with 1:1 hexane:acetone. The extracts were used for both congener-specific PCB analysis and the determination of total tissue lipids. Analyses of PCB concentrations and congener distributions in both Aroclors and rat tissues were performed by Northeast Analytical Environmental Lab Services, Schenectady, NY, using the 118-peak DB-1 capillary gas chromatographic procedure described by Frame et al. (1996a). TEQ values were calculated using the 2005 consensus TEF values for PCB congeners (Van den Berg et al., 2006).
Biochemical and ROS analyses.
Methods of enzymology procedures were used for the analyses of glutathione reductase (GR) (Racker, 1955); quinone reductase (QR), also known as DT-diaphorase (Ernster, 1967); and microsomal superoxide (O2•–) (O'Brien, 1984). Cytosolic superoxide dismutase (SOD; presumably very largely Cu, Zn-SOD) was measured by competition with cytochrome c for superoxide produced by the action of xanthine oxidase on xanthine in 0.005M phosphate buffer at pH 7.4 (Geller and Winge, 1984; Hassan, 1984) and the response calibrated by measurements on dilutions of a 1 mg/ml (5800 units/ml) commercial SOD sample. Commercial test kits were used for the analyses for CYP1A1 protein (Amersham Life Sciences ELISA kit No. RPN 269); CYP2B1/2 proteins (Amersham Life Sciences ELISA kit No. RPN 270); glutathione (GSH) (Calbiochem kit No. 354102); and glutathione peroxidase (GPx) (Calbiochem kit No. 354104). Literature procedures were used for the analyses for hydrogen peroxide (H2O2) (Black and Brandt, 1974); glutathione S-transferase (GST) (Habig et al., 1974); ethoxy-, methoxy-, pentoxy-, and benzyloxy resorufins (EROD, MROD, PROD, and BROD) (Burke et al., 1985); thiobarbituric acid reactive substances (TBARS) (Yagi, 1994); catechol O-methyl transferase (COMT) (Reenila et al., 1995); porphyrins (van Birgelen et al., 1996); and 
-tocopherol (Bachowski et al., 1998).
Using these procedures, analyses for COMT, GSH, GPx, GR, GST, total porphyrins, QR, and SOD were conducted in triplicate on hepatic cytosol fractions from the same rats as used for the PCB analyses. The hepatic microsome fractions were similarly analyzed for CYP1A1, CYP2B1/2, EROD, MROD, PROD, BROD, TBARS, -tocopherol, and production of O2•– and H2O2 in response to NADPH and O2.
Determination of Relative Rates of Metabolism for PCB Congeners
The pharmacokinetic analysis presented in the Supplemental Materials demonstrates that for any two congeners, i, j, present at concentrations ci and cj, the Aroclor-normalized ratio R(i/j), defined as (ci/cj)rat/(ci/cj)aro, at steady state in an animal, such as the rat, with a sizeable nonmetabolic clearance rate, kn, could be related to the metabolic clearance rates, kmi and kmj, as follows:
 |
Thus, by selecting as reference congener i one with a low metabolic rate relative to that of congener j, R'(i/j) becomes a measure of relative kmj that can be used as an indicator of the PCB oxidase (PCBox; i.e., aromatic hydroxylase) activities present. These fall into two broad types: Type A, which is active against only non-ortho and mono-ortho PCB congeners, and formerly referred to (correctly) as P450 1A–like (Brown, 1994) and Type B, which is at least as active against multi-ortho–substituted PCB congeners as well, and formerly referred to (incorrectly) as P4502B like (Brown, 1994). The bases for selecting particular i, j congener combinations as indicators of Types A or B PCBox activities are presented in the Supplemental Materials and Table S-1. Also described is a definition of the term R''(i/j) used to indicate the fractional contributions of Type A or B metabolism to the total (used in Figs. 6a and 6b).
|
Cytosolic Redox-Cycling Activity as an Indicator of Superoxide (O2•–) Production Potential
Since it was known that the cytosol preparations from treated animals contained glutathione (which might reduce quinones originally present to diols) at levels comparable to those in the controls (Brown et al., 2001
), and also that the oxidized and glutathionylated estrogens present were in the reduced (i.e., catechol) form, which is inactive for redox-cycling (Butterworth et al., 1998), all cytosols were subjected to a mild (re)oxidation before redox-cycling assay. For this, frozen (– 70°C) 1.5-ml portions of the cytosols were thawed, passed through a Centricom-10, or Centricon-3 membrane filter (passing molecules < 10 or < 3 kDa, respectively, but with no difference in filtrate activity), then treated with 10 µl of 3% H2O2 plus 10 µl of 1 mg/ml horseradish peroxidase for 30 min., then with 10 µl of 0.l mg/ml catalase (to remove excess H2O2) and again passed through the membrane filters. For the assay, microtiter plate wells were first loaded with 100 µl of the (re)oxidized, membrane-filtered cytosols, followed by 100 µl of a stock solution prepared from 60 µl of a 3.7 mg/ml suspension of hepatic microsomes from a PB-dosed male S-D rat, 500 µl of a 22 mg/ml solution of succinylated cytochrome c (Kuthan et al., 1982), 260 µl of a 1.0 mg/ml catalase solution, and 4600 µl of 0.02M tris–KCl, pH 7.60, buffer, and finally with 10 µl of 10mM NADPH to start the reaction. The reduction of the cytochrome c was followed spectrophotometrically by loss of absorption at 550 nm (Kuthan et al., 1982). All measurements were made in triplicate. The measurements of redox-cycling activity on hepatic cytosols from the 18- and 24-month sacrifice groups were made within a few days of each other, and found similar in both magnitude and correlation with HT (Fig. 3, below). Measurements made on the cytosols from the 6- and 12-month sacrifice groups were made 3 months later using a test mixture derived from a different PB-induced microsome suspension. These measurements were also similar to each other and correlated with HT, but 25- to 30-fold lower in absolute magnitude (Fig. 3). It was not possible to determine the source of this uniform technical error. However, when the remaining cytosols from all Aroclor 1254 test groups were reassessed with a single microsome preparation, there were no significant differences among the redox-cycling activities of the 6-, 12-, 18-, and 24-month samples for any one dose group. Therefore, it was concluded that the cytosolic redox-cycling quinone (RCQ) levels did not vary much over the course of the bioassay. However, since it was not possible to standardize the activities of the different microsome preparations that had been used the data obtained must be regarded as demonstrating relative rather than absolute measures of RCQ activity. All of the above measurements were conducted in triplicate for each Aroclor, dose, sex, and time point.
|
Succinylated cytochrome c bleaching is basically an assay for the formation of small molecules capable of carrying out 1-electron reductions and is usually regarded as a measure of O2•– production. However, in the in vitro assay procedure used, we have a system where the O2•–is being produced from another small molecule 1-electron reducing agent, namely, a semiquinone radical anion, Q•–. Thus, it is uncertain whether the electron was being transferred from Q•– to O2 and then to the cytochrome c, or from Q•– to cytochrome c directly. However, the difference is irrelevant to the question of potential for O2•– production in vivo since the measured cytochrome c reduction rate depends upon the rate of Q•– production (and hence redox-cycling activity) in either case, and any such redox-cycling activity in vivo would lead directly to O2•– formation because of the high rate of the nonenzymatic Q•– to O2 electron transfer (O'Brien, 1991). Thus, the measured activities of the hepatic cytosols can be taken as indicators of relative potential for O2•– production in the liver.
To characterize one family of the reduced RCQs present, 30 membrane-filtered cytosols from the Aroclor 1242- or 1254-dosed female rats were combined and subjected to ion exchange (Supelco Superclean LC-SAX) and reversed phase (Superclean LC-18 SPE) chromatography to obtain a broad fraction that would contain any glutathionylated estrogen catechols present. This was subjected to high-performance liquid chromatography (HPLC), using a DIONEX DX 500 chromatograph, a ZORBAX SB-octyl (4.6 mm x 25 cm) HPLC column, and an ED50 electrochemical detection system, to resolve the individual redox-active species present. For species identification, the resulting chromatographic patterns were compared with those prepared by the oxidation and glutathionylation of the 2,3- and 3,4-estrogen catechols, as described by Butterworth et al. (1996).
Statistical Analysis
The data sets for the various group mean chemical, biochemical, or histological responses measured in this or the earlier (Mayes et al., 1998; Whysner and Wang, 2001) studies were assessed for covariance with HT by determination of r2 and p values for simple linear regressions. Variation of individual values from group means was determined for some responses, but not indicated by error bars on Figures 1–6
since the primary concern was with the correlations for the data sets as a whole.
|
|
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Overview of Aroclor Bioassay Results
While readers are encouraged to refer to Mayes et al. (1998)
for a description of the study (including treatment of Aroclor 1254 to remove PCDFs) and the details of the results, the principal findings are briefly summarized below. Table 1 presents the effects of Aroclor type, dose, and sex on total PCB and TEQ intake; mean midlife adipose and liver PCB and TEQ levels; and the incidence of hepatic and nonhepatic neoplasms, as well as additional data from Mayes et al. (1999). Briefly, hepatic tumors were always a minor part of the total. In the males, liver tumor incidence was not significantly increased by Aroclors 1016, 1242, or 1254 but was in the 100 ppm Aroclor 1260 dose group. For females, there was a significant and generally dose-related increased incidence of hepatic neoplasms (primarily adenomas) in all treatment groups, except for the 50 ppm Aroclor 1016 group. The magnitude of the increase was greatest for Aroclor 1254, followed by Aroclor 1260 
Aroclor 1242 > Aroclor 1016 (Table 1). Consistent with the results of many other studies, the Aroclors also produced decreases in the incidence of many types of constitutive tumors, as well as in total tumor incidence in males (Mayes et al., 1999; Supplemental Tables S-4, S-5) which indicates that the PCB mixtures tested had both tumorigenic and antitumorigenic activity. As also shown in Table 1, the Aroclors used varied widely in their tendency to bioaccumulate and in the TEQ of the tissue-accumulated PCBs.
|
Tissue Accumulation of
PCB and TEQ and HTThe time course of tissue PCB accumulation is illustrated in Figure 1, which shows the mean dose- and lipid-normalized total PCB concentrations in mammary adipose tissues of each group of Aroclor-treated rats at various time points of continuous or discontinuous (i.e., stop-study) dosing. Replicate measurements of
PCB in spiked adipose and liver specimens showed 96.1–100.2% recoveries and coefficients of variation (CVs) of 0.67–2.16%. Interanimal variations in lipid-normalized adipose PCB within the usual three-animal/dose/sex/time groups exhibited CVs that averaged 28% (range, 17–47%) in the continuously dosed animals or 54% (range, 48–60%) in the stop-study animals. Because of these sizeable CVs, individual animal values used for calculating group means were normalized and combined where feasible, in order to increase the number averaged from three animals to six or nine animals. As shown in Figure 1, steady-state PCB tissue concentrations had generally been attained by 6–12 months. Between 18 and 24 months, PCB tissue levels generally rose for all Aroclors, possibly reflecting losses of body fat or reduced non-metabolic (i.e., fecal) clearance in the senescent animals. The differences in bioaccumulation tendencies of the four Aroclors paralleled those reported for man (Brown, 1994). Cessation of dosing resulted in substantial loss of tissue PCB from Aroclors 1016 or 1242, but less of that from Aroclors 1254 or 1260. The accumulated PCBs consisted mainly of the persistent or semipersistent (generally, 4,4'-substituted) congeners with only traces of those not so substituted (Brown, 1994). The lipid-normalized accumulation of PCBs, which reached a near plateau after about 12 months, demonstrated that mean midlife accumulation (i.e., averaged group means from 6, 12, and 18 month data sets) was a valid indicator of mean internal exposure. This measure was also less variant than data from the three-animal subsets analyzed for interim determinations of biochemical indicators. Consequently, this metric (see Table 1) was used in all of the correlations described below.
A plot of HT versus mean midlife adipose PCB for male rats of all Aroclor dose groups (Fig. 2b) showed a reasonably smooth curve with a PCB accumulation threshold for increased HT over control levels at approximately 500 mg PCB/kg lipid, and a linear increase in HT for higher concentrations of adipose PCB. The fact that the data points for the high TEQ Aroclor 1242 and 1254 accumulations fell on the same curve as those for the low TEQ Aroclors 1016 and 1260 residues indicates absence of correlation between HT and tissue TEQ in S-D males (see Table 1). This finding is consistent with the established absence of hepatotumorigenic activity for TCDD in male S-D rats (Kociba et al., 1978). Conversely, in the female rats similar plots (Fig. 2a) showed a much higher incidence of HT relative to PCB for rats dosed with Aroclors 1242 or 1254, which produce elevated hepatic TEQ levels (Table 1). Again, this finding is consistent with the data for TCDD, which is a potent hepatotumorigen in S-D females (Kociba et al., 1978). However, Aroclors 1016 and 1260, both low contributors of hepatic TEQ (Table 1) induced significant HT in the females, albeit with very different HT versus PCB profiles (Fig. 2a).
Mathematical modeling showed that the lifetime incidence of HT (all tumors) in all Aroclor-dosed rats could be similarly correlated with PCB in both sexes and additionally with hepatic TEQ accumulation in females only (r2 = 0.89) (Brown et al., 1996). This showed that there must be at least two major processes by which PCBs contribute to HT in S-D rats, one that operates in both sexes and is dependent only on the accumulation of persistent PCB congeners (which are > 98% noncoplanar) and one that is TEQ dependent, but operates only in females.
Other components of the Aroclors were also evaluated, but neither the persistent mono-ortho-chlorinated congeners in liver, some of which express dioxin-equivalency, nor liver TEQ less the PCB126 component correlated with the liver tumor response. Also, tissue levels of PCB 153, which accumulates equally in both liver and adipose tissue, did not correlate with the liver tumor response. Finally, as shown by both mathematical modeling and by the lack of increased tumors in males dosed with Aroclors 1016 or 1242 (each > 90% metabolized in rats), the extent of PCB metabolism was not correlated with the development of liver tumors, implying that PCB metabolites did not contribute to HT.
Correlation between HT and Cytosol-Catalyzed Superoxide (O2•–) Production
The most striking finding from this study was the discovery that soluble low molecular weight components present in the stored liver cytosols from either Aroclor-dosed or undosed rats were readily (re)oxidized to species that could catalyze redox-cycling (i.e., permitting a reaction between NADPH and O2 to produce O2•–), and that this redox-cycling activity was consistently highly correlated with HT, regardless of time point, sex, Aroclor, or MFO system involved.
As illustrated in Figures 2a–d, in the continuously dosed groups the variations in potential for cytosolic O2•– production with accumulated tissue PCB followed the same pattern (albeit with different units) as the variations in HT for all Aroclors and both sexes. The magnitude of the responses (i.e., RCQ activity) was the same in individual rats that did not carry foci or tumors as in those animals that did, indicating that the increase in O2•– production potential was not a consequence of tumorigenesis. Figure 3 shows plots of the relationships between O2•– production potential and HT for all Aroclors and both sexes at all four time points with r2 values at 6, 12, 18, and 24 months of 0.93, 0.91, 0.95, and 0.89, respectively (all p < 0.0001). It is noteworthy that even though the relationships between HT and PCB accumulation were nonlinear, indicative of thresholds and saturations, and also Aroclor dependent and sexually dimorphic (Figs. 2a and 2b), those between HT and O2•– were linear, passed nearly through the origins, indicating direct proportionality above very low O2•– production thresholds, and held equally for all Aroclors and both sexes. In short, although different types of PCBs, different induced aromatic hydroxylases (i.e., MFOs), and different redox-cylers were likely involved in generating O2•–, once that O2•– (or some very close correlate thereof) was formed the subsequent steps to HT were no longer affected by those differences. It should also be noted that the correlation between late life tumor incidence (largely, adenoma), and relative O2•– formation was just as strong at 6 or 12 months, before tumors were detectable (Brunner et al., 1997), as it was later on after tumors appeared. Thus, formation of cytosolic redox-cycling activity, implying corresponding O2•– production, was an early predictor, as well as a close correlate, of liver tumor incidence.
Reversed phase chromatography of a pooled female rat cytosol fraction from Aroclor-dosed rats that had been isolated by adsorption on reversed phase and anion exchange columns revealed a set of four strong redox-active peaks that coeluted with those of authentic glutathionylated estrogen catechols of known elution sequence (Butterworth et al., 1996), prepared from the 2,3- and 3,4-estrogen catechols. The four matched HPLC peaks and proportions of matched peak areas were 2-hydroxy-1-glutathion-S-yl-17ß-estradiol (15%); 2-hydroxy-4-glutathion-S-yl-17ß-estradiol (24%); 2-hydroxy-1,4-bisglutathion-S-yl-17ß-estradiol (9%); and 4-hydroxy-1-glutathion-S-yl-17ß-estradiol (52%). However, the amounts of tissue samples available were insufficient for the determination of redox-cycling activity in each individual Aroclor/dose/sex group. Consequently, it was not possible to determine whether variations in estrogen catechol/quinone formation were correlated with HT.
As illustrated in Figure 4, increased Aroclor-induced O2•– production potential was, however, correlated with other responses, some of which have MOA implications. By way of comparison, Figures 4a and 4b, which are very similar to Figures 3a, 3b and 3c, 3d, respectively, again show the simple linear increase in HT with relative O2•– production at 6–12 months and 18–24 months Conversely, group mean serum alanine aminotransferase (ALT), one of the most specific and widely recognized indicators of liver cell damage, showed significant increases over the controls only at 24 months (data not shown), and then only in animals with hepatocellular carcinomas (Brunner et al., 1997). At 12 months, there was only a slight (insignificant) upward trend in mean ALT with increasing O2•– production (Fig. 4c). Mean hepatic levels of GPx, a major scavenger of H2O2, decreased with increased O2•–/H2O2 production only slightly in the Aroclor-dosed males, while in the females it was decreased significantly at 12 months and still more so at 24 months (Fig. 4d), at which time GPx levels in a few individual rats in the highest O2•–-producing groups had dropped nearly to zero. The incidence of mitosis in normal hepatocytes, as indicated by the % of cells carrying the proliferating cell nuclear antigen (PCNA), at 12 months, although initially increasing in proportion to O2•– production, eventually reached a limiting value (Fig. 4e) (Whysner and Wang, 2001). Finally, in all dose groups except the three with the highest O2•– production and HT, the incidence of hepatic carcinoma (rather than adenoma) remained at control levels (Fig. 4f), indicating the absence of either carcinoma cell promotion or adenoma cell progression. However, in the three groups with the highest O2•–/H2O2 production (which included the animals lacking GPx), there were statistically significant increases in carcinoma (Table 1), suggesting that a threshold for tumor progression had been exceeded.
Use of PCB Congener Ratios to Indicate and Differentiate MFO Aromatic Hydroxylase Activities
The gas chromatograms of the PCBs retained in the tissues of Aroclor-dosed rats differed from those in the administered Aroclors in showing not only near-total losses of nonpersistent (i.e., rapidly metabolized) congeners, but also partial losses of more slowly metabolized PCB congeners (Brown, 1994, Fig. 1). The pharmacokinetic analysis detailed in the Supplemental Materials and briefly summarized in the Methods section showed that the Aroclor-normalized intercongener ratio minus one, R'(i/j), could be a measure of the relative metabolic rate of congener j and hence an indicator of the PCBox activity (i.e., aromatic hydroxylase activity), present. R'(i/j) values for a number of PCB congeners from the four Aroclors showed two limiting congener selection patterns, designated Type A and Type B, and some intermediate patterns, designated Type AB. For Type A, the relative metabolic rates were PCB 28 > 66 > 74 > 118 > 47, 85, 99 zero, indicating metabolism of only the mono-ortho–substituted congeners. For Type B, the relative metabolic rates were PCB 47 > 66 > 85 > 28 > 74 > 99 > 118, indicating somewhat greater susceptibilities for the di-ortho–substituted congeners. The intermediate congener susceptibility profiles characterizing the AB patterns are presented in Supplemental Table S-2. The Type A and B selection profiles, along with practical considerations of GC peak measurability, indicated several peak ratios that could be used as practical measures of either Type A (or A-like Type AB) or of Type B (or B-like Type AB) activities as shown in Supplemental Table S-1.
On the bases of the observed type of congener selection profile (i.e., Type A, Type B, or Type AB), whether induced constitutively (c), by tissue PCB (p), or by tissue TEQ (t), whether expressed in adult males (m), adult females (f), or senescent animals (s), it was possible to distinguish six types of PCBox activities and name them according to the designations noted above. These six PCBox activities are presented in Table 2 along with the Aroclor and dose groups in which activity (induction and repression) was observed, approximate EC50 for induction, and references to illustrative figures. The characteristics of the individual PCBox activities are described in greater detail in the Supplemental Materials.
|
The CYP1A1, EROD, and MROD responses were 50–100% greater in males than in females (Supplemental Table S-3) but correlated with the PCBox-Amft indicator, R'(99/118) in both sexes (Fig. 5b), although with HT and O2•–production only in the females. The covariance with the CYP1A2 indicator, MROD, indicated that the active MFO for aromatic hydroxylation could have been either CYP1A1, 1A2, or both. CYP1B1 could not be detected in the livers of the chronically dosed rats (Silkworth et al., 1999
). There was little CYP1A1 or EROD induction in the 1016-dosed males, and none in the females, indicating that the PCBox activities with Type AB patterns did not come simply from mixtures of MFOs with Type A and B activities.
The CYP2B1/2, PROD, and BROD responses (Supplemental Table S-3) were about 50% higher in the males and exhibited very low EC50 values and saturation points (Fig. 5c); but were not correlated with PCBox activities, O2•–, or HT in either sex. Since the major CYPs induced by PCBs are known to be of the 2B and 3A families (Bandiera, 2001), this implicates the latter as mediators of PCBox-Bmp activity and possibly the PCBox-ABfp activity as well. The rodent CYP3A MFOs are known to be repressed by agonists for CYP1A1 induction (Lee et al., 2006; Shaban et al., 2005) as was observed for the PCBox-ABfc indicators R'(99/47), R'(99/85), and R'(153/128) in female rats dosed with Aroclors 1242 or 1254 (Fig. 5d). The major constitutive CYPs in male and female rat livers are known to be CYP2C11 and CYP2C12, respectively (Bandiera, 2001), so these likely contributed to the PCBox-ABmc and PCBox-ABfc activities, respectively. The CYP2C11 is known to be repressed by PCBs (Ngui and Bandiera, 1999) just as was PCBox-ABmc (Table 2, Fig. 6a).
To check on the validity of R'(i/j) ratios as measures of relative MFO activity the mean midlife female values for the Type A indicator R'(99/118) were compared with those for hepatic CYP1A1 and showed the agreement illustrated in Figure 5b. However, it should be noted that R‘(i/j) values become close measures of 
km/kn only under steady-state conditions. Such conditions were probably met for the tri-, tetra-, and pentachlorobiphenyls in the tissue residues from Aroclors 1016, 1242, and 1254, but a midlife steady state was less certain for Aroclor 1260 (Fig. 1) implying that the hexachlorobiphenyl-based R'(153/128) values might underestimate the actual PCBox- Type B activities. Notable, however, is that the seven-congener selection profile deduced from Type B R' values in rats (Supplemental Table S-2) was identical to that indicated by absolute rates of Type B congener clearance (Brown, 1994).
Correlation of MFO Activities with Redox-Cycling Activities and Hepatotumorigenesis
A striking feature of Figure 2 is that the HT versus PCB and O2•– versus PCB profiles, while very similar to each other for any one Aroclor-sex combination, varied considerably from one Aroclor-sex combination to another. As summarized in Table 2, the data show that these variations arose because of differences between the sets of MFOs (indicated by PCBox activities) that were induced or repressed by the particular combinations of PCB and TEQ accumulations from the different Aroclors, as well as the sex-dependent responses to these accumulations.
In the males, where TEQ-driven induction of CYP1A1/2 and PCBox-Amft activity does not contribute to hepatic RCQ/ROS formation, there was, for all Aroclors, an approximately linear-through-zero increase in PCBox-Bmp activity with increasing tissue PCB, and parallel contributions to O2•– production potential (Fig. 5f). However, there was also a PCB-dependent decrease in constitutive PCBox-ABmc activity, with a corresponding decrease in its contribution to O2•– production potential (Figs. 5f and 6a), so that the net O2•–production potential remained nearly constant (Figs. 2d, 5f, and 6a) until no further losses of the constitutive activity could occur. This resulted in a PCB accumulation threshold ( 500 ppm), above which O2•– rose in parallel with the PCBox-Bmp indicator, R'(99/85) (Fig. 5f).
Conversely, in the females, where TEQ-driven induction of CYP1A1/2 and PCBox-Amft activity does produce elevation in O2•– and HT (Figs. 2a and 2c; Table 1), this RCQ/ROS source became the dominant one for animals dosed with the high-TEQ Aroclors 1242 and 1254. With these Aroclors, the HT versus PCB and O2•– versus PCB showed a response curve (Figs. 2a and 2c) that was congruent with that for CYP1A1 versus TEQ (Fig. 5a), indicating HT-mediation by CYP1A1/2.
In the adult, but not the prepubertal, female controls there was appreciable PCBox-ABfc activity that was also PCB-repressed (Figs. 6b and 6d) by Aroclor 1016. However, this appeared to be less productive of RCQs than the male PCBox-ABmc activity, so that production of O2•– and HT was seven- to eightfold lower than in the male controls (Fig. 2, Table 1 and Supplemental Table S-3). Thus, with the increasing PCB and PCBox-ABfp activity induced by Aroclor 1016 (Figs. 6b and 6d) or Aroclor 1260 (Fig. 5d) the repressive effects of PCB on PCB-ABfc–mediated O2•– production were much sooner exhausted, and an accumulation threshold for increased O2•– production and HT was reached at lipid PCB levels around 50 ppm (Figs. 2a, 2c, and 6b) rather than around 500 ppm (Figs. 2b, 2d, and 5f).
In the Aroclor 1260–dosed females the mean midlife (6–18 months) responses exhibited by the PCBox Type B indicators R'(99/47), R'(99/85) and R'(153/128) were considerably lower than in the males (Figs. 5d versus 5f; Supplemental Table S-3), and also lower in those dosed with the CYP1A1-inducing Aroclor 1242 and 1254 than in those dosed with 1016 or 1266 (Fig. 5d). The latter findings were in line with previously noted reports that CYP1A1 agonists cause CYP3A repression in rodents (Lee et al., 2006; Shaban et al., 2005). However, by the 24 months time point about 30% of the more heavily PCB-burdened females had shown large (up to 20-fold) increases in both Type A and B PCBox activities so that the mean values of both (Supplemental Table S-3) became in line with the productions of O2·- and HT (Figs. 5d and 5e).
While no determinations of MFO activity in extrahepatic tissue were made, we did observe that in the males the repressions of pituitary, pancreatic, prostate, adrenal, and total extrahepatic tumors were all statistically significant and correlated with tissue TEQ but not PCB (Supplemental Table S-4), while in the females the repression of pituitary, mammary, and total extrahepatic tumors were statistically significant and correlated with both tissue TEQ and PCB (Supplemental Table S-5). Thus, the PCBs could produce not only bimodal effects on MFOs—inducing some while repressing others, as shown in Table 2, with resultant nonlinear dose–response relationships and accumulation thresholds for HT—but also simultaneous carcinogenic and anticarcinogenic effects at other sites.
In summary, the data for all Aroclors, doses, and sexes showed substantial overall correlations with hepatic RCQ activity production for the two major Aroclor-induced MFO activities, i.e., PCBox-Bmp and PCBox-Amft, in the S-D males and females, respectively. Examination of the data for PCB congener selection patterns showed, however, that additional types of MFO activity, some PCB-induced and some PCB-repressed (Table 2) were also present and correlated with RCQ production (Figs. 5 and 6), especially in the animals dosed with Aroclor 1016. As shown in Figures 2a–d, the relationships between RCQ production (i.e., cytosolic O2•–) or HT and PCB were complex, sex specific, and Aroclor dependent. For each of the four sex-specific relationships between PCB and HT or cytosolic O2•–-producing (i.e., RCQ) activity in Aroclor-dosed rats (Fig. 2) there were congruent relationships between one or more PCBox activities and either HT or RCQ. Thus, the Aroclor and sex variations in HT or RCQ versus PCB or TEQ in Figure 2 were correlated with similar variations in different types of PCB-modulated MFO activity as shown in Figures 5 and 6. In short, the forms of the dose–response patterns shown in Figures 2a–d for both HT and cytosolic O2•–-production (RCQ) can be explained by the corresponding dose–response patterns for multiple PCB-modulated MFOs.
Other Correlates and Noncorrelates of HT
The P450 cytochromes are known to have some NADPH oxidase (NOX) activity (producing O2•– and H2O2) that is generally attributed to futile cycling. Aroclor-dosed rat liver microsomes, which would contain most, if not all, the P450s, were indeed found to have such activity, with total ROS (mostly H2O2) production rates ranging up to 25 nmol/min/mg liver protein, as compared to rates ranging up to 400 nmol/min/mg protein for cytosol-catalyzed O2•– production (Fig. 2c). The microsomal ROS production appeared largely associated with CYP1A1, and therefore correlated with HT in the females, but not the males. This suggested that futile cycling by PCB-induced P450s did not make more than a very minor contribution to overall O2•–/H2O2 production and HT in Aroclor-dosed rats.
Cytosolic superoxide (Cu, Zn-SOD) levels (in units/mg protein) at 24 months were 0.65 in the male controls and ranged from 0.51 to 3.48 in the Aroclor-dosed males, but were not correlated with HT. In the females, the corresponding levels were 1.21 and 1.40–3.64 with weak (r2 = 0.54) correlation with HT. However, in a regression of HT against both cytosolic SOD and cytosolic O2•–, the degree of correlation (r2 = 0.89) was not altered by inclusion of the SOD term. Thus, although SOD-mediated conversion of O2•– to H2O2 may be an important step in the tumorigenic signaling cascade, there was enough SOD activity in the rats so that the O2•– conversion was not a rate-limiting step.
There was no clear evidence of a significant PCB effect on apoptosis as measured by TUNEL-stainable cells at 6 or 12 months and consequently no correlation with HT (Whysner and Wang, 2001). The Phase 2 enzymes GST, QR, and GR were generally only moderately (< twofold) increased in the Aroclor-dosed groups. The inductions were significantly correlated with HT in both sexes (r2 = 0.63–0.84), especially at the 24 months time point; however, the limited extent of the inductions and the late timing for the correlations argued against any sizeable contribution to the rate of tumor development.
-Tocopherol levels, measured only in the 100 ppm dose groups at 18 months, were unaffected by Aroclors 1016, 1242, 1254, or 1260 in the males, but roughly doubled in the females dosed with Aroclors 1242, 1254, or 1260 and increased two- to fivefold in those dosed with Aroclor 1016, suggesting overcompensation for oxidative stress. TBARS were quite variable, but averaged about three times the controls in the Aroclor-dosed females, and less than that in the males, with no obvious correlations with Aroclor type, dose, or HT in either sex. COMT activity, measured only at 12 months, was not significantly altered in the Aroclor-dosed animals. Finally, it should be noted that the progressive accumulation of hepatic nonheme, acid-soluble, ferric iron (presumably, as hemosiderin) reported by Whysner and Wang (2001) showed a pattern of sex and Aroclor dependency very similar to that for ROS production, suggesting a dependency on that process.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
As described in the introductory section, a number of biochemical responses have been hypothesized as involved in the tumorigenic process. However, the present study is the first ever to survey such hypothesized responses for their correlation with late life tumorigenesis. This survey revealed correlations of total HT (mainly adenomas) with tissue
PCB/TEQ accumulations, activities of at least six types of MFOs, and the production of RCQ/ROS activity. Significant increases in malignant hepatic tumors (i.e., carcinomas) occurred only at the highest levels of the latter activity, while extrahepatic tumors were TEQ suppressed. Responses found not correlated with HT included futile cycling by microsomes, formation of PCB metabolites, and indicators of cytotoxicity. The tumorigenesis-correlated responses and their interconnection to define a biologically plausible multistep MOA for the tumorigenic process are discussed below.
PCB and TEQ Accumulation in Aroclor-Dosed Rats
The four Aroclors used in this study differed widely in their contents of individual PCB congeners (Frame et al., 1996a,b) which differed widely in bioaccumulation potential (Brown, 1994) and in TEQ (Van den Berg et al., 2006). In the chronically dosed animals tissue PCB accumulations were 10- to 20-fold higher in animals dosed with Aroclors 1254 or 1260 than in those given the same doses of Aroclors 1016 or 1242 (Table 1, Fig. 1). The differences in TEQ accumulation covered an even wider range, but in a different pattern namely, 1254 > 1242 > 1260 > 1016 zero (Table 1). Total HT was found correlated with above accumulation threshold levels of tissue PCB in both sexes and additionally with tissue TEQ in females (Table 1, Fig. 2). The lipid-normalized tissue PCB accumulation thresholds for increased HT were about 50 ppm in the females and 500 ppm in the males (Figs. 5f and 6b). The suppression of extrahepatic tumors (e.g., those of the mammary, prostate, pancreas, pituitary, and adrenals) was associated with tissue TEQ accumulations in the males and with both TEQ and PCB in the females (Table 1, Supplemental Tables S-4, S-5).
Relevant Characteristics of Rat Liver MFOs
The rat genome is known to contain 88 putatively functional P450 genes (Nelson et al., 2004). In adult rat liver, about a dozen constitutively active P450 cytochromes with MFO activity have thus far been identified, with the expression of most being sexually dimorphic. These constitutive MFOs include the male-specific CYP2A2, CYP2C11 (the major CYP present in male rats), CYP2C13 and CYP3A2; the female-specific CYP2C12; and the female-predominant CYP2A1, CYP2C6, and CYP2C7 (Bandiera, 2001). It is also well established that the sexual dimorphism in adult rat liver MFO expression is determined by the pattern of growth hormone (GH) release from the pituitary (Agrawal and Shapiro, 2000; Shapiro et al., 1995; Waxman and O'Connor, 2006). This GH and sex hormone signaling stimulates formation of nuclear factors having transcriptional activities like those of the ligand-activated nuclear receptors (Waxman and O'Connor, 2006). In general, in male rats GH is secreted from the pituitary in episodic bursts while in females GH pulses are more frequent and irregular and of lower amplitude than in males. Thus, while expression of the female-specific CYP2C12 requires continuous presence of GH, the male-specific isoforms (i.e., CYP2A2, CYP2C11, CYP2C13, and CYP3A2) are repressed when GH is present continuously, but induced when it is not. The sexually dimorphic nature of constitutive MFO expression may explain the marked differences between HT in undosed male and female S-D rats (Table 1).
Some of the sex-specific constitutive MFOs, notably CYP2C11 and CYP3A2, are already known to be downregulated by persistent, lipophilic xenobiotics such as the PCBs and DDT (Kushida et al., 2005; Ngui and Bandiera, 1999; Okey, 1990). In addition, there is another large and well-studied group of rat liver MFOs, e.g., CYPs 1A1, 1A2, 1B1, 2B1, 2B2, and 3A1 that are upregulated by a wide variety of persistent lipophilic xenobiotics, including the Aroclors, all persistent PCB congeners, DDT, and the dioxins (Bandiera, 2001; Nebert, 1994, Nebert et al., 2000; Parkinson et al., 1983).
The present findings regarding the six groups of PCB-induced or constitutive PCBox activities (Table 2) are in agreement with previous reports of CYP protein subtype expression in several important ways. First, in both sexes, some constitutive MFO activities (present in undosed controls), and even some PCB-induced activities, were sex specific, suggesting regulation by GH. Second, in both sexes, there were indications that some of these constitutive MFO activities were downregulated by tissue PCB (Figs. 6a and 6b). Third, in both sexes, other types of MFO (i.e., as indicated by CYP1A1 as well as various PCBox subtypes) activities were both induced and repressed by tissue PCB or TEQ (Table 2, Fig. 5).
Although it was possible to establish correlations between redox-cycling metabolite formation and various subtypes of PCB oxidase activity, it was not possible to determine the relative importance of the various coinduced CYPs as contributors to PCBox activity. Nevertheless, on the basis of inducer type (i.e., constitutive vs. PCB vs. TEQ) sex selectivity, and previously reported CYP induction patterns (Bandiera, 2001), it is plausible that the contributors to the PCBox-Amft activity included CYP1A1 and/or CYP1A2; those to the male-specific constitutive PCBox-ABmc included CYP2C11 and/or CYP3A2; to the female-specific constitutive PCBox-ABfc, CYP2C12; and to PCBox-Bmp, CYP3A1, and/or CYP3A2. The correlations of these various PCBox subtypes with the generation of RCQ (O2•– production potential) and HT are illustrated in Figures 5 and 6. These PCBox activities are of particular interest because they provide an integrated measure of the same type of action (i.e., aromatic hydroxylation) as that required for the conversion of an ambient aromatic (endogenous or dietary) to an RCQ (or other quinone).
Oxidative Stress as a Mediator of PCB-Induced Tumor Promotion
An association between tumor promotion (which depends upon mitotic stimulation and/or GJIC inhibition) and "oxidative stress" (i.e., formation of ROS) has been long recognized and extensively reviewed (Athar, 2002; Cerutti, 1985; Gopalakrishna and Jaken, 2000; Kensler and Trush, 1984; Klaunig et al., 1995, 1998; Stohs, 1990). More recently, both the mitotic stimulation produced by a variety of polypeptide growth factors (Finkel, 2003; Rhee et al., 2003) as well as GJIC inhibition (Upham et al., 1997) have been shown to be mediated by H2O2 (or its precursor, O2•–). H2O2 performs this intracellular "second messenger" function by selectively (and reversibly) oxidizing low pka thiol groups on signaling proteins, e.g., the protein tyrosine phosphatases (Finkel, 2003; Rhee et al., 2003) and protein tyrosine kinases (Nakashima et al., 2005) that mediate mitosis. The conversion of the initially produced O2•– to H2O2 is mediated by SOD. In PCB-dosed rats, total SOD activity was found to be moderately elevated and weakly HT correlated, but not an additional contributor to HT risk in a bivariant regression. This indicated that the O2•– 
H2O2 conversion was not a rate-limiting step in ROS-mediated tumor promotion.
Several routes to intracellular O2•–/H2O2 formation have been demonstrated or suggested. For example, polypeptide growth factors (e.g., epidermal growth factor [EGF], nerve growth factor [NGF], platelet-derived growth factor [PDGF]) can bind to and activate ligand-specific cell membrane receptor kinases, and thereby activate phospholipase C (PLC) and protein kinase C (PKC) (Gomperts et al., 2002b). Noncoplanar PCBs in tissue culture have been shown to activate less specific cell membrane receptors (e.g., ryanodine receptor) which can then induce calcium ion influx leading to activation of PLC and PKC (Gomperts et al., 2002a; Machala et al., 2003). In either scenario, PKC activation would be followed by phosphorylation/activation of NOX and finally NOX-mediated O2 reduction to O2•– and H2O2 (Gopalakrishna and Jaken, 2000; Lambeth 2004; Rhee et al., 2003). Alternatively, a wide range of strongly to weakly lipophilic organic chemicals, including PCBs, chlorinated pesticides, aromatic hydrocarbons, many drugs, cigarette tar, and even alcohol, are capable of activating nuclear receptors (e.g., AhR, constitutive androstane receptor [CAR], or pregnane X receptor [PXR]) to induce P450 cytochromes (Bandiera, 2001; Nebert, 1994). These P450s can attack a variety of endogenous biochemical species, such as estrogen (Butterworth et al., 1998; Liehr and Roy, 1990) or other aromatics (Bolton et al., 2000; Nebert et al., 2000) to produce ROM including quinones. Nebert et al. (2000) hypothesized that ROS formation would then occur via either futile cycling by the P450s or redox-cycling by the quinones. Finally, in adult rodent liver GH produces not only induction of sex-specific P450s, as noted above, but also induction and release of insulin-like growth factor 1, another polypeptide mitogen. Thus, GH has the potential for stimulating ROS production via either the MFO/RCQ- or NOX-mediated routes.
In the past, the source of whatever ROS contributed to tumorigenesis was usually attributed to futile cycling at mitochondrial or endoplasmic membranes (Nebert et al. 2000). In this regard, some H2O2 and O2•– production was observed from our test rat hepatic microsomes (mixtures of mitochondrial, endoplasmic, and plasma membrane fragments) when treated with NADPH and O2. However, substantially more O2•– equivalents were produced when membrane-filtered and reoxidized test rat hepatic cytosols were treated with NADPH and O2 in the presence of hepatic microsomes from PB-dosed rats. This RCQ-attributed ROS production was highly correlated with HT over the entire range of experimental variables examined, e.g., in both sexes, in Aroclor-dosed as well as in undosed rats, in conjunction with a number of different PCB-induced (and repressed) MFOs, and for all time points examined (Fig. 3). This strongly suggests that rat liver tumor cell proliferation, whether spontaneous or Aroclor-induced, is ROS mediated, just as is mitogen-induced cell proliferation in normal cells (Finkel, 2003; Rhee et al., 2003, 2005). However, the source of the ROS formation appears to be cytosolic ROM having the chemical characteristics of RCQs.
RCQs are known to be species that can undergo one-electron reduction by the NADPH-reduced forms of membrane flavoproteins, including P450 reductase (Iyanagi, 1990), thereby producing ion-radicals (semiquinones) that can transfer the electron to molecular oxygen (O2), thus producing superoxide (O2•–) and regenerating the quinone (redox-cycling) (Bolton, 2000; Iyanagi, 1990; O'Brien, 1991). Formation of quinones having redox-cycling activity has been demonstrated to result from the metabolism of estrogen, hydroquinones and other endogenous and exogenous aromatics (Bolton et al., 2000; Butterworth et al., 1998; Liehr and Roy, 1990). A striking finding of the present study, as well as an earlier study with TCDD in S-D rats (Kociba et al., 1978) was the enhanced RCQ and/or HT production in the females, which was correlated with tissue levels of TEQ and CYP1A1. The usual substrates for the CYP1 MFOs are near-coplanar lipophiles (Juchau, 1990). Consequently, since the only near-coplanar oxidizable aromatic present in females but not males is estrogen, a known RCQ precursor, it is plausible that the sexual dimorphism in the S-D rat hepatotumorigenic response was due to estrogen metabolites, such as the glutathionylated estrogen catechols (readily reoxidized to quinones), which were detected in the hepatic cytosols. However, since it was not possible to quantify the contribution of estrogen-derived RCQs to the total RCQ activity observed, and because there was sex specificity in PCB-mediated MFO induction (Table 2) it was not possible to establish whether the sex specificity of the TEQ-mediated O2•–/HT response in the S-D females arose mainly from a female-specific RCQ precursor, or a strain- and female-specific induced MFO, or both. Also notable is that Aroclors 1016 and 1242, which are both extensively metabolized to oxygenated products (Safe, 1994) of types shown to produce ROS in vitro (McLean et al., 2000; Srinivasan et al., 2001) nevertheless produced little or no increase in O2•– production (Figs. 5f and 6a) or HT (Table 1) over the controls in male S-D rats. This suggests that the substrates for RCQ formation in the PCB-dosed males must also have been endogenous aromatics rather than PCB metabolites.
Tumor Promotion versus Tumor Repression
It is well established that Aroclors and related substances (e.g., dioxins), like almost all other rodent carcinogens (Gray et al., 2002; Haseman and Johnson, 1996; Linkov et al., 1998) can produce either increases, decreases, or no effects on rat tumor incidence, depending upon organ, sex, and strain. Thus, dioxin (TCDD) in female S-D rats decreased the incidence of tumors of the pituitary, uterus, mammary gland, pancreas, and adrenal medulla, while increasing those of the liver, lung, tongue, and hard palate/nasal turbinates (Kociba et al., 1978). Conversely, in similarly treated male S-D rats none of the latter increases occurred (Kociba et al., 1978). In male Wistar rats Clophen A60 (equivalent to Aroclor 1260) increased tumors of the liver, while decreasing the incidence of thymomas and all other neoplasias (mainly those of the skin, testes, and pituitary) (Schaeffer et al., 1984). Clophen A30, which has a composition intermediate between those of Aroclors 1016 and 1242, in the Wistar males produced only nonsignificant increases in liver tumors, but was protective against thymomas and other tumors with effects comparable to those produced by Clophen A60 (Schaeffer et al., 1984). Aroclor 1260 in female Sherman rats produced increased incidence of liver tumors along with decreased incidence of mammary tumors (Kimbrough et al., 1975). Aroclor 1254, despite producing a 56% liver tumor incidence in S-D females (Mayes et al., 1998), produced no significant liver tumor increases in F344 females (NCI, 1977).
In the present S-D rat study, the more heavily Aroclor-burdened animals showed increases in hepatic tumors, but significant decreases in several types of extrahepatic tumors, e.g., those of the pituitary, mammary gland, adrenals, prostate, and pancreas (Table 1, Supplemental Tables S-4, S-5). The hepatic tumor increases correlated with tissue PCB, but not TEQ in the males, but with both PCB and TEQ in the females. Conversely, the extrahepatic tumor decreases correlated with tissue TEQ, but not PCB in the males, and again with both TEQ and PCB in the females. Thus, tumorigenic outcome depended on organ, sex, and Aroclor component in complementary patterns.
The interstrain, intersex, and interorgan differences in response to Aroclors are easily rationalized, since it is known that there are substantial interspecies, interstrain, intersex, and interorgan differences in the induction of Phase 1 and Phase 2 drug/xenobiotic-metabolizing enzymes, and consequently in the production of reactive metabolites (Bandiera, 2001; Conney, 1967; Creel et al., 1976; Nebert et al., 1982; Okey, 1990), which are needed for the MFO–RCQ–ROS mitotic signaling. However, there remains the challenge of explaining how Aroclor components can inhibit as well as induce tumorigenesis.
The decreases in extrahepatic tumors in the more heavily Aroclor-burdened rats of the present study (Table 1, Supplemental Tables S-4, S-5) as well as the decreases in constitutive signaling for HT at low Aroclor 1016 accumulations (Table 2, Fig. 6) occurred without decreases in food intake (as occurs in dioxin-dosed rats) (Table 1 and Brunner et al., 1997), indicating that they could not be attributed to reduced nutrition. Additionally, none of the four parameters indicative of cytotoxicity, i.e., liver cell apoptosis, necrosis, fibrosis, or elevated serum ALT (Fig. 4e) was found significantly increased by any Aroclor in either sex (Brunner et al., 1997; Whysner and Wang, 2001). This indicated that neither cell death nor stimulation of regenerative proliferation was a likely contributor to either the inhibition or promotion of tumorigenesis.
Of greater potential relevance are recent reports that AhR agonists can decrease rodent expression of MFOs of the CYP 2B, 2C, 2D, and 3A families, which are mediated by the CAR and the PXR (Lee et al., 2006; Shaban et al., 2005). It has also been found that at least three other steroid receptors of the same basic helix-loop-helix (bHLH)/Per-Arnt-Sim (PAS) family, namely the estrogen receptors ER- and ER-ß and the androgen receptor AR, are downregulated by AhR agonists, and that this occurs via the liganded AhR-mediated formation of a multiprotein complex including culin 4B (CUL4B) that catalyzes the ubiquitination of the steroid receptors, thus leading to their proteosome-mediated degradation (Ohtake et al., 2007). Thus, it is plausible that PCB congeners with TEQ activity, acting through the AhR, can reduce levels of ER-, ER-ß, AR, CAR, and PXR, thereby reducing the tumor promoting (mitotic) activity of both exogenous PCB and endogeneous sex hormones. The latter would explain why the TEQ-mediated suppression of extrahepatic tumorigenesis is particularly effective on hormone-dependent tumors, such as those of the mammary and prostate (Supplemental Tables S-4, S-5). Furthermore, the fact that the CAR/PXR agonist PCB 153 can downregulate AhR-mediated responses (Chen and Bunce, 2004) and that the AhR is also in the bHLH/PAS nuclear receptor family and also degraded by the ubiquitin/proteosome pathway (Roberts and Whitelaw, 1999) suggests that PCB accumulation, acting through the CAR or PXR receptors, might downregulate AhR levels and AhR-mediated tumorigenesis by a similar culin/ubiquitination pathway. In short, liganded nuclear receptor–mediated degradation of other nuclear receptors or factors may provide a general explanation for the antitumorigenic effects of rodent tumorigens.
These mechanistic considerations suggest that the net tumorigenic response to PCBs in any one organ will depend upon the distribution of nuclear receptors constitutively present therein, the availabilities of RCQ precursors, and the agonist activities of the administered dose. They also predict multimodal dose–response curves, with possibilities of hormesis and/or thresholds and/or high dose saturation, since the same ligand-induced receptor activation leads to both induction of new MFOs and degradation of the nuclear receptors/factors responsible for the constitutive MFOs; however, the possible extent of each process is limited by the amounts of receptors originally present in the target involved.
Intracellular Signaling Pathways and Rat Liver Tumorigenesis
The above-discussed relationships between Aroclor administration and tissue PCB/TEQ accumulation, between PCB/TEQ accumulation and modulation of MFO activity, between MFO activity and production of RCQ/ROS, and between RCQ/ROS production and tumor promotion, in addition to the observed correlations of HT with PCB/TEQ, MFO activities, and RCQ/ROS activities, indicate the two similar types of multistep hepatotumorigenic signaling pathways diagrammed in Figure 7. In the first type of signaling cascade, which dominates in undosed or less heavily PCB-dosed (but still ad libitum fed) animals, the deduced sequence of mediators is (1) pituitary-derived GH, (2) GH-induced but PCB/TEQ-repressed "constitutive" MFOs, (3) RCQs formed by MFO action on unidentified endogenous aromatics (4) O2•–, and (5) H2O2 (Fig. 7, left side). In the other type of signaling cascade, which dominates in the more heavily PCB-dosed animals, the postulated sequence of mediators is (1) administered Aroclor with subsequent PCB/TEQ tissue accumulation, (2) PCB/TEQ-induced MFOs, (3) generation of RCQs, formed by MFO-mediated metabolism of estrogen or other suitable substrates, (4) O2•–, and (5) H2O2 (Fig. 7, right side). In either case, the effect of the H2O2 produced is to stimulate the proliferation of spontaneously initiated cells to produce adenomas (Fig. 7, bottom). Thus, as tissue PCB and/or TEQ levels increase, the O2•– H2O2 mitotic signal produced by the first sequence of mediators will decrease and that produced by the second sequence will increase. As a result of these competing responses, the net mitotic signal will remain near-constant until there is no longer any more constitutive MFO to be repressed. At that point a PCB accumulation threshold will be exceeded, leading to increased tumorigenesis. At very high tissue PCB/TEQ levels, leading to excessive ROS production, normal H2O2 controls by catalase and GPx may be overwhelmed, resulting in production of hydroxyl radicals (•OH), DNA damage, oncogene mutation, and progression to carcinoma (Fig. 7, lower right), as was observed (Fig. 4f).
|
In summary, an examination of the biochemical correlates of tumor development in PCB-dosed rats has produced evidence for a novel, epigenetic, carcinogenic process, where tumor growth and development is controlled by the net activity of multiple MFO–RCQ–ROS mitotic signaling cascades. This MOA offers a rational and biologically plausible explanation for (1) the observed interstrain, intersex, interorgan, and inter-Aroclor differences in the tumorigenic responses; (2) the parallel occurrence of tumor promotion and tumor repression; (3) the variety of HT versus
PCB/TEQ relationships observed; (4) the dosage accumulation thresholds; (5) the correlations of HT with tissue PCB/TEQ and MFO induction; and (6) the unusually close and consistent correlation of HT with earlier RCQ/ROS activity (Fig. 3). |
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Supplemental data are available online at www.toxsci.oxfordjournals.org.
|
FUNDING |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
General Electric Company.
|
ACKNOWLEDGMENTS |
|---|
The authors are indebted to Dr Kenneth M. Fish, Joanne Smith, and Eric Williams for the biochemical measurements reported here.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Agency for Toxic Substances and Disease Registry (ATSDR). In: Toxicological Profile For Polychlorinated Biphenyls (PCBs) (2000) U.S. Department Of Health And Human Services, Public Health Service.
Agrawal AK, Shapiro BH. Differential expression of gender-dependent hepatic isoforms of cytochrome P-450 by pulse signals in the circulating masculine episodic growth hormone profile of the rat. J. Pharmacol. Exp. Ther. (2000) 292(1):228–237.
Ames BN, Gold LS. Chemical carcinogenesis: Too many rodent carcinogens. Proc. Natl. Acad. Sci. USA (1990) 87:7772–7776.
Arif JM, Lehmler HJ, Robertson LW, Gupta RC. Interaction of benzoquinone- and hydroquinone derivatives of lower chlorinated biphenyls with DNA and nucleotides in vitro. Chem. Biol. Interact. (2003) 142(3):307–316.[CrossRef][Web of Science][Medline]
Athar M. Oxidative stress and experimental carcinogenesis. Ind. J. Exp. Biol. (2002) 40(6):656–667.[Medline]
Bachowski S, Xu Y, Stevenson DE, Walborg EF Jr, Klaunig JE. Role of oxidative stress in the selective toxicity of dieldrin in the mouse liver. Toxicol. Appl. Pharmacol. (1998) 50:301–309.
Bager Y, Kato Y, Kenne K, Warngard L. The ability to alter the gap junction protein expression outside GST-P positive foci in liver of rats was associated to the tumour promotion potency of different polychlorinated biphenyls. Chem. Biol. Interact. (1997) 103(3):199–212.[CrossRef][Web of Science][Medline]
Bandiera SM. Cytochrome P450 enzymes as biomarkers of PCB exposure and modulators of toxicity, In PCBs. In: Recent Advances in Environmental Toxicology and Health Effects—Robertson LW, Hansen LG, eds. (2001) Lexington, KY: University of Kentucky Press. 185–192.
Black MJ, Brandt RB. Spectrofluoroimetric analysis of hydrogen peroxide. Anal. Biochem. (1974) 58:246–254.[CrossRef][Web of Science][Medline]
Bolton JL, Trush MA, Penning TM, Dryhurst G, Monks TJ. Role of quinones in toxicology. Chem. Res. Toxicol. (2000) 13:135–160.[CrossRef][Web of Science][Medline]
Brown JF Jr. Determination of PCB metabolic, excretion and accumulation rates for use as indicators of biological responses and relative risk. Environ. Sci. Technol. (1994) 28:2295–2305.
Brown JF Jr, Fish KM, Mayes BA, Silkworth JB, Hamilton SB, Whysner J. PCB effects on epigenetic carcinogenic processes, In PCBs. In: Recent Advances in Environmental Toxicology and Health Effects—Robertson LW, Hansen LG, eds. (2001) Lexington, KY: University of Kentucky Press. 329–336.
Brown JF Jr, Silkworth JB, Mayes BA. Characterization of PCB composition, tissue accumulation, and correlations with tumorigenesis in chronically dosed male and female Sprague-Dawley rats. In: In Report on Peer Review Workshop on PCBs: Cancer-Dose Response Assessment and Application to Environmental Mixtures—Koller L, ed. (1996) Washington, D.C: U.S. Environmental Protection Agency, National Center for Environmental Assessment, Office of Research and Development. 35–42.
Brunner MJ, Sullivan TM, Singer AW, Ryan MJ, Toft JD, Menton RG, Graves SW, Peters AC. An assessment of the Chronic Toxicity and Oncogenicity of Aroclor–1016, Aroclor–1242, Aroclor–1254, and Aroclor–1260 Administered in Diet to Rats, Vols. I–XII. Study No. SC920192 (1997) Columbus, OH: Battelle Laboratory.
Buchmann A, Ziegler S, Wolf A, Robertson LW, Durham SK, Schwarz M. Effects of polychlorinated biphenyls in rat liver: Correlation between primary subcellular effects and promoting activity. Toxicol. Appl. Pharmacol. (1991) 111:454–468.[CrossRef][Web of Science][Medline]
Burke MD, Thompson S, Elscombe CR, Halpert J, Haaparanta T, Mayer RT. Ethoxy-, pentoxy- and benzyloxyphenoxazones and homologues: Series of substrates to distinguish between different induced cytochromes P–450. Biochem. Pharmacol. (1985) 34:3337–3345.[CrossRef][Web of Science][Medline]
Butterworth M, Lau SS, Monks TJ. 17 beta-Estradiol metabolism by hamster hepatic microsomes: Comparison of catechol estrogen O-methylation with catechol estrogen oxidation and glutathione conjugation. Chem. Res. Toxicol. (1996) 9(4):793–799.[CrossRef][Web of Science][Medline]
Butterworth M, Lau SS, Monks TJ. 2-Hydroxy-4-glutathion-S-yl-17beta-estradiol and 2-hydroxy-1-glutathion-S-yl-17beta-estradiol produce oxidative stress and renal toxicity in an animal model of 17beta-estradiol-mediated nephrocarcinogenicity. Carcinogenesis (1998) 19(1):133–139.
Cerutti PA. Prooxidant states and tumor promotion. Science (1985) 227:375–381.
Chen G, Bunce NJ. Interaction between halogenated aromatic compounds in the Ah receptor signal transduction pathway. Environ. Toxicol. (2004) 19:480–489.[CrossRef][Web of Science][Medline]
Conney AH. Pharmaceutical implications of microsomal enzyme induction. Pharmacol. Rev. (1967) 19:317–366.
Creel D, Shearer DE, Hall PF. Differences in cytochrome P-450 of various strains of rats following chronic administration of pentobarbital. Pharmacol. Biochem. Behav. (1976) 5(6):705–707.[CrossRef][Web of Science][Medline]
Ernster L. DT diaphorase. Methods Enzymol. (1967) 10:309–317.[Medline]
Finkel T. Oxidant signals and oxidative stress. Curr. Opin. Cell Biol. (2003) 15:247–254.[CrossRef][Web of Science][Medline]
Fish KM, Silkworth JB, Brown JF Jr. The induction of oxidizing agents in S-D rats chronically dosed with Aroclors 1016, 1242, 1254, and 1260. Organohalogen Compd. (1997) 34:296–300.
Frame GM, Cochran JW, Bowadt SS. Complete PCB congener distributions for 17 Aroclor mixtures determined by 3 HRCG systems optimized for comprehensive, quantitative, congener-specific analysis. J. High Resolut. Chromatogr. (1996b) 19:657–668.[CrossRef]
Frame GM, Wagner RE, Carnahan JC, Brown JF Jr, May RJ, Smullen LA, Bedard DL. Comprehensive, quantitative, congener-specific analyses of eight Aroclors and complete PCB congener assignments on DB-1 capillary GC columns. Chemosphere (1996a) 33(4):603–623.
Geller BL, Winge DR. Subcellular distribution of superoxide dismutases in rat liver. Methods Enzymol. (1984) 105:105–114.[Web of Science][Medline]
Glauert HP, Robertson LW, Silberhorn EM. PCBs and Tumor Promotion. In PCBs. In: Recent Advances in Environmental Toxicology and Health Effects—Robertson LW, Hansen LG, eds. (2001) Lexington, KY: University of Kentucky Press. 355–371.
Gomperts BD, Kramer LM, Tathem PER. Chapter 8. Calcium signaling. In: In Signal Transduction (2002a) London: Elsevier Academic Press. 171–187.
Gomperts BD, Kramer LM, Tatham PER. Chapter 11. Signaling pathways operated by receptor protein tyrosine kinases. In: In Signal Transduction (2002b) London: Elsevier Academic Press. 257–282.
Gopalakrishna R, Jaken S. Protein kinase C signaling and oxidative stress. Free Radic. Biol. Med. (2000) 28(9):1349–1361.[CrossRef][Web of Science][Medline]
Gray GM, Huang H, Linkov I, Polkanov M, Wilson R. The effect of different tumor groupings on findings of anticarcinogenic responses in long term rodent bioassays. Regul. Toxicol. Pharmacol (2002) 36(2):139–148. (Review).[CrossRef][Web of Science][Medline]
Guengerich FD. Analysis and characterization of enzymes. In: In Principles and Methods of Toxicology—Hayes AW, ed. (1994) New York: Raven Press. 1259.
Habig WH, Pabst MJ, Jakoby WB. Glutathione S-transferases. The first enzymatic step in mercapturic acid formation. J. Biol. Chem. (1974) 249:7130–7139.
Halliwell B, Gutteridge JMC. Free Radicals in Biology and Medicine (1999) 3rd ed. NY: Oxford University Press.
Hansen LG. The ortho Side of PCBs. Occurrence and Disposition (1999) Norwell, MA: Kluwer Academic Publishers.
Haseman JK, Johnson FM. Analysis of National Toxicology Program rodent bioassay data for anticarcinogenic effects. Mutat. Res. (1996) 350(1):131–141.[Web of Science][Medline]
Hassan HM. Determination of microbial damage caused by oxygen free radicals, and the protective role of superoxide dismutase. Methods Enzymol. (1984) 105:404–414.[Web of Science][Medline]
Hassoun EA, Li F, Abushaban A, Stohs SJ. Production of superoxide anion, lipid peroxidation, and DNA damage in the hepatic and brain tissues of rats after subchronic exposure to mixtures of TCDD and its congeners. J. Appl. Toxicol. (2001) 21(3):211–219.[CrossRef][Web of Science][Medline]
Iyanagi T. On the mechanism of 1-electron reduction of quinones by microsomal flavin enzymes: The kinetic analysis between cytochrome B5 and menadione. Free Radic. Res. Commun. (1990) 8(4–6):259–268.[Web of Science][Medline]
Juchau MR. Substrate specificities and functions of the P450 cytochromes. Life Sci. (1990) 47:2385–2394.[CrossRef][Web of Science][Medline]
Kensler TW, Trush MA. Role of oxygen radicals in tumor promotion. Environ. Mutagen. (1984) 6:593–616.[Web of Science][Medline]
Kimbrough RD, Squire RA, Strandberg JD, Montali RJ, Burse VW. Induction of liver tumors in Sherman strain female rats by PCB Aroclor 1260. J. Natl. Cancer Inst. (1975) 55(6):1453–1456.[Web of Science][Medline]
Klaunig JE, Xu Y, Bachowski S, Ketcham CA, Isenberg JS, Kolaja KL, Baker TK, Walborg EF Jr, Stevenson DE. Oxidative stress in nongenotoxic carcinogenesis. Toxicol. Lett. (1995) 82–83:683–691.[CrossRef]
Klaunig JE, Xu Y, Isenberg JS, Bachowski S, Kolaja KL, Jiang J, Stevenson DE, Walborg EF Jr. The role of oxidative stress in chemical carcinogenesis. Environ. Health Perspect. (1998) 106(Suppl. 1):289–295.[CrossRef][Web of Science][Medline]
Kociba RJ, Keyes DG, Beyer JE, Carreon RM, Wade CE, Dittenber DA, Kalnins RP, Frauson LE, Park CN, Barnard SD, et al. Results of a two-year chronic toxicity and oncogenicity study of 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin in rats. Toxicol. Appl. Pharmacol. (1978) 46:279–303.[CrossRef][Web of Science][Medline]
Kushida M, Sukata T, Uwagawa S, Ozaki K, Kinoshita A, Wanibuchi H, Morimura K, Okuno Y, Fukushima S. Low dose DDT inhibition of hepatocarcinogenesis initiated by diethylnitrosamine in male rats: Possible mechanisms. Toxicol. Appl. Pharmacol. (2005) 208(3):285–294.[CrossRef][Web of Science][Medline]
Kuthan H, Ullrich V, Estabrook RW. A quantitative test for superoxide radicals produced in biological systems. Biochem. J. (1982) 203:551–558.[Web of Science][Medline]
Lambeth JD. NOX enzymes and the biology of reactive oxygen. Nat. Rev. Immunol. (2004) 4(3):181–189.[CrossRef][Web of Science][Medline]
Lee AJ, Cai MX, Thomas PE, Conney AH, Zhu BT. Characterization of oxidative metabolites of 17-beta-estradiol and estrone formed by 15 selectively expressed human P450 isoforms. Endocrinology (2003) 144:3382–3398.
Lee C, Hutson JR, Tzau VK, Riddick DS. Regulation of constitutive mouse hepatic cytochromes P450 and growth hormone signaling components by 3-methylcholanthrene. Drug Metab. Dispos. (2006) 34:1530–1538.
Lehmann L, Esch H, Kirby P, Robertson LW, Ludewig G. 4-monochlorobiphenyl (PCB3) induces mutations in the livers of transgenic Fisher 344 rats. Carcinogenesis (2006) 28(2):471–478.[CrossRef][Web of Science][Medline]
Liehr JG, Roy D. Free radical generation by redox cycling of estrogens. Free Radic. Biol. Med. (1990) 8(4):415–423.[CrossRef][Web of Science][Medline]
Lin PH, Sangaiah R, Ranasinghe A, Upton PB, La DK, Gold A, Swenberg J. Formation of quinonoid-derived protein adducts in the liver and brain of Sprague-Dawley rats treated with 2,2',5,5'-tetrachlorobiphenyl. Chem. Res. Toxicol. (2000) 13(8):710–718.[CrossRef][Web of Science][Medline]
Linkov I, Wilson R, Gray GM. Anticarcinogenic responses in rodent cancer bioassays are not explained by random effects. Toxicol. Sci. (1998) 43(1):1–9.
Machala M, Blaha L, Vondracek J, Trosko JE, Scott J, Upham BL. Inhibition of gap junctional intercellular communication by noncoplanar polychlorinated biphenyls: Inhibitory potencies and screening for potential mode(s) of action. Toxicol. Sci. (2003) 76(1):102–111.
Mayes BA, Fish KM, Brown JF Jr, Silkworth JB. PCB tumorigenicity in rodents: How should it be measured? The Toxicologist (1999) 48:338.
Mayes BA, McConnell EE, Neal BH, Brunner MJ, Hamilton SB, Sullivan TM, Peters AC, Ryan MJ, Toft JD, Singer AW, et al. Comparative carcinogenicity in Sprague-Dawley rats of the polychlorinated biphenyl mixtures Aroclors 1016, 1242, 1254, and 1260. Toxicol. Sci. (1998) 41(1):62–76.
McLean MR, Twaroski TP, Robertson LW. Redox cycling of 2-(x'-mono, -di, and -trichlorophenyl)-1,4-benzoquinones, oxidation products of polychlorinated biphenyls. Arch. Biochem. Biophys. (2000) 376(2):449–455.[CrossRef][Web of Science][Medline]
Nakashima I, Takeda K, Kawamoto Y, Okuno Y, Kato M, Suzuki H. Redox control of catalytic activation of membrane-associated protein tyrosine kinases. Arch. Biochem. Biophys. (2005) 434:3–10.[CrossRef][Web of Science][Medline]
Nath RG, Randerath E, Randerath K. Short term effects of tumor-promoting polychlorinated biphenyl mixture, Aroclor 1254, on I-compounds in liver, kidney, and lung DNA of male Sprague-Dawley rats. Toxicology (1991) 68(3):275–289.[CrossRef][Web of Science][Medline]
National Cancer Institute (NCI). Bioassay of Aroclor 1254 for Possible Carcinogenicity. Carcinogenesis Technical Report Series No. 38, DHEW Publ. No. (NIH) 78-538. (1978) Bethesda, MD: U.S. Department of Health, Education and Welfare, National Institutes of Health.
Nebert DW. Drug-metabolizing enzymes in ligand-modulated transcription. Biochem. Pharmacol. (1994) 47:25–37.[CrossRef][Web of Science][Medline]
Nebert DW, Jensen NM, Shinozuka H, Kunz HW, Gill TJ 3rd. The Ah phenotype. Survey of forty-eight rat strains and twenty inbred mouse strains. Genetics (1982) 100(1):79–87.
Nebert DW, Roe AL, Dieter MZ, Solis WA, Yang Y, Dalton TP. Role of the aromatic hydrocarbon receptor and [Ah] gene battery in the oxidative stress response, cell cycle control, and apoptosis. Biochem. Pharmacol. (2000) 59(1):65–85.[CrossRef][Web of Science][Medline]
Nelson DR, Zeldin DC, Hoffman SM, Maltais LJ, Wain HM, Nebert DW. Comparison of cytochrome P450 (CYP) genes from the mouse and human genomes, including nomenclature recommendations for genes, pseudogenes and alternative splice variants. Pharmacogenetics (2004) 14(1):1–18.[CrossRef][Web of Science][Medline]
Ngui JS, Bandiera SM. Induction of CYP2B is a more sensitive indictor of exposure to Aroclor 1260 than CYP1A in male rats. Toxicol. Appl. Pharmacol. (1999) 161:160–170.[CrossRef][Web of Science][Medline]
Norback DH, Weltman RH. Polychlorinated biphenyl induction of hepatocellular carcinoma in the Sprague-Dawley rat. Environ. Health Perspect. (1986) 60:97–105.[CrossRef][Web of Science]
O'Brien PJ. Superoxide production. Methods Enzymol. (1984) 105:370–378.[Web of Science][Medline]
O'Brien PJ. Molecular mechanisms of quinone toxicity. Chem. Biol. Interact. (1991) 80:1–41.[CrossRef][Web of Science][Medline]
Ohtake F, Baba A, Takada I, Okada M, Iwasaki K, Miki H, Takahashi S, Kouzmenko A, Nohara K, Chiba T, et al. Dioxin receptor is a ligand-dependent E3 ubiquitin ligase. Nature (2007) 446:7135–566, 562.
Okey AB. Enzyme induction in the P-450 system. Pharmacol. Ther. (1990) 45:341–398.
Parkinson A, Robertson L, Safe LE, Safe S. Polychlorinated biphenyls as inducers of microsomal enzymes: Structure-activity rules. Chem. Biol. Interact. (1980) 30:271–285.[CrossRef][Web of Science][Medline]
Parkinson A, Safe SH, Robertson LW, Thomas PE, Ryan DE, Reik LM, Levin W. Immunochemical quantitation of cytochrome P-450 isozymes and epoxide hydrolase in liver microsomes from polychlorinated or polybrominated biphenyl-treated rats. A study of structure-activity relationships. J. Biol. Chem. (1983) 258(9):5967–5976.
Pereg D, Tampal N, Espandiari P, Robertson LW. Distribution and macromolecular binding of benzo[a]pyrene and two polychlorinated biphenyl congeners in female mice. Chem. Biol. Interact. (2001) 137(3):243–258.[CrossRef][Web of Science][Medline]
Pitot HC, Campbell HA, Maronpot R, Bawa N, Rizvi TA, Xu YH, Sargent L, Dragan Y, Pyron M. Critical parameters in the quantitation of the stages of initiation, promotion, and progression in one model of hepatocarcinogenesis in the rat. Toxicol. Pathol. (1989) 17(4 Pt 1):594–611.[Medline]
Pitot HC, Dragan YP, Teeguarden J, Hsia S, Campbell H. Quantitation of multistage carcinogenesis in rat liver. Toxicol. Pathol. (1996) 24:119–128.
Racker E. Glutathione reductase (liver and yeast). Methods Enzymol. (1955) 2:722–725.[CrossRef][Web of Science]
Reenila I, Tuomainen P, Mannisto PT. Improved assay of reaction products to quantitate catechol-O-methyltransferase activity by high-performance liquid chromatography with electrochemical detection. J. Chromatogr. B Biomed. Appl. (1995) 663:137–142.[CrossRef][Web of Science][Medline]
Rhee SG, Chang TS, Bae YS, Lee SR, Kang SW. Cellular regulation by hydrogen peroxide. J. Am. Soc. Nephrol. (2003) 14(8 Suppl. 3):S211–S215.
Rhee SG, Kang SW, Jeong W, Chang TS, Yang KS, Woo HA. Intracellular messenger function of hydrogen peroxide and its regulation by peroxiredoxins. Curr. Opin. Cell Biol. (2005) 17(2):183–189.[CrossRef][Web of Science][Medline]
Roberts BJ, Whitelaw ML. Degradation of the basic helix-loop-helix/Per-ARNT-Sim homology domain dioxin receptor via the ubiquitin/proteosome pathway. J. Biol. Chem. (1999) 274:36351–36356.
Robertson LW, Hansen LG, eds. PCBs. In: Recent Advances in Environmental Toxicology and Health Effects (2001) Lexington, KY: University Press of Kentucky.
Safe S. Polychlorinated biphenyls (PCBs): Mutagenicity and carcinogenicity. Mutat. Res. (1989) 220(1):31–47.[CrossRef][Web of Science][Medline]
Safe SH. Polychlorinated biphenyls (PCBs): Environmental impact, biochemical and toxic responses, and implications for risk assessment. CRC Crit. Rev. Toxicol. (1994) 24:87–149.
Schaeffer E, Greim H, Grossner W. Pathology of chronic polychlorinated biphenyl feeding in rats. Toxicol. Appl. Pharmacol. (1984) 75:278–288.[CrossRef][Web of Science][Medline]
Schilderman PA, Maas LM, Pachen DM, deKok TM, Kleinjans JC, van Schooten FJ. Induction of DNA adducts by several polychlorinated biphenyls. Environ. Mol. Mutagen. (2000) 36(2):79–86.[CrossRef][Web of Science][Medline]
Schulte-Hermann R, Kraupp-Grasl B, Bursch W, Gerbracht U, Timmermann-Trosiener I. Effects of non-genotoxic hepatocarcinogens phenobarbital and nafenopin on phenotype and growth of different populations of altered foci in rat liver. Toxicol. Pathol. (1989) 17(4 Pt 1):642–649.[Medline]
Schulte-Hermann R, Ohde G, Schuppler J, Timmermann-Trosiener I. Enhanced proliferation of putative preneoplastic cells in rat liver following treatment with the tumor promoters phenobarbital, hexachlorocyclohexane, steroid compounds, and nafenopin. Cancer Res. (1981) 41(6):2556–2562.
Shaban Z, Soliman M, El-Shazly S, El-Bohi K, Abdelazeez A, Kehelo K, Kim HS, Muzandu K, Ishizuka M, Kazusaka A, et al. AhR and PPAR alpha antagonistic effects on CYP2B and CYP3A, and additive inhibitory effects on CYP2C11. Xenobiotica (2005) 35:51–68.[CrossRef][Web of Science][Medline]
Shapiro BH, Agrawal AK, Pampori NA. Gender differences in drug metabolism regulated by growth hormone. Int. J. Biochem. Cell Biol. (1995) 27(1):9–20.[CrossRef][Web of Science][Medline]
Shirai T. A medium term rat liver bioassay as a rapid in vivo test for carcinogenic potential: A historical review of model development and summary of results from 291 tests. Toxicol. Pathol. (1997) 25:453–460.
Silberhorn EM, Glauert HP, Robertson LW. Carcinogenicity of polyhalogenated biphenyls: PCBs and PBBs. Crit. Rev. Toxicol. (1990) 20(6):440–496.[Medline]
Silkworth JB, Sutter TR, Kim JH, Fish KM, Brown JF Jr, Mayes BA. CYP1A1 and CYP1B1 expression in Sprague-Dawley rats fed Aroclors 1016, 1242, 1254 and 1260 for 6 months. Organohalogen Compd. (1999) 42:481–484.
Srinivasan A, Lehmler HJ, Robertson LW, Ludwig G. Production of DNA strand breaks in vitro and reactive oxygen species in vitro and in HL-60 cells by PCB metabolites. Toxicol. Sci. (2001) 60:92–102.
Stohs SJ. Oxidative stress induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Free Radic. Biol. Med. (1990) 9(1):79–90.[Web of Science][Medline]
Upham BL, Kang KS, Cho HY, Trosko JE. Hydrogen peroxide inhibits gap junction intercellular communication in glutathione-sufficient but not glutathione-deficient cells. Carcinogenesis (1997) 18:37–42.
U.S. Environmental Protection Agency (EPA). Integrated Risk Information System (IRIS). (1997) Polychlorinated biphenyls (PCBs) (CASRN 1336-36-3).
van Birgelen AP, DeVito MJ, Akins JM, Ross DG, Diliberto JJ, Birnbaum LS. Relative potencies of polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls derived from hepatic porphyrin accumulations in mice. Toxicol. Appl. Pharmacol. (1996) 138:98–109.[CrossRef][Web of Science][Medline]
Van den Berg M, Birnbaum LS, Denison M, DeVito M, Farland W, Feeley M, Fiedler H, Hakanson H, Hanberg A, Haws L, et al. The 2005 World Health Organization reevaluation of human and mammalian toxic equivalency factors for dioxins and dioxin-like compounds. Toxicol. Sci. (2006) 93(2):223–241.
Ward JM. Increased susceptibility of livers of aged F344 rats to the effects of Phenobarbital on the incidence, morphology and histochemistry of hepatocellular foci and neoplasms. J. Natl. Cancer Inst. (1983) 71:815–822.[Web of Science][Medline]
Waxman DJ, O'Connor C. Growth hormone regulation of sex-dependent liver gene expression. Mol. Endocrinol. (2006) 20(11):2613–2629.
Wetzel LT, Luempert LG III, Breckenridge CB, Tisdel MO, Stevens JT, Thakur AK, Extrom PC, Eldridge JC. Chronic effects of atrazine on estrus and mammary tumor formation in female Sprague-Dawley and Fischer 344 rats. J. Toxicol. Environ. Health (1994) 43:169–182.[Web of Science][Medline]
Whysner J, Montandon F, McClain RM, Downing J, Verna LK, Steward RE 3rd, Williams GM. Absence of DNA adduct formation by phenobarbital, polychlorinated biphenyls, and chlordane in mouse liver using the 32P-postlabeling assay. Toxicol. Appl. Pharmacol. (1998) 148(1):14–23.[CrossRef][Web of Science][Medline]
Whysner J, Wang CX. Hepatocellular iron accumulation and increased cell proliferation in polychlorinated biphenyl-exposed Sprague-Dawley rats and the development of hepatocarcinogenesis. Toxicol. Sci. (2001) 62(1):36–45.
World Health Organization (WHO). Monograph for atrazine. In: IARC Monogr. Eval. Carcinog. Risks Hum. (1999) 73:59–113.[Medline]
World Health Organization (WHO). Polychlorinated Biphenyls: Human Health Aspects. Concise International Chemical Assessment Document (CICAD) 55 (2003) Geneva: World Health Organization.
Yagi K. Lipid peroxides and related radicals in clinical medicine. Adv. Exp. Med. Biol. (1994) 366:1–15.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. B. Silkworth, E. A. Carlson, C. McCulloch, K. Illouz, S. Goodwin, and T. R. Sutter Toxicogenomic Analysis of Gender, Chemical, and Dose Effects in Livers of TCDD- or Aroclor 1254-Exposed Rats Using a Multifactor Linear Model Toxicol. Sci., April 1, 2008; 102(2): 291 - 309. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||