the chemistry vol. no. u. s. a.the journal of biological chemistry 0 1993 by the american society...

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 268, No. 27, Issue of September 25, pp. 20185-20190,1993 Printed in U. S. A .

Metabolism of 2,3-Butanediol Stereoisomers in the Perfused Rat Liver*

(Received for publication, February 16, 1993, and in revised form, June 2, 1993)

Jane A. MontgomerySj, France Davidv, Michel GarneauS, and Henri Brunengraberll From the $Department of Nutrition, University of Montreal, Montreal Quebec, H3C 357 Canada and the TDepartment of Nutrition, Case Western Reserve University, Cleveland Ohio 44106

The identification of 2,3-butanediol in sera of alcoholics led to the hypothesis that it may be a specific marker of alcohol abuse. We have investigated the metabolism of the individual isomers of 2,3-butanediol (2R,3R-, 2S,3S-, meso-2,3-butanediol and racemic 2,3-butanediol) in perfused livers from fed rats. Rates of uptake of the isomers decrease in the order (i) 2R,3R-, (ii) meso-, (iii) 2S,3S-2,3-butanediol. We observed interconversion of isomers and oxidation to acetoin with 2R,3R- and meso- but not with 2S,3S- 2,3-butanediol. In perfusions conducted in deuterium oxide, interconversion of isomers was accompanied by incorporation of deuterium. Thus, interconversion of isomers occurs via a reversible oxidation to acetoin with incorporation of hydrogen from water. In perfusions with either 2R,3R- or meso-[2-“C]2,3-butanediol, the substrates were converted to labeled acetate, R-3-hydroxybutyrate and COz, suggesting that 2,3- butanediol is oxidized to acetyl-coA via acetoin.

Although the epidemiology of some cases of alcoholism suggests a genetic predisposition, to date there are few, if any, reliable tests to identify susceptible individuals (1,2). Positive identification of both prealcoholic and alcoholic tendencies would benefit from the availability of biochemical markers (3, 4). Reports of elevated 2,3-butanediol (butanediol)’ in the serum and urine from inebriated alcoholics suggest that this compound could be of diagnostic value (5-8). It has been proposed that the plasma concentration of butanediol could be used as a marker of ethanol abuse and as an index of compliance with treatment of alcoholism (1).

2,3-Butanediol is formed in mammalian cells (9-17) and in microorganisms (18-22) by the reduction of acetoin (3-hy- droxybutan-2-one), a minor metabolite of pyruvate (23-31). In mammals, acetoin is formed by a side reaction of pyruvate dehydrogenase (23), as illustrated in Fig. 1, whereby hydroxyethylthiamine pyrophosphate (derived from pyruvate) con-

* This work was supported by National Institutes of Health Grant DK35543 (to H. B.) and by a fellowship from the Medical Research Council of Canada (to J. A. M.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact.

f To whom correspondence should be addressed Research Center, Notre-Dame Hospital, 1560 Sherbrooke St. E., Montreal, Quebec, H2L 4M1, Canada. Fax: 514-876-6630.

’The abbreviations and conventions used are: butanediol, 2,3- butanediol; RR-butanediol, 2R,3R-, D(-) or threo(-)-2,3-butanediol; SS-butanediol, 2S,3S-, L(+), or threo(+)-2,3-butanediol; meso-butanediol, R, S- or erythro-2,3-butanediol; RR,SS-butanediol, a mixture of coeluting 2R,3R- and 28,3S-2,3-butanediol in unknown proportions; GC, gas chromatography, GC-MS, gas chromatography-mass spectrometry, TMS, trimethylsilyl.

r . . f *

CH,-CO-COOH PY ruvate x TPP-PDH ~, )(.. CH,-CO-CHOH-CH, Acetoin

, - LAcetyl-dihydrollpoamide

” Lipoamide

t o 2 HETPP-PDH CH3-CH0 Acetaldehyde

FIG. 1. Pyruvate dehydrogenase-mediated synthesis of acetoin. TPP and HETPP, thiamine pyrophosphate and hydroxyethylthiamine pyrophosphate, respectively, are both bound to pyruvate dehydrogenase (PDH). The dotted line represents the normal transfer from hydroxyethylthiamine to lipoamide through the El subunit of the pyruvate dehydrogenase complex.

denses with acetaldehyde (derived mostly from ethanol oxidation). We have recently shown the general nature of acyloin (3-hydroxyalkan-2-one) formation by demonstrating that pyruvate dehydrogenase catalyzes the condensation of short to medium chain saturated aldehydes with hydroxyethylthiamine pyrophosphate to form the corresponding Cn+2 acyloins (32).

The mammalian metabolism of butanediol (23-27) is not as well documented as that in microorganisms (18-22). Two asymmetric carbon centers in butanediol give rise to three possible isomers, 2R,3R- (RR), 2S,3S- (SS), and R,S- (meso)- butanediol. The first two are enantiomers, and the third is a meso form. Plasma and urine samples from alcoholics were extracted, derivatized with achiral reagents, and analyzed by either gas chromatography-mass spectrometry (GC-MS) (5, 8) or gas chromatography (GC) (6, 33). Two peaks were identified as the meso-isomer and a coeluting mixture of the RR- and SS-butanediol enantiomers in unknown proportions. It has been suggested that this latter peak is more specific to alcoholism (6) as it remains elevated in sera from abstinent alcoholics with liver cirrhosis (34, 35). Using a sensitive GC- MS assay for butanediol, we demonstrated that both diastereomeric pairs are minor components of blood from nonalco- holics and that the concentrations of both increase after alcohol ingestion (36), implying that at least two isomers are related to ethanol metabolism.

To elucidate further the metabolism of butanediol, we perfused rat livers with individual isomers of butanediol (either unlabeled or 14C-labeled). We demonstrate that all three isomers are taken up by the liver, the RR-isomer at the highest rate. RR- and meso-butanediol are reversibly oxidized to acetoin with some conversion to their diastereomeric coun- terpart. In addition, butanediol is oxidized to COz and acetate. Our data are consistent with butanediol being metabolized in the liver to acetyl-coA, presumably via the reversal of the acetoin condensation reaction catalyzed by pyruvate dehydrogenase.

20185

20186 Metabolism of 2,3-Butanediol Isomers in Liver

EXPERIMENTAL PROCEDURES

Materials"BR,BR- and 2S,3S-2,3-butanediol, R,S-acetoin, 2,3-butanedione, acetaldehyde, and (-)-menthylchloroformate were obtained from Aldrich Chemical Co. Racemic 2,3-butanediol was purchased from Pfaltz and Bauer, Waterbury, CN. Meso-2,3-butanediol was obtained from Fluka Chemical Corp., Ronkonkoma, NY. Sodium borohydride was provided by Fisher Scientific. TRI-SIL/BSA was obtained from Pierce Chemical Co. Sodium borodeuteride (98% 'H) and deuterium oxide (99.9% 'H) were purchased from Merck, Sharp and Dohme Isotopes. [2-"C]Pyruvate and [14C]toluene (internal counting standard) were obtained from Du Pont-New England Nu- clear. Pyruvate dehydrogenase (pig heart), lactate dehydrogenase, NADH, and pyruvate were purchased from Boehringer Mannheim. Pyruvate decarboxylase (yeast) and thiamine pyrophosphate were purchased from Sigma. All aqueous solutions were made with water purified by a "Milli-Q" system (Millipore). Racemic [2,3-'H2]-2,3- butanediol was synthesized by sodium borodeuteride reduction of 2,3- butanedione (36).

[2-14C]Acetoin was synthesized enzymatically by a modification of the procedure of Gabriel et al. (10). The incubation contained 10 units of pig heart pyruvate dehydrogenase, 25 pmol (370 pCi) of sodium [2- "C]pyruvate, 120 pmol of acetaldehyde, 100 pg of thiamine pyrophosphate, 2 pmol of MgC12, 2 pmol of MnC12, and 150 pmol of potassium phosphate buffer, pH 6.5, in a final volume of 1 ml. The incubation was maintained at 37 "C until pyruvate concentration, assayed enzymatically (37), was undetectable (6 h). Then the reaction mixture, containing R-[2-14C]acetoin, was deionized on a column of mixed AG 1-X8-C1- and AG 50-X8-H+ resins. R-[2-'4C]Acetoin was reduced with NaBH4 to a mixture of RR-[2-l4C]butanedio1 and meso- [2-"C]butanediol. The isomeric composition of the product was ver- ified by the GC-MS assay of chiral mono-(-)menthylcarbonate-mono- TMS derivatives (38). RR- and mes0-[2-'~C]butanediol were sepa- rated on an Aminex HPLC HPX-87C column-Ca2+ column (300 X 7.8 mm, Bio-Rad) developed with deionized water (85 "C, 0.6 ml/ min). Meso- and RR-butanediol eluted a t 21 and 28 min, respectively.

When we tried the above synthesis with yeast pyruvate decarboxylase, the end products were RR-, SS-, and meso-butanediol, identified as the chiral derivatives. We conclude that yeast pyruvate decarboxylase yields partially racemic acetoin. Crout et al. (39) have demonstrated rearrangement of the keto group in acyloins by yeast pyruvate decarboxylase. This could result in racemization. Since there is no preparative procedure to separate RR- from SS-butanediol, we synthesized the I4C tracers using pig heart pyruvate dehydrogenase. Therefore, our investigations with "C tracers were conducted only with the RR- and meso-isomers.

Liver Perfusions-Male Sprague-Dawley rats (Charles River Lab- oratories) were fed ad libitum with Purina rat chow. For the series of liver perfusions with unlabeled isomers of butanediol, the rats weighed 210-300 g. For the series of perfusions with [2-'4C]butanediol, the rats weighed 160-210 g. Livers from fed rats were perfused with 150 ml of recirculating Krebs-Ringer bicarbonate buffer containing 4% dialyzed bovine serum albumin (fraction V, fatty acid-poor, Miles Scientific) and 15 mM glucose as initial substrate. The surgical procedure and perfusion apparatus were as described previously (40), with the following modifications: (i) Teflon tubing replaced Silastic tubing wherever possible, and (ii) gas flow through the oxygenator was stopped, and the oxygenator was sealed when butanediol was added to the perfusate. Earlier studies indicated that enough oxygen is present in the sealed oxygenator for more than 90 min of perfusion (41). The perfusate was maintained at pH 7.4 by titration with 0.3 N NaOH.

Perfusion Protocols-After 30 min of equilibration, a 2 mM concentration of either RR- or SS- or racemic-butanediol was added to the perfusate. For the series using [2-14C]butanediol, the specific activity of RR-[2-"C]butanediol was 35,000 dpm/pmol, 2 mM, and meso-[2- "C]butanediol was 64,000 dpm/pmol, 1 mM. Six livers were used with each substrate.

Two livers were perfused with buffer made up in deuterium oxide (99.9% 'H) containing 15 mM glucose and a 2 mM concentration of either RR-butanediol or SS-butanediol. In addition, to determine whether the liver would produce butanediol from ethanol, a liver was perfused with 15 mM glucose and 20 mM ethanol for 2 h. After 1 h, 5 mM pyruvate was added to the perfusate. In this perfusion, no exogenous butanediol was added.

Analytical Procedures-Lactate (42), pyruvate (37), R-3-hydroxybutyrate (43), acetoacetate (44), and acetate (45) were assayed enzymatically. Butanediol was quantitated by isotope dilution GC-MS

under ammonia chemical ionization conditions as described previously (36). Briefly, samples were (i) spiked with racemic [2,3-'H2] butanediol as internal standard; (ii) treated with sodium borodeuter-

extracted and derivatized using TRI-SIL/BSA. The [M+H]+ ions for ide to produce [2-2Hl]butanediol from endogenous acetoin; and (iii)

the TMS derivative of butanediol (m/z 235), acetoin as 'H1-butanediol (m/z 236) and the internal standard ('H2-butanediol, m/z 237) were monitored, and the ratios of peak areas (after correction for natural abundance and isotopic impurities) were used to quantitate butanediol and acetoin in the perfusates.

In certain experiments, the three butanediol isomers were analyzed by GC-MS as the mono-(-)menthylcarbonate-mono-TMS derivatives after treating extracts with (-)menthylchloroformate (38), followed by TRI-SIL/BSA. These derivatives elute on the HP-5 column in the order RR + meso, SS- and meso-butanediol. Meso-butanediol forms two distinct derivatives in equal proportions, depending on whether (-)menthylchloroformate reacts with the R- or S-hydroxyl group. The RR component of the first peak (composed of RR + meso- butanediol) can be calculated by deduction of its meso portion which is equal to the area of the third peak. The preparation of the chiral derivatives proved to be difficult because of the instability of the (-)menthylchloroformate reagent. Therefore, this assay was used only to assess the purity and isomeric composition of substrates, tracers, and internal standards.

l4COZ Production-Production of l4COZ was measured as described previously (41). Aliquots (10 ml) of the alkalinized final perfusate were acidified, and the evolved 14C02 was retrapped in 2 ml of 2 N NaOH in a scintillation vial. Standards of NaH"C03 and [2-"C] butanediol were analyzed separately to assess recovery of "COP (96- 100%) and the lack of contamination of trapped I4CO2 by [2-"C] butanediol, respectively.

Isolation of 14C-labeled Acids-To identify acidic products of [2- 14C]butanediol metabolism, perchloric acid extracts of perfusate were gassed with COz and N, and neutralized prior to chromatography on an AG 1-X8-C1- column (15 X 1 cm). The column was developed with a linear gradient made up of 75 ml of H20 and 75 ml of 40 mM HC1. One major peak containing I4C coeluted with a standard of [3H] acetate added to the sample. The identity of the peak was confirmed by converting the acetate to citrate (41) and rechromatography on a similar column developed with 60 ml of 15 mM HC1, followed by 60 ml of 60 mM HCl. Then 80% of the acetate radioactivity was trans- ferred to citrate. The 10-20% remaining was identified as 3-hydroxybutyrate by enzymatically converting it to acetoacetate using R-3- hydroxybutyrate dehydrogenase. Upon rerunning the column, the residual radioactivity was then associated with acetoacetate. All counts of 14C were converted to dpm by recounting with an internal standard of [14C]toluene.

To measure the specific activity of acetate in the final perfusate, neutralized perchloric acid extracts of 1.5-ml aliquots of perfusate were analyzed on a ClS reversed phase HPLC column (Varian), 300 X 8 mm, mobile phase 0.01 N H2S04, flow rate 1.5 ml/min. Retention times were: acetate, 11.5 min; pyruvate, 11.6 min; lactate, 12.6 min; HC03,17.0 min; acetoacetate, 19.3 min; 3-hydroxybutyrate, 20.5 min.

Statistics-Data were analyzed using one-way analysis of variance, followed by the Bonferroni test to identify significant differences between groups. Perfusions with labeled butanediol were analyzed separately from perfusions with unlabeled butanediol because of the difference in weights in the rats. The significance level was set a t p < 0.05.

RESULTS

Analysis of perfusates of control livers from fed rats indicated that any endogenous production of butanediol or acetoin was below the detection limit of the assay (1 pM). In a preliminary perfusion with 20 mM ethanol, neither acetoin nor butanediol was produced during the 1st h. However, when 5 mM pyruvate was added, at that point, acetoin and butanediol accumulated to 15 p~ over the 2nd h.

In perfusions with unlabeled butanediol isomers, uptake rates were linear with the rate for RR- being significantly higher than that for SS-butanediol (Table I). In perfusions with racemic butanediol, the uptake rate for meso-butanediol was twice that of RR,SS-butanediol, reflecting the rela- tive isomeric composition in the commercial product (meso:RR,SS, 2:l). A racemic butanediol with high meso

Metabolism of 2,3-Butanediol Isomers in Liver 20187

TABLE I Uptake of 2$butanediol isomers and production of related compounds by perfused rat livers

Livers from fed rats were perfused with 2 mM initial concentration of the listed 2,3-butanediol isomers. In the case of racemic butanediol, the initial concentrations of its RR,% and meso components were 0.8 and 1.2 mM, respectively, and uptake is presented for each of the two diastereomers. All rates are expressed as nmol X min-’ X g-i (mean + SE.).

Substrate (n) Butanediol uptake Acetoin production Isomer conversion

nmol/min/g RR-Butanediol (6) 171 + 21 14.6 f 3.8 8.5 -t 0.7 SS-Butanediol (6) 74 f 9 0.3 r 0.3 ND” Racemic butanediol (6) 4.6 +- 1.7 NMb

RR,SS 29 f 5 meso 57 f 3

One-way ANOVA’ for unlabeled butanediols RR veraus SS*, p < 0.002 RR uersus SS, p < 0.01 RR versus SS, p < 0.001 RR uersvs racemic, p <

0.05

RR-[“ClButanediol (6) Meso-[“C]Butanediol (6) One-way ANOVA for i4C-butanediols

125 f 8 49 f 3

RR versus meso, p < 0.0001

28.6 + 5.7 9.2 k 2.9

RR vers’sus meso, p < 0.02

5.7 k 0.8 3.1 + 0.8 RR versus meso, p <

0.05

’ ND, not detectable. ’ NM, not measurable because of racemic nature of butanediol. ’ ANOVA, analysis of variance. d Data for racemic butanediol were not included in the statistical analysis.

content was used since pure meso-butanediol was not available at the time of these early experiments. The presence of more than one butanediol isomer did not affect the uptake rates of individual isomers.

In perfusions with 2 mM RR-[2-Wlbutanediol or 1 mM meso-[2-Wlbutanediol, uptake of RR- was significantly higher than that of meso-[Wlbutanediol (Table I). The uptake of the meso-isomer was comparable to that seen in perfusions with racemic butanediol where the meso-diastereomer concentration was initially 1.2 mM. Again, the rate of uptake of 1 mM meso-[14C]butanediol was approximately half that of RR-butanediol (either Y-labeled or unlabeled).

The TMS (achiral) derivatives of RR,SS- and meso-butanediol form two separate GC peaks because of the presence of two asymmetric centers. Thus, it was technically possible to follow the conversion of RR- or SS-butanediol to meso- butanediol, and of meso-butanediol to RR- and/or SS-butanediol (the actual isomer could not be positively identified) (Fig. 2). Similarly, it was not possible to monitor the interconversion of diastereomers in racemic butanediol which con- tains all three isomers producing two GC peaks. Only RR- and meso-butanediol were converted to other isomers, and the processes were linear with time. In perfusions with SS-butanediol, there was no accumulation of meso-butanediol. In fact, there was a slight decrease in meso-butanediol (a small amount of which was present in the commercial SS-butanediol) indicating that it may have been partially metabolized. Therefore, because of the lack of conversion of SS-butanediol to meso-butanediol, we assumed that the isomer accumulating in perfusions with the meso-butanediol substrate was RR- butanediol. Analysis of an extract of per&sate from an RR- butanediol perfusion, after reaction with the chiral-derivatiz- ing reagent, indicated that meso-butanediol accumulates with little, if any, further conversion to SS-butanediol (Fig. 3). In the second series of perfusions, the rate of production of meso- [Wlbutanediol from RR-[Ylbutanediol (Table I) is significantly higher than (approximately double) the rate of formation of RR,SS-[Ylbutanediol from the meso-isomer.

Oxidation of butanediol produces acetoin and possibly 2,3- butanedione (diacetyl). Acetoin was measured by GC-MS as [2-‘Hilbutanediol (M+l) following sample treatment with sodium borodeuteride. There was no evidence for the produc-

L

30 40 50 60 70 80 90 100 110 120

TIME (min)

FIG. 2. Interconversion of 2,3-butanediol isomers. Livers were perfused with 2 mM RR-2,3-butanediol (0, the meso form was assayed), 2 mM SS-2,3-butanediol (v, the meso form was assayed), or 1 mM meso-2,3-butanediol (m, the RR,SS form was assayed). Data are mean f S.E. (n = 6).

tion of [2,3-‘HJbutanediol (M+2) which would be formed from 2,3-butanedione. RR-butanediol, which exhibited the highest uptake rate, also displayed a significantly greater production of acetoin (Table I). In keeping with a lower uptake, SS-butanediol produced minimal (if any) amounts of acetoin, and racemic butanediol showed intermediate rates. In the second series of perfusions, acetoin production was significantly higher with RR-[i4C]butanediol than with meso- [‘4C]butanediol. Acetoin concentrations reached a plateau at the midpoint of the perfusion, indicating that it is probably not an end product of butanediol metabolism (Fig. 4).

Two livers were perfused with either RR- or SS-butanediol in buffer prepared with deuterium oxide to determine whether isomer interconversion occurs with the incorporation of hydrogen from water. In the perfusion with RR-butanediol,


RR-23BD

d and meso-23BD

SS-23BD

meso-23BD

12:s I

13.0 Retention Time (min)

FIG. 3. 2,3-Butanediol isomer resolution in liver perfusate with 2R,3R-2,3-butanediol as substrate. The mono-(-)menthylcarbonate-mono-TMS derivatives form three peaks. The peak eluting at 12.8 min is a mixture of RR- and meso-butanediol (23BD), the second peak at 12.95 min is SS-butanediol, and the third peak at 13 min is meso-butanediol.

T

I I I I I I I I I I

30 40 50 60 70 80 90 100 110 120

TIME (min) FIG. 4. Production of acetoin by perfused livers. Acetoin was

measured in perfusate as [2-2Hl]butanediol, after sample reduction with sodium borodeuteride. Livers were perfused with 2 mM RR-2,3- butanediol (O), 2 mM SS-2,3-butanediol (V), or 1 mM meso-2,3- butanediol (W). Data are mean f S.E. ( n = 6).

meso-[*Hl]butanediol accumulated to 10 p~ after 60 min and remained at this concentration for the following 30 min. A smaller amount of RR,SS-[2Hl]butanediol accumulated (3 p ~ ) . There was no evidence for the incorporation of two deuterium atoms in either diastereomer. In the perfusion with SS-butanediol, no meso-[2Hl]butanediol accumulated, and no deuterium was incorporated into RR,SS-butanediol.

Perfusate lactate, pyruvate, 3-hydroxybutyrate, and acetoacetate were measured as indices of the cytosolic and mitochondrial NADH/NAD+ redox states. The accumulation of the redox indicators and the ratios [lactate]/[pyruvate] and [3-hydroxybutyrate]/[acetoacetate] were not significantly affected by any butanediol isomer.

Acetate concentrations rose to 0.8 f 0.1 and 0.7 f 0.1 mM after 90 min of perfusion with RR-[14C] or meso-['*C]butanediol, respectively. Acetate accumulation was greater than

in (i) a control liver perfused under similar conditions (0.4 mM) and (ii) fasted livers perfused with 4 mM acetone (41). The specific activity of acetate (2,740 f 300 dpm/pmol with RR-butanediol and 2,730 f 400 dpm/pmol with meso-butanediol) was 10-20 times lower than the initial specific activity of the butanediol substrates.

Table I1 shows the partial balance of the uptake of RR-[2- "C] and mes0-[2-'~C]butanediol. Production of 14C02, acetate, ketone bodies, acetoin, and other isomers of butanediol accounted for approximately one-third of the label uptake.

DISCUSSION

Control livers from fed rats perfused with 15 mM glucose did not produce detectable amounts of butanediol (<I p ~ ) . In these experiments, pyruvate accumulated in the perfusate to 0.5 f 0.1 mM (n = 6). Dawson et al. (16) have reported that acetoin is produced through pig heart pyruvate dehydrogenase incubated with 5 mM pyruvate as sole substrate. This mechanism must be inactive in intact liver cells in the presence of pyruvate concentrations 10 times higher than those occurring in uiuo. This suggests that acetaldehyde is required for acetoin formation in the liver. However, the addition of 20 mM ethanol (precursor of acetaldehyde) to the perfusate does not in itself produce the necessary conditions. Only the addition of pyruvate (5 mM) to the perfusion with ethanol stimulated both acetoin and butanediol formation. Oxidation of ethanol to acetaldehyde increases the [NADH]/[NAD+] ratio and de- creases pyruvate concentration. Addition of pyruvate provides the second substrate necessary for acetoin production. NADH concentration (an inhibitor of pyruvate dehydrogenase) is reduced by the addition of pyruvate, which stimulates pyruvate dehydrogenase activity (46). Thus, the increase in substrate availability, possibly combined with increased pyruvate dehydrogenase activity, results in acetoin formation. These observations are consistent with the hypothesis of Veech et al. (11) that during ethanol metabolism the liver is incapa- ble of producing acetoin via pyruvate dehydrogenase. This results from the inhibition of the citric acid cycle by a high [NADH]/[NAD+] ratio, as confirmed by a decrease in CO, production.

Butanediol and acetoin are the reduced and oxidized forms of a redox couple. Thus, supply of butanediol to the liver could produce a shift in the cytosolic or mitochondrial NADH/NAD+ redox potential. However, the [lactate]/[pyruvate] and [3-hydroxybutyrate]/[acetoacetate] ratios in the perfusate were not affected by the addition of butanediol. Oxidation of butanediol to acetoin yields 1 NADH which consumes 1 oxygen atom to produce 3 ATP. The maximum butanediol uptake (0.17 pmol X g" X rnin") corresponds to

TABLE I1 Partial balance of the uptake of 14C-labeled RR- and

meso-Z,3-butanediol Data are presented as percent of the uptake of label from RR- and

meso-["C]2,3-butanediol, perfused at an initial concentration of 2 or 1 mM, respectively (mean * S.E., n = 6). Incorporation of label in CO,, acetate, and R-3-hydroxybutyrate was measured directly. Incor- poration of label in acetoin and other isomeds) was calculated from the concentrations of these species, assuming that they had the same sDecific activitv as the I"C12.3-butanediol substrate.

Substrate "'OZ hydroxybutyrate Acetoin conversion Acetate -+ R-3- Isomer

% RR-["C]2,3- 9 f 2 11 + 2 1 2 f 4 6 + 2

butanediol

meso-I"C1-2.3- 10 * 1 8 + 1 7 + 1 5 + 1 butanediol

Metabolism of 2,3-Butanediol Isomers in Liver 20189

the production of 0.5 @mol of ATP, which is only 4% of the rate of ATP turnover (based on oxygen uptake of 2 @mol X g" X min"). Such a low production of NADH is insufficient to affect the shuttles of reducing equivalents in the liver cell.

In perfusions with specific isomers of butanediol, we have demonstrated significant differences in their uptake rates. High rates of uptake correlate with increased production of acetoin and increased rates of conversion to other diastereomers of butanediol. An exception is SS-butanediol, which appears to be metabolically inert. However, differences in metabolite production disappear when label incorporation into acetoin is normalized to the uptake of label from RR- or me~o-['~C]butanediol (Table 11). Note that the observed rate of isomer conversion is a minimal estimate, since acetoin can be reduced to either RR- or meso-butanediol. Thus, the actual conversion could be up to twice the observed rate.

The simplest mechanism for the interconversion of the butanediol isomers involves the enzymatic oxidation of butanediol to R-acetoin and subsequent reduction of acetoin back to butanediol. Deuterium incorporation during isomer conversion with RR-butanediol as substrate confirms the reversible oxidation-reduction of butanediol. Veech et al. (11) demonstrated the reduction of acetoin to butanediol in the liver without addressing the issue of stereoisomerism. Gabriel et al. (10) suggested the presence of an R-acetoin reductase in liver extracts which produced both diastereomers of butanediol. Our observation that deuterium is incorporated into both diastereomers is consistent with the data of Gabriel and co- workers and suggests that the reduction of acetoin is either nonstereospecific or is catalyzed by a second enzyme with stereospecificity opposite to the first as illustrated in Fig. 5. As very little acetoin or meso-butanediol accumulated during perfusions with SS-butanediol, the latter appears to be metabolized very slowly, if a t all, by the enzymes catalyzing the interconversion of isomers. Chiral analysis of butanediol iso- lated following liver perfusion with RR-butanediol indicated that accumulation of SS-butanediol accounted for only 1-2% of the butanediol isomers as illustrated in Fig. 2. This is also consistent with the lack of production of acetoin in perfusions with SS-butanediol.

The possible spontaneous racemization of R-acetoin, via ketoenol tautomerism, was tested by dissolving acetoin in deuterium oxide at neutral pH. After 24 h, acetoin was reduced with borohydride, and the resulting butanediol was analyzed for deuterium incorporation. No incorporation was observed, but the sensitivity of the assay was such that a deuterium enrichment of 1-2% would have been difficult to determine. I t is also possible that keto group rearrangement from C2 to C3 could occur (39) with chiral inversion and no deuterium incorporation from the medium. However, we did not observe keto group displacement in acyloins produced from saturated aldehydes where borodeuteride reduction resulted in asymmetric 2,3-alkanediols (32) with no evidence of deuterium incorporation into the C3 position. However, we cannot rule out the possibility that SS-butanediol is derived from small amounts of S-acetoin produced via racemization of R-acetoin.

The combined radioactivity in l4CO2 and [14C]acetate rep-

7% 7% y 3

HO-C-H I

I

c-0 H-C-OH I

H-C-OH I

H-C-OH *"-* H.&H c"--*

CH3 CH3 I

CH3

RR-2,3-Butanedm R-Acetoln meso-2 3-Butaned~al

diol isomers. FIG. 5. Proposed enzymatic interconversion of 2,3-butane-

resents 20% of the uptake of 14C from butanediol and is independent of the isomer (Table 11). The percentage of butanediol uptake found as l4CO2 is comparable to that observed in Gabriel's study with liver minces incubated with [2,3-14C]acetoin where 0.5% of the acetoin was oxidized to COz (10). At supraphysiological concentrations, substrate OX-

idation rates expressed as percentages of the uptake are largely dependent on substrate concentration. In incubations with lower concentrations of acetoin, a much higher fractional conversion to COz was found (10).

Although acetate is a known metabolite of butanediol and acetoin in microorganisms (20,21), we believe that this is the first confirmed report of acetate production from butanediol in a mammalian species. Gabriel and colleagues (12) suggested that acetate might be a product of acetoin oxidation in rat liver minces. Production of l4CO2, [14C]acetate, and R-3-hydroxy-[14C]butyrate from [14C]butanediol in our experiments strongly suggests that butanediol is metabolized to acetyl- CoA. Gabriel et al. (10) reported scrambling of 14C between the two Cz units of acetoin synthesized from [2-14C]pyruvate and unlabeled acetaldehyde. This could result from a partial reversal of the pyruvate decarboxylase reaction whereby acetoin is oxidized to hydroxyethylthiamine pyrophosphate and acetaldehyde (47). In our experiments on acyloin production from saturated aldehydes in uitro, we observed small amounts of compounds resulting from the condensation of either two Cz units or two C, units where n is the number of carbons in the aldehyde added to the incubation (32). This is consistent with a partial reversal of the pyruvate dehydrogenase reaction with an acyloin as substrate. Although Alkonyi et al. (17) found that acetyl-coA could not be produced using acetoin as substrate in the presence of required cofactors and pyruvate dehydrogenase from pigeon breast muscle, this does not pre- clude a partial reversal to produce acetaldehyde which could be oxidized to acetate.

If we assume that there was some label scrambling during the synthesis of [14C]butanediol resulting in [2,3-14C]butanediol, the production of [14C]acetate in liver occurs probably by two concurrent processes, each one carrying label from one of the Cp moieties of the diol. [2,3-14C]Butanediol is oxidized to [2,3-14C]acetoin which is split by pyruvate dehydrogenase to hydr~xy-['~C]ethylthiamine pyrophosphate and [14C]acetaldehyde. Hydroxy- [14C]ethylthiamine pyrophosphate, via pyruvate dehydrogenase, yields ['4C]acetyl-CoA which is hydro- lyzed to [14C]acetate, presumably by acetyl-coA hydrolase (48, 49). [14C]Acetaldehyde is oxidized to [14C]acetate by acetaldehyde dehydrogenase. Simultaneously, [14C]acetyl-CoA can be formed by three mechanisms: (i) from hydroxy-[14C] ethylthiamine pyrophosphate via pyruvate dehydrogenase, as above; (ii) from dismutation of [2,3-14C]acetoin; and (iii) by activation of [14C]acetate. The low specific activity of ["C] acetate (5-10% that of [2,3-14C]butanediol) reflects the dilution of the [14C]acetyl-CoA pool by unlabeled acetyl-coA derived from glucose and endogenous fatty acids.

In summary, there are marked differences in the metabolism of the individual butanediol isomers in rat liver. RR- and meso-butanediol are reversibly oxidized to acetoin resulting in interconversion of isomers. SS-Butanediol does not readily convert to the other isomers. The conversion of [14C]butanediol to l4COZ and [14C]acetate shows that production of butanediol, by whatever mechanism(s), does not lead to dead-end products. However, there is still no simple explanation for the persistence of elevated plasma concentrations of butanediol in cirrhotics who abstain from alcohol (34, 35). Our results emphasize the need to investigate butanediol metabolism in cirrhosis to determine the mechanism(s) leading to its accu-


mulation. Extension of our work to studies of butanediol metabolism in humans will require (i) synthesizing the three isomers labeled with 13C, and (ii) a practical assay of chiral derivatives for precise GC-MS measurements of their concentration and I 3 C enrichment.

REFERENCES 1. Thacker, S. B., Veech, R. L., Vernon, A. A., and Rutstein, D. D. (1984)

2. Skinner, H. A., Holt, .??, Sheu, W. J., and Israel, Y. (1986) Br. Med. J . 2 9 2 , Alcoholism Clin. Ex Res 4,375-383

1703-1708 3. Eckhardt, M. J., Rawlings, R. R., and Martin, P. R. (1986) Prog. Neuro-

4. Schuckit, M. A. (1986) Prog. Neuro-psychopharmacol. & Biol. Psychiatry psychopharmacol. & Biol. Psychiatry 10,135-144

1 n ~ 1 q 1 - 1 w 5. Mamer, 0. A,, Montgomery, J. A,, 'Ijoa, S. S., Crawhall, J. C., and Feld-

6. Rutstein, D. D., Nickerson, R. J., Vernon, A. A., Kishore, P., Veech, R. L., kamp, C. S. (1978) Biomed. Mass Spectrom. 5,287-290

Felver, M. E., Needham, L. L., and Thacker, S. B. (1983) Lancet 2,534-

", " _ _ _ _

527 7. Wglf, S., Felver, M. E., Altschule, M. D., Werthessen, N. T., Gerner, R,,

8. Sisfontes, L., Nyborg, G., Jones, A. W., and Blomstrand, R. (1986) Clin.

9. Westerfeld, W. W., and Berg, R. L. (1943) J. Biol. Chem. 148 , 523-528

and Veech, R. L. (1983) Br. J . Psychiatry 142,388-390

Chim. Acta 155 , 117-122

10. Gabriel, M. A,, Jabara, H., and AI-Khalidi, U. A. S. (1971) Biochem. J .

11. Veech, R. L., Felver, M. E., Lakshmanan, M. R., and Huang, M.-T. (1981)

12. Gabriel, M. A,, Ilhawi, M., and AI-Khalidi, U. A. S. (1972) Comp. Biochem.

13. Lemoigne, M., and Monguillon, P. (1930) C. R. Acad. 191,80-82 14. Dawson, J., and Hullin, R. P., (1954) Biochem. J. 5 7 , 177-180

16. Dawson, J., Hullin, R. P., and Walker, M., (1957) Biochem. J . 67,456-463 15. Dawson, J., and Hullin, R. P., (1954) Biochem. J. 5 7 , 180-185

17. Alkonyi, I., Bolygo, E., Gyosci, L., and Szabo, D., (1976) Eur. J . Biochem.

18. Stanier, R. Y., and Fratkin, S. B. (1944) Can. J . Res. Sect. B 2 2 , 140-153 19. Happold, F. C., and Spencer, C. P. (1952) Biochim. Biophys. Acta 8, 18-29 20. Sebek, 0. K., and Randles, C. I. (1952) Arch. Biochem. Biophys. 4 0 , 373-

21. Juni, E., and Heym, G. A. (1956) J. Bacteriol. 71,425-432 22. Taylor, M. B., and Juni, E. (1960) Biochim. Biophys. Acta 39, 448-457 23. Green, D. E., Westerfeld, W. W., Vennesland, B., and Knox, W. E. (1942)

124, 793-800

Curr. Top. Cell. Regul. 18 , 151-179

Physiol. 41B, 493-502

66,551-557

380

J . Bid. Chem. 145,69-84

24. Westerfeld, W. W., Stotz, E., and Berg, R. L. (1942) J. Biol. Chem. 144 ,

25. Lubin, M., and Westerfeld, W. W. (1945) J. Biol. Chem. 161,503-512 26. Juni, E. (1952) J. Biol. Chem. 195 , 727-734 27. Schreiher, G., Kohlhaw, G., Goedde, H. W., and Holzer, H. (1963) Biochem.

28. Juni, E. (1952) J. Biol. Chem. 195,715-726 29. Ui, S., Masuda, H., and Muraki, H., (1984) J. Ferment. Techno!. 6 2 , 151-

30. Tanko, B., Munk, L., and Abonyi, I., (1940) Hoppe-Seyler's 2. Physiol.

31. Juni, E., and Heym, G., (1956) J. Bid. Chem. 218,365-378 32. Montgomery, J. A., Des Rosiers, C., and Brunengraber, H . (1992) Bid.

33. Needham, L. L., Hill, R. H., Orti, D. L., Felver, M. E., and Liddle, J. A.

34. Casazza, J. P., Song, B. J., and Veech, R. L. (1990) Trends Biochem. Sci.

657-665

Z. 339,83-93

156

Chem. 264,91-107

Mass Spectrom. 2 1 , 242-248

(1982) J. Chromatogr. 233,9-17

15. 3fi-20 35. Casazza. J.. Frietas. J.. Stamhuk. D.. Morean. M.. and Veech. R.. (1987)

",__ "

Alco&lAlcohol. 1, (suppl.) 607k09 ' '

36. Montgomery, J. A,, Jette, M., and Brunengraber, H. (1990) Anal. Biochem. 185,71-76

37. Czok, R., and Lamprecht, W. (1974) in Methods of Enzymatic Analysis (Bergmeyer, H. U., ed) 2nd Ed., Vol. 3, p. 1446, Academic Press, New

, . .

York 38. W&$, J. W., and Halpern, B. (1968) J. Org. Chem. 33, 3978-3980 39. Crout, D., Dalton, H., Hutchinson, D., and Miyagoshi, M., (1991) J. Chem.

40. BrunenEraher, H., Boutry, M., and Lowenstein, J. M. (1973) J. Biol. Chem. SOC. Perkin Trans. 1, 1329-1334

41. Gavino, V. C., Somma, J., Philbert, L., David, F., Garneau, M., Belair, J., 248,-2656-2669

42. Guttmann, I., and Wahlefeld, A. (1974) in Methods of Enzymatic Analysis and Brunengraber, H. (1987) J. Biol. Chem. 262,6735-6740

(Bergmeyer, H. U., ed) 2nd Ed., Vol. 3, p. 1465, Academic Press, New

43. Williamson, D. H., and Mellanby, J. (1974) in Methods of Enzymatic York

Analysis (Bergmeyer, H. U., ed) 2nd Ed., Vol. 3, pp.1836-1839, Academic Press, New York

44. Mellanby, J., and Williamson, D. H. (1974) in Methods of Enzymatic Analysls (Bergmeyer, H. U., ed) 2nd Ed., Vol. 3, p. 1840, Academic Press, New York

45. Guynn, R. W., and Veech, R. L. (1975) Methods Enzymol. 35,302-307 46. Cooper, R., Randle, P., and Denton, R. (1974) Biochem. J . 143,625-641 47. Walsh, D., Cooper, R., Denton. R., Bridges, B., and Randle, P. (1976)

48. Prass, R. L., Isohashi, F., and Utter, M. F. (1980) J. Biol. Chem. 2 5 5 ,

49. Knowles, S. E., Jarrett, I. G., Filsell, 0. H., and Ballard, F. J. (1974)

Biochem. J. 157,41-67

5215-23

Biochem. J. 142,401-411

the chemistry vol. no. u. s. a.the journal of biological chemistry 0 1993 by the american society...

Documents