Chem. Res. Toxicol. 2002, 15, 723-733
723
Biotransformation of 1,1,1,3,3-Pentafluoropropane (HFC-245fa) Tanja Bayer,† Alexander Amberg,† Ru¨diger Bertermann,‡ George M. Rusch,§ M. W. Anders,# and Wolfgang Dekant*,† Institut fu¨ r Toxikologie, Universita¨ t Wu¨ rzburg, Versbacher Strasse 9, 97078 Wu¨ rzburg, Germany, Institut fu¨ r Anorganische Chemie, Universita¨ t Wu¨ rzburg, Am Hubland, 97074 Wu¨ rzburg, Germany, Honeywell, P.O. Box 1057, Morristown, New Jersey 07962-1057, and Department of Pharmacology and Physiology, University of Rochester Medical Center, 601 Elmwood Avenue, Box 711, Rochester, New York 14642 Received January 3, 2002
1,1,1,3,3-Pentafluoropropane (HFC-245fa) is being developed as a CFC substitute. 1,1,1,3,3Pentafluoropropane has a low potential for toxicity: the only remarkable toxic effect seen in rats after inhalation exposure to 1,1,1,3,3-pentafluoropropane in concentrations of up to 50,000 ppm for 90 days was an increased incidence of diffuse myocarditis. To elucidate the possible role of biotransformation in 1,1,1,3,3-pentafluoropropane-induced cardiotoxicity, the biotransformation of 1,1,1,3,3-pentafluoropropane was investigated in rats after inhalation exposure and in rat and human liver microsomes. Male and female rats were exposed by inhalation to 50 000, 10 000, and 2000 ppm 1,1,1,3,3-pentafluoropropane for 6 h, urine was collected for 72 h, and metabolites excreted were identified by 19F NMR spectroscopy and quantified by GC/ MS. Trifluoroacetic acid and inorganic fluoride were identified as major urinary metabolites of 1,1,1,3,3-pentafluoropropane; 3,3,3-trifluoropropanoic acid and 1,1,1,3,3-pentafluoropropane2-ol were minor metabolites. The extent of 1,1,1,3,3-pentafluoropropane biotransformation after inhalation was dependent on exposure concentrations. Neither 3,3,3-trifluoropropanoic acid nor 3,3,3-trifluoropyruvic acid were metabolized to trifluoroacetic acid in vitro or in rats. In rat and human liver microsomes, 1,1,1,3,3-pentafluoropropane was biotransformed by a cytochrome P450-dependent reaction to trifluoroacetic acid and 3,3,3-trifluoropropanoic acid. Rates of trifluoroacetic acid formation were 99.2 ( 20.5 pmol (mg of protein)-1 min-1 and of 3,3,3-trifluoropropanoic acid formation were 17.5 ( 4.0 pmol (mg of protein)-1 min-1 in liver microsomes from male rats. In human liver microsomes, rates of trifluoroacetic acid formation ranged from 0 to 11.6 pmol (mg of protein)-1 min-1, and rates of 3,3,3-trifluoropropanoic acid formation ranged from 0.7 to 7.6 pmol (mg of protein)-1 min-1. The results show that 1,1,1,3,3pentafluoropropane is metabolized at low rates in vivo and in vitro. The toxic effects of 1,1,1,3,3pentafluoropropane may be associated with the formation of the minor metabolite 3,3,3trifluoropropanoic acid, which is highly toxic in rats.
Introduction 1,1,1,3,3-Pentafluoropropane (HFC-245fa) is a higherboiling hydrofluorocarbon and is being considered for use in applications such as foam blowing, refrigeration, and precision cleaning. 1,1,1,3,3-Pentafluoropropane shows little potential for ozone depletion and, hence, its globalwarming potential is also reduced. As with several other hydrofluorocarbons, 1,1,1,3,3-pentafluoropropane has a low potential for toxicity and is neither mutagenic nor teratogenic (1-6). In a 90-day study, the only histopathological lesion observed in rats exposed to 1,1,1,3,3-pentafluoropropane by inhalation was an increased incidence of mild myocarditis, which was seen in all animals exposed to 50 000 ppm and in the majority of animals exposed to 10 000 ppm (8 h/day, 5 days/week) (7). * Address correspondence to Dr. W. Dekant, Department of Toxicology, University of Wu¨rzburg, Versbacher Str. 9, 97078 Wu¨rzburg, Germany. Tel.: +49(931)201 3449. Fax: +49(931)201 3446. E-mail:
[email protected]. † Institut fu ¨ r Toxikologie, Universita¨t Wu¨rzburg. ‡ Institut fu ¨ r Anorganische Chemie, Universita¨t Wu¨rzburg. § Honeywell. # University of Rochester Medical Center.
Chemically induced myocarditis is an uncommon finding in experimental animals, and cardiac lesions have not been reported with other chlorofluorocarbon substitutes (8, 9). The mechanism of observed selective toxicity of 1,1,1,3,3-pentafluoropropane for the heart is not known. In general, however, selective organ damage seen after long-term administration of chemicals is often associated with biotransformation to reactive electrophiles or to stable, but toxic, metabolites (10). Some hydrofluorocarbons and hydrochlorofluorocarbons are metabolized by cytochrome P450-catalyzed C-H bond oxidation to acylating agents, which may form protein adducts (11, 12). Stable metabolites, e.g., halogenated aliphatic alcohols, aldehydes, or carboxylic acids, may be formed by hydrolysis of these reactive intermediates. Some stable metabolites, such as trihaloacetic acids, are peroxisome proliferators, and their formation may also contribute to toxic effects of the parent compounds (1317). To elucidate possible mechanism of 1,1,1,3,3-pentafluoropropane-induced myocarditis, its biotransformation was investigated in rats in vivo and in vitro. These
10.1021/tx025505c CCC: $22.00 © 2002 American Chemical Society Published on Web 04/19/2002
724
Chem. Res. Toxicol., Vol. 15, No. 5, 2002
studies revealed an unusual biotransformation reaction, i.e., the cleavage of a carbon-carbon bond in 1,1,1,3,3pentafluoropropane to give trifluoroacetic acid as the major metabolite; this biotransformation, which involves C-C bond cleavage, has apparently not been previously observed with halogenated hydrocarbons.
Material and Methods Chemicals. 1,1,1,3,3-Pentafluoropropane, 99.8% pure (based on FID GC analysis) was supplied by Honeywell (Morristown, NJ). 3,3,3-Trifluoropropanoic acid, trifluoroacetic acid, difluoroacetic acid, and other chemicals were purchased from SigmaAldrich (Deisenhofen, Germany) in the highest purity available. Enzymes and cofactors were also obtained from Sigma-Aldrich. Sodium 3,3,3-trifluoropyruvate was obtained by hydrolysis of ethyl 3,3,3-trifluoropyruvate (Lancaster Synthesis Ltd., Windham, NH). Ethyl 3,3,3-trifluoropyruvate was added to 20 mL of 1 M sulfuric acid, and the mixture was stirred for 3 h at room temperature. The reaction mixture was extracted three times with 50 mL of ether and three times with 50 mL of ethyl acetate. The combined organic layers were concentrated under reduced pressure, and the residue was taken up in 20 mL of water. The pH of the solution was brought to 7.8 with 1 M sodium hydroxide, and the solution was lyophilized to give 543 mg of a white solid, which was used without further purification. Animals and Treatment. Sprague-Dawley rats (CharlesRiver Wiga, Sulzfeld, Germany, 220-250 g of body weight) were used for all studies. To induce CYP2E1, rats were given pyridine (100 mg/kg i.p. dissolved in isotonic sodium chloride solution) once daily for 5 days. All animals were fasted 18 h before sacrifice and preparation of microsomes (18-20). To study the fate of possible metabolites of 1,1,1,3,3-pentafluoropropane, two male rats were given 10 mg/kg 3,3,3-trifluoropyruvic acid or 5 mg/kg 3,3,3-trifluoropropanoic acid, dissolved in water, by gavage. Animals were transferred to metabolic cages, and urine was collected for 48 h. Urine samples were analyzed by 19F NMR spectroscopy. Incubation Conditions. Hydroxylation of p-nitrophenol was determined as described previously (12, 21). The absorbance of 4-nitrocatechol was measured spectrophotometrically at 510 nm (a ) 14.6 L/mmol‚cm). Final protein concentrations were 1 and 2 mg/mL for rat and human liver microsomes, respectively. Diethyldithiocarbamate, a selective inhibitor of CYP2E1 (22), was used in final concentrations of 100 and 300 µM. For all reaction mixtures, microsomes, a NADPH-generating system (23), and diethyldithiocarbamate were incubated for 5 min at 37 °C before addition of substrate. Reaction mixtures with 1,1,1,3,3-pentafluoropropane contained microsomes and substrates or inhibitors as noted in a total volume of 1.1 mL of 0.1 M phosphate buffer containing 1 mM EDTA (pH 7.4). Microsomes and the corresponding amount of buffer were placed in sealed 2-mL GC vials. 1,1,1,3,3-Pentafluoropropane (10 µL of liquid at 4 °C) was added through the septum with a microliter syringe. Reaction mixtures were incubated at 37 °C in a shaking water bath for 20 min. The vials were submerged in water to ensure constant temperatures in the vials. The reactions were stopped by placing the vials on ice. Each reaction was repeated four times. The 6-hydroxylation of chlorzoxazone was determined in reaction mixtures containing 0.25 mM chlorzoxazone, microsomal protein, and NADPH-generating system in a final volume of 1.0 mL of 100 mM phosphate buffer (pH 7.4) (18). The source and characterization of the human liver samples have been described previously (18). Exposure of Rats to 1,1,1,3,3-Pentafluoropropane. Five male (210-230 g, 12-weeks-old) and five female rats (190-210 g, 12-weeks-old) were exposed to targeted concentrations of 2000, 10 000, and 50 000 ppm 1,1,1,3,3-pentafluoropropane in a dynamic exposure chamber consisting of a 20.6-L desiccator, a mixing chamber, and connections to compressed air and a tank of 1,1,1,3,3-pentafluoropropane fitted with flow meters. Metered amounts of 1,1,1,3,3-pentafluoropropane were mixed with air
Bayer et al. in the mixing chamber and introduced into the exposure chamber. Chamber concentrations of 1,1,1,3,3-pentafluoropropane were monitored at 15-min intervals by taking 1-mL samples of the chamber atmosphere with a gastight syringe. The content of 1,1,1,3,3-pentafluoropropane in these samples was determined by GC/MS. Quantification of 1,1,1,3,3-pentafluoropropane was based on calibration curves with air samples containing known concentrations of 1,1,1,3,3-pentafluoropropane. After the end of the exposure, the animals were transferred to metabolism cages, and urine was collected on ice for 72 h at 6-h intervals. Quantification of Inorganic Fluoride Concentrations. To quantify excretion of inorganic fluoride in 1,1,1,3,3-pentafluoropropane-exposed rats, 5 mL of urine was combined with an equal volume of total-ionic-strength-adjustment buffer (TISAB: 1 M acetic acid, 1 M sodium chloride, 0.012 M (()trans-1,2-diaminocyclohexane-N,N,N′,N′-tetraacetic acid monohydrate in deionized water, pH 5.5) and analyzed with a fluoride-selective electrode and a reference electrode (Metrohm, Herisau, CH). The fluoride-selective electrode was calibrated daily with freshly prepared solutions containing 0.1, 0.2, 0.5, 0.75, 1, and 10 ppm sodium fluoride. Samples were constantly stirred during analysis, and values for response (mV) were read after 10 min, and calibration curves were prepared by plotting the response vs the fluoride concentration. Instrumental Analyses. 19F NMR spectra were recorded with a Bruker DRX 300 NMR spectrometer with a 5-mm fluorine probe operating at 282.4 MHz. 19F chemical shifts were referenced to external CFCl3. 19F NMR spectra were recorded with a 90° pulse length of 11 µs and a recycle delay of 1 s. The acquisition time was 1.5 s and 2000 up to 5000 scans were recorded for a good signal-to-noise (S/N) ratio. For comparison purposes, the 19F NMR spectra were acquired with and without proton decoupling. Before the Fourier transformation, a line broadening of 1 Hz was applied. To record NMR spectra, 720 µL of rat urine was diluted with 80 µL of D2O, and samples were analyzed directly. To analyze incubation mixtures, proteins were sedimented by centrifugation at 100000g for 20 min, and 80 µL of D2O was added to the supernatants (720 µL). The mixtures were analyzed by 19F NMR spectroscopy without further workup. GC/MS analyses were performed with an Agilent 5973 mass spectrometer coupled to an Agilent 6890 GC equipped with a CTC Combi-PAL autoinjector with capability for headspace injections. Quantification of Trifluoroacetic Acid and 3,3,3-Trifluoropropanoic Acid Formation. Samples of urine or supernatants (1 mL) of microsomal incubations were mixed with 80 µL of 0.1 M NaOH and 100 nmol of difluoroacetic acid in 50 µL of H2O as an internal standard. Samples were then taken to dryness in an evacuated desiccator containing anhydrous P2O5. The organic acids in the dried residues were converted to methyl esters by addition of 100 µL of methanol and 100 µL of concentrated (97%) sulfuric acid and heating for 1 h at 80 °C in gastight reaction vials. Headspace samples (250 µL) were removed with a warmed (80 °C) gastight syringe, and the concentrations of methyl esters were quantified by GC/MS. Compounds were separated on Agilent Q-Plot fused-silica capillary column (30 m × 0.32 mm i.d.; film thickness, 20 µm) with the following conditions: linear temperature program from 100 to 220 °C with a heating rate of 15 °C/min; injector and detector temperatures, 250 °C; helium, 2 mL/min; and split injection with a split ratio of 5:1. During the chromatographic separation, the intensities of m/z 51, 59, 60, 69, 83, 111, and 142 were monitored with a dwell time of 100 ms. Retention times were 4.3 min for trifluoroacetic acid, 5.8 min for difluoroacetic acid, and 7.0 min for 3,3,3-trifluoropropanoic acid. Quantification was performed relative to the content of difluoroacetic acid (m/z 51) and referenced to calibration curves prepared with samples of urine or microsomal supernatant samples containing 0-350 nmol/mL trifluoroacetic acid (quantified at m/z 69) or 0-15 nmol/mL 3,3,3-trifluoropropanoic acid
Biotransformation of HFC-245fa
Chem. Res. Toxicol., Vol. 15, No. 5, 2002 725
Figure 1. 19F NMR spectrum (proton decoupled) of urine from a male rat exposed to 50 000 ppm 1,1,1,3,3-pentafluoropropane for 6 h. The urine sample was collected from 0 to 12 h after the end of the exposure. (quantified at m/z 111). This method permitted the quantification of 0.5 nmol/mL trifluoroacetic acid and 0.5 nmol/mL 3,3,3trifluoropropanoic acid with a signal-to-noise ratio of 10:1. Deviations between repeatedly analyzed reference samples were C18) 3-methyl-fatty acids (24, 25), but the R-oxidation of shortchain fatty acids has apparently not been reported. A second possible route for the biotransformation of 1,1,1,3,3pentafluoropropane 1 to trifluoroacetic acid 11 involves the oxidation of 2-hydroxy-3,3,3-propanoic acid 12, which may be formed by hydrolysis of 2-hydroxy-3,3,3-trifluoropropanoyl-CoA 9 or by direct hydroxylation, to 3,3,3-trifluoropyruvic acid 13 followed by pyruvate decarboxylase-catalyzed decarboxylation to give trifluoroacetic acid 11. Hence, the fate of 3,3,3-trifluoropropanoic acid 4 and 3,3,3-trifluoropyruvic acid 13 was studied in liver and heart homogenates and in rats. To test for the involvement of the R-hydroxylation pathway, dialyzed rat liver and heart homogenates from rats were incubated with 3,3,3-trifluoropropanoic acid and the cofactors required for the several steps in the R-oxidation pathway (24, 25), and biotransformation was assessed by 19F NMR spectroscopy and GC/MS analysis. Very small amounts of trifluoroacetic acid and inorganic fluoride were present in all samples, but no fluorinecontaining products were formed, as indicated by 19F NMR spectroscopy. Moreover, the concentrations of trifluoroacetic acid, as determined by GC/MS analysis, were identical in all samples irrespective of cofactors present, indicating that 3,3,3-trifluoropropanoic acid was not biotransformed to trifluoroacetic acid by the R-hydroxylation pathways in vitro (data not shown). The oral administration of 3,3,3-trifluoropyruvic acid 13 to rats did not result in the excretion of trifluoroacetic acid. The only fluorine-containing metabolite identified
in urine showed a 19F NMR spectrum (Figure 4) consistent with the structure of 2-hydroxy-3,3,3-trifluoropropanoic acid 12 (Scheme 2). The 19F NMR spectrum of the urine samples showed a doublet at δ ) -75.71 with a H-F coupling constant of 7.8 Hz and an increase in the intensity of the resonance for inorganic fluoride at δ ) -118.8. No resonances for trifluoroacetic acid 11 (δ ) -75.92, s) or 3,3,3-trifluoropyruvic acid 13 (δ ) -83.21, s) were observed, indicating that 3,3,3-trifluoropyruvic acid 13 is reduced to 2-hydroxy-3,3,3-trifluoropropanoic acid 12 and, to a small extent, converted to inorganic fluoride. In addition, oral administration of a low dose of 3,3,3-trifluoropropanoic acid 4 resulted in the recovery of unchanged 3,3,3-trifluoropropanoic acid 4, but no trifluoroacetic acid 11 was found in the urine (Figure 5). The resonance at δ ) -63.9 ppm was a triplet in protoncoupled spectra with a coupling constant of 11.4 Hz and was identical in chemical shift and coupling constant with that of authentic 3,3,3-trifluoropropanoic acid 4. The smaller resonance at δ ) -63.46 ppm had identical H-F coupling constants and may represent a conjugate of 3,3,3-trifluoropropanoic acid, since it was also seen in low intensity in urine samples of rats exposed to 1,1,1,3,3pentafluoropropane (Figure 1). These in vivo data further support the negative in vitro results, indicating that the pathways shown in Scheme 2 are not involved in the formation of trifluoroacetic acid 11 from 3,3,3-trifluoropropanoic acid 4. Incubation of Liver Microsomes with 1,1,1,3,3Pentafluoropropane. Alternative pathways to account for trifluoroacetic acid formation from 1,1,1,3,3-pentafluoropropane may involve the cytochrome P450-catalyzed oxidation of 1,1,1,3,3-pentafluoropropane resulting directly in the formation of trifluoroacetic acid or in the formation of an intermediate that is converted to trifluoroacetic acid. To test these possible mechanisms and to characterize the biotransformation of 1,1,1,3,3-pentafluoropropane by cytochrome P450, 1,1,1,3,3-pentafluoropropane was incubated with rat and human liver microsomes. Incubation of 1,1,1,3,3-pentafluoropropane
Biotransformation of HFC-245fa
Chem. Res. Toxicol., Vol. 15, No. 5, 2002 729
Figure 5. 19F NMR spectrum (proton coupled) of a urine sample (collected for 24 h) from a male rat after oral administration of 5 mg/kg 3,3,3-trifluoropropanoic acid. The resonance at δ ) -63.90 was identical in chemical shift and H-F coupling (1JH-F ) 11.4 Hz) with the resonance obtained with authentic 3,3,3-trifluoropropanoic acid. The smaller resonance was also a triplet with a H-F coupling constant of 11.4 Hz. Table 2. Biotransformation of 1,1,1,3,3-Pentafluoropropane in Liver Microsomes from Ratsa microsome source
conc (nmol/mL)
rate of formation [pmol (mg of protein)-1 min-1]
p-nitrophenol oxidation [nmol (mg of protein)-1 min-1]
Trifluoroacetic Acid male rats -NADPH +NADPH male rats, pyridine -NADPH +NADPH female rats -NADPH +NADPH female rats, pyridine -NADPH +NADPH
1.99 ( 0.26 5.96 ( 0.73*
99.2 ( 20.5
0.18 ( 0.05
1.3 ( 0.5 5.0 ( 1.3**
92.8 ( 32
3.22 ( 1.2
1.65 ( 0.24 4.80 ( 0.76*
80.1 ( 19.8
0.11 ( 0.02
1.6 ( 0.16 7.7 ( 2.1**
151 ( 53
7.14 ( 1.67
Acidb
male rats, pyridine female rats, pyridine
no background
3,3,3-Trifluoropropanoic 17.5 ( 4.0 26.7 ( 4.8
a Incubations were performed at 37 °C for 20 min. Concentrations of trifluoroacetic acid in samples with NADPH (n ) 5) were significant different when compared with samples lacking NADPH (*p < 0.05, **p < 0.001). Rates were calculated after subtraction of background. b 3,3,3-Trifluoropropanoic acid was not detected after incubation of 1,1,1,3,3-pentafluoropropane with liver microsomes from control rats without NADPH.
in the presence of microsomal protein from control rats did not result in the formation of detectable amounts of 3,3,3-trifluoropropanoic acid; based on the detection limit of the GC/MS-assay used, rates of formation of 3,3,3trifluoropropanoic acid were 0.05) as compared with reaction mixtures lacking NADPH after an incubation time of 20 min (Figure 6), but not after 10 min. The concentration of 3,3,3-trifluoropropanoic acid in all samples was below the sensitivity of the 19F NMR spectroscopic
730
Chem. Res. Toxicol., Vol. 15, No. 5, 2002
Bayer et al.
Figure 6. GC/MS analysis of reaction mixtures containing 1,1,1,3,3-pentafluoropropane and rat liver microsomes in the absence or presence of a NADPH-generating system. The upper panel shows the combined intensity of m/z 51, 59, 60, 69, 83, 111, and 142 during the chromatographic separation; the lower panel shows the intensity of m/z 111. The peak at a retention time of 4.3 min (upper panel) represents methyl trifluoroacetate, and the peak at 5.6 min (upper panel) represents the internal standard methyl difluoroacetate. The peak at a retention of 7.0 min (lower panel) represents methyl 3,3,3-trifluoropropanoate. The upper traces in each panel were from reaction mixtures containing NADPH, and the lower traces were from reaction mixtures lacking NADPH. Reaction mixtures were incubated for 20 min.
method, but the increases in the concentrations of trifluoroacetic acid could also be observed by 19F NMR spectroscopy (data not shown). GC/MS analysis of the gas phase from the reaction mixtures did not show the formation of volatile metabolites. Time- and protein-concentration-dependent increases in trifluoroacetic acid and 3,3,3-trifluoropropanoic acid formation were observed by GC/MS analysis in microsomes from rats given the CYP2E1-inducer pyridine. On the basis of the measured concentrations of trifluoroacetic acid, rates of formation of trifluoroacetic acid were calculated (Table 2). The rates of formation of trifluoroacetic acid from 1,1,1,3,3-pentafluoropropane were not increased in microsomal fraction from pyridinetreated rats as compared with control rats, despite a significant increase in p-nitrophenol oxidase activity in the microsomal fractions. Biotransformation of 3,3,3trifluoropropanoic acid to 2-hydroxy-3,3,3-trifluoropropanoic acid or other fluorine-containing metabolites was not observed by 19F NMR spectroscopy in rat liver microsomes incubated in the presence of a NADPHgenerating system, indicating that 3,3,3-trifluoropropanoic acid is not a substrate for cytochromes P450. The oxidation of p-nitrophenol and 1,1,3,3,3-pentafluoropropane was studied in seven human liver microsome samples (Table 3). The human liver microsome samples showed p-nitrophenol oxidase activities similar to those reported previously (26). In some of the samples of liver microsomes, increased concentrations of trifluoroacetic acid were formed in reaction mixtures incubated in the presence of NADPH as compared with controls lacking NADPH, indicating that these samples catalyzed the formation of trifluoroacetic acid from 1,1,3,3,3-pentafluoropropane. Enzymatic formation of 3,3,3-trifluoropropanoic acid from 1,1,1,3,3-pentafluoropropane was observed at low rates in all samples. As with rat liver
Table 3. Oxidation of 1,1,3,3,3-Pentafluoropropane in Human Liver Microsomes rate of oxidation of 1,1,3,3,3-pentafluoropropane to rate of oxidation of 3,3,3-trifluorop-nitrophenol trifluoroacetic propanoic acid [nmol (mg of acid [pmol (mg of [pmol (mg of -1 -1 -1 -1 samples protein) min ] protein) min ] protein)-1 min-1)b HL11 HL19 HL2a HL13 HL15 HL1b HL3a
2.17 ( 0.06 1.19 ( 0.27 1.23 ( 0.21 1.6 ( 0.57 0.76 ( 0.06 0.49 ( 0.03 0.93 ( 0.04
nd 11.5 ( 3.5a nd nd 11.6 ( 5.3a 7.7 ( 0.9a nd
2.9 ( 0.5 4.1 ( 1.0 7.6 ( 2.0 2.8 ( 0.7 3.2 ( 0.3 1.2 ( 0.3 1.4
a Trifluoroacetic acid concentrations in incubation mixtures with and without NADPH were significantly different (p < 0.05); nd, no significant difference in trifluoroacetic acid concentrations between samples with or without NADPH. b Absence of background for 3,3,3-trifluoropropanoic acid permitted quantification of low rates of biotransformation.
microsomes, trifluoroacetic acid was the predominant metabolite formed by oxidation of 1,1,1,3,3-pentafluoropropane.
Discussion These results demonstrate that trifluoroacetic acid 11, 3,3,3-trifluoropropanoic acid 4, inorganic fluoride, and, perhaps, 1,1,1,3,3-pentafluoro-2-propanol 5 and 1,1,1,3,3pentafluoro-2-propanone 6 or its hydrate 7, or both, are metabolites formed during the biotransformation of 1,1,1,3,3-pentafluoropropane 1 in rats after inhalation exposure. The formation of 3,3,3-trifluoropropanoic acid 4 from 1,1,1,3,3-pentafluoropropane 1 may be rationalized by a cytochrome P450-catalyzed oxidation of 1,1,1,3,3pentafluoropropane 1 at C-3 to give the geminal fluorohydrin 1,1,3,3,3-pentafluoro-1-propanol 2, which may lose
Biotransformation of HFC-245fa
Chem. Res. Toxicol., Vol. 15, No. 5, 2002 731 Scheme 3
HF to give 3,3,3-trifluoropropanoyl fluoride 3. Hydrolysis of acyl fluoride 3 would give 3,3,3-trifluoropropanoic acid 4, which is excreted in the urine (Scheme 1). The cytochrome P450-catalyzed oxidation of 1,1,1,3,3-pentafluoropropane 1 at C-2 would give 1,1,1,3,3-pentafluoro2-propanol 5, which may be conjugated with glucuronic acid; oxidation of alcohol 5 would give 1,1,1,3,3-pentafluoro-2-propanone 6, which may exist as the hydrate 7 (Scheme 1). The quantification of metabolites indicated that 1,1,1,3,3-pentafluoropropane undergoes little biotransformation, even after exposure to high concentrations. The low extent of biotransformation was expected due to the high stability of the C-H bond in hydrofluorocarbons. For example, pentafluoroethane is also metabolized slowly in rats (27, 28). The formation of trifluoroacetic acid 11 from 1,1,1,3,3pentafluoropropane 1 was an unexpected finding, and such reactions have apparently not been previously observed in the biotransformation of aliphatic halogenated hydrocarbons. Trifluoroacetic acid 11 was the major metabolite of 1,1,1,3,3-pentafluoropropane 1 at all exposure concentrations, whereas 3,3,3-trifluoropropanoic acid 4 was a relatively minor metabolite. The metabolism of 1,1,1,3,3-pentafluoropropane 1 to both trifluoroacetic acid 11 and 3,3,3-trifluoropropanoic acid 4 was dependent on time and exposure concentrations. 19F NMR spectroscopic analysis of urine of rats exposed to 1,1,1,3,3pentafluoropropane 1 showed that 1,1,1,3,3-pentafluoro2-propanol 5 and 1,1,1,3,3-pentafluoro-2-propanone 6 or its hydrate 7 were also metabolites (Scheme 1). Trifluoroacetic acid excretion was apparently slow, perhaps due to plasma protein binding; as described previously (13, 26), trifluoroacetic acid is slowly excreted after animal and human exposure to chemicals that are biotransformed to trifluoroacetic acid. The recovery of inorganic fluoride was lower than expected, based on the amount of trifluoroacetic acid and 3,3,3-trifluoropropanoic acid recovered in urine; less than two equivalents of fluoride were recovered based on organofluorine metabolite excretion. The low excretion of fluoride may be attributed to its incorporation into bone, which may amount to up to 50% of the dose after oral administration (29). The mechanism of formation of trifluoroacetic acid 11 from 1,1,1,3,3-pentafluoropropane 1 is not readily apparent from the data presented herein. A role for the R-oxidation pathway could not be confirmed experimentally. The microsomal biotransformation experiments, after specific precautionary methods to reduce trifluoroacetic acid contamination, indicated conclusively that trifluoroacetic acid 11 was formed by the cytochrome P450-dependent oxidation of 1,1,1,3,3-pentafluoropropane 1 or a metabolite formed by cytochrome P450 from 1,1,1,3,3-pentafluoropropane. Known mechanisms of cytochrome P450-catalyzed oxidations do not suggest a pathway for the biotransformation of 1,1,1,3,3-pentafluoropropane to trifluoroacetic acid (30). A carboncarbon bond cleavage reaction has been proposed for
haloalkenes, such as trichloroethene, where the epoxide intermediate may undergo C-C fission (31, 32). With 1,1,1,3,3-pentafluoropropane, however, this reaction would require prior olefin formation from 1,1,1,3,3-pentafluoropropane followed by oxidation and C-C bond cleavage. Neither 1,1,1,3-tetrafluoroprop-2-ene nor 1,1,1,3,3pentafluoroprop-2-ene were observed in incubation mixtures by 19F NMR spectroscopy or in the headspace of incubation mixtures in closed vials by GC/MS analysis. Carbon-carbon fission reactions have been observed in steroid biosynthesis, but this reaction apparently involves a Baeyer-Villiger rearrangement (33, 34), which requires a carbonyl group that is converted to an intermediate peroxide that rearranges to give an ester. CYP-catalyzed oxidation of the C-H bond in the putative metabolite 1,1,1,3,3-pentafluoro-2-propanol 5 may give an intermediate ferryl peroxide intermediate 14 that is analogous to the peroxide intermediates in the Baeyer-Villiger rearrangement. Ferryl peroxide intermediate 14 may rearrange via carbocationic intermediate 15 to give trifluoroacetic acid 11 after reaction with water (Scheme 3). Attempts to prepare 1,1,1,3,3-pentafluoro-2-propanol 5 by synthesis were unsuccessful; hence, this proposed mechanism could not be tested. In human liver microsomes, the extent of biotransformation of 1,1,1,3,3-pentafluoropropane to trifluoroacetic acid was not correlated with CYP2E1 activity, as determined by p-nitrophenol oxidation. CYP2E1 is the major cytochrome P450 involved in the biotransformation of hydrochlorofluorocarbons and hydrofluorocarbons and a number of other low molecular-weight compounds (22). Because the molecular size of 1,1,1,3,3-pentafluoropropane is similar to other CYP2E1 substrates, it is likely that 1,1,1,3,3-pentafluoropropane would also be a substrate for CYP2E1. The lack of correlation of trifluoroacetic acid formation and CYP2E1 activity may indicate the involvement of a sequential mechanism, as proposed above for the formation of trifluoroacetic acid from 1,1,1,3,3-pentafluoro-2-propanol. Such a sequential mechanism may involve more than one CYP. The present investigations on the biotransformation of 1,1,1,3,3-pentafluoropropane were aimed at elucidating the role of metabolites and possible reactive intermediates in the myocarditis induced by 1,1,1,3,3-pentafluoropropane. The intermediate acyl fluoride 3 that is likely formed during oxidation of 1,1,1,3,3-pentafluoropropane 1 may react with nucleophilic sites in proteins. Protein binding of reactive acyl fluorides has been implicated as a potential mechanism of toxicity for several halogenated aliphatic hydrocarbons (35). With 1,1,1,3,3-pentafluoropropane, the role of protein acylation by 3,3,3-trifluoropropanoyl fluoride 3 is uncertain, but its low rate of formation argues against a major role. Moreover, cardiac cytochrome P450 activities are low, indicating limited formation of 3,3,3-trifluoropropanoyl fluoride in the heart compared with other organs with higher cytochrome P450 activity, such as the liver (36). Hepatotoxicity of 1,1,1,3,3pentafluoropropane has, however, not been reported (7).
732
Chem. Res. Toxicol., Vol. 15, No. 5, 2002
In addition, the probable short half-life of 3,3,3-trifluoropropanoyl fluoride may limit its distribution to the heart after formation in other organs. Therefore, other metabolites likely mediate the cardiotoxicity of 1,1,1,3,3-pentafluoropropane. Of the stable metabolites identified, trifluoroacetic acid shows little toxicity, but does induce peroxisome proliferation in the liver. Induction of myocarditis is not associated with other chemicals that are efficiently metabolized to trifluoroacetyl halides, such as halothane and 1,1,1-trifluoro-2,2-dichloroethane (8, 37, 38). 3,3,3-Trifluoropropanoic acid, even though it is a minor metabolite of 1,1,1,3,3-pentafluoropropane, may be a precursor of a toxic metabolite that may induce myocarditis. 3,3,3-Trifluoropropanoic acid is highly toxic in rats with an LD50 of less than 10 mg/kg (Amberg et al., unpublished observations). We speculate that the toxicity of 3,3,3-trifluoropropanoic acid may be associated with its biotransformation by mitochondrial β-oxidation to 3,3difluoroacryloyl-CoA, a highly reactive electrophilic R,βunsaturated carbonyl compound. Both 3-fluoropropanoylCoA and 3,3-difluoropropanoyl-CoA are substrates for butyryl-CoA dehydrogenase from Megasphaera elsdenii and give an R,β-unsaturated thioester as a product. (39). Reaction of 3,3-difluoroacryloyl-CoA with nucleophilic sites in mitochondria may induce mitochondrial dysfunction and myocarditis. The heart is highly dependent on mitochondrial fatty acid oxidation for energy production (40), and energy deprivation in the heart may be associated with the myocarditis seen after exposure to 1,1,1,3,3pentafluoropropane. Due to the low rate of metabolism of 1,1,1,3,3-pentafluoropropane to 3,3,3-trifluoropropanoic acid, high and prolonged exposure to 1,1,1,3,3-pentafluoropropane may be needed to induce toxicity. Moreover, the lower rates of 3,3,3-trifluoropropanoic acid formation in human liver microsomes as compared with rat liver microsomes indicate that humans would form less 3,3,3trifluoropropanoic acid than rats and, thus, may be at low risk for potential adverse effects after exposure to 1,1,1,3,3-pentafluoropropane. Support for this proposal is found in the observation that prolonged inhalation exposure of rats to 1,1,1,3tetrachloropropane (90-day inhalation exposure to 200 ppm) induces myocarditis (41). 1,1,1,3-Tetrachloropropane would be expected to be metabolized by cytochromes P450 and an aldehyde oxidase to 3,3,3-trichloropropanoic acid, which may be biotransformed by mitochondrial β-oxidation to 3,3-dichloroacryloyl-CoA. The analogous chloropropanoic acid is a substrate for acetyl coenzyme A synthetase (42). Indeed, 3-chloropropanoyl-CoA apparently inhibits fatty acid synthase via the formation of acryloyl thioester, which reacts with an active site nucleophile (43). The observed increased toxicity of 1,1,1,3-tetrachloropropane as compared with 1,1,1,3,3pentafluoropropane may be due to higher blood concentrations achieved during inhalation and to greater biotransformation to 3,3,3-trichloropropanoic acid. Experiments designed to investigate the roles of 3,3dihaloacrylic acids in the toxicity of 3,3,3-trihalopropanoic acids are warranted.
Acknowledgment. Research described in this article was funded by Honeywell Inc., Morristown, NJ. The excellent technical assistance of Marion Friedewald is gratefully acknowledged.
Bayer et al.
References (1) Collins, M. A., Rusch, G. M., Sato, F., Hext, P. M., and Millischer, J. R. (1995) 1,1,1,2-Tetrafluoroethane: repeat exposure inhalation toxicity in the rat, developmental toxicity in the rabbit, and genotoxicity in vitro and in vivo. Fundam. Appl. Toxicol. 25, 271280. (2) Rusch, G. M., Millischer, R. J., Derooij, C., Brooker, A. J., Hughes, E., and Coombs, D. (1995) Inhalation teratology and two-generation reproduction studies with 1,1-dichloro-1-fluoroethane (HCFC141b). Food Chem. Toxicol. 33, 285-300. (3) Brock, W. J., Trochimowicz, H. J., Millischer, R. J., Farr, C., Kawano, T., and Rusch, G. M. (1995) Acute and subchronic toxicity of 1,1-dichloro-1-fluoroethane (HCFC-141b). Food Chem. Toxicol. 33, 483-490. (4) Kawano, T., Trochimowicz, H. J., Malinverno, G., and Rusch, G. M. (1995) Toxicological evaluation of 1,1,1,2,2-pentafluorethane (HFC-125). Fundam. Appl. Toxicol. 28, 223-231. (5) Brock, W. J., Trochimowicz, H. J., Farr, C. H., Millischer, R.-J., and Rusch, G. M. (1996) Acute, subchronic, and developmental toxicity and genotoxicity of 1,1,1-trifluoroethane (HFC-134a). Fundam. Appl. Toxicol. 31, 200-209. (6) Malley, L. A., Frame, S. R., Elliott, G. S., Bentley, K. S., Brock, W. J., Trochimowicz, H. J., and Rusch, G. M. (1998) Chronic toxicity, oncogenicity, and mutagenicity studies with chlorotetrafluoroethane (HCFC-124). Drug Chem. Toxicol. 21, 417-447. (7) Rusch, G. M., Coombs, D., and Hardy, C. (1999) The acute, genetic, developmental, and inhalation toxicology of 1,1,1,3,3pentafluoropropane (HFC 245fa). Toxicol. Sci. 52, 289-301. (8) Rusch, G. M. (1991) Toxicity studies with new refrigerant gases. Toxicologist 11, 102. (9) Dekant, W. (1996) Toxicology of chlorofluorocarbon replacements. Environ. Health Perspect. 104, 75-83. (10) Andrews, L. S., and Snyder, R. (1991) Toxic effects of solvent and vapors. In Casarett and Doull’s Toxicology. The Basic Science of Poisons (Amdur, M. O., Doull, J., and Klaassen, C. D., Eds.) pp 681-722, Pergamon Press, New York. (11) Anders, M. W. (1993) Metabolism and toxicity of hydrofluorocarbons: current knowledge and needs for the future. Environ. Health Perspect. 96, 185-191. (12) Harris, J. W., Pohl, L. R., Martin, J. L., and Anders, M. W. (1991) Tissue acylation by the chlorofluorocarbon substitute 1,1-dichloro2,2,2-trifluoroethane (HCFC-123). Proc. Natl. Acad. Sci. U.S.A. 88, 1407-1410. (13) Reinhold: R. W. (1993) Trifluoroacetic acid toxicity. Internal Report, Hoechst AG, Frankfurt. (14) Kaminsky, L. S., Fraser, J. M., Seaman, M., and Dunbar, D. (1992) Rat liver metabolism and toxicity of 2,2,2-trifluoroethanol. Biochem. Pharmacol. 44, 1829-1837. (15) Davis, M. E. (1990) Subacute toxicity of trichloroacetic acid in male and female rats. Toxicology 63, 63-72. (16) Bull, R. J., Sanchez, I. M., Nelson, M. A., Larson, J. L., and Lansing, A. J. (1990) Liver tumor induction in B6C3F1 mice by dichloroacetate and trichloroacetate. Toxicology 63, 341-359. (17) Elcombe, C. R. (1985) Species differences in carcinogenicity and peroxisome proliferation due to trichloroethylene: a biochemical human hazard assessment. Arch. Toxicol. Suppl. 8, 6-17. (18) Urban, G., Speerschneider, P., and Dekant, W. (1994) Metabolism of the chlorofluorocarbon substitute 1,1-dichloro-2,2,2-trifluoroethane by rat and human liver microsomes: the role of cytochrome P450 2E1. Chem. Res. Toxicol. 7, 170-176. (19) Herbst, J., Ko¨ster, U., Kerssebaum, R., and Dekant, W. (1994) Role of P4502E1 in the metabolism of 1,1,2,2-tetrafluoro-1-(2,2,2trifluoroethoxy)-ethane. Xenobiotica 24, 507-516. (20) Ko¨ster, U., Speerschneider, P., Kerssebaum, R., Wittmann, H., and Dekant, W. (1994) Role of cytochrome P450 2E1 in the metabolism of 1,1,2,3,3,3-hexafluoropropyl methyl ether. Drug Metab. Dispos. 22, 667-672. (21) Koop, D. R. (1986) Hydroxylation of p-nitrophenol by rabbit ethanol-inducible cytochrome P-450 isoenzyme. Mol. Pharmacol. 29, 399-404. (22) Guengerich, F. P., Kim, D.-H., and Iwasai, M. (1991) Role of human cytochrome P-450 IIE1 in the oxidation of many low molecular weight cancer suspects. Chem. Res. Toxicol. 4, 168179. (23) Werner, M., Birner, G., and Dekant, W. (1995) The role of cytochrome P4503A1/2 in the sex-specific sulfoxidation of the hexachlorobutadiene metabolite, N-acetyl-S-(pentachlorobutadienyl)-L-cysteine in rats. Drug Metab. Dispos. 23, 861-868.
Biotransformation of HFC-245fa (24) Croes, K., Foulon, V., Casteels, M., Van Veldhoven, P. P., and Mannaerts, G. P. (2000) Phytanoyl-CoA hydroxylase: recognition of 3-methyl-branched acyl-coAs and requirement for GTP or ATP and Mg(2+) in addition to its known hydroxylation cofactors. J. Lipid Res. 41, 629-636. (25) Foulon, V., Antonenkov, V. D., Croes, K., Waelkens, E., Mannaerts, G. P., Van Veldhoven, P. P., and Casteels, M. (1999) Purification, molecular cloning, and expression of 2-hydroxyphytanoyl-CoA lyase, a peroxisomal thiamine pyrophosphatedependent enzyme that catalyzes the carbon-carbon bond cleavage during R-oxidation of 3-methyl-branched fatty acids. Proc. Natl. Acad. Sci. U.S.A. 96, 10039-10044. (26) Urban, G., and Dekant, W. (1994) Metabolism of 1,1-dichloro2,2,2-trifluoroethane in rats. Xenobiotica 24, 881-892. (27) Harris, J. W., Jones, J. P., Martin, J. L., LaRosa, A. C., Olson, M. J., Pohl, L. R., and Anders, M. W. (1992) Pentahaloethanebased chlorofluorocarbon substitutes and halothane: correlation of in vivo hepatic protein trifluoroacetylation and urinary trifluoroacetic acid excretion with calculates enthalpies of activation. Chem. Res. Toxicol. 5, 720-725. (28) Yin, H., Anders, M. W., Korzekwa, K. R., Higgins, L., Thummel, K. E., Kharasch, E. D., and Jones, J. P. (1995) Designing safer chemicals: predicting the rates of metabolism of halogenated alkanes. Proc. Natl. Acad. Sci. U.S.A. 92, 11076-11080. (29) Whitford, G. M. (1989) The metabolism and toxicity of fluoride. Monogr. Oral Sci. 13, 1-160. (30) Guengerich, F. P. (2001) Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity. Chem. Res. Toxicol. 14, 611-650. (31) Miller, R. E., and Guengerich, F. P. (1982) Oxidation of trichloroethylene by liver microsomal cytochrome P-450: evidence for chlorine migration in a transition state not involving trichloroethylene oxide. Biochemistry 21, 1090-1097. (32) Henschler, D., Hoos, W. R., Fetz, H., Dallmeier, E., and Metzler, M. (1979) Reactions of trichloroethylene epoxide in aqueous systems. Biochem. Pharmacol. 28, 543-548. (33) Swinney, D. C., and Mak, A. Y. (1994) Androgen formation by cytochrome P450 CYP17. Solvent isotope effect and pL studies
Chem. Res. Toxicol., Vol. 15, No. 5, 2002 733
(34)
(35)
(36) (37)
(38)
(39)
(40)
(41)
(42) (43)
suggest a role for protons in the regulation of oxene versus peroxide chemistry. Biochemistry 33, 2185-2190. Fischer, R. T., Trzaskos, J. M., Magolda, R. L., Ko, S. S., Brosz, C. S., and Larsen, B. (1991) Lanosterol 14-R-methyl demethylase. Isolation and characterization of the third metabolically generated oxidative demethylation intermediate. J. Biol. Chem. 266, 61246132. Pumford, N. R., and Halmes, N. C. (1997) Protein targets of xenobiotic reactive intermediates. Annu. Rev. Pharmacol. Toxicol. 37, 91-117. Van Stee, E. W., Eds. (1982) Cardiovascular Toxicology, Raven Press, New York. Malley, L. A., Carakostas, M. C., Hansen, J. F., Trochimowicz, H. J., and Rusch, G. M. (1991) Chronic toxicity of hydrochlorofluorocarbon HCFC-123. Toxicologist 11, 103. Rusch, G. M., Trochimowicz, H. J., Malley, L. J., Kelly, D. P., Peckham, J., Hansen, J., and Charm, J. B. (1994) Subchronic inhalation toxicity studies with hydrofluorocarbon 123 (HCFC 123). Fundam. Appl. Toxicol. 23, 169-178. Fendrich, G., and Abeles, R. H. (1982) Mechanism of action of butyryl-CoA dehydrogenase: reactions with acetylenic, olefinic, and fluorinated substrate analogues. Biochemistry 21, 66856695. Neely, J. R., and Morgan, H. E. (1974) Relationship between carbohydrate and lipid-metabolism and energy-balance of heartmuscle. Annu. Rev. Physiol. 36, 413-459. Kolesar, G. B., Siddiqui, W. H., Crofoot, S. D., Evans, M. G., and Meeks, R. G. (1995) Subchronic inhalation toxicity of 1,1,1,3tetrachloropropane in rats. Fundam. Appl. Toxicol. 25, 52-59. Patel, S. S., and Walt, D. R. (1987) Substrate specificity of acetyl coenzyme A synthetase. J. Biol. Chem. 262, 7132-7134. Miziorko, H. M., and Behnke, C. E. (1986) Active site directed inactivation of rat mammary gland fatty acid synthase by 3-chloropropionyl coenzyme A. Biochemistry 25, 468-473.
TX025505C
