TY - JOUR
AU - Wallace, K. B.
AB - Abstract Perfluorooctanoate (PFOA) and perfluorooctanesulfonate (PFOS) are thought to induce peroxisome proliferation and interfere with mitochondrial metabolic pathways. Direct measurements revealed that PFOA and the unsubstituted sulfonamide of perfluorooctane (FOSA) uncouple mitochondrial respiration by increasing proton conductance. The purpose of this investigation was to characterize structural determinants responsible for the mitochondrial uncoupling effect of several structurally related fluorochemicals. Included in the study were PFOA, PFOS, FOSA, the N-acetate of FOSA (perfluorooctanesulfonamidoacetate, FOSAA), N-ethylperfluorooctanesulfonamide (N-EtFOSA), and the N-ethyl alcohol [2-(N-ethylperfluorooctanesulfonamido)ethyl alcohol, N-EtFOSE] and N-acetic acid (N-ethylperfluorooctanesulfonamidoacetate, N-EtFOSAA) of N-EtFOSA. Each test compound was dissolved in ethanol and added directly to an incubation medium containing substrate-energized rat liver mitochondria. Mitochondrial respiration and membrane potential were measured concurrently using an oxygen electrode and a TPP+-selective electrode, respectively. All of the compounds tested, at sufficiently high concentrations, had the capacity to interfere with mitochondrial respiration, albeit via different mechanisms and with varying potencies. At sufficiently high concentrations, the free acids PFOA and PFOS caused a slight increase in the intrinsic proton leak of the mitochondrial inner membrane, which resembled a surfactant-like change in membrane fluidity. Similar effects were observed with the sulfonamide N-EtFOSE. Another fully substituted sulfonamide, N-EtFOSAA, at high concentrations caused inhibition of respiration, the release of cytochrome c, and high-amplitude swelling of mitochondria. The swelling was prevented by cyclosporin A or by EGTA, indicating that this compound induced the mitochondrial permeability transition. The unsubstituted and mono-substituted amides FOSA, N-EtFOSA, and FOSAA all exerted a strong uncoupling effect on mitochondria resembling that of protonophoric uncouplers. Among these compounds, FOSA was a very potent uncoupler of oxidative phosphorylation, with an IC50 of approximately 1 μM. These data suggest that the protonated nitrogen atom with a favorable pKa is essential for the uncoupling action of perfluorooctane sulfonamides in mitochondria, which may be critical to the mechanism by which these compounds interfere with mitochondrial metabolism to induce peroxisome proliferation in vivo. mitochondria, perfluoroalkanes, structure-activity, peroxisome proliferator Perfluorinated carboxylic and sulfonic acids have long been used in a wide variety of commercial applications, including water repellants on assorted fabrics, flame retardants and extinguishers, surfactants, waxes and gloss finish enhancers, anticorrosion agents and lubricants. Though once considered to be biologically inert, detection of nonionic fluorine and perfluorooctanoate (PFOA) in human sera (Gilliland and Mandel, 1996; Olsen et al., 1999; Taves, 1968), combined with evidence of cachexia in experimental animals exposed to PFOA, has raised concerns of possible long-term health effects of various derivatives of perfluorinated acids. The classic effects of both acute and subchronic administration of PFOA, perfluorodecanoic acid (PFDA), or perfluorooctanesulfonate (PFOS) are a reduction in food consumption (hypophagia), weight loss, and hepatomegaly due primarily to parenchymal cell swelling (Borges et al., 1992; Pastoor et al., 1987; Vahden Heuvel, 1996; Van Rafelghem et al., 1987). Histological examination of the liver of PFOA- and PFDA-exposed rats reveals proliferation of mitochondrial membranes, the endoplasmic reticulum, and peroxisomal bodies (Goecke-Flora and Reo, 1996; Pastoor et al., 1987; Sohlenius et al., 1992, 1993), which are associated with proportionate increases in protein content and enzyme activities of the respective subcellular fractions (Goecke-Flora and Reo, 1996; Harrison et al., 1988; Kawashima et al., 1995; Pastoor et al., 1987; Permadi et al., 1992, 1993; Witzmann et al., 1994). The wasting syndrome, or cachexia, observed with perfluorinated acid exposures in experimental animals is a dose-limiting event and attributed to a metabolic disorder at the cellular level. Biochemical manifestations of PFOA, PFDA, and PFOS exposure are indicative of altered lipid metabolism and include a lowering of serum triglycerides and cholesterol and the accumulation of lipid droplets in liver cells (Haughom and Spydevold, 1992; Pastoor et al., 1987; Sohlenius et al., 1993; Vahden Heuvel, 1996). Enzyme analysis reveals an inhibition of HMG-CoA reductase and acyl-CoA:cholesterol acyltransferase activities in livers of rats fed either PFOA or PFOS (Sohlenius et al., 1993). Histomorphology reveals an expansion of peroxisomal bodies that is accompanied by the stimulation of peroxisome-related enzyme activities such as acyl CoA oxidase and catalase (Pastoor et al., 1987; Permadi et al., 1992, 1993; Sohlenius et al., 1992; Vahden Heuvel, 1996; Van Rafelghem et al., 1987). From a structure-activity standpoint, longer-chain perfluorinated fatty acids are more potent than short-chain acids (Goecke-Flora and Reo, 1996). In view of these observations, perfluorinated acids have been categorized as peroxisome proliferators and hypothesized to act as structural mimics of fatty acids, thereby being competitive inhibitors of mitochondrial β-oxidation. The reported increase in mitochondrial protein content in liver of exposed rats may represent either a condensation and change in buoyancy of individual mitochondria (Permadi et al., 1993) or a compensatory mitochondrial biogenesis. Regardless, the mitochondrion has emerged as a primary intracellular target for perfluorinated acid-induced hepatotoxicity. In addition to the expansion of the mitochondrial fraction within individual hepatocytes (Pastoor et al., 1987; Permadi et al., 1992, 1993; Sohlenius et al., 1992), a number of perfluorinated acids have been reported to uncouple mitochondrial respiration. For example, PFOA and PFDA stimulate state 4 respiration and uncouple state 3 respiration in isolated rat liver mitochondria (Keller et al., 1992; Langley, 1990). This uncoupling of mitochondrial respiration has also been demonstrated for isolated perfused liver from rats exposed in vivo to PFOA in their diet (Keller et al., 1992). Schnellmann and Manning (1990) demonstrated that perfluorooctane sulfonamide (FOSA) is a protonophoric uncoupler of oxidative phosphorylation in isolated rat kidney mitochondria and suggested that this may account for the nephrotoxicity caused by the insecticide sulfluramid (N-ethyl-FOSA). Based on the results to date, it can be proposed that the critical molecular event that mediates perfluorinated carboxylic and sulfonate-induced interference with hepatic lipid metabolism and stimulation peroxisome proliferation is that they act as protonophoric uncouplers of mitochondrial respiration. On a molecular basis, this requires certain physical chemical features of the individual compounds: They must be sufficiently hydrophobic to traverse the inner mitochondrial membrane, possess an ionizable atom with a pKa in the range of 5–7, and contain a fairly strong electron-withdrawing moiety (Wallace and Starkov, 2000). The purpose of the current investigation was to more thoroughly describe the specific mechanisms by which the various perfluorooctanes interfere with mitochondrial bioenergetics and to identify the key structural/physical chemical features responsible for these activities. Structures of the compounds selected for this investigation are illustrated in Table 1. MATERIALS AND METHODS Mitochondria were isolated from liver of adult male Sprague-Dawley rats (∼200 g body weight) by conventional differential centrifugation techniques (Zhou and Wallace, 1999). Animals were killed by decapitation and the liver was immediately excised, weighed, and cooled in 40 ml isolation medium (210 mM mannitol, 10 mM sucrose, 5mM HEPES-KOH [pH 7.4], 1 mM EGTA). The cooled liver was then minced with scissors and washed twice with 20 ml isolation medium, then diluted with the same medium and homogenized for 1 min with a motor-driven Teflon/glass Potter homogenizer. The tissue-to-medium ratio was 1:8 (g:ml). The homogenate was filtered through gauze and centrifuged for 10 min at 700 × g and 4°C. The mitochondrial pellet was then recovered from the supernatant by centrifugation at 10,000 × g for 10 min. The pellet was resuspended in 10 ml washing medium (210 mM mannitol, 10 mM sucrose, 5 mM HEPES-KOH, pH 7.4) supplemented with bovine serum albumin (BSA, 1 mg/ml). The mitochondrial suspension was diluted to 35 ml with the same medium without BSA and centrifuged at 10,000 × g for 10 min. The final mitochondrial pellet was resuspended in washing medium to a protein concentration of 70–80 mg/ml and stored on ice. Mitochondrial membrane potential (ΔΨ) was estimated from TPP+ ion distribution measured with a TPP+- selective electrode constructed according to Kamo et al. (Kamo et al., 1979). Mitochondrial membrane potential was calculated as described elsewhere (Custodio et al., 1998; Rottenberg, 1984). The rate of oxygen consumption by mitochondria was measured with a Clark-type oxygen electrode. Both the mitochondrial membrane potential and the respiration rate were recorded simultaneously using a multichannel incubation chamber equipped with a magnetic stirrer. The volume of the chamber was 1.8 ml. All experiments were performed at room temperature (25°C). The TPP+-sensitive electrode was calibrated by sequential additions of known amounts of TPP+Cl− before the addition of mitochondria (Fig. 1). Preliminary experiments were performed to ensure that the ethanol, pH, or the compounds themselves did not interfere with the TPP+ or oxygen electrodes. Respiration rates were calculated assuming the initial oxygen concentration to be 240 μM. Protein concentration was determined by the Bradford assay (Bradford, 1976) using BSA as a standard. All perfluorinated test compounds were dissolved in absolute ethanol (except for PFOA, which was dissolved in deionized water). Preliminary experience revealed that all compounds are very hydrophobic and tend to precipitate in the incubation medium. Therefore, a set of dilutions was made for each compound to obtain an array of concentrations from 6.25 to 100 μM. The perfluorinated test compounds were added to mitochondria as 1.8 μl of the appropriate stock solution to give a final ethanol concentration of 0.1%. Reagents. All perfluorinated compounds were synthesized, characterized, and provided gratis by The 3M Company, St. Paul, MN. Ultra Pure sucrose was purchased from ICN Biomedicals, Inc. (www.icnbiomed.com), and all other reagents were from Sigma-Aldrich (www.sigma-aldrich.com). BSA was essentially fatty acid–free. Statistical analysis. All experiments were repeated at least three times using freshly isolated hepatic mitochondria from separate animals for each experimental repetition. The results were analyzed by the Student's paired t-test, using a probability of p < 0.05 as the criterion for statistical significance. RESULTS Figure 1 illustrates the progress of a typical experimental run testing the effects of FOSA on mitochondrial respiration and membrane potential, measured concurrently. All test chemicals were subjected to the same experimental protocol. The experiment began by making sequential additions of TPP+ (up to a final total concentration of 2 μM) to a MOPS-buffered high ionic strength medium supplemented with phosphate and glutamate plus malate, but lacking mitochondria. This allowed for calibration of the TPP+ electrode (lower trace). The reaction was initiated by adding mitochondria to a final concentration of approximately 1 mg protein/ml, which caused an immediate redistribution of TPP+ from the incubation medium into the mitochondria (upward pen deflection), accompanied by the initiation of a slow rate of oxygen consumption indicative of state 4 respiration. Once a steady state was established, ADP (100 μM) was added to initiate state 3 respiration, which was transient and reverted to state 4 respiration once all of the added ADP had been phosphorylated. Induction of state 3 is indicated by the abrupt but brief increase in respiration rate coupled with the transient depolarization of membrane potential, which repolarized upon return to state 4. From this initial series of rates, we calculated the respiratory control ratio (RCR; state 3/state 4) and the ADP/O ratio according to standard procedures (Estabrook, 1967). With the reaction continuing, we then added the test chemical at the desired final concentration and allowed the reaction to proceed to establish a linear state 4 respiration rate. This was followed by a second addition of ADP and the recording of state 3, from which we calculated the posttreatment RCR and ADP/O values. The reaction was terminated by adding sufficient 2,4-dinitrophenol (DNP) to completely uncouple mitochondrial respiration, which is reflected by the complete depolarization of mitochondrial membrane potential and exhaustion of dissolved oxygen. Data from each experimental run were collected and tabulated for statistical comparisons (Table 2). Significant drug effects were determined by paired analyses of the rates recorded before and after addition of the respective test compound. The advantage of this repeated measures method is that it provides a well-controlled process for evaluating the effects of test agents on the fundamental properties of mitochondrial performance. With this method, it is possible to detect disturbances in mitochondrial function caused by inhibitors of the electron transport chain, membrane transporters, or the Na+/K+-ATPase activity, as well as the effects of either uncouplers or other less-specific acting agents that increase membrane ion conductance. For the example illustrated, FOSA caused an immediate decrease in mitochondrial membrane potential, which was associated with an increase in both state 4 and state 3 respiration rates and a decrease in the RCI from 5.8 to 4.2. The results of this initial screen for effects by the various test compounds on mitochondrial respiration are compiled in Table 2. Initial titration experiments were performed to establish the concentration for each compound that elicits a detectable effect without completely uncoupling mitochondrial respiration. In the case of uncoupling agents, the concentration was set as that which causes a significant stimulation of state 4 respiration rate. The concentration for each compound is listed in the table, inspection of which gives an indication of the relative potency of each. The purpose of the screen was to explore the potential for an effect on mitochondrial bioenergetics, not to reveal specific mechanisms of action. The general conclusions from these screens are: With the exception of the carboxylate and sulfonate (PFOA and PFOS, respectively), all of the perfluorochemicals tested were capable of decreasing the coupling efficiency (RCI) of rat liver mitochondria in vitro. PFOA, PFOS, and the fully saturated amide, N-EtFOSE, are relatively weak mitochondrial toxicants. The secondary amides FOSAA and N-EtFOSA stimulate state 4 respiration and uncouple oxidative phosphorylation at concentrations near 5 μM. In contrast, the primary amide FOSA is an extremely potent uncoupler of mitochondrial oxidative phosphorylation. Additional experiments were then performed to investigate the specific molecular mechanism by which each of these agents interfere with mitochondrial respiration. The first to be studied was the very potent uncoupler FOSA. Figure 2 illustrates the results of an experiment wherein both mitochondrial membrane potential and state 4 respiration were titrated with successive additions of FOSA. Each successive addition of FOSA caused a progressive depolarization of membrane potential and stimulation of state 4 respiration. The fact that this mimics very closely the response observed for agents such as 2,4-dinitrophenol (DNP) is highly suggestive of a protonophoric uncoupling effect, which is precisely what was reported by Schnellmann and Manning (Schnellmann and Manning, 1990). Similar responses to increasing concentrations of the secondary amides N-EtFOSA and FOSAA were also observed (data not shown), suggesting that they too might be classified as protonophoric uncouplers. To help verify whether these three compounds elicit an uncoupling effect comparable to DNP, the data were analyzed in terms of quantifying the dependence of state 4 respiration on membrane potential (Fig. 3). As would be expected, depolarization of mitochondrial membrane potential caused by successive additions of DNP was associated with proportionate increases in the rate of respiration. This linear relationship held true over a 50-mV range of membrane potentials, which indicates that none of these compounds directly inhibit the mitochondrial electron transport chain. The fact that similar relationships for membrane potential–dependent mitochondrial respiration existed for FOSA, N-EtFOSA, and FOSAA suggests that all three agents are DNP-like protonophoric uncouplers of mitochondrial respiration (Fig. 3). The fact that they share the characteristic of being primary or secondary amides suggests that it is the protonated nitrogen atom that is key to this mechanism of uncoupling of mitochondrial respiration by the perfluorooctanesulfonamides. PFOA, PFOS, N-EtFOSE, and N-EtFOSAA did not demonstrate a similar linear relationship between respiration rate and membrane potential. The relative potencies of the uncouplers of oxidative phosphorylation are illustrated in Figure 4, where the stimulated rate of state 4 respiration is plotted as a function of the concentration of uncoupler added to the reaction chamber, using DNP as a reference standard. The data reveal that on a molar basis, both FOSA and N-EtFOSA are much stronger uncouplers of mitochondrial respiration than is DNP. FOSAA was only a weak uncoupler, whereas N-EtFOSA was nearly twice as potent as DNP, and FOSA was 6–7 fold more potent as an uncoupler of mitochondrial respiration. Although the data are not shown, we found FOSA to be nearly equipotent as carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP), which is one of the most potent “classical” uncouplers of mitochondrial respiration known. Repeating these experiments in a low ionic strength buffer (225 mM mannitol, 5 mM HEPES [pH 7.4], 4 mM KH2PO4) yielded qualitatively the same results (data not shown), indicating that this uncoupling of respiration is not a secondary effect due to induction of the mitochondrial permeability transition by FOSA. As suggested by Figure 4, at sufficiently high concentrations the relationship between respiration rate and uncoupler concentration begins to deviate from linearity. In fact, sufficiently high concentrations of N-EtFOSAA actually inhibit mitochondrial respiration (data not shown), which also occurs with high concentrations of DNP. Using FOSAA as a prototype, we found that the mechanism of inhibition of state 3 respiration by 80 μM FOSAA reflects an interference with mitochondrial integrity (swelling). Further experiments with the agents that were found not to be specific uncouplers of mitochondrial respiration revealed that at sufficiently high concentrations, PFOA, PFOS, and N-EtFOSE caused a small increase in resting respiration rate and slightly decreased the membrane potential, with no effect on the phosphorylating respiration or on the coupling efficiency of mitochondria. This might be tentatively attributed to induction of a slight increase in intrinsic proton leak of the mitochondrial inner membrane due to changes in its fluidity, which is a surfactant-like effect of these agents. The basis for the differential effect on states 3 and 4 respiration is that the intrinsic proton leak is membrane potential-dependent and thus greatest during state 4 when membrane potential is greatest. Adding ADP to stimulate state 3 respiration causes a transient depolarization of membrane potential (Fig. 1), thereby reducing protonmotive force responsible for the proton leak (Nicholls 1974a,b; 1997). This reduction in protonmotive force was reflected by the decrease in resting (state 4) membrane potential observed with these three compounds (data not shown). A particularly interesting observation was that at concentrations greater than 25 μM, the di-substituted sulfonamide N-EtFOSAA caused strong inhibition of state 3 respiration. Figure 5 illustrates this effect, where it is shown that adding 50 μM N-EtFOSAA inhibited state 3 respiration by 50%. The fact that N-EtFOSAA inhibited both succinate- and glutamate/malate-supported respiration suggests that N-EtFOSAA does not directly inhibit complex I or complex II of the respiratory chain. Adding DNP at a concentration that completely uncouples mitochondrial oxidative phosphorylation did not stimulate respiration beyond that recorded for N-EtFOSAA by itself. The fact that adding cytochrome c back to the reaction chamber restored respiration indicates that the inhibition by N-EtFOSAA is due to its causing the release of mitochondrial cytochrome c. Adding the cytochrome back to the reaction restored cytochrome oxidase activity (complex IV) and reestablished respiratory capability to the mitochondria. Such an effect suggests that N-EtFOSAA may be acting as an inducer of the mitochondrial permeability transition pore (MPTP). To test this, we next observed the effects of N-EtFOSAA on mitochondrial swelling. Figure 6 demonstrates that 25 μM N-EtFOSAA induces a very rapid permeabilization of the mitochondrial membranes and that adding cyclosporin A (CsA), which is a potent and specific inhibitor of the MPTP, provides complete protection against N-EtFOSAA-induced mitochondrial swelling. We observed similar effects when 2.4 mM EGTA was added to the reaction. The fact that N-EtFOSAA did not inhibit ascorbate-TMPD–supported respiration (data not shown) is evidence that it does not have a direct effect on cytochrome c binding to the inner mitochondrial membrane. It is well known that in high ionic strength buffer, mitochondrial swelling and physical disruption of the outer membrane due to the mitochondrial permeability transition results in the release of cytochrome c from the intermembrane space. When these same experiments were repeated in a low ionic strength buffer [210 mM mannitol, 10 mM sucrose, 5 mM HEPES-Tris (pH 7.4, n mM MgCl2)], state 4 respiration was not inhibited even at concentrations of up to 100 μM N-EtFOSAA (data not shown). Essentially similar results were obtained with 80 μM FOSAA (Fig. 7), indicating that it shares this same property of inducing the mitochondrial permeability transition in vitro, albeit at higher concentrations than those that cause uncoupling of mitochondrial oxidative phosphorylation (5–40 μM). In this case, induction of the mitochondrial permeability transition is assessed on the basis of the physical disruption of the outer mitochondrial membrane, as assessed by the reduction of exogenously added ferricytochrome c. Mitochondrial swelling and cytochrome c reduction followed the same time course to yield mirror images of the traces for absorbance at 550 and 540 nm, respectively (top two traces in panels A and B). Adding FOSAA caused an almost immediate decrease in light scattering and an increase in absorbance at 550 nm, which was not obvious until azide was added to inhibit the reoxidation of cytochrome c by cytochrome oxidase. In panel A, azide was added prior to FOSAA, and thus the changes in light scattering and cytochrome c were apparent immediately after adding FOSAA. In panel B, however, azide was not added until after three additions of FOSAA. In the absence of azide, FOSAA caused no apparent change in either light scattering at 540 nm or cytochrome c reduction at 550 nm. Subsequent addition of azide, however, resulted in an immediate increase in absorbance at 550 nm, which was associated with a mirror image decrease in absorbance at 540 nm. These effects of FOSAA on mitochondrial swelling and disruption of the outer mitochondrial membrane are accentuated by plotting the difference between absorbance at 550 nm and 540 nm, which mathematically accounts for changes in light scattering to give a more quantitative measure of cytochrome c reduction. DISCUSSION Though once considered to be biologically inert, the perfluorinated alkyl acids have since been found to cause metabolic wasting in some, but not all, species; rats and mice are sensitive, whereas guinea pigs and primates are not. Signs of intoxication include decreased body weight and food consumption associated with a fairly substantial hepatomegaly. Indications for this being a metabolic disorder are the proliferation of peroxisomal and possibly mitochondrial and sarcoplasmic membranes; the accumulation of both lipid droplets and glycogen stores in the individual liver parenchymal cells; and the lowering of serum glucose, cholesterol, and triglycerides. In view of the pivotal role of mitochondria in fatty acid, cholesterol, and glucose metabolism, the mitochondrion has been identified as a suspected target of perfluorinated alkane toxicity. However, few definitive studies have been performed to date. The data presented herein indicate that at sufficiently high concentrations, all of the tested perfluorinated octanyl acids and their derivatives are able to increase the nonselective permeability of mitochondrial membranes. This should not be surprising in view of the intended use of many of these agents as surfactants and surface-active water repellants. This is particularly true for PFOA and PFOS, which demonstrated no other specific effects on mitochondrial bioenergetics. Whether these effects have any physiological relevance remains to be determined and relies on the accumulation of the individual perfluorochemical in the targeted membrane at concentrations sufficient to interfere with membrane structure and fluidity. Besides this general, nonspecific effect observed at high concentrations, selected perfluorooctanes caused fairly specific and sometimes highly potent effects on mitochondrial bioenergetics. One of the more common effects shared by the primary and secondary amides is the uncoupling of rat liver mitochondrial oxidative phosphorylation, apparently via a protonophoric mechanism similar to that caused by DNP. This same effect has been previously reported for PFOA and PFDA, but at much higher concentrations (150 μM–0.5 mM; Keller et al., 1992; Langley, 1990). We found that the ionizable perfluorooctanesulfonamides (FOSA, FOSAA, and N-EtFOSA) elicit a strong uncoupling of mitochondrial oxidative phosphorylation at concentrations of 5–50 μM. Such an effect is evident from the stimulation of state 4 respiration combined with the linear relationship between respiration rate and membrane potential. Schnellmann and Manning (1990) demonstrated specific protonophoric uncoupling of rabbit renal cortical mitochondria with low concentrations of FOSA (1–25 μM) and suggested that this might mediate the cytotoxicity of the insecticide sulfluramid (N-EtFOSA), which is metabolized in part to FOSA. Similar mechanisms have been implicated for the herbicide perfluidone (1, 1, 1-trifluoro-N-[2-methyl-4-(phenylsulphonyl)-phenyl]methanesulphonamide, which required only 5 μM to cause a doubling of state 4 respiration and a 50% decrease in RCR (Olorunsogo and Malomo, 1985; Olorunsogo et al., 1985). It is suggested that the ionizable amide, with a favorable pKa, shuttles protons back into the mitochondrial matrix, thereby dissipating the protonmotive force generated by the electron transport chain. The fully substituted amides N-EtFOSE and N-EtFOSAA, which lack the protonated amide, were found not to be uncouplers of mitochondrial respiration. Of particular concern is the striking potency of FOSA in uncoupling rat liver mitochondrial respiration. As illustrated in Figure 4, FOSA is approximately five times more potent than DNP in stimulating state 4 respiration in isolated rat liver mitochondria. As little as 0.5 μM FOSA caused significant stimulation of mitochondrial respiration. This is of the same order of potency as FCCP, which is one of the most potent uncouplers of mitochondrial oxidative phosphorylation currently known. This concentration of FOSA translates to 0.25 ppm, a concentration that under certain conditions could conceivably occur from occupational or environmental exposures. The other interesting finding is that high concentrations of FOSAA and N-EtFOSAA inhibit mitochondrial respiration by causing the release of cytochrome c from the inner mitochondrial membrane, thereby inhibiting the activity of cytochrome oxidase (COX). Replenishing with exogenous cytochrome c restored both COX activity and respiration rate. The basis for this is that both compounds induce what is known as the mitochondrial permeability transition (MPT), a phenomenon whereby the exquisitely controlled permeability of the inner mitochondrial membrane is lost and the mitochondria become nonselectively permeable to solutes of up to 1.5 kD (Bernardi et al., 1999; Petronilli et al., 1994; Zoratti and Szabo, 1995). Associated with this is the rapid equilibration of solutes across the mitochondrial membranes, leading to depolarization of the membrane potential and osmotic swelling. This, in turn, is proposed to cause physical disruption of the outer mitochondrial membrane to permit the release of cytochrome c and other apoptogenic factors that reside in the intermembrane space (Kroemer and Reed, 2000). It is the accompanying inhibition of ATP synthesis and/or release of apoptogenic factors that has been the basis for implicating induction of the MPT in numerous and assorted cytopathies, including both chemical-induced as well as ischemia/reperfusion injuries (Lemasters et al., 1998; Wallace et al., 1997). Whether induction of the MPT occurs in response to exposures to the perfluorooctanesulfonamides in vivo and its role in mediating any ensuing metabolic dysfunction or tissue injury has yet to be defined. In summary, the perfluorooctanes were found to elicit three basic effects on mitochondrial bioenergetics. All of the compounds at sufficiently high concentrations caused a nonspecific increase in ion permeability of the mitochondrial membrane, an effect that resembles that caused by membrane detergents. At much lower concentrations, the ionizable amides caused a specific and potent uncoupling of mitochondrial oxidative phosphorylation. In the case of FOSA, the potency compares with some of the most potent uncouplers of mitochondrial respiration known to date. The third mechanism by which selected perfluorooctanesulfonamides interfere with mitochondrial bioenergetics is by inducing the MPT, leading to inhibition of ATP synthesis and release of apoptogenic factors. In view of the primary role of mitochondria in glucose, fatty acid, and cholesterol metabolism in the cell, it is possible that any one of these effects could account for the metabolic disorders and cachexia observed in conjunction with perfluorooctane exposure to rats in vivo. Whether mitochondrial dysfunction is a factor in assessing human health risks associated with perfluorooctanyl exposure depends on whether similar effects are manifested in vivo and on the absence of remarkable species-related differences in sensitivity of the response. TABLE 1 Structures of the Perfluorooctanes View Large TABLE 2 Effect of Fluorochemicals on Mitochondrial Respiration Compound  μM  State 3  State 4  RCI  (%)  ADP/O  (%)  Note. Values represent the mean ± SEM for 3–6 individual mitochondrial preparations. Values for State 3 and State 4 are given in nmol O2/min/mg protein. Standard rates for state 4 and state 3 respiration in the absence of test compound were 19.5 ± 2.2 nmol O2/min/mg protein and 89.5 ± 8.6 nmol O2/min/mg protein, respectively. Concentrations of the individual fluorochemicals were selected on the basis of a concentration that caused a detectable effect, but not so severe that it prohibited precise calculation of RCI.  *Statistically significant difference compared with the paired control value recorded in the absence of the test compound (p < 0.05).  PFOA  100  94.6 ± 6.0  17.4 ± 2.0  6.0 ± 0.4  (117 ± 1)  1.86 ± 0.12  (98 ± 2)  PFOS  10  98.3 ± 5.1  18.5 ± 1.9  5.5 ± 0.4  (126 ± 10)  1.82 ± 0.06  (101 ± 2)  FOSA  0.50  61.9 ± 0.4*  22.0 ± 3.2*  2.9 ± 0.4*  (70 ± 3)  1.81 ± 0.14*  (91 ± 3)  FOSAA  5  84.8 ± 7.4  26.0 ± 2.0*  3.3 ± 0.5*  (79 ± 2)  1.80 ± 0.10  (86 ± 2)  N-EtFOSA  6.25  87.0 ± 3.7  31.8 ± 3.1*  2.8 ± 0.3*  (54 ± 8)  1.40 ± 0.10*  (79 ± 11)  N-EtFOSE  62.5  83.7 ± 9.6  25.8 ± 1.9*  3.3 ± 0.3*  (71 ± 10)  1.84 ± 0.10  (100 ± 4)  N-EtFOSAA  6.25  81.7 ± 7.0  23.5 ± 3.4  3.87 ± 0.6*  (78 ± 17)  1.60 ± 0.08*  (88 ± 2)  Compound  μM  State 3  State 4  RCI  (%)  ADP/O  (%)  Note. Values represent the mean ± SEM for 3–6 individual mitochondrial preparations. Values for State 3 and State 4 are given in nmol O2/min/mg protein. Standard rates for state 4 and state 3 respiration in the absence of test compound were 19.5 ± 2.2 nmol O2/min/mg protein and 89.5 ± 8.6 nmol O2/min/mg protein, respectively. Concentrations of the individual fluorochemicals were selected on the basis of a concentration that caused a detectable effect, but not so severe that it prohibited precise calculation of RCI.  *Statistically significant difference compared with the paired control value recorded in the absence of the test compound (p < 0.05).  PFOA  100  94.6 ± 6.0  17.4 ± 2.0  6.0 ± 0.4  (117 ± 1)  1.86 ± 0.12  (98 ± 2)  PFOS  10  98.3 ± 5.1  18.5 ± 1.9  5.5 ± 0.4  (126 ± 10)  1.82 ± 0.06  (101 ± 2)  FOSA  0.50  61.9 ± 0.4*  22.0 ± 3.2*  2.9 ± 0.4*  (70 ± 3)  1.81 ± 0.14*  (91 ± 3)  FOSAA  5  84.8 ± 7.4  26.0 ± 2.0*  3.3 ± 0.5*  (79 ± 2)  1.80 ± 0.10  (86 ± 2)  N-EtFOSA  6.25  87.0 ± 3.7  31.8 ± 3.1*  2.8 ± 0.3*  (54 ± 8)  1.40 ± 0.10*  (79 ± 11)  N-EtFOSE  62.5  83.7 ± 9.6  25.8 ± 1.9*  3.3 ± 0.3*  (71 ± 10)  1.84 ± 0.10  (100 ± 4)  N-EtFOSAA  6.25  81.7 ± 7.0  23.5 ± 3.4  3.87 ± 0.6*  (78 ± 17)  1.60 ± 0.08*  (88 ± 2)  View Large FIG. 1. View largeDownload slide Screening protocol for the effects of perfluorooctanes on mitochondrial bioenergetics. The various perfluorooctanes were screened for activity against mitochondrial respiration by conducting successive estimates of respiratory control and coupling efficiency first in the absence, then in the presence of the desired test compound. Freshly isolated rat liver mitochondria were incubated at 1 mg protein/ml at 25°C in 125 mM KCl, 10 mM MOPS (pH 7.4), and 4 mM KH2PO4 containing 5 mM each of glutamate and malate. Oxygen consumption (upper trace) was quantified with an oxygen electrode and membrane potential (lower trace) was monitored simultaneously with a TPP+-specific electrode. Additions were TPP+ at 0.2–0.8 μM; Mito, (rat liver mitochondria) at 1 mg/ml; 100 μM ADP; 0.5 μM FOSA; and 40 μM 2,4-dinitrophenol. Traces are typical of 3–6 individual experiments, each using a separate mitochondrial preparation. The numbers indicate the rates of respiration after each successive addition (nmol O2/min × mg protein). An effect of the perfluorooctane on mitochondrial bioenergetics was assessed by comparing the state 4 and state 3 respiration rates and the RCI and ADP:O ratios before and after addition of the respective perfluorooctane in a paired fashion. FIG. 1. View largeDownload slide Screening protocol for the effects of perfluorooctanes on mitochondrial bioenergetics. The various perfluorooctanes were screened for activity against mitochondrial respiration by conducting successive estimates of respiratory control and coupling efficiency first in the absence, then in the presence of the desired test compound. Freshly isolated rat liver mitochondria were incubated at 1 mg protein/ml at 25°C in 125 mM KCl, 10 mM MOPS (pH 7.4), and 4 mM KH2PO4 containing 5 mM each of glutamate and malate. Oxygen consumption (upper trace) was quantified with an oxygen electrode and membrane potential (lower trace) was monitored simultaneously with a TPP+-specific electrode. Additions were TPP+ at 0.2–0.8 μM; Mito, (rat liver mitochondria) at 1 mg/ml; 100 μM ADP; 0.5 μM FOSA; and 40 μM 2,4-dinitrophenol. Traces are typical of 3–6 individual experiments, each using a separate mitochondrial preparation. The numbers indicate the rates of respiration after each successive addition (nmol O2/min × mg protein). An effect of the perfluorooctane on mitochondrial bioenergetics was assessed by comparing the state 4 and state 3 respiration rates and the RCI and ADP:O ratios before and after addition of the respective perfluorooctane in a paired fashion. FIG. 2. View largeDownload slide Effects of FOSA on mitochondrial bioenergetics. The reaction conditions were as those described in Figure 1, except that oligomycin was present at 2 μg/ml. The additions of FOSA were 0.5 μM, 1 μM, 1 μM, 1 μM, 2 μM, and 2 μM, respectively. Trace is typical of 3–6 individual mitochondrial preparations. FIG. 2. View largeDownload slide Effects of FOSA on mitochondrial bioenergetics. The reaction conditions were as those described in Figure 1, except that oligomycin was present at 2 μg/ml. The additions of FOSA were 0.5 μM, 1 μM, 1 μM, 1 μM, 2 μM, and 2 μM, respectively. Trace is typical of 3–6 individual mitochondrial preparations. FIG. 3. View largeDownload slide Effect of perfluorochemicals on mitochondrial respiration as a function of membrane potential. The data were gathered from experiments conducted as described in Figure 2. Each point represents the mean ± SEM of 3–6 individual mitochondrial preparations, each performed with a range of concentrations of the respective perfluorooctanes that produced the indicated dose-dependent changes in respiration rate and membrane potential. The concentration ranges were 0.5–5 μM FOSA; 2–25 μM N-EtFOSA; 2–50 μM 2,4-DNP; and 10–50 μM FOSAA. The lines were plotted by least-squares linear regression. FIG. 3. View largeDownload slide Effect of perfluorochemicals on mitochondrial respiration as a function of membrane potential. The data were gathered from experiments conducted as described in Figure 2. Each point represents the mean ± SEM of 3–6 individual mitochondrial preparations, each performed with a range of concentrations of the respective perfluorooctanes that produced the indicated dose-dependent changes in respiration rate and membrane potential. The concentration ranges were 0.5–5 μM FOSA; 2–25 μM N-EtFOSA; 2–50 μM 2,4-DNP; and 10–50 μM FOSAA. The lines were plotted by least-squares linear regression. FIG. 4. View largeDownload slide Dose-dependent uncoupling of mitochondrial respiration by perfluorooctane sulfonamides. Rat liver mitochondria were incubated according to the conditions described for Figure 2 in the presence of 2 μg/ml oligomycin. Each point represents the mean ± SEM for three to six individual mitochondrial preparations. The horizontal dashed line indicates the mean state 3 respiration rate recorded in the absence of added perfluorochemical and is included strictly as a reference point for illustrative purposes. FIG. 4. View largeDownload slide Dose-dependent uncoupling of mitochondrial respiration by perfluorooctane sulfonamides. Rat liver mitochondria were incubated according to the conditions described for Figure 2 in the presence of 2 μg/ml oligomycin. Each point represents the mean ± SEM for three to six individual mitochondrial preparations. The horizontal dashed line indicates the mean state 3 respiration rate recorded in the absence of added perfluorochemical and is included strictly as a reference point for illustrative purposes. FIG. 5. View largeDownload slide Cytochrome c release-dependent inhibition of mitochondrial respiration by N-EtFOSAA. Rat liver mitochondria were incubated according to the conditions described in Figure 1. Additions were 50 μM N-EtFOSAA; 40 μM 2,4-DNP; and 5 μM cytochrome c. Trace is representative 3–6 individual mitochondrial preparations. FIG. 5. View largeDownload slide Cytochrome c release-dependent inhibition of mitochondrial respiration by N-EtFOSAA. Rat liver mitochondria were incubated according to the conditions described in Figure 1. Additions were 50 μM N-EtFOSAA; 40 μM 2,4-DNP; and 5 μM cytochrome c. Trace is representative 3–6 individual mitochondrial preparations. FIG. 6. View largeDownload slide Induction of mitochondrial swelling by N-EtFOSAA. Rat liver mitochondria were incubated according to the conditions described for Figure 1. Mitochondrial swelling was monitored by light scattering at 540 nm. All reagents, including N-EtFOSAA, were added prior to initiating the reaction with glutamate/malate. Additions were 25 μM N-EtFOSAA ± 1 μM cyclosporine A (CyA). Trace is representative of 3–6 repetitions, each using a separate mitochondrial preparation. FIG. 6. View largeDownload slide Induction of mitochondrial swelling by N-EtFOSAA. Rat liver mitochondria were incubated according to the conditions described for Figure 1. Mitochondrial swelling was monitored by light scattering at 540 nm. All reagents, including N-EtFOSAA, were added prior to initiating the reaction with glutamate/malate. Additions were 25 μM N-EtFOSAA ± 1 μM cyclosporine A (CyA). Trace is representative of 3–6 repetitions, each using a separate mitochondrial preparation. FIG. 7. View largeDownload slide Disruption of the outer mitochondrial membrane by FOSAA. Rat liver mitochondria were incubated according to the conditions described for Figure 1. Induction of the permeability transition and disruption of the outer mitochondrial membrane were assessed concurrently by monitoring the reduction of exogenously added 50 μM ferricytochrome c at 550 nm (uppermost trace in each panel) and mitochondrial swelling at 540 nm (middle trace). The difference spectra (540–550 nm; lower trace) provides a quantitative measure of cytochrome c reduction by accounting for the background change in absorbance due to mitochondrial swelling. The additions were Mito, 0.2 mg/ml mitochondrial protein; 10 mM azide; and 80 μM FOSAA. Traces are representative of 3–6 repetitions, each using a separate mitochondrial preparation. FIG. 7. View largeDownload slide Disruption of the outer mitochondrial membrane by FOSAA. Rat liver mitochondria were incubated according to the conditions described for Figure 1. Induction of the permeability transition and disruption of the outer mitochondrial membrane were assessed concurrently by monitoring the reduction of exogenously added 50 μM ferricytochrome c at 550 nm (uppermost trace in each panel) and mitochondrial swelling at 540 nm (middle trace). The difference spectra (540–550 nm; lower trace) provides a quantitative measure of cytochrome c reduction by accounting for the background change in absorbance due to mitochondrial swelling. The additions were Mito, 0.2 mg/ml mitochondrial protein; 10 mM azide; and 80 μM FOSAA. Traces are representative of 3–6 repetitions, each using a separate mitochondrial preparation. 1 To whom correspondence should be addressed. Fax: (218) 726-8014. E-mail: kwallace@d.umn.edu. This work was supported by a grant from the 3M Company. REFERENCES Bernardi, P., Scorrano, L., Colonna, R., Petronilli, V., and Di Lisa, F. ( 1999). Mitochondria and cell death. Mechanistic aspects and methodological issues. Eur. J. Biochem.  264, 687–701. Google Scholar Borges, T., Robertson, L. W., Peterson, R. E., and Glauert, H. P. ( 1992). Dose-related effects of perfluorodecanoic acid on growth, feed intake and hepatic peroxisomal beta-oxidation. Arch. Toxicol.  66, 18–22. Google Scholar Bradford, M. M. ( 1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem.  72, 248–254. Google Scholar Custodio, J. B., Palmeira, C. M., Moreno, A. J., and Wallace, K. B. ( 1998). Acrylic acid induces the glutathione-independent mitochondrial permeability transition in vitro. Toxicol. Sci.  43, 19–27. Google Scholar Estabrook, R. W. ( 1967). Mitochondrial respiratory control and the polarographic measurement of ADP/O ratios. Methods Enzymol.  10, 41–47. Google Scholar Gilliland, F. D., and Mandel, J. S. ( 1996). Serum perfluorooctanoic acid and hepatic enzymes, lipoproteins, and cholesterol: A study of occupationally exposed men. Am. J. Ind. Med.  29, 560–568. Google Scholar Goecke-Flora, C. M., and Reo, N. V. ( 1996). Influence of carbon chain length on the hepatic effects of perfluorinated fatty acids. Chem. Res. Toxicol.  9, 689–695. Google Scholar Harrison, E. H., Lane, J. S., Luking, S., Van Rafelghem, M. J., and Andersen, M. E. ( 1988). Perfluoro-n-decanoic acid: Induction of peroxisomal β-oxidation by a fatty acid with dioxin-like toxicity. Lipids  23, 115–119. Google Scholar Haughom, B., and Spydevold, O. ( 1992). The mechanism underlying the hypolipemic effect of perfluorooctanoic acid (PFOA), perfluorooctane sulphonic acid (PFOSA) and clofibric acid. Biochim. Biophys. Acta  1128, 65–72. Google Scholar Kamo, N., Muratsugu, M., Hongoh, R., and Kobatake, Y. ( 1979). Membrane potential of mitochondria measured with an electrode sensitive to tetraphenyl phosphonium and relationship between proton electrochemical potential and phosphorylation potential in steady state. J. Membr. Biol.  49, 105–121. Google Scholar Kawashima, Y., Kobayashi, H., Miura, H., and Kozuka, H. ( 1995). Characterization of hepatic responses of rat to administration of perfluorooctanoic and perfluorodecanoic acids at low levels. Toxicology  99, 169–178. Google Scholar Keller, B. J., Marsman, D. S., Popp, J. A., and Thurman, R. G. ( 1992). Several nongenotoxic carcinogens uncouple mitochondrial oxidative phosphorylation. Biochim. Biophys. Acta  1102, 237–244. Google Scholar Kroemer, G., and Reed, J. C. ( 2000). Mitochondrial control of cell death. Nat. Med.  6, 513–519. Google Scholar Langley, A. E. ( 1990). Effects of perfluoro-n-decanoic acid on the respiratory activity of isolated rat liver mitochondria. J. Toxicol. Environ. Health  29, 329–336. Google Scholar Lemasters, J. J., Nieminen, A. L., Qian, T., Trost, L. C., Elmore, S. P., Nishimura, Y., Crowe, R. A., Cascio, W. E., Bradham, C. A., Brenner, D. A., and Herman, B. ( 1998). The mitochondrial permeability transition in cell death: a common mechanism in necrosis, apoptosis and autophagy. Biochim. Biophys. Acta  1366, 177–196. Google Scholar Nicholls, D. G. ( 1974). Hamster brown-adipose-tissue mitochondria. The control of respiration and the proton electrochemical potential gradient by possible physiological effectors of the proton conductance of the inner membrane. Eur. J. Biochem.  49, 573–583. Google Scholar Nicholls, D. G. ( 1974). The influence of respiration and ATP hydrolysis on the proton-electrochemical gradient across the inner membrane of rat-liver mitochondria as determined by ion distribution. Eur. J. Biochem.  50, 305–315. Google Scholar Nicholls, D. G. ( 1997). The non-Ohmic proton leak—25 years on. Biosci. Rep.  17, 251–257. Google Scholar Olorunsogo, O. O., and Malomo, S. O. ( 1985). Sensitivity of oligomycin-inhibited respiration of isolated rat liver mitochondria to perfluidone, a fluorinated arylalkylsulfonamide. Toxicology  35, 231–240. Google Scholar Olorunsogo, O. O., Malomo, S. O., and Bababunmi, E. A. ( 1985). Protonophoric properties of fluorinated arylalkylsulfonamides. Observations with perfluidone. Biochem. Pharmacol.  34, 2945–2952. Google Scholar Olsen, G., Burris, J. M., Mandel, J. H., and Zobel, L. R. ( 1999). Serum perfluorooctane sulfonate and hepatic and lipid clinical chemistry tests in fluorochemical production employees. J. Occup. Environ. Med.  41, 799–806. Google Scholar Pastoor, T. P., Lee, K. P., Perri, M. A., and Gillies, P. J. ( 1987). Biochemical and morphological studies of ammonium perfluorooctanoate-induced hepatomegaly and peroxisome proliferation. Exp. Mol. Pathol.  47, 98–109. Google Scholar Permadi, H., Lundgren, B., Andersson, K., and DePierre, J. W. ( 1992). Effects of perfluoro fatty acids on xenobiotic-metabolizing enzymes, enzymes which detoxify reactive forms of oxygen and lipid peroxidation in mouse liver. Biochem. Pharmacol.  44, 1183–1191. Google Scholar Permadi, H., Lundgren, B., Andersson, K., Sundberg, C., and DePierre, J. W. ( 1993). Effects of perfluoro fatty acids on peroxisome proliferation and mitochondrial size in mouse liver: Dose and time factors and effect of chain length. Xenobiotica  23, 761–770. Google Scholar Petronilli, V., Nicolli, A., Costantini, P., Colonna, R., and Bernardi, P. ( 1994). Regulation of the permeability transition pore, a voltage-dependent mitochondrial channel inhibited by cyclosporin A. Biochim. Biophys. Acta  1187, 255–259. Google Scholar Rottenberg, H. ( 1984). Membrane potential and surface potential in mitochondria: Uptake and binding of lipophilic cations. J. Membr. Biol.  81, 127–138. Google Scholar Schnellmann, R. G., and Manning, R. O. ( 1990). Perfluorooctane sulfonamide: A structurally novel uncoupler of oxidative phosphorylation. Biochim. Biophys. Acta  1016, 344–348. Google Scholar Sohlenius, A. K., Andersson, K., and DePierre, J. W. ( 1992). The effects of perfluoro-octanoic acid on hepatic peroxisome proliferation and related parameters show no sex-related differences in mice. Biochem. J.  285, 779–783. Google Scholar Sohlenius, A. K., Eriksson, A. M., Hogstrom, C., Kimland, M., and DePierre, J. W. ( 1993). Perfluorooctane sulfonic acid is a potent inducer of peroxisomal fatty acid β-oxidation and other activities known to be affected by peroxisome proliferators in mouse liver. Pharmacol. Toxicol.  72, 90–93. Google Scholar Taves, D. R. ( 1968). Evidence that there are two forms of fluoride in human serum. Nature  217, 1050–1051. Google Scholar Vahden Heuvel, J. P. ( 1996). Perfluorodecanoic acid as a useful pharmacologic tool for the study of peroxisome proliferation. Gen. Pharmacol.  27, 1123–1129. Google Scholar Van Rafelghem, M. J., Mattie, D. R., Bruner, R. H., and Andersen, M. E. ( 1987). Pathological and hepatic ultrastructural effects of a single dose of perfluoro-n-decanoic acid in the rat, hamster, mouse, and guinea pig. Fundam. Appl. Toxicol.  9, 522–540. Google Scholar Wallace, K. B., Eells, J. T., Madeira, V. M., Cortopassi, G., and Jones, D. P. ( 1997). Mitochondria-mediated cell injury. Symposium overview. Fundam. Appl. Toxicol.  38, 23–37. Google Scholar Wallace, K. B., and Starkov, A. A. ( 2000). Mitochondrial targets of drug toxicity. Annu. Rev. Pharmacol. Toxicol.  40, 353–388. Google Scholar Witzmann, F. A., Parker, D. N., and Jarnot, B. M. ( 1994). Induction of enoyl-CoA hydratase by LD50 exposure to perfluorocarboxylic acids detected by two-dimensional electrophoresis. Toxicol. Lett.  71, 271–277. Google Scholar Zhou, S., and Wallace, K. B. ( 1999). The effect of peroxisome proliferators on mitochondrial bioenergetics. Toxicol. Sci.  48, 82–89. Google Scholar Zoratti, M., and Szabo, I. ( 1995). The mitochondrial permeability transition. Biochim. Biophys. Acta  1241, 139–176. Google Scholar © 2002 Society of Toxicology
TI - Structural Determinants of Fluorochemical-Induced Mitochondrial Dysfunction
JF - Toxicological Sciences
DO - 10.1093/toxsci/66.2.244
DA - 2002-04-01
UR - https://www.deepdyve.com/lp/oxford-university-press/structural-determinants-of-fluorochemical-induced-mitochondrial-9Ug32O9mL1
SP - 244
EP - 252
VL - 66
IS - 2
DP - DeepDyve
ER -