TY - JOUR
AU - Denny, J. S.
AB - Abstract The toxicity of four quinones, 2,3-dimethoxy-1,4-naphthoquinone (DMONQ), 2-methyl-1,4-naphthoquinone (MNQ), 1,4-naphthoquinone (NQ), and 1,4-benzoquinone (BQ), which redox cycle or arlyate in mammalian cells, was determined in isolated trout (Oncorhynchus mykiss) hepatocytes. More than 70% of cells died in 3 h when exposed to BQ or NQ; 50% died in 7 h when exposed to MNQ, with no mortality compared to controls after 7 h DMONQ exposure. A suite of biochemical parameters was assessed for ability to discriminate these reactivity pathways in fish. Rapid depletion of glutathione (GSH) with appearance of glutathione disulfide (GSSG) and increased dichlorofluoroscein fluorescence were used as indicators of redox cycling, noted with DMONQ, MNQ, and NQ. Depletion of GSH with no GSSG accumulation, and loss of free protein thiol (PrSH) groups (nonreducible) indicated direct arylation by BQ. All toxicants rapidly oxidized NADH, with changes in NADPH noted later (BQ, NQ, MNQ) or not at all (DMONQ). Biochemical measures including cellular energy status, cytotoxicity, and measures of reactive oxygen species, along with the key parameters of GSH and PrSH redox status, allowed differentiation of responses associated with lethality. Chemicals that arylate were more potent than redox cyclers. Toxic pathway discrimination is needed to group chemicals for potency predictions and identification of structural parameters associated with distinct types of reactive toxicity, a necessary step for development of mechanistically based quantitative structure-activity relationships (QSARs) to predict chemical toxic potential. The commonality of reactivity mechanisms between rodents and fish was also demonstrated, a step essential for species extrapolations. toxicity pathways, cross-species extrapolation, quinones, glutathione, reactive oxygen species, protein thiol, Oncorhynchus mykiss The U.S. Environmental Protection Agency (EPA), under a variety of legislation, is charged with the responsibility of assessing the hazard of chemicals to human health and the environment, and in doing so relies upon predictive techniques in its decision-making process when empirical toxicological data is unavailable (Auer et al., 1990). In the field of aquatic toxicology, quantitative structure-activity relationships (QSAR) have been used as effective models for predicting the acute and chronic toxicity of organic chemicals (Auer et al., 1990; Hermens, 1986). The proper application of QSAR techniques requires that, where possible, chemical similarity be assessed based on commonality in toxic action (Bradbury, 1995), the assumption being that chemicals categorized by their ability to initiate toxicity through a common pathway will have similar chemical structural properties. A series of QSAR models for estimating toxic responses (Veith and Broderius, 1987; Veith et al., 1983) were eventually developed based on systematically collected fish acute lethality information on over 600 chemicals. Various techniques were employed to elucidate distinct toxic pathways initiated by these compounds for assignment of chemicals to specific toxicity groups. Groups were assigned by (1) critical analysis of behavioral data and dose-response curves from the single chemical acute exposures (Drummond and Russom, 1990; Drummond et al., 1986); (2) an assessment of physiological responses associated with intoxication (McKim et al., 1987a,b,c); and (3) evaluation of joint toxic action in mixture studies (Broderius et al., 1995, and references therein). Such categorization resulted in successful development of mechanistically based QSARs for several pathways of toxicity. The approach also resulted in identification of numerous chemicals that elicit responses consistent with reactive chemical-based toxicity. However, as yet, development of quantitative models predictive of reactive chemical toxicity has not been achievable. Further advancement of these models for prediction of reactive toxicity in aquatic organisms is dependent on the ability to discriminate among multiple pathways of reactivity (Hermens, 1990; Russom et al., 1997). Once pathways are discriminated and chemicals are assigned to reactive toxicity groups, quantum chemical descriptors can be selected to relate electrophile/proelectrophile chemical structure to the propensity of a chemical to interact with nucleophilic sites on target biomolecules (Karabunarliev et al., 1996). Two reactive mechanisms widely studied in mammalian systems for which characteristic biochemical responses have been elucidated are redox cycling and arylation (Comporti, 1989; O‘Brien, 1991). For example, rapid depletion of glutathione (GSH) with formation of glutathione disulfide (GSSG) has been shown to occur upon redox cycling of menadione (2-methyl-1,4-naphthoquinone; MNQ) in isolated rat hepatocytes. Alternatively, exposure of rat cells to 1,4-benzoquinone (BQ) resulted in rapid GSH depletion but no formation of GSSG, a response characteristic of arylation (Rossi et al., 1986). Relatively less is known about the significance of these mechanisms in aquatic organisms, although several important contributions have been made (DiGiulio et al., 1989; Hasspieler et al., 1994; Ribera et al., 1991; Winston and DiGiulio, 1991) that demonstrate similar quinone-based reactive pathways likely occur in aquatic species, with significant parallels in the biochemical responses noted (DiGiulio et al., 1989). The present work was an attempt to apply knowledge of reactive toxicity in mammalian systems (and the more limited information in aquatic species) to the task of discriminating redox cycling and arylation pathways in aquatic vertebrates for eventual development of mechanistically based QSAR models predictive of reactive toxicity. The approach taken was conceptually similar to that of McKim et al., 1987a,b,c) who used physiological responses characteristic of known toxic mechanisms (uncoupling of oxidative phosphorylation, inhibition of acetyl cholinesterase, etc.) to discriminate major pathways of industrial chemical toxicity in fish. The resulting fish acute toxicity syndromes were used in a multivariate classification scheme to categorize chemicals, which were then examined for unique structural characteristics associated with toxic action. This was further used to classify untested chemicals as to likely toxic pathway, and subsequently for selection of appropriate QSAR model to predict toxic potential (McKim et al., 1987c). In the current study, a series of biochemical assays were tested for their ability to discriminate redox cycling and arylation by identifying the sequence and magnitude of biochemical responses noted upon exposure of isolated trout hepatocytes to reactive chemicals of known mechanism. An in vitro model was used in the current study because whole organism responses were previously determined inadequate to discriminate reactive toxicities (Russom et al., 1997). Chemicals were chosen for study (2,3-dimethoxy-1,4-naphthoquinone [DMONQ], 1,4-naphthoquinone [NQ], MNQ, and BQ) to cover a range of redox potentials (Fig. 1) and thiol reactivities. MATERIALS AND METHODS Reagents. All chemicals and reagents were purchased from common commercial sources with the exception of DMONQ, which was synthesized by R. Carlson (University of Minnesota, Duluth, MN). Fluorescent probes, monobromobimane (4-bromomethyl-3,6,7-trimethyl-1,5-diazabicyclo[3.3.0]octa-3,6-diene-2,8-dione; mBBr), and 2`,7`-dichlorofluorescin diacetate (DCF-DA) were obtained from Molecular Probes Inc. (Eugene, OR). Fish. Immature male and female rainbow trout (Oncorhynchus mykiss, 158–548 g) were obtained from the Seven Pines Fish Hatchery (Lewis, WI) and held for at least 2 weeks in flow-through 815-l tanks with sand-filtered Lake Superior water (2 μM filtered, UV treated, 4 l/min, pH 7.7, hardness 45 mg/l) at 11°C. Trout were fed commercial Silver Cup trout pellets from Nelson and Sons Inc. (Murray, UT) three times a week at a rate of 1.2% body weight/day and held under a 16-h light:8-h dark photoperiod. Hepatocyte isolation and incubation. Hepatocytes were isolated by a collagenase perfusion technique, yielding approximately 4 × 108 cells from two 2.5–5 g livers from trout fasted 48 h (Tapper et al., 2000). After several rinsing steps (see Tapper et al., 2000 for details), cells were suspended in Hanks Balanced Salt Solution (HBSS) containing 5.5 mM glucose and 1% bovine serum albumin, hereafter referred to as cell control buffer. Following hepatocyte isolation the cell suspension was microscopically examined to ascertain cell concentration by counting cells in a Neubauer improved hemacytometer. Cell viability was established by LDH leakage (Fariss et al., 1985), DNA pellet analysis of viable cell fraction (Erwin et al., 1981; Fariss et al., 1985), and trypan blue dye exclusion (initial viability was > 90% in all cases). Experimental design. Hepatocyte exposures were begun by gently pelleting cells (100 × g, 2 min), resuspending cells in control buffer or toxicant solution (100 μM and 400 μM NQ; 200 μM and 600 μM BQ; 200 μM and maximum solubility of MNQ, and DMONQ), and placing in round bottom flasks (2–4 × 106 cells/ml; 20–50 ml/flask) and shaking (orbital shaker, 11°C) under an atmosphere of 0.25% CO2:balance air for 7 h. The following parameters were monitored in isolated trout hepatocytes exposed to each quinone: ATP, ADP, and AMP for computation of ATP/ADP ratio and cellular energy charge; NADP+, NADPH, NAD+, and NADH for assessment of pyridine nucleotide availability; GSH and GSSG for glutathione redox status; concentration of free protein thiol (PrSH) and portion of any noted PrSH loss recoverable by addition of a reducing agent; dichlorofluoroscein (DCF) fluorescence as a measure of total reactive oxygen species (ROS); and measures of cell number and viability including exclusion of trypan dye. Cells in control and exposure flasks were monitored for most biochemical parameters at 0.1, 1.5, 3, 4.5, 6, and 7 h with additional early sample points added at 0.5 and 1 h for GSH and GSSG, and 0.25 h for DCF. Extracellular and intracellular toxicant concentrations were measured at 0.5, 3, and 6 h. The majority of biochemical parameters listed were monitored in each experiment, with samples withdrawn from control, low toxicant, or high toxicant flasks at times specified previously. (Due to consistency of responses between low and high concentrations for each toxicant, only cell responses to high toxicant concentrations are reported and discussed.) Data reported for NQ, MNQ, and DMONQ exposures are means of three replicate experiments for each, with two replicates for BQ, using a homogenous mixture of hepatocytes freshly isolated from at least two trout for each experiment. For determination of PrSH depletion additional experiments were performed: three replicate exposures with four flasks each, including control, and DMONQ, MNQ, and NQ high concentrations; and three replicate experiments with three flasks each including control, low BQ, and high BQ. Samples were taken at 0.1, 1.5, and 6 h for all PrSH experiments. Toxicant concentration, cell viability, GSH, and GSSG were measured in these additional PrSH experiments to ensure comparability of responses with previous single toxicant experiments during which the rest of the biochemical parameters were measured. Preliminary investigations revealed that saturated solutions of MNQ and DMONQ were needed to achieve the highest concentrations attainable, in hopes of observing a toxic response in trout hepatocyte suspensions in 7 h. Concentrations below saturation were used for NQ and BQ, to achieve cell lethality in and biochemical response during the 7 h exposures. Toxicant preparation and analysis. BQ and NQ stocks (900–1000 μM, nominal) were prepared by adding weighed quantities of NQ or BQ to control buffer, stirring in the dark, and sonicating. The NQ and BQ stock solutions were filtered (0.2 μ), and diluted with control buffer, 11°C, to achieve nominal concentrations of 400 and 600 μM, respectively. DMONQ and MNQ were prepared as saturated solutions in control buffer (11°C) and filtered (0.2 μ) prior to cell exposure, and sampled for analysis. Due to light sensitivity of quinones, all chemical solutions and cell exposure flasks were protected from light, to the extent possible, throughout the experiment. Concentrations of the stock solutions (NQ, BQ) and saturated solutions (DMONQ, MNQ) were determined prior to dilutions with cells to achieve desired concentrations. Hepatocytes (500 μl, in duplicate) were centrifuged (14,400 × g for 30 s) and the supernatant subsampled (400 μl) for analysis of extracellular toxicant concentrations. The cell pellet was aspirated to dryness and placed on ice for measurement of intracellular toxicant concentrations. The extracellular analytes were stabilized by mixing with 400 μl acetonitrile (ACN), centrifuged (14,400 × g for 30 s), and the supernatant sampled and placed in a refrigerated autosampler for immediate HPLC analysis. Cell pellets for intracellular chemical analyses were resuspended in 300 μl ACN:deionized water (50:50), sonicated to completely disrupt cells, centrifuged (14,400 × g for 30 s), and supernatant sampled for immediate HPLC analysis. To minimize sample degradation, all samples were kept on ice and processed as quickly as possible, i.e., 2 min from sampling to stabilization for extracelluar and within 10 min from sampling to stabilization for intracellular toxicant analyses. Analyses of BQ and HQ concentrated stock solutions, extracellular concentrations (i.e., 600 μM nominal exposure solution), and intracellular concentrations were previously reported by Tapper et al.(2000). Stock concentrations, extracellular, and intracellular concentrations of NQ, DMONQ, and MNQ were determined on a similar isocratic reverse phase HPLC system with a mobile phase of ACN:deionized water (44:56) using UV detection at 340 nm. All toxicant concentrations were quantified using external standards, fortified samples, and replicate samples. Glutathione analysis. Hepatocytes (800 μl) were sampled in duplicate and viable cells separated from nonviable by centrifugation through a di-n-butylphthalate layer into 70% perchloric acid (PCA) thus lysing live cells and releasing the thiols of interest (modified from Fariss et al., 1985). The PCA layer was subsampled and derivatized with dansyl chloride and analysed by HPLC with fluorescence detection as described by Hammermeister et al.(2000). Stability of the analytes was assured by use of cold reagants, immediately icing samples, rapid centrifugation into PCA and initial sulfhydryl carboxymethylation. All steps were completed within 8–10 min of sampling. External standards, and fortified and replicate samples were used for quantification. Detection limits at 3:1 signal to noise ratio were typically 1.0 pmol on column for all analytes. Protein thiol analysis. The determination of intracellular PrSH and dithiotreitol (DTT) recoverable PrSH was based on a modification of the method of Cotgreave and Moldeus (1986) as described in Tapper et al.(2000). Content of “free” PrSH (i.e., that portion available to react with monobromobimane) was reported as nmol GSH equivalents/106 cells. The contribution of oxidation to the total depletion of cellular PrSH was determined by treating samples with 100 mM DTT for 30 min at room temperature in 10% Triton X-100 prior to derivatization with mBBr (20 mM). Assuming oxidation and arylation are primarily responsible for the depletion of cellular PrSH, reversing the apparent oxidative depletion with DTT allows the remaining loss, attributed to arylation, to be quantified. Adenine and pyridine nucleotide analysis. Hepatocytes (800 μl, in duplicate) were centrifuged to separate extracellular and intracellular fractions as previously described for GSH, with the exception of xanthine used in the acid (10% PCA layer) as an internal standard (6–8 μM). The PCA layer was subsampled (140 μl), neutralized with 10 M KOH (22 μl) and 2.5 M K2HPO4 (pH 7.4; 18 μl), and analyzed by HPLC following a method of Jones (1981). Due to the labile nature of both samples and standards, samples were collected and prepared in an ice bath. Standard solutions of nucleotides (ATP, ADP, AMP, NADPH, NADH; 200–500 μM) were prepared in 1% PCA for NAD+ and NADP+ while all others were prepared in 0.045 M KOH, with appropriate dilutions to bracket sample ranges (0.2–25 μM). Acid-fraction nucleotide samples and standards (40 μl) were analyzed by HPLC (Shandon Hypersil 5 μ 4.6 × 250 mm C18 RP-HPLC column; Alltech Associates, Deerfield, IL) equipped with a refrigerated autosampler and a diode array UV detector at 260 nm. Elution was achieved using the following mobile phase (1.7 ml/min): A (0.2M NaH2PO4, pH 5.8) and B (0.2M NaH2PO4, pH 5.8 with 25% methanol), and a gradient of 0–3.25 min at 0% B; 3.25–7.75 min to 25% B; 7.75–9 min to 97% B; 9–15 min at 97% B; 15–16 min to 0% B; 16–20 min at 0% B. Nucleotides in samples were identified based on identical retention times and quantified based on areas of standards. Integrity of the samples under conditions of holding and analyses was verified by reanalysis of samples up to 8 h after immediate analysis without any appreciable degradation. For NADH and NADPH analyses, cells were sampled (400 μl) in duplicate, centrifuged (30 s, 9000 × g, 4°C), and lysed by addition of chilled 0.1M KOH (125 μl). Pellets were mixed vigorously into the KOH and allowed to digest on ice for 15 min, followed by neutralization with 15 μl chilled 1M NaH2PO4. Lysate was vortexed and centrifuged, the supernatant (140 μl) filtered (0.45 μ nitrocellulose membrane), and filtrate (> 15 μl) analyzed by HPLC (Jones, 1981) at 340 nm utilizing a mobile phase of (0.3 ml/min) 200 mM NaH2PO4 in 5% methanol, pH 5.8. Elapsed time from cell sampling to injection of filtrate was < 30 min with all sample processing steps performed on ice or in chilled autosamplers. Reactive oxygen species analysis. DCF superstock (SS) was prepared by dissolving 4–5 mg in 2 ml methanol, adding 10 ml control buffer, sonicating 15 min, and diluting to 25 ml with control buffer. A working stock of DCF (0.07–0.08 μM) was prepared by diluting SS in control buffer. A DCF (0–40 pmol) in control buffer baseline curve was generated on a spectrofluorophotometer (RF5000U, Shimadzu Corp., Kyoto, Japan) in fixed wavelength mode (λex = 490 nm, λem = 520 nm, bandwidthex = 10 nm, bandwidthem = 10 nm, 11°C). To prepare DCF-DA stock 2–3 mg was dissolved in 0.25 ml dimethylformamide and diluted to 5 ml in control buffer. Freshly prepared hepatocytes were loaded with DCF-DA (20 μM in control buffer) for 30 min. Cells were then centrifuged and resuspended in control buffer (control) or toxicant solution (quinones in control buffer). DCF standard curves were generated for each control or toxicant exposure flask to adjust for slight differences in cell numbers and their influence on the curve. At time zero, DCF (0–40 pmol) was combined with 0.8 ml cell suspension and control buffer to bring total volume to 3.0 ml, vortexed, transferred to cuvette, and fluorescence measured. Subsequent duplicate 0.8 ml cell suspension samples were taken at various time intervals, combined with 2.2 ml control buffer, and fluorescence measured immediately as above. All buffers and reagents were kept at 11°C throughout the experiment. RESULTS Cell Death and Toxicant Concentration Toxicant concentrations to which cells were exposed were chosen based upon preliminary studies (data not reported) focused on achieving measurable cell death in 7 h, detectable change in biochemical parameters (with at least one observation falling between 0 and 100% change, if possible), with the maximum achievable toxicant concentration in cell incubation buffer not to exceed solubility. Cells were exposed in separate flasks to a low concentration of each chemical (100 μM NQ, and 200 μM BQ, MNQ, and DMONQ), a high concentration of each (400 μM NQ, 600 μM BQ, and saturated solutions of MNQ and DMONQ), or no added chemical (controls). Results of low concentration exposures for a given toxicant and a given parameter either did not show a detectable change from control in 7 h, or exhibited the same response seen at the high concentration but to a lesser extent or at a later point in time (data not shown). Consequently, only results of high concentration exposures are presented. Seventy to 90% of hepatocytes exposed to nominal concentrations of 400 μM NQ or 600 μM BQ died in 3 h, determined by trypan blue dye exclusion (Fig. 2). Saturated solutions of MNQ and DMONQ resulted in 50% death in 7 h, or no measurable cell death, respectively. Additional measures of cell viability included LDH leakage, and DNA content of pellets prepared from viable cells separated from nonviable cells by centrifugation through a density gradient. DNA measures agreed well with dye exclusion, while LDH leakage proved to be a relatively ambiguous measure of trout hepatocyte viability (data not shown). Measured total, extracellular, and intracellular chemical concentrations to which trout cells were exposed help explain noted differences in hepatocyte susceptibility to quinones (Table 1). Nominal (target) concentrations, measured concentration in buffer prior to mixing with hepatocytes, measured extracellular, and detectable intracellular toxicant concentrations over time (0.5 and 6 h) are presented. The apparent solubility of DMONQ in suspension medium at 11°C in the absence of cells (∼500 μM) was somewhat less than that of MNQ (∼800 μM), dictating the achievable extracellular exposure concentrations. Comparison of initial (0.5 h) extracellular concentrations in the presence of cells, for the four toxicants, reveal that similar concentrations of two quinones, 229 μM NQ and 306 μM BQ (as BQ + HQ), resulted in a rapid and substantial cell toxicity, while the maximum achievable concentration of DMONQ (484 μM) resulted in no effect even after 7 h (Fig. 2; Table 1). The measured MNQ extracellular concentration of 715 μM resulted in a gradual decline in cell viability but to a lesser extent than that due to NQ or BQ. Measured intracellular chemical concentrations also followed this trend, with relatively more measurable MNQ than DMONQ, especially at 6 h. There were no, and low, detectable intracellular concentrations of NQ and BQ, respectively. Direct measurement of quinone concentrations was somewhat complicated by buffer components. As noted in Table 1, the BQ concentration represents the sum of measured BQ plus HQ. Upon mixing of BQ with cell suspension buffer, about half of the detectable chemical is present as the reduced form, HQ. Shortly after addition to hepatocytes this equilibrium is further altered with 100% of the toxicant measurable only as HQ (see Tapper et al., 2000, for a more complete discussion). A similar equilibrium between oxidized and reduced toxicant form may have occurred with NQ although no attempt was made to measure a possible reduced form of this chemical. The instability of NQ and BQ in comparison with MNQ and DMONQ, no doubt related to their inherent reactivities, is perhaps most strikingly reflected in the measured intracellular concentrations of these quinones. There was no measurable intracellular NQ in any of the experiments, ∼1 nmol/million cells measured HQ, and 5–10 nmol/million cells measurable intracellular DMONQ and MNQ (Table 1). Glutathione Glutathione depletion occurred rapidly, within minutes, in response to all toxicants, however the magnitude of the response was different among the quinones (Fig. 3A). There was an initial immediate GSH decline of 70–80% in DMONQ and MNQ exposed cells, and 50% in NQ exposed cells by 0.5 h. Further decline in detectable GSH continued through 7 h for MNQ and NQ exposed cells, while cells exposed to DMONQ maintained ∼30% of the initial GSH throughout the exposure. The depletion of GSH in BQ-treated cells was so rapid that the initial sample (0.1 h) was completely depleted of GSH compared to control. Concurrent with the rapid GSH depletion there was a notable rise in intracellular GSSG (0.1 h) with DMONQ > MNQ > NQ (Fig. 3B). There was no detectable GSSG in trout cells exposure to 600 μM BQ. Protein Thiols Cellular PrSH were also depleted rapidly upon exposure to three of the quinones, with near 80% depletion noted with NQ and BQ by 0.1 h and 6 h, respectively. MNQ depleted PrSH by 40% at 1.5 h, while no change was noted due to the di-substituted DMONQ (Fig. 4A). The time course for the total loss of “free” PrSH groups was similar to that of GSH depletion, with the initial insult occurring in minutes. When DTT was added to the PrSH samples to reduce oxidized thiols, there was complete recovery of PrSH groups in cells exposed to MNQ, and NQ (Fig. 4B). There was some recovery of oxidized thiols from cells exposed to BQ, but close to 50% of the protein-associated thiols remained unavailable even after reaction with DTT. Pyridine Nucleotides Pyridine nucleotides in trout cells also changed dramatically upon exposure to the quinones. NADH declined immediately and drastically in cells exposed to NQ and BQ, and notably in cells exposed to MNQ and DMONQ (Fig. 5A). Immediate increases in oxidized nucleotides (NAD(P)+) were noted for all toxicants, with more dramatic changes noted for NQ and BQ exposure (Figs. 5B and 5D). In contrast, intracellular concentrations of NADPH were largely unaffected with DMONQ and MNQ exposures, but rapid decline was noted with NQ and BQ (Fig. 5C). Assessment of total NADH + NAD+ (Fig. 6A) shows that increases in NAD+ largely counterbalance losses in NADH in NQ exposed cells, maintaining levels until near total cell death at 3 h. Total NADH + NAD+ declined gradually in DMONQ and MNQ exposed cells but never to the extent seen with NQ and BQ. This was not the case with NADPH + NADP+, which was maintained throughout exposures to DMONQ and MNQ, and in NQ treated cells until death at 3 h (Fig. 6B). All pyridine nucleotide levels decreased with rapidly declining cell viability in BQ exposed cells. Reactive Oxygen A fluorescence assay was used as an indicator of total ROS. A separate experimental design was used in which one isolate of cells was exposed in separate flasks to DMONQ, MNQ, and NQ. BQ exposures were conducted separately due to insufficient cell volumes. This multiple exposure (single toxicant per flask) was done to maximize comparability between toxicants as fluorescent (DCF) background signal can vary significantly in different isolates (data not shown). The mean response (of replicate measures) noted with each toxicant in one representative experiment is presented in Figure 7. For the three toxicants using the same cell isolate (NQ, MNQ, DMONQ) overall production of reactive species measured by DCF was greatest for NQ, followed by MNQ and DMONQ. For each of these toxicants, cells exposed to lower concentrations produced lesser amounts of DCF in a concentration-dependent manner (data not shown). However, this was not the case with BQ. In a separate experiment, when hepatocytes were exposed to 90, 150, 200, and 600 μM BQ, there was no increase in DCF signal over control values up to 200 μM. However, a large DCF signal was measured when trout hepatocytes were exposed to 600 μM BQ (200 and 600 μM concentrations are shown in Fig. 7). Adenine Nucleotides Adenine nucleotides were also measured in trout hepatocytes exposed to these quinones. A drastic decrease in ATP/ADP ratios (75%) were observed within minutes of exposure to NQ and BQ. These decreases resulted in an immediate 20–30% decline in cellular energy charge, which dropped dramatically by 3 h and corresponded with cell death (Figs. 8A and 8B; Fig. 2). MNQ caused a 30% decrease in the ATP/ADP ratio in trout cells by 1.5 h, which proceeded to 70% by 7 h (Fig. 8A). This translated into a gradual decline in cellular energy charge throughout the exposure (Fig. 8B). Although DMONQ caused a gradual decrease (maximum 30%) in ATP/ADP ratio, there was no overall decline in cellular energy charge throughout the exposure, consistent with cell viability measures. DISCUSSION It is essential to accurately measure chemical exposure concentrations to ensure solubility limits are not exceeded when working with complex media especially at low temperatures. The identification of chemical form present is also important when distinguishing toxic pathways to group chemicals for QSAR development. As this study indicates, rapid conversions among complex chemical forms may be encountered when making such measurements. For instance, BQ is known to react with protein in the media and is rapidly converted to HQ in media alone and to a greater extent in the presence of trout hepatocytes (Tapper et al., 2000). Similar association with proteins and rapid nonenzymatic conversion between oxidized and reduced forms may have occurred with NQ (although not determined in present study). Measured concentrations of MNQ and DMONQ are likely more representative of total chemical available as these chemical forms are less reactive with proteins in the media and somewhat less susceptible to autooxidation than BQ and NQ. Overall, measurement of chemical concentrations allowed a more accurate determination of relative potency of arylating and redox cycling toxicants than use of nominal concentrations, with BQ and NQ more potent than MNQ or DMONQ. Attempts to make direct comparisons of measured toxicant concentrations and chemical form in the current study with similar studies using isolated mammalian cells were unsuccessful due to the lack of available comparative data. Studies utilizing isolated rat hepatocytes generally rely on nominal exposure concentrations despite the presence of complex media components and cellular metabolism, which can significantly alter chemical form and available concentration. Given the lack of measured data, potency comparisons based on reported nominal concentrations indicate that trout hepatocytes at 11°C generally survived longer upon exposure to higher concentrations than did isolated rat hepatocytes exposed to the same four quinones at 37°C. Trout hepatocytes showed no decreased viability after 7 h exposure to 500 μM DMONQ, exhibited 50% cell death after 7 h in 800 μM MNQ, and 80% death in 3 h with 400 μM NQ. In contrast, Gant et al.(1988) reported an 80% decrease in isolated rat hepatocyte viability after 4 h at 500 μM DMONQ, or 2 h at 200 μM MNQ. Rat hepatocytes died in 4 h when exposed to 200 μM MNQ or 100 μM NQ as reported by Miller et al.(1986). The slowed rate of metabolism in fish at the lower physiological temperature may, in part, be responsible for these differences (e.g., Kolanczyk et al., 1999), although differences in measured versus nominal concentrations may also play a role. Regardless of differences in time to death, similarities in toxic pathways are apparent in trout and rat cells exposed to reactive quinones. The in vitro isolated hepatocyte model and biochemical assays used in this study were able to discriminate predominant redox or arylation driven reactive toxicities in a cold-water fish species. Quinones were chosen for the study because chemical structures can be found along redox cycling and arylation continuums (Fig. 1). DMONQ was chosen as a representative “pure” redox cycler because of its strong negative redox potential and its inability to arylate due to double substitution on the quinone ring (Gant, 1988). In contrast, BQ, described by some as a “pure” arylator, has four quinone ring positions available for direct interaction with thiols and other groups (Alt and Eyer, 1998) but has limited potential to redox cycle due to its positive one-electron redox potential (Powis and Appel, 1980). MNQ and NQ both have sites available for arylation and both have redox potentials favorable for cycling with molecular oxygen at –155 mV (Wardman, 1989). Therefore, biochemical responses expected from MNQ and NQ exposures should be intermediate between those characteristic of a pure arylator (BQ) and a pure redox cycler (DMONQ), as was observed in the present study. Measured alterations in GSH and PrSH cellular concentrations and redox states are as useful for discriminating reactive toxicity in fish as in mammalian cells. This is in spite of the fact that control cell GSH concentrations can be significantly lower in aquatic species (Tapper et al., 2000) requiring development of analytical methods with more sensitive detection limits (Hammermeister et al., 2000). Based on previous investigations (Comporti, 1989; O‘Brien, 1991; Smith and Mitchell, 1989) redox cycling compounds would be expected to deplete GSH while forming GSSG, deplete free PrSH in the formation of protein mixed disulfides (recoverable by DTT treatment), deplete reduced pyridine nucleotides and ATP, and possibly show an elevation of DCF fluorescence as an indicator of ROS (Cathcart et al., 1983; LeBel et al., 1992; Roberts et al., 1996). Chemicals that primarily arylate would be expected to reduce available GSH with no formation of GSSG, and deplete PrSH (nonrecoverable by reduction with DTT), with no change in reducing equivalents prior to cell death, and no observable changes in ROS. In the present study, the most dramatic change for the redox cycler DMONQ was an immediate GSH depletion due to rapid and extensive conversion to GSSG. The rapid appearance of GSSG distinguished DMONQ from the pure arylator BQ (Rossi et al., 1986). As expected, the immediate depletion of GSH by BQ was not accompanied by detectable GSSG, but likely by extensive BQ-SG conjugation. Arylation was also indicated as PrSH loss, largely unrecoverable through DTT reduction, an expected consequence of BQ reactivity with protein sulfhydryl groups. Naphthoquinone and MNQ also rapidly depleted GSH with some conversion to oxidized GSSG, the degree of which seemed to correlate with their redox potential, that is, the more negative the redox potential the larger the measured GSSG increase. These two chemicals also depleted PrSH but apparently largely through oxidation and not conjugation (DTT recoverable). Therefore it would seem that both NQ and MNQ redox cycle to some extent in fish cells. Taken together, conversion of GSH to GSSG, free PrSH conversion to mixed disulfides recoverable with a strong reducing agent, and production of ROS are indicative of oxidative stress in trout cells produced as a consequence of redox cycling. The fluorescence (DCF) assay for ROS serves as a qualitative indicator of redox cycling, with rank order potency of NQ > MNQ > DMONQ. Lemaire et al.(1996) also showed NQ driven production of ROS to be greater than MNQ in an assay that measured NADH-dependent apparent one-electron reduction using rainbow trout purified DT-diaphorase (an enzyme usually thought of as catalyzing two-electron reductions). However, if GSSG production in the current study were used as an indicator of redox cycling, the rank order of response was DMONQ > MNQ > NQ. Regardless of the relative redox activity, GSSG formation and DCF signal were both adequate qualitative measures of toxicant induced oxidative stress for the quinones studied here, with the possible exception of BQ. Exposure of trout cells to a nontoxic (200 μM) concentration of BQ showed no response in the DCF assay yet 600 μM BQ eventually yielded a large DCF signal. However, no increase in GSSG was noted with 600 μM BQ. Recalling that the majority of measured chemical present during this exposure was HQ (Table 1) it is likely that a significant amount of HQ conjugation occurred resulting in formation of HQ-SG conjugates. The plausibility of glutathionyl-hydroquinone conjugates undergoing redox cycling has been previously described by Rao (1996), which could account for the delayed DCF signal occurring upon cell exposures to high BQ concentrations in the current study. The delayed nature of the response, however, would indicate an indirect effect likely not indicative of the primary cause of cell death. This does however emphasize the importance of determining time and concentration dependency of any noted response. Given this caution, it appears that the propensity for a chemical to redox cycle as a primary mode of action in trout cells can be determined by measuring GSSG appearance, DTT recoverable PrSH depletion, and/or DCF signal. Chemical rank order potency associated with cell death is important to consider when discriminating pathways of toxic action. The rank order of cytotoxicity in fish cells (based on measured extracellular concentrations) of BQ ≥ NQ > MNQ > DMONQ, assessed in the context of the biochemical responses measured, indicate a greater chemical potency when arylation is the primary pathway of reactive toxicity when compared with redox cycling, a determination consistent with relative toxic potential in rodent hepatocytes (Gant et al., 1988; Henry and Wallace, 1995; Miller et al., 1986). Cytotoxicity, loss of energy charge, and depletion of PrSH and GSH were similar for NQ and BQ in rat hepatocytes. Additionally, NQ exposed cells generated GSSG and DTT recoverable PrSH, a response not seen with BQ and indicative of some degree of redox cycling. The toxicity of NQ to trout hepatocytes was greater than MNQ, although both redox cycle and arylate to some extent. A study by Henry and Wallace (1995) measuring rat liver mitochondrial oxygen consumption as a measure of redox cycling indicated that NQ may redox cycle to a greater degree than MNQ. Although GSH depletion seemed somewhat less for NQ, the extent of PrSH depletion was greater than with MNQ. It is likely the difference in toxicity can be linked to differences in targeted proteins. In associated studies (not shown), there were notable differences in cytoskeletal components of isolated trout hepatocytes that were targeted by NQ and MNQ (K. Flynn, personal communication). This additional information may prove essential in discriminating primary pathways of toxicity for chemicals for which multiple mechanisms are operable. These distinctions may be necessary for the selection of appropriate stereoelectronic chemical parameters associated with different types of reactive toxic pathways needed for development of predictive models. Conclusions A suite of biochemical responses were successfully used to discriminate redox cycling and arylation pathways of reactive chemical toxicity in fish. Cellular energy status, cytotoxicity, and measures of reactive oxygen species production, along with the key parameters of glutathione redox status (requiring measurement of GSH and GSSG) and PrSH status (including DTT recoverable portion), allowed differentiation of responses associated with lethality. Once additional chemicals are tested and reactivity pathways discriminated, application of multivariate statistics may prove useful in classifying chemicals as to major toxic pathway. These classifications are needed to group chemicals for determination of structural parameters associated with distinct types of reactive toxicity, for development of mechanistically based QSARs to predict toxic potential of untested chemicals. Additionally, the current study further substantiates a functional concordance between rodents and fish for the two pathways of reactive toxicity under study, a step essential for extrapolation of data among species for the assessment of ecological and human health risk. TABLE 1 Differences in Hepatocyte Susceptibility to Quinones   Concentration (μM)  Extracellular (μM)  Intracellular (nmol/106 cells)    Nominal  Measureda  0.5 hb  6 hc  0.5 hd  6 he  Note. Nominal and measured concentration of four quinones in cell suspension media prior to exposure of hepatocytes, and in extracellular and intracellular fractions sampled during exposure of isolated rainbow trout hepatocytes are shown.  aConcentration measured in cell medium without cells, immediately prior to exposure.  bConcentration measured in cell medium 0.5 h after mixing with hepatocytes.  cConcentration measured in cell medium 6 h after mixing with hepatocytes.  dIntracellular concentration measured 0.5 after mixing with hepatocytes.  eIntracellular concentration measured 6 h after mixing with hepatocytes.  fMean ± SE (number of experiments).  gNot measured.  hNone detected at estimated NQ detection limit of 0.10 nmol/106 cells.  iMeasured concentration prior to addition of hepatocytes contained 50% 1,4-benzoquinone with 1,4-hydroquinone(HQ); after 30 min mixing with hepatocytes, all measured concentration was as HQ.  DMONQ  Saturation  503 ± 12 (3)f  484 ± 6 (3)  451 ± 13 (3)  4.99 ± 1.36 (3)  4.61 ± 1.01 (3)  MNQ  Saturation  798 ± 42 (3)  715 ± 45 (3)  572 ± 8 (3)  6.28 ± 0.86 (2)  10.76 ± 0.50 (2)  NQ  400  —g  229 ± 3 (3)  120 ± 10 (4)  < 0.10h  < 0.10  BQ + HQi  600  334 ± 11 (2)  306 ± 12 (2)  251 ± 31 (2)  0.79 ± 0.09 (2)  1.10 ± 0.11 (2)    Concentration (μM)  Extracellular (μM)  Intracellular (nmol/106 cells)    Nominal  Measureda  0.5 hb  6 hc  0.5 hd  6 he  Note. Nominal and measured concentration of four quinones in cell suspension media prior to exposure of hepatocytes, and in extracellular and intracellular fractions sampled during exposure of isolated rainbow trout hepatocytes are shown.  aConcentration measured in cell medium without cells, immediately prior to exposure.  bConcentration measured in cell medium 0.5 h after mixing with hepatocytes.  cConcentration measured in cell medium 6 h after mixing with hepatocytes.  dIntracellular concentration measured 0.5 after mixing with hepatocytes.  eIntracellular concentration measured 6 h after mixing with hepatocytes.  fMean ± SE (number of experiments).  gNot measured.  hNone detected at estimated NQ detection limit of 0.10 nmol/106 cells.  iMeasured concentration prior to addition of hepatocytes contained 50% 1,4-benzoquinone with 1,4-hydroquinone(HQ); after 30 min mixing with hepatocytes, all measured concentration was as HQ.  DMONQ  Saturation  503 ± 12 (3)f  484 ± 6 (3)  451 ± 13 (3)  4.99 ± 1.36 (3)  4.61 ± 1.01 (3)  MNQ  Saturation  798 ± 42 (3)  715 ± 45 (3)  572 ± 8 (3)  6.28 ± 0.86 (2)  10.76 ± 0.50 (2)  NQ  400  —g  229 ± 3 (3)  120 ± 10 (4)  < 0.10h  < 0.10  BQ + HQi  600  334 ± 11 (2)  306 ± 12 (2)  251 ± 31 (2)  0.79 ± 0.09 (2)  1.10 ± 0.11 (2)  View Large FIG. 1. View largeDownload slide Chemical structure, Log octanol:water partition coefficient (Log P), and one-electron redox potential for four quinones tested in the present study. Log P values for DMONQ (calculated) and MNQ, NQ, BQ (measured) are from the CLOGP™ program version 3.4 and STARLIST database, respectively, within the UDRIVE system version 3.53, 1988, from Pomona College Medicinal Chemistry Project, Claremont, CA. One electron reduction potentials are from aBradburyet al.(1995), bOllinger et al.(1990), and cWardman (1989). FIG. 1. View largeDownload slide Chemical structure, Log octanol:water partition coefficient (Log P), and one-electron redox potential for four quinones tested in the present study. Log P values for DMONQ (calculated) and MNQ, NQ, BQ (measured) are from the CLOGP™ program version 3.4 and STARLIST database, respectively, within the UDRIVE system version 3.53, 1988, from Pomona College Medicinal Chemistry Project, Claremont, CA. One electron reduction potentials are from aBradburyet al.(1995), bOllinger et al.(1990), and cWardman (1989). FIG. 2. View largeDownload slide Viability of isolated rainbow trout hepatocytes exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 7 h. Exposure concentrations and number of replicate experiments are specified in Table 1; chemical names and parameters are specified in Figure 1. Data are expressed as percentage of cells no longer capable of excluding trypan blue dye relative to control cells (mean ± SE). FIG. 2. View largeDownload slide Viability of isolated rainbow trout hepatocytes exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 7 h. Exposure concentrations and number of replicate experiments are specified in Table 1; chemical names and parameters are specified in Figure 1. Data are expressed as percentage of cells no longer capable of excluding trypan blue dye relative to control cells (mean ± SE). FIG. 3. View largeDownload slide Change in intracellular concentrations of (A) GSH or (B) GSSG in isolated rainbow trout hepatocytes exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 7 h. Exposure concentrations and number of replicate experiments are specified in Table 1; chemical names and parameters are specified in Figure 1. Data are expressed as percentage change relative to control values measured at the same time (mean ± SE). FIG. 3. View largeDownload slide Change in intracellular concentrations of (A) GSH or (B) GSSG in isolated rainbow trout hepatocytes exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 7 h. Exposure concentrations and number of replicate experiments are specified in Table 1; chemical names and parameters are specified in Figure 1. Data are expressed as percentage change relative to control values measured at the same time (mean ± SE). FIG. 4. View largeDownload slide Decline in measurable (A) total free cellular protein thiols, and (B) that portion of PrSH groups recoverable upon addition of the reducing agent dithiothreitiol to samples taken from isolated rainbow trout hepatocyte suspensions exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 24 h (data shown through 6 h). Data shown are mean ± SE of five experiments expressed relative to controls. FIG. 4. View largeDownload slide Decline in measurable (A) total free cellular protein thiols, and (B) that portion of PrSH groups recoverable upon addition of the reducing agent dithiothreitiol to samples taken from isolated rainbow trout hepatocyte suspensions exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 24 h (data shown through 6 h). Data shown are mean ± SE of five experiments expressed relative to controls. FIG. 5. View largeDownload slide Change in intracellular concentrations of (A) NADH, (B) NAD+, (C) NADPH, or (D) NADP+ in isolated rainbow trout hepatocytes exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 7 h. Exposure concentrations and number of replicate experiments are specified in Table 1; chemical names and parameters are specified in Figure 1. Data are expressed as percentage change relative to control values measured at the same time (mean ± SE). FIG. 5. View largeDownload slide Change in intracellular concentrations of (A) NADH, (B) NAD+, (C) NADPH, or (D) NADP+ in isolated rainbow trout hepatocytes exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 7 h. Exposure concentrations and number of replicate experiments are specified in Table 1; chemical names and parameters are specified in Figure 1. Data are expressed as percentage change relative to control values measured at the same time (mean ± SE). FIG. 6. View largeDownload slide Change in intracellular concentrations of (A) NADH+NAD+ and (B) NADPH+NADP+ in isolated rainbow trout hepatocytes exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 7 h. Exposure concentrations and number of replicate experiments are specified in Table 1, chemical names and parameters are specified in Figure 1. Data are expressed as percentage change relative to control values measured at the same time (mean ± SE). FIG. 6. View largeDownload slide Change in intracellular concentrations of (A) NADH+NAD+ and (B) NADPH+NADP+ in isolated rainbow trout hepatocytes exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 7 h. Exposure concentrations and number of replicate experiments are specified in Table 1, chemical names and parameters are specified in Figure 1. Data are expressed as percentage change relative to control values measured at the same time (mean ± SE). FIG. 7. View largeDownload slide Increase in fluorescence of dichlorofluorescein as a measure of reactive oxygen species in isolated rainbow trout hepatocytes exposed to: control (open square), 500 μM DMONQ (circle), 800 μM MNQ (filled square), 400 μM NQ (filled diamond; one representative three toxicant experiment with cells from the same isolation); or 200 μM BQ (filled triangle), or 600 μM BQ (open triangle; one representative BQ experiment with cells from the same isolation); for 6 h. FIG. 7. View largeDownload slide Increase in fluorescence of dichlorofluorescein as a measure of reactive oxygen species in isolated rainbow trout hepatocytes exposed to: control (open square), 500 μM DMONQ (circle), 800 μM MNQ (filled square), 400 μM NQ (filled diamond; one representative three toxicant experiment with cells from the same isolation); or 200 μM BQ (filled triangle), or 600 μM BQ (open triangle; one representative BQ experiment with cells from the same isolation); for 6 h. FIG. 8. View largeDownload slide Change in the intracellular (A) ATP/ADP ratio or (B) energy charge of isolated rainbow trout hepatocytes exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 7 h. Energy charge is: [ATP + (0.5ADP)/(ATP + ADP + AMP)]. Exposure concentrations and number of replicate experiments are specified in Table 1; chemical names and parameters are specified in Figure 1. Data are expressed as percentage change relative to control values measured at the same time (mean ± SE). FIG. 8. View largeDownload slide Change in the intracellular (A) ATP/ADP ratio or (B) energy charge of isolated rainbow trout hepatocytes exposed to DMONQ (circle), MNQ (square), NQ (diamond), or BQ (triangle) for 7 h. Energy charge is: [ATP + (0.5ADP)/(ATP + ADP + AMP)]. Exposure concentrations and number of replicate experiments are specified in Table 1; chemical names and parameters are specified in Figure 1. Data are expressed as percentage change relative to control values measured at the same time (mean ± SE). 1 To whom correspondence should be addressed. Fax: (218) 529-5003. E-mail: schmieder.patricia@epa.gov. The research reported here has been subjected to review by the EPA‘s National Health and Environmental Effects Research Laboratory and approved for publication. Approval does not signify that the contents reflect the views of the EPA, nor does mention of trade names or commercial products constitute endorsement. The authors would like to thank J. M. McKim III, T. R. Henry, and three anonymous reviewers for their thoughtful review of the manuscript. REFERENCES Alt, C., and Eyer, P. ( 1998). Ring addition of the alpha-amino group of glutathione increases the reactivity of benzoquinone thioethers. Chem. Res. Toxicol.  11, 1223–1233. Google Scholar Auer, C. M., Nabholz, J. V., and Baetcke, K. P. ( 1990). Mode of action and the assessment of chemical hazards in the presence of limited data: Use of structure activity relationships (SAR) under TSCA Section 5. Environ. Health Perspect.  87, 183–187. Google Scholar Bradbury, S. P. ( 1995). Quantitative structure-activity relationships and ecological risk assessment: An overview of predictive aquatic toxicology research. Toxicol. Lett.  79, 229–237. Google Scholar Bradbury, S. P., Mekenyan, O. G., Veith, G., and Zaharieva, N. ( 1995). SAR models for futile metabolism: One-electron reduction of quinones, phenols, and nitrobenzene. SAR QSAR Environ. Res.  4, 109–124. Google Scholar Broderius, S. J., Kahl, M. D., and Hoglund, M. D. ( 1995). Use of joint toxic response to define the primary mode of toxic action for diverse industrial organic chemicals. Environ. Toxicol. Chem.  9, 1591–1605. Google Scholar Cathcart, R., Schwiers, E., and Ames, B. N. ( 1983). Detection of picomole levels of hydroperoxides using a fluorescent dichlorofluorescein assay. Anal. Biochem.  134, 111–116. Google Scholar Comporti, M. ( 1989). Three models of free radical-induced cell injury. Chem. Biol. Interact.  72, 1–56. Google Scholar Cotgreave, I. A., and Moldeus, P. ( 1986). Methodologies for the application of monobromobimane to the simultaneous analysis of soluble and protein thiol components of biological systems. J. Biochem. Biophys. Methods  13, 231–249. Google Scholar DiGiulio, R. T., Washburn, P. C., Wenning, R. J., Winston, G. W., and Jewell, C. S. ( 1989). Biochemical responses in aquatic animals: A review of determinants of oxidative stress. Environ. Toxicol. Chem.  8, 1103–1123. Google Scholar Drummond, R. A., and Russom, C. L. ( 1990). Behavioral toxicity syndromes: A promising tool for assessing toxicity mechanisms in juvenile fathead minnows. Environ. Toxicol. Chem.  9, 37–46. Google Scholar Drummond, R. A., Russom, C. L., Geiger, D. L., and Defoe, D. L. (1986). Behavioral and morphological changes in fathead minnow (Pimephales promelas) as diagnostic endpoints for screening chemicals according to mode of action. In Aquatic Toxicology and Environmental Fate (T. M. Poston and R. Purdy, Eds.), Vol. 9, STP 921, pp. 415–435. American Society for Testing and Materials, Philadelphia, PA. Google Scholar Erwin, B. G., Stoscheck, C. M., and Florini, J. R. ( 1981). A rapid fluorometric method for the estimation of DNA in cultured cells. Anal. Biochem.  110, 291–294. Google Scholar Fariss, M. W., Brown, M. K., Schmitz, J. A., and Reed, D. J. ( 1985). Mechanism of chemical-induced toxicity. I. Use of a rapid centrifugation technique for the separation of viable and non-viable hepatocytes. Toxicol. Appl. Pharmacol.  79, 283–295. Google Scholar Gant, T. W., Rao, D. N. R., Mason, R. P., and Cohen, G. M. ( 1988). Redox cycling and sulphydryl arylation: Their relative importance in the mechanism of quinone cytotoxicity to isolated hepatocytes. Chem. Biol. Interact.  65, 157–173. Google Scholar Hammermeister, D. E., Serrano, J., Schmieder, P. K., and Kuehl, D. W. ( 2000). Characterization of dansylated glutathione, glutathione disulfide, cysteine, and cystine by narrow bore liquid chromatography—electrospray ionization mass spectrometry. Rapid Commun. Mass Spectrom.  14, 503–508. Google Scholar Hasspieler, B. M., Behar, J. V., Carlson, D. B., Watson, D. E., and DiGiulio, R. T. ( 1994). Susceptibility of channel catfish (Ictalurus punctatus) and brown bullhead (Ameriurus nebulosus) to oxidative stress: A comparative study. Aquat. Toxicol.  28, 53–64. Google Scholar Henry, T. R., and Wallace, K. B. ( 1995). Differential mechanisms of induction of the mitochondrial permeability transition by quinones of varying chemical reactivities. Toxicol. Appl. Pharmacol.  134, 195–203. Google Scholar Hermens, J. L. M. ( 1986). Quantitative structure-activity relationships in aquatic toxicology. Pestic. Sci.  17, 287–296. Google Scholar Hermens, J. L. M. ( 1990). Electrophiles and acute toxicity to fish. Environ. Health Perspect.  87, 219–225. Google Scholar Jones, D. P. ( 1981). Determination of pyridine dinucleotides in cell extracts by high-performance liquid chromatography. J. Chromatogr.  225, 446–449. Google Scholar Karabunarliev, S., Mekenyan, O. G., Karcher, W., Russom, C. L., and Bradbury, S. P. ( 1996). Quantum-chemical descriptors for estimating the acute toxicity of electrophiles to the fathead minnow (Pimephales promelas): An analysis based on molecular mechanisms. Quant. Struct.-Act. Relat.  15, 302–310. Google Scholar Kolanczyk, R., Schmieder, P., Bradbury, S., and Spizzo, T. ( 1999). Biotransformation of 4-methoxyphenol in rainbow trout (Oncorhynchus mykiss) hepatic microsomes. Aquat. Toxicol.  45, 47–61. Google Scholar LeBel, C. P., Ischiropoulos, H., and Bondy, S. C. ( 1992). Evaluation of the probe 2`,7`-dichlorofluorescin as an indicator of reactive oxygen species formation and oxidative stress. Chem. Res. Toxicol.  5, 227–231. Google Scholar Lemaire, P., Sturve, J., Forlin, L., and Livingstone, D. R. ( 1996). Studies on aromatic hydrocarbon quinone metabolism and DT-diaphorase function in liver of fish species. Mar. Environ. Res.  42, 317–321. Google Scholar McKim, J. M., Bradbury, S. P., and Niemi, G. J. ( 1987). Fish acute toxicity syndromes and their use in the QSAR approach to hazard assessment. Environ. Health Perspect.  71, 171–186. Google Scholar McKim, J. M., Schmieder, P. K., Carlson, R. W., Hunt, E. P., and Niemi, G. J. ( 1987). Use of respiratory-cardiovascular responses of rainbow trout (Salmo gairdneri) in identifying acute toxicity syndromes in fish: Part 1. Pentachlorophenol, 2,4-dinitrophenol, tricaine methanesulfonate and 1-octanol. Environ. Toxicol. Chem.  6, 295–312. Google Scholar McKim, J. M., Schmieder, P. K., Niemi, G. J., Carlson, R. W., and Henry, T. R. ( 1987). Use of respiratory-cardiovascular responses of rainbow trout (Salmo gairdneri) in identifying acute toxicity syndromes in fish: Part 2. Malathion, carbaryl, acrolein and benzaldehyde. Environ. Toxicol. Chem.  6, 313–328. Google Scholar Miller, M. G., Rodgers, A., and Cohen, G. M. ( 1986). Mechanisms of toxicity of naphthoquinones to isolated hepatocytes. Biochem. Pharmacol.  35, 1177–1184. Google Scholar O‘Brien, P. J. ( 1991). Molecular mechanisms of quinone cytotoxicity. Chem. Biol. Interact.  80, 1–41. Google Scholar Ollinger, K., Buffinton, G. D., Ernster, L., and Cadenas, E. ( 1990). Effect of superoxide dismutase on the autoxidation of substituted hydro- and semi-naphthoquinones. Chem. Biol. Interact.  73, 53–76. Google Scholar Powis, G., and Appel, P. L. ( 1980). Relationship of the single-electron reduction potential of quinones to their reduction by flavoproteins. Biochem. Pharmacol.  29, 2567–2572. Google Scholar Rao, G. S. ( 1996). Glutathionyl hydroquinone: A potent pro-oxidant and a possible toxic metabolite of benzene. Toxicology  106, 49–54. Google Scholar Ribera, D., Narbonne, J. F., Michel, X., Livingstone, D. R., and O‘Hara, S. ( 1991). Responses of antioxidants and lipid peroxidation in mussels to oxidative damage exposure. Comp. Biochem. Physiol.  100C, 177–181. Google Scholar Roberts, A. E., Miller, D. M., Emery, M., and O‘Neill, T. J. ( 1996). Evaluation of oxidative stress in rat hepatocytes using the intracellular probe dichlorofluorescin diacetate and flow cytometry. In Vitro Toxicol.  9, 125–133. Google Scholar Rossi, L., Moore, G., Orrenius, S., and O‘Brien, P. J. ( 1986). Quinone toxicity in hepatocytes without oxidative stress. Arch. Biochem. Biophys.  251, 25–35. Google Scholar Russom, C. L., Bradbury, S. P., Broderius, S. J., Hammermeister, D. E., and Drummond, R. A. ( 1997). Predicting modes of toxic action from chemical structure: Acute toxicity in the fathead minnow (Pimephales promelas). Environ. Toxicol. Chem.  16, 948–967. Google Scholar Smith, C. V., and Mitchell, J. R. (1989). Glutathione-Chemical, Biochemical and Medical Aspects Part B, Chapter 1. J. Wiley & Sons, New York. Google Scholar Tapper, M. A., Sheedy, B. R., Hammermeister, D. E., and Schmieder, P. K. ( 2000). Depletion of cellular protein thiols as an indicator of arylation in isolated trout hepatocytes exposed to 1,4-benzoquinone. Toxicol. Sci.  55, 327–334. Google Scholar Veith, G. D., and Broderius, S. J. (1987). Structure-toxicity relationships for industrial chemicals causing type (II) narcosis syndrome. In QSAR in Environmental Toxicology—II (K. L. E. Kaiser, Ed.), pp. 385–391. D. Reidel, Dordrecht, Netherlands. Google Scholar Veith, G. D., Call, D. J., and Brooke, L. T. ( 1983). Structure-toxicity relationships for the fathead minnow. I. Narcotic industrial chemicals. Can. J. Fish. Aquat. Sci.  40, 743–748. Google Scholar Wardman, P. ( 1989). Reduction potentials of one-electron couples involving free radicals in aqueous solution. J. Phys. Chem. Ref. Data  18, 1637–1745. Google Scholar Winston, G. W., and Di Giulio, R. T. ( 1991). Prooxidant and antioxidant mechanisms in aquatic organisms. Aquat. Toxicol.  19, 137–161. Google Scholar © 2003 Society of Toxicology
TI - Discriminating Redox Cycling and Arylation Pathways of Reactive Chemical Toxicity in Trout Hepatocytes
JF - Toxicological Sciences
DO - 10.1093/toxsci/kfg016
DA - 2003-03-01
UR - https://www.deepdyve.com/lp/oxford-university-press/discriminating-redox-cycling-and-arylation-pathways-of-reactive-ASmui853gj
SP - 66
EP - 76
VL - 72
IS - 1
DP - DeepDyve
ER -