Article pubs.acs.org/est
Disruption of Oxidative Phosphorylation (OXPHOS) by Hydroxylated Polybrominated Diphenyl Ethers (OH-PBDEs) Present in the Marine Environment Jessica Legradi,†,§ Anna-Karin Dahlberg,‡,§ Peter Cenijn,† Göran Marsh,‡ Lillemor Asplund,‡ Åke Bergman,‡ and Juliette Legler*,† †
Institute for Environmental Studies, VU University Amsterdam, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands Environmental Chemistry Unit, Department of Materials and Environmental Chemistry, Stockholm University, SE-106 91 Stockholm, Sweden
‡
S Supporting Information *
ABSTRACT: Hydroxylated polybrominated diphenyl ethers (OH-PBDEs) are of growing concern, as they have been detected in both humans and wildlife and have been shown to be toxic. Recent studies have indicated that OH-PBDEs can be more toxic than PBDEs, partly due to their ability to disrupt oxidative phosphorylation (OXPHOS), an essential process in energy metabolism. In this study, we determined the OXPHOS disruption potential of 18 OH-PBDE congeners reported in marine wildlife using two in vitro bioassays, namely the classic rat mitochondrial respiration assay, and a mitochrondrial membrane potential assay using zebrafish PAC2 cells. Single OH-PBDE congeners as well as mixtures were tested to study potential additive or synergistic effects. An environmental mixture composed of seven OH-PBDE congeners mimicking the concentrations reported in Baltic blue mussels were also studied. We report that all OH-PBDEs tested were able to disrupt OXPHOS via either protonophoric uncoupling and/ or inhibition of the electron transport chain. Additionally we show that OH-PBDEs tested in combinations as found in the environment have the potential to disrupt OXPHOS. Importantly, mixtures of OH-PBDEs may show very strong synergistic effects, stressing the importance of further research on the in vivo impacts of these compounds in the environment.
1. INTRODUCTION Hydroxylated polybrominated diphenyl ethers (OH-PBDEs) are of growing concern, as they have been reported in both humans and wildlife and have been shown to be toxic. OHPBDEs originate from natural production or via metabolic biotransformation of anthropogenically produced polybrominated diphenyl ethers (PBDEs).1 OH-PBDEs have been detected in mussels,2,3 marine fish,4,5 freshwater fish,6,7 birds8,9 and marine mammals,10,11 as well as humans.12−16 Naturally produced OH-PBDEs are mainly found in the marine environment, where they are considered to be synthesized by marine organisms such as marine bacteria,17−19 filamentous macroalgae18−20 and marine sponges.21−23 Naturally produced OH-PBDEs identified so far share a common structural characteristic in that the hydroxyl group is attached in ortho position relative to the diphenyl ether bond,4 while OH-PBDEs formed via metabolic transformation from anthropogenic PBDEs tend to have the hydroxyl group in either meta or para position,24 with a few known exceptions (e.g., 2′-OHBDE28, 2′-OH-BDE66, 6-OH-BDE47 studied herein).24,25 PBDEs have been used as additive flame retardants in various materials and produced since the 1970s in three technical © 2014 American Chemical Society
mixtures (i.e., Penta-, Octa- and DecaBDE mixtures) depending on their degree of bromination. However, PBDEs have shown to display reproductive toxicity, developmental neurotoxicity and endocrine disrupting effects as reviewed by Legler et al.,26 and today commercial Penta- and OctaBDE are included in the Stockholm Convention (under Annex A)27 and banned within the European Union28 and the use of DecaBDE is restricted.29 In the United States, Penta- and OctaBDEs were voluntarily phased out in 2004 and several companies committed to phase out DecaBDE by the end of 2013.30 Nevertheless, due to their extensive use, PBDEs have now become ubiquitous to our environment. Recent studies have indicated that OH-PBDEs can be more biologically active and potentially more toxic than PBDEs.26 6OH-BDE47 has been reported to bind to transthyretin (TTR), the thyroid hormone transporting protein,31 thus potentially disrupting thyroid hormone homeostasis. 6-OH-BDE47 has Received: Revised: Accepted: Published: 14703
August 15, 2014 October 27, 2014 November 12, 2014 November 25, 2014 dx.doi.org/10.1021/es5039744 | Environ. Sci. Technol. 2014, 48, 14703−14711
Environmental Science & Technology
Article
Supporting Information) of which several have been found in the Baltic Sea (Table S1, Supporting Information). To study OXPHOS disruption, we used two in vitro bioassays: the classic rat mitochondrial respiration assay (TPP assay) and a mitochondrial potential assay in zebrafish PAC2 cells (TMRM assay). The TMRM assay provides a rapid means to screen large number of compounds whereas the traditional TPP assay provides more mechanistic information regarding the mode of OXPHOS disruption (i.e., protonophoric uncoupling and/or inhibition). They were complemented with a 24 h cytotoxicity test and a LDH leakage test. Single OH-PBDEs congeners as well as mixtures were tested in the TMRM assay to study potential mixture effects. An environmental mixture composed of seven OH-PBDE congeners mimicking the concentrations in Baltic blue mussels (Mytilus edulis) reported by Loefstrand et al.3 was also tested using the TMRM assay.
also been reported to inhibit aromatase activity in human adrenocortical carcinoma cells32 and placental cells.33 Furthermore, 6-OH-BDE47 has been shown to disturb calcium (Ca2+) homeostasis in pneochromocytoma cells34 and to cause cytotoxic effects (e.g., reduced cell viability, increased apoptosis rate, block cell cycle and DNA damage) and oxidative stress in human hepatoma cells.35 DNA damage has also been reported in chicken cell lines, and suggested to be caused by oxidative base damage through generation of reactive oxygen species (ROS).36 Cytotoxicity, elevated ROS levels and altered oxidative stress responses have been seen in human hepatocyte cells after exposure to 6-OH-BDE85, which structurally resembles 6-OH-BDE47.37 Recent studies have reported 6OH-BDE47 to be acutely toxic to developing and adult zebrafish38 and induces developmental arrest in zebrafish.39 The potency to induce developmental arrest in zebrafish has also been shown for 3-OH-BDE47 and 5-OH-BDE47.39 Interestingly, van Boxtel and co-workers showed that 6-OH-BDE47 is a potent disrupter of mitochondrial oxidative phosphorylation (OXPHOS), a mode of action that could explain the acute toxicity observed in zebrafish.38 OXPHOS is the metabolic pathway used to produce energy in aerobic conditions. It is based on the electron transport chain (ETC), which is located in the inner mitochondrial membrane where it is coupled with adenosine triphosphate (ATP) synthase, which phosphorylates adenosine diphosphate (ADP) to ATP.40,41 The ETC consists of four complexes (I− IV) that transfer electrons via a series of redox reactions. This electron transfer by ETC results in an efflux of protons from the mitochondrial matrix into the intermembrane space, generating a proton gradient called mitochondrial membrane potential. The energy stored in this electrochemical gradient is utilized to power the ATP synthase.40,41 Many compounds are known to affect the rate of formation of ATP via disruption of OXPHOS.41 Some compounds act by dissipating the electrochemical gradient across the mitochondrial membrane via transfer of protons from the intermembrane space to the mitochondrial matrix or interfering with the mitochondrial membrane structure, working against the proton gradient created by the ETC and abolishing the link between substrate oxidation and ATP synthesis.42 To describe proton translocation across the mitochondrial membrane by protonophoric uncouplers both monomolecular and bimolecular (phenol/phenolate complex) models have been proposed.43 Disruption of OXPHOS can also be mediated by xenobiotics through inhibition of the electron transporting complexes, which depending on the inhibited complex, will hamper or stop the electron transfer and proton efflux, thus preventing an electrochemical gradient to be formed.38 6-OH-BDE 47, for example, has been shown to disrupt OXPHOS via inhibition of complex II.38 Rapid depletion of ATP due to complete disruption of OXPHOS can be acutely toxic whereas a low dose exposure over a prolonged period of time can lead to less efficient energy metabolism, which can cause severe body weight loss and increased thermogenesis.41 Given the reported OXPHOS disruption capability of 6-OHBDE47, it is important to assess the OXPHOS disruption potential of other OH-PBDE congeners present in the marine environment. In the eutrophicated Baltic Sea, a large number of ortho substituted OH-PBDEs of likely natural origin have been identified in marine biota. In this study, we determined the OXPHOS disruption potential of 18 OH-PBDE congeners of both natural and anthropogenic (metabolic) origin (Figure S1,
2. EXPERIMENTAL SECTION 2.1. Chemicals. 2′-Hydroxy-2,4,4′-tribromodiphenyl ether (2′-OH-BDE28), 2′-hydroxy-2,3′,4,4′-tetrabromodiphenyl ether (2′-OH-BDE66), 2′-hydroxy-2,3′,4,5′-tetrabromodiphenyl ether (2′-OH-BDE68), 2′-hydroxy-6′-chloro-2,3′,4,5′-tetrabromodiphenyl ether (2′-OH-6′-Cl-BDE68), 2-hydroxy2′,3,4,4′,5-pentabromodiphenyl ether (2-OH-BDE123), 3hydroxy-2,2′,4,4′-tetrabromodiphenyl ether (3-OH-BDE47), 3-hydroxy-2,2′,4,4′,5,5′-hexabromodiphenyl ether (3-OHBDE153), 3′-hydroxy-2,2′,4,4′,5,6′-hexabromodiphenyl ether (3′-OH-BDE154), 3-hydroxy-2,2′,4,4′,6,6′-hexabromodiphenyl ether (3-OH-BDE155), 4′-hydroxy-2,2′,4-tribromodiphenyl ether (4′-OH-BDE17), 5-hydroxy-2,2′,4,4′-tetrabromodiphenyl ether (5-OH-BDE47), 6-hydroxy-2,2′,4,4′-tetrabromodiphenyl ether (6-OH-BDE47), 6-hydroxy-5-chloro-2,2′,4,4′-tetrabromodiphenyl ether (6-OH-5-Cl-BDE47), 6′-hydroxy-2,2′4,5′tetrabromodiphenyl ether (6′-OH-BDE49), 6-hydroxy2,2′,3,4,4′-pentabromodiphenyl ether (6-OH-BDE85), 2,2′,3,4′,5-pentabromodiphenyl ether (6-OH-BDE90), 6-hydroxy-2,2′,4,4′,5-pentabromodiphenyl ether (6-OH-BDE99), 6hydroxy-2,2′,3,4,4′,5-hexabromodiphenyl ether (6-OHBDE137) and 2,2′,4,4′-tetrabromodiphenyl ether (BDE-47) were synthesized in-house as described elsewhere.16,44,45 Their purities were >98%, as determined by gas chromatography− mass spectrometry (GC−MS) using electron ionization (EI) as described in Rydén et al.16 Carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP) (≥98% purity), 3,5-di(tert-butyl)-4-hydroxybenzylidenemalononitrile (SF6847), 2,4-dinitrophenol (DNP, Pestanal grade) and pentachlorophenol (PCP, 98% purity) were purchased from Sigma-Aldrich (St. Louis, MO, USA). All compounds were dissolved in dimethyl sulfoxide (DMSO, Arcos Organics, Geel, Belgium) and stored in darkness at room temperature. 2.2. Triphenylphosophonium (TPP) Assay. Animals were sacrificed according to ethical procedures and the livers immediately collected and kept in cold 0.9% NaCl(aq). The rat liver was homogenized and the mitochondria were isolated and prepared as described elsewhere.38 The mitochondrial respiration and inner membrane potential were measured simultaneous by incubating the mitochondria in reaction medium in a closed and stirred acrylic vessel at 25 °C. The container was equipped with a Clarke-type oxygen electrode and a tetraphenylphosphonium ion (TPP+) electrode connected to a reference electrode. Coupling between respiration (O2), membrane potential and (Δψ) and ATP production in the 14704
dx.doi.org/10.1021/es5039744 | Environ. Sci. Technol. 2014, 48, 14703−14711
Environmental Science & Technology
Article
normalized setting the 100% effect levels to the same level as the FCCP control. Concentration addition mixture effects were analyzed by calculating toxic units. One toxic unit (TU) corresponds to the sum of the actual exposure concentrations divided by the EC10’s (Table S3, Supporting Information) or EC50’s (Table S4, Supporting Information) of the compounds. When substances act via the same mode of action according to concentration addition, the EC50 of the mixture should be one TU. The same holds for EC10’s. Calculations for the mixture experiments were performed as described by Van Gestel et al.46 Mixture effects based on independent action were calculated as described by Hadrup et al.47 Independent action is commonly assumed when compounds act via different modes of action. Therefore, toxicity of a mixture can be calculated statistically from the responses to the individual substances. All calculations were performed using Graphpad Prism 5. Four (2′-OH-6′-Cl-BDE68, 6-OH-BDE47, 6-OH-5-ClBDE47 and 6-OH-BDE99) and three compound mixtures (without 6-OH-BDE 47) were combined at equal potent EC10 and EC50 levels and tested in dose−response experiments covering six dilutions in order to obtain the mixture EC10 and EC50 (Table S3, Supporting Information). In addition to mixtures based on fixed ratio of compounds of equal potency, an environmentally relevant mixture was also studied. The concentrations of OH-PBDEs in this mixture was selected based on reported levels in Baltic blue mussels by Loefstrand et al.3 (Table S5, Supporting Information). OH-PBDEs in this blue mussel mix were tested at levels 1- to 1000-times abovereported concentrations (assuming; 1 g of mussel fresh weight = 1 mL). 2.4. MTT Cytotoxicity Test. To determine the effects of OH-PBDEs on mitochondrial-mediated toxicity, the MTT assay was applied. PAC2 cells were exposed for 24 h to the compounds and the amount of formazan measured with a fluorescence reader (SpectraMax Gimini EM fluorescence reader, Molecular Devices). A day before the start of exposure 30 000 PAC-2 cells in 100 μL per well were plated in a flat bottom 96-well plate (Greiner Bio-One). Exposure medium was prepared diluting compounds in growing media (as described for the TMRM assay) to reach a final concentration in the well of 50 μM. The next day, 100 μL of exposure media was added and the plate incubated at 28°. After 24 h, 20 μL of a solution containing 5.5 mg/mL methylthiazolyldiphenyltetrazolium bromide (MTT) (Sigma-Aldrich) dissolved in phosphate buffer saline (PBS) from Oxoid limited (Basingstoke, Hampshire, UK) was added to the wells. After 1 h, the media was removed and the formazan crystals where dissolved in 100 μL of DMSO (Arcos Organics). After shaking, the absorbance was measured at 562 nm. Tributyltin at a concentration of 10 μM was used as a positive control and DMSO as a solvent control. Each compound was tested with five replicates. 2.5. LDH Leakage. To ensure that OXPHOS disruption in the TMRM assay was not due to nonspecific cytotoxicity caused by other mechanisms than disruption of OXPHOS, a subset of 16 compounds was tested for LDH leakage. The PAC2 cells were cultured and plated as described for the TMRM assay at concentrations shown in Figure S5 (Supporting Information). The cell culture media was replaced with 100 μL of test media and the cells were incubated at 28 °C for 45 min (without CO2 supplement). After incubation, 50 μL of the exposure media was transferred to an empty 96-well plate with clear bottom, and 100 μL of sodium pyruvate (Sigma-
isolated mitochondria was demonstrated by addition of succinate (1 mM) and ATP (10 mM) to initiate OXPHOS. Rotenone was added to block complex I (NADH-UQ reductase) of the ETC, which allows to study inhibitory effects on the ETC as well as protonophoric uncoupling in parallel. All test compounds were added after the membrane potential was stable (i.e., when state 4-respiration had been reached) and each compound was measured at different concentrations. Four classical disrupters of OXPHOS; FCCP, SF6847, PCP and DNP were included as positive controls and BDE-47 as a negative control. All data were recorded digitally using a PowerLab/4SP system connected to Chart v. 3.6s software (ADInstruments Pty Ltd., United Kingdom). The lowest observed effect concentration (LOEC) was defined as the lowest concentration where a reduction of the membrane potential in combination with an effect (positive or negative) on the oxygen consumption was detected. The analysis of the recorded measurements was performed using Excel software (2007). The different modes of action were identified by protonophoric uncoupling causing a decrease in membrane potential and an increase in respiration rate in a dosedependent manner, while inhibition of ETC reduced the membrane potential and decreased the respiration rate (as exemplified in Figure S2, Supporting Information). 2.3. TMRM Assay. The zebrafish embryonic fibroblast (PAC2) cells were kindly provided by Professor H. Spaink, Leiden University, The Netherlands. The cells were plated in a 96-well black polystyrene plate with clear bottom (Greiner BioOne, The Netherlands) at a density of 30 000 cells/well (100 μL). The test media was prepared by diluting the test compounds in assay medium (50 mM HEPES, 150 mM KCl, 160 mM NaCL, 50 mM D-glucose, pH 7.4). All ingredients of the assay medium were purchased from Sigma-Aldrich. The nominal concentration of test compound in the test media is given in Table S2 (Supporting Information). 100 μL of test medium was added to the plated cells and plates were incubated at 28 °C for 30 min without CO2 supplement. 20 μL of tetramethylrhodamine methyl ester perchlorate (TMRM) (≥95% purity; Sigma-Aldrich) was added to the exposure medium with a final concentration of 500 nM. The plate was incubated for 20 min at room temperature. Afterward, 20 μL of 16 mM Brilliant Black (0.14 gr/10 mL assay medium; SigmaAldrich) was added. Fluorescence was measured immediately from the bottom on a fluorescence reader (SpectraMax Gimini EM Fluorescence reader, Molecular Devices, CA, USA) at an excitation wavelength of 530 nm and emission wavelength of 580 nm using SoftMax Pro 5.2 Software (Molecular Devices, CA, USA). The mean fluorescence and standard deviation was calculated from at least three experiments, each performed in triplicate. A FCCP dose−response curve was used as positive control throughout the study, along with solvent blanks. Dose− response curves were fitted and half maximal effect concentration (EC50) values were calculated using Graphpad Prism 5 (Graph Pad Software, La Jolla, USA). The lowest observed effect concentration (LOEC) was determined using a Kruskal−Wallis test (P < 0.05) to check for normality followed by a One-Way ANOVA with a bonferroni posthoc test (P < 0.05) using Graphpad Prism 5. Individual and mixture dose response experiments were carried out within the same experiment. All tested mixtures had a maximum total DMSO concentration below 0.5%. For each experiment, FCCP was applied as positive control and the data 14705
dx.doi.org/10.1021/es5039744 | Environ. Sci. Technol. 2014, 48, 14703−14711
Environmental Science & Technology
Article
Table 1. Potency of OH-PBDEs to Uncouple and/or Inhibit Oxidative Phosphorylation in the TPP- and TMRM Assay TPP assayb
compound FCCP SF6847 PCP DNP 2′-OH-BDE28 2′-OH-BDE66 2′-OH-BDE68 2′-OH-6′-Cl-BDE68 2-OH-BDE123 3-OH-BDE47 3-OH-BDE153 3′-OH-BDE154 3-OH-BDE155 4′-OH-BDE17 5-OH-BDE47 6-OH-BDE47 6-OH-5-Cl-BDE47 6′-OH-BDE49 6-OH-BDE85 6-OH-BDE90 6-OH-BDE99 6-OH-BDE137 BDE-47
pKaa 5.9 ± 7.6 ± 4.7 ± 4.0 ± 7.8 ± 6.2 ± 6.6 ± 5.7 ± 5.7 ± 6.1 ± 5.1 ± 5.5 ± 5.5 ± 8.8 ± 7.1 ± 6.8 ± 5.3 ± 6.2 ± 6.2 ± 5.7 ± 5.2 ± 4.7 ± n/ad
0.1 0.4 0.3 0.2 0.4 0.4 0.5 0.5 0.5 0.2 0.2 0.3 0.3 0.2 0.2 0.4 0.5 0.4 0.5 0.5 0.5 0.5
uncoupling
inhibition
a
LOEC (μM)
LOEC (μM)
LOEC (μM)
EC50 (μM)
0.6 0.4 0.4 0.2 0.9 0.9 0.9 1.0 1.0 0.6 0.7 0.7 0.7 0.6 0.6 0.9 1.0 0.9 1.0 1.0 1.0 1.0 0.6
0.05 0.02 0.05 1.0 3.8 1.0 0.5 1.0 0.7 0.01 2.5 nee 3.2 4.4 0.1 ne 0.4 0.8 0.01 1.3 0.5 1.0 ne
1.0 0.38 2.0 80 16 10 5.0 3.8 3.8 2.0 8.8 1.3 32 34 ne 0.02 3.8 4.0 1.8 ne 0.7 2.5 ne
0.75 2.5 30 n.d.f n.d. 50 n.d. 12 12 n.d. 50 20 n.d. 75 n.d. 6.0 16 50 8.0 20 14 10 n.d.
1.0 2.1 33 >90 >40 39 >100 19 15 >100 56 28 >100 60 100 6.0 21 41 5.6 17 27 21 >100
log Kow 3.7 4.6 5.1 1.7 5.6 6.3 6.7 7.2 7.2 6.2 7.4 7.5 7.3 4.9 6.4 5.9 6.5 5.9 6.6 6.8 6.6 7.4 6.7
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
TMRM assayc altered membrane potential
a
Calculated using Advanced Chemistry Development (ACD/Laboratories) software v11.02. bValues based one experiment (n = 1). cValues based on at least three individual experiments (n ≥ 3), each performed in triplicate. dNot applicable (n/a). eNo effect (ne). fNot detected (n.d.).
3. RESULTS
Aldrich) dissolved in phosphate buffer saline (PBS) from Oxoid limited (Basingstoke) to 1.5 mM was added, followed by 100 μL of nicotinamide adenine dinucleotide (NADH) from Applichem (Darmstadt, Germany) dissolved in phosphate buffer saline (PBS) to 0.5 mM. The wells plate was then measured on a SpectraMax 340pc (Molecular Devices) measuring the absorbance at a wavelength of 340 nm. SoftMax Pro 5.2 software (Molecular Devices) was used as the computer program. Compounds were tested in triplicate and DMSO was used as a control. 2.6. ATP Measurements. To illustrate that a decrease in MMP results in a decrease in cellular ATP, the ATP level in PAC2 cells was measured after exposure to FCCP (0.1, 2.8, 50 μM) and 6-OH-BDE137 (0.1, 27, and 100 μM). The PAC2 cells were cultured and plated in as described for the TMRM assay. The cell culture media was replaced with 100 μL of test media and the cells were incubated at 28 °C for 45 min (without CO2 supplement). After incubation, the test medium was discarded and 50 μL ATP reagent SL (Cellular ATP Kit HTS, BioThema, Sweden) containing D-luciferin and luciferase added. Luciferase catalyzes the oxidation of D-luciferin to the luminescent product oxyluciferin via dephosphorylation of adenosine triphosphate (ATP) to adenosine monophosphate (AMP). The bioluminescence was measured using a Lucy 2 luminometer (Anthos Labtec Instruments, Salzburg, Austria). Each concentration was tested in triplicate and DMSO was used as a control. The relative decrease in light emission with increasing exposure concentration compared to control was calculated.
3.1. TPP Assay. All studied OH-PBDEs disrupt OXPHOS either by protonophoric uncoupling and/or via inhibition of ETC within the concentration range tested (Table 1). The potency to disrupt OXPHOS differed by a factor of 100, with 3OH-BDE47 (LOEC = 0.01 μM), 6-OH-BDE47 (LOEC = 0.02 μM) and 6-OH-BDE85 (LOEC = 0.01 μM) being the most potent ones, with effect concentrations in the same order of magnitude as the positive controls FCCP (LOEC = 0.05 μM) and SF6847 (LOEC = 0.02 μM) (Table 1). 6-OH-BDE47 and 3′-OH-BDE154 (LOEC = 1.3 μM) were found to act solely as inhibitors of ETC. 5-OH-BDE47 (LOEC = 0.1 μM) and 6OH-BDE90 (LOEC = 1.3 μM) were found to act only via protonophoric uncoupling within the concentration range tested whereas all other OH-PBDEs were found to disrupt OXPHOS through both protonophoric uncoupling and inhibition, hence exhibiting two modes of action. This double mode of action was also noted for FCCP, SF6847 and PCP (LOEC = 0.05 μM) whereas DNP (LOEC=1.0 μM) was observed to inhibit only ETC at higher (80 μM) concentrations. The parent compound BDE-47 did not disrupt OXPHOS at test concentrations up to 6 μM. 3.2. TMRM Assay. Because TMRM accumulates in the mitochondrial matrix in proportion to its mitochondrial membrane potential, the measured TMRM fluorescence was expected to decrease when the MMP was disrupted (as shown for FCCP and 6-OH-BDE47 in Figure S3, Supporting Information). The repeatability of the TMRM method was evaluated using FCCP. Six individual experiments, each performed in triplicate, resulted in an average EC50 value for 14706
dx.doi.org/10.1021/es5039744 | Environ. Sci. Technol. 2014, 48, 14703−14711
Environmental Science & Technology
Article
FCCP of 1.0 ± 0.05 μM, showing low intratest variation (Figure S3a, Supporting Information). Based on the dose−response curves obtained for the test compounds (shown in Figure S4, Supporting Information), LOEC and EC50 values were calculated (Table 1). 6-OHBDE47 was the most potent disrupter of OXPHOS among the OH-PBDEs tested (LOEC = 6.0 μM; EC50 = 6.0 μM). The structurally related compound 6-OH-BDE85 (LOEC = 8.0 μM; EC50 = 5.6 μM) was almost as potent as 6-OH-BDE47. These values were about 6 to 8 times higher than those for FCCP (LOEC = 0.75 μM; EC50 = 1.0 μM) and about 3 times higher than those for SF6847 (LOEC = 2.5 μM; EC50 = 2.1 μM). Furthermore, 2′-OH-BDE28, 2′-OH-BDE66, 2′-OH-6′-ClBDE68, 2-OH-BDE123, 3-OH-BDE153, 3′-OH-BDE154, 4′OH-BDE17, 6-OH-5-Cl-BDE47, 6′-OH-BDE49, 6-OHBDE90, 6-OH-BDE99 and 6-OH-BDE137 disrupted OXPHOS in PAC2 cells, but with less potency than 6-OH-BDE85 and 6OH-BDE47. No disruption of OXPHOS was found in the TMRM assay with 2′-OH-BDE68, 3-OH-BDE47, 3-OHBDE155, 5-OH-BDE47 and BDE-47, within the concentration range tested. 3.3. Cytotoxicity of OH-PBDEs. None of the tested compounds showed an increase in LDH leakage (Figure S5, Supporting Information). In the MTT assay, 2′-OH-BDE28, 2′OH-BDE68, 2′-OH-6′-Cl-BDE68, 2-OH-BDE123, 5-OHBDE47, 6-OH-BDE47, 6-OH-5-Cl-BDE47, 6-OH-BDE85, 6OH-BDE90 and 6-OH-BDE99 showed signs of mitochondrial toxicity with a decrease in formazan production compared to the control at 50 μM (Figure S6, Supporting Information). 2′OH-BDE66, 3-OH-BDE47, 6-OH-BDE137 and BDE-47 showed no significant cytotoxic effect whereas 3-OHBDE153, 3′-OH-BDE154, 3-OH-BDE155, 4′-OH-BDE17 and 6′-OH-BDE49 actually showed an increased mitochondrial metabolism (i.e., increased formazan production compared to the control. 3.4. ATP Measurements. FCCP showed a dose-dependent decrease in ATP compared to the control with 50 μM showing an 82% decrease (Figure S7, Supporting Information). 6-OHBDE137 showed a 67% decrease in ATP at 100 μM whereas 27 μM showed no difference in ATP level compared to the control. 3.5. OXPHOS Disruption of OH-PBDE Mixtures. Mixture experiments were performed with 2′-OH-6′-ClBDE68, 6-OH-BDE47, 6-OH-5-Cl-BDE47 and 6-OH-BDE99, compounds that were positive in the TPP and TMRM assay and therefore showed potential to disrupt OXPHOS. 2′-OH-6′Cl-BDE68, 6-OH-5-Cl-BDE47 and 6-OH-BDE99 were selected based on their similar concentration−response curves in the TMRM assay (Figure 1). 6-OH-BDE47 was selected as one of the most potent compounds tested, with a slightly different shape of concentration−response curve (Figure 1). Fixed ratio mixtures based on EC50 levels were tested in the TMRM assay. The calculated mixture EC50 was 0.55 TU (95% confidence interval 0.45−0.65), which is significantly lower than 1, suggesting strong synergistic effects. The potency of the mixture dose−response curve (Figure 2a) differed from the curve shape of the individual compounds, indicating that the toxicity of the mixture at low and high concentrations may be different from that at EC50 levels. For this reason, we also conducted a comparison of the mixture effects at EC10 levels, which resulted in a mixture EC10 of 0.34 TU (95% confidence interval 0.25−0.43) (Figure 2a). This low TU level indicated an
Figure 1. Concentration−response curves for the effect on mitochondrial membrane potential of the four individual OH-PBDEs used in the mixture experiments in the TMRM assay. The experiments were performed in triplicate and the result is presented as mean ± standard deviation of the induced effect.
even stronger synergistic effect of the OH-PBDEs at lower concentrations. We also investigated the toxicity of mixtures without 6-OHBDE47, because this compound was active only via inhibition in the TPP assay, and was clearly more potent than the other compounds. The concentration addition mixture model was applied to EC50 and EC10 fixed ratio mixtures of 2′-OH-6′-ClBDE68, 6-OH-5-Cl-BDE47 and 6-OH-BDE99. The three compound mixture EC50 was 0.96 TU, representing unity, whereas the EC10 was 0.22 (95% confidence interval 0.10− 0.53), indicating synergistic effects at lower concentrations (Figure 2b). Because most OH-PBDEs tested acted as both uncouplers and inhibitors in the TPP assay depending on their concentration, we hypothesized that the concentration addition mixture model may not be appropriate to use as it assumes that all compounds in the mixture have the same mode of action. For this reason, we also applied the mixture model of independent action (IA) on the four compound mixture. Mixture effects exceeded the predicted effects based on IA (Figure 3), again confirming the very strong synergistic effects. In the Baltic blue mussel OH-PBDE mixture experiment, disruption of OXPHOS (decrease in MMP in the TMRM assay) was found with mixtures 100 and 1000 times more concentrated than the reported levels in blue mussels (Figure 4). It should be noted that the concentrations used for the 100 times blue mussel mix were around or below the measured NOEC based on individual compounds (Table S5, Supporting Information). Application of the concentration addition mixture model resulted in a mixture EC50 of 0.22 TU, showing an even higher synergistic effect than with the four compound study based on equal potency. These results seem to indicate that the synergistic effect increases with increasing number of OH-PBDEs. The mixture model of independent action could not be applied due to the low concentrations used (
