Purification and characterization of glutathione reductase from Rhodospirillum rubrum

ARCHIVES

OF BIOCHEMISTRY

AND

BIOPHYSICS

Vol. 298, No. 1, October, pp. 247-253, 1992

Purification and Characterization of Glutathione Reductase from Rhodospirillum rubrum Carlos A. Libreros-Minotta,

Juan P. Pardo, Guillermo

Departamento de Bioquimica, Facultad de Medicina, Apartado Postal 70-159, 04510, M&co, D.F. M&co

Received February

Universidad

Academic

Press,

National

and Juan L. Rendbn’

Auto’noma de M&co,

27, 1992, and in revised form May 28, 1992

Glutathione reductase (NAD(P)H:GSSG oxidoreductase EC 1.6.4.2.) was purified 1160-fold to homogeneity from the nonsulfurous purple bacteria Rhodospirillum rubrum (wild type). Specific activity of the pure preparation was 102 U/mg. The enzyme displayed a typical flavoprotein absorption spectrum with maxima at 274,365, and 459 nm and an absorbance ratio A,80/A,,, of 7.6. The amino acid analysis revealed an unusually high content of glycine and arginine residues. Titration of the enzyme with 5,5’-dithiobis(2-nitrobenzoic acid) showed a total of two free thiol groups per subunit, one of which is made accesible only under denaturing conditions. An isoelectric point of 5.2 was found for the native enzyme. K, values, determined at pH 7.5, were 6.1 and 90 PM for NADPH and GSSG, respectively. NADH was about 2% as active as NADPH as an electron donor. The enzyme’s second choice in disulfide substrate was the mixed disulfide of coenzyme A and glutathione, for which the specific activity and K, values were 5.1 U/mg and 3.4 mM, respectively. A native molecular weight of 118,000 was found, while denaturing electrophoresis gave a value of 54,400 per subunit, thus suggesting that R. rubruna glutathione reductase exists as a dimeric protein. Other physicochemical constants of the enzyme, such as Stokes radius (4.2 nm) and sedimentation coefficient (5.71 S), were also consistent with a particle of 110,000. 0 1992

Mendoza-Hernbndez,

Inc.

Glutathione reductase (NAD(P)H:GSSG oxidoreductase EC 1.6.4.2.) is an almost ubiquitous flavoenzyme whose catalytic action maintains a high GSH/GSSG concentration ratio in cells (1). It is a member of the disulfide oxidoreductase family, and has been purified to homogeneity from a variety of sources . The studies carried out with crystals of human erythrocyte glutathione reductase 1 To whom correspondence

should be addressed.

0003-9861/92 $5.00 Copyright Q 1992 by Academic Press, All rights of reproduction in any form

have been particularly significant, since a considerable amount of information, concerning both mechanistic and structural aspects of the catalytic cycle, has been obtained (2-4). Results from such studies have revealed that the enzyme is constituted by two identical polypeptide chains (3), a trait that extends to the majority of glutathione reductases from other sources (5-9). Also, a 1:l stoichiometry between subunits and active sites has been found; it is interesting to note that both subunits contribute with residues essential to the formation of catalytic centers (2), thus rendering the dimeric state of the enzyme a prerequisite for its function. Nevertheless, although the enzyme has been purified to homogeneity from a variety of eukaryotic sources, information concerning glutathione reductase from prokaryotic sources is scarce, in spite of the great physiological diversity of the group. The purification and characterization of the enzyme from Escherichia coli has been reported (9), and the gene coding for this flavoenzyme has been cloned and its primary structure determined (10). However, in photosynthetic bacteria, homogeneous glutathione reductase has been obtained only from the cyanobacteria Anabaena sp. (5) and Spirulina maxima (11). In regard to Rhodospirillum rubrum, a member of the nonsulfurous purple bacteria (family Rhodospirillaceae), an earlier report showed some properties of glutathione reductase (12). In that work, an unusually low molecular weight of 63,000 was determined for the native enzyme, this figure being very close to the known subunit molecular weight of the enzyme from other sources (about 55,000), and thus suggesting a monomeric protein. Unfortunately, no information concerning the subunit composition was given as molecular properties were determined in a very raw preparation. We undertook the purification and characterization of R. rubrum glutathione reductase in order to gain insight into the properties of the enzyme from photosynthetic 247

Inc. reserved.

248

LIBREROS-MINOTTA

prokaryotes, as well as to elucidate the true oligomeric nature of the R. rubrum enzyme. MATERIALS AND METHODS Materials. All organic and biochemical reagents were obtained from Sigma Chemical Co. (St. Louis). Inorganic salts and buffers were of the highest quality available and were purchased from Merck de Mexico, S.A. All chemicals were used without further purification. Reverse-osmosis-purified water was employed in the preparation of solutions, except for amino acid analysis and isoelectric focusing, where deionized water was used. Biological material. R. rubrum (wild type) extracts from photoautotrophically grown cultures were a gift of Dr. Heliodoro Celis (Instituto de Fisiologia Celular, UNAM). Culture media and growth conditions have been described elsewhere (13). Cells were harvested by centrifugation at the late logarithmic phase of growth and resuspended at a ratio of 1O:l (v/w) in 25 mM Tris-HCI buffer (pH 7.5) containing DNAase (1 mg per 100 g bacteria). The cell suspension was sonicated for 2 min at maximum power in an ice bath, and the homogenate was centrifuged for 20 min at 20,OOOgin order to remove cell fragments. The supematant fraction was then ultracentrifuged for 80 min at 100,OOOg.The resulting chromatophore-free supernatant was frozen and stored until purification of the enzyme. Enzyme assays. Glutathione reductase activity was measured in a DU-70 visible-ultraviolet spectrophotometer by following the decrease in absorbance of NADPH at 340 nm. The standard assay mixture contained, in a final volume of 0.5 ml, 0.1 M potassium phosphate buffer (pH 7.5) plus 1 mM EDTA (buffer A). Enzyme aliquots were preincubated in the presence of NADPH for 1 min in order to trace the baseline, after which GSSG was added to start the reaction. Measurement of activity was carried out at 25°C. A unit of enzyme activity is defined as the amount of protein needed to oxidize 1 pmol NADPH per minute at pH 7.5 and 25°C. Kinetic parameters were derived from the initial rate data by nonlinear regression analysis, fitting the Michaelis-Menten rate equation as described in (14). Protein determination. Protein content was estimated according to the Bradford dye-binding assay (15) using purified S. maxima glutathione reductase as a standard, with a concentration previously determined by the dry weight method (16). Ultracentrifugation protein profiles were also determined by the Bradford technique. Amino acid analysis. The amino acid composition was performed at the Yale University Protein Chemistry Facility using a 7300 Beckman amino acid analyzer with a cation-exchange column. Before hydrolysis, the enzyme was treated as follows: a 500-pg sample was precipited with 5% trichloroacetic acid. Protein was recovered by centrifugation and the pellet resuspended in 0.2 M ammonium bicarbonate (pH 8) containing was then added and samples were incubated 8 M urea; 5 mM dithiothreitol at 30°C for 60 min. After this, the enzyme was treated with 10 mM iodoacetic acid at room temperature for an additional 60-min period, followed by exhaustive dialysis against deionized water (three changes of 10 h each). Hydrolyses were performed at 15O’C at 2, 4, and 6 h. Tryptophan content was determined spectrophotometrically after dissolving the protein in 6 M guanidine hydrochloride (17). Electrophoresis. Native electrophoresis at a constant acrylamide concentration (T = 0.08) was performed in 90 X 6-mm cylindrical gels prepared in buffer A. The same buffer system was used in both cathode and anode reservoirs. Electrophoreses were run for 7 h at a constant current of 5 mA per gel. Determination of the molecular weight under denaturing conditions, in the presence of both SDS’ and 2-mercaptoethanol, was performed

* Abbreviations focusing; DTNB,

used: SDS, sodium dodecyl sulfate; IEF, isoelectric 5,5’-dithiobis(2-nitrobenzoic acid).

ET AL. according to Weber and Osborn (18). Pore-gradient electrophoresis under nondenaturing conditions was carried out in 80 X 2.7-mm gel slabs prepared in 90 mM Tris/bO mM boric acid buffer (pH 8.35). A linear concentration gradient of acrylamide (4 to 30%) at a constant cross-linkage of 4% was used. Electrophoreses were run at 125 V for 15 h. For both native and denaturing condition gels were stained and unstained by conventional methods. Isoelectric point determination. Preparative isoelectric focusing was performed in a Sephadex IEF gel plate previously prepared by mixing an enzyme sample (100 pg) with both ampholytes 1:15 (pH range 3-10) and Sephadex beads. Runs were performed at 4’C on a horizontal electrophoresis apparatus (Multiphor, LKB) for 5 h. A constant power of 30 W was used. Immediately after this, the gel was fractioned and the pH and enzyme activity were determined for each fraction. Ultracentrifugation. The sedimenting behavior of the enzyme was analyzed by zonal ultracentrifugation under isokinetic conditions in a L5-65 Beckman preparative ultracentrifuge. Aliquots of the enzyme were layered on top of 5-ml linear sucrose density gradients (5 to 20%), previously prepared in buffer A and ultracentrifuged in a SW 50.1 rotor at 45,000 rpm (average rcf = 190,OOOg)for 13 h. Calibration proteins were run in parallel tubes. Following this, all tubes were punctured at the bottom and fractions of about 80 pl were collected. The sedimentation coefficient was estimated according to Martin and Ames (19), and the molecular weight was calculated from a log-log plot as described elsewhere (20). Gel filtration studies. Native molecular weights, as well as data for the determination of Stokes radius and diffusion coefficient, were obtained by gel filtration chromatography in a G-150 Sephadex column (1.6 X 63 cm) previously equilibrated in buffer A. Samples consisting of R. rubrum glutathione reductase and seven calibration proteins were applied onto the column and eluted at a flow rate of 10 ml/h. From the elution volume of the enzyme, molecular weight was estimated as described by Andrews (21), and Stokes radius was determined according to Laurent and Killander (22). Diffusion coefficient of the enzyme was calculated by plotting V./V,, against D20,W(23). Determination of the free sulfhydryl content. The number of free sulfhydryl groups of the enzyme was estimated with DTNB (5,5’-dithiobis(2-nitrobenzoic acid)) according to Ellman (24). A typical reaction mixture, in a final volume of 0.5 ml, contained 2 mM DTNB in 0.1 M Tris-HCl buffer (pH 8.0) plus 1 mM EDTA. To start the reaction, the enzyme was added (3-4 jtM final concentration) and the time-course of absorbance at 412 nm was followed. After absorbance reached a plateau, a 4 jd-aliquot of 10% SDS was added and the additional increase in absorbance was recorded. An extinction coefficient of 14,140 M-’ cm-’ at 412 nm for DTNB (25) was used in the calculations. Purification of enzyme. All buffer solutions used in the purification protocol were made with 1 mM of EDTA. With the exception of affinity chromatography, which was performed at room temperature, all operations were carried out at 4°C. R. rubrum glutathione reductase was purified following the procedure described by Rendon et al. (11) -with minor modifications. The chromatophore-free supernatant was brought to 30% saturation with solid ammonium sulfate under continuous stirring and the protein precipitate was discarded by centrifugation at 15,000g for 20 min; then, ammonium sulfate crystals were added to the supernatant solution until a 60% saturation was reached. After a 60-min period, the protein fraction was recovered as mentioned above and resuspended in a minimal volume with 25 mM imidazole-HCl buffer (pH 6.2). Ammonium sulfate was removed by exhaustive dialysis against the same imidazole buffer (in this step an inactive precipitate was formed and removed by centrifugation). The retentate was applied onto a DEAE-Sephacel column (2.6 X 10 cm) previously equilibrated in imidazole buffer. After exhaustive washing, the enzyme was eluted by means of a linear NaCl gradient (O0.5 M) prepared in imidazole buffer. Under these conditions, the enzyme eluted in a symmetrical activity peak at an ionic strength of 0.23 M. Fractions containing glutathione reductase activity were pooled, dialyzed against buffer A, and applied onto a 2’,5’-ADP-Sepharose 4B column

GLUTATHIONE

REDUCTASE

FROM

Rhodospirillum

249

rubrum

(1.6 X 1.5 cm) previously equilibrated in buffer A. The column was washed until no absorbance at 280 nm was detected and the enzyme was recovered by application of a linear NADPH gradient (O-50 ELM). Those fractions containing reductase activity were pooled, dialyzed exhaustively against buffer A in order to remove NADPH, and stored at -20°C for further use. Miscellaneous. All data reported in the present paper represent the average of at least three independent determinations. When necessary, data were fitted into a linear trace by the least-squares method. RESULTS

Purification of the enzyme. Table I shows the results from a representative purification procedure. The homogeneity of preparations was checked by electrophoresis. As can be seen in Fig. 1, a single protein band was obtained for the enzyme in the last purification step, under both native and denaturing conditions. As regards to specific activity of the pure enzyme, typical values were in the range of loo-105 U/mg. However, when protein content was calculated on the basis of flavin content, an increase in specific activity was obtained (Table I). As regards enzyme stability, no loss in reductase activity was detected after 3 months of storage at -20°C and lower. Amino acid composition and spectral properties. In Table II the amino acid content of R. rubrum glutathione reductase is shown. From these results, a partial specific volume of 0.724 cm3/g was calculated by summation of individual amino acid-partial specific volumes, as described by Cohn and Edsall (26). The oxidized enzyme displayed a typical flavoprotein absorption spectrum, with maxima at 274,365, and 459 nm (Fig. 2). The absorbance ratios of A274/A459,A280/A459,and A365/A459were 8.35, 7.6, and 0.96, respectively. Isoelectric focusing of the native enIsoelectric point. zyme gave an homogeneous peak with maximum activity at pH 5.2 (data not shown). Catalytic properties and substrate specificity. Preincubation of the enzyme in the presence of NADPH (200 PM) results in no inhibition; identical initial velocity data were obtained when the preincubation was extended up to 10 min. The Michaelis-Menten constants, obtained

s L

Volume (ml)

Step Chromatophore-free supernatant Ammonium sulfate

1169

(30-60%)

DEAE-Sephacel 2',5' ADP-Sepharose

4B

62 40 1.82

Procedure

0.2

0.3

0.4

0.5

0.6

I RELATIVE

MOBILITY

FIG. 1. Electrophoretic patterns of purified glutathione reductase. (A) Native gel electrophoresis. An enzyme sample (25 rg) was loaded on top of a native gel (T = 0.08, C = 0.04) previously prepared in buffer A. (B) SDS-polyacrylamide gel electrophoresis. An amount of 10 pg protein was loaded on top of a denaturing gel. Molecular weight markers were as follows: (a) transferrine (78,000), (b) bovine serum albumin (SS,OOO), (c) immunoglobulin heavy chain (53,000), (d) ovalbumin (43,000), (e) carbonic anhydrase (29,000), (f) immunoglobulin light chain (25,000) and(g) myoglobin (17,200). The arrow point the relative mobility of glutathione reductase. Other conditions were as described under Materials and Methods.

from the initial velocity patterns at pH 7.5, were 6.1 PM and 90 PM for NADPH and GSSG, respectively (results not shown). From the maximal velocity, and taking a molecular weight of 54,500 per subunit, a turnover number of 5600 min-’ was calculated. NADPH was the preferred electron donor; however, NADH was also recognized by the enzyme, although with much lower efficiency. Maximum activity, obtained with 400 PM NADH, was only 2% of that obtained with an identical concentration of NADPH at pH 7.5. Interestingly, the optimum pH of the reaction was the same for both NADH and NADPH (data not shown). As regards specificity toward its disulfide substrate, glutathione re-

TABLE Purification

I 0.1

of Rhodospirillum

I

rubrum Glutathione

Total protein (md

Specific activity W/md

2683

0.088

568 209 1.6

’ Amounts rated for 130 g of biomass, wet weight. b Specific activity in parenthesis was based on flavine content, determined

0.41 0.95 102 (157)b

spectrophotometrically.

Reductase” Purification factor (-fold)

1 4.7 10.8 1159

Yield (So)

100 99.2 84.3 68.9

250

LIBREROS-MINOTTA

ET AL.

TABLE II Amino

Acid Composition

of Glutathione

Reductase

from

Rhodospirillum rubrum Residues per subunit of molecular weight 53,660” 2h

4h

6h

Average

Nearest integer

47.07

46.53

29.48

27.90

46.63 31.23b

47 31

Ser

17.62

Glx Pro

36.96 21.73 82.19 59.02

15.40 36.63

46.28 26.07 13.74 37.46

Amino acid Asx Thr

Gb

Ala Val Met Be Leu Tyr

Phe Hi& LYS

Arg CYS Trp

43.03

12.78 29.15

19.46b

19

37.02

37 22 82 58 45’

22.29

22.29

22.10

81.70 57.83 43.94

81.58 57.81 44.72

81.82 58.22 43.90

12.60

12.50

12.63

13

33.61 10.26

30.26 33.24 10.20

30.53 33.53 10.25

29.98 33.46 10.24

31’ 33

15.16

15.20

15.40

15.25

6.83 16.05 38.78 4.05

10 15

6.96

6.90

6.90

7

16.39

16.35

16.26

16

38.76 3.89

38.44 3.86

38.66 3.93 5.32e

39 4 5

’ Excluding flavine. b Extrapolated to zero time. ’ The integer of the highest value was taken. d Taken as reference to calculate molecular weight. e Determined spectrophotometrically.

ductase from R. rubrum was also capable of recognizing and reducing the mixed disulfide of coenzyme A and glutathione (CoASSG). V,,, and K,,, values for the latter, at a fixed saturating concentration of NADPH, were 5.1 U/ mg and 3.4 mM, respectively. Cystine up to concentrations of 5 mM was not a substrate at all. Protein structure and hydrodynamic properties. When the enzyme was analyzed through a variety of techniques for molecular weight determination under native conditions, similar figures were consistently obtained. Thus, gel filtration chromatography provided a value of 122,500 _t 11,000 (Fig. 3A), while pore gradient electrophoresis gives a molecular weight of 116,000 + 9000 (results not shown). The slightly lower figure of 92,000 +- 7500 was obtained by zonal ultracentrifugation in sucrose density gradients. No oligomeric state other than a dimer was observed. On the other hand, denaturing electrophoresis revealed a molecular weight of 54,400 + 1100 (Fig. lB), about half the molecular weight of the native enzyme. This figure is also consistent with the 53,660 obtained from the amino acid composition of the enzyme (Table II). The sedimentation coefficient of the pure enzyme was estimated at 5.71 f 0.25s (Fig. 4). From the elution volume obtained with the Sephadex G-150-calibrated column, the Stokes radius of the enzyme was determined as 4.2 f 0.15 nm

WAVE LENGTH

(nm)

FIG. 2. Visible and ultraviolet spectra of oxidized glutathione reductase. Spectra were recorded at 25°C in buffer A. An enzyme concentration of 0.9 mg/ml was used.

(Fig. 3B), while a figure of 51.2 -t 2.7 pm2 set-’ for the diffusion coefficient was calculated from either a V,,V,, against D20,w (data not shown) or the Stokes-Einstein equation (27). The frictional coefficient ratio (flfO) was obtained from the equation given by Erlichman et al. (28). For a molecular weight of 109,000, a figure of 1.34 was found. Sulfhydryl group content. R. rubrum glutathione reductase contains four half-cystine residues per subunit (Table II). Titration of free SH groups in the native enzyme revealed 1 mol SH per mole of enzyme subunit. In the presence of SDS, however, a second free thiol group

0.5. - 0.7 ; Y

z. s

0.3 -

B I

-05

0.1 -

1, ’ 42

io.3 > 4.6 log

5.0 MW

5.4

2

3 STOKES

4 RADIUS

5 hd

FIG. 3. Determination of molecular weight (A) and Stokes radius (B) of glutathione reductase by gel filtration chromatography. Plots were traced from the elution volumes data obtained as described under Materials and Methods. Protein standards (with molecular weight and Stokes radius in bracket) are (a) myoglobin (17,200, 1.91 nm), (b) carbonic anhydrase (30,000,2.01 nm), (c) ovalbumin (43,000,2.83 nm), (d) transferrin (78,000,3.61 nm), (e) alkaline phosphatase (86,000,3.3 nm), (f) immunoglobulin G (156,000,5.23 nm), and (g) catalase (240,000,5.23 nm). The arrows point to the values corresponding to glutathione reductase.

GLUTATHIONE

E c

0.6 -

!g In

0.5 -

2

0.4-

s 2

0.3 -

E! 0

0.2 -

$

REDUCTASE

0.1 /P "IO

20

30 FRACTION

40

NUMBER

FIG. 4. Sedimentation profile of glutathione reductase based on enzyme activity. Samples were centrifuged in a swinging bucket rotor (SW 50.1) as described under Materials and Methods. Calibration proteins were run in either parallel tubes or mixed with the enzyme. Inset: estimation of the sedimentation coefficient of glutathione reductase. Calibration proteins, with their sedimentation coefficients (in bracket) are: (a) myoglobin (1.75 S), (b) ovalbumin (3.55 S), (c) immunoglobulin G (6.7 S), (d) glucosamine-6-phosphate deaminase (9.0 S), and (e) catalase (11.3 S). (0) Protein profile, (0) glutathione reductase activity.

reacted with DTNB (results not shown). Taken in conjunction with amino acid composition data, these results suggest that a total of two free sulfhydryl groups exist per subunit, one of which is unaccesible to solvent in the native enzyme. The remaining two half-cystines appear to be involved in a disulfide bond. In Effect of denaturing reagents on enzyme activity. order to gain insight into the structural and catalytic stability of the enzyme, the effect of denaturing reagents on the activity of R. rubrum glutathione reductase was studied. Figure 5 shows the relative activity at pH 7.5 as a function of either urea or guanidine hydrochloride concentration. The denaturing reagent concentrations at which enzyme activity is half of maximum were 0.24 and 1.55 M for guanidine hydrochloride and urea, respectively. Ultracentrifugation experiments, carried out at the two urea concentrations at which enzyme activity is half of maximum (1.5 M) or totally abolished (3 M), showed no change in sedimentation coefficient as compared to the untreated enzyme.

FROM

Rhodospirillum

251

rubrum

from a chromatophore-free supernatant with a protein content significantly lower than the whole homogenate. As can be seen in Table I, the largest contribution to overall purification was due, by far, to affinity chromatography. The relatively low specific activity obtained for the pure enzyme is noteworthy; although significantly higher as compared to the preparation obtained by Boll (12), the figure 102 U/mg is about half of that reported for the enzyme from other sources (5,6,8,9,11). Even the flavinbased specific activity of 157 U/mg is below values typical of other glutathione reductases. In principle, the presence of proteolytic activity in crude extracts and/or a loosely bound FAD could be invoked as a possible origin of the low activity found. However, neither purification from a fresh extract nor addition of exogenous FAD in the assay mixture had the result of increasing specific activity. Furthermore, the absorbance ratio A280/A459 of 7.6 for the pure enzyme, comparable to reported findings for the analogous enzyme from other sources (7-g), constitutes additional evidence that flavine coenzyme is tightly bound to protein. A number of reports concerning purified glutathione reductase with a relatively low specific activity are known (7, 29,30). Hence, the figure 102 U/mg for R. rubrum glutathione reductase can be considered an intrinsic property of the enzyme. The comparison of amino acid composition from the R. rubrum enzyme and other prokaryotic and eukaryotic glutathione reductases revealed some interesting differences. An unusually high content of glycine and arginine residues was found, while the number of lysine and cysteine residues was relatively low. The isoelectric point of

Gu-HCI

hl)

DISCUSSION

In this work, the three-step purification protocol followed for obtainment of pure glutathione reductase from the cyanobacterium S. maxima (11) has been successfully applied, with only minor modifications, to the case of the R. rubrum enzyme. Although, in principle, this fact would seem astonishing, we must take into account that the purification of R. rubrum glutathione reductase was started

UREA

(M)

FIG. 5. Inactivation profiles of glutathione reductase. Enzyme samples were incubated at pH 7.5 for 1 min in the presence of NADPH at the indicated concentration of either urea or guanidine hydrochloride; GSSG was then added to start the reaction. Activities are expressed as percentage of control assays performed in the absence of denaturing reagent. (0) Guanidine hydrochloride, (0) urea.

252

LIBREROS-MINOTTA

5.2 for the R. rubrum enzyme, relatively high when compared to that of other prokaryotic glutathione reductases (5, ll), could be correlated with the high content of arginine residues. In regard to cysteines, the four half-cystine residues found in the present work represent the lowest number of cysteines so far reported for a glutathione reductase. In the mammalian enzyme, numbers as high as 12 half-cystines per monomer have been reported (31). Unfortunately, little information is available concerning the redox state of cysteines in other glutathione reductases (3, 32, 33). In Spirulina sp. there are only two DTNBreacting groups, one of which is exposed exclusively in the presence of a denaturing reagent (33). This situation is identical to that found in the present work for R. rubrum glutathione reductase. In this sense, kinetic data previously available on the inhibition of the R. rubrum enzyme by sulfydryl reagents (12), suggest that enzyme activity is strongly dependent on SH groups. This information is consistent with the results reported in this paper, which reveal a total of two free sulfhydryls and two nonreactive cysteines. Thus, the existence of a disulfide bridge in the subunit of R. rubrum glutathione reductase is strongly suggested. As in other glutathione reductases, this could correspond with the redox disulfide of the active site. Although NADPH was preferred over NADH as an electron donor, the pH-dependence of the two rates of activity was essentially identical, reaching a maximum at pH 7.5. This contrasts with the behavior of the enzyme from other sources (7, 8, 29, 31), where the maximum activity with NADH is found at a more acid pH compared to that with NADPH. Concerning enzyme specificity toward its disulfide substrate, our results showed that R. rubrum glutathione reductase displays activity toward CoASSG, although with less efficiency than with GSSG. In cases where it has been assayed, CoASSG has been proven to be the second most preferred substrate of enzyme (8, 30, 31, 34). Table III shows a comparison of kinetic parameters for purified glutathione reductase from

ET AL.

various sources, with either GSSG or CoASSG as a substrate. With the exception of the cyanobacterium S. maxima, in which a high activity ratio was reported, no significant difference between the prokaryotic and eukaryotic enzyme was observed. Although no specific metabolic role for CoASSG has been proposed, an inhibitory effect of this compound on a number of enzymes is known (35 37). In E. coli cells subjected to a variety of metabolic conditions, an important fraction of the COA pool is converted to CoASSG (38). In R. rubrum, where the presence of both GSH and CoA has been demonstrated (39), the formation of CoASSG is therefore feasible. Consequently, the existence of CoASSG reductase activity in glutathione reductase could be explained as a mechanism by which the intracellular level of this compound is regulated. As regards the oligomeric nature of R. rubrum glutathione reductase, there was a significant discrepancy between our data and those reported by Boll (12). This author reported a molecular weight of 63,000 for the native enzyme at pH 7.5, while in the present work a native molecular weight above 100,000, determined by both gel filtration and pore-gradient electrophoresis, was consistently obtained. The relatively low 92,000 obtained by ultracentrifugation, as compared to the about 120,000 determined by other mass-transport techniques, can be adscribed to the frictional properties of the enzyme. These data, taken in conjunction with the figure of 54,400 obtained per subunit, strongly support the existence of a dimeric protein. Other hydrodynamic parameters, such as Stokes radius and diffusion coefficient, are also consistent with a particle of about 110,000 and are comparable to those reported for other dimeric glutathione reductases (5, 7, 29, 31). In conclusion, the present results do not support the existence of a monomeric glutathione reductase in R. rubrum. Finally, the relatively high sensitivity of our preparation to urea compared to other glutathione reductases is worth mentioning. While enzyme activity from the rabbit liver

TABLE

III

Kinetic Parameters of Glutathione Reductase from Both Prokaryotic and Eukaryotic Sources with either GSSG or CoASSG as Substrate Km(PM)

Escherichia coli Spirulinn maxima Rhodospirillum rubrum Saccharomyces cerevisiue Rat liver * Bovine ciliar body

V, (U/mid

GSSG

CoASSG

GSSG

57 120

1234 3300 3400 200 230 1010 1320

66.6 238 102 260 207 40 84

90

55 57 85 98

a Data for E. coli were obtained by us in a partially purified preparation. * Two distinct forms of glutathione reductase were reported in this tissue.

CoASSG

5.7 3.0 5.1 28.6 17.0 12.0 20.0

Reductase activity ratio (GSSG/CoASSG)

11.7 79 20 9.1

12.2 3.3 4.2

Ref.

This work

(34) (8) (30)

GLUTATHIONE

REDUCTASE

(40) and Phycomyces blakesleanus (29) enzymes is essentially unaltered by 8 M urea, the enzyme activity of R. rubrum glutathione reductase is completely abolished by experiments performed in 3 M urea. Ultracentrifugation the presence of urea revealed that inactivation is not correlated to changes in quaternary structure. In this sense, it has been demonstrated that minor changes in tertiary structure can lead to a considerable change in enzyme activity (41). Thus, the effect of both guanidine hydrochloride and urea on the activity of R. rubrum glutathione reductase can be explained as a consequence of local disturbances in enzyme structure, probably involving substrate and/or catalytic sites. REFERENCES 1. Schirmer, R. H., Krauth-Siegel, R. L., and Schulz, G. E. (1989) in Glutathione (Dolphin, D., AvramoviC, O., and Poulson, R., Eds.), part A, pp. 553-596, Wiley, New York. 2. Pai, E. F., and Schulz, G. E. (1983) J. BioZ. Chem. 258,1752-1757. 3. Thieme, R., Pai, E. F., Schirmer, Mol. BioZ. 152, 763-782. 4. Karplus,

R. H., and Schulz, G. E. (1981) J.

P. A., and Schulz, G. E. (1987) J. Mol. Biol. 195,701-729.

5. Serrano, A., Rivas, J., and Losada, M. (1984) J. Bacterial. 324. 6. Mavis, R. D., and Stellwagen, 814. I. Lopez-Barea, 8. Carlberg, 5480.

158,317-

E. (1968) J. BioZ. Chem. 243, 809-

J., and Lee, C. (1979) Eur. J. Biochem. 98, 487-499.

I., and Mannervik,

B. (1975) J. BioZ. Chem. 250, 5475-

9. Mata, A. M., Pinto, M. C., and Lopez-Barea, J. (1984) 2. Naturforsch. 39c, 908-915. 10. Greer, S., and Perham, R. N. (1986) Biochemistry,

25,2736-2742.

11. Rendon, J. L., CaIcagno, M., Mendoza-Her&ndez, G., and Ondarza, R. N. (1986) Arch. Biochem. Biophys. 248, 215-223. 12. Boll, M. (1969) Arch. Mikrobiol.

66, 374-388.

13. Cohen-Bazire, G., Sistrom, W. B., and Stainer, R. Y. (1957) J. Cell. Comp. Physiol. 49, 29-68. 14. Caceci, M. S., and Cacheris, W. P. (1984) Byte 9, 340-362. 15. Bradford,

M. M. (1976) Anal. Biochem. 72, 248-254.

FROM

Rhodospirillum

rubrum

253

16. Kupke, D. W., and Dorrier, T. E. (1978) in Methods in Enzymology (Hirs, C. H. W., and Timasheff, S. N., Eds.), Vol. 48, pp. 155-162, Academic Press, New York. 17. Edelhoch, H. (1967) Biochemistry 6, 1948-1954. 18. Weber, K., and Osborn, M. (1969) J. BioZ. Chem. 244,4406-4412. 19. Martin, R. G., and Ames, B. N. (1961) J. BioZ. Chem. 236, 13721379. 20. Rendon, J. L., and Calcagno, M. (1985) Experientia 41,382-383. 21. Andrews, P. (1965) Biochem. J. 96, 595-606. 22. Laurent, T. C., and Killander, J. (1964) J. Chromatogr. 14, 317330. 23. Kulbe, K. D., Bojanovski, M., and Lamprecht, W. (1975) Eur. J. Biochem. 52,239-254. 24. Ellman, G. L. (1959) Arch. Biochem. Biophys. 82, 70-77. 25. Collier, B. H. (1973) Anal. Biochem. 56, 310-311. 26. Cohn, E. J., and Edsall, J. T. (1943) Proteins, Amino Acids, and Peptides, pp. 370-381, Van Nostrand-Reinhold, New Jersey. 27. Tanford, C. (1961) Physical Chemistry of Macromolecules, pp. 346363, Wiley, New York. 28. Erlichman, J., Rubin, C. S., and Rosen, 0. M. (1973) J. BioZ. Chem. 248,7607-7609. 29. Montero, S., De Arriaga, D., and Soler, J. (1988) Biochim. Biophys. Acta 952, 56-66. 30. Ng, M. C., and Shichi, H. (1986) Exp. Eye Res. 43, 477-489. 31. Boggaram, V., Larson, K., and Mannervik, B. (1978) Biochem. Biophys. Acta 527,337-347. 32. Massey, V., and Williams, C. H. (1965) J. Biol. Chem. 240, 44704480. 33. Cui, J. Y., Wakabayashi, S., Wada, K., Fukuyama, K., and Matsubara, H. (1989) J. Biochem. 105, 390-394. 34. Carlberg, I., and Mannervik, B. (1977) Biochim. Biophys. Actu 484, 268-274. 35. Bees, W. C. H., and Loewen, P. C. (1979) Can. J. Biochem. 57,336345. 36. Brandwein, H. J., Lewicki, J. A., and Murad, F. (1981) J. BioZ. Chem. 256,2958-2962. 37. Gilbert, H. F. (1982) J. BioZ. Chem. 257, 12086-12091. 38. Loewen, P. C. (1978) Can. J. Biochem. 56, 753-759. 39. Fahey, R. C., Buschbacher, R. M., and Newton, G. L. (1987) J. Mol. EvoZ. 25, 81-88. 40. Zanetti, G. (1979) Arch. Biochem. Biophys. 198,241-246. 41. Yao, Q., Tian, M., and Tsou, C. L. (1984) Biochemistry 23, 27402744.