Proc. Natl. Acad. Sci. USA Vol. 74, No. 11, pp. 4942-4946, November 1977 Biochemistry Enzymatic formation of glutathione-citryl thioester by a mitochondrial system and its inhibition by (-)erythrofluorocitrate (glutathione-S-citryl ester/metalloprotein/inner mitochondrial membrane/fluorocitrate toxic mechanism) ERNEST KUN, EVA KIRSTEN, AND MANOHAR L. SHARMA Departments of Pharmacology, Biochemistry, and Biophysics and the Cardiovascular Research Institute, University of California, San Francisco, Ca ifornia 94143 Communicated by Van R. Potter, August 29, 1977

ABSTRACT A soluble extract of the mitochondrial com- see ref. 11). The inability to demonstrate an inhibition of citrate partment composed of the inner membrane and matrix catalyzes efflux by fluorocitrate in mitochondria that had been loaded the enzymatic synthesis and hydrolysis of the 1:1 adduct of citric acid and glutathione. The adduct was identified as the thioester with citrate (9) was also explained by the stringent requirement by isolation with single and double isotope labeling ([14C]citric of preincubation of mitochondria with fluorocitrate in the ab- acid and [35S]glutathione) and by conversion to the monohy- sence of millimolar concentrations of citrate (cf. 7, 8), allowing droxamate of citric acid and comparison with the synthetic for the formation of two protein-fluorocitrate thioesters (8). product by thin layer chromatography and high voltage elec- The present report is concerned with the identification of a trophoresis. The enzymatic formation of the thioester (pH op- mitochondrial enzyme system that catalyzes the formation of timum 7.39 at 300) requires oxidized glutathione and citrate; both substrates exhibit a Michaelis-Menten kinetics. During the 1:1 adduct of glutathione and citric acid. The adduct was the enzymatic reaction equimolar quantities of thioester and identified as the thioester of glutathione, and kinetic evidence glutathione sulfinic acid are formed. After gel filtration or salt was also obtained that the mitochondrial extract contained, in fractionation the enzyme system requires Mn2+ (or Mg2+, which addition to the thioester-forming enzyme system, a hydrolase is less effective) for maximal activity. When extracts ofmitoplast capable of degrading the thioester. The citryl-glutathione- are tested, the time course of reaction is biphasic due to the forming enzyme system proved to be the most sensitive catalytic rapid synthesis of the product by the thioester-forming system (molecular weight 171,000) followed by its decay by the hydro- component of mitochondria towards the irreversible inhibitory lase (molecular weight 71,000). The two systems were separated action of (-)erythrofluorocitrate, suggesting that the molecular by molecular filtration on Sephadex G-200 and by precipitation site of action of this highly toxic agent has been identified. with (NH4)2SO4. The thioester-forming system is inhibited by preincubation with 0.5 mM mersalyl. Other inhibitors are METHODS 1,2,3propane tricarboxylic acid, 10 mM Ca2+, 200 mM K+, and the free radical trapping agent, phenazine methosulfate. The Lysosome-free mitochondria and mitoplasts (defined as the citrate-glutathione thioester formation is irreversibly and spe- matrix surrounded by the inner membrane) were isolated from cifically inhibited by (-)erythrofluorocitrate (50% inhibition rat livers as described (12). An extract of mitoplasts was pre- at 25 pmol of added fluorocitrate per mg of protein), which pared by stirring (at 00) of a suspension of mitoplasts (30 mg forms a trichloroacetic acid-stable adduct with the enzyme of protein per in M protein (at 50% inhibition, 0.8 pmol is bound to 1 mg of protein). ml) 0.25 sucrose/mannitol medium (12) Synthesis of malyl-glutathione thioester by inner membrane containing 1 mg of Brij 56 (Sigma Chemicals) per 10 mg of vesicles is selectively inhibited by (-)erythrofluoromalate. mitoplast protein for 15 min. The resulting lysate was diluted 1:3 with the medium; after removal of large granules (unbroken We have previously identified the biologically active species mitoplasts) by centrifugation at 10,000 X g (10 min at 40), the of the four possible isomers of monofluorocitric acid as supernatant (containing soluble extract of the inner membrane (-)erythrofluorocitrate (1, 2); its structure subsequently was and soluble proteins of the matrix and inner membrane vesicles) confirmed by x-ray crystallography (3, 4). Kinetic evidence was subjected to ultracentrifugation (145,000 X g for 1 hr at 20). indicated that (-)erythrofluorocitrate, when preincubated with The clear supernatant was concentrated by centrifugation (at isolated mitochondria at concentrations of 20-50 pmol of 20) in Amicon CF-25 membrane cones at 100 X g to one-fourth fluorocitrate per mg of mitochondrial protein, irreversibly in- of the original volume. Approximately 250 mg of protein ex- hibited bidirectional citrate transport (5-8). Whereas the high tract was obtained by this technique from 500 mg (protein) of sensitivity of citrate influx to low concentrations of fluorocitrate mitoplast starting material. The concentrated extract was used in intact mitochondria has been confirmed (9), an unusually directly in most experiments, unless specified, whereas the low competitive ki value for fluorocitrate was also reported for sediment containing the inner membrane vesicles was reho- the aconitase activity of disrupted mitochondria that had been mogenized in the suspending medium. Both fractions were preincubated with 1 mM Mg2+ (9). However, we have shown stored at -15° until used. The enzymatic activity of the soluble subsequently that mitochondrial aconitase is inactivated by extract decayed slowly at -150 (about half of the activity was Mg2+ in the presence of low concentrations of fluorocitrate (10) lost in 2 weeks) but much more rapidly after repeated thawing and apparent ki values calculated under these conditions have and refreezing. The activity was completely recovered when no bearing on the linearly competitive and reversible inhibition the stored extract was preincubated for 5-10 min with 5 mM of purified aconitase by (-)erythrofluorocitrate (ki = 290,M, reduced glutathione (GSH), demonstrating the essential role The costs of publication of this article were defrayed in part by the Abbreviations: GSH, reduced glutathione; GSSG, oxidized glutathione; payment of page charges. This article must therefore be hereby marked Hepes, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid; "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate citryl-SG and malyl-SG, glutathione thioesters of citric and malic acids, this fact. respectively. 4942 Downloaded by guest on October 2, 2021 Biochemistry: Kun et al. Proc. Natl. Acad. Sci. USA 74 (1977) 4943

@ - 1000 (2400 cpm)

E E 0. GSH 500 un U5 A Ia' GSSG

A (1,

- A 'A nAnA'A Qa 0 0 0 10 20 30 40 Number of strips (1 cm each) FIG. 1. Electrophoretic isolation of the citric acid-glutathione adduct containing either 14C or 35S as radioactive label. Unreacted citric acid was removed in the preceding step (Dowex 50-H+). Citric acid placed on the paper migrated quantitatively into the buffer reservoir at the anode

at pH 3.5 in 150 min. 0, 14C; A, 35S.

of protein-bound -SH groups for catalytic activity. As shown GSH, 15 mM GSSG, 220,uM citrate, Tris/N-2-hydroxyethyl- later, GSH was not a substrate of the enzyme, but served as a piperazine-N'-2-ethanesulfonic acid (Hepes) buffer (10 mM, thiol-protecting agent. pH 7.4), and sucrose mannitol (0.25 M) in a final volume of 600 The glutathione-citrate adduct was quantitatively assayed ,ul. After incubation for 5 min at 300, the reaction was stopped by its isolation as follows. After incubation of the enzyme with with 2 ml of 10% trichloroacetic acid. The product was ad- oxidized glutathione (GSSG), GSH, and citrate (for details see sorbed to Dowex 50 H+ (see Methods) from six 100-,l batches Results) containing [1,5-'4C]citric acid (98 mCi/mmol; New of the incubation mixture and eluted five times with 1 M pyr- England Nuclear) in a final volume of 100 pl, the reaction was idine (2 ml for each batch). The pyridine eluates were collected stopped with 300 yl of 10% trichloroacetic acid and the pre- into one-half volume of 2 M acetic acid (at 00) in order to avoid cipitated proteins were removed by centrifugation. An aliquot alkali-catalyzed rearrangement of the thioester (16) and of the trichloroacetic acid supernatant (300 .l) was directly freeze-dried, and the product was isolated by high voltage transferred onto a small column (1 X 2 cm, containing 1.5 ml electrophoresis at pH 3.5 (0.2 M acetate/formate) and 4000 V of resin bed) of Dowex 50 H+ equilibrated with 1 M formic for 150 min at 200 (14). In the experiment shown in Fig. 1 either acid/Na formate of pH 2.0 and the unreacted citric acid was citrate, 14C, or glutathione, 35S, was labeled. Both labels ap- quantitatively removed by 12 ml of formate (1 M, pH 2.0), as peared in the same radioactive peak, corresponding to a 1:1 monitored by radiochemical analysis. After the residual formate molar ratio of citric acid and GSH. From the above large-scale was eluted from the column with 3 ml of H20, the adduct was preparative experiment 200 nmol of citrate glutathione (1:1) quantitatively eluted with 6 ml of 1 M NaOH and an aliquot adduct was obtained, which was used in part for conversion to (1.5 ml), after neutralization with 1 M HCI, was analyzed by the hydroxamate (see Methods). Comparison of the NH2OH scintillation spectrometry with toluene/Triton X-100/Omni- derivative of the enzymatically synthesized product and the fluor (1 liter:1 liter:18 g) scintillator. Labeled GSH (15S, 19 hydroxamate prepared from the chemically synthesized mCi/mmol, obtained from Schwartz/Mann) was used as the monomethylester of citric acid showed that the two products second label for the product. Hydroxamate derivatives were migrated with identical mobility, RF of 0.27, whereas RF of the prepared from synthetic monomethyl ester of citric acid (pre- unreacted citric acid was 0.18 and the RF of the 1:1 adduct was pared by the diazomethane procedure, with 0.9 mol of diazo- 0.36. The hydroxamic acid derivatives of citric acid, obtained methane per mol of citrate in methanol) and from the enzy- either from the monomethylester or from the enzymatic matically synthesized product by reacting it directly (after product, were also electrophoretically separated at pH 1.85 (1.0 freeze-drying) with a methanolic solution of NH2OH (13) for M formic acid). Both hydroxamic acid derivatives moved 1 hr at 400 after 16 hr at room temperature. Hydroxylamino- towards the cathode at a rate of 2 cm/hr, whereas citric acid lysis was 60% under these conditions. The excess reagent was remained at the origin. Both synthetic and enzymatically removed by freeze-drying. Thin layer chromatography was formed hydroxylamine derivatives gave a reddish-violet spot carried out on silica gel sheets (Eastman) with butanol/acetic on the silica plate when sprayed with ethanolic FeCl3, typical acid/H20 (4:1:1) as developing solvent. The technique of high of hydroxamic acids. These results identified the enzymatic voltage electrophoresis on Whatman no. 3 paper and analytical product as citryl-glutathione thioester. methods have been described (14). (-)Erythrofluoro-[3- The time course of the formation of the GS-citric acid 614C]citric acid was synthesized by enzymatic condensation thioester catalyzed by the unfractionated extract of mitoplasts of [1-414C]oxalacetic acid with synthetic fluoroacetyl-coenzyme was characteristically biphasic (Fig. 2), indicating a rapid A (1, 2). The products was purified by high voltage electro- synthesis and, at a critical concentration of the product, a pro- phoresis at pH 1.85, resulting in homogeneous fluorocitrate (50 gressive hydrolysis to GSH and citric acid. It is also apparent mCi/mmol). Unlabeled (-)erythrofluoromalate was synthe- that the enzyme system is predominantly in the extract and only sized as described (15). traces of the thioester synthetase remain in membrane vesicles. Protein fractions containing either synthetic or hydrolytic ac- RESULTS tivities were readily separated from the extract of mitoplasts by molecular filtration on Sephadex G-200, developed with 50 The electrophoretic isolation of the enzymatically formed ci- mM Tris/Hepes, pH 7.4 at 4°. The citryl-SG synthetase activity trate-glutathione adduct is shown in Fig. 1. A typical reaction appeared in the molecular fraction exhibiting a mass of 171,000 mixture contained 1.76 mg (protein) of mitoplast extract, 5 mM daltons, whereas the hydrolase had a molecular mass of 71,000 Downloaded by guest on October 2, 2021 4944 Biochemistry: Kun et al. Proc. Natl. Acad. Sci. USA 74 (1977) Table 1. Effects of activators and inhibitors on enzymatic

4000- formation of GS-citryl thioester 0 / Extract L 3500 pmol cm E GS-thioester 3000 formed/mg Inhibitors or protein 2500 activators in 5 min 2000 2 None 2180 G~1500 0.5 mM Mn2+ 4260 1 mM Mg2+ 3770 1000 1 mM Ca2+ 1660

~~~~~~Vesicles 0 10 mM Ca2+ 80 500 1 mM 1,2,3-Propane tricarboxylate 405 1 mM Phenazine methosulfate 362 2 4 6 10 15 20 25 30 35 40 45 200 mM KCl 471 Time, min 0.5 mM Mersalyl FIG. 2. The time course of the enzymatic formation of GS-citryl (preincubation) 104 thioester catalyzed by an extract of mitoplast and by inner membrane vesicles. Amount of protein per test, 0.7 mg; GSH, 10 mM; GSSG, 10 An extract of mitoplast (0.4 mg of protein) was incubated with 10 mM; citrate, 100 1AM. 30°, pH = 7.4. Total volume, 100 ,gl. Small mM GSH, 10 mM GSSG, and 14C-labeled citrate (final molarity of 100 nonenzymatic rates were always subtracted. GS-citrate, citryl glu- AM) for 5 min in the presence of the agents listed. The effect of mersalyl tathione thioester. (0.5 mM) was tested by preincubation for 5 min prior to the enzymatic assay. daltons. These protein fractions represent a 450- and 250-fold tion of citryl glutathione by 96%. Phenazine methosulfate, a purification, respectively, as compared to total liver tissue. free radical trapping agent, inhibited thioester formation. Recombination of the two protein fractions reconstituted the Preincubation of the extract with the organomercurial reagent biphasic time course characteristic for the unfractionated ex- mersalyl inhibited citryl glutathione synthesis, in agreement tract. The time course of the reaction catalyzed by the protein with previous results (see Methods), which point to the essential fraction of 171,000 daltons was sustained for hours provided role of protein-bound thiol groups in catalysis. These results are 0.5 mM MnCl2 was also present (Fig. 3). These results illustrate shown in Table 1. that molecular filtration rendered the citryl-glutathione-syn- The enzymatic synthesis of citryl-glutathione thioester would thesizing enzyme system dependent on Mn2+. Separation of be expected to be an energy-requiring process; therefore a si- the thioester synthetase system from the hydrolase was also multaneous oxidation of S in glutathione to a higher oxidation accomplished by fractionation with (NH4)2SO4. The synthetase state was anticipated. Sulfinic acid of cysteine is not adsorbed system was recovered in the fraction between 0 and 35% on Dowex 50 H+ under our conditions (20, 21); therefore this (wt/vol). This preparation also exhibited dependence on Mn2+ technique provided a quantitative assessment of the formation for the synthesis of the thioester. No other cation except Mg2+ of glutathione sulfunic acid during the synthesis of citryl-glu- could replace Mn2+; Mg2+ was somewhat less effective, and tathione thioester. Within an experimental error of ±12%, 1 mol Ca2+ was inhibitory. The enzyme system was inhibited 80% by of sulfinic acid was generated for each mole of thioester formed, 200 mM K+. Since intramitochondrial concentration of K+ in indicating the simultaneous contribution of an oxygenase type the matrix can be as high as 180 mM (12), it is probable that of reaction (21) coincidental with the formation of the citryl- variation in the mitochondrial K+ content exerts a regulatory glutathione thioester. During all enzymatic reactions reagent effect on the activity of this enzyme system. A typical inhibitor blanks showed the formation of a small but measurable quantity of mitochondrial citrate translocation, 1,2,3-propane tricar- of thioester (0.5-1% of the enzymatic rate) and significant boxylate (17-19), at 1 mM concentration inhibited the forma- quantities of sulfinic acid. Sulfinic acid was detectable in an aqueous solution of GSSG (about 8%). The rates in the reagent blanks exhibited no temperature dependence between 00 and 0 C 400, while the enzymatic rates were more than doubled for each so4 - °0/ 10° within this temperature range. The pH optimum of the 0. synthetase reaction was 7.39 at 300. Typical Michaelis-Menten kinetics was obtained with citrate _0, 3 /Mn2 (Fig. 4) and GSSG (Fig. 5). The concentration of GSSG was maintained at 10 mM for the determination of the apparent Km E 2-/ of citrate because this concentration of glutathione corre- <,n/ sponded to known intracellular concentrations of total gluta- thione (22). Since the apparent Km for GSSG was 9 mM (at a .~ 1_ O_ O No Mn2+ fixed concentration of citrate of 220 ,M), it is evident that the rate of reaction is very sensitive to fluctuations in GSSG con- centrations, suggesting that the intracellular GSSG/GSH ratio 1 2 2.5 Time, hr and the absolute concentration of glutathione are powerful cellular regulators of this enzyme system. FIG. 3. The effect of 0.5 mM MnCl2 on the rates of GS-citryl When the extract of mitoplasts was preincubated with in- thioester formation by a protein fraction (molecular weight 171,000) of the extract of mitoplasts obtained by molecular filtration on Se- creasing concentrations of '4C-labeled (-)erythrofluorocitrate phadex G-200; 0.35 mg of protein per test system. Conditions were for 5 min at 300, the rates of formation of the GS-citrate the same as described in the legend of Fig. 2. thioester were progressively inhibited (Fig. 6). The protein- Downloaded by guest on October 2, 2021 IBiochemistry: Kun et al. Proc. Natl. Acad. Sci. USA 74 (1977) 4945 500 r-

c E 400 LoSf

0, ° 300 C a Q 6 .0 a; t 4 .~200 nmol/mg Km = 36 PM 0 . . 25 50 100 200 I20P0 30 40 pmol added/mg protein 100 1 Fcj, mM Citrate FIG. 6. Correlation between the inhibitory effect of (-)erythro- fluorocitrate on rates of GS-citrate thioester formation and the co- I I valent binding (8) of fluorocitrate to protein. Left ordinate shows the 100 200 300 400 500 synthetase activity (5 min) after preincubation of 0.9 mg of (protein) Citrate, ,M extract of mitoplasts for 10 min with 10 mM GSH + 10 mM GSSG and varying concentrations of fluorocitrate (0). Conditions were the FIG. 4. The effect of increasing concentrations of citrate on the same as described in the legend of Fig. 2. Right ordinate indicates the rates (5 min) of GS-citryl thioester formation catalyzed by an quantity ofprotein-bound fluorocitrate (M) determined under com- (NH4)2SO4 fraction (0-35%) of the extract. GSSG, 10 mM; MnCl2, parable conditions by incubating the extract with the same concen- mg per test conditions were the 0.5 mM; 0.3 of protein system. Other tration of fluorocitrate. Protein-bound fluorocitrate was determined same as described in the legend of Fig. 2, except no GSH was by precipitation of proteins with 10% trichloroacetic acid, redissolu- present. tion of precipitate in 0.1 M NaOH, reprecipitation with 10% trichlo- roacetic acid, and glass fiber filtration (8, 14) followed by scintillation bound fluorocitrate was simultaneously determined by pre- spectrometry. Fci, (-)erythrofluorocitric acid. cipitation of the protein with 10% trichloroacetic acid and ra- diochemical assay for bound fluorocitric acid by the glass fiber a reaction catalyzed by inner membrane vesicles, was selectively filtration technique (14). A correlation was found between the inhibited fluoromalate but not fluorocitrate. These results degree of inactivation of the GS-citryl thioester synthetase by by are shown in Table 2. The formation of malyl-SG or of other system and the quantity of acid-stable protein-fluorocitrate acid thioesters was the same method as product. About 50% inactivation occurred when 0.8 pmol of carboxylic assayed by described for citryl-glutathione except that L-[14C]malate or fluorocitrate were bound to 1 mg of protein, corresponding to other carboxylic acids were substituted for citrate. It is impor- 25 pmol of added fluorocitrate per mg of protein. The forma- tant that the malyl-SG-synthesizing system remained attached tion of the trichloroacetic acid-stable protein adduct of fluo- to the inner membrane vesicles (see Table 2), whereas the rocitrate continued at a linear rate even after almost complete citryl-SG synthetase system was extracted by the technique used inhibition of the synthetase. This phenomenon is readily ex- (see Methods; compare Table 2 with Fig. 2). Experiments plained by the fact that fluorocitrate was bound not only to the completed at the present time show that specific enzyme sys- synthetase, but also to the hydrolase, the latter being present tems exist in mitochondria that catalyze the GS-thioester for- in 3fold excess over the synthetase in the extract. The maximal mation of succinic, malic, a-ketoglutaric, glutamic, pyruvic, capacity of the extract to bind fluorocitrate in a trichloroacetic and isocitric acids and of glutamine; thus, the reaction described acid-stable form was 80 pmol/mg of protein, corresponding for citrate may have general metabolic significance. to the total amount of reactive sites in both enzymes. Synthetic (-)erythrofluoromalate (15), even at millimolar concentrations, DISCUSSION had no effect on the citryl-SG thioester-synthesizing system. On the other hand, the formation of malyl-glutathione thioester, The two mitochondrial proteins that form thioesters with (-)erythrofluorocitrate (8) were identified as citryl-S-gluta-

Table 2. Effects of fluorocitrate and fluoromalate on GS c 4 thioester-synthetase activities of submitochondrial preparations CL with citrate and L-malate as substrates.* E0 3 Enzyme Inhibitors preparation and Fluorocitrate, Fluoromalate, E c carboxylic 125 pmol/ 50 nmol/ a; acid substrate Control mg protein mg protein Extract + citrate 2770 110 2802 I aI Membrane vesicles + L-malate 305 298 0 Conditions for the assay for GS-citrate thioester formation were the 4 8 16 same as described in the legend of Fig. 2, except that the extract (1 mg GSSG, mM of protein) was preincubated with or without the inhibitors for 10 min FIG. 5. The effects ofincreasing concentrations of GSSG on the at 300. The concentration of L-malate (77,000 cpm/nmol) was 68 AM rate (5 min) of GS-citryl thioester formation by an extract of mito- in the GS-malyl thioester synthetase system; otherwise, conditions were plasts. Citrate, 100MuM; MnC12, 0.5 mM; protein = 0.2 mg/test system. the same as for the GS-citryl thioester-forming system. Other conditions were the same as described in the legend of Fig. * Results are expressed as pmol of thioester formed in 5 min per mg 4. of protein. Downloaded by guest on October 2, 2021 4946 Biochemistry: Kun et al. Proc. Natl. Acad. Sci. USA 74 (1977) thione synthetase and hydrolase, an observation that suggests no biochemical role for these substances has been proposed (27, that new sites of action of this toxic agent have been identified. 28). The enzymatic formation of thioesters involves GS' formed The concentrations of fluorocitrate that inhibit citrate transport by the homolytic cleavage of GSSG. The existence of GS' is well and the citryl-S-glutathione enzymes are of the order of mag- documented (29). nitude found in mitochondria of lethally poisoned rats (23). The exact role of the citryl-S-glutathione enzymes in mitochondrial This work was supported by National Institutes of Health Grants metabolism is unknown, but at least two possible functions may GM-20552 and HL-6285. Ernest Kun is a recipient of the Research be postulated. The thioester synthetase is also inhibited by Career Award of the U.S. Public Health Service. 1,2,3-propane tricarboxylate (Table 1), which is generally 1. Fanshier, D. W., Gottwald, L. K. & Kun, E. (1962) J. Biol. Chem. recognized as a specific inhibitor of citrate transport in intact 237,3588-3596. mitochondria (17-19). Coincidence of inhibition of citrate 2. Fanshier, D. W., Gottwald, L. K. & Kun, E. (1964) J. Biol. Chem. transport in mitochondria with inhibition of the soluble ci- 239,425-433. tryl-S-glutathione thioester synthetase tends to suggest that the 3. Carrell, H. L., Glusker, J. P., Villafranca, J. J., Mildvan, A. S., thioester synthetase may be a constituent of the mitochondrial Dummel, R. J. & Kun, E. (1970) Science 170, 1412-1414. 4. Carrell, H. L. & Glusker, J. P. (1973) Acta Crystallogr. Sect. B transport system. It has been reported that various dicarboxylic 29,674-682. acids exhibit competitive inhibition with respect to citrate 5. Kun, R (1972) "Carbon-Fluorine Compounds," Ciba Foundation uptake in intact mitochondria (18). This apparent competition Symposia (Elsevier, New York), pp. 70-76. cmn be readily demonstrated with inner membrane vesicles 6. Eanes, R. Z., Skilleter, D. N. & Kun, E. (1972) Biochem. Blophys. containing all di- and tricarboxylic acid-S-glutathione thioester Res. Commun. 46,1618-1622. synthetases, when a limiting concentration of GSSG is exposed 7. Kun, E. (1976) ACS Symposium Series 28, ed. Filler, R. (Amer- simultaneously to citric acid and dicarboxylic acids. In this ican Chemical Soc., Washington, DC), pp. 1-22. system apparent competition for GSSG as the second substrate 8. Kirsten, E., Sharma, M. L. & Kun, E. (1977) Mol. Pharmacol., of the reaction takes place. The citrate-malate exchange dif- in press. fusion in mitochondria has been reported to be sensitive to thiol 9. Brand, M. D., Evans, S. M., Mendes-Mourao, J. & Chappell, J. B. (1973) Biochem. J. 134,217-224. reagents (24), but observations to the contrary also exist (25), 10. Eanes, R. Z. & Kun, E. (1974) Mol. Pharmacol. 10, 130-139. whereas the soluble thioester synthetase requires preincubation 11. Glusker, J. P. (1971) "Aconitase" in The Enzymes, ed. Boyer, P. with mersalyl for inhibition (Table 1). Protein-thiol groups, D. (Academic Press, New York), Vol. 5, 3rd Ed., pp. 413-433. which are essential for citrate transport, may be not readily 12. Kun, E. (1976) Biochemistry 15,2328-2336. accessible to thiol, reagents; this inaccessibility can explain 13. Knappe, E. & Yekundi, K. G. (1964) Z. Anal. Chem. 203, 87- variable results. It has been shown that glutamate-aspartate 92. exchange is not sensitive to.ionic thiol reagents, but is inhibited 14. Kun, E., Chang, A. C. Y., Sharma, M. L., Ferro, A. M. & Nitecki, by uncharged -SH reagents (26). The apparent Vmax for ci- D. (1976) Proc. Natl. Acad. Sci. USA 73,3131-3135. was calculated to be 7-8* of 15. Skilleter, D. N., Dummel, R. J. & Kun, E. (1972) Mol. Pharmacol. trate-malate exchange nmol/mg 8,139-148. protein per min (18). The Vx estimated from Fig. 5 for the 16. Jaenicke, L. & Lynen, F. (1960) "Coenzyme A" in The Enzymes, soluble synthetase can account for only 15 of citrate transport eds. Boyer, P. D., Lardy, H. A. & Myrback, K. (Academic Press, in intact mitochondria. However, the Vmax of the soluble en- New York), Vol. 3,2nd Ed., pp. 4-101. yzme was found to be increased by 3-fold (on a mg of protein 17. Robinson, B. H., Williams, G. R., Halperin, M. L. & Leznoff, C. basis) when membrane vesicles were added back to the soluble C. (1970) Eur. J. Biochem. 15,263-272. enzyme, demonstrating that nonenzymatic membrane con- 18. Palmieri, F., Stipani, J., Quagliariello, E. & Klingenberg, M. stituents, e.g., lipids, have a decisive effect on the rate of en- (1972) Eur. J. Btochem. 26,587-594. zymatic catalysis. Therefore, the efficiency of the soluble en- 19. Meijer, A. J. & Van Dam, K. (1974) Biochim. Biophys. Acta 346, zyme for citrate transport cannot be estimated at this time. The 213-244. for was 20. Lombardini, J. B., Turini, P., Biggs, D. R. & Singer, T. P. (1969) apparent Km citrate in intact mitochondria calculated Physiol. Chem. Phys. 1, 1-23. from 20- to 60-sec rates at 90 to be 120 AM (18), whereas the 21. Lombardini, J. B., Singer, T. P. & Boyer, P. D. (1969) J. Biol. apparent Km of the soluble enzyme determined at 10 mM GSSG Chem. 244, 1172-1175. from 5-min rates at 300 was 36MAM (Fig. 4). Differences in these 22. Arias, I. M. & Jacobi W. B., eds. (1975) Glutathlone: Metabolism values are probably due to different experimental condi- and Function (Raven Publ., New York). tions. 23. Gal., E. M. (1972) "Carbon fluorine compounds," Ciba Foun- A secQnd possible mitochondrial role of the glutathione dation Symposia (Elsevier, New York), pp. 77-93. carboxylic acid thioester enzyme system can be predicted from 24. Stipani, I., Prezioso, G., Genchi, G. & Palmieri, F. (1976) Boll. present results and may explain a special function of the un- Soc. Ital. Biol. Sper. 52, 1288-1293. 25. Meijer, A. J. & Tager, J. M. (1969) Biochim. Biophys. Acta 189, usually high cellular concentration of glutathione (22). The 136-139. .concentration of all free carboxylic acid substrates of the Krebs 26. Meyer, J. & Vignais, P. M. (1973) Biochim. Biophys. Acta 325, cycle can be directly regulated by the thioester synthetase- 375-384. hydrolase system, an effect that may profoundly influence 27. Uotila, L. (1973) Biochemistry 12,3938-3943. cellular metabolism. The existence of glutathione S-formyl and 28. Uotila, L. (1973) Biochemistry 12,3944-3951. succinyl thioesters in human liver has been demonstrated, but 29. Henriksen, T., Melo, T. B. & Saxebol, G. (1976) in Free Radicals in Biology, ed. Pryor, W. A. (Academic Press, New York), Vol. * Without mersalyl. 2, pp. 213-254. Downloaded by guest on October 2, 2021