Chem. Res. Toxicol. 1996,8, 182-187
182
Articles Analysis of Butadiene, Butadiene Monoxide, and Butadiene Dioxide in Blood by Gas Chromatography/Gas ChromatographylMass Spectroscopy William E. Bechtold," Michael R. Strunk, Janice R. Thornton-Manning, and Rogene F. Henderson Inhalation Toxicology Research Institute, Lovelace Biomedical and Environmental Research Institute, P.O. Box 5890, Albuquerque, New Mexico 87185 Received March 10, 1994; Revised Manuscript Received November 4, 1994@
A new method was developed to quantify the levels of 1,3-butadiene (BD), butadiene monoxide (BDO), and butadiene diepoxide (BD02) in blood. The method was based on vacuum distillation of tissues followed by analysis of the distillates using multidimensional GC/MS. Metabolites isolated from blood by vacuum distillation were condensed into a cold trap. After warming the traps to room temperature, BD and BDO were sampled from the trap vapor phase. BD02 was extracted from the codistilled water phase using ethyl acetate. Samples were analyzed using a multidimensional GC system equipped with a custom-built interface. The method was validated by analysis of 0.75-mL aliquots of mouse blood spiked with 5.0, 3.4, and 0.55 nmol of BD, BDO, and BD02, respectively. The recoveries of analytes were 96 f 18%, 125 ik 15%, and 98 f 12%, respectively (mean f SD, n = 6). Kinetic studies indicated no loss of BDO and BDOz in blood held at room temperature in closed containers for u p to 1 h. The method was applied to blood samples from B6C3F1 mice and Sprague-Dawley rats exposed by inhalation (nose-only) to 100 ppm BD for 4 h. Blood levels of BD and BDO in exposed rats were 4.1 f 1.0 and 0.10 f 0.06 pM,respectively (mean f SD, n = 6). Levels of BD02 were below the limits of detection (0.01 nmoVmL). Blood levels of BD, BDO, and BDOz in mice exposed to 100 ppm BD for 4 h were 2.9 f 1.3, 0.38 f 0.14, and 0.33 f 0.19 pM, respectively (mean f SD, n = 6). The results for BDO and BDO2 compare favorably with those previously determined by a less specific method a t our Institute. Overall, the results suggest that although both rats and mice have detectable levels of BDO in their blood following similar BD exposures, only mice have detectable levels of BDOz. This metabolite may therefore be critical in explaining the marked species difference in BD toxicity. Future studies will confirm these results over a variety of time points and exposure concentrations.
Introduction 1,3-Butadiene (BDY is a monomer used in the production of synthetic rubber. In 1992, production of BD was ranked twenty-first among industrial chemicals ( 1). BD is also a product of automotive exhaust (2) and is found in cigarette smoke (3). Because of the industrial importance of BD and its prevalence in the environment, BD has been the object of several carcinogenicity studies. The carcinogenicity of BD has been determined in life-span inhalation studies in Sprague-Dawley rats (4)and B6C3F1 mice (5). Results from these studies suggest a remarkable species difference in the carcinogenic effects of BD. For, example, female mice exposed to concentrations of BD as low as 6.25 ppm exhibited increased alveolar/ bronchiolar neoplasms. In contrast, BD was only a weak * Address correspondence to this author at the Inhalation Toxicology Research Institute, P.O. Box 5890,Albuquerque, NM 87185 (505) 8451027. @Abstractpublished in Advance ACS Abstracts, January 1, 1995. Abbreviations: 1,3-butadiene (BD); 1,2-dihydroxy-4-(N-acetylcystein8-y1)butane (M-I); (1 or 2)-hydroxy(N-acetylcystein-S-yl)-3-butene (M-11); butadiene monoepoxide (BDO); butadiene diepoxide (BD02); Hewlett-Packard (HP); glutathione (GSH);gas chromatography/gas chromatography/mass spectroscopy (GCIGCIMS).
carcinogen in Sprague-Dawley rats, which were observed to have an increase in only mammary tumors after exposure to 1000 ppm BD. The disparity between species in the toxicity may in part be a result of differences in the metabolism of BD to reactive intermediates. We previously analyzed the urinary metabolites of butadiene and have identified two metabolites, 1,2-dihydroxy-4-(N-acetylcystein-S-yl)-butane (M-I), and (1or 2)-hydroxy(N-acetylcystein-S-y1)-3butene (M-111, in the urine of B6C3F1 mice, SpragueDawley rats, cynomolgus monkeys, and humans exposed by inhalation to butadiene ( 6 , 7 ) . M-I1 is the mercapturic acid formed from conjugation of glutathione (GSH)with butadiene monoepoxide (BDO), while M-I may be a mercapturate formed from GSH conjugation with the hydrolysis product of the monoepoxide. These two metabolites constituted between 50% and 90% of the total urinary butadiene equivalents for mice, rats, and monkeys. When comparing species, the ratios of excreted M-I relative to the total of M-I M-I1 were qualitatively related to the hepatic epoxide hydrolase clearance rates, with mice displaying the lowest and humans the highest ratios. These results suggest that humans metabolize butadiene, or its metabolites, differently than mice, the
+
This article not subject to U.S.Copyright. Published 1995 by the American Chemical Society
Chem. Res. Toxicol., Vol. 8, No. 2, 1995 183
Volatile Butadiene Metabolites in Blood
rodent species known to be sensitive to butadiene's carcinogenic effects. Studies a t our Institute and elsewhere have explored species differences in the metabolism of P4C1BD to volatile metabolites in blood (8-10). These studies have indicated large differences between Sprague-Dawley rats and B6C3F1 mice. After comparable inhalation exposures, blood levels of volatile [l4C1 metabolites were measured in the two species by vacuum line distillation and I4C counting of blood distillates (8). Radioactivity tentatively assigned as the monoepoxide of BD was higher in the mouse than in the rat by 2-3-fold. Subsequently, methods were developed that allowed selective determination of specific BD metabolites in blood and tissue (11-13). These methods also incorporated the distillation of blood samples as a primary step in the isolation of volatile metabolites. However, prior to distilling the blood and tissue samples, deuterated internal standards corresponding to the anticipated metabolites were added. After the distillation process was complete, the contents of the distilled traps were analyzed by GC/MS. BDO was measured in the blood and tissues of B6C3F1 mice exposed by inhalation to 100 ppm butadiene for 4 h (11). However, greater sensitivity was needed to measure the highly toxic butadiene diepoxide (BD02) a t this exposure level in blood and tissues. One of the major limiting factors in achieving the desired sensitivity for the analysis of blood distillates is the volume of analyte that can be injected onto the GC (11). For practical reasons, no more than 1pL of solution can be injected; no improvement in sensitivity is achieved with larger volumes. However, after processing tissues for BD metabolites, about a 1-mL aqueous sample is generated (11);thus, only a small fraction (0.1%) of the sample can be analyzed. To improve the overall sensitivity of the GC/MS, a new technique was needed that would allow more of the solution to be injected. This paper reports the development of a sensitive, specific method for the analysis of volatile metabolites of butadiene in blood and tissues, and the results of application of the method to the analysis of blood from rats and mice exposed t o 100 ppm BD for 4 h. One goal was to develop a method with sufficient sensitivity to measure BD, BDO, and BDO2 in tissues of rats and mice after exposures to 100 ppm BD or less. The increase in sensitivity was achieved by two means: by extraction and concentration of the analyte from an organic solvent, and by the use of multidimensional GC/GC/MS. We describe the development of a rugged, versatile interface for column switching that is relatively easy to set up. The new technique allowed us to analyze blood samples from BD-exposed B6C3F1 mice and Sprague-Dawley rats for the genotoxic metabolites BDO and BD02. Only the blood from mice, the more sensitive species in BD carcinogenesis bioassays, was found to have detectable levels of BD02, the more genotoxic of the two metabolites.
Materials and Methods Reagents. BD (CAS #106-99-0),2BDO (CAS #930-22-3),and BDOz (CAS #298-18-0)were acquired from Aldrich (Milwaukee, WI) in the highest available purities. BD-ds (CAS #106-99-0) was acquired from Cambridge Isotope Laboratories (Woburn, MA). Deuterated analogues of BDO and BDOz were synthesized as previously described (14).The deuterated standards were Registry nos. were supplied by the author.
'
identified by their coelution with unlabeled standards, and by 111-scan mass spectra. The deuterated standards were partially purified by distilling the contents of the reaction mixture through sequential traps held at -45 and -196 "C. The BDOz condensed in the former trap while BDO condensed in the latter. Ethyl acetate was purchased from Burdick and Jackson (Muskegon, MI) and was greater than 99% pure. All water was Milli-& water. Reagent purities were determined prior to every validation or exposure. Purity of BD was determined by direct injection of a 1-pL aliquot of neat BD gas onto the GC/MS. The column was a Restek RTx-200(Bellefonte, PA), 30 m x 0.25 mm, with a 0.25 pm film thickness. Purity was assessed as the area of the BD peak divided by the sum of all integrated peaks. BDO and BDO2 purities were determined in a similar fashion. In addition, nonvolatileimpurities were determined by evaporating an aliquot of BDO or BDOz under a stream of nitrogen at room temperature and weighing the residues that remained. Overall, purities were greater than 98%, 93%, and 99% for BD, BDO, and BD02, respectively. Stock solutions of deuterated BDO and BDOz diluted in deionized water were used for standard curves. The concentrations of the internal standards were determined indirectly against primary standard solutions of unlabeled analytes. Stock solutions were separated into smaller aliquots and frozen in liquid nitrogen for subsequent use. Instrument Design. All analyses were conducted using GC/ GC/MS. There are five basic components t o the GC/GC/MS: a Hewlett Packard (HP) 5890 gas chromatograph, an HP 5970 mass selective detector (North Hollywood, CAI, an HP-UX datastation, a Nelson Analytical datastation with control and data storage devices (Cupertino, CAI, and a custom-built interface for column switching and cryogenic focusing. The interface is diagrammatically shown in Figure 1and is composed of pneumatically controlled Valco six- and four-port mechanical switchingvalves (Valco, Houston, Tx)housed in a separate oven. The transfer line between the two switching valves (Restek deactivated fused silica, 0.53 mm i d . ) passes through a cryogenic focuser (SGE glassware, Austin, TX). The first column used for isolation is an Alltech AT-Wax capillary column (Deerfield, IL), 15 m x 0.54 mm i d . , with a 2.5 pm film thickness. The second column is a Restek RT,-200 capillary column, 30 m x 0.25 mm i d . , with a 0.25 pm film thickness. The valve oven is maintained at an elevated temperature (230 "C) by the use of valve heaters. Temperatures in the oven are controlled by using thermocouples. All timed events for the switching valves and the cryofocuser are controlled by the Nelson Analytical datastation. The GC/GC interface operates as follows. An injection is made onto the wide-bore GC column with the relatively high flow rate of about 20 mumin. Initially, the column eluant passes out of the GC oven into the six-port valve and directly to a flame ionization detector (Figure 1A). Just prior t o the emergence of analyte from the column, the valve is switched, and the eluant is directed through and captured in the cryogenically cooled transfer line (cryofocused)(Figure 1B). Appropriate times for switching valves are determined by the injection of individual standards. Once analyte has been cryofocused onto the loop, the six-port valve is switched back to the original position (Figure 1A). This process is repeated until all analytes of interest have emerged from the first column. When all analytes have been cryofocused, the GC oven is cooled to ambient or subambient temperatures. Then, the fourport valve is switched, and the cryofocusor is turned off (Figure 1C). The captured analytes are rapidly desorbed from the cryofocusingloop and pass onto the second GC column. Because the second GC column has been cooled, the analytes cryogenically refocus at the head of the second column. Finally, the oven temperature of the GC column is increased for elution of the analytes. Flow through the second column is typically maintained at less than 1 mumin. General Study Design. The method was first validated by removing blood from control animals and adding analyte a t the
184 Chem. Res. Toxicol., Vol. 8, No. 2, 1995
Bechtold et al. GCIGCIMS
irogen Cooled CO,
I
I
._I__.__
I
I
flF l L
0 [1HJ Standards
u
-1
B6C3F1 Mics
Control Blood (Valldations)
or
1
El ~-
r
Vacuum Dirtiliation
--
-H a I t -5 mL)
% G S s
Butadlane Expoaura 100 ppm, 4 hr
I
U
Sacrifice
035 mm m
Withdraw Blood, Tissue
GUGUMS
Figure 2. Assay for measuring BD and metabolites in blood.
Llquld Nltrogrn Cooled GO2
I L
I
-
Inlector
L I W !J3S mm id
0 injector
Figure 1. Schematic diagram for a gas chromatography/gas chromatography/mass spectroscopy using a 2-valve,cryofocusing interface. The two GC columns and the two injectors are housed within the oven of a Hewlett Packard 5890 GC. The 2 valves are housed in a separate oven. (A) Inject/FID analysis position; (B) peak capture position; (C) GCMS analysis position. expected levels based on prior studies (8).Percent recoveries were determined as a result of these studies. Next, both mice and rats were exposed by inhalation in nose-only chambers to approximately 100 ppm BD for 4 h. At the end of the exposure, six exposed mice and rats, and six control mice and rats (exposed to filtered air in nose-only chambers), were anesthetized by an
ip injection of pentobarbital while breathing the exposure atmosphere, and blood was withdrawn by heart puncture, through a special port in the exposure tube, by a syringe containing heparin as an anticoagulant. The mice and rats were added to the exposure system in an incremental fashion to allow time for sacrifice and blood collection between animals. The time of exposure was defined as the time between when the nose-only exposure tube containing the animals was attached to the exposure system and the time that blood was withdrawn. Animals. Male Sprague-Dawley rats were purchased from Taconic (Germantown, NY).Male B6C3F1 mice were acquired from Charles River Laboratories (Kingston, NY,and Portage, MI). Animals were shipped in filtered shipping containers and quarantined for 2 weeks upon arrival before use. Animals were housed in polycarbonate cages (two rats or four mice per cage) containing hardwood chip bedding and filter caps. Rooms were maintained at 70 & 5'F with a relative humidity of 20-60% and a 12 h light-dark cycle with light starting at 0600. Food (Lab-Blox,Allied Mills, Chicago, IL) and water from bottles with sipper tubes were provided ad libitum, except during inhalation exposures. Animals were observed twice daily. All animals were exposed at 11-14 weeks of age. Exposure Atmospheres. BD atmospheres were generated directly from commercially available cylinders of BD. The cylinders were fitted with two-stage regulators and rotometers. The BD gas was diluted with filtered house air to achieve the desired concentration. BD concentrations in the exposure chambers were quantitated by a GC (GOW-MAC Series 750) equipped with a flame ionization detector. The GC was fitted with a gas-sampling valve (1 mL sample loop) and a Porapak 80/100-mesh column (6 ft x ' / 6 in.). The carrier gas was helium a t a flow rate of 30 mumin. The column was maintained isothermally a t 100 "C. Samples were taken either by syringe from a sample port on the exposure plenum or by a transfer line from the plenum directly to the gas-sampling valve. Samples were taken at least every 15 min throughout the exposure. Peak areas from the GC were quantitated using a Spectra Physics 4270 integrator. Peak areas from the integrator were compared to standard curves to determine the concentration of BD in the chambers. The BD concentration in the exposure chambers was also continuously monitored by IR spectroscopy (Miran). The wavelength was set a t 3.4 pm. Because of the level of noise, the IR signal was not used for quantitation, but rather as a measure of atmosphere stability. Method. The method for analysis of BD and metabolites in blood is schematically shown in Figure 2. Approximately 0.75 mL of blood from control (for validations) or exposed animals was added to 50-mL round-bottomed flasks. These flasks had been stoppered, evacuated, and tared immediately prior to use. An internal standard solution of BDO-ds and BDOz-ds was added, aliquots of diluted BD-& gas were added using a gas-
Chem. Res. Toxicol., Vol. 8, No. 2, 1995 185
Volatile Butadiene Metabolites in Blood Table 1. Assignment of Ions for Quantitation and Confirmation of BD and Metabolites deuterated internal analyte analyte ions standard ions 42,60 BD 39,54 46,74 BDO 39,41,69 30,58 BDOz 29,55 tight syringe, and the contents were well mixed and frozen on liquid nitrogen. The volatile BD metabolites were isolated from nonvolatile impurities by vacuum-line distillation. A 10-port manifold was Torr. evacuated with a two-stage vacuum pump to below Contents of the 50-mL round-bottomed flask were distilled through a 10-mL U-tube modified with a sampling port (septaport U-traps) as previously described (9). The septaport U-traps were held at -196 "C by liquid nitrogen. Blood samples in the round-bottomed flasks were distilled until visually dry. The contents of the traps were analyzed for BD, BDO, and BDOz using the GC/GC/MS apparatus described above. Preliminary experiments showed that BD and BDO partitioned significantly into the septaport U-trap headspace, while BDOz remained in the water that had codistilled. BD and BDO were measured from the headspace of the septaport U-traps held at room temperature. Headspace (2-10 mL) from the septaport U-trap was removed with a glass syringe and injected directly onto the first column. At the initial oven temperature (50 "C), little retention of BD or BDO occurred on the first column. Therefore, after injection of headspace, the first 2 min of eluant was directed to the cryofocuser (Figure 1B). After the two analytes had eluted, the six-port switching valve was rotated (Figure lC), thereby precluding the transfer of water vapor to the cryofocusor. The four-port valve was then rotated, the cryofocuser was turned off, and the analytes were transferred to the second column. The oven temperature was increased to 250 "C at 10 "C/min. The BD and BDO peaks eluted a t 10.0 and 12.9 min, respectively. The MS was operated in the selected ion mode. Ions that were monitored for BD, BDO, BD-ds, and BDO-ds are shown in Table 1. Peaks were integrated, and ion ratios were calculated to both internal standards and t o each of the other ions from the same analyte. The former set of ion ratios was used for quantitation, while the latter set was used for confirmation of analyte identity. Standard curves were created by adding graded amounts of analyte to septaport U-traps with a consistent amount of water and internal standard, followed by analysis as described above. The aqueous phase remaining in the septaport U-trap was removed for analysis of the BDOz. The aqueous phase was extracted three times with separate 2-mL aliquots of ethyl acetate, and the organic phases were pooled. The organic extract was transferred t o 2-mL autosampler vials and reduced in volume to between 50 and 100 pL by passing a stream of nitrogen gas over the sample at room temperature. The organic extract was analyzed for BDOz by GC/GC/MS using the same columns described above. Injections of up to 10 pL indicated that peak areas were linear with injection volume. For convenience, 5 pL of the extract was injected. Initially, the switching valves directed eluate to the flame ionization detector (Figure 1A). The initial oven temperature was 50 "C for 2.5 min, followed by an increase of 15 "C/min to 150 "C. The oven was held a t this temperature for 1min. BDOz was shown to have a retention time of 7.0 min under these conditions. Therefore, at 6.5 min, the six-port valve was rotated to direct column eluate to the prechilled cryofocuser (Figure 1B). This process effectively excluded any solvent. At 7.5 min, the six-port valve was rotated back t o the original position (Figure 1A). After completion of the oven temperature gradient, the oven was cooled to 25 "C and held for 5 min. The cryofocuser was turned off at 16 min, and the four-port valve was rotated at 16.5 min t o transfer the BDOz to the second column. The oven temperature was increased at 15 "C/min t o 150 "C and
Table 2. Method Validation for BD, BDO, and BDOz in Blood
analyte BD BDO BDOz =
spike levels (nmol) 4.96 3.42 0.545
analyte recovery (% f SD 95.6 & 17.7 124.6 f 14.5 9 8 f 12
internal analyte standard ion ratios: recoveries standards (% f SEM vs samples 1.30 vs 1.25= 41 f 9.0b 52 f 9Bd 7.63 vs 8.2W 2.24 vs 1.88e 62 f 15f
a Ion pair 39/54. n = 9; based on ion 42. Ion pair 39/69. n 9; based on ion 46. e Ion pair 29/55. f n = 6; based on ion 58.
held a t this temperature for 4.5 min. The BDOz peak eluted a t 23.4 min under these conditions. The ions monitored for BDOz and BDOZ-dsare shown in Table 1. Metabolite Stability. The stabilities of BDO and BDOz were determined in blood for short time periods (1 h) when held at room temperature, and for BDO, for extended time periods when blood was held a t -80 "C. To do so, a 25-mL volume of control blood was spiked with BDO and BDOz a t nominal concentrations of 5 and 0.5 nmoUmL, respectively. For shortterm stability studies, 750-pL aliquots were removed at specified times and placed in 50-mL round-bottomed flasks. Internal standard was added, and the mixture was frozen in liquid nitrogen. The mixture was distilled on the vacuum line, and the distillate was analyzed for BDO and BD02. For long-term stability studies, three, 750-pL aliquots were removed and placed into individual 50-mL round-bottom flasks. Internal standard solution was added, and samples were immediately frozen and stored in a -80 "C freezer. After a storage time of 6 weeks, the samples were removed from the freezer, distilled, and analyzed by GC/GC/MS. The stability of BDOz a t 6 weeks was not determined.
Results Measurement of BD and Metabolites in Blood. Results of the method validation experiments, which are shown in Table 2, allow assessment of both the accuracy and reproducibility of the method. In general, the method gave values within 25% of those expected. The error associated with the measurements was typically less than 20%. The consistency for internal ion ratios confirms the identity of the analytes as BD and metabolites. Overall, distillation efficiencies varied between about 40% and 60%, as measured by the internal standard recoveries. The limits of sensitivity for the assay were 100, 20, and 10 pmol/mL for BD, BDO, and BD02, respectively. Retention times for the potential BD metabolites crotonaldehyde and 3-butene-1,2-diol were 1.13 relative to BDO and 1.06 relative to BD02, respectively. Metabolite Stability. Knowledge of the short-term stability of BDO and BDOz in blood was important because a t least some time lag occurs (1-2 min) between when animals were sacrificed on the inhalation exposure plenum and when internal standard was added to the blood. Because of the potentially reactive nature of these epoxides, degradation could potentially occur in the interim and artificially lower the observed values. Knowledge of the long-term stability will be important for studies performed with human samples for which extended times can occur between collection and analysis of blood. Results for the short-term stability of BDO and BDOz are shown in Table 3. Over a time span of 1h, no degradation of the analyte occurred. After a storage time of 6 weeks, recovery of the BDO in the blood sample was 92 + 8% (n = 3, SD). Exposure of Mice and Rats to BD. The target exposure concentration was 100 ppm, while the actual
186 Chem. Res. Toxicol., Vol. 8, No. 2, 1995
Bechtold et al.
Table 3. Stability of BDO and BD02 in Blood Held at Room Temperature
time point (min) 0 2 4 8
analyte
time
point
%BDO" %BDOzb 123 117 124 122
(mid
%BDW
%BDOzb
BDO rats
16 32 62
120 125 97
94 108 109
mice BDO2
100 108 111 97
a Spike level for BDO was 4.96 nmol. Values are given relative to the means shown in Table 2 for BDO spiking experiments. Spike level for BDO2 was 0.545 nmol. Values are given relative to the means shown in Table 2 for BDO2 spiking experiments.
A
Table 5. Comparison of Blood Levels of BD and Volatile Metabolites
B
GC/GC/MS"
Himmelstein et aLb
*
0.1 f 0.06 0.38 f 0.14
0.07 0.01 0.56 f 0.04
rats
ND'
ND
mice
0.33 f 0.19
0.65 f 0.10
a Present study. Reference 16; exposures were for 6 h to 62.5 ppm, units are in nmol/mL. Not detected.
umn 4) confirms the identity of the respective analytes. Internal standard recoveries (column 5) were in general similar to those reported in the validations.
28000 24000
Discussion
40000
2OOOO
0 16000 f 9
t
30000
E9 -
1 m 8wo
lwoD UHH)
22.4
h.8
G.2
T h e (mln)
A.6
24.0
22.4
d.8 24.2 "e(mln)
A.6
2i.O
Figure 3. GC/GC/MS selected ion chromatogram (ion 55) from the blood of (A) control and (B) exposed mice. Mice were exposed by inhalation to nominal atmospheres of 100 ppm butadiene. After sacrifice by heart puncture, blood was withdrawn, spiked with internal standard, and processed as shown in Figure 2. Table 4. Summary of Exposure Results: Analysis of BD, BDO, and BDO2 in the Blood of Mice and Rats Exposed by Inhalation to Butadiene exposure analyte analyte ion internal levels blood level ratios: standard (ppm) (nmoUmL) standards vs recoveries analytea (mean f SD (mean f SD samples (B f SEM) BD rats 96.5 f 5.5 4.1 f 1.0 1.3 vs 1.3 28 f 2.7 mice 101 f 4 2.9 =k 1.3 1.2 vs 1.2b 36 f 4.8c BDO rats 96.5 f 5.5 0.10 f 0.06 7.4vs 7.2 32 f 2.6 mice 101 f 4 0.38 f 0.14 8.1 vs 7.5d 48 f 5.0" BDOz rats 93 f 13 NDh ND 75 f 3.3 mice 93 f 13 0.33 f 0.19 1.3 vs 1.g 68 f 48 a n = 6 for both rats and mice. Ion pair 39/54. n = 12; based on ion 42. Ion pair 39/69. e n = 12; based on ion 46. f Ion pair 29/55. n = 12; based on ion 58. ND, not detected.
measured values varied between 92 and l01,ppm. The stability of this generation was evaluated a t least every 15 min over the 4-h exposure period and typically gave a variation of less than f20% of the mean concentration. Blood was analyzed for chemical-specific metabolites as described above. New standard curves were generated (triplicate determinations with five points per curve) for each set of validations and after exposure of animals. Figure 3 shows selected ion chromatograms for BDO2 in the blood of control and exposed mice. Results from the exposure are shown in Table 4. In all cases in which values are reported for BD and metabolites in blood, measured values for exposed animals were significantly different than those for controls (Student's t-test, p < 0.05). The consistency between analyte ion ratios (col-
In this study, a method previously used to measure volatile epoxides in blood (11-13) was dramatically improved for greater sensitivity. Some of the increase in sensitivity was achieved for BDO2 by extracting the aqueous distillate with organic solvents and concentrating the organic phase. This approach has proven successful elsewhere for measuring BDO2 in microsomal metabolic systems (15).However, a significant increase in sensitivity for all three analytes was achieved by the adaptation of multidimensional GC techniques. To perform the column switching, we built an interface largely from commercially available components. The interface was relatively easy to construct, versatile, and rugged. Of primary importance, the interface allows for the facile transfer of analytes from a capillary column operating a t a high flow rate (20 mumin) to one operating at a low flow rate (1mumin). This low flow rate is necessary for the typical pumping capacities of bench-top MS. In contrast, the high flow rate is typical of large bore columns (0.53 mm i.d.1 that have greater injection capacities (10 pL). The dual-valved, multidimensional system therefore allows large volumes of solutions to be quantitatively injected without compromising chromatographic resolution or mass spectral sensitivity. As a result, by the use of the interface, a n improvement of about 2 orders of magnitude was achieved for measuring these metabolites relative to our previous method. After incorporating the extraction step, a n improvement of approximately 3 orders of magnitude was achieved for measuring BDO2 (limits of detection of 10 nmol/mL vs 10 pmol/mL). Several hundred injections of head space and organic extracts have been made without any failures of the interface. An added benefit is that the MS requires less maintenance, as little superfluous background materials reach the source. All three analytes were clearly evident in the blood distillates from exposed but not control mice. The levels of BDO and BDO2 measured in the blood of exposed mice compare well with recently reported values as determined by GC/MS (Table 5) (16). Although some BDO was detected in rat blood in both the previous and current studies, little or no BDO2 was detected in either study. Of interest is the comparison of blood levels of BDO2 between rats and mice. The limits of detection for our assay were about 10 pmol/mL. Therefore, for a given level of exposure and time of collection, mice had a t least a 30-fold excess of BDO2 relative to rats. Consistent with the conclusions of Himmelstein et al. (161, these results suggest that BDO2 may contribute to the marked species difference in the toxicity of BD. However, the results
Volatile Butadiene Metabolites in Blood must be confirmed by measuring levels of BDOz in target tissues such as lung aRer various exposure levels and over greater collection times. In summary, multidimensional GCIGCNS has proven to be a method capable of dramatically enhancing the sensitivity of our analysis of BD and metabolites in blood. After mice were exposed for 4 h to 100 ppm BD, the parent compound, BDO and BDOz concentrations were determined in blood. The values compared favorably with previously published estimates. BDOz was observed in mouse, but not rat blood samples. In upcoming experiments, we will determine levels of these reactive metabolites in the blood from humans occupationally exposed to BD.
Acknowledgment. The authors gratefully acknowledge contributions to this work from members of the staff a t the Inhalation Toxicology Research Institute. In particular, we thank Margo Allen and Louise Archuleta for their excellent technical assistance. This research was sponsored by the Chemical Manufacturers Association under Funds-in-Agreement DE-FIOY-91AL66351, and by the U.S.Department of Energy, Office of Health and Environmental Research, under Contract DE-AC0476EV01013, in facilities fully accredited by the American Association for the Accreditation of Laboratory Animal Care.
References (1) Chem. Eng. News (1992) Chemical production held steady in 1991. Chem. Eng. News 70, 35. (2) Neligan, R. E. (1962) Hydrocarbons in the Los Angeles atmosphere. Arch. Environ. Health. 5, 67-77. (3) Brunnemann, K. D., Kagan, M. R., Cox, J. E., and Hoffman, D. (1990) Analysis of 1,3-butadiene and other selected gas-phase components in cigarette mainstream and sidestream smoke by gas chromatography-mass selective detection. Carcinogenesis 11, 1863-1868. (4) Hazelton Laboratories Europe, Ltd. (1981) The toxicity and carcinogenicity of butadiene gas administered to rats by inhalation for approximately 24 months. Prepared for the International
Chem. Res. Toxicol., Vol. 8, No. 2, 1995 187 Institute of Synthetic Rubber Producers, New York, NY. (5) Melnick, R. L., Shackelford, C. C., and Huff, J. (1993) Carcino-
genicity of 1,3-butadiene. Environ. Health Perspect. 100, 227236. ( 6 ) Sabourin, P. J.,Burka, L. T., Bechtold, W. E., Dahl, A. R., Hoover, M. D., Chang, I.-Y., and Henderson, R. F. (1992) Species differences in urinary butadiene metabolites; identification of 1,2dihydroxy-4-(N-acetylcysteinyl)butane, a novel metabolite of butadiene. Carcinogenesis 13, 1633-1638. (7) Bechtold, W. E., Strunk, M. R., Chang, I.-Y., Ward, J. B., and Henderson, R. F. (1994) Species differences in urinary butadiene metabolites: comparisons of metabolite ratios between mice, rats, and humans. Toxicol. Appl. Pharmacol. 127, 44-49. (8) Bond, J. A., Dahl, A. R., Henderson, R. F., Dutcher, J. S., Mauderly, J. L., and Birnbaum, L. S. (1986) Species differences in the disposition of inhaled butadiene. Toxicol.App1. Pharmacol. 84,617-627. (9) Schmidt, U., and Loeser, E. (1985) Species difference in the formation of butadiene monoxide from 1,3-butadiene and its reactive metabolites. Arch. Toxicol. 57, 222-225. (10) Kreiling, R., Laib, R. J., Filser, J. G., and Bolt, H. M. (1986) Species differences in butadiene metabolism between mice and rats evaluated by inhalation pharmacokinetics. Arch. Toxicol. 58, 235-238. (11) Maples, K. R., Bechtold, W. E., Dahl, A. R., and Henderson, R. F. (1993) Butadiene monoepoxide levels are greater in bone marrow than in blood following inhalation of 1,3-butadiene in B6C3F1 mice. Toxicologist 13, 335. (12) Maples, K. R., and Dahl, A. R. (1991) Blood levels of propylene oxide during propylene inhalation and effect on hepatic and nasal cytochrome P-450 concentrations. Drug Metab. Dispos. 19,835837. (13) Maples, K. R., and Dahl, A. R. (1993) Levels of epoxides in blood during inhalation of alkenes and alkene oxides. Inhalation Toxicol. 5, 43-54. (14) Maples, K. R., Lane, J. L., and Dahl, A. R. (1992) An efficient method for synthesizing (d(6))-butadienemonoepoxide.J.Labelled Compd. Radiopharm. 31,469-475. (15) Cheng, X. Q., and Ruth, J. A. (1993) A simplified methodology for quantitation of butadiene metabolites-Application to the study of 1,3-butadiene metabolism by rat liver microsomes. Drug Metab. Dispos. 21, 121-124. (16) Himmelstein, M. W., Turner, M. J., Asgharian, B., and Bond, J. A. (1994) Comparison of blood concentrations of 1,3-butadiene and butadiene epoxides in mice and rats exposed to 1,3-butadiene by inhalation. Carcinogenesis 15, 1479-1486.
TX940024H
