ANAL,YTLCAL
BIOCHEMISTRY
186,101-107
(1990)
Quantitation of 1,3-Butanediol by Gas Chromatography-Mass
and Its Acidic Metabolites Spectrometry
Sylvain Desrochers,* Jane A. Montgomery,? Christine Beth C. Lincoln,? and Henri Brunengraber*-tsl Departments
Received
October
of *Biochemistry
and TNutrition,
University
Des Rosiers,?
of Montreal,
Montreal,
Quebec, Canada H3C 357
4,1989
A number of problems present themselves during the gas chromatographic-mass spectrometric assay of R,S1,3-butanediol as its bis-tert-butyldimethylsilyl ether. To circumvent these problems, three labeled internal standards were synthesized: (i) R,S-1,3-[3,4-‘3C2]butanediol, (ii) R,S- 1,3-[ 1,l ,3-‘H3]butanediol, and (iii) R,S-1,3-[1,1,3-2H3,3,4-‘3C2]butanediol. Theavailability of internal standards with different degrees of labeling allows (i) assaying of either unlabeled or r3Clabeled R,S-1,3-butanediol and (ii) analysis of 1,3-butanediol in either blood or urine samples. Reproducible standard curves were obtained using both electron impact and ammonia chemical ionization modes. The latter provides greater sensitivity and a lower limit of detection (5 PM). We have also designed an indirect assay of S-3-hydroxybutyrate, a catabolite of R,S-1,3-butanediol, which is difficult to analyze by conventional methods. This assay relies on the difference between (i) the concentration of R,S-3-hydroxybutyrate assayed by gas chromatography-mass spectrometry and (ii) the concentration of R-3-hydroxybutyrate assayed enzymatically. B 1990 Academic Press, Inc.
R,S-1,3-Butanediol (BD)‘z3 is a potential nutrient for oral and parenteral nutrition (3). It is an industrial solvent, a fungicide, and a potential food preservative. i To whom correspondence should be addressed at Research Center, Notre Dame Hospital, 1560 Sherbrooke St. East, Room M-5210, Montreal, Quebec, Canada H2L 4Ml. ’ Abbreviations used: BD, R,S-1,3-butanediol; [‘%,]BD, R,S-1,3[3,4-i3Cz]butanediol; [‘HJBD, R,S-1,3-[1,1,3-*H,]butanediol; [‘Ha, i3Cz]BD, R,S-1,3-[1,1,3-*H3, 3,4-WJbutanediol; AcAc, acetoacetate; BHB, 3-hydroxybutyrate; CI, chemical ionization; EI, electron impact; GC-MS, gas chromatograph-mass spectrometer; THF, tetrahydrofuran; TBDMS, tert-butyldimethylsilyl. 3 Except where indicated, the word 1,3-butanediol (BD) refers to the racemic RS mixture.
0003.2697/90 $3.00 Copyright Q 1990 by Academic Press, All rights of reproduction in any form
BD has also been proposed as a therapeutic agent for the ethanol withdrawal syndrome (4). Lastly, it was reported (5) that the S-, but not the R-isomer, normalizes blood glucose in streptozotocin-diabetic rats, when added to the diet. Since BD is a substrate for equine liver alcohol dehydrogenase (6), it can be assayed with this enzyme by monitoring the reduction of NADf to NADH (7). However, this assay suffers from a large and variable drift of NADH absorbance. We found this drift manageable with standards, but not with biological samples, since the extent of drift could not be predicted from the blanks or the standards. BD has also been assayed by gas chromatography (8) using an external standard of 1,2-propanediol. Because of its high limit of detection (0.5 mM), this assay is not suitable for metabolic or environmental studies. In addition, it is not desirable to use 1,2-propanediol as an external standard because of the ubiquitous nature of this compound. As part of a study of the metabolism of BD and R,S1,3-[3,4-‘3C2]butanediol in isolated liver and in live animals, we set up an isotope dilution assay of this compound by gas chromatography-mass spectrometry (GC-MS). Briefly, BD is derivatized to its bis-tert-butyldimethylsilyl (TBDMS) ether. Quantitation is achieved with the use of one of three internal standards which we have synthesized. The two enantiomers of BD are metabolized in the liver to the corresponding R- and S-3-hydroxybutyrates (BHB)4 via alcohol and aldehyde dehydrogenases. The urinary excretion of physiological ketone bodies R-BHB and acetoacetate (AcAc), assayed enzymatically using R-BHB dehydrogenase, has been used to monitor in4 Note that the physiological D-BHB in the recent literature. Merck Index, it is called L-BHB.
enantiomer R-BHB is designated as In early publications (12) and in the S-BHB is the unnatural enantiomer.
101 Inc. reserved.
102
DESROCHERS
dustrial exposure to BD (9). The unnatural enantiomer S-BHB, derived from the S component of R,S-1,3-butanediol, is not detected by the enzymatic assay. We attempted to assay S-BHB using L-3-hydroxyacid dehydrogenase isolated from pig kidney (10). However, this assay suffers from major drifts, generated in part by the contamination of the enzyme with lactate dehydrogenase. We therefore designed an assay for S-BHB which uses the difference between (i) the GC-MS assay of R,SBHB and (ii) the enzymatic assay of R-BHB. MATERIALS
AND
METHODS
Materials Ethyl-AcAc, sulfosalicylic acid, tetrabydrofuran (THF), sodium sulfate, and diethyl ether were bought from Fisher Scientific (Montreal, Quebec). LiA1H4 and R,S1,3-butanediol were obtained from Aldrich Chemical Co. (Milwaukee, WI). Celite (No. 501) was obtained from Johns Manville (Denver, CO). Ion-exchange resin AG501-X8 was obtained from Bio-Rad (Richmond, CA). Ethyl-[3,4-‘3C2]AcAc (99 at.% 13C)and lithium aluminium deuteride (LiAl’H,; 98 at.% 2H) were obtained from MSD Isotopes (Dorval, Quebec). The derivatizing agent, N-methyl-N-(t-butyldimethylsilyl)-trifluoroacetamide was obtained from Regis Chemical Co. (Morton Grove, IL). Anhydrous ammonia gas for chemical ionization (CI; 99.99% minimum purity) was obtained from Matheson Gas Products Canada (Montreal, Quebec). R-3-Hydroxybutyrate dehydrogenase and coenzymes were purchased from Boehringer-Mannheim Canada, Ltd. (Dorval, Quebec). Synthesis of Labeled Internal Standards of R,S-1,3-Butanediol R,S-1,3-[3,4-‘3C2]Butanediol ([13C2]BD) is prepared by reduction of ethyl-[3,4-‘3C2]AcAc with LiAIHI. A suspension of 18 mmol of LiAlH, in 150 ml of freshly distilled THF is introduced into a three-neck reaction vessel, equipped with a reflux condenser. The vessel is purged with anhydrous nitrogen. After slow addition of 11.7 mmol ethyl-[3,4-13C2]AcAc dissolved in 35 ml THF, the suspension is magnetically stirred and refluxed for 6 h in the dark. Excess LiA1H4 is destroyed by sequential additions of 0.35 ml HzO, 0.35 ml 15% NaOH, and 1.0 ml H20 (11). After another 45 min of stirring, the suspension is vacuum filtered on celite. The precipitate of LiA1H4 is rinsed and extracted for 30 min with 25 ml of THF. Combined filtrates are evaporated under nitrogen, and the residue is dissolved in distilled water. The resulting solution is deionized on an AG-501-X8 column and filtered on a 0.22~pm nylon membrane. Since the reduction of ethyl-[3,4-13C,]AcAc is conducted in THF, one must avoid peroxidation of THF
ET
AL.
which would lead to 1,4-butanediol after reduction. Production of toxic 1,4-butanediol would render [13C2]BD unsuitable for use as a labeled substrate in vivo. Therefore it is essential that this synthesis be conducted using THF that has been freshly distilled under nitrogen from a suspension of LiAlH,. Mass spectrometric assay of [i3C2]BD as its bis-TBDMS derivative did not detect any ethyl-[3,4-13C2]AcAc or 1,4-butanediol. Internal standards of R,S-1,3-[1,1,3-2H,]butanediol ( [2H3]BD) and R,S-1,3-[1,1,3-2H3,3,4-13Cz]butanediol (L2H3, 13Cz]BD) were similarly synthesized by LiA12H4 reduction of ethyl-AcAc and ethyl-[3,4-13C,]AcAc, respectively. Yields of synthesis of the three internal standards were 75 to 85%. Isotopomer composition of the internal standards, assayed in the CI mode, were for [13C2]BD, 97.3% M + 2, 2.10% M + 1, 0.28% M; for [2H3]BD, 95.7% M + 3, 3.81% M + 2, 0.35% M + 1, 0.19% M; for [2H3, 13C2]BD, 93.9% M + 5,5.07% M + 4, 0.94% M + 3, 0.13% M + 2, 0.02% M + 1, 0.01% M. Solutions of standards are kept at -80°C until use. Preparation of Standards and Samples Working standards containing 10 nmol to 5 pmol of BD with 3.70 pmol [13C2]BD, 2.97 pmol [2H3]BD, or 1.95 lrmol i2H3, 13C2]BD are prepared in 1 ml of distilled water. Standards are saturated with NaCl, acidified with 50 ~1 of saturated sulfosalicylic acid, and extracted three times with 5 ml diethyl ether. The ether extracts are combined, dried with anhydrous Na2S04, and evaporated under nitrogen down to about 10 ~1. Evaporation must be stopped before residue is completely dry to avoid major loss of BD. The yield of BD isolation is approximately 60%. N-Methyl-N-(t-butyldimethylsilyl)-trifluoroacetamide (50 ~1) is added to each residue. After overnight incubation, 1 ~1 is injected into the gas chromatograph. Samples are analyzed in duplicate. Blood and urine samples (1 ml) are spiked with 0.5 pmol of an appropriate internal standard of BD (seeDiscussion). Also, 1 pmol of R,S-3-hydroxy-[2,2,3,4,4,42Hs]butyrate, an internal standard, is added to allow quantitation of R,S-3-hydroxybutyrate (12). Samples are deproteinized with sulfosalicylic acid (50 pi/ml) and centrifuged at 3000g for 20 min. The supernatant is saturated with NaCl and processed as above. A second set of blood samples (1 ml) is collected into perchloric acid (final concentration 3%). R-BHB and AcAc are assayed enzymatically in neutralized extracts (13). GC-MS Analysis of R,S-1,3-Butanediol and R,S-3-Hydroxybutyrate Analysis of the TBDMS derivatives of R,S-BD was performed on an HP 5988A GC-MS (Hewlett-Packard Canada, Pointe-Claire, Quebec). The gas chromatograph is equipped with a HP-5 fused silica column (25
GAS
CHROMATOGRAPHY-MASS
SPECTROMETRY
m X 0.2 mm, 0.33 pm film thickness, Hewlett-Packard). Operating conditions included: helium flow rate, 0.7 ml/ min; split ratio, 1:lO; injection port temperature, 270°C. The temperature of the column was programmed for 5 min isothermal at 15O”C, increased by 5”C/min for 5 min. Bis-TBDMS-BD elutes 6.8 min after sample injection followed by the bis-TBDMS derivative of BHB at 8.6 min. The column was baked at 250°C for 3 min between sample injections. The mass spectrometer parameters in the positive ion electron impact mode included: ion source temperature, 200°C; transfer line temperature, 265”C, emission current, 300 yA; electron energy, 70 eV. The analyte and internal standard of BD are analyzed by selected ion monitoring of the [M-57]+ ions (m/z 261 and 263, 264, or 266 corresponding to BD and to the different isotopomers synthesized). The [M-57]+ fragment is formed by the loss of a t-butyl group from the TBDMS radical of the bis-TBDMS derivative of BD. R,S-BHB is quantitated by monitoring at m/z 159 and 163, as described previously (12). Mass spectrometer conditions in the positive chemical ionization mode were modified as follows: ammonia pressure, 1 to 2 X 10m4 Torr; electron energy, loo-250 eV. Samples are analyzed by selected ion monitoring of the quasi-molecular [M + H]+ ions (m/z 319 and 321, 322, or 324 as described above). Standards and samples are analyzed in duplicate. Areas under fragmentograms are determined by interactive computer integration and are corrected for naturally occurring heavy isotopes and light isotopic impurities as described previously (12). In Vivo Experiment
ASSAY
OF
103
1.3.BUTANEDIOL
v CH,-OH-CH*--CH,-O-TBDMS
169 /
I
261 /
loo6 s
60-
>r rg 5
60-
-c p ‘G 0 z
TBDMS b *CH,=~“-C”,CH,O-TBDMS
40219 \/
20-
235
263 /
234
264 /
TBDMS s x Y 2 s c ? ‘SI)
1,
80-
D
CH,-LCH,-LM-BDMS d
60-
d
40-
z
A dog (24 h starved, 18 kg) was anesthetized by an intravenous injection of thiopental (6 mg/kg) and succinylcholine (1 mg/kg). After tracheal intubation, the animal was ventilated with a mixture of 30% O2 and 0.8% halothane in nitrogen. A solution of saline was infused through a femoral vein catheter. Blood samples were taken from a femoral artery catheter on the opposite leg. Urine was collected through a bladder catheter. Starting at t = 0, a 1.0 M solution of BD was administered intravenously over 2.25 min (2.5 mmol/kg). Blood samples (1 ml) were taken from -30 to 180 min at various time intervals.
233
219 \/
221 \I
1001
,
D
TBDMS I
60s
$
;ii
,,,,,_,
II
1
,,
1
;2~,,36,
f3;
RESULTS 100
Figure 1 shows the electron impact (El) mass spectra of bis-TBDMS-BD and of its three isotopomers. The heaviest ion detected in the spectrum of the unlabeled species (Fig. lA), at m/z 261, is formed by the loss of a tbutyl radical from the molecular ion. This spectrum also
150
200
250
m/z FIG. 1. Electron impact mass spectra silyl derivatives of (A) R,S-1,3-butanediol, tanediol, (C) R,S-1,3-[1,1,3-‘HJbutanediol, ‘Ha, 3,4-‘3C2]butanediol.
of the bis-tert-butyldimethyl(B) R,S-1,3-[3,4-‘3C,Jbuand (D) R,S-1,3-[1,1,3-
104
DESROCHERS
includes peaks at m/z 233 and 219 which correspond to rearrangement ions formed from the [M-57]+ ion. The m/z 233 fragment is formed by the loss of ethylene (C-l and C-2 of BD) and the transfer of 0-TBDMS to C-3. The resulting ion contains C-3 and C-4 of the intact molecule. Confirmation of this rearrangement is provided by the spectra of each isotopomer which show a fragment incorporating the labeling on these two terminal carbons. The m/z 219 fragment is formed by the loss of propylene (C-2, C-3, and C-4 of BD) and transfer of O-TBDMS to C-l. The masses of the corresponding ions in the spectra of the isotopomers are also consistent with this loss. Figure 2 shows the CI mass spectra of the same compounds. The base peak for unlabeled BD is at m/z 319, corresponding to the [M + HI+ ion. There is also a minor peak at m/z 261, corresponding to the [M-57]+ EI fragment described above. The CI mass spectra of the three isotopomers produce intense [M + HI+ ions and reflect the extent of label incorporation. Figures 3A and 3B show four standard curves of unlabeled BD under EI and CI conditions, using [‘H3]BD and [‘H3, 13Cz]BD as internal standards. Note that the two panels of Fig. 3 cover different concentration ranges. The four curves are linear in the concentration range of 0.1 to 5 mM under EI conditions, and in the concentration range of 0.01 to 5 mM under CI conditions (r = 0.998 for all four curves). The slopes of the two EI curves are virtually identical ([2H3]BD 0.343 f 0.004 vs [2H3, 13C,]BD 0.337 + 0.003). This is also the case for the two CI curves ( t2H3]BD 0.397 + 0.002 vs [2H3, 13Cz]BD 0.413 + 0.003). However, for each internal standard, the slopes of the EI and CI curves differ by 15%. In contrast, the standard curves obtained using the [13C,]BD internal standard (not shown) are identical in the EI and CI modes (slopes = 0.340 t 0.001 and 0.340 + 0.003, r = 0.999) and are identical to the EI standard curves obtained using the [2H3]BD and [‘H3, 13C,]BD internal standards (Fig. 3). Figure 4 shows the uptake of BD by an anesthetized dog after bolus intravenous injection. Identical BD concentrations were obtained by analyzing the extracts of whole blood under EI and CI conditions. The intraassay coefficients of variation for duplicate assays of BD in blood samples were 4.8 and 2.3% under EI and CI conditions, respectively. Following bolus injection, BD concentration decreased rapidly showing that this substrate is well metabolized. The semilogarithmic plot shows an apparent biexponential decline of BD concentration. The linear portion of the graph corresponds to metabolism of BD. Extrapolation of this linear portion to t = 0 yields an “initial concentration” of 3.47 mM which corresponds to a distribution volume of 72% of body weight. Figure 5 shows other data from the same in vivo experiment, in particular the accumulation of physiological ketone bodies R-BHB and AcAc (the third physiological
ET
AL.
100,
A s
>r .% r 3 -r p J 0 B
31;
BO-
TBDMS d
60-
CH,-dH-CH,-CH,-0-TBDMS
40-
20-
07
‘.
L,
.
261 / .’ I
.
J
L
loo-
B
32;
60Iit x .$
TBDMS I *y *CH,-CH-CH,-CH1-O-TBDMS
60-
5 E $! ‘3 -0
40-
2
20.
0-t.
!
.I.
I
263 / .‘I
.
1. . (
loo-
C
32
BIOCHEMISTRY
186,101-107
(1990)
Quantitation of 1,3-Butanediol by Gas Chromatography-Mass
and Its Acidic Metabolites Spectrometry
Sylvain Desrochers,* Jane A. Montgomery,? Christine Beth C. Lincoln,? and Henri Brunengraber*-tsl Departments
Received
October
of *Biochemistry
and TNutrition,
University
Des Rosiers,?
of Montreal,
Montreal,
Quebec, Canada H3C 357
4,1989
A number of problems present themselves during the gas chromatographic-mass spectrometric assay of R,S1,3-butanediol as its bis-tert-butyldimethylsilyl ether. To circumvent these problems, three labeled internal standards were synthesized: (i) R,S-1,3-[3,4-‘3C2]butanediol, (ii) R,S- 1,3-[ 1,l ,3-‘H3]butanediol, and (iii) R,S-1,3-[1,1,3-2H3,3,4-‘3C2]butanediol. Theavailability of internal standards with different degrees of labeling allows (i) assaying of either unlabeled or r3Clabeled R,S-1,3-butanediol and (ii) analysis of 1,3-butanediol in either blood or urine samples. Reproducible standard curves were obtained using both electron impact and ammonia chemical ionization modes. The latter provides greater sensitivity and a lower limit of detection (5 PM). We have also designed an indirect assay of S-3-hydroxybutyrate, a catabolite of R,S-1,3-butanediol, which is difficult to analyze by conventional methods. This assay relies on the difference between (i) the concentration of R,S-3-hydroxybutyrate assayed by gas chromatography-mass spectrometry and (ii) the concentration of R-3-hydroxybutyrate assayed enzymatically. B 1990 Academic Press, Inc.
R,S-1,3-Butanediol (BD)‘z3 is a potential nutrient for oral and parenteral nutrition (3). It is an industrial solvent, a fungicide, and a potential food preservative. i To whom correspondence should be addressed at Research Center, Notre Dame Hospital, 1560 Sherbrooke St. East, Room M-5210, Montreal, Quebec, Canada H2L 4Ml. ’ Abbreviations used: BD, R,S-1,3-butanediol; [‘%,]BD, R,S-1,3[3,4-i3Cz]butanediol; [‘HJBD, R,S-1,3-[1,1,3-*H,]butanediol; [‘Ha, i3Cz]BD, R,S-1,3-[1,1,3-*H3, 3,4-WJbutanediol; AcAc, acetoacetate; BHB, 3-hydroxybutyrate; CI, chemical ionization; EI, electron impact; GC-MS, gas chromatograph-mass spectrometer; THF, tetrahydrofuran; TBDMS, tert-butyldimethylsilyl. 3 Except where indicated, the word 1,3-butanediol (BD) refers to the racemic RS mixture.
0003.2697/90 $3.00 Copyright Q 1990 by Academic Press, All rights of reproduction in any form
BD has also been proposed as a therapeutic agent for the ethanol withdrawal syndrome (4). Lastly, it was reported (5) that the S-, but not the R-isomer, normalizes blood glucose in streptozotocin-diabetic rats, when added to the diet. Since BD is a substrate for equine liver alcohol dehydrogenase (6), it can be assayed with this enzyme by monitoring the reduction of NADf to NADH (7). However, this assay suffers from a large and variable drift of NADH absorbance. We found this drift manageable with standards, but not with biological samples, since the extent of drift could not be predicted from the blanks or the standards. BD has also been assayed by gas chromatography (8) using an external standard of 1,2-propanediol. Because of its high limit of detection (0.5 mM), this assay is not suitable for metabolic or environmental studies. In addition, it is not desirable to use 1,2-propanediol as an external standard because of the ubiquitous nature of this compound. As part of a study of the metabolism of BD and R,S1,3-[3,4-‘3C2]butanediol in isolated liver and in live animals, we set up an isotope dilution assay of this compound by gas chromatography-mass spectrometry (GC-MS). Briefly, BD is derivatized to its bis-tert-butyldimethylsilyl (TBDMS) ether. Quantitation is achieved with the use of one of three internal standards which we have synthesized. The two enantiomers of BD are metabolized in the liver to the corresponding R- and S-3-hydroxybutyrates (BHB)4 via alcohol and aldehyde dehydrogenases. The urinary excretion of physiological ketone bodies R-BHB and acetoacetate (AcAc), assayed enzymatically using R-BHB dehydrogenase, has been used to monitor in4 Note that the physiological D-BHB in the recent literature. Merck Index, it is called L-BHB.
enantiomer R-BHB is designated as In early publications (12) and in the S-BHB is the unnatural enantiomer.
101 Inc. reserved.
102
DESROCHERS
dustrial exposure to BD (9). The unnatural enantiomer S-BHB, derived from the S component of R,S-1,3-butanediol, is not detected by the enzymatic assay. We attempted to assay S-BHB using L-3-hydroxyacid dehydrogenase isolated from pig kidney (10). However, this assay suffers from major drifts, generated in part by the contamination of the enzyme with lactate dehydrogenase. We therefore designed an assay for S-BHB which uses the difference between (i) the GC-MS assay of R,SBHB and (ii) the enzymatic assay of R-BHB. MATERIALS
AND
METHODS
Materials Ethyl-AcAc, sulfosalicylic acid, tetrabydrofuran (THF), sodium sulfate, and diethyl ether were bought from Fisher Scientific (Montreal, Quebec). LiA1H4 and R,S1,3-butanediol were obtained from Aldrich Chemical Co. (Milwaukee, WI). Celite (No. 501) was obtained from Johns Manville (Denver, CO). Ion-exchange resin AG501-X8 was obtained from Bio-Rad (Richmond, CA). Ethyl-[3,4-‘3C2]AcAc (99 at.% 13C)and lithium aluminium deuteride (LiAl’H,; 98 at.% 2H) were obtained from MSD Isotopes (Dorval, Quebec). The derivatizing agent, N-methyl-N-(t-butyldimethylsilyl)-trifluoroacetamide was obtained from Regis Chemical Co. (Morton Grove, IL). Anhydrous ammonia gas for chemical ionization (CI; 99.99% minimum purity) was obtained from Matheson Gas Products Canada (Montreal, Quebec). R-3-Hydroxybutyrate dehydrogenase and coenzymes were purchased from Boehringer-Mannheim Canada, Ltd. (Dorval, Quebec). Synthesis of Labeled Internal Standards of R,S-1,3-Butanediol R,S-1,3-[3,4-‘3C2]Butanediol ([13C2]BD) is prepared by reduction of ethyl-[3,4-‘3C2]AcAc with LiAIHI. A suspension of 18 mmol of LiAlH, in 150 ml of freshly distilled THF is introduced into a three-neck reaction vessel, equipped with a reflux condenser. The vessel is purged with anhydrous nitrogen. After slow addition of 11.7 mmol ethyl-[3,4-13C2]AcAc dissolved in 35 ml THF, the suspension is magnetically stirred and refluxed for 6 h in the dark. Excess LiA1H4 is destroyed by sequential additions of 0.35 ml HzO, 0.35 ml 15% NaOH, and 1.0 ml H20 (11). After another 45 min of stirring, the suspension is vacuum filtered on celite. The precipitate of LiA1H4 is rinsed and extracted for 30 min with 25 ml of THF. Combined filtrates are evaporated under nitrogen, and the residue is dissolved in distilled water. The resulting solution is deionized on an AG-501-X8 column and filtered on a 0.22~pm nylon membrane. Since the reduction of ethyl-[3,4-13C,]AcAc is conducted in THF, one must avoid peroxidation of THF
ET
AL.
which would lead to 1,4-butanediol after reduction. Production of toxic 1,4-butanediol would render [13C2]BD unsuitable for use as a labeled substrate in vivo. Therefore it is essential that this synthesis be conducted using THF that has been freshly distilled under nitrogen from a suspension of LiAlH,. Mass spectrometric assay of [i3C2]BD as its bis-TBDMS derivative did not detect any ethyl-[3,4-13C2]AcAc or 1,4-butanediol. Internal standards of R,S-1,3-[1,1,3-2H,]butanediol ( [2H3]BD) and R,S-1,3-[1,1,3-2H3,3,4-13Cz]butanediol (L2H3, 13Cz]BD) were similarly synthesized by LiA12H4 reduction of ethyl-AcAc and ethyl-[3,4-13C,]AcAc, respectively. Yields of synthesis of the three internal standards were 75 to 85%. Isotopomer composition of the internal standards, assayed in the CI mode, were for [13C2]BD, 97.3% M + 2, 2.10% M + 1, 0.28% M; for [2H3]BD, 95.7% M + 3, 3.81% M + 2, 0.35% M + 1, 0.19% M; for [2H3, 13C2]BD, 93.9% M + 5,5.07% M + 4, 0.94% M + 3, 0.13% M + 2, 0.02% M + 1, 0.01% M. Solutions of standards are kept at -80°C until use. Preparation of Standards and Samples Working standards containing 10 nmol to 5 pmol of BD with 3.70 pmol [13C2]BD, 2.97 pmol [2H3]BD, or 1.95 lrmol i2H3, 13C2]BD are prepared in 1 ml of distilled water. Standards are saturated with NaCl, acidified with 50 ~1 of saturated sulfosalicylic acid, and extracted three times with 5 ml diethyl ether. The ether extracts are combined, dried with anhydrous Na2S04, and evaporated under nitrogen down to about 10 ~1. Evaporation must be stopped before residue is completely dry to avoid major loss of BD. The yield of BD isolation is approximately 60%. N-Methyl-N-(t-butyldimethylsilyl)-trifluoroacetamide (50 ~1) is added to each residue. After overnight incubation, 1 ~1 is injected into the gas chromatograph. Samples are analyzed in duplicate. Blood and urine samples (1 ml) are spiked with 0.5 pmol of an appropriate internal standard of BD (seeDiscussion). Also, 1 pmol of R,S-3-hydroxy-[2,2,3,4,4,42Hs]butyrate, an internal standard, is added to allow quantitation of R,S-3-hydroxybutyrate (12). Samples are deproteinized with sulfosalicylic acid (50 pi/ml) and centrifuged at 3000g for 20 min. The supernatant is saturated with NaCl and processed as above. A second set of blood samples (1 ml) is collected into perchloric acid (final concentration 3%). R-BHB and AcAc are assayed enzymatically in neutralized extracts (13). GC-MS Analysis of R,S-1,3-Butanediol and R,S-3-Hydroxybutyrate Analysis of the TBDMS derivatives of R,S-BD was performed on an HP 5988A GC-MS (Hewlett-Packard Canada, Pointe-Claire, Quebec). The gas chromatograph is equipped with a HP-5 fused silica column (25
GAS
CHROMATOGRAPHY-MASS
SPECTROMETRY
m X 0.2 mm, 0.33 pm film thickness, Hewlett-Packard). Operating conditions included: helium flow rate, 0.7 ml/ min; split ratio, 1:lO; injection port temperature, 270°C. The temperature of the column was programmed for 5 min isothermal at 15O”C, increased by 5”C/min for 5 min. Bis-TBDMS-BD elutes 6.8 min after sample injection followed by the bis-TBDMS derivative of BHB at 8.6 min. The column was baked at 250°C for 3 min between sample injections. The mass spectrometer parameters in the positive ion electron impact mode included: ion source temperature, 200°C; transfer line temperature, 265”C, emission current, 300 yA; electron energy, 70 eV. The analyte and internal standard of BD are analyzed by selected ion monitoring of the [M-57]+ ions (m/z 261 and 263, 264, or 266 corresponding to BD and to the different isotopomers synthesized). The [M-57]+ fragment is formed by the loss of a t-butyl group from the TBDMS radical of the bis-TBDMS derivative of BD. R,S-BHB is quantitated by monitoring at m/z 159 and 163, as described previously (12). Mass spectrometer conditions in the positive chemical ionization mode were modified as follows: ammonia pressure, 1 to 2 X 10m4 Torr; electron energy, loo-250 eV. Samples are analyzed by selected ion monitoring of the quasi-molecular [M + H]+ ions (m/z 319 and 321, 322, or 324 as described above). Standards and samples are analyzed in duplicate. Areas under fragmentograms are determined by interactive computer integration and are corrected for naturally occurring heavy isotopes and light isotopic impurities as described previously (12). In Vivo Experiment
ASSAY
OF
103
1.3.BUTANEDIOL
v CH,-OH-CH*--CH,-O-TBDMS
169 /
I
261 /
loo6 s
60-
>r rg 5
60-
-c p ‘G 0 z
TBDMS b *CH,=~“-C”,CH,O-TBDMS
40219 \/
20-
235
263 /
234
264 /
TBDMS s x Y 2 s c ? ‘SI)
1,
80-
D
CH,-LCH,-LM-BDMS d
60-
d
40-
z
A dog (24 h starved, 18 kg) was anesthetized by an intravenous injection of thiopental (6 mg/kg) and succinylcholine (1 mg/kg). After tracheal intubation, the animal was ventilated with a mixture of 30% O2 and 0.8% halothane in nitrogen. A solution of saline was infused through a femoral vein catheter. Blood samples were taken from a femoral artery catheter on the opposite leg. Urine was collected through a bladder catheter. Starting at t = 0, a 1.0 M solution of BD was administered intravenously over 2.25 min (2.5 mmol/kg). Blood samples (1 ml) were taken from -30 to 180 min at various time intervals.
233
219 \/
221 \I
1001
,
D
TBDMS I
60s
$
;ii
,,,,,_,
II
1
,,
1
;2~,,36,
f3;
RESULTS 100
Figure 1 shows the electron impact (El) mass spectra of bis-TBDMS-BD and of its three isotopomers. The heaviest ion detected in the spectrum of the unlabeled species (Fig. lA), at m/z 261, is formed by the loss of a tbutyl radical from the molecular ion. This spectrum also
150
200
250
m/z FIG. 1. Electron impact mass spectra silyl derivatives of (A) R,S-1,3-butanediol, tanediol, (C) R,S-1,3-[1,1,3-‘HJbutanediol, ‘Ha, 3,4-‘3C2]butanediol.
of the bis-tert-butyldimethyl(B) R,S-1,3-[3,4-‘3C,Jbuand (D) R,S-1,3-[1,1,3-
104
DESROCHERS
includes peaks at m/z 233 and 219 which correspond to rearrangement ions formed from the [M-57]+ ion. The m/z 233 fragment is formed by the loss of ethylene (C-l and C-2 of BD) and the transfer of 0-TBDMS to C-3. The resulting ion contains C-3 and C-4 of the intact molecule. Confirmation of this rearrangement is provided by the spectra of each isotopomer which show a fragment incorporating the labeling on these two terminal carbons. The m/z 219 fragment is formed by the loss of propylene (C-2, C-3, and C-4 of BD) and transfer of O-TBDMS to C-l. The masses of the corresponding ions in the spectra of the isotopomers are also consistent with this loss. Figure 2 shows the CI mass spectra of the same compounds. The base peak for unlabeled BD is at m/z 319, corresponding to the [M + HI+ ion. There is also a minor peak at m/z 261, corresponding to the [M-57]+ EI fragment described above. The CI mass spectra of the three isotopomers produce intense [M + HI+ ions and reflect the extent of label incorporation. Figures 3A and 3B show four standard curves of unlabeled BD under EI and CI conditions, using [‘H3]BD and [‘H3, 13Cz]BD as internal standards. Note that the two panels of Fig. 3 cover different concentration ranges. The four curves are linear in the concentration range of 0.1 to 5 mM under EI conditions, and in the concentration range of 0.01 to 5 mM under CI conditions (r = 0.998 for all four curves). The slopes of the two EI curves are virtually identical ([2H3]BD 0.343 f 0.004 vs [2H3, 13C,]BD 0.337 + 0.003). This is also the case for the two CI curves ( t2H3]BD 0.397 + 0.002 vs [2H3, 13Cz]BD 0.413 + 0.003). However, for each internal standard, the slopes of the EI and CI curves differ by 15%. In contrast, the standard curves obtained using the [13C,]BD internal standard (not shown) are identical in the EI and CI modes (slopes = 0.340 t 0.001 and 0.340 + 0.003, r = 0.999) and are identical to the EI standard curves obtained using the [2H3]BD and [‘H3, 13C,]BD internal standards (Fig. 3). Figure 4 shows the uptake of BD by an anesthetized dog after bolus intravenous injection. Identical BD concentrations were obtained by analyzing the extracts of whole blood under EI and CI conditions. The intraassay coefficients of variation for duplicate assays of BD in blood samples were 4.8 and 2.3% under EI and CI conditions, respectively. Following bolus injection, BD concentration decreased rapidly showing that this substrate is well metabolized. The semilogarithmic plot shows an apparent biexponential decline of BD concentration. The linear portion of the graph corresponds to metabolism of BD. Extrapolation of this linear portion to t = 0 yields an “initial concentration” of 3.47 mM which corresponds to a distribution volume of 72% of body weight. Figure 5 shows other data from the same in vivo experiment, in particular the accumulation of physiological ketone bodies R-BHB and AcAc (the third physiological
ET
AL.
100,
A s
>r .% r 3 -r p J 0 B
31;
BO-
TBDMS d
60-
CH,-dH-CH,-CH,-0-TBDMS
40-
20-
07
‘.
L,
.
261 / .’ I
.
J
L
loo-
B
32;
60Iit x .$
TBDMS I *y *CH,-CH-CH,-CH1-O-TBDMS
60-
5 E $! ‘3 -0
40-
2
20.
0-t.
!
.I.
I
263 / .‘I
.
1. . (
loo-
C
32
