Catalytic and Regulatory Properties of the Heavy Subunit of Rat ...

Vol. 268,No. 26,Issue of September 15,pp. 19675-19680, 1993 Printed in U.S.A.

THEJOURNALOF BIOLOGICAL CHEMISTRY 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc

Catalytic and Regulatory Propertiesof the Heavy Subunit of Rat Kidney y-Glutamylcysteine Synthetase* (Received for publication, April 7, 1993)

Chin-Shiou Huang, Long-Sen Chang, Mary E. Anderson, and Alton Meister From the Department of Biochemistry, Cornel1 University Medical College, New York, New York 10021

y-Glutamylcysteine synthetase (rat kidney), which Inhibition of the holoenzyme by GSH is competitive with recatalyzes the first step of GSH synthesis, can bedissoci- spect to glutamateand thus appears to involvethe glutamateated into subunits (M, 73,000 and 27,700) by native gel binding siteof the enzyme (9). Notably, however,the thiol group electrophoresis after treatment with dithiothreitol of GSH seems to be required for inhibition; thus, ophthalmic (DTT);the heavy subunit, which exhibits catalytic activ- acid (a GSH analog in which the thiol is replaced by a methyl ity and feedback inhibition byGSH (Seelig, G. E , Si- group) inhibits the holoenzyme only slightly. mondsen, R. p., and Meister,A. (1984)J. Biol. Chem. 259, The cDNA for the heavy subunit was isolated and found to 9345-9347), was cloned and sequenced (Yan, N., and Mei- code for a protein of 637 amino acid residues(M,72,614) (8). In ster, A. (1990) J. Biol. Chem. 265, 1588-1593). Here, the this work, we expressed the cDNA for the heavy subunit in cDNA for the heavysubunitwasexpressed in Es- Escherichia coli and purified the recombinant heavy subunit. cherichia coli, and the recombinant enzyme was sepa- The latter canbe distinguished fromE. coli y-glutamylcysteine rated fromE. coli y-glutamylcysteine synthetase and pusynthetase by makinguse of the fact that the kidney enzyme is rified. The recombinant enzyme and the isolated heavy markedly inhibitedby cystamine, whereasthe E. coli enzyme is subunit have muchloweraffinityforglutamateand higher sensitivity to GSH inhibition than the holoen- not (10).We have examined some catalytic and regulatory properties of the recombinant heavy subunit; these are similar to zyme, suggesting that the heavy subunit alone would the holoenzyme, b u t not be very active in vivo. A GSH analog, y-Glu-a-ami- those of the heavy subunit obtained from nobutyryl-Gly (ophthalmic acid), inhibits only slightly, are significantly different from those of the isolated holoenbut inhibits much more after treatment of the holoen- zyme. zyme with Dm. In contrast, ophthalmic acid inhibits the MATERIALSANDMETHODS recombinant and isolated heavy subunit enzymes substantially without DTT treatment. We conclude that (a) y-Glutamylcysteine Synthetase the light subunit has a regulatory function affecting the This activity was determined (4) by measuring the rate of formation affinity of the enzyme for glutamate and GSH and ( b ) of ADP spectrophotometrically using a coupled assay with pyruvate feedback inhibition byGSH involves reduction of the kinase and lactate dehydrogenase. The reaction mixture (1 ml) conenzyme and also competition between GSH and gluta- tained Tris-HC1 (100 mM,pH 8.2), sodium L-glutamate (10 mM), L-CXmate for the glutamate site. aminobutyrate (10 mM), magnesium chloride (20mM), disodium ATP (5 m),sodium phosphoenolpyruvate(2 mM), potassium chloride (150mM), , kinase (5 units; bovine heart type III), and NADH (0.2 m ~ ) pyruvate GSH, a widely distributed tripeptide thiol that has impor- lactate dehydrogenase(10 units; rabbit heart type 11).The reaction was tant functions in cellular metabolism and protection, is synthe- initiated by adding the enzyme, and the rate of decrease in absorbance at 340 nm was followedat 37 “C. One unit of enzyme activity is defined sized intracellularly by the consecutive actions o f y-glutamylas the amount that catalyzes the formation of 1 pmol of product%. cysteine and GSH synthetases (Reactions 1 a n d 2) (1-3). Specific activity is expressed as unitdmilligram of protein. Protein was determined by the method of Lowry et al. (11);bovine serum albumin L-Glutamate + L-cysteine + ATP a L-y-glutamyl-L-cysteine+ ADP + Pi was used as the standard. For the determination of K,,, values, the standard assay was used, with various substrate concentrations. The 1 REACTION standard assay was validated by carrying out the reactions with ~-[~~Slcysteine and measuring the rate of formation of the labeled dipepL-y-Glutamyl-L-cysteine+ glycine + ATP e GSH + ADP + Pi tide (4). In the studies in which the enzyme was treated with [35S]GSH, the holoenzyme (10nmol, 10 pl) was mixed with 1 ml of 20 mM [35S]GSH REACTION 2 (1pCi/mmol);after 1min, the solution was applied to a column (1.2 x 60 y-Glutamylcysteine synthetase isolated from rat kidney has a cm) of Sephadex G-50equilibrated with 50 mM sodium acetate buffer molecular weight of 100,000 and canbe dissociated intosub- (pH 5.0), and the protein was thus separated from the low molecular units with molecular weights of -73,000 and 27,700 (4-6). The weight peak. of the isolated enheavy subunit, isolated after dissociation Inhibition Studies zyme under nondenaturing conditions the in presence of 50 mM Inhibition by cystamine was carried out by incubating the enzyme dithiothreitol, exhibits all of the catalytic activity of the isocystamine (0.1mM) in Tris-HCl(100mM, pH 8.2) at 37 “C for5 min; GSH (7). The with lated enzyme and also feedback inhibition by a small portion of this solution was added tothe standardassay solution function of the light subunit, which is not itself enzymatically (10, 12, 13). Cystamine inhibits the rat kidney enzyme, but not the E. active, has not yet been established, and t h e possibility that it coli enzyme (10).Inactivation byL-buthionine (SJZ)-sulfoximinewas is not an integral part of the enzyme has been considered (8). carried out by preincubating the enzyme with 1 mM buthionine sulfoximine in a solution containing Tris-HCl(100m,pH 8.2),MgClz (20 mM), * This work was supported in part by National Institutes of Health and disodium ATP (5 mM) at 37 “C for 10 min; a small portion of this Grant 2-R37-DK-12034. The costs of publication of this article were solution was added to the standard assay solution (14). Inhibition by defrayed in part by the payment of page charges. This article must y-methylene glutamate was carried out by preincubation with this therefore be hereby marked “aduertisement” in accordance with 18 amino acid (10 mM) in Tris-HCl(100m,pH 8.2) containing MnCI, (0.25 U.S.C. Section 1734 solely to indicate this fact. mM) at 37 “C for 1 h (15). Inthe studies on inhibition by glutathione or

-

19675

19676

y-Glutamylcysteine Heavy Synthetase

Subunit

ophthalmic acid, the enzyme was added to standard assay mixtures that contained varying concentrations of these peptides.

SDS-Polyacrylamide Gel Electrophoresis This was performedby the methodof Laemmli (16). The holoenzyme was treated with varying concentrations of glutathione, immediately added to an equal volume of loading buffer (0.125M Tris-HC1 (pH 6.8), 4% SDS, and 20% glycerol), and then applied to a7.5% acrylamide gel. Electrophoresis was carried out at 120 V.

pRGCSH R a t GCS

H subunlt

Bacteria and Plasmids The bacterial hosts for growing M13 phage and pUC18 plasmid derivatives were E. coli XL1-Blue and HB101, respectively. The host for the expression experiments was E. coli BL211DE3) (17). DE3 is a A phage derivative that contains a DNA fragment inserted into the int gene that codes for T7 RNApolymerase;the lac U V 5 promoter controls the T7 polymerase and is induced by isopropyl-1-thio-B-D-galatopyranoside (18-22).

\

Ndel

TI promoter

Construction of Expression Plasmid for Heavy Subunit of y-Glutamylcysteine Synthetase

FIG.1. Structure of pRGCSH expression plasmid. The NdeIHindIII fragmentfrom pUC18/pGCSH (shaded box) coding for the

heavy subunit of y-glutamylcysteine synthetase was inserted into the GCS243, a 1.5-kilobase cDNA clone that contains thestart codon for expression vector pT7-7 immediately downstream of the T7 promoter the heavy subunit of y-glutamylcysteine synthetase, was cloned into (910). M13mp18 phage using the EcoRI restriction site. An NdeI site was created at the start codon by in vitro mutagenesis (Mutagene,Bio-Rad) TABLEI using 5'-AGCCCCATATGCGCGTCC-3' as a primer (23). The resulting Isolation of recombinant rat y-glutamylcysteinesynthetase mutant, GCS243A, was subcloned intopUC18. The full-lengthcDNAfor heavy subunit the heavy subunit was constructedby digesting pUC18/GCS243A and cDNA GCS258, which contains the 3'-end (8), with HindIII and BglII; Step Volumeactivity Specific Activity Protein the 1-kilobase fragment from GCS258 was ligated to pUC18/GCS243A ml mg unitslmg units t o give pUC18IGCSH. After amplification, pUC18/GCSH was cleaved 1.Extract" 55 1760 3.76 468 with NdeI and HindIII. A 2-kilobase full-length cDNA was purified and 1300 2.DEAE-Sephacel 120 115 11.3 ligated t o pT7-7 that was previously digested with NdeI and HindIII. 3.ATP-agarose 14.2 1350 3.7 385 (1030)* This full-length clone, pRGCSH, is immediately downstream (Fig. 1) from the T7 promoter (910).pRGCSH (1 ng) was used t o transform E. * From 9.8 g of packed cells, wet weight. coli BL21(DE3). * When assayed at 50mM L-glutamate.

Expression and Purification of Recombinant y-Glutamylcysteine Synthetase Heavy Subunit

MOPS (50 mM, pH 7.0) containing L-glutamate (5mM), MgClz (5 mM), and disodium ATP ( 1 mM). Fractions containing enzyme activity were pooled and concentrated to-5 ml using an Amicon concentrator witha Y"10 filter. The enzyme solution was dialyzed against 4 liters of imidazole buffer (10mM, pH 7.4 (with NaOH)) containingEDTA (1 m).

Ten colonies of E. coli BL21(DE3) containing pRGCSH were incubated in 10 ml of Luna broth containing ampicillin (0.2 mg/ml). The medium was changed every 1.5 h; the total incubation time was 5 h. This stock culture was added to 3 liters of Luria broth containing ampicillin (0.2 mgiml) and grown a t 37 "C until the absorbancea t 550 nm lsolation of Heavy Subunit from RatKidney Holoenzyme reached 0.8. Isopropyl-1-thio-P-D-galatopyranoside (1 mM) was added, and the incubation was continued for a n additional 5 h. The cells were The holoenzyme was prepared as described (4) and treated with 50 harvested and washed with0.85% NaCI. mM DTT at 4 "C for 1 h, followed by preparative polyacrylamide gel electrophoresis (7.5% gel) under nondenaturing conditions (7).The proPurification of Recombinant Heavy Subunit tein band corresponding to the heavy subunit, obtained -10% in yield, was excised and eluted in imidazolebuffer (10 mM, pH 7.4) containing All steps were carried outa t 4 "C. g glutamate (5 mM) and MgCl, (5 m)for 48 h. Step I-Cells (-10 g, wet weight) were suspended in 50 ml of Tris-

HCl (50 m ~ pH , 7.4) containing L-glutamate (5 mM) and MgCl, (5 mM) RESULTS (Buffer A). The cells were sonified (80 watts, 20 min), and the cellular RecombinantRatKidneyy-GlutamylcysteineSynthetase debris was removed by centrifugation (10,000 x g, 30 min). Step 2-The resulting supernatant was appliedto a DEAE-Sephacel Heauy Subunit-We were surprised to find that the recombicolumn (2.4x 10 cm) previously equilibrated with Buffer A. The column nant enzymeexhibitedrelatively low activity, i.e. a specific was washed with50 ml of Buffer A, and the enzyme was eluted with a activity of 385 units/mg (micromoles/hour/milligram), when linear gradient established between 400 ml of each Buffer A and Buffer measured in the standard assay, which contained 10 mM LA containing0.4 M NaCI. The flow rate was 50 mVh; fractions of 10 ml glutamate. However, we found that by increasing the L-glutawere collected. Two peaks containing enzyme activity were obtained. mate level to 50 mM in the assay, a specific activity of 1030 The peak containing the recombinant enzyme (fractions 33-40] was identified by its sensitivity to cystamine inhibition; the E. coli enzyme, units/mg was obtained (Table I). The corresponding values of which is not inhibited by cystamine, elutes in fractions 2628. The V, for the recombinant heavy subunit enzyme and the isorecombinant enzyme solution was concentrated t o -10 ml using an lated holoenzyme are -1700 and 1500 pmoWmg, respectively. Amicon concentrator equipped with a Y " 1 0 filter and dialyzed against The isolated recombinant heavy subunit was apparently ho4 liters of MOPS1 (50 m ~pH , 7.0) containing L-glutamate(5 m). Step 3-Solid MnClz was added toa final concentrationof 5 mM, and mogeneous on SDS gel electrophoresis and exhibited a molecthe enzyme solution was appliedt o a n ATP-agarose column (0.5 x 4 cm; ular weight of -73,000, which is close to that calculated from N-6, Sigma) previously equilibrated with starting buffer (MOPS(50 mM, the translated cDNA sequence (72,614) (8). NH,-terminal sepH 7.0) containing L-glutamate(5 m ~ and ) MnCl, (5m)). The flow rate quence analysis (15 residues) revealed that the sequence was was 30 ml/h; fractions of 2 ml were collected. The column was washed identical to thatdeduced from the cDNA sequence, except that with 24 ml of starting buffer and then with 24 ml of starting buffer it lacked the NH,-terminal methionine residue, which was precontaining NaCl(0.7M) and glycerol (10%). The enzyme was eluted with

sumably removed during synthesis in E. coli. Comparison of

The abbreviations used are: MOPS, 4-morpholinepropanesulfonic the isolated holoenzyme with the recombinant heavy subunit acid; DTT, dithiothreitol; HPLC, high pressure liquid chromatography. enzyme showed that both enzyme preparations were inhibited 1

19677

y-Glutamylcysteine Synthetase Heavy Subunit by cystamine (lo), y-methylene glutamate (13,and buthionine sulfoximine plus ATP (14), indicating that therecombinant enzyme, like the holoenzyme, has a thiol close to or at the active site (12) and that both enzymes use a catalytic pathway involving intermediate formation of enzyme-bound y-glutamyl phosphate (24). Apparent K,,, Values-The apparent K , values for L-a-aminobutyrate and L-cysteine for the recombinant heavy subunit, obtained from double-reciprocalplots, were similar to the corresponding values obtained for the isolated holoenzyme (Table 11).However, the recombinant heavy subunit enzyme exhibited a much higher apparent K , value for L-glutamate (18.2 m ~ compared with the isolated holoenzyme (1.4mM). This finding suggests that the affinity of the recombinant heavy subunit enzyme for glutamate is much lower than of the isolated holoenzyme. The possibility was considered that therecombinant enzyme expressed in E. coli might not be folded correctly as compared to the heavy subunit component of the isolated holoenzyme. As an approach to this issue, we separated the heavy subunit from the isolated holoenzyme by treating the holoenzyme with 50 mM dithiothreitol, followed by native gel electrophoresis (7). The apparent K , value for L-glutamate for this preparation of the enzyme (12.4mM) was not far from the value obtained for the recombinant heavy subunit enzyme (18.2mM). (That there is asmall difference in these values may probably be ascribed to a slight contamination of the heavy subunit isolated with dithiothreitol with the holoenzyme.) It thus appears that both the recombinant heavy subunit enzyme and the heavy subunit enzyme obtained by isolation from the holoenzyme exhibit significantly lower affinity for glutamate than found for the isolated holoenzyme. Feedback Inhibition Studies-The isolated holoenzyme is inhibited by GSH, and inhibition can be overcome in a competitive fashion by increasing the concentration of glutamate (9).In contrast, ophthalmic acid produces only little inhibition. However, we found that when the holoenzyme was treated with 50 mM dithiothreitol at 4 "C for 1h, its sensitivity to inhibition by ophthalmic acid increases substantially. Thus, 10 mM ophthalmic acid inhibited the holoenzyme by -5%, whereas the dithiothreitol-treated holoenzyme was inhibited by -25% (Fig. 2). Dithiothreitol(50 mM) alone had no significant effect on activity. This result suggests that feedback inhibition of the enzyme by GSH involves a reductive step, which can thus be carried out by GSH or dithiothreitol. It is of interest that inhibition of the dithiothreitol-treated enzyme by GSH was greater by a small but significant extent than that of the untreated holoenzyme (Fig. 2). Notably, both heavy subunit enzyme preparations are effectively inhibited by ophthalmic acid and GSH. Inhibition of the heavy subunit enzyme preparations was greater than that observed with the isolated holoenzyme. Forexample, the holoenzyme exhibited -65% of control activity, i.e. 35% inhibition, when assayed with 10 mM each L-glutamate and GSH, whereas TABLEI1 Apparent K,,, values for y-glutamylcysteine synthetase K". Enzyme preparation Ahtamate

Aminobutyrate L-Cl-

L-Cysteine

mM

Holoenzyme 1.2 Holoenzyme (DTT)" Heavy subunit (isolated) Recombinant0.2 heavy 0.8 subunit Recombinant heavy subunit a

1.4 2.8 12.4 18.2 18.2

Treated with 50 mM DTT at 4 "C for 1 h prior to assay.

0.2

100 I

,?c

eo 80

b

x

c

S

70

2-

K

)

60 HOLOENZYME

0

2

4

/

6

8

10

Inhibitor ( m M )

FIG.2. Inhibition of y-glutamylcysteine synthetase holoenzyme by GSH and ophthalmic acid. The isolated holoenzyme (5 units) and the holoenzyme that had been treated with 50 m~ DTT for 1 h at 4 "C were added to the standard assay solution containing 10 mM glutamate inthe presence of various concentrations of GSH or ophthalmic acid (OPH.1.

the heavy subunitpreparations showed-80% inhibition. Treatment of the recombinant heavy subunit enzyme with dithiothreitol did not significantly affect the sensitivity of the enzyme to inhibition by GSH or ophthalmic acid (Fig. 3). These observations indicate that the heavy subunit enzymes (recombinant and isolated) are highly sensitive to inhibition by GSH and that unlike the holoenzyme, inhibition does not require a reductive step. Inhibition of the dithiothreitol-treated holoenzyme by ophthalmic acid, like inhibition of the holoenzyme by GSH, was competitive with respect to glutamate (Fig. 4 and Table 111). Inhibition of the recombinant heavy subunit by ophthalmic acid was also competitive, with a similar apparent Kivalue (Table 111).However, the recombinant heavy subunit preparation was inhibited by GSH noncompetitively (Fig. 41, with a relatively low apparent Ki value (1.8 mM). This might reflect binding of GSH to both the glutamate and to another enzyme site, possibly the site for cysteine. When the GSH-treated recombinant heavy subunit preparation was dialyzed, full activity was restored. Reduction of Holoenzyme byGSH-When the isolated holoenzyme was treated with SDS and then subjected to SDS gel electrophoresis, -30% of the holoenzyme was dissociated into heavy and light subunits (Fig. 5). When GSH was added prior t o SDS gel electrophoresis, the amount of enzyme that dissociated increased significantly. Thus, after treatment for 1 min with GSH at 2, 6, and 10 mM, the respective extents of dissociation of the total enzyme present were 62, 80, and 88%. On the basis of the amount of GSH-dependent SDS-dissociable enzyme present, 50%dissociation required -1.5 m~ GSH (Fig. 6). DIT was less effective than GSH; treatment with 10 m~ DTT for 1h led to 50%dissociation. No dissociation was found when 10 mM ophthalmic acid was added in place of GSH. Reduction of the enzyme by treatment with GSH was neither prevented by addition of 10 mM L-glutamate nor affected by preincubation of the enzyme with ATP and buthionine sulfoximine. Studies in which reduction was carried out with [35SlGSH showedno significant labeling of the protein subsequently separated by gel filtration. Reduction of the enzyme by GSH is readily reversible; thus, dialysis of the GSH-reduced enzyme against Tris-HC1 (0.05 M, pH 8.2) containing 5 mM L-glutamate for 18 h at 5 "C led to restoration of its original form. A similar result was obtained after treating the GSHtreated enzyme with 10 mM hydrogen peroxide.In each of these

19678

y-Glutamylcysteine Subunit Heavy Synthetase

FIG.3. Inhibition of heavy subunit of y-glutamylcysteine synthetase by GSH and ophthalmic acid.The activities of the isolated heavy subunit (A) and the recombinant heavy subunit ( B )were determined in assay solutions containing 10 mM glutamate and various concentrations of GSH or ophthalmic acid (OPH. ). An experiment was also carried out with the recombinant heavy subunit and the recombinant heavy subunit that had been treated with 50 mM D'R' a t 4 "C for 1 h (B). 0

2

4

6

10

8

2

0

4

Inhibitor (mu)

Holo

-

8

0

s

10

7

8

Inhibitor (mM)

1

2

3

.r..

4

..

~

5

6

9

~

H-

L-

F'IG.

5. SDS gel electrophoresis of yglutamylcysteine synthe-

tase. The isolated holoenzyme (10 pg, 10 pl) was treated with 10 pl of

"is-HC1 (0.125 M, pH 6.8) containing 4% SDS and 20% glycerolin the presence of 0,0.5, 1,2,4,6,8, and 10mM GSH (lunes 1-8, respectively) for 1 min. The mixtures were immediately analyzed by SDS gel electrophoresis. The enzyme was also incubated with the same buffer containing 0.7 M 2-mercaptoethanol at 100 "C for 5 min (lune 9 ) (100% dissociation). H,heavy subunit; L,light subunit.

found no such separation after treatment of the holoenzyme with 10 m~ GSH. When the holoenzyme was treated with 10 m~ GSH or 50 m~ dithiothreitol and then subjected to HPLC gel filtration or standard gel filtration on Sephadex G-100, only a component corresponding to M, 100,000 was found, indicat[Glul-l [GI"]" ing no dissociation to form the separated heavy subunit. These Frc. 4. Inhibition of y-glutamylcysteine synthetase by GSH observations are in accord with the finding that treatment of and ophthalmic acid. Enzyme activity was determined by adding the the holoenzyme with 50 m~ dithiothreitol did not greatly inholoenzyme (A), the dithiothreitol-treated holoenzyme ( B ) ,or the re- crease the apparent K,,,value for glutamate; a large increase combinant heavy subunit (Cand D )(-5 units) to assay solutions containing varying glutamate concentrations with no added tripeptide (0) would be expected ifan appreciable amount of free heavy subor with 5 (0)or 10 (A) m~ GSH (A and C)or ophthalmic acid ( B andD). unit had been formed.Treatment with 50 m~ dithiothreitol did not greatly affect enzyme activity, whereas treatment with 10 TABLE I11 mM GSH inhibited it by -35% (Fig. 2). Cross-linking of the Apparent Ki values for glutathione and Ophthalmic acid holoenzyme (at different protein concentrations) with dimethyl Enzyme Inhibitor Ki a suberimidate (25) followed by SDS-polyacrylamide gel electrophoresis leads only to a species of M, -100,000 without the mM Holoenzyme* GSH 8.2 appearance of other protein species (6). Ophthalmic acid 11.4 Holoenzyme (DTT)**' These findings indicate that -70% of the enzyme is isolated Recombinant heavy subunitd 1.8 GSH in a form that can be reversibly reduced by GSH and that the Recombinant heavy subunit* Ophthalmic acid 12.5 subunits of this form as well as those of the form that does not a Apparent Kivalues calculated from double-reciprocal plots. have covalent intersubunit linkage are probably linked by hy* Competitive inhibition. drophobic interactions that persist aRer treatment of the enTreated with 50 mM DTT at 4 "C for 1h prior to assay. zyme with 10 m~ GSH. The available data indicate that both Noncompetitive inhibition. forms are active. The isolated holoenzyme has essentially the experiments, enzyme activity returned to the original level and same activity before and after treatment with dithiothreitol. as did the degree of dissociability on treatment with SDS. Complete inactivation of the isolated holoenzyme byincubation Although treatment of the holoenzyme with 50 mM dithio- with L-methionine (SJE)-sulfoximine,Mn2+,and ATP leads to threitol followed by native gel electrophoresis leads to partial binding of 1.00-1.05 mol of methionine sulfoximine phosphate/ separation of the heavy subunit from the holoenzyme (71, we mol of enzyme (26);a similar resultwas obtained with buthion-

y-Glutamylcysteine Subunit Heavy Synthetase

19679

would most likely besubstantially less active than the holoenzyme under physiological conditionsin which glutamate levels are -3 m.Thus, both heavy subunit preparationsexhibit very much lower affinity for glutamate than the isolated holoen0 zyme and are highly sensitive to inhibition by GSH. The findings support the view that thelight subunitplays a significant regulatory role under physiological conditions.2 This study also elucidates aspects of the feedback inhibition by GSH. The finding that ophthalmic acid, which inhibits the isolated holoenzyme only slightly, inhibits to a much greater extent after the holoenzyme is treated with dithiothreitol suggests that a reductive step isinvolved in inhibition by GSHand that this stepmakes the glutamate-binding site more accessible to interaction with tripeptides. Thus, GSH seems to act in two ways, the effects being synergistic. It is relevant to note that treatment of the enzyme with dithiothreitol led to slight but significant increase in inhibition by GSH (Fig. 2) and to an increase in the apparent K , value for glutamate (Table 11). Notably, the separated heavy subunit enzymes are not signifi2o o ~ ' ~ cantly affected~by dithiothreitol, again ~ ' interacsuggesting that 0 2 4 6 8 10 tion between heavy and light subunits is involved in the physiological inhibition by GSH. The reductive reaction involved in GSH ( m M ) FIG.6. Dissociation of y-glutamylcysteine synthetaseby SDS. inhibition by GSH appears to affect interactions of glutamate The experiment was performed as described for Fig. 5 . The protein in and tripeptides at the active site. Although detailed underthe bands corresponding to the heavy subunit on treatment with GSH standing of these phenomena requires further study, these was determined by densitometry. findings are consistent with and may be summarized by the tentative scheme givenin Fig. 7, according to which an oxidized state of the enzyme has a conformation of the heavy and light subunit combination that favors a glutamate-binding site with relatively high affinity and high specificity for glutamate. Interaction of this form of the holoenzyme with GSH reduces S HS intersubunit disulfide linkage; this is associated with conformational changes that favor a decrease in affinity for glutamate aswell as interaction of GSH (and ophthalmic acid) at the glutamate-binding site of the enzyme, thus permittingcompetitive inhibition. Thus, although inhibition by GSH is nonallosteric (91, it is greatly facilitated by a reductive event that seems to occur at anenzyme site different from the active site. Since FIG.7. Tentative scheme for oxidized and reduced forms of holoenzyme. The glutamate-binding site is indicated by the arrow. relatively low levels of GSH sufficeto reduce the intersubunit GSH also interacts a t this site as well as atanother site. The binding disulfide form of the isolated enzyme (Figs. 5 and 61, it would sites for glutamate, cysteine, and ATP are on the heavy subunit (H) (see appear that in vivoa substantialportion of the enzyme must be the text). L, light subunit. in a form in which the subunits are closely associated, but not disulfide-linked. Other studiesindicate that in vivothe enzyme ine sulfoximine (27). Complete inactivation of the isolated ho- is partially inhibited by GSH and that it is released from this type of inhibition when cellular levels of GSH are decreased (9). loenzyme by treatment with ~-2-amino-4-0~0-5-chloropentanoate (121, S-sulfocysteine (281, y-methylene glutamate The isolated holoenzyme is obtained in a state that can be (15),or cystamine (12) leads to stoichiometric binding of these partially inhibited by adding GSH. It may be concluded that active-site reagents to the enzyme. The enzyme formthat does reversible reduction of the enzyme by GSH and competitive not have intersubunit disulfide linkage has thus far been found interactions of glutamate and GSH at the active site are sigin all preparations of the holoenzyme. It was found in isolations nificant phenomena involved in regulation of the activity of the in which 2-mercaptoethanol was added to the isolation buffers enzyme. It has long been thought that GSH can affect the (usual procedure) as well as inisolations in which it was not (in activities of certain enzymes (29,301, and there has been recent which yields of enzyme were low). Attempts to separate two additional consideration of this idea (see, for example, Refs. forms of the holoenzyme by native gel electrophoresis, DEAE- 31-33). Although this study has not yet led to a detailed mechcellulose chromatography (HPLC), and chromatography on hy- anism, it indicates a way in which the cellular level of GSH can drophobic interaction columns were not successful. The sub- provide a signal that influences y-glutamylcysteine synthetase unitsare presumably synthesized in the thiol forms, and activity. oxidation might lead to various types of disulfides including Acknowledgment-We thank Dr. Stanley W. Tabor for the gift of the inter- and intrasubunit disulfides. Such forms might be propT7-7 vector and for very helpful advice. duced during enzyme isolation, but might occur in vivo;further study is needed. REFERENCES

I ' " " " " ' I

1

I

;i 'i

I

DISCUSSION

1. Dolphin, D., Poulson, R., and Avramovic, 0.(eds) (1989) Glutathione: Chemical, Biochemical, and Medical Aspects, Coenzymes and Cofactors, Vol. 111,

This study supportsthe conclusion (7) that theheavy subunit This conclusion is further verified by studies (Huang, C.-S., Anderof rat kidney y-glutamylcysteine synthetase contains the strucson, M. E., and Meister, A. (1993)J.Biol. Chern., in press) in which the tural requirements for enzymatic activity and also for feedback subunits were (1) coexpressed and (2) separately expressed and then inhibition by GSH, but indicates that theheavy subunit alone mixed.

~

19680

y-Glutamylcysteine Synthetase Heavy Subunit

Parts A and B, John Wiley & Sons, New York 2. Taniguchi, N., Taniguchi, N., Higashi, T., Sakamoto, Y., and Meister, A. (eds) (1989) Glutathione Centennial: Molecular Perspectives and Clinical Implications, Academic Press, Inc., San Diego, CA 3. Meister, A. J. (1988) J. Biol. Chem. 263,17205-17208 4. Seelig, G. F., and Meister, A. (1985) Methods Enzymol. 113, 379390 5. Richman, P. G. (1975) y-Glutamylcysteine Synthetase and Related Studies on the Metabolism of the y-Glutamyl Group. Ph.D thesis, Cornell University Graduate School of Medical Sciences 6. Sekura, R., and Meister, A. (1977) J. Biol. Chem. 252,2599-2605 7. Seelig, G . F., Simondsen, R. P., and Meister, A. (1984) J. Biol. Chem. 259, 9345-9347 8. Yan, N., and Meister, A. (1990) J. Biol. Chem. 288, 1586-1593 9. Richman, P. G., and Meister, A. (1975) J. Biol. Chem. 250, 1422-1426 10. Huang, C.-S., Moore, W., and Meister, A. (1988) Proc. Natl. Acad. Sci. U.S. A. 86,2464-2468 11. Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275 12. Beamer, R. L., Griffith, 0. W., Gass, J. D., Anderson, M. E., and Meister, A. (1980) J. Biol. Chem. 285,11732-11736 13. Seelig, G.F., and Meister, A. (1982) J. Biol. Chem. 267, 5092-5096 14. Griffith, 0.W., Anderson, M. E., and Meister, A. (1979) J. Biol. Chem. 254, 1205-1210 15. Simondsen, R. P., and Meister, A. (1986) J. Biol. Chem. 261, 17134-17137 16. Laemmli, U.K. (1970) Nature 227, 680-685

17. Studier, F. W., Rosenberg, A. H., Dunn, J. J., and Dubendorff, J. W. (1990) Methods Enzymol. 186, 60-89 18. Rosenberg, A. H., Lade, B. N., Chui, D.-S., Lin, S.-W., D u m , J. J.,and Studier, F. W. (1987) Gene (Amst.) 56, 125-135 19. Studier, F. W., and Moffatt, B. A. (1986) J. Mol. Biol. 189, 113-130 20. Tabor, C. W., and Tabor, H. (1987) J. Biol. Chem. 262, 16037-16040 21. O’Mahony, D. J., Wang, H. Y., McConnell, D. J., Jia,F., Zhou, S . W., and Qi. S. (1990) Gene (Amst.1 91,275-279 22. Tabor, S . W., and Richardson, C. C. (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 1074-1078 23. Kunkel, T. A. (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 488-492 24. Meister, A. (1992) ChemTracts: Biochem. Mol. Biol. 3, 75-106 25. Davies, G. E., and Stark, G. R. (1970) Proc. Natl. Acad. Sei. U.S.A. 66, 651-656 26. Richman, P. G., Orlowski, M., and Meister, A. (1973) J. Biol. Chem. 248, 6684-6690 27. Griffith, 0.W. (1982) J. Biol. Chem. 257, 13704-13712 28. Moore, W., Wiener, H. L., and Meister, A. (1987) J. Eiol. Chem. 262, 1677116777 29. Hopkins, F. G., and Dixon, M. (1922) J. Biol. Chem. 64,527-563 30. Guzman-Barron, E. S . (1951)Adu. Enzymol. 11,201-266 31. Gilbert, H. F. (199O)Adu.Enzymol. Relat. Areas Mol. Biol. 63, 6%172 32. Wells, W.W., Yang, Y., Gan, 2.-R., and Deits,T. L. (1993)Adu. Enzymol. Relat. Areas Mol. Biol. 68,149-202 33. Ziegler, D. M. (1985) Annu. Reu. Biochem. 54,305-329