Toxicology and Applied Pharmacology 263 (2012) 32–38

Contents lists available at SciVerse ScienceDirect

Toxicology and Applied Pharmacology

j ourna l homepage: www.e lsev ie r .com/ locate /ytaap

Biotransformation of 2,3,3,3-tetrafluoropropene (HFO-1234yf) in male, pregnant andnon-pregnant female rabbits after single high dose inhalation exposure

Tobias Schmidt a, Rüdiger Bertermann b, George M. Rusch c, Gary M. Hoffman d, Wolfgang Dekant a,⁎a Institut für Toxikologie, Universität Würzburg, Versbacher Str. 9, 97078 Würzburg, Germanyb Institut für Anorganische Chemie, Universität Würzburg, Am Hubland, 97074 Würzburg, Germanyc Honeywell, P.O. Box 1057, Morristown, NJ 07962–1057, United Statesd Huntingdon Life Sciences., East Millstone, NJ, United States

⁎ Corresponding author at: Department of ToxicolVersbacher Str. 9, 97078 Würzburg, Germany. Fax: +49

E-mail address: dekant@toxi.uni-wuerzburg.de (W.

0041-008X/$ – see front matter © 2012 Elsevier Inc. Alldoi:10.1016/j.taap.2012.05.019

a b s t r a c t

a r t i c l e i n f o

Article history:Received 2 March 2012Revised 25 May 2012Accepted 26 May 2012Available online 1 June 2012

Keywords:Acute toxicityBiotransformationRabbitPregnant rabbitsHaloolefinsHFO-1234yfMercapturic acids

2,3,3,3-Tetrafluoropropene (HFO-1234yf) is a novel refrigerant intended for use in mobile air conditioning. Itshowed a low potential for toxicity in rodents studies with most NOAELs well above 10,000 ppm in guidelinecompliant toxicity studies. However, a developmental toxicity study in rabbits showed mortality at exposurelevels of 5,500 ppm and above. No lethality was observed at exposure levels of 2,500 and 4,000 ppm. Never-theless, increased subacute inflammatory heart lesions were observed in rabbits at all exposure levels. Sincethe lethality in pregnant animals may be due to altered biotransformation of HFO-1234yf and to evaluate thepotential risk to pregnant women facing a car crash, this study compared the acute toxicity and biotransfor-mation of HFO-1234yf in male, female and pregnant female rabbits. Animals were exposed to 50,000 ppmand 100,000 ppm for 1 h. For metabolite identification by 19F NMR and LC/MS-MS, urine was collected for48 h after inhalation exposure. In all samples, the predominant metabolites were S-(3,3,3-trifluoro-2-hydroxypropanyl)-mercaptolactic acid and N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine. Sinceno major differences in urinary metabolite pattern were observed between the groups, only N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine excretion was quantified. No significant differences in recovery be-tween non-pregnant (43.10±22.35 μmol) and pregnant female (50.47±19.72 μmol) rabbits were observed,male rabbits exposed to 100,000 ppm for one hour excreted 86.40±38.87 μmol. Lethality and clinical signs oftoxicity were not observed in any group. The results suggest that the lethality of HFO-1234yf in pregnant rabbitsunlikely is due to changes in biotransformation patterns or capacity in pregnant rabbits.

© 2012 Elsevier Inc. All rights reserved.

Introduction

2,3,3,3-Tetrafluoropropene (HFO-1234yf) is a non-ozone-depletingfluorocarbon replacement with a low global warming potential (GWP,4 relative to CO2), absence of capacity for ozone depletion and a shortatmospheric life time of only 11 days (Nielsen et al., 2007). It is devel-oped as a replacement for halocarbons such as 1,1,1,2-tetrafluoroethane(HFA-134a), which is currently themost common refrigerant in mobileair conditioning systems (McCulloch, 1999).

Despite repeated exposure toxicity studies of HFO-1234yf in rodentswith inhalation NOAELs>10,000 ppm (Hamner Institute for HealthSciences, 2007) and a rat developmental toxicity study with a NOAELof 50,000 ppm (Waalkens-Berendsen, 2011), a developmental toxicitystudy in rabbits showed mortality at exposure levels of 5,500 ppmand above. In contrast, no lethality was observed at exposure levels of

ogy, University of Würzburg,931 201 3446.

Dekant).

rights reserved.

2,500 and 4,000 ppm of HFO-1234yf (WIL Research Laboratories,2008). Histopathology examination (WIL Research Laboratories, 2011)revealed subacute inflammation of the heart as a predominant effectboth in surviving and deceased maternal rabbits at all exposure levels.In addition, non dose-dependent mineralization was detected in theheart and liver of a few maternal rabbits.

Since toxicities of haloalkenes other than central nervous systemdepression are thought to be mediated by biotransformation reactions,the higher sensitivity of rabbits, in comparison to rats, may be due tospecies-specific biotransformation reactions or an influence of pregnan-cy on the extent or pathways of biotransformation of HFO-1234yf.However, studies comparing the biotransformation of HFO-1234yf inrats and female rabbits did not find major differences in metabolitepatterns or suggest the formation of cardiotoxic metabolites (Schuster,2009; Schuster et al., 2008, 2010). In rabbits (female), rats (male)and mice (male), the predominant metabolites of HFO-1234yf were S-(3,3,3-trifluoro-2-hydroxypropanyl)-mercaptolactic acid and N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine, both formed byrenal processing of S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine(Scheme 1). Signals attributed to these compounds represented over90% of total peak areas in 19F NMR. A sulfoxide of N-acetyl-S-(3,3,3-

F3CH

F

H

HFO 1234yf

CYP 450 2E1

F3CH

F

H

O

1. glutathione-S-transferase

F3CSG

OH

HH

2. reductase

F3CS

OH

H H

HNH

CO2H

CH3O

F3CS

OH

H H

OH

CO2H F3COH

OH

H H

1. γγ-glutamyl-transpeptidase2. dipeptidase

3. S-conjugate transaminase

1. γ-glutamyl-transpeptidase2. dipeptidase

3. renal cysteine S-conjugate β–lyase

1. γ-glutamyl-transpeptidase2. dipeptidase

3. N-acetyl-transferase4. hydrolysis

*

*** *

???

F3C O

OH

1

2

3

4 5 6

7

*

sulfoxidationsulfoxidation

4. reductase

*

Scheme 1. Biotransformation of HFO-1234yf. 1, 2,3,3,3-tetrafluoropropene (HFO-1234yf); 2, 2,3,3,3-tetrafluoro-1,2-epoxypropane; 3, S-(3,3,3-trifluoro-2-hydroxypropanyl)glutathione; 4,N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine; 5, S-(3,3,3-trifluoro-2-hydroxypropanyl)mercaptolactic acid; 6, 3,3,3-trifluoro-1,2-dihydroxypropane; 7, 3,3,3-trifluoroacetic acid.

33T. Schmidt et al. / Toxicology and Applied Pharmacology 263 (2012) 32–38

trifluoro-2-hydroxypropanyl)-L-cysteine represented a minor metab-olites of HFO-1234yf and the only species differences observedwere the absence of the trace metabolites 3,3,3-trifluorolactic acid,3,3,3-trifluoroacetic acid, 3,3,3-trifluoroacetone and 3,3,3-trifluoro-1-hydroxyacetone in rabbit urine (Schuster et al., 2008, 2010).

Regarding potentially changedmetabolism during pregnancy, infor-mation onmaternal hepatic enzyme activities of cytochrome P 450 2E1or glutathione S-transferase in pregnant rabbits or humans (Anderson,2005) is not available. These enzymes are involved in the biotransfor-mation of HFO-1234yf. However, pregnant rats show decreased mater-nal capacity of CYP 2E1 (Czekaj et al., 2005; He et al., 2005; Songand Cederbaum, 1996), but an increased detoxification of epoxidesby glutathione S-transferase (Polidoro et al., 1981). Therefore, thesechanges in biotransformation capacities may influence the toxicityprofile of HFO-1234yf in pregnant animals.

An acute inhalation toxicity study was designed to investigatewhether there is a gender-based or pregnancy-based difference inacute toxic response and biotransformation of HFO-1234yf in rabbits.Pregnant rabbits were exposed on gestation day (GD) 12, since mortal-ity in the 50,000 ppm exposure group started to occur at GD 13.Furthermore, this short term high concentration exposure was aimedto evaluate a potential risk to pregnant women facing a car crash witha rupture in the air conditioning system resulting in a release of highconcentrations of HFO-1234yf into the passenger compartment.

Materials and methods

Chemicals. 2,3,3,3-Tetrafluoroprop-1-ene(HFO-1234yf, purity of 99,99%)was supplied by Honeywell (Morristown, New Jersey) and 3,3,3-trifluoro-1,2-epoxypropane was purchased from Matrix Scientific(Columbia, South Carolina). All other chemicals were purchased fromSigma-Aldrich (Taufkirchen, Germany) in the highest available purity.

Synthesis of N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine.This compound was synthesized as previously described (Schusteret al., 2008). For purification, the reaction product was precipitated asdicyclohexylammonium salts and recrystallized from ethyl acetate. Twodiastereomers of N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine were obtained in a ratio of 5:1 according to HPLC-analysiswith UV-detection at 225 nm. Purity was determined>96% basedon 1H-NMR (500 MHz, [D6]-DMSO): δ=1.02–1.99 (m, 20 H, cyclo-hexane –CH2–), 1.82 (s, 3 H, –COCH3), 2.60 (dd, JAB=13.6 Hz, JAX=10.0 Hz, 1 H, HA), 2.73 (dd, JBA=13.6 Hz, JBX=3.3 Hz, 1 H, HB), 2.82(dd, JAB=13.5 Hz, JAX1=6.1 Hz, 1 H, HA), 2.97 (dd, JBA=13.5 Hz,JBX1=5.1 Hz, 1 H, HB), 3.92–3.00 (m, 2 H, cyclohexane –NCH–), 4.01 (m,1 H,HX), 4.1 (m, 1 H, HX1), 7.58 (d, J=7.25 Hz, 1 H, NHCOCH3).

Animals and treatment. Male and female (pregnant and non-pregnant)albino New ZealandWhite Rabbits (Robinson Services, Inc., Mocksville,

34 T. Schmidt et al. / Toxicology and Applied Pharmacology 263 (2012) 32–38

North Carolina) were used for all exposures. Average animal weightswere 2.994 kg (range 2.622 kg to 3.677 kg) for male rabbits, 3.258 kg(range 3.024 kg to 3.548 kg) for non-pregnant female rabbits and3.445 kg (range 3.119 kg to 3.699 kg) for presumed pregnant femalerabbits at the initiation of inhalation exposures. Age at the receipt wasapproximately 5 to 6 months and the pregnant animals were on GD 2where GD 0 was the day of mating. All animals were stabilized for atleast 5 days and examined during stabilization period to confirm suit-ability for this study.

Exposure of rabbits to HFO-1234yf. The single inhalation exposureby whole-body inhalation and urine sampling were conducted byHuntingdon Life Sciences (100Mettlers Road, EastMillstone, New Jersey).

Non-pregnant female, pregnant female (GD 12) and male NewZealand White Rabbits were exposed by whole-body inhalation for1 h to 2,3,3,3-tetrafluoropropene (HFO-1234yf) at two different expo-sure concentrations (n=5–6/group). Non-pregnant rabbits were notexposed to 50,000 ppm as toxicity and biotransformation under thisexposure conditions were previously determined (Schuster et al.,2010).

The whole-body exposure chambers (volume approximately1 m3) were operated at a minimum flow rate of 200 L/min. HFO-1234yf was metered into the inflowing air and thus diluted to thetarget exposure concentrations. These settings provided at least 12 airchanges per hour. Chamber oxygen levels were measured every 30 minwith an oxygen analyzer and maintained at a concentration of at least19% (v/v) throughout each inhalation experiment. Determination of theexposure levels was performed by a MIRAN® Ambient Air analyzerequippedwith a strip chart recorder. The test atmospherewasmonitoredcontinuously and exposure levels were determined by comparison of themeasured absorbance to a calibrated response curve.

Urine was collected at approximately 12 h before initiation of theinhalation exposure and for approximately 48 hours after exposure, atapproximately 12 h intervals from 5 males/group, up to 6 pregnantfemales/group and 5 non-pregnant females/group. Urine was collectedin ice chilled containers and volume was measured for each interval.

Macroscopic and histopathologic examinations. Following an observa-tion period of 14 days after inhalation exposure, the animals weresacrificed by Phenobarbital injection. Macroscopic observationswere performed on all animals and the pregnancy status was deter-mined for the presumed pregnant animals. The external surface andall orifices as well as the external surfaces of the brain and spinalcord were examined. In addition, organs and tissues of the cranial,thoracic, abdominal and pelvic cavities and neck were examined.For histopathology studies, tissue samples of the heart, kidneys,liver, lungs and gross lesions were preserved and slides of thesewere examined microscopically.

Quantification of inorganic fluoride concentrations. Inorganic fluoridein urine samples was quantified as described previously (Schusteret al., 2008). Achieved R2-values of the calibration curves were >0.98.

Instrumental analyses. 19F NMR spectra were recorded with a BrukerDRX 300 NMR spectrometer with a 5 mm fluoride probe operatingat 282.4 MHz. 19F chemical shifts were referenced to external CFCl3.Spectra were recorded with a 30° pulse, a pulse length of 13.6 μs anda cycle delay of 6 μs. The acquisition time was 3 sec and 10240 scanswere recorded to obtain a good signal to noise ratio (S/N). The 19Fspectra were acquired with proton coupling and a spectral width of200 ppm (10 to −190 ppm). Before the integration of peak areas, anaccurate baseline correction and spectra phasing was performedbetween −70 and −85 ppm chemical shift. Sample preparation:After thawing at 4 °C, 1 mL of a rabbit urine sample was vortexed andthen centrifuged for 15 min at 14,000 rpm and 4 °C. An aliquot(720 μL) of the obtained supernatant was added to 80 μL of deuterium

oxide. No further sample workup was performed before 19F NMR anal-ysis. NMR spectra were analyzed with TopSpin 3.0 (Bruker). In eachspectrum signals were normalized to the integral of S-(3,3,3-trifluoro-2-hydroxypropanyl)-mercaptolactic acid (set to 1).

LC-MS/MS analyses were performed with an API 3000 mass spec-trometer (Applied Biosystems, Darmstadt, Germany) coupled to anAgilent 1100 HPLC-pump equipped with an Agilent 1100 autosampler(Agilent, Waldbronn, Germany).

Quantification ofN-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine.Rabbit urine was analyzed by LC-MS/MS using electrospray ionization.Before analysis, frozen urine samples were thawed at 4 °C, a 1 mL aliquotwas vortexed and then centrifuged for 15 min at 14,000 rpm and 4 °C.The obtained supernatants were diluted with water (1:10), 10 μL of thedilutions were added to 90 μL of a 500 pg/μL internal standard solution(N-acetyl-S-(2-chloro-1,1,2-trifluoroethyl)-L-cysteine), vortexed, trans-ferred into glass autosampler vials, sealed and aliquots (10μL) wereanalyzed. If sample values were outside the linear range of the calibrationcurve, an additional dilution by 100 or 1,000 fold was performed.Quantifications were performed as described previously (Schusteret al., 2010). The R2-values of the calibration curves were >0.997.The limit of quantification (LOQ) for N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine was 0.45 μg/L (1.63 nmol/L) with a limitof detection (LOD) of 0.14 μg/L (0.51 nmol/L) in rabbit urine.

Statistical analysis. All data was analyzed by Prism 5.01 (GraphPadSoftware, Inc, La Jolla, CA 92037 USA).

Results

Biotransformation of HFO-1234yf in male, non-pregnant and pregnantrabbits

To obtain information on differences in the biotransformationof HFO-1234yf in male, non-pregnant and pregnant female rabbits,animals were exposed to 0 ppm, 50,000 ppm and 100,000 ppm bywhole-body inhalation. Measured exposure chamber concentrationswere 47,000±1,530 ppm, 102,000±7,640 ppm for non-pregnantand pregnant female and 45,000±4,000 ppm, 100,000±0 ppm formale rabbits. Collected urine samples were analyzed by 19F NMRspectroscopy and LC-MS/MS.

Metabolite identification by 19F NMR

Biotransformation products of HFO-1234yf were identified by 19FNMR spectrometry, comparing the chemical shifts and F-H couplingconstants with the 19F NMR spectra of synthetic standards (Schuster,2009; Schuster et al., 2010). 19F NMR signals were not presentin urine samples of control animals. Signals found in the spectra ofurine fractions collected within 48 hours after the inhalation wereassigned to 3,3,3-trifluoro-propan-1,2 diol (δ=−77.2 ppm; d, JHF=7.4 Hz), N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine (δ=−78.7 ppm; d, JHF=6.6 Hz), S-(3,3,3-trifluoro-2-hydroxypropanyl)mercaptolactic acid (δ=−78.9 ppm; d, JHF=6.9 Hz) and 3,3,3-trifluoroacetic acid (δ=−75.4 ppm, s) (Fig. 1). Beside these knownmetabolites, there were two minor unidentified metabolites.

The 19F chemical shifts (δ=−77.1 and −78.4 ppm) of these indi-cate that the F3C-group is still present in the structure (Berger et al.,1994). Metabolite b (Fig. 1) is likely a conjugation product of 3,3,3-tri-fluoro-propan-1,2-diol due to a very similar 19F chemical shift, multi-plicity and 3JHF-coupling constant. The slightly more positive 19Fchemical shift as well as the lower 3JHF-coupling constant may indicatethat the conjugate represents a sulfate or glucuronide of the hydroxidegroup resulting in a higher polarity of the chemical bond. Metabolite d(Fig. 1) likely is an alternative derivate of the 2,3,3,3-tetrafluoroprop-1-ene glutathione-S conjugation pathway due to a similar 19F chemical

Fig. 1. 19F NMR spectra of (A) male, (B) pregnant female, (C) non-pregnant female rabbit urine (100,000pmm, 0–12 h); resonances were assigned to (a) 3,3,3-trifluoroacetic acid(d=−75.4 ppm, s), (b) unidentified metabolite (d=−77.1 ppm; d, JHF=7.2 Hz), (c) 3,3,3-trifluoro-propan-1,2-diol (d=−77.2 ppm; d, JHF=7.4 Hz), (d) unidentified metabolite(d=−78.4 ppm; d, JHF=6.8 Hz), (e) N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine (d=−78.7 ppm; d, JHF=6.6 Hz), (f) S-(3,3,3-trifluoro-2-hydroxypropanyl)-mercaptolactic acid (d=−78.9 ppm; d, JHF=6.9 Hz).

35T. Schmidt et al. / Toxicology and Applied Pharmacology 263 (2012) 32–38

shift, multiplicity and 3JHF-coupling constant as S-(3,3,3-trifluoro-2-hydroxypropanyl)-mercaptolactic acid and N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine. The more positive 19F chemical shiftmay again indicate the introduction of a more electronegativesubstituent.

To assess possible differences in the biotransformation of HFO-1234yf between the groups of rabbits, 19F NMR data were compared.Fig. 1 illustrates that there were no significant differences betweenthe 19F spectra of urine from female, pregnant female and male rabbitsafter a single one hour exposure to 100,000 ppm. All spectra showedidentical major signals with very similar relative intensities. The deter-mined urinary pattern of fluorine-containingmetabolites was similar tothat previously reported (Schuster et al., 2010). The only difference ob-served was that S-(3,3,3-trifluoro-2-hydroxypropanyl)-mercaptolacticacid was the major biotransformation product instead of N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine (Schuster et al., 2010).

For further characterization of metabolism, relative peak areas(related to area of the signal of S-(3,3,3-trifluoro-2-hydroxypropanyl)-mercaptolactic acid) of the six major metabolic products were de-termined and compared between the different study groups. Asshown in Fig. 2, therewere nomajor differences inmetabolite excretionbetween male and female animals, between the dose groups50,000 ppm and 100,000 ppm and between non-pregnant femaleand pregnant female animals at all sampling times.

Urine concentratios of N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine

In all urine samples, concentrations of N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine were determined by LC-MS/MS in orderto evaluate differences in the excretion kinetics of the glutathione-Sconjugation pathway (Fig. 3). Peak urine excretions were reachedwithin

12 hours after termination of the exposures (male 100,000 ppm–63.55±29.00 μmol, male 50,000 ppm–53.76±16.37 μmol, non-pregnant female100,000 ppm–32.51±20.47 μmol, pregnant female 50,000 ppm–45.07±13.11 μmol) except for the study group pregnant females 100,000 ppm.In this group, peak excretion occurred between 12 and 24 hours afterthe end of exposure (27.17±11.76 μmol). Metabolite eliminationfollowed 1st order kinetics with a half-life of approx. 6 hours in allgroups and >83% of N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine excreted was recovered within 24 hours. Elimination withinthe first 12 hours in the 100,000 ppm pregnant female group (35.9±24.4% of total metabolite excretion) was slower compared to all othergroups (71.8±14.5%). However, this is likely due to significantly lowerurine output in pregnant female rabbits within the first 12 hours afterinhalation compared to the non-pregnant female group (Table 1).

The total urinary recovery of N-acetyl-S-(3,3,3-trifluoro-2-hydro-xypropanyl)-L-cysteine within 48 hours after exposure was higher inmales (male 100,000 ppm–86.40±38.87 μmol; male 50,000 ppm–

67.86±22.95 μmol) as compared to females (pregnant 100,000 ppm–

50.47±19.72 μmol; pregnant 50,000 ppm–60.97±13.05 μmol; non-pregnant 100,000 ppm–43.10±22.35 μmol). No significant differences(Table 1) were observed between non-pregnant females and pregnantfemales as well as between pregnant and male (P=0.08, Student'st-test) and non-pregnant and male (P=0.06, Student's t-test) rabbitsin the 100,000 ppm group. Since the quantity of metabolite recoveredat 100,000 ppm is close to the quantity recovered at 50,000 ppm, uptakeand/or metabolism of HFO-1234yf are likely saturated above concentra-tions of 50,000 ppm, as observed for similar substances in previousstudies (Gargas et al., 1986).

Urinary excretion of inorganic fluoride within 48 hours (Table 1) alsoshowed no significant differences (P=0.1, Student's t-test) betweenpregnant (3.03±1.76 μmol) and non-pregnant (4.44±1.65 μmol)rabbits (100,000 ppm exposure).

A B

C D

E F

Fig. 2. Relative amount ofmetabolite excretion in 12 hour intervals for (one animal per group): (A) S-(3,3,3-trifluoro-2-hydroxypropanyl)-mercaptolactic acid, (B)N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine, (C) unidentified metabolite (d=−78.4 ppm), (D) 3,3,3-trifluoro-1,2-dihydroxypropane, (E) unidentified metabolite (d=−77.1 ppm), (F) 3,3,3-trifluoroacetic acid.

36 T. Schmidt et al. / Toxicology and Applied Pharmacology 263 (2012) 32–38

Macroscopic and histopathologic examinations

There were neither macroscopic nor microscopic exposure relatedfindings. Macroscopic abnormalities in all groupswere of low incidenceand were rated as common incidental findings in the test species. Thisincluded the finding of discolored lungs in a small number of femalesexposed to HFO-1234yf or in controls. Microscopic findings consistedof interstitial inflammatory cell infiltrates in the lungs and minimalfocal alveolar ossification.

Discussion

The biotransformation reactions for HFO-1234yf in this studyare in good agreementwith previousmetabolic studies confirming bio-transformation by CYP P450 2E1 oxidation followed by conjugation ofthe intermediate epoxide with glutathione (Scheme 1). In contrast toprevious studies in rabbits, the predominant metabolite in urine was

S-(3,3,3-trifluoro-2-hydroxypropanyl)-mercaptolactic acid instead ofN-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine. Both me-tabolites are formed by renal processing of S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine, either by N-acetylation or by ß-oxidation.Possible explanations for the difference in renal processing could be therabbit strain, housing conditions or diet. Besides this, the recorded 19FNMR spectra showed no major differences in metabolite patternsbetween non-pregnant and pregnant female animals (Figs. 1, 2).Among the biotransformation products, metabolites suggesting changesin toxication pathways were not identified. The formation of glutathioneS-conjugationmetabolites (Anders andDekant, 1998; Anders et al., 1992;Dekant, 1996), derived from HFO-1234yf, do not provide a conclusiveexplanation for myocardial inflammation (WIL Research Laboratories,2008) since the very small amounts of epoxide formed from HFO-1234yf in the liver are expected to be rapidly detoxified. Heart tissueonly has very low CYP 2E1 activity (Gottlieb, 2003; Thum and Borlak,2000) and thus epoxide formation in the heart likely is very limited. It

A B C

Fig. 3. Kinetics of urinary elimination of N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine (A, B) and inorganic fluoride (C) within 48 hours after whole-body inhalation ofmale (▲), non-pregnant female (●) and pregnant female (■) rabbits to 100,000 or 50,000 ppm of HFO-1234yf.

37T. Schmidt et al. / Toxicology and Applied Pharmacology 263 (2012) 32–38

is also considered unlikely that the low amounts of metabolites formedby ring-opening of the epoxide are involved in the cardiac effects inducedby HFO-1234yf in rabbits, since they are highly polar, have little chemicalreactivity and are rapidly excreted.

Based on urine recovery of N-acetyl-S-(3,3,3-trifluoro-2-hydro-xypropanyl)-L-cysteine, HFO-1234yf-metabolites are rapidly excreted inall study groups. More than 83% of the metabolite in male and morethan 93% in non-pregnant female/pregnant female animals was excretedwithin 24hours after inhalation exposure (50,000 ppmand100,000 ppmexposures). However, pregnant females in the 100,000 ppm groupeliminated approximately only half of the metabolite amount withinthe first 12 hours after exposure compared to non-pregnant femalesexposed to the same concentration of HFO-1234yf. On the other hand,total excretion of N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine within 48 hours was not significantly (P=0.58, Student'st-test) different between non-pregnant female and pregnant femaleanimals. Analysis of the free inorganic fluoride kinetics of pregnantand non-pregnant female rabbits exposed to 100,000 ppm also showedthe same characteristics with a rapid excretion of more than 70% of thewhole excretion recovery determined within 48 h in the first 24 h andno significant excretion differences (P=0.2, Student's t-test) within48 hours after termination of inhalation exposure (Fig. 3). The calculat-ed dosage of inorganic fluoride resulting of a one hour exposure wasbetween 0.9 and 1.7 mg/kg/KG. These were far below the NOAELs for

Table 1Urinary excretion of inorganic fluoride, N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-c24 hours after inhalation exposure to HFO-1234yf.

Fraction Compound and specimen

0–12 h Inorganic fluoride – F− [μmol]12–24 h24–36 h36–48 h0–48 h0–12 h N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine [μmol]12–24 h24–36 h36–48 h0–48 h0–12 h Urine volume [mL]12–24 h24–36 h36–48 h0–48 h

Statistical analysis was performed by Student's t-test. Significant changes between non-pre

inorganicfluoride regardingmaternal (18 mg/kg/KG) and developmenttoxicity NOAELs (29 mg/kg/KG)(Heindel et al., 1996).

The low fluoride excretion in comparison to the content of N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine in the rabbiturine samples, is likely due to the disposition of inorganic fluoridesince only about 15% of fluoride intake is excreted in the urine (Hallet al., 1977). In addition, rabbit blood and urine is high in calcium(Redrobe, 2002) which may further reduce fluoride concentrationsin urine due to formation of insoluble calcium fluoride.

Based on mean body weights of 2,994 g in male, 3,258 g in non-pregnant female and 3,444 g in pregnant female rabbits, respiratoryminute volume was calculated as 1.548 L min−1, 1.663 L min−1 and1,744 L min−1 respectively (Alexander et al., 2008). Therefore, the totaldose of HFO-1234yf received within one hour of whole-body inhalationexposure was calculated as 386 mmol (male 100,000 ppm), 193 mmol(male 50,000 ppm), 415 mmol (non-pregnant female 100,000 ppm),435 mmol (pregnant female 100,000 ppm) and 217 mmol (pregnantfemale 50,000 ppm) (Table 2).

The amount of HFO-1234yf metabolized was calculated to beless than 0.1% of the received dose in all groups. These results andthe absence of clinical signs of toxicity suggest that lethality of HFO-1234yf in pregnant rabbits after inhalation exposure is unlikely dueto changes in biotransformation patterns or capacity of pregnantrabbits.

ysteine and urine output by non-pregnant (F) and pregnant female (P) rabbits within

F 100,000 ppm P 100,000 ppm Student's t-test,p-value

2.7±1.3 1.5±1.3 0.170.6±0.3 0.8±0.4 0.370.6±0.3 0.5±0.2 0.200.5±0.4 0.3±0.1 0.104.4±1.7 3.0±1.8 0.12

32.5±20.5 19.9±14.9 0.298.1±4.0 27.2±11.8 0.01 **2.0±1.5 5.3±3.2 0.110.9±1.0 2.8±1.6 0.05

43.1±22.4 50.5±19.7 0.58193.2±98.2 44.7±41.6 0.01 **66.4±44.3 64.5±36.6 0.9487.2±79.8 63.8±43.1 0.2493.6±68.0 45.8±34.7 0.20

446.4±209.3 200.5±136.0 0.04 *

gnant and pregnant rabbits are indicated as *, pb0.05 and **, pb0.01.

Table 2Biotransformation extent of HFO-1234yf in male (M), female (F) and pregnant female (P) rabbits after one hour exposure.

Group N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteinemean excretion within48 hours [μmol]

N-acetyl-S-(3,3,3-trifluoro-2-hydroxypropanyl)-L-cysteine mean fraction of the five mainmetabolites [%]a

Total fluoro metabolitesexcretion [μmol]

Amount of HFO-1234yfinhaled [μmol]b

Metabolizedamount [%]

M 50,000 ppm 75.24±15.67 35.83 209.99 192,976 0.11P 50,000 ppm 60.97±13.11 37.27 163.59 217,415 0.08M 100,000 ppm 86.40±38.68 32.70 264.22 385,952 0.07P 100,000 ppm 50.47±19.72 31.33 161.09 434,830 0.04F 100,000 ppm 43.10±22.35 29.10 148.11 414,755 0.04

a Based on the metabolic pattern of one rabbit per group.b Mole volume of HFO-1234yf was calculated to be 24.06 L/ mol under standard conditions (293.15 K, 1013.25 hPa).

38 T. Schmidt et al. / Toxicology and Applied Pharmacology 263 (2012) 32–38

Conflict of interest

No competing interests.

Acknowledgement

This work was supported by Honeywell International Inc.,Morristown, NJ.We gratefully thank JustineM. Damiano for conductingand coordinating the animal exposures and Heike Keim-Heusler for herexcellent technical assistance.

References

Alexander, D.J., Collins, C.J., Coombs, D.W., Gilkison, I.S., Hardy, C.J., Healey, G.,Karantabias, G., Johnson, N., Karlsson, A., Kilgour, J.D., McDonald, P., 2008. Associa-tion of Inhalation Toxicologists (AIT) working party recommendation for standarddelivered dose calculation and expression in non-clinical aerosol inhalation toxi-cology studies with pharmaceuticals. Inhal. Toxicol. 20, 1179–1189.

Anders, M.W., Dekant, W., 1998. Glutathione-dependent bioactivation of haloalkenes.Annu. Rev. Pharmacol. Toxicol. 38, 501–537.

Anders, M.W., Dekant, W., Vamvakas, S., 1992. Glutathione-dependent toxicity.Xenobiotica 22, 1135–1145.

Anderson, G.D., 2005. Pregnancy-induced changes in pharmacokinetics: a mechanistic-based approach. Clin. Pharmacokinet. 44, 989–1008.

Berger, S., Braun, S., Kalinowski, H.-O., 1994. NMR-Spektroskopie von Nichtmetallen,19F-NMR-Spektroskopie. George Thieme Verlag, Stutgart, New York.

Czekaj, P., Wiaderkiewicz, A., Florek, E., Wiaderkiewicz, R., 2005. Tobacco smoke-dependent changes in cytochrome P450 1A1, 1A2, and 2E1 protein expressionsin fetuses, newborns, pregnant rats, and human placenta. Arch. Toxicol. 79,13–24.

Dekant,W., 1996. Toxicology of chlorofluorocarbon replacements. Environ. Heal. Perspect.104, 75–83.

Gargas, M.L., Andersen, M.E., Clewell III, H.J., 1986. A physiologically based simulationapproach for determining metabolic constants from gas uptake data. Toxicol. Appl.Pharmacol. 86, 341–352.

Gottlieb, R.A., 2003. Cytochrome P450: major player in reperfusion injury. Arch.Biochem. Biophys. 420, 262–267.

Hall, L.L., Kilpper, R.W., Smith, F.A., Morken, D.A., Hodge, H.C., 1977. Kinetic model offluoride metabolism in the rabbit. Environ. Res. 13, 285–302.

Hamner Institute for Health Sciences, 2007. Toxicogenomic assessment of the carcinogenicpotential of 2,3,3,3-tetrafluoropropene, Report no. 06014, conducted for HoneywellInternational. Hamner Institute for Health Sciences, Research Triangle Park, NC.

He, X.J., Ejiri, N., Nakayama, H., Doi, K., 2005. Effects of pregnancy on CYPs proteinexpression in rat liver. Exp. Mol. Pathol. 78, 64–70.

Heindel, J.J., Bates, H.K., Price, C.J., Marr,M.C., Myers, C.B., Schwetz, B.A., 1996. Developmentaltoxicity evaluation of sodiumfluoride administered to rats and rabbits in drinkingwater.Fundam. Appl. Toxicol. 30, 162–177.

McCulloch, A., 1999. CFC and Halon replacements in the environment. J. Fluor. Chem.100, 163–173.

Nielsen, O.J., Javadi, M.S., Andersen, M.P.S., Hurley, M.D., Wallington, T.J., Singh, R.,2007. Atmospheric chemistry of CF3CF=CH2: Kinetics and mechanisms of gas-phasereactions with Cl atoms, OH radicals, and O-3. Chem. Phys. Lett. 439, 18–22.

Polidoro, G., Diilio, C., Arduini, A., Federici, G., 1981. Effect of Pregnancy on HepaticGlutathione S-Transferase Activities in the Rat. Biochem. Pharmacol. 30, 1859–1860.

Redrobe, S., 2002. Calciummetabolism in rabbits. Semin. Avian Exot. Pet Med. 11, 94–101.Schuster, P., 2009. Biotransformation of trans-1,1,1,3-tetrafluoropropene, 2,3,3,3-tetra-

fluoropropene and 1,2,3,3,3-pentafluoropropene. Department of Pharmacology andToxicology. University of Wuerzburg, Wuerzburg.

Schuster, P., Bertermann, R., Snow, T.A., Han, X., Rusch, G.M., Jepson, G.W., Dekant, W.,2008. Biotransformation of 2,3,3,3-tetrafluoropropene (HFO-1234yf). Toxicol. Appl.Pharmacol. 233, 323–332.

Schuster, P., Bertermann, R., Rusch, G.M., Dekant, W., 2010. Biotransformation of 2,3,3,3-tetrafluoropropene (HFO-1234yf) in rabbits. Toxicol. Appl. Pharmacol. 244, 247–253.

Song, B.J., Cederbaum, A.I., 1996. Ethanol-inducible cytochrome P450 (CYP2E1):Biochemistry, molecular biology and clinical relevance: 1996 update. Alcohol. Clin. Exp.Res. 20, A138–A146.

Thum, T., Borlak, J., 2000. Gene expression in distinct regions of the heart. Lancet 355,979–983.

Waalkens-Berendsen, D.H., 2011. Inhalatory two-generation reproduction toxicitystudy with HFO-1234yf in Wistar rats, study code V7931, conducted for HoneywellInternational. TNO, Utrechtseweg 48, P.O Box 360, 3700 AJ Zeist, The Netherlands.

WIL Research Laboratories, 2008. An inhalation range-finding prenatal developmentaltoxicity study of HFO-1234yf (2,3,3,3-tetrafluoropropene) in rabbits, study no.WIL-447021, conducted for Honeywell International. WIL Research Laboratories,LLC, 1407 George Road, Ashland, OH. 44805–8946.

WIL Research Laboratories, 2011. An inhalation prenatal developmental toxicity study ofHFO-1234yf (2,3,3,3-tetrafluoropropene) in rabbits, revised final report, study no.WIL-447022, conducted for Honeywell International. WIL Research Laboratories,LLC, 1407 George Road, Ashland, OH. 44805–8946.