Journal of Neurochemistry
`Raven Press, New York
`0 1987 International Society for Neurochemistry
`
`Evidence for the Participation of a Cytosolic NADPf-
`Dependent Oxidoreductase in the Catabolism of
`y-Hydroxybutyrate In Vivo
`
`Elaine E. Kaufman and Thomas Nelson
`
`Laboratory of Cerebral Metabolism, National Institute ofMental Health, U.S. Public Health Service,
`Department of Health and Human Services, Bethesda, Maryland, U S A .
`
`Abstract: The concentration of y-hydroxybutyrate (GHB)
`in brain, kidney, and muscle as well as the clearance of [ 1-
`14C]GHB in plasma have been found to be altered by the
`administration of a number of metabolic intermediates and
`drugs that inhibit the NADP+-dependent oxidoreductase,
`“GHB dehydrogenase,” an enzyme that catalyzes the oxida-
`tion of GHB to succinic semialdehyde. Administration of
`valproate, salicylate, and phenylacetate, all inhibitors of
`GHB dehydrogenase, significantly increased the concentra-
`tion of GHB in brain; salicylate increased GHB concentra-
`tion in kidney, and cu-ketoisocaproate increased GHB levels
`in kidney and muscle. The half-life of [1-I4C]GHB in
`
`plasma was decreased by D-ghcuronate, a compound that
`stimulates the oxidation of GHB by this enzyme and was
`increased by a competitive substrate of the enzyme, L-gulo-
`nate. The results of these experiments suggest a role for
`GHB dehydrogenase in the regulation of tissue levels of en-
`dogenous GHB. Key Words: y-Hydroxybutyrate dehydro-
`genase-Succinic
`semialdehyde dehydrogenase-D-Glu-
`curonate-Sodium valproate. Kaufman E. E. and Nelson T.
`Evidence for the participation of a cytosolic NADP+-depen-
`dent oxidoreductase in the catabolism of y-hydroxybutyr-
`ate in vivo. J. Neurochem. 48, 1935-1941 (1987).
`
`Studies on the metabolic fate of y-hydroxybutyric
`acid (GHB), a naturally occurring compound present
`in both brain (Roth and Giarman, 1969; Roth, 1970)
`and peripheral tissues (Nelson et al., 198 I), have es-
`tablished that this compound is largely disposed of by
`oxidation to C02 and water (Walkenstein et al., 1964).
`Furthermore, it has been found that most of the car-
`bon skeleton enters the citric acid cycle as succinate
`(Doherty et al., 1975; Mohler et al., 1976) rather than
`as acetyl-CoA derived from @-oxidation as previously
`proposed by Walkenstein et al. (1964). The recent dis-
`covery of a metabolic disease in which GHB and suc-
`cinic semialdehyde (SSA) are markedly elevated in
`both blood and urine due to a block in SSA dehydro-
`genase (Jakobs et al., 1981; Gibson et al., 1983) adds
`evidence to support a degradative pathway in which
`GHB is oxidized to SSA, which in turn is oxidized to
`succinate.
`These findings strongly suggest that the main degra-
`dative pathway for GHB proceeds through the follow-
`ing series of steps.
`
`(3) succinate - - - - CO2 + H20
`
`(1) GHB-SSA
`(2) SSA-
`
`succinate
`
`The reactions in step (3) are catalyzed by the enzymes
`of the citric acid cycle and in step (2) by SSA dehydro-
`genase, the enzyme that is either low or missing in pa-
`tients with GHB aciduria (Gibson et al., 1983). Al-
`though step (1) is always depicted as an essential part
`of this scheme, until recently an enzyme or enzymes
`that could catalyze this reaction had not been identi-
`fied.
`We have reported the purification and characteriza-
`tion of a cytosolic NADP+-dependent oxidoreduc-
`tase, “GHB dehydrogenase,” which can catalyze the
`oxidation of GHB to SSA (step 1) in vitro under con-
`ditions that approximate those in the cytosol of both
`brain and some peripheral tissues (Kaufman et al.,
`1979; Kaufman and Nelson, 1981). The physical
`characteristics as well as substrate and inhibitor speci-
`ficity of this enzyme indicate that the ability to cata-
`
`Received August 5 , 1986; revised manuscript accepted December
`15, 1986.
`Address correspondence and reprint requests to Dr. E. E. Kauf-
`man at Laboratory of Cerebral Metabolism, National Institute of
`
`Mental Health, 36/1A-05, 9000 Rockville Pike, Bethesda, MD
`20205 U.S.A.
`Abbreviations used: GHB, y-hydroxybutyric acid; SSA, succinic
`semialdehyde.
`
`1935
`
`Ranbaxy Ex. 1016
`IPR Petition - USP 9,050,302
`
`

`
`I936
`
`E. E. KAUFMAN AND T. NELSON
`
`lyze the oxidation of GHB may represent a previously
`unreported activity for the NADP+-dependent oxi-
`doreductase (EC 1.1.1.19), commonly known as D-
`glucuronate reductase (York et al., 1961).
`Whether this enzyme plays a significant role in the
`oxidation of GHB to SSA in vivo has been studied
`by the administration to animals of compounds that
`either inhibit or increase the activity ofthis enzyme in
`vitro. Determination of tissue levels of GHB or of the
`half-life of G H B in plasma was used to assess the effect
`of the inhibition of this enzyme on the catabolism of
`GHB in vivo. The results suggest that factors that
`affect the activity of this enzyme in vitro alter both
`tissue levels of GHB and the half-life of GHB in vivo
`in directions consistent with their effects on the en-
`zyme in vitro.
`
`MATERIALS AND M E T H O D S
`Materials and animals
`Sodium a-ketoisocaproate, sodium phenylacetate, y-
`aminobutyric acid (GABA), and sodium GHB were pur-
`chased from Sigma Chemical (St. Louis, MO, U.S.A.). Val-
`proic acid was obtained from Saber Laboratories (Morton
`Grove, IL, U.S.A.) and salicylic acid from Fisher Scientific
`(Fairlawn, NJ, U.S.A.). Sprague-Dawley male rats weighing
`350-450 g were obtained from Taconic Farms (Ger-
`mantown, NY, U.S.A.). [ I-I4C]GHB was obtained from
`Research Products
`International
`(Mount Prospect,
`IL, U.S.A.).
`Effects of inhibitors of GHB dehydrogenase
`In studies in which the animals were infused with solu-
`tions of either a-ketoisocaproate, phenyl acetate, or 0.9%
`sodium chloride, a catheter was implanted in the femoral
`vein under light halothane anesthesia. The animals were al-
`lowed to recover from the anesthesia for at least 2 h before
`the experiment was started. The solutions were infused for
`2 h intravenously at constant rates with a Harvard infusion
`pump, model number 600 (Harvard Apparatus, Dover,
`MA, U.S.A.). Immediately after the infusions the animals
`were killed by decapitation and the brains, kidneys, and
`quadriceps muscle were removed and quickly frozen in liq-
`uid nitrogen. The organs were stored at -80°C until assayed
`for GHB content as described by Nelson et al. (198 1).
`In the studies in which either salicylate or valproate was
`administered, the drug or physiological saline solution was
`administered intrapentoneally to the experimental and con-
`trol animals, respectively. The animals were killed by decap-
`itation 1 h after receiving salicylate and 2 h after valproate.
`Brain, kidney, and quadriceps muscle were rapidly frozen
`in liquid N2 and assayed for GHB as described above.
`Effects of D-ghCUrOnate or L-gulonate on the tIl2 of
`[1-I4C]GHB in plasma
`Catheters were implanted in both the femoral vein and
`artery of Sprague-Dawley male rats under halothane anes-
`thesia and the animals were allowed to recover from the sur-
`gery and anesthesia for at least 2 h. The experiment was
`started by the injection into the venous catheter of approxi-
`mately 4 pCi of [1-'4C]GHB (sp act 4.5 pCi/pmol) in 0.4 ml
`of 0.9% NaCI. In one series of experiments D-glucuronate
`(333 mg) was administered (intravenously) as a bolus in sa-
`line just prior to the start of the experiment followed by a
`
`J . Neurochem , Vol. 48, No. 6, 1987
`
`constant infusion (556 mg/h) during the experiment. In a
`second series of experiments, L-gulonate (240 mg) was given
`as a bolus just prior to the start of the experiment followed
`by a constant infusion of 320 mg/h for 3 h. Control rats
`received a matching bolus and constant infusion of physio-
`logical saline. Blood samples were withdrawn at various
`times into heparinized tubes and centrifuged. The plasma
`was deproteinized by the addition of 100 pl of ice-cold abso-
`lute ethanol to 50 pl of plasma. Sixty microliters of the de-
`proteinized plasma were counted in Aquasol (New England
`Nuclear, Boston, MA, U.S.A.), and the remainder was used
`for paper chromatography. Chromatographic separation of
`GHB was camed out by spotting approximately 20 pl of the
`deproteinized plasma on Whatman 3MM Chr paper. The
`solvent system was ethanol/H20/NH40H, 98:2: 1, by vol.
`The chromatogram was developed in the ascending direc-
`tion for approximately 18 h, cut into 1-cm strips, and
`counted in a Packard Tri-carb scintillation counter, Model
`3375 (Packard Instrument, Downers Grove, IL, U.S.A.) in
`10 ml of Aquasol and 1 ml of H20. The fraction of total of
`I4C in each plasma sample that migrated with authentic
`GHB was measured and used to determine the concentra-
`tion of [I4C]GHB found in the plasma sample in the pres-
`ence of radioactive metabolites of GHB. The plasma
`[ I4C]GHB concentrations were plotted on semilogarithmic
`paper against time. The straight line that was obtained after
`the initial equilibration with the tissues was extrapolated to
`zero time and used to calculate the tlI2 for the disappearance
`of GHB from plasma.
`Preparation and assay of GHB dehydrogenase and of
`SSA dehydrogenase
`GHB dehydrogenase was prepared from the livers of
`adult male golden Syrian hamsters as previously described
`by Kaufman et al. (1979). Inasmuch as the previous work
`on the kinetics of GHB dehydrogenase was done with the
`enzyme purified from hamster liver, it is important to note
`that the GHB dehydrogenase from rat brain and from rat
`kidney has been shown to cross-react with an antibody to
`the purified hamster liver enzyme (unpublished results).
`GHB dehydrogenase was assayed at 37°C in a reaction mix-
`ture containing 0.08 A4 potassium phosphate, pH 7.6,
`0.0025 A4 NADP', purified enzyme, and 0.01 M sodium
`GHB. Rat brain SSA dehydrogenase was prepared accord-
`ing to Whittle and Turner (1978); the purification was car-
`ried out through the ammonium sulfate fractionation step,
`and was assayed as described by Whittle and Turner (1 978).
`Rat brain and kidney cytosol (100,000 g supernatant frac-
`tion) were prepared according to the method of Sokoloff and
`Kaufman (196 1).
`
`RESULTS
`Effects of a-ketoisocaproate or phenyl acetate on
`tissue levels of GHB
`Phenylacetate and a-ketoisocaproate, metabolic
`products of phenylalanine and leucine, are potent in-
`hibitors of GHB dehydrogenase (Table 1). If this en-
`zyme plays a role in the disposition of GHB in vivo,
`then these inhibitors, when administered to a n ani-
`mal, should decrease the rate of degradation of GHB
`and thereby increase the tissue level of GHB. The
`effects of infusions of these compounds are shown in
`Table 2. a-Ketoisocaproate produced a small but not
`
`

`
`REGULATION OF ENDOGENOUS 7-HYDROXYBUTYRATE
`
`1937
`
`TABLE 1. Comparison ofKi values for compounds
`that are inhibitors of both GHB dehydrogenase
`and SSA dehydrogenase
`
`Ki values
`
`Compound
`
`Valproate
`
`Salicylate
`a-Ketoisocaproate
`Phenyl acetate
`
`SSA dehydrogenase GHB dehydrogenase
`4.0 x
`M=
`4.8 x ~ O - ~ M =
`1.2 x I O - ~ M
`1.7 X lo-' M
`3.5 x 10-3 M
`
`5.7 X lO-'M
`1.15 x 1 0 - 4 ~
`1.7 x 1 0 - 4 ~ ~
`5.0 x
`~b
`
`The Ki of4.8 X 1 0-3 Mfor sodium valproate for SSA dehydroge-
`nase is taken from Whittle and Turner (1978), and the value of 4.0
`Mfrom Maitre et al. (1976).
`X
`*The Ki values of a-ketoisocaproate and phenyl acetate for GHB
`dehydrogenase are taken from Kaufman et al. (1983).
`
`statistically significant increase in the level of GHB in
`the brain. It did, however, produce a twofold increase
`in the level of GHB in both kidney and muscle. Phe-
`nyl acetate, by contrast, increased the level of GHB in
`brain by 2.4-fold but did not increase the level in kid-
`ney or muscle.
`Effects of sodium valproate and sodium salicylate on
`tissue levels of GHB
`The drugs sodium salicylate and sodium valproate
`are also excellent inhibitors of GHB dehydrogenase
`(Table 1) and therefore might also be expected to alter
`tissue levels of GHB. Administration (intraperitoneal)
`of sodium salicylate to rats produced a twofold in-
`crease in the level of GHB in brain and a 1.5-fold in-
`crease in kidney (Table 3). Sodium valproate pro-
`duced a 1.4-fold increase in the GHB level in brain
`and a small but insignificant decrease in the kidney.
`Three of the four inhibitors that were used produced
`statistically significant increases in the concentration
`of GHB in brain. The fourth inhibitor, y-ketoisoca-
`proate, produced a small increase that did not reach
`statistical significance (Tables 2 and 3). These data
`support a role for GHB dehydrogenase in brain.
`In contrast to the effects of these inhibitors of GHB
`
`dehydrogenase on GHB levels in brain, two of these
`did not increase GHB levels in kidney and quadriceps
`muscle and therefore raise questions about the contri-
`bution of GHB dehydrogenase to the metabolism of
`GHB in these particular peripheral tissues. One expla-
`nation would be that there are different concentra-
`tions of valproate or phenyl acetate in the cytosol of
`these tissues. Secondly, the question of whether equal
`concentrations of these inhibitors would exhibit
`different magnitudes of inhibitory effects in these tis-
`sues was addressed by testing the ability of these
`compounds to inhibit GHB dehydrogenase activity in
`rat brain cytosol and rat kidney cytosol. The inhibi-
`tion found in cytosol was compared to that found with
`the purified enzyme (Table 4). All four inhibitors,
`when tested at two concentrations, produced a degree
`of inhibition of GHB dehydrogenase activity in the
`cytosol from both brain and kidney that was compara-
`ble to that found with the purified enzyme (Table 4).
`These data provide further evidence of the identity of
`the enzyme in rat brain cytosol and in rat kidney cyto-
`sol with purified hamster liver enzyme, an identity
`previously established by titration with an antibody to
`the purified enzyme. Finally, it should be noted that
`since tissue levels represent a balance between synthe-
`sis and degradation, an effect of these compounds on
`the rate of synthesis could either magnify or obscure
`an effect on the rate of degradation.
`Effects of D-glucuronate and L-gulonate on the tlI2 of
`(1-'4qGHB in plasma
`Since the results obtained by examining the effects
`of inhibitors of GHB dehydrogenase on the tissue con-
`centrations of GHB in specific peripheral tissues such
`as kidney were not conclusive, a method that would
`measure the sum of GHB metabolism occurring in the
`whole animal, namely, a measure of the tIl2 of GHB
`in plasma, was also included.
`In vitro experiments in which the oxidation of GHB
`was coupled to the reduction of D-ghCUrOnate dem-
`onstrated that the addition of D-glucuronate mark-
`edly accelerates the oxidation of GHB catalyzed by
`GHB dehydrogenase under conditions approximat-
`
`TABLE 2. Efects of a-ketoisocaproate and ofphenyl acetate on tissue levels of GHB
`
`Saline
`infusion
`
`nmol GHB/g
`tissue
`
`2.6 f 0.3 (5)
`27.8 f 3.2 (5)
`22.2 2 3.0 (6)
`
`Tissue
`
`Brain
`Kidney
`Muscle
`
`a-Ketoisocaproate infusion
`
`Phenyl acetate infusion
`
`nmol GHB/g
`tissue
`
`3.2 f 0.1 (3)
`55.0 t 8.2 (4)"
`46.4 k 9.9 (4)'
`
`Percent of control
`
`123
`198
`209
`
`nmol GHB/g
`tissue
`
`6.1 2 1.0(4)*
`18.4 f 1.3 (4)"
`16.7 2 7.1 (4)
`
`Percent of control
`
`235
`66
`75
`
`Phenyl acetate (1.0 M ) was given intravenously as an initial bolus of 1.5 ml followed by a constant infusion of g2.0 ml/h for 2 h.
`a-Ketoisocaproate (0.5 M ) was given intravenously as an initial bolus of 0.6 ml followed by a constant infusion of g2.0 ml/h for 2 h. At the
`end of the infusion the animals were killed and the tissues were removed and assayed for GHB as described under Materials and Methods. All
`values are means & SEM, numbers of animals in parentheses.
`a p < 0.05.
`bp <0.01.
`
`J. Neurochem.. Vol. 48, No. 6, 1987
`
`

`
`I938
`
`E. E. KAUFMAN AND T. NELSON
`
`TABLE 3. Eflects of salicylate and of vulproate on tissue levels of GHB
`
`Saline
`(control)
`
`nmol GHB/g
`tissue
`
`Tissue
`
`Salicylate
`
`Valproate
`
`nmol GHB/g
`tissue
`
`Percent of control
`
`nmol GHB/g
`tissue
`
`Percent of control
`
`Brain
`
`Kidney
`
`5.8 f 0.5 (4)b
`
`52.8 f 2.5 (5)"
`
`200
`
`153
`
`2.9 t 0.5 (6)
`2.6 t 0.2 (4)
`34.5 t 5.7 (6)
`30.0 k 4. I (4)
`79
`23.7 f 3.8 (4)
`Sodium valproate (100 mg/kg, i.p.) was given 2 h prior to decapitation; sodium salicylate (500 mg/kg, i.p.) was given 1 h prior to decapitation.
`Tissues were removed and assayed as described under Materials and Methods. All values are means f SEM, numbers ofanimalsin parentheses.
`a p < 0.05.
`p < 0.01.
`
`3.7 t 0.2 (4)b
`
`142
`
`ing those existing in the cytosol of brain, kidney, and
`muscle (Kaufman and Nelson, 198 1). L-Gulonate, the
`product of D-glucuronate reduction, competes with
`GHB as a substrate for GHB dehydrogenase and
`therefore inhibits GHB oxidation. We have examined
`the effects of both of these compounds on the tl/2 of
`[ 1-14C]GHB in plasma and found that the administra-
`tion of D-glucuronate in vivo decreases the tlI2 by 33%
`whereas administration of L-gulonate increases it by
`33% (Fig. 1). The decrease in the tlR is consistent with
`an increased rate of oxidation of GHB by GHB dehy-
`drogenase in the presence O f D-glUCUrOnate; similarly,
`the increase in tl,2 caused by L-gulonate suggests a de-
`crease in the rate of oxidation of GHB. These results
`are consistent with our in vitro findings.
`
`Inhibition of SSA dehydrogenase and of GHB
`dehydrogenase by valproate, salicylate,
`a-ketoisocaproate, and phenyl acetate
`Since an increase in tissue levels of GHB produced
`by phenyl acetate, a-ketoisocaproate, valproate, and
`salicylate could be attributed to either an increase in
`the rate of synthesis of GHB or to a decrease in its
`
`rate of degradation or to a combination of both, we
`examined the inhibition of both SSA dehydrogenase
`and GHB dehydrogenase by valproate, salicylate, a-
`ketoisocaproate, and phenyl acetate. Inhibition of
`SSA dehydrogenase could theoretically lead to higher
`tissue levels of SSA and, therefore, an increased rate
`of synthesis of GHB, whereas inhibition of the GHB
`dehydrogenase activity could increase GHB levels di-
`rectly. Valproate, salicylate, a-ketoisocaproate, and
`phenyl acetate are all competitive inhibitors for GHB
`dehydrogenase with GHB as the variable substrate
`(Kaufman et al., 1983). Ki values for SSA dehydroge-
`nase were similarly determined for salicylate, phenyl
`acetate, and a-ketoisocaproate with SSA as the vari-
`able substrate. All of these compounds were competi-
`tive inhibitors of SSA dehydrogenase (data not
`shown). Whittle and Turner (1 978) have reported that
`sodium valproate is a noncompetitive inhibitor of
`SSA dehydrogenase.
`The Ki values for these compounds with both en-
`zymes are given in Table 1. In every case the Ki of
`these inhibitors for GHB dehydrogenase was one to
`two orders of magnitude lower than that for SSA de-
`hydrogenase.
`
`TABLE 4. Effects of inhibitors ofpurijied NADPi-dependent
`GHB dehydrogenase on the NADPi-dependent oxidation
`of GHB in rat bruin cytosol and in rut kidney cytosol
`
`Inhibitor
`
`a-Ketoisocaproate
`a-Ketoisocaproate
`Valproate
`
`Salicylate
`
`Phenyl acetate
`
`Percent inhibition
`
`Concentration
`(M)
`
`Purified GHB
`dehydrogenase
`
`~~
`
`Rat brain Rat kidney
`cvtosol
`cvtosol
`
`2
`5
`2
`5
`2
`5
`2
`5
`
`82
`96
`93
`100
`92
`94
`67
`93
`
`100
`100
`87
`100
`90
`100
`51
`12
`
`76
`93
`91
`100
`84
`91
`47
`84
`
`GHB dehydrogenase activity was assayed as described in Materials and Methods. In-
`hibitors were added to the reaction mixture in the concentrations indicated in the table.
`
`J. Neurochem., Vol. 48, No. 6, 1987
`
`

`
`REG ULA TION OF ENDOGENOUS 7-HYDROXYB UTYRA TE
`
`1939
`
`C O M P O U N D
`INFUSED
`
`SALINE
`0-GLUCURONATE
`L-GULONATE
`
`t 1 / 2 for[I4C]GHB
`Imin)
`mean 2 S.E.M.
`In)
`6 0 f 3.6
`4 0 5 9 I
`81
`0 . 5 s
`
`161
`( 2 1
`( 2 )
`
`~......a L-GULONATE
`v NaCL
`~ - - - L I D-GLUCURONATE
`
`0
`
`20
`
`40
`
`60
`
`80
`
`100 120 140
`
`160
`
`180
`
`200
`
`TIME (min)
`
`FIG. 1. Animals that received D-glucuronate received a bolus containing 240 mg followed by an infusion of 320 mg/h for 3 h. Animals
`receiving L-gulonate received a bolus of 333 mg and an infusion of 556 mg/h for 3 h. The tl,2 for the disappearance of ['4C]GHB from
`plasma was determined as described under Materials and Methods. *p < 0.05; ""p < 0.025.
`
`DISCUSSION
`
`This investigation was designed to determine
`whether GHB dehydrogenase, an enzyme that cata-
`lyzes the oxidation of GHB to SSA in vitro, plays a
`significant role in the disposition of GHB in vivo. We
`have also examined the possibility that the informa-
`tion obtained in this study might provide a reason-
`able explanation for the significant increases in GHB
`levels in brain brought about by the acute administra-
`tion of sodium valproate and other drugs used in the
`treatment of petit ma1 epilepsy (Snead et al., 1980).
`This would be of interest since investigations of the
`inhibition of aminobutyrate aminotransferase (EC
`2.6.1.19), SSA dehydrogenase (EC 1.2.1.16), as well
`as an NADPH-dependent aldehyde reductase (EC
`1.1.1.2) by sodium valproate (Godin et al., 1969; Har-
`vey et al., 1975; Sawaya et d., 1975; Whittle and
`Turner, 1978), have failed to produce a satisfactory
`explanation for the elevated GHB levels following an
`acute dose of sodium valproate.
`Marked inhibition of SSA dehydrogenase might be
`expected to lead to increased tissue levels of SSA and
`thereby to increased synthesis of GHB. Indeed, in pa-
`tients with GHB aciduria, a genetic disease in which
`this dehydrogenase is low or missing (Gibson et al.,
`1983), exactly such increases do occur. Little or no
`inhibition of SSA dehydrogenase would be expected
`
`to result from relatively high doses ( 100 mg/kg) of so-
`dium valproate since the Ki of sodium valproate for
`SSA dehydrogenase has been reported to be 4.0 X l 0-3
`M by Maitre et al., (1 976) or 4.8 X 1 0-3 M by Whittle
`and Turner (1 978). At tissue concentrations of 0.1
`mM- 1 .O mM, the highest concentrations likely to be
`found in clinical use (Sawaya et al., 1975), inhibition
`would vary from a negligible amount at 0.1 mM to
`12- 15% inhibition at 1 .O mM.
`In vitro studies of GHB dehydrogenase (Kaufman
`et al., 1979) have provided the first clue to the meta-
`bolic basis for the increased level of GHB in brain fol-
`lowing acute administration of sodium valproate
`(Snead et al., 1980). The activity of SSA dehydroge-
`nase in the brain of the adult rat (Pitts and Quick,
`1967) is approximately 1,000 times greater than that
`of GHB dehydrogenase (Kaufman et al., 1979).
`Therefore, in the sequence of steps leading to the for-
`mation of succinate from GHB, GHB dehydrogenase
`would catalyze the rate-limiting step. If GHB dehy-
`drogenase is the first enzyme in one of the quantita-
`tively significant pathways for the degradation of
`GHB, then administration of inhibitors of this en-
`zyme might lead to increased tissue levels of GHB.
`Phenyl acetate, a-ketoisocaproate, valproate, and sali-
`cylate, all potent inhibitors of GHB dehydrogenase
`with K, values in the range of 10-4-10-5 M, were se-
`lected to test this hypothesis. Indeed, administration
`
`J. Neurochem., Vol. 48, No. 6, I987
`
`

`
`1940
`
`E. E. KAUFMAN AND T. NELSON
`
`of these compounds increased the endogenous con-
`centration of GHB in one or more of the tissues that
`were assayed (Tables 2 and 3). The failure of a-keto-
`isocaproate to increase brain levels significantly may
`indicate that it does not readily cross the blood-brain
`barrier.
`The effect of sodium valproate on SSA dehydroge-
`nase and GHB dehydrogenase can now be compared
`under conditions seen in the clinical use of sodium
`valproate as an anticonvulsant drug. Sawaya et al.
`( 1975) have reported plasma concentrations of 1.1
`mM-0.07 mM in clinical use. In addition, Loscher
`and Frey (1 977) have reported that brain levels, in
`mice, are approximately 30% of the plasma levels. We
`can therefore calculate that, at a tissue concentration
`of 0.3 mM sodium valproate, GHB dehydrogenase
`would be 85% inhibited, a very marked inhibition
`compared to the 5% inhibition of SSA dehydrogenase
`which would occur at a 0.3 mM tissue concentration
`of the drug.'
`Even at a tissue concentration of 0.05 mM sodium
`valproate, a tissue concentration at which there would
`be essentially no inhibition of SSA dehydrogenase,
`GHB dehydrogenase still would be 48% inhibited.'
`Inasmuch as SSA levels could be elevated if phenyl
`acetate, a-ketoisocaproate, or salicylate were potent
`inhibitors of SSA dehydrogenase, Ki values for these
`compounds were determined for SSA dehydrogenase.
`It is apparent from a comparison of the Ki values (Ta-
`ble 1) that, like valproate, these compounds are also
`much better inhibitors of GHB dehydrogenase than
`of SSA dehydrogenase. In each case the Ki for GHB
`dehydrogenase is such that if the tissue level of any of
`these compounds reached a concentration of 5-10
`M, there would be significant inhibition ofthat
`X
`portion of the degradative pathway that goes through
`GHB dehydrogenase step with very little inhibition of
`SSA dehydrogenase.
`It is likely, therefore, that the increased tissue levels
`of GHB that follow the acute administration of val-
`proate and of other compounds of which salicylate,
`a-ketoisocaproate, and phenyl acetate are examples,
`result from inhibition of the oxidation of GHB to
`SSA, the first step in the degradative pathway to succi-
`nate and ultimately to COz , rather than from the inhi-
`bition of SSA dehydrogenase. The data presented in
`Tables 2 and 3 are consistent with a significant role
`for GHB dehydrogenase in the overall degradation of
`GHB in brain.
`
`' The relative velocities with and without inhibitor were calcu-
`lated from the Michaelis-Menten equation in which the K,,, term
`has been multiplied by the factor (1 + [I]/K,) in the presence of a
`competitive inhibitor. In the case of SSA dehydrogenase an arbi-
`trary value of 1 X
`M was assumed for substrate concentration
`M for the K,,, for SSA at pH 7.6 (unpublished results
`and 5 X
`from this laboratory). This value is similar to the K , of 2.7 X
`for SSA at pH 8.4 found by Albers and Koval(l961). In the case of
`M for the K, for GHB
`GHB dehydrogenase, values of 4.5 X
`were used. The K, values are those given in Table 1.
`
`J. Neurochem., Vol. 48, No. 6,1987
`
`Another in vivo reflection of the action of com-
`pounds that stimulate or inhibit the rate of a degrada-
`tive enzyme such as GHB dehydrogenase would be a
`change in the tl,Z for GHB in plasma after the admin-
`istration of such compounds. This method has the ad-
`vantage of examining the effect of such compounds
`on GHB metabolism in the whole body rather than in
`any one specific tissue; furthermore it can reveal
`effects on degradation uncomplicated by effects on
`synthesis since a trace dose of [I4C]GHB is used to fol-
`low the rate of degradation. In vitro, D-glucuronate
`markedly stimulated the GHB dehydrogenase-cata-
`lyzed oxidation of GHB by participating in a coupled
`oxidation-reduction reaction (Kaufman and Nelson,
`198 1). In vivo, the infusion of D-ghcuronate led to a
`33% decrease in the half-life of GHB in plasma, a re-
`sult consistent with an increased rate of degradation
`of GHB in body tissues. On the other hand, infusion
`of L-gulonate, which we have shown to be a competi-
`tive substrate of GHB for GHB dehydrogenase (un-
`published data), produced a 33% increase in the half-
`life for the plasma disappearance of GHB, indicating
`a decreased rate of degradation.
`Preliminary evidence from this laboratory indicates
`that in both brain and kidney there is at least one other
`enzyme, in addition to GHB dehydrogenase, that can
`convert GHB to SSA. The relative contribution of
`these two enzymes to the metabolism of GHB has not
`yet been determined. However, both the increase in
`tissue levels of GHB following administration of in-
`hibitors of GHB dehydrogenase and the change in the
`tIl2 for the plasma disappearance curve after infusion
`of either an inhibitor or a compound that stimulates
`the reaction are findings that are consistent with a role
`for GHB dehydrogenase in the overall degradation of
`GHB in vivo.
`
`Acknowledgment: We would like to thank Dr. Louis So-
`koloff for his most generous help and support in carrying
`out this project. The expert technical assistance of John
`Nine is gratefully acknowledged.
`
`REFERENCES
`Albers R. W. and Koval G. K. (196 1) Succinic semialdehyde dehy-
`drogenase: purification and properties of the enzyme from
`monkey brain. Biochim. Biophys. Acta 52,29-35.
`Doherty J. D., Stout R. W., and Roth R. H. (1975) Metabolism of
`[ l-14C]y-hydroxybutyric acid by rat brain after intraventricu-
`lar injection. Biochem. Pharmacol. 24,469-474.
`Gibson K. M., Sweetman L., Nyhan W. L., Jakobs C., Rating D.,
`Siemes H., and Hanefeld F. (1 983) Succinic semialdehyde de-
`hydrogenase deficiency: an inborn error of gamma-aminobu-
`tyric acid metabolism. Clin. Chim. Acta 133,33-42.
`Godin Y., Heinler H. L., Mark J., and Mandel P. (1969) Effect of
`dipropylacetate, an anticonvulsant compound on GABA me-
`tabolism. J. Neurochem. 16,869-873.
`Harvey P. K. P., Bradford H. F., and Davison A. (1975) The inhibi-
`tory effect of dipropylacetate on the degradative enzymes of the
`GABA shunt. FEBS Lett. 52,25 1-254.
`JakobsC., Bojasch M., Monch E., Rating D., Siemes H., and Hane-
`feld F. (1 98 1) Urinary excretion of gamma-hydroxybutyric
`acid in a patient with neurological abnormalities. The proba-
`
`

`
`REGULATION OF ENDOGENOUS 7-HYDROXYBUTYRATE
`
`1941
`
`bility of a new inborn error of metabolism. Clin. Chim. Acta
`111,169-178.
`Kaufman E. E. and Nelson T. (1981) Kinetics of coupled y-hy-
`droxybutyrate oxidation and D-glucuronate reduction by an
`NADP+-dependent oxidoreductase. J. Biol. Ckem. 256,6890-
`6894.
`Kaufman E. E., Nelson T., Goochee C., and Sokoloff L. (1979) The
`purification and characterization of an NADP+-linked alcohol
`oxidoreductase which catalyzes the interconversion of y-hy-
`droxybutyrate and succinic semialdehyde. J. Neurochem. 32,
`699-7 12.
`Kaufman E. E., Relkin N., and Nelson T. (1983) Regulation and
`properties of an NADP' oxidoreductase which functions as a
`y-hydroxybutyrate dehydrogenase. J. Neurochem. 40, 1639-
`1646.
`L6scher W. and Frey H. H. (1977) On the mechanism of action of
`valproic acid. Arzneim. Forsch./Drug Res. 21, 108 1-1082.
`Maitre M., Ossola L., and Mandel P. (1976) In vitro studies into the
`effect of inhibition of rat brain succinic semialdehyde dehydro-
`genase on GABA synthesis and degradation. FEBS Lett. 72,
`53-57.
`Mohler H., Patel A. J., and Balks R. (1976) Gamma-hydroxybu-
`tyrate degradation in the brain in viuo: negligible direct conver-
`sion to GABA. J. Neurochem. 27,253-258.
`Nelson T., Kaufman E., Mine J., and Sokoloff L. (198 1) The extra-
`neural distribution of y-hydroxybutyrate. J. Neurochem. 37,
`1345-1 348.
`
`Pitts F. N. and Quick C. (1967) Brain succinate semialdehyde dehy-
`drogenase-11, changes in the developing rat brain. J. Neuro-
`chem. 14,561-570.
`Roth R. H. (1970) Formation and regional distribution of y-hy-
`droxybutyric acid in mammalian brain. Biochem. Pharmacol.
`19,3013-3019.
`Roth R. H. and Giarman N. J. (1969) Conversion of y-aminobu-
`tync to y-hydroxybutyric acid in the rat. Biochem. Pharmacol.
`18,247-250.
`Sawaya M. C. B., Horton R. W., and Meldrum B. S. (1975) Effects
`of anticonvulsant drugs on cerebral enzymes metabolising
`GABA. Epilepsia 16,649-655.
`Snead 0. C., Bearden L. J., and Pegram V. (1980) Effect of acute
`and chronic anticonvulsant administration on endogenous y-
`hydroxybutyrate in rat brain. Neuropharmacology 19,47-52.
`Sokoloff L. and Kaufman S. (1961) Thyroxine stimulation of
`amino acid incorporation into protein. J. Biol. Chem. 236,
`795-803.
`Walkenstein S., Wiser R., Gudmundsen C., and Kimel H. (1964)
`Metabolism of y-hydroxybutync acid. Biockim. Biophys. Acta
`86,640-642.
`Whittle S. R. and Turner A. J. (1978) Effects of the anticonvulsant
`sodium valproate on y-aminobutyrate and aldehyde metabo-
`lism in ox brain. J. Neurochem. 31, 1453-1459.
`York J. L., Grollman A. P., and Bublitz C. (1961) TPN-L-gulonate
`dehydrogenase. Biochim. Biophys. Acta 47,298-306.
`
`J. Neurochem., Vol. 48, No. 6, 1987