|Home | About | Journals | Submit | Contact Us | Français|
The polybrominated diphenyl ethers (PBDEs) are ubiquitous environmental contaminants whose residues are increasing in fish, wildlife and human tissues. However, relatively little is known regarding the mechanisms of cell injury caused by PBDE congeners in fish. In the present study, we employed flow cytometry-based analyses to understand the onset and mechanisms of cell injury in rainbow trout gill cells (RTgill-W1 cells) exposed to 2,2′,4,4′-tetrabromodiphenyl ether (BDE 47). Substantial optimization and validation for flow cytometry protocols were required during assay development for the trout gill cell line. Exposure to micromolar concentrations of BDE 47 elicited a significant loss in RTgill-W1 cell viability that was accompanied by a decrease in NAD(P)H autofluorescence, a marker associated with disruption of cellular redox status. This loss in NAD(P)H content was accompanied by a decrease in nonylacridine orange fluorescence, indicating mitochondrial membrane lipid peroxidation. Furthermore, low doses of BDE 47 altered cellular forward angle light scatter (FS, a measure of cell diameter or size) and side light scatter properties (SS, a measure of cellular internal complexity), consistent with the early stages of apoptosis. These changes were more pronounced at higher BDE 47 concentrations, which lead to an increase in the percentage of cells undergoing frank apoptosis as evidenced by sub-G1 DNA content. Apoptosis was also observed at a relatively low dose (3.2 μM) of BDE 47 if cells were exposed for an extended period of time (24 hr). Collectively, the results of these studies indicate that exposure of rainbow trout gill cells to BDE47 is associated with the induction of apoptosis likely originating from disruption of cellular redox status and mitochondrial oxidative injury. The current report extends observations in other species demonstrating that oxidative stress is an important mechanism of BDE 47 mediated cellular toxicity, and supports the use of oxidative stress-associated biomarkers in assessing the sublethal effects of PBDEs and their replacements in fish. The application of flow cytometry endpoints using fish cell lines should facilitate study of the mechanisms of chemical injury in aquatic species.
Polybrominated diphenyl ethers (PBDEs) are halogenated compounds used as flame retardants in textiles, plastics and electronics that represent a novel group of persistent environmental contaminants (McDonald, 2002). Because these compounds are not chemically bound into the polymers, they readily migrate into environmental media (McDonald, 2002; Sjodin et al., 2003). The high lipid solubility, low vapor pressure and general resistance to degradation of these compounds has contributed to their environmental persistence and subsequent bioaccumulation throughout food chains (Birnbaum and Cohen Hubal, 2006; Darnerud, 2003; Gill et al., 2004; Hites, 2004; McDonald, 2002). The fact that PBDEs are also commonly detected in ambient air in industrialized regions (Butt et al., 2004), remote Arctic locations (de Wit et al., 2006), as well as in indoor air (Schecter et al., 2005; Stapleton et al., 2005), underscores the widespread distribution and concerns associated with these emerging contaminants.
Of the various PBDE congeners, 2,2′,4,4′-tetrabromodiphenyl ether (IUPAC, BDE 47) is of lower molecular weight and typically the major PBDE congener detected in highest concentrations in fish and human tissues (Darnerud, 2003; Mazdai et al., 2003; Schecter et al., 2007). Interestingly, while BDE 47 often predominates in the tissues of fish, it constitutes a minor contribution to global PBDE production and usage (Darnerud, 2003; Hale et al., 2003). Collectively, PBDE congeners 47, 49, 99, and 100 account for the highest proportion of total PBDE concentrations in the aquatic environment and biota (Betts, 2002; Burreau et al., 1997; Gustafsson et al., 1999; Hites, 2004; Sjodin et al., 2003; Stapleton and Baker, 2003). Furthermore, PBDEs in the edible portions of fish tissues may lead to dietary exposures and increase the PBDE body burdens in humans (McDonald, 2002; Sjodin et al., 2003). Although the human and ecological health risks associated with PBDE exposures are unknown, studies in rodents have shown developmental, reproductive, neurological and endocrine toxicity (Birnbaum and Staskal, 2004; de Wit, 2002; Eriksson et al., 2001; Zhou et al., 2001; Zhou et al., 2002).
Mechanisms of PBDE toxicity appear to be varied and include disruption of thyroid hormone homeostasis, disregulation of genes involved in endocrine function (Birnbaum and Staskal, 2004; Darnerud et al., 2001), and biotransformation reactions (Holm et al., 1993; Tjarnlund et al., 1998). Although less studied relative to their mammalian counterparts, these mechanisms also appear to be relevant in aquatic species. For example, chronic exposure of sticklebacks (Gasterosteus aculeatus) to a commercial penta-BDE mixture (Bromkal 70-5DE) reduces spawning success and causes mild hepatic lipid accumulation (Holm et al., 1993), whereas chronic exposures of rainbow trout to tetra- and penta-BDE are associated with reductions in cytochrome P4501A and glutathione reductase activities, as well as perturbations in blood glucose and hematocrit (Tjarnlund et al., 1998). In addition, embryonic exposure of zebrafish to BDE 47 has been shown to cause developmental toxicity, and impair cardiovascular function in zebrafish (Lema et al., 2007). It is also apparent that oxidative stress is an important component of PBDE toxicity in vivo (Fernie et al., 2005; Lam et al., 2006; van Boxtel et al., 2008), as well as in vitro in primary human fetal liver hematopoietic stem cells (Shao et al., 2008b) and rat neuronal cells (Costa and Giordano, 2007; Giordano et al., 2008). However, oxidative stress encompasses a broad spectrum of subcellular perturbations and these mechanisms of PBDE toxicity have not well characterized on a cellular level in aquatic species.
We have previously reported that transformed gill and liver cells from rainbow trout increase reactive oxygen species production in the presence of BDE 47 (Shao et al., 2008a). In the current study, we have optimized several flow cytometry assays for use in RTgill-W1 gill epithelia cells to more closely investigate the mechanisms of cellular injury of a common PBDE congener found in fish. Fish gills receive continuous exposure to waterborne toxicants and comprise a first line of defense against the toxicity of aquatic chemicals (McKim et al., 1985; Wood et al., 2002) Gill epithelial cells in particular, perform important physiological functions such as gas exchange, ion transport, acid–base regulation, osmoregulation, and excretion of endogenous metabolic byproducts. Others have reported that gill epithelial cells are more sensitive than cultured hepatocytes to environmental toxicants (Zhou et al., 2006). Although primary fish gill cells more closely approximate in vivo function, the use of immortalized gill cell lines such as RTgill-W1 cells avoids logistical issues associated with microbial contamination as well as the variability among cell preparations which can arise from the use of individual animals (Bols et al., 1994). We focused on flow cytometry endpoints such as pyridine nucleotide redox status, mitochondrial cardiolipin content, mitochondrial membrane potential, and apoptosis to investigate the roles of mitochondrial injury and oxidative stress as mechanisms of BDE 47 toxicity.
Leibovitz’s L-15 media was purchased from Sigma-Aldrich (St. Louis, MO). Fetal Bovine Serum (FBS), heat-inactivated, was purchased from Atlanta Biologicals (Lawrenceville, GA). L-Glutamine and 0.05% Trypsin-EDTA were purchased from GibcoBRL (Life Technologies Inc., Gaithersburg, MD). Penicillin-Streptomycin was purchased from Fisher Scientific (Pittsburgh, PA). BDE 47 (2,2′,4,4′-tetrabromodiphenyl ether, >99% purity) was purchased from Chem Service, Inc. (West Chester, PA). DMSO (dimethylsulfoxide) was purchased from Sigma-Aldrich. Acridine orange 10-nonyl bromide (nonyl acridine orange, NAO), JC-1 (5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethyl-benzimidazolo-carbocyanineiodide), MitoTracker Red and propidium iodide (PI) were purchased from Invitrogen (Carlsbad, CA). Vented tissue culture flasks were purchased from Corning Inc. (Corning, NY). All other chemicals were purchased from either Sigma-Aldrich or Fisher Scientific.
RTgill-W1 cells originated from histologically normal gill lamella of a rainbow trout Oncorhynchus mykiss (Bols et al., 1994) were purchased from American Type Culture Collection (Manassas, VA). Gill cells were maintained without CO2 incubation in free gas exchange with air at 19°C in Leibovitz’s L-15 medium supplemented with 10% FBS (v/v), 2 mM L-glutamine, 100 IU/ml penicillin and 100 μg/ml streptomycin according to the manufacturer’s instructions. The medium was filtered through a Durapore 0.22 μm filter (Millipore, Billerica, MA).
All experimental treatments were carried out in serum-free L-15/ex solution (Lee et al., 2008). Cells were plated into 25 cm2 or 75 cm2 flasks at a density of 1 × 107 or 2 × 107cells, respectively, and allowed to attach and acclimate for 48 hours prior to chemical exposures. BDE 47 was dissolved in DMSO and prepared in L-15/ex solution for final concentrations of 0 μM (0.02% DMSO, throughout all experiments), 0.2 μM, 0.8 μM, 3.2 μM, 12.5 μM or 50 μM. These concentrations were based upon preliminary range-finding experiments establishing a dose response for cell injury and similar to other studies of PBDE toxicity in human, rodent and fish cells (He et al., 2008; Hu et al., 2007; Shao et al., 2008b). In all experiments, 2 mM hydrogen peroxide (H2O2) was used as positive control for cell toxicity and validation of flow cytometry assays. Following the appropriate exposure times, the cells were trypsinized and transferred to sterile 15 ml conical tubes. The samples were then centrifuged at 4°C for 5 min at 250 × g. The supernatant was removed and the pelleted cells were resuspended in L-15/ex solution for further analysis via flow cytometry.
A commonly used quantitative method for cell viability is propidium iodide (PI) staining. PI is a non-permeant dye which stains the DNA of dead or dying cells and thus viable cells with intact cell membranes exclude PI. We used PI in conjunction with flow cytometry and confirmed the effects of BDE 47 on cell viability using a biochemical assay based upon alamarBlue reduction (Shao et al., 2008a). AlamarBlue is an indicator dye which measures the proliferative ability of cells based upon oxidation-reduction reflecting mitochondrial function. Following BDE 47 exposures, media from each treatment (including cells detached during the incubation) was transferred to 15 ml polypropylene tubes. Gill cells which remained attached were trypsinized and combined with their respective “floater” population. These samples represented total cell populations for each treatment and were centrifuged at 150 × g for 5 min at 4°C prior to resuspension to a density of 1 × 106 cells/ml. The cells were then incubated with 1 μg/ml PI and placed at 4°C. PI positive populations were quantified via flow cytometry using 488 nm excitation and fluorescence emissions were collected using a 645 nm long pass filter.
Gill cells contain a variety of endogenous fluorescent molecules including aromatic amino acids, flavins and the reduced pyridine nucleotides NADH and NADPH. Excitation of NAD(P)H at ~360nm gives rise to fluorescent emission signal which peaks at 450 nm (blue) (del et al., 2008; Laing et al., 1980; Masters et al., 1989). While NAD(P)H molecules are fluorescent, the oxidized forms of NAD(P)H (NAD+ and NAD+P) lose their fluorescence characteristics (Laing et al., 1980). Accordingly, changes in NAD(P)H autofluorescence reflect alterations in cellular pyridine nucleotide redox status (Laing et al., 1980; Masters et al., 1989). NAD(P)H autofluorescence intensities were measured in conjunction with other indicators via flow cytometry on a Coulter EPICS Elite flow cytometer (Beckman Coulter, Miami, FL) using 355 nm (UV) excitation. The resulting blue fluorescence emission signals were collected using a 450/35 band-pass filter.
Cells undergoing apoptosis or necrosis display distinct physiological and morphological characteristics, which can be readily measured by flow cytometry (Darzynkiewicz et al., 1992; Tomer et al., 1988). One such technique involves analysis of the cell’s light scattering properties. Forward angle light scatter (FS) is a measure of cell diameter or size while side angle, or 90°, light scatter (SS) is a measure of cellular internal complexity (Darzynkiewicz et al., 1992). One morphological characteristic of early apoptosis is cellular shrinkage reflected by a decrease in the FS signal (Darzynkiewicz et al., 1992). As membrane integrity is not compromised in early apoptosis, the SS signal either remains unchanged, or can increase as a result of chromatin condensation or mitochondrial swelling. In contrast to apoptotic cells, cells undergoing necrosis display an initial increase in FS as the SS signal decreases. Accordingly, the effect of BDE 47 on cell morphology was determined by monitoring the FS and SS properties of treated cells in relation to DMSO (vehicle control) treated cells via flow cytometry. Briefly, a gate was placed around the untreated, control population in a FS vs. SS bivariate plot. Shifts in the FS and SS of BDE47 treated cell populations were then assessed, relative to the gate established for control cell populations.
The percentage of apoptotic cells was determined by propidium iodide (PI) staining of semi-permeabilized cells conducted as described previously with minor modifications (Nicoletti et al., 1991). In this assay, apoptotic cells were identified on the basis of hypodiploid DNA content that results from DNA fragmentation (Nicoletti et al., 1991; Pozarowski et al., 2004). Cells having DNA content less than those cells in the G1 phase of the cell cycle (sub-G1) were assumed to be apoptotic. In these experiments, approximately 1×106 gill cells from each treatment group were detached with Trypsin-EDTA for 5 min @ 37°C, transferred to 15 ml polypropylene tubes and centrifuged at 150 × g for 5 min at 4°C. The cell pellet was resuspended in 200 μl 1X phosphate buffered saline (PBS) at room temperature (Chemicon International, Temecula, CA) and cells were fixed by the addition of 2 ml of ice-cold 70% ethanol/30% PBS. The cells were incubated on ice for 30 min, centrifuged at 1000 × g for 5 min, and resuspended in 800 μl 1X PBS. The isolated nuclei were then incubated with 1 μl RNase H (2 U/μl) and 40 μl PI (1 mg/μl) in a 37°C water bath for 30 min, and filtered through a 70 micron pore filter prior to flow cytometry analysis. PI stained nuclei were excited at 488 nm and fluorescence emission was detected at >605 nm.
A decrease in mitochondrial membrane potential is often associated with mitochondrial injury and can serve as an early indicator of apoptosis. Two fluorescent dyes were used to determine the effects of BDE 47 on trout gill cell mitochondrial membrane potential. Specifically, JC-1 is a lipophilic, cationic dye, which selectively incorporates into mitochondrial membranes, exhibiting differential fluorescence characteristics dependent upon the mitochondrial membrane potential (Cossarizza et al., 1993). In polarized membranes, JC-1 molecules form red fluorescent J-aggregates. Upon membrane depolarization, these J-aggregates increasingly shift to a monomeric form exhibiting green fluorescence (Cossarizza et al., 1993). Briefly, approximately 1×106 gill cells from each treatment group were detached with Trypsin-EDTA for 5 min @ 37°C, transferred to 15 ml polypropylene tubes and centrifuged at 250 × g for 5 min at 4°C. The cell pellet was resuspended in 800 μl 1X PBS (Chemicon International, Temecula, CA). JC-1 was prepared in DMSO and added to the cell suspension to a final concentration of 1 μg/ml and incubated for 45 min at room temperature. The fluorescence intensities were acquired on a Coulter Elite flow cytometer (Beckman Coulter, Miami, FL) using 488 nm excitation. Green (JC-1 monomer) and red (JC-1 aggregate) fluorescence emission were acquired with 525/40-nm band-pass and 590-nm long-pass filters, respectively.
Changes in mitochondrial membrane potential were further evaluated using MitoTracker Red (CMXRos), a mitochondrion-specific probe, which passively diffuses across the plasma membrane and is sequestered into functional mitochondria (Poot and Pierce, 1999). Mitochondria with a decreased membrane potential are less able to retain the probe and are therefore less fluorescent. Approximately 1 × 106 gill cells from each treatment group were detached with Trypsin-EDTA for 5 min @ 37°C, transferred to 15 ml polypropylene tubes and centrifuged at 250 × g for 5 min at 4°C. The cell pellet was resuspended in 800 μl 1X PBS (Chemicon International, Temecula, CA). MitoTracker Red was prepared in DMSO and added to the cell suspension at a final concentration of 125 μg/ml and incubated for 15 min at 37°C. MitoTracker Red was excited with 488 nm, and fluorescence emission was collected using a 645 nm long pass filter.
Cardiolipin is an important component of the inner mitochondrial membrane, where it constitutes about 20% of the total lipid composition (Wright et al., 2004). Peroxidation of cardiolipin leads to a collapse of mitochondrial membrane integrity and release of cytochrome c from mitochondria into the cytosol (Petit et al., 1992; Wright et al., 2004; McMillin and Dowhan, 2002). The fluorescent dye acridine orange 10-nonyl bromide (nonyl acridine orange, NAO) binds cardiolipin with a higher affinity for cardiolipin than other molecules, thus providing a quantitative measure of mitochondrial lipid peroxidation and mitochondrial membrane integrity (Petit et al., 1992; McMillin and Dowhan, 2002; Wright et al., 2004). In this assay, approximately 1 × 106 gill cells from each treatment group were collected for NAO staining. NAO (Chemicon International, Temecula, CA) was prepared in DMSO and diluted 1:4 with culture medium to a concentration of 125 μg/ml and 1.0 μl of NAO working stock was added to cells that were resuspended in 800 μl 1X PBS prior to incubation for 30 min at 37°C. NAO fluorescence was determined by flow cytometry using 488 nm excitation, and a 645 nm long pass filter to measure fluorescence emission.
Experimental values for all oxidative stress and viability indices represent the mean of a minimum of triplicate replications performed in three experiments. All statistical analyses were done using GraphPad Prism version 4.0c for Macintosh, (GraphPad Software, San Diego, California USA). Treatment-related effects on cell injury parameters were assessed using one-way Analysis of Variance (ANOVA) followed by Dunnett’s post-hoc test. Treatment-related differences were considered statistically significant at p ≤ 0.05.
Treatment of RTgill-W1 cells with low concentrations of BDE 47 (<3.2 μM) led to a substantial increase in the adherence of cells to culture dishes. At these dosages, collection of cells required the addition of trypsin over incubation, likely as a result of increases in the production of gill cell surface proteins that modulate cell-cell and cell-extracellular matrix interactions (Luo et al., 2007). Presented in Figure 1 is a dose-response of the acute toxicity of BDE 47 in rainbow trout cells. As observed, BDE 47 caused a dose-dependent increase in PI staining relative to DMSO controls. However, the effects of BDE 47 were only statistically significant at the 50 μM exposure concentration. As shown in Figure 2, significant decreases in NAD(P)H autofluorescence were observed in live cells at concentrations of 12.5 and 50 μM BDE 47. Specifically, incubation of cells with 12.5 μM BDE 47 led to a 45% decrease in NAD(P)H autofluorescence, a concentration below those that resulted in acute cell death. Exposure to 50 μM BDE 47, a concentration that was associated with significant cell toxicity, resulted in a 68% decrease in NAD(P)H autofluorescence in live cells .
As observed in Figure 3, exposure to 12.5 μM BDE 47, a concentration that was not associated with acute cell toxicity caused a moderate reduction in the FS signal and a concomitant increase in the SS signal, suggesting the presence of chromatin condensation and mitochondrial swelling. Treatment with the highest dose (50 μM) caused a more dramatic decrease in FS signal and a significant increase in SS signal, strongly suggesting stimulation of the early phase of apoptosis. To further examine whether BDE 47-induced cell death was mediated through apoptosis, we analyzed the DNA content of the RTgill cells by flow cytometry. As shown in Figure 4, at 6 hours post BDE 47 exposure, the sub-G1 cell populations were 10% and 22% for the 12.5 μM and 50 μM BDE 47 concentrations, respectively. At 24 hours post exposure, the observed apoptotic (sub-G1) populations were significantly increased at lower exposure levels (12% increase at 3.2 μM BDE 47) and apoptopic cells constituted more than 75% of total gill cell populations at both the 12.5 μM and 50 μM concentrations.
In contrast to our previous study in human cells, we were unable to acquire a reliable J-aggregate (red signal) using JC-1 (Shao et al., 2008b). This may have been due to several factors, including; an intrinsically lower mitochondrial potential of gill cells relative to those from other tissues (O’Dowd et al., 2006) which would not favor the formation of JC-1 aggregates, a low lipid composition of the gill cell membranes precluding the formation of J-aggregates, or quenching of red fluorescence of the aggregates when formed. However, we observed decreased membrane potential associated with an increase in JC-1 green fluorescence on exposure to BDE 47 and thus report the increased in JC-1 monomer (green signal) alone as reflective of mitochondrial depolarization. As shown in Figure 5A, exposure to BDE 47 at doses below 50 μM did not change the percentage of cells exhibiting JC-1 green fluorescence, which in these cells is associated with an intact mitochondrial redox potential. These effects were most pronounced at the highest dose of BDE 47 where a 71% increase in JC-1 green fluorescence was observed. As shown in Figure 5B, a 56% decrease in MitoTracker Red staining was observed in cells exposed to 50 μM BDE 47, further indicating a reduction in mitochondrial membrane potential and consistent with the results obtained using JC-1. The reduction in mitochondrial membrane potential at 50 μM BDE 47 was associated with an increase in mitochondrial lipid peroxidation as evidenced by decreased NAO fluorescence (3% stained cells, Figure 6), compared to control cells.
We have previously reported dose-response characteristics associated with BDE 47 toxicity in transformed rainbow trout gill and liver cells (Shao et al., 2008a). The dose-response curves previously observed for BDE 47 were strikingly similar to other studies using mammalian primary and transformed cells exposed to either BDE 47 or to PBDE congener mixtures (Costa and Giordano, 2007; Giordano et al., 2008; Shao et al., 2008b; Yu et al., 2008), suggesting that these agents may have similar mechanisms of toxicity in fish and mammalian cells. However, the biochemical events underlying the progression of BDE 47 cell injury in fish have not been well characterized. The application of flow cytometry approaches and use of fluorescently labeled probes that sensitively detect losses of cell function provides a valuable approach to test the hypothesis that the acute toxicity of BDE 47 in trout gill cells is associated with mitochondrial oxidative injury leading to apoptosis. To our knowledge, the flow cytometry probes used in the present study have not been applied to studies using fish cells, at least in a comprehensive approach to identify toxicants that disrupt cellular redox status.
In the current study, the use of the PI, a dye that binds to double-stranded DNA, enabled us to determine the effect of BDE 47 on the percentage of gill cells in the sub G1 phase of the cell cycle while concomitantly allowing for comparison of results obtained by flow cytometry and biochemically. We conducted the two viability assays (i.e., PI and alamarBlue) under the same conditions with similar results (Shao et al., 2008a). The BDE 47-induced loss of NADPH autofluorescence was consistent with biochemical changes associated with the loss of mitochondrial energetics using the alamarBlue reduction assay that reported in our earlier studies (Shao et al., 2008a; Shao et al., 2008b). Because the alamarBlue assay measures the metabolic integrity of the mitochondria based upon the function of mitochondrial reductases (Ahmed et al., 1994), the rate of alamarBlue reduction reflects oxidation-reduction activity of respiratory chain components in mitochondria, a major intracellular source of ROS (Lee and Wei, 2007). Therefore, our observations of decreased mitochondrial cardiolipin content and mitochondrial membrane potential on exposure to low concentrations of BDE 47 is consistent with the generation of oxidative stress within the mitochondria as a mechanism of cell injury, and indicates that mitochondria are a subcellular target of PBDEs in fish.
A close examination of the flow cytometry data reveals the nature of mitochondrial perturbations caused by BDE 47. For example, the BDE 47-induced reduction in the FS signal, and concomitant increase in the SS signal, reflects morphological characteristics associated with the onset of apoptosis. The results obtained by using the JC-1 fluorescent assay were consistent with those by MitoTracker Red indicating a loss of mitochondrial redox potential which is typically correlated with apoptosis. These data are consistent with analysis of sub G1 DNA content analysis by flow cytometry which indicated that exposure to BDE 47 increased the percentage of cells undergoing frank apoptosis. Endonucleases activated during apoptosis cleave sections of internucleosomal DNA, causing extensive DNA fragmentation, a primary indicator of apoptosis (Nagata, 2000). As a result of DNA fragmentation and chromatin condensation, apoptotic cells show reduced DNA staining with PI, reflecting lower quantitative DNA content relative to G1 phase nuclei (Gong et al., 1994). The shedding of apoptotic bodies containing fragments of nuclear chromatin may further contribute to the loss of DNA from apoptotic cells (Pozarowski et al., 2004). Accordingly, the emergence of sub-G1 peaks seen with increasing levels of BDE 47 treatment indicates that these cells were undergoing apoptosis. These results are consistent with other studies in primary human fetal liver hematopoietic stem cells (Shao et al., 2008b) and primary mouse neurons and astrocytes (Giordano et al., 2008) which also demonstrated that exposure to PBDEs lead to cellular injury through induction of apoptotic cell death.
A substantial body of work has shown that mitochondrial membrane depolarization plays an important role in regulating apoptosis (Green and Kroemer, 2004; Marchetti et al., 1996). Chemical exposures can destabilize the mitochondrial membrane and cause a loss of mitochondrial integrity and cell survival (Loh et al., 2006; Zorov et al., 2006). In the current study, we observed a marked loss of mitochondrial membrane potential associated with the degradation of cardiolipin, a marker of mitochondrial membrane lipid oxidation, in gill cells treated with BDE 47, further supporting our hypothesis that BDE 47 targets the mitochondria and stimulates apoptosis via mitochondrial oxidative injury (Lugli et al., 2005). Cardiolipin is a phospholipid located in the inner mitochondrial membrane, that when oxidized, plays a crucial role in cytochrome C release from mitochondria leading to apoptosis (Basova et al., 2007). Normally, cytochrome C is localized to the outer surface of the inner mitochondrial membrane via electrostatic interactions with anionic lipids such cardiolipin (Demel et al., 1989). BDE 47 mediated cardiolipin oxidation likely disrupted the interaction between cytochrome C and cardiolipin, followed by its dissociation from the membrane and escape through the Bax-mediated permeabilized outer mitochondrial membrane (Nomura et al., 2000; Ott et al., 2002; Petrosillo et al., 2001). It is also possible that BDE 47 triggered mitochondrial swelling and a rupture of the outer mitochondrial membrane (as reflected by light scattering measurements), which could have then caused cytochrome C release (Goldstein et al., 2000), leading to the activation of caspase-dependent and independent mitochondrial death pathways (Antonsson, 2004).
It is important to note that all of the fluorescently-labeled commercial dyes used in this study required substantial optimization methodologies during assay validation. Furthermore, in addition to the battery of cell indicators discussed (i.e. JC-1, MitoTracker, NAO, PI staining, NADPH autofluorescence), we initially investigated other probes, including BODIPY FL C11, diphenyl-1-pyrenylphosphine (DPPP) and Annexin-V-Fluorescein (Annexin-V-FL). Specifically, BODIPY FL C11 is a lipophilic fluorophore that localizes to cell membranes and emits a fluorescence profile reflective of the oxidative state of the plasma membrane. Accordingly, as BODIPY FL C11 becomes oxidized, there is a concomitant shift from the red fluorescent excimer form to a green fluorescent monomeric form associated with peroxidation state of the membrane. By contrast, Annexin-V-FL is a phospholipid-binding protein with strong affinity for phophotidylserine (PS). During the early stages of apoptosis, PS is translocated from the inner part of the plasma membrane to the external surface of the cell. Thus, Annexin-V-FL binds exposed PS, providing a direct measure of apoptotic events. Although these aforementioned probes were used effectively in our studies addressing the mechanisms of BDE 47 injury to human fetal liver hematopoietic stem cells (Shao et al., 2008b), they yielded inconsistent results in trout gill cells. The unsuccessful application of some of these flow cytometry probes may have been a result of the gill cell phospholipid content, which appears to be relatively low in trout gill cell membranes and can be further reduced during thermal acclimation (Hazel and Carpenter, 1985; Kraffe et al., 2007). Furthermore, the fact that the trout gill cells were maintained at 19°C, a higher than normal physiological temperature for rainbow trout, may have further decreased the content of unsaturated phospholipids allowing for detection of lipid peroxidation by BODIPY FL C11 and DPPP. The lower mitochondrial potential of rainbow trout gill cells relative to other tissues may have contributed to the inability of JC-1 to form aggregates (O’Dowd et al., 2006). Although we were not able to capture the red JC-1 signal by flow cytometry, the observed increase of the corresponding JC-1 green signal is similar to results observed using fetal liver hematopoietic stem cells, and in the current study using the MitoTracker Red dye.
As discussed, the dosing range of BDE 47 used in this study (3.2 μM -50 μM) allowed for comparison to other studies and were reasonable given that a only a single acute dose was employed, likely underestimating the environmental scenario involving chronic exposures. This range of exposures encompassed somewhat higher BDE 47 levels that caused cytotoxicity as well as lower BDE 47 levels that elicited subtle biochemical changes that allowed us to clearly evaluate with mechanisms of toxicity. The cellular effects observed with low micromolar concentrations of BDE 47 suggests a moderate sensitivity of trout gill cells to this, and potentially to other, PBDE congeners. BDE 47 toxicity to primary hepatocytes from rainbow trout show similar sensitivities which may be partially dependent upon the poor detoxification capabilities of the cells (Nakari and Pessala, 2005). Although the role of detoxification gene expression has not been clearly identified in detecting against PBDE toxicity, a number of genes critical to cellular protection are down-regulated in the liver of rainbow trout exposed to BDE 47 in vivo (Hook et al., 2006). The sensitivity of trout gill epithelial cells to BDE 47 may be due to their poor ability to detoxify reactive oxygen species via cellular antioxidant pathways (Shao et al., 2008a). However, we did not characterize the baseline expression of antioxidant enzymes in the gill cells, or determine if BDE 47 exposure modulates the expression of genes that protect against oxidative stress. These studies are ongoing in our laboratory.
In summary, BDE 47 is a predominant PBDE congener detected in fish that causes injury to cultured rainbow trout gill cells that is associated with increased cellular oxidative stress, mitochondrial injury, and apoptosis. Although the resulting cell death occurs at a level that is above PBDE concentrations encountered in the environment, we have not examined the effect of multiple or chronic exposures which may be more relevant to environmental exposures. Furthermore, the fact that BDE 47 bioaccumulates in fish relative to the other higher molecular weight congeners (Betts, 2002; Browne et al., 2009; Sjodin et al., 2003) is important in extrapolating our in vitro results to more relevant environmental scenarios. Collectively, our studies also indicate that oxidative stress-associated biomarkers may be useful in assessing the sublethal effects of PBDEs in fish, as well as in other species. To this end, the application of flow cytometry-based analyses and fluorescent probes that are sensitive markers of cell injury from aquatic species are of utility for broader use in the field of aquatic toxicology.
This work was supported by grants from the National Oceanic and Atmospheric Association (NOAA) Oceans and Human Health Initiative NA05NOS4781253, the National Institute of Environmental Health Sciences (NIEHS) Superfund Basic Sciences Research Program P42-ES-04696, and the UW NIEHS sponsored Center for Ecogenetics and Environmental Health, P30ES07033.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.