- Email: [email protected]
Expression of glyoxalase, glutathione peroxidase and glutathione S-transferase isoenzymes in different bovine tissues
Biochimica et Biophysica Acta, 994 (1989) 21-29 Elsevier
21
BBA33285
Expression of glyoxalase, glutathione peroxidase and glutattfione S-transferase isoenzymes in different bovine tissues John D. Hayes, Susan W. Milner and Simon W. Walker University Department of Clinical Chemistry, The Royal Infirmary, Edinburgh ( U.K.) (Received 12 August 1988)
Key words: Glutathione peroxidase; Glutathione S-transferase; Glyoxalase; (Bovine)
(1) The tissue-specific expression of various glutathione-dependent enzymes, including glutathione S-transferase (GST), glutathlone peroxidase and glyoxalase |, has been studied in bovine adrenals, brain, heart, kidney, liver, lung and spleen. Of the organs studied, liver was found to p.,~sess the greatest GST and glyoxalase I activity, and spleen the greatest glutathione peroxidase activity. The adrenals contained large amounts of these glutathione-dependent enzymes, but significant differences were observed between the cortex and medulla. (2) GST and glyoxalase | activity were isolated by $-hexyiglutathione affinity chromatography. Glyoxalase | was found in all the organs examined, but GST exhibited marked tissue-specific expression. (3) The or, ~t and ~t classes of GST (i.e., those that comprise respectively Ya/Yc, Y b / Y n and Yf subunits) were all identified in bovine tissues. However, the Ya and Yc subunits of the a class GST were not co-ordinately regulated nor were the Yb and Yn subunits of the /t class GST. (4) Bovine Ya subunits (25.5-25.7 kDa) were detected in the adrenal, liver and kidney, but not in brain, heart, lung or spleen. The Yc subunit (26.4 kDa) was expressed in all those organs which expressed the Ya subunit, but was also found in lung. The # class Yb (27.0 kDa) and Yn (26.1 kDa) subunits were present in all organs; however, brain, lung and spleen contained significantly more Yn than Yb type subunits. The ¢t class Yf subunit (24.8 kDa) was detected in large amounts in the adrenals, brain, heart, lung and spleen, but not in kidney or liver. (5) Gradient affinity elution of S-hexylglutathioneSepharose showed that the bovine proteins that bind to this matrix elute in the order Ya/Yc, Yf, Yb/Yn and glyoxalase I. (6) In condusion, the present investigation has shown that bovine GST are much more complex than previously supposed; Asaoka (J. Biochem. 95 (1984) 685-696) reported the purification of # class GST but neither a nor ~t class GST were isolated.
Introduction
Glutathione is the major intracellular non-protein thiol; it exists in reduced (GSH) and oxidised (GSSG) forms. In addition to its participation in normal metabolic processes, GSH helps protect against damage caused by free radicals and plays several roles in detoxication of xenobiotics [1]. Important GSH-dependent detoxication enzymes include glyoxalase, glutathione $-transferase and glutathione peroxidase. The glyoxalase system uses GSH to protect the cell against 2-oxoaldehydes. This system involves two enzymes, glyoxalases I and II. Glyoxalase I catalyses the
Abbreviations: CDNB, 1-chloro-2,4-dinitrobenzene; SDS-PAGE, sodium dodecyl sulphate-polyacrylamide gel electrophoresis. Correspondence: J.D. Hayes, University Department of Clinical Chemistry, The Royal Infirmary, Edinburgh, EH3 9YW, U.K.
formation of a thiol ester between thiolesterase GSH and the 2-oxoaldehyde. Glyoxalase II is a thiolesterase; its hydrolytic activity results in the formation of a 2-hydroxy acid and regeneration of GSH [2]. Glutathione S-transferase (GST) forms a thiol-ether linkage between GSH and a large spectrum of electrophiles including herbicides, insecticides, chemotherapeutic drugs and carcinogens [3-5]. This conjugation reaction is the first step in mercapturic acid synthesis [6]. GST exists in multiple forms; cytosolic GST are encoded by three multi-gene families, designated by Mannervik et al. [7] as a, # and ~r. These families display distinct catalytic properties and play separate detoxication roles. Glutathione peroxidase uses GSH to protect against oxidative damage caused by oxygen radicals. In this reaction peroxides are reduced to alcohols and GSH is oxidised to GSSG [8]. Two types of glutathione peroxidase exist; these differ in their requirement for selenium. The selenium-dependent enzyme (GPx) is ac-
0167-4838/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
22 tive with both hydrogen peroxide and organic hydroperoxides, (e.g. cumene hydroperoxide). The selenium-independent glutathione peroxidase also possesses GST activity and is active with organic hydroperoxides, but not with hydrogen peroxide. This selenium-independent activity is principally expressed by a class GST. The level of these three glutathione-requiring enzymes within cells probably reflect their ability to deal with oxidative stress and/or cytotoxic insult. Indeed, in the case of GST, differences in enzyme levels have been correlated with drug resistance [3]. Despite the pharmacological and toxicological consequences that result from alterations in the intracellular levels of GSH-requiring enzymes, the biological control of these enzymes has been little investigated [9]. Marked tissue-specific expression of GSH-requiring enzymes is known to occur, and is presumed to reflect differences in the metabolic functions and physiological stresses encountered by different organs. The differences that have been described may partly explain the intrinsic resistance of certain :ells to cytotoxic drugs and oxidative damage. The synthesis of steroid hormones by the adrenal cortex results in the production of free radicals, but little is known about the defence mechanisms the adrenal cortex has evolved to compensate for this unique form of oxidative stress. Most work describing the tissuespecific expression of GSH-metabolising enzymes has so far employed the rat as the model system [10-16]. Although the rat is often used to study the effects of diet and the administration of xenobiotics, its size limits studies of the control of GSH-metabolising enzymes in steroidogenic tissues. It is also essential to study the adrenal cortex and medulla separately since they are both metabolically and embryologically distinct. The physical separation of these two tissues is not easy in the rat but is much simpler if bovine adrenal glands are used as the tissue source. A primary cell culture derived from bovine adrenal cortex has been established in this department [17,18]; it provides an ideal model for studying the biological control of GSH-metabolising enzymes. However, these enzymes have not been well characterised in bovine organs. For example, it is unclear whether a and r, classes of GST are expressed in this species. To date, GST have only been studied in bovine liver, from which organ Asaoka [19] isolated two closely si~nilar enzymes, more recently identified as members of the/~ class of GST [7]. The presence o f ~ ~ ~, class GST in extrahepatic bovine tissues has not yet been described. It is important to define which enzymes are expressed in bovine organs before commencing a longerterm study of the biological control of GSH-metabolising enzymes, using primary cell cultures derived from the adrenal cortex and other organs. As tissue differences in the content of GSH-requiring enzymes may
reflect their embryological origin, organs of ecto-, mesoand endodermal origin were chosen for the present study. Adrenal medulla and brain were chosen as examples of organs of ectodermal origin, and liver and lung as examples or organs of endodermal origin. The remaining four organs studied (adrenal cortex, heart, kidney and spleen) are all of mesodermal origin. The GST and glyoxalase I subunits present in different organs have been isolated by affinity chromatography on Shexylglutathione-Sepharose to enable direct comparison of their electrophoretic properties.
Materials and Methods
Chemicals These were all of analytical grade and readily available commercially. The S-hexylglutathione-Sepharose 6B was made by the method of Mannervik and Guthenberg [201.
Preparation of organ samples Bovine organs were obtained from a local slaughterhouse. The organs, obtained within 30 min of death, were transported to the department on ice. Except for the adrenal glands, organs were immediately stored at - 8 5 ° C , until processed. Adrenal cortex and medulla were manually dissected from one another before being stored separately at - 85 o C. Homogenates (1:4 (w/v)) of bovine tissues were prepared in ice-cold 50 mM-Tris-HC1 buffer (pH 7.8) that contained 200 mM-NaCI (buffer A). The 100 000 × g supernatant fraction (cytosol) from the organ extracts was prepared at 4 ° C using standard centrifuge techniques. The cytosol (300 ml) was filtered through a plug of glass wool, to remove lipid, and dialysed for 16 h at 4°C against 4 litres of buffer A. Following dialysis, the cytosol from each organ was re-centrifuged (15 000 x g, 30 rain, 4°C) to remove precipitated material; the supernatant was either subjected to chromatography immediately or stored ( - 8 5 o C) for later analysis.
Isolation of glutathione S-transferases and glyoxalase These enzymes were prepared from each of the cytosols by affinity chromatography on S-hexylglutathioneSepharose 6B. Columns (1.6 × 15 cm) of the affinity matrix were equilibrated at 4°C with buffer A. Dialysed cytosol (400-500 ml; 0.6-2~55 g of protein) was applied to these columns and the non-specifically adsorbed protein removed by washing the affinity matrix with at least 400 ml of buffer A. Specifically bound material was eluted using two different methods. The first method, used for the Western blot study, entailed a single-step elution of adsorbed material with a solution of 5 mM S-hexylglutathione in buffer A (as described in Ref. 20). The second method, used to prepare hepatic and adrenal GST, employed gradient elution of the
23 affinity matrix in two steps with 0-0.25 mM Shexylglutathione followed by 0.25-5.0 mM S-hexylglutathione [21]. The material (20 mi; 3-20 mg of protein) that was eluted from S-hexylglutathione-Sepharose 6B in a single step was dialysed at 4 ° C against 2 litres of buffer A before storage at - 85 o C prior to Western blot analysis.
TABLE I
Activities of GSH-dependent enzymes in the cytosolic fraction from various bovine tissues Values given are #mol of product formed per min per mg protein
Organ Enzyme and protein determination All enzyme assays were performed at 37 ° C. Glyoxalase I activity was measured by the method of Racker [22], using methylglyoxal as substrate. Glutathione peroxidase activities, towards both hydrogen peroxide and cumene hydroperoxide, were determined by the coupled assay system of Reddy et al. [23]. Glutathione S-transferase activities were assayed by the methods described by Habig and Jakoby [24]. Protein concentrations were measured by the dye-binding method of Bradford [25]. The 1-chloro-2,4-dinitrobenzene, cumene hydroperoxide, hydrogen peroxide and Coomassie blue G-250 protein assays were all carried out using a centrifugal analyser; all other assays were carried out manually. The assays were performed in triplicate or quintuplicate and the mean value was calculated and used for further calculations. The precision of these assays has been described before [21,26,~.7]. To avoid inter-batch analytical variation from contributing towards observed catalytic differences in the bovine organs the enzyme assays (for a given substrate) were all performed on the same day using the same stock solutions. Electrophoretic methods The SDS-polyacrylamide gel electrophoresis (SDSPAGE) and Western blotting techniques have been described elsewhere [11,28]. Rat GST were employed as M r standards for the SDS-PAGE and Western blot analyses; purified GST from rat fiver and rat lung provided Ya, Yb and Yc subunits and Yf, Yb and Yc subunits, respectively. The majority of SDS-PAGE analyses performed employed a 12~ polyacrylamide resolving gel containing C Bis 2.6~. However, during attempts to improve the resolution of Ya and Yc subunits, a 15% polyacrylamide resolving gel containing low levels of cross-linker (C ais 0.6~) was used [28]. Results
Cytosolic activity of GSH-requiring enzymes Significant glyoxalase I, GST and glutathione peroxidase activities were found in all the bovine organs studied, but the relative amounts varied considerably. The range of specific activities for glyoxalase I calculated in the different organs was notably less than the range of activities for GST and glutathione peroxidase. Table I shows that liver, spleen and kidney possessed highest glyoxalase activity. The fiver, adrenal cortex and
Heart Spleen Brain Liver Cortex Medulla Lung Kidney
Substrate:
Enzymeactivity glyoxalase GST GSH-peroxidase methyl CDNB cumene hydrogen glyoxal hydro- peroxide peroxide 0.87 0.10 0.07 0.05 1.50 0.20 0.71 0.56 0.88 0.08 0.07 0.09 1.97 0.37 0.21 0.01 0.77 0.23 0.28 0.18 0.58 0.07 0.16 0.15 0.47 0.23 0.20 0.08 1.27 0.04 0.27 0.07
lung had highest GST activity towards 1-chloro~2,4-dinitrobenzene (CDNB). Spleen and adrenal cortex had high glutathione peroxidase activity towards both hydrogen peroxide and cumene hydroperoxide. By contrast, both liver and kidney were found to exhibit high peroxidase activity towards cumene hydroperoxide but not towards hydrogen peroxide. Comparisons between GST activity and the two peroxidase activities allowed an estimate of the ~mount of selenium-dependent glutathione peroxidase in the different organs as the a class GST subunits (i.e., Ya and Yc) are active with cumene hydroperoxide but not with hydrogen peroxide. For example, liver cytosol has high CDNB-GSH-conjugating activity and moderate peroxidase activity for cumene hydroperoxide, but little activity for hydrogen peroxide, which suggests that this organ expresses little selenium-dependent glutathione peroxidase activity. Adrenal cortex and medulla exhibit similar peroxidase activities towards hydrogen peroxide, but the data in Table I indicate that the medulla contains considerably less selenium-independent glutathione peroxidase activity than the cortex as the activities towards both cumene hydroperoxide and CDNB are significantly lower in the medulla than in the cortex. Activity of affinity-purified enzymes S-Hexylglutathione-Sepharose 6B affinity chromatography was exploited to isolate glyoxalase I and GST from the various organs to allow the molecular basis for the tissue differences in enzyme activities to the explored; the selenium-dependent form of glutathione peroxidase does not bind to this matrix. The proportion of cytosolic protein that specifically bound to the affinity matrix varied between 0.2 and 1.0% (Table II). The percentage of GST activity towards CDNB that was absorbed to the S-hexylglutathione affinity column from the various organ cytosols was as
24 TABLE II Analysis of affinity-purified GST using model substrates The coefficient of variation obtained using these assays is as follows: p-nitrobenzyl chloride, 10.5%; ethacrynic acid 8.5~; trans-4-phenyl-3-butene2-one, 15~; 1,2-dichloro-4-nitrobenzene, 14.5%; cumene hydroperoxide, 6.5~; 1-chloro-2,4-dinitrobenzene, 6.0%; AS-androstene-3,17-dione, 9.0%. The CVs are higher for trans-4-phenyl-3-butene-2-one and 1,2-dichloro-4-nitrobenzene than normally found in this laboratory owing to the low :~pecfic activities of the bovine GST pools for those substrates.Abbreviations: n.d., not detected; p-NBC, para-nitrober~zyl chloride; EA, ethacrynic acid; t-PBO, trans-4-phenyl-3-butene-2-one; DCNB, 1,2-dichloro-4-nitrobenzene, CuOOH, cumene hydroperoxide; CDNB, 1-chloro2,4-dinitrobenzene; ASADD, AS-androstene-3,17-dione. Tissue
Heart Spleen Brain Liver Adrenal cortex Adrenal medulla Lung Kidney
g of protein applied to the affinity matrix
% protein bound to affinity column
GST substrate (/~mol of product formed per min per mg protein) p-NBC
EA
t-PBO
DCNB
CuOOH
CDNB
AS-ADD
1.43 0.64 0.61 2.55 1.92 2.19 1.48 2.34
0,9 0.2 0.2 1.0 0.8 0.9 0.8 0.9
0.73 0.71 0.50 0.38 0.27 0.61 0.53 0.21
0.78 0.30 0.72 0.07 0.11 0.15 0.11 0.08
n.d. 0.01 0.005 0.02 0.01 0.03 0.01 0.02
0.05 n.d. 0.01 0.003 0.001 n.d. n.d. n.d.
n.d. 0.24 0.09 4.17 1.76 0.43 9.20 0.20
12.43 10.89 11.89 7.80 6.15 9.50 17.27 6.37
0.11 0.04 0.04 0.07 0.18 0.11 0.07 0.16
follows: adrenal cortex, 85%; adrenal medulla, 88%; brain, 90%; heart, 96%; kidney, 67%; fiver, 79%; lung, 77%; spleen, 85%. The enzyme activities of these affinity-purified pools were examined with the range of GST substrates used to discriminate between the isoenzymes in rat and man (for a review, see Ref. 4). The results (Table II) reveal marked inter-organ differences in specific GST activity for a given substrate; with p-nitrobenzyl chloride and 1-chloro-2,4-dinitrobenzene, a 3-fold range in specific activities was observed, whereas with cumene hydroperoxide, differences in specific activity of at least 100-fold were noted between organs. The peroxidase activity towards cumene hydroperoxide is presumed to be due entirely to the selenium-independent forms since no activity towards hydrogen peroxide was detected. Differences were observed in the spectrum of GST substrates that the organs can metabolise. For example, heart and brain possess high activity for ethacrynic acid but tittle activity for cumene hydroperoxide, whereas liver and lung express high activity for cumene hydroperoxide but tittle activity for ethacrynic acid. Of all the organs studied, adrenal cortex and kidney expressed the highest activity for androstene-3,17-dione, but possessed markedly different activities for cumene hydroperoxide; in the S-hexylglutathione affinity-purified enzyme pools from human and mouse liver, these two substrates are specific for the Ya subunit. A similar divergence in these two activities was observed with liver and lung.
SDS /polyacrylamide.gel electrophoresis The major polypeptides that the affinity-purified pools comprise have mobilities during SDS-PAGE similar to those of rat GST subunits. Using this method, at
least seven subunits of different molecular mass can be resolved. Assuming that rat Ya, Yb, Yc and Yf subunits have molecular masses of 25.5, 26.3, 27.5 and 24.8 kDa, respectively [26], then bovine affinity-purified polypeptides have apparent masses of 24.0-27.0 kDa in gels comprising 0.32% (w/v) N,N-methylenebisacrylamide. T',e relative electrophoretic mobilities of these polypeptides were found to vary according to the amount of cross-tinker in the resolving gel. When electrophoresis was performed in gels composed of 0.32% (w/v) N, Nmethylenebisacrylamide, the bovine Yf-type subunit was found to migrate faster than the Ya-type subunit. However, the converse was true when the procedure was conducted in gels comprising 0.09% (w/v) N, N-methylenebisacrylamide. Fig. 1 shows this cross-tinker-dependent electrophoretic mobility, and demonstrates clear differences in the subunits expressed by the various organs that bind to the S-hexylglutathione affinity matrix.
Immunoblotting of glyoxalase I The protein that bound to S-hexylglutathione-Sepharose 6B was probed with antiserum raised against mouse glyoxalase I. The blots obtained showed that glyoxalase I was present in all bovine organs but no obvious tissue specific differences in the amount of this enzyme were detected (data not shown).
Immunoblotting of GST subunits Western blotting experiments were carried out using antisera raised against rat and human enzymes of the a (Ya/Yc), # (Yb/Yn) and ~r (Yf) classes, to investigate the organ-specific differences in GST content (Table III). Fig. 2 shows that the Yf subunit is present in significant amounts in all the organs examined except
25 1
2
3
~
5
6
7
small amounts in all organs except heart. The levels of the Yn subunit were more variable than Yb, with Yn being expressed in greatest amounts in spleen, brain and lung. The Yc subunit was expressed in bovine liver and lung but was not readily demonstrated in other organs. The bovine Yc subunit cross-rezcted with antisera raised against the rat Ya subunit but not the human Ya subunit (i.e., from GST BIB 1). Antisera against the rat Yk subunit, which is a member of the a class of GST [11], failed to react with any of the bovine GST. Identification of bovine Yf by Western blot analysis was found to correlate with the GST ethacrynic acid assays in Table If. Similarly, the blotting data for the Yc subunit correlates with the cumene hydroperoxide assays (Table II). Blots for Ya are also in broad agreement in all organs, except heart, with the A%androstene3,17-dione assays. No activity was identified that segregated with the Yb/Yn subunits, and it is assumed that the appropriate substrate has yet to be identified.
8
A
Yc ¥o yf /
.....
Yc Yb Yf Ya
w
Resolution Of glutathione S-transferase subunits and glyoxalase I by gradient elution of S-hexylglutathioneSepharose
C 0"6%
CBis
2"6% Yb
j ~ y c
Yb
G ~
J
~
Yh
•q - - - ~ YC, Y n
~
~-------- Y 1
Yo. Y I
*V¢
Fig. 1. Electrophoretic analysis of glutathione S-transferases purified by affinity chromatography. SDS-polyacrylamide gel electrophoresis was performed as described in the text. Panel A shows electrophoresis using a 12% (w/v) polyacrylamide resolving gel containing Cais 2.6% (w/w). Panel B shows electrophoresis of the same samples performed using a 15% (w/v) polyacrylamide resolving gel containing Cais 0.6% (w/w). In both instances the purified pools were loaded in the same order as follows: 1, heart; 2, spleen; 3, brain; 4, liver; 5, adrenal cortex; 6, adrenal medulla; 7, lung; 8, kidney. The positions of the rat Ya (25.5 kDa), Yb (26.3 kDa), Yc (27.5 kDa) and Yf (24.8 kDa) subunits are shown. Panel C shows a line diagram depicting the relative mobilities of the bovine GST subunits Ya, Yb, Yc, Yf, YI and Yn in polyacrylamide-gels containing Cnis 0.6% (w/w) ,nd Cni s 2.6% (w/w). The most anodal band, designated G, represents glyoxatase I.
liver and kidney. The Ya-type subunit was found in liver, adrenal cortex, adrenal medulla and kidney, but not in heart, spleen, brain or lung. However, the Ya-type subunit was resolved into two components (25.5 kDa and 25.7 kDa) in polyacrylamide-gels containing larger amounts of cross-linker, C,is 2.6%, but not in gels that incorporated lesser amounts of cross-linker, C Bis 0.6%. Antisera to the # class Yb GST, which crossreact with both Yb and Yn subunits, showed Yb to be present in
A detailed examination of the bovine GST subunits in the adrenal cortex and the liver was performed by eluting the enzymes which bound to S-hexylglutathione-Sepharose 6B with a gradient of counter-ligand. Comparison between the S-hexylglutathione-Sepharose 6B elution profiles obtained from these two organs revealed marked differences (Fig. 3). The adrenal cortex produced a single major peak of GST activity which eluted at a S-hexylglutathione concentration of 0.060 mmol/litre. By contrast, liver produced three major peaks of GST activity which eluted at S-hexylglutathione concentrations of 0.048, 0.075 and 0.130 mmol/litre. Both organs also yielded a protein-containing peak that eluted from the affinity column at 1.10 mmol/litre S-hexylglutathione; this protein lacked GST activity. Fractions from the chromatograms from adrenal cortex and liver were combined into 5 pools, as shown in Fig. 3, and the pooled material subjected to SDSPAGE (Fig. 4). Western blotting (data not shown) showed that the polypeptides were distributed as follows: Ya, in pools 1 and 2; Yb, in pools 2, 3 and sometimes 4; Yc, in pool 1; Yf, in pool 2; Yn in pool 3, 4 and sometimes 5; glyoxalase I, pool 5. Western blotting revealed quantitative differences in the subunit content of these pools. For example, the Yf subunit was not detected in pools obtained from bovine liver, whereas adrenal cortex did not contain as much Yb or Yc as liver. These data indicated that the polypeptides were eluted from the affinity matrix in the order Yc, Ya, Yf, Yb, Yn and, lastly, glyoxalase I at S-hexylglutathione concentrations of about 0.01-0.06, 0.01-0.10,
26 TABLE III SDS-PAGE analysis of polypeptides isolated by S-hexylglutathione-Sepharose 6B chromatography Identification of the subunits was achieved by Western blotting using antisera raised against rat and human GST. The level of each of the affinity-purified polypeptides were given a relative score. Bands present in highest concentration were assigned a score of + + + . Trace amounts are indicated by tr. Subunit kDa calculated using gel comprising: Organ Heart Spleen
Subunit type: Subunit class:
T 12%, Cei s T 15%, CBis
2.6% 0.6%
24.0 23.5
24.8 24.0
25.5 23.7
25.7 23.7
26.1 25.8
26.4 25.8
27.0 26.1
glyoxalase
Yf ~t
Ya a
Ya a
Yn /t
Yc a
Yb /t
+ tr
+ + +
-
+
-
-
+ + + +
-
tr +
+
Brain
+
+ +
-
-
+ + +
-
+
Liver Adrenal cortex Adrenal medulla Lung Kidney
tr tr tr tr +
+ + + + + + tr
tr + + + tr tr tr
+ + + + +
+ + + + + + -a + +
+ + + tr + + + +
+ + + + +
anti-Ya
anti-Y¢ 1
2
3
4
S
6
7
9
10
c
W
anti-Yb
anti-Yf
1
2
3
4
5
6
7
8
9
10
Fig. 2. Immunochemical identification of bovine cytosolic glutathione S-transferase subunits. Immunoblots of the affinity-purified pools were performed as described elsewhere [10]. SDS-PAGE was carried out in 12% ( w / v ) polyacrylamide gels that incorporated 0.32% ( w / v ) N, N-methylenebisacuiamide (i.e. C8~ 2.6% (w/w)). Tracks 1 and 10 were loaded with the purified immunogen whilst tracks 2 - 9 contain the affinity-purified pools from heart, spleen, brain, liver, adrenal cortex, adrenal medulla, lung and kidney, respectively. Panels a, b, c and d show the blots probed with anti-Ya, anti-Yb, anti-Yc and anti-Yf IgG, respectively.
27 0-5 (o)
1
2
3
4
,L
5
!
17"5 0"4
E
C 0 oO
E ._c 15"0 E "6 E ::L 12"5
0.3
m
z o u
0 u C 0 .,O
1,-50 A
3"50 10"0
2"50
C; 0 "r"
0.2
7.5 ¢:: 0
C
!
5.0
Ot ;3
0"1
u,
t
0'50
_=
0"25
A,
.~, 0
E E .9 ¢-
'6 •~
m
#
2"5 0 (b)
2
I
3
4
5
0.8 _. 1
30"0
E ~50 .-
E 25"0
0.6
"o
15"0
1
o
10.0
JO50
E
E t-"
- "
350
20'0
z 1:3
¢0
(J
o 0.4 u C 0
tO "I" Ul
< 0.2-
E
250
E
1.50
0
3
i-
.o
~
d~
5.0
0
_
O
~
0
JO 10
20
0.05-0.10, 0.08-0.20, 0 . 1 0 - 0 . 5 0 mmol/litre, respectively.
and
30
40 50 60 70 80 90 Froction no. Fig. 3. Gradient elution of bovine GST and glyoxalase ! from S-hexylglutathione-Sepharose 6B. Cytosols from adrenal cortex and liver were dialysed against two changes, each of 5 litres of buffer A, 50 mM-Tris-HCl buffer (pH 7.8) containing 200 mM NaCI. The dialysed material was re-centrifuged (30 min, 20000x g, 4°C) and applied to the S-hexylglutathione affinity columns (1.6×20.0 cm). After extensive washing, the columns were developed in two stages, using two separate gradients of S-hexylglutathione (see Ref. 21 for details). Panels a and b show the profiles obtained form bovine adrenal cortex and liver, respectively. Fractions (5.0 ml) were collected and GST activity with CDNB (•) and absorbance at 280 nm (o) measured. The S-hexylglutathione gradient is represented by a solid line. !.n each panel the numbered horizontal bars indicate the fractions which were combined for SDS-PAGE (Fig. 4).
1.00-1.20
Discussion This study was undertaJ~ea to identify the GST present in bovine organs before starting to investigate the
biological control of these transferases in primary cell cultures derived from different organs. Previous work has shown the existence of only two bovine GST, isolated from livec cytosol, catalytically similar and possessing identical N-terminal amino acid sequences [19]; subsequently, both have been designated # class GST [7]. Although quantitatively bovine organs are a
28 b) Liver
a) Adrenal cortex 1
2
/.
5
1
2
3
l,
5
Fig, 4. SDS-PAGE of bovine proteins resolved by gradient elution of $.hexylglutathione-Sepharose 6B. Fractions from chromatograms of adrenal cortex and liver proteins eluted from the affinity matrix were combined into 5 pools (shown in Fig. 3). Each of these pools analysed by SDS-PAGE in 12~ (w/v) polyacrylamide gels comprising 0.32~ (w/v) N,N-methylenebisacrylamide. The gels were all loaded with about 4/~g of protein. Panel a shows pools 1-5 from the adrenal cortex, and panel b shows pools 1-5 from the liver.
good source of cells, they clearly could not provide a good model to study GST expression if only one of the three GST gene families is represented in this animal. The data described in this paper show that bovine organs contain all three classes of cytosolic GST (a, and ~r) that have been reported in rodents and in man [7,11,21]. The Ya- and Yc-type subunits (a class GST), the Yb- and Yn-type subunits (/~ class GST) and the Yf-type subunits (vr class GST) have all been identified, thereby demonstrating the existence of at least five distinct basic subunit types. The bovine GST system is thus more complex than previously supposed and primary cell cultures of bovine organs, such as the adrenal cortex [17,18], should allow a wide cross-section of GST enzymes to be studied. Small differences in the electrophoretic properties of the Ya-type subunits in adrenal cortex and fiver were observed. During SDS-PAGE in gels comprising C Bis 2.6~ (w/w), the Ya-type subunit in adrenal cortex was found to migrate slightly faster than the Ya-type subunit in bovine liver; this difference was not observed in gels comprising C ais 0.6~ (w/w). In the rat, both fetal fiver and adult kidney have been shown to express a "slow Ya-type subunit', related immunochemically to the "fast Ya-type subunit' [21,29,30]. In rat kidney, this slower migrating polypeptide has been referred to as Y1 [21]; in the rat, the Ya and YI subunits have apparent
molecular masses of 25.5 and 25.7 kDa, respectively, in gels comprising C his 2.6% (w/w), but in gels comprising C nis 0.6~ Ya and YI have identical electrophoretic mobilities. It is possible that the 'slow Ya-type subunit' in bovine liver is equivalent to the YI subunit in rat kidney and fetal liver. The rat YI subunit (25.7 kDa) has higher peroxidase activity with cumene hydroperoxide as substrate than the Ya subunit (25.5 kDa). During the present study, the affinity-purified hepatic GST pool exhibited a significantly higher peroxidase activity than the GST pool from the adrenal cortex. It is clear that further structural data are required to establish the relationship between the bovine hepatic Ya-type subunit and the rat YI subunit. Sheehan and Mantle [31] have described two distinct Ya subunits in the rat and it remains to be determined whether either of these corresponds to YI. As in other species, bovine GST are expressed in an organ-specific fashion. The spectrum of subunits that the various bovine organs contain is similar to that noted in the rat [10-14], especially in the adrenals, liver and lung. Of all the subunits, the pattern of expression of bovine Ya shows closest similarity to the rat. However, differences in expression of the other subunits between these two species are apparent. For example, bovine kidney lacks the Yf subunit, which represents a major renal polypeptide in the rat. Similarly, the Yc subunit is absent from bovine spleen but is readily detected in rat spleen. In addition to similarities in their immunochemical properties, bovine GST subunits appear to share certain other characteristics with rat GST. For example, the Ya subunit possesses a mobility during SDS-PAGE that is dependent on the amount of cross-linker incorporated into the resolving gel (cf. Ref. 28). When gels comprising CBi s 0.6t~ are employed, the Ya subunit migrates faster towards the anode than Yf, whereas the converse is true when gels comprising Cais 2.6% are used to resolve Ya and Yf. The order of elution of bovine GST from the $-hexylglutathione affinity matrix is closely similar to that observed in the rat (cf. Ref. 21). Bovine GST, like their rat counterparts, appear to be able to form both homodimers and heterodimers (cf. Refs. 32 and 33). Chromatofocusing of pool 1 from the adrenal cortex (Fig. 3) yielded a purified peak of the bovine YaYc heterodimer at pH 8.6, and chromatofocusing of pool 3 yielded a purified peak of "~'bYn heterodimer at pH 8.1 (Milner, S.W., Walker, S.W. and Hayes, J.D., unpublished data). Several differences between bovine and rat GST were observed during the present study. Bovine GST have little activity towards 1,2-dichloro-4-nitrobenzene or trans-4-phenyl-3-buten-2-one, whereas rat Ybl and Yh2 subunits are active with both these substrates [4,7,33]. Also, the bovine Yc subunit was found to possess a significantly faster anodal mobility during SDS-PAGE
29 than its rat homologue. The reason for this latter difference is not known, but it is recognised that the rats Yc subunit has a much slower electrophoretic mobility than would be predicted from its primary structure; the rat Ya subunit comprises 222 amino acid residues and the Yc subunit 221 residues [34,35] but their apparent molecular masses, as determined by SDS-PAGE, are 25.5 and 27.5 kDa, respectively [28]. It is possible that the slower mobility of rat Yc is the result of an Ala -~ Pro substitution at position 171 (for a similar example see Refs. 36 and 37). If this hypothesis is correct, it might be expected that a different substitution has occurred at position i71 in the bovine Yc subuniL The purification strategy devised by Asaoka [19] for bovine hepatic GST resulted in the successful purification of mu class enzymes but not of the a class enzymes. The loss of a class GST can be ascribed to the use of the triazine dye agarose, Orange A-agarose, as an affinity purification step. It can therefore be concluded that/t class but not a class GST bind to this triazine dye matrix. Substituted S-chlorotriazines, atrazine and simazine, are used to control annual weeds and a major detoxication pathway for these herbicides is by conjugation to GSH [3]. It would be interesting to know whether the bovine ~t class GST can metabolise the S-chlorotriazine herbicides since they bind to triazine dye agarose. Acknowledgements We thank Professor L.G. Whitby for critically reading this script. The financial support of the MRC (G8622978 CA) and The Wellcome Trust is gratefully acknowledged. We thank Mrs. E. Ward and Miss D. Fisher for help with the preparation of this manuscript. References 1 Meister, A. and Anderson, M.G. (1983) Annu. Rev. Biochem. 52, 711-760. 2 Mannervik, B. (1980) in Enzymatic Basis of Detoxication, Vol. 2. (Jakoby, W.B., ed.), pp. 263-273, Academic Press, New York. 3 Hayes, J.D. and Wolf, C.R. (1988) in Glutathione Conjugation: Its Mechanism and Biological Significance (Sies, H. and Ketterer, B., eds.), pp. 315-355, Academic Press, New York. 4 Mannervik, B. (1985), Adv. Enzymol. Relat. Areas Mol. Biol. 57, 357-417. 5 Ketterer, B. (1986) Xenobiotica 126, 957-973. 6 Chasseaud, L.F. (1979) Adv. Cancer Res. 29, 175-274.
7 Mannervik, B., Afin, P., Guthenberg, C., Jennson, H., Tahir, M.K., Warholm, M. and Jornvall, H. (1985) Proc. Natl. Acad. Sci. USA 82, 7202-7206. 8 Wendel, A. (1980) in Enzymatic Basis of Detoxication, Voi. 1 (Jakoby, W.B., ed.), pp. 333-353, Academic Press, New York. 9 Haifiwell, B. and Gutteridge, J.M.C. (1985) Free Radicals in Biology and Medicine, Clarendon Press, Oxford. 10 Tu, C.-P.D., Weiss, M.J., Li, N. and Reddy, C.C. (1983) J. Biol. Chem. 258, 4659-4662. 11 Hayes, J.D. and Mantle, T.J. (1986) Biochem. J. 233, 779-788. 12 Pemble, S.E., Taylor, J.B. and Ketterer, B. (1986) Biochem. J. 240, 885-889. 13 Scully, N.C. and Mantle, T.J. (1981) Biochem. J. 193, 367-370. 14 Li, N.-Q., Reddanna, P., Thyagaraju, K., Reddy, C.C. and Tu, C.-P.D. (1986) J. Biol. Chem. 261, 7596-7599. 15 Hanlgan, M.J. and Pitot, H.C. (1985) Carcinogenesis 6, 167-172. 16 Tate, S.S. (1980) in Enzymatic Basis of Detoxication, Vol. 2 (Jakoby, W.B., ed.), pp. 95-120, Academic Press, New York. 17 Williams, B.C., Lightly, E.R.T., Ross, A.R., Bird, I.M. and Walker, S.W. (1988) J. Endocrinol., in press. 18 Walker, S.W., Lightly, E.R.T., Milner, S.W. and Williams, B.C. (1988) Mol. Ceil. Endocrinol. 57, 139-147. 19 Asaoka, K. (1984) J. Biochem. 95, 685-696. 20 Mannervik, B. and Guthenberg, C. (1981) Methods Enzymol. 77, 231-235. 21 Hayes, J.D. (1988) Biochem. J. 225, 913-922. 22 Racker, E. (1951) J. Biol. Chem. 190, 685-696. 23 Reddy, C.C., Tu, C.-P.D., Burgess, J.R., Ho, C.Y., Scholz, R.W. and Massaro, E.J. (1981) Biochem. Biophys. Res. Commun. 101, 970-978. 24 Habig, W.H. and Jakoby, W.B. (1981) Methods Enzymol. 77, 398 -405. 25 Bradford, M. (1976) Anal. Biochem. 72, 248-254. 26 Hayes, J.D. and Clarkson, G.H.D. (1982) Biochem. J. 207, 459-470. 27 McLellan, L.I. and Hayes, J.D. (1987) Biochem. J. 245, 399-406. 28 Hayes, J.D. and Mantle, T.J. (1986) Biochem. J. 237, 731-740. 29 Meyer, D.J., Tan, K.H., Christodoulides, L.G. and Ketterer, B. (1985) in Free Radicals in Liver Injury (Poll, G., Cheeseman, K.H., Dianzani, M.U. and Slater, T.F., eds.), pp. 221-224, IRL Press, Oxford. 30 Scott, T.R. and Kirsch, R.E. (1987) Biochim. Biophys. Acta 926, 264-269. 31 Sheehan, D. and Mantle, T.J. (1984) Biochem. J. 218, 893-897. 32 Hayes, J.D., Strange, R.C. and Percy-Robb, I.W. (1981) Biochem. J. 197, 491-502. 33 Hayes, J.D. (1984) Biochem. J. 224, 839-852. 34 Telakowski-Hopkins, C.A., Rodkey, J.A., Bennett, C.D., Lu, A.Y.H. and Pickett, C.B. (1985) J. Biol. Chem. 260, 5820-5825. 35 Tu, C.-P.D. Lai, H.-C. and Reddy, C.C. (1987) in Glutathione S-transferases and Carcinogenesis (Mantle, T.J., Pickett, C.B. and Hayes, J.D., eds.), pp. 87-~ 10, Taylor and Francis, London. 36 De Jong, W.W., Zweers, A. and Cohen, L.H. (1978) Biochem. Biophys. Res. Commun. 82, 532-539. 37 Carstens, E.B., Krebs, A. a~d Gallerneault, C.E. (1986) J. Virol. 58, 684-688.
21
BBA33285
Expression of glyoxalase, glutathione peroxidase and glutattfione S-transferase isoenzymes in different bovine tissues John D. Hayes, Susan W. Milner and Simon W. Walker University Department of Clinical Chemistry, The Royal Infirmary, Edinburgh ( U.K.) (Received 12 August 1988)
Key words: Glutathione peroxidase; Glutathione S-transferase; Glyoxalase; (Bovine)
(1) The tissue-specific expression of various glutathione-dependent enzymes, including glutathione S-transferase (GST), glutathlone peroxidase and glyoxalase |, has been studied in bovine adrenals, brain, heart, kidney, liver, lung and spleen. Of the organs studied, liver was found to p.,~sess the greatest GST and glyoxalase I activity, and spleen the greatest glutathione peroxidase activity. The adrenals contained large amounts of these glutathione-dependent enzymes, but significant differences were observed between the cortex and medulla. (2) GST and glyoxalase | activity were isolated by $-hexyiglutathione affinity chromatography. Glyoxalase | was found in all the organs examined, but GST exhibited marked tissue-specific expression. (3) The or, ~t and ~t classes of GST (i.e., those that comprise respectively Ya/Yc, Y b / Y n and Yf subunits) were all identified in bovine tissues. However, the Ya and Yc subunits of the a class GST were not co-ordinately regulated nor were the Yb and Yn subunits of the /t class GST. (4) Bovine Ya subunits (25.5-25.7 kDa) were detected in the adrenal, liver and kidney, but not in brain, heart, lung or spleen. The Yc subunit (26.4 kDa) was expressed in all those organs which expressed the Ya subunit, but was also found in lung. The # class Yb (27.0 kDa) and Yn (26.1 kDa) subunits were present in all organs; however, brain, lung and spleen contained significantly more Yn than Yb type subunits. The ¢t class Yf subunit (24.8 kDa) was detected in large amounts in the adrenals, brain, heart, lung and spleen, but not in kidney or liver. (5) Gradient affinity elution of S-hexylglutathioneSepharose showed that the bovine proteins that bind to this matrix elute in the order Ya/Yc, Yf, Yb/Yn and glyoxalase I. (6) In condusion, the present investigation has shown that bovine GST are much more complex than previously supposed; Asaoka (J. Biochem. 95 (1984) 685-696) reported the purification of # class GST but neither a nor ~t class GST were isolated.
Introduction
Glutathione is the major intracellular non-protein thiol; it exists in reduced (GSH) and oxidised (GSSG) forms. In addition to its participation in normal metabolic processes, GSH helps protect against damage caused by free radicals and plays several roles in detoxication of xenobiotics [1]. Important GSH-dependent detoxication enzymes include glyoxalase, glutathione $-transferase and glutathione peroxidase. The glyoxalase system uses GSH to protect the cell against 2-oxoaldehydes. This system involves two enzymes, glyoxalases I and II. Glyoxalase I catalyses the
Abbreviations: CDNB, 1-chloro-2,4-dinitrobenzene; SDS-PAGE, sodium dodecyl sulphate-polyacrylamide gel electrophoresis. Correspondence: J.D. Hayes, University Department of Clinical Chemistry, The Royal Infirmary, Edinburgh, EH3 9YW, U.K.
formation of a thiol ester between thiolesterase GSH and the 2-oxoaldehyde. Glyoxalase II is a thiolesterase; its hydrolytic activity results in the formation of a 2-hydroxy acid and regeneration of GSH [2]. Glutathione S-transferase (GST) forms a thiol-ether linkage between GSH and a large spectrum of electrophiles including herbicides, insecticides, chemotherapeutic drugs and carcinogens [3-5]. This conjugation reaction is the first step in mercapturic acid synthesis [6]. GST exists in multiple forms; cytosolic GST are encoded by three multi-gene families, designated by Mannervik et al. [7] as a, # and ~r. These families display distinct catalytic properties and play separate detoxication roles. Glutathione peroxidase uses GSH to protect against oxidative damage caused by oxygen radicals. In this reaction peroxides are reduced to alcohols and GSH is oxidised to GSSG [8]. Two types of glutathione peroxidase exist; these differ in their requirement for selenium. The selenium-dependent enzyme (GPx) is ac-
0167-4838/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
22 tive with both hydrogen peroxide and organic hydroperoxides, (e.g. cumene hydroperoxide). The selenium-independent glutathione peroxidase also possesses GST activity and is active with organic hydroperoxides, but not with hydrogen peroxide. This selenium-independent activity is principally expressed by a class GST. The level of these three glutathione-requiring enzymes within cells probably reflect their ability to deal with oxidative stress and/or cytotoxic insult. Indeed, in the case of GST, differences in enzyme levels have been correlated with drug resistance [3]. Despite the pharmacological and toxicological consequences that result from alterations in the intracellular levels of GSH-requiring enzymes, the biological control of these enzymes has been little investigated [9]. Marked tissue-specific expression of GSH-requiring enzymes is known to occur, and is presumed to reflect differences in the metabolic functions and physiological stresses encountered by different organs. The differences that have been described may partly explain the intrinsic resistance of certain :ells to cytotoxic drugs and oxidative damage. The synthesis of steroid hormones by the adrenal cortex results in the production of free radicals, but little is known about the defence mechanisms the adrenal cortex has evolved to compensate for this unique form of oxidative stress. Most work describing the tissuespecific expression of GSH-metabolising enzymes has so far employed the rat as the model system [10-16]. Although the rat is often used to study the effects of diet and the administration of xenobiotics, its size limits studies of the control of GSH-metabolising enzymes in steroidogenic tissues. It is also essential to study the adrenal cortex and medulla separately since they are both metabolically and embryologically distinct. The physical separation of these two tissues is not easy in the rat but is much simpler if bovine adrenal glands are used as the tissue source. A primary cell culture derived from bovine adrenal cortex has been established in this department [17,18]; it provides an ideal model for studying the biological control of GSH-metabolising enzymes. However, these enzymes have not been well characterised in bovine organs. For example, it is unclear whether a and r, classes of GST are expressed in this species. To date, GST have only been studied in bovine liver, from which organ Asaoka [19] isolated two closely si~nilar enzymes, more recently identified as members of the/~ class of GST [7]. The presence o f ~ ~ ~, class GST in extrahepatic bovine tissues has not yet been described. It is important to define which enzymes are expressed in bovine organs before commencing a longerterm study of the biological control of GSH-metabolising enzymes, using primary cell cultures derived from the adrenal cortex and other organs. As tissue differences in the content of GSH-requiring enzymes may
reflect their embryological origin, organs of ecto-, mesoand endodermal origin were chosen for the present study. Adrenal medulla and brain were chosen as examples of organs of ectodermal origin, and liver and lung as examples or organs of endodermal origin. The remaining four organs studied (adrenal cortex, heart, kidney and spleen) are all of mesodermal origin. The GST and glyoxalase I subunits present in different organs have been isolated by affinity chromatography on Shexylglutathione-Sepharose to enable direct comparison of their electrophoretic properties.
Materials and Methods
Chemicals These were all of analytical grade and readily available commercially. The S-hexylglutathione-Sepharose 6B was made by the method of Mannervik and Guthenberg [201.
Preparation of organ samples Bovine organs were obtained from a local slaughterhouse. The organs, obtained within 30 min of death, were transported to the department on ice. Except for the adrenal glands, organs were immediately stored at - 8 5 ° C , until processed. Adrenal cortex and medulla were manually dissected from one another before being stored separately at - 85 o C. Homogenates (1:4 (w/v)) of bovine tissues were prepared in ice-cold 50 mM-Tris-HC1 buffer (pH 7.8) that contained 200 mM-NaCI (buffer A). The 100 000 × g supernatant fraction (cytosol) from the organ extracts was prepared at 4 ° C using standard centrifuge techniques. The cytosol (300 ml) was filtered through a plug of glass wool, to remove lipid, and dialysed for 16 h at 4°C against 4 litres of buffer A. Following dialysis, the cytosol from each organ was re-centrifuged (15 000 x g, 30 rain, 4°C) to remove precipitated material; the supernatant was either subjected to chromatography immediately or stored ( - 8 5 o C) for later analysis.
Isolation of glutathione S-transferases and glyoxalase These enzymes were prepared from each of the cytosols by affinity chromatography on S-hexylglutathioneSepharose 6B. Columns (1.6 × 15 cm) of the affinity matrix were equilibrated at 4°C with buffer A. Dialysed cytosol (400-500 ml; 0.6-2~55 g of protein) was applied to these columns and the non-specifically adsorbed protein removed by washing the affinity matrix with at least 400 ml of buffer A. Specifically bound material was eluted using two different methods. The first method, used for the Western blot study, entailed a single-step elution of adsorbed material with a solution of 5 mM S-hexylglutathione in buffer A (as described in Ref. 20). The second method, used to prepare hepatic and adrenal GST, employed gradient elution of the
23 affinity matrix in two steps with 0-0.25 mM Shexylglutathione followed by 0.25-5.0 mM S-hexylglutathione [21]. The material (20 mi; 3-20 mg of protein) that was eluted from S-hexylglutathione-Sepharose 6B in a single step was dialysed at 4 ° C against 2 litres of buffer A before storage at - 85 o C prior to Western blot analysis.
TABLE I
Activities of GSH-dependent enzymes in the cytosolic fraction from various bovine tissues Values given are #mol of product formed per min per mg protein
Organ Enzyme and protein determination All enzyme assays were performed at 37 ° C. Glyoxalase I activity was measured by the method of Racker [22], using methylglyoxal as substrate. Glutathione peroxidase activities, towards both hydrogen peroxide and cumene hydroperoxide, were determined by the coupled assay system of Reddy et al. [23]. Glutathione S-transferase activities were assayed by the methods described by Habig and Jakoby [24]. Protein concentrations were measured by the dye-binding method of Bradford [25]. The 1-chloro-2,4-dinitrobenzene, cumene hydroperoxide, hydrogen peroxide and Coomassie blue G-250 protein assays were all carried out using a centrifugal analyser; all other assays were carried out manually. The assays were performed in triplicate or quintuplicate and the mean value was calculated and used for further calculations. The precision of these assays has been described before [21,26,~.7]. To avoid inter-batch analytical variation from contributing towards observed catalytic differences in the bovine organs the enzyme assays (for a given substrate) were all performed on the same day using the same stock solutions. Electrophoretic methods The SDS-polyacrylamide gel electrophoresis (SDSPAGE) and Western blotting techniques have been described elsewhere [11,28]. Rat GST were employed as M r standards for the SDS-PAGE and Western blot analyses; purified GST from rat fiver and rat lung provided Ya, Yb and Yc subunits and Yf, Yb and Yc subunits, respectively. The majority of SDS-PAGE analyses performed employed a 12~ polyacrylamide resolving gel containing C Bis 2.6~. However, during attempts to improve the resolution of Ya and Yc subunits, a 15% polyacrylamide resolving gel containing low levels of cross-linker (C ais 0.6~) was used [28]. Results
Cytosolic activity of GSH-requiring enzymes Significant glyoxalase I, GST and glutathione peroxidase activities were found in all the bovine organs studied, but the relative amounts varied considerably. The range of specific activities for glyoxalase I calculated in the different organs was notably less than the range of activities for GST and glutathione peroxidase. Table I shows that liver, spleen and kidney possessed highest glyoxalase activity. The fiver, adrenal cortex and
Heart Spleen Brain Liver Cortex Medulla Lung Kidney
Substrate:
Enzymeactivity glyoxalase GST GSH-peroxidase methyl CDNB cumene hydrogen glyoxal hydro- peroxide peroxide 0.87 0.10 0.07 0.05 1.50 0.20 0.71 0.56 0.88 0.08 0.07 0.09 1.97 0.37 0.21 0.01 0.77 0.23 0.28 0.18 0.58 0.07 0.16 0.15 0.47 0.23 0.20 0.08 1.27 0.04 0.27 0.07
lung had highest GST activity towards 1-chloro~2,4-dinitrobenzene (CDNB). Spleen and adrenal cortex had high glutathione peroxidase activity towards both hydrogen peroxide and cumene hydroperoxide. By contrast, both liver and kidney were found to exhibit high peroxidase activity towards cumene hydroperoxide but not towards hydrogen peroxide. Comparisons between GST activity and the two peroxidase activities allowed an estimate of the ~mount of selenium-dependent glutathione peroxidase in the different organs as the a class GST subunits (i.e., Ya and Yc) are active with cumene hydroperoxide but not with hydrogen peroxide. For example, liver cytosol has high CDNB-GSH-conjugating activity and moderate peroxidase activity for cumene hydroperoxide, but little activity for hydrogen peroxide, which suggests that this organ expresses little selenium-dependent glutathione peroxidase activity. Adrenal cortex and medulla exhibit similar peroxidase activities towards hydrogen peroxide, but the data in Table I indicate that the medulla contains considerably less selenium-independent glutathione peroxidase activity than the cortex as the activities towards both cumene hydroperoxide and CDNB are significantly lower in the medulla than in the cortex. Activity of affinity-purified enzymes S-Hexylglutathione-Sepharose 6B affinity chromatography was exploited to isolate glyoxalase I and GST from the various organs to allow the molecular basis for the tissue differences in enzyme activities to the explored; the selenium-dependent form of glutathione peroxidase does not bind to this matrix. The proportion of cytosolic protein that specifically bound to the affinity matrix varied between 0.2 and 1.0% (Table II). The percentage of GST activity towards CDNB that was absorbed to the S-hexylglutathione affinity column from the various organ cytosols was as
24 TABLE II Analysis of affinity-purified GST using model substrates The coefficient of variation obtained using these assays is as follows: p-nitrobenzyl chloride, 10.5%; ethacrynic acid 8.5~; trans-4-phenyl-3-butene2-one, 15~; 1,2-dichloro-4-nitrobenzene, 14.5%; cumene hydroperoxide, 6.5~; 1-chloro-2,4-dinitrobenzene, 6.0%; AS-androstene-3,17-dione, 9.0%. The CVs are higher for trans-4-phenyl-3-butene-2-one and 1,2-dichloro-4-nitrobenzene than normally found in this laboratory owing to the low :~pecfic activities of the bovine GST pools for those substrates.Abbreviations: n.d., not detected; p-NBC, para-nitrober~zyl chloride; EA, ethacrynic acid; t-PBO, trans-4-phenyl-3-butene-2-one; DCNB, 1,2-dichloro-4-nitrobenzene, CuOOH, cumene hydroperoxide; CDNB, 1-chloro2,4-dinitrobenzene; ASADD, AS-androstene-3,17-dione. Tissue
Heart Spleen Brain Liver Adrenal cortex Adrenal medulla Lung Kidney
g of protein applied to the affinity matrix
% protein bound to affinity column
GST substrate (/~mol of product formed per min per mg protein) p-NBC
EA
t-PBO
DCNB
CuOOH
CDNB
AS-ADD
1.43 0.64 0.61 2.55 1.92 2.19 1.48 2.34
0,9 0.2 0.2 1.0 0.8 0.9 0.8 0.9
0.73 0.71 0.50 0.38 0.27 0.61 0.53 0.21
0.78 0.30 0.72 0.07 0.11 0.15 0.11 0.08
n.d. 0.01 0.005 0.02 0.01 0.03 0.01 0.02
0.05 n.d. 0.01 0.003 0.001 n.d. n.d. n.d.
n.d. 0.24 0.09 4.17 1.76 0.43 9.20 0.20
12.43 10.89 11.89 7.80 6.15 9.50 17.27 6.37
0.11 0.04 0.04 0.07 0.18 0.11 0.07 0.16
follows: adrenal cortex, 85%; adrenal medulla, 88%; brain, 90%; heart, 96%; kidney, 67%; fiver, 79%; lung, 77%; spleen, 85%. The enzyme activities of these affinity-purified pools were examined with the range of GST substrates used to discriminate between the isoenzymes in rat and man (for a review, see Ref. 4). The results (Table II) reveal marked inter-organ differences in specific GST activity for a given substrate; with p-nitrobenzyl chloride and 1-chloro-2,4-dinitrobenzene, a 3-fold range in specific activities was observed, whereas with cumene hydroperoxide, differences in specific activity of at least 100-fold were noted between organs. The peroxidase activity towards cumene hydroperoxide is presumed to be due entirely to the selenium-independent forms since no activity towards hydrogen peroxide was detected. Differences were observed in the spectrum of GST substrates that the organs can metabolise. For example, heart and brain possess high activity for ethacrynic acid but tittle activity for cumene hydroperoxide, whereas liver and lung express high activity for cumene hydroperoxide but tittle activity for ethacrynic acid. Of all the organs studied, adrenal cortex and kidney expressed the highest activity for androstene-3,17-dione, but possessed markedly different activities for cumene hydroperoxide; in the S-hexylglutathione affinity-purified enzyme pools from human and mouse liver, these two substrates are specific for the Ya subunit. A similar divergence in these two activities was observed with liver and lung.
SDS /polyacrylamide.gel electrophoresis The major polypeptides that the affinity-purified pools comprise have mobilities during SDS-PAGE similar to those of rat GST subunits. Using this method, at
least seven subunits of different molecular mass can be resolved. Assuming that rat Ya, Yb, Yc and Yf subunits have molecular masses of 25.5, 26.3, 27.5 and 24.8 kDa, respectively [26], then bovine affinity-purified polypeptides have apparent masses of 24.0-27.0 kDa in gels comprising 0.32% (w/v) N,N-methylenebisacrylamide. T',e relative electrophoretic mobilities of these polypeptides were found to vary according to the amount of cross-tinker in the resolving gel. When electrophoresis was performed in gels composed of 0.32% (w/v) N, Nmethylenebisacrylamide, the bovine Yf-type subunit was found to migrate faster than the Ya-type subunit. However, the converse was true when the procedure was conducted in gels comprising 0.09% (w/v) N, N-methylenebisacrylamide. Fig. 1 shows this cross-tinker-dependent electrophoretic mobility, and demonstrates clear differences in the subunits expressed by the various organs that bind to the S-hexylglutathione affinity matrix.
Immunoblotting of glyoxalase I The protein that bound to S-hexylglutathione-Sepharose 6B was probed with antiserum raised against mouse glyoxalase I. The blots obtained showed that glyoxalase I was present in all bovine organs but no obvious tissue specific differences in the amount of this enzyme were detected (data not shown).
Immunoblotting of GST subunits Western blotting experiments were carried out using antisera raised against rat and human enzymes of the a (Ya/Yc), # (Yb/Yn) and ~r (Yf) classes, to investigate the organ-specific differences in GST content (Table III). Fig. 2 shows that the Yf subunit is present in significant amounts in all the organs examined except
25 1
2
3
~
5
6
7
small amounts in all organs except heart. The levels of the Yn subunit were more variable than Yb, with Yn being expressed in greatest amounts in spleen, brain and lung. The Yc subunit was expressed in bovine liver and lung but was not readily demonstrated in other organs. The bovine Yc subunit cross-rezcted with antisera raised against the rat Ya subunit but not the human Ya subunit (i.e., from GST BIB 1). Antisera against the rat Yk subunit, which is a member of the a class of GST [11], failed to react with any of the bovine GST. Identification of bovine Yf by Western blot analysis was found to correlate with the GST ethacrynic acid assays in Table If. Similarly, the blotting data for the Yc subunit correlates with the cumene hydroperoxide assays (Table II). Blots for Ya are also in broad agreement in all organs, except heart, with the A%androstene3,17-dione assays. No activity was identified that segregated with the Yb/Yn subunits, and it is assumed that the appropriate substrate has yet to be identified.
8
A
Yc ¥o yf /
.....
Yc Yb Yf Ya
w
Resolution Of glutathione S-transferase subunits and glyoxalase I by gradient elution of S-hexylglutathioneSepharose
C 0"6%
CBis
2"6% Yb
j ~ y c
Yb
G ~
J
~
Yh
•q - - - ~ YC, Y n
~
~-------- Y 1
Yo. Y I
*V¢
Fig. 1. Electrophoretic analysis of glutathione S-transferases purified by affinity chromatography. SDS-polyacrylamide gel electrophoresis was performed as described in the text. Panel A shows electrophoresis using a 12% (w/v) polyacrylamide resolving gel containing Cais 2.6% (w/w). Panel B shows electrophoresis of the same samples performed using a 15% (w/v) polyacrylamide resolving gel containing Cais 0.6% (w/w). In both instances the purified pools were loaded in the same order as follows: 1, heart; 2, spleen; 3, brain; 4, liver; 5, adrenal cortex; 6, adrenal medulla; 7, lung; 8, kidney. The positions of the rat Ya (25.5 kDa), Yb (26.3 kDa), Yc (27.5 kDa) and Yf (24.8 kDa) subunits are shown. Panel C shows a line diagram depicting the relative mobilities of the bovine GST subunits Ya, Yb, Yc, Yf, YI and Yn in polyacrylamide-gels containing Cnis 0.6% (w/w) ,nd Cni s 2.6% (w/w). The most anodal band, designated G, represents glyoxatase I.
liver and kidney. The Ya-type subunit was found in liver, adrenal cortex, adrenal medulla and kidney, but not in heart, spleen, brain or lung. However, the Ya-type subunit was resolved into two components (25.5 kDa and 25.7 kDa) in polyacrylamide-gels containing larger amounts of cross-linker, C,is 2.6%, but not in gels that incorporated lesser amounts of cross-linker, C Bis 0.6%. Antisera to the # class Yb GST, which crossreact with both Yb and Yn subunits, showed Yb to be present in
A detailed examination of the bovine GST subunits in the adrenal cortex and the liver was performed by eluting the enzymes which bound to S-hexylglutathione-Sepharose 6B with a gradient of counter-ligand. Comparison between the S-hexylglutathione-Sepharose 6B elution profiles obtained from these two organs revealed marked differences (Fig. 3). The adrenal cortex produced a single major peak of GST activity which eluted at a S-hexylglutathione concentration of 0.060 mmol/litre. By contrast, liver produced three major peaks of GST activity which eluted at S-hexylglutathione concentrations of 0.048, 0.075 and 0.130 mmol/litre. Both organs also yielded a protein-containing peak that eluted from the affinity column at 1.10 mmol/litre S-hexylglutathione; this protein lacked GST activity. Fractions from the chromatograms from adrenal cortex and liver were combined into 5 pools, as shown in Fig. 3, and the pooled material subjected to SDSPAGE (Fig. 4). Western blotting (data not shown) showed that the polypeptides were distributed as follows: Ya, in pools 1 and 2; Yb, in pools 2, 3 and sometimes 4; Yc, in pool 1; Yf, in pool 2; Yn in pool 3, 4 and sometimes 5; glyoxalase I, pool 5. Western blotting revealed quantitative differences in the subunit content of these pools. For example, the Yf subunit was not detected in pools obtained from bovine liver, whereas adrenal cortex did not contain as much Yb or Yc as liver. These data indicated that the polypeptides were eluted from the affinity matrix in the order Yc, Ya, Yf, Yb, Yn and, lastly, glyoxalase I at S-hexylglutathione concentrations of about 0.01-0.06, 0.01-0.10,
26 TABLE III SDS-PAGE analysis of polypeptides isolated by S-hexylglutathione-Sepharose 6B chromatography Identification of the subunits was achieved by Western blotting using antisera raised against rat and human GST. The level of each of the affinity-purified polypeptides were given a relative score. Bands present in highest concentration were assigned a score of + + + . Trace amounts are indicated by tr. Subunit kDa calculated using gel comprising: Organ Heart Spleen
Subunit type: Subunit class:
T 12%, Cei s T 15%, CBis
2.6% 0.6%
24.0 23.5
24.8 24.0
25.5 23.7
25.7 23.7
26.1 25.8
26.4 25.8
27.0 26.1
glyoxalase
Yf ~t
Ya a
Ya a
Yn /t
Yc a
Yb /t
+ tr
+ + +
-
+
-
-
+ + + +
-
tr +
+
Brain
+
+ +
-
-
+ + +
-
+
Liver Adrenal cortex Adrenal medulla Lung Kidney
tr tr tr tr +
+ + + + + + tr
tr + + + tr tr tr
+ + + + +
+ + + + + + -a + +
+ + + tr + + + +
+ + + + +
anti-Ya
anti-Y¢ 1
2
3
4
S
6
7
9
10
c
W
anti-Yb
anti-Yf
1
2
3
4
5
6
7
8
9
10
Fig. 2. Immunochemical identification of bovine cytosolic glutathione S-transferase subunits. Immunoblots of the affinity-purified pools were performed as described elsewhere [10]. SDS-PAGE was carried out in 12% ( w / v ) polyacrylamide gels that incorporated 0.32% ( w / v ) N, N-methylenebisacuiamide (i.e. C8~ 2.6% (w/w)). Tracks 1 and 10 were loaded with the purified immunogen whilst tracks 2 - 9 contain the affinity-purified pools from heart, spleen, brain, liver, adrenal cortex, adrenal medulla, lung and kidney, respectively. Panels a, b, c and d show the blots probed with anti-Ya, anti-Yb, anti-Yc and anti-Yf IgG, respectively.
27 0-5 (o)
1
2
3
4
,L
5
!
17"5 0"4
E
C 0 oO
E ._c 15"0 E "6 E ::L 12"5
0.3
m
z o u
0 u C 0 .,O
1,-50 A
3"50 10"0
2"50
C; 0 "r"
0.2
7.5 ¢:: 0
C
!
5.0
Ot ;3
0"1
u,
t
0'50
_=
0"25
A,
.~, 0
E E .9 ¢-
'6 •~
m
#
2"5 0 (b)
2
I
3
4
5
0.8 _. 1
30"0
E ~50 .-
E 25"0
0.6
"o
15"0
1
o
10.0
JO50
E
E t-"
- "
350
20'0
z 1:3
¢0
(J
o 0.4 u C 0
tO "I" Ul
< 0.2-
E
250
E
1.50
0
3
i-
.o
~
d~
5.0
0
_
O
~
0
JO 10
20
0.05-0.10, 0.08-0.20, 0 . 1 0 - 0 . 5 0 mmol/litre, respectively.
and
30
40 50 60 70 80 90 Froction no. Fig. 3. Gradient elution of bovine GST and glyoxalase ! from S-hexylglutathione-Sepharose 6B. Cytosols from adrenal cortex and liver were dialysed against two changes, each of 5 litres of buffer A, 50 mM-Tris-HCl buffer (pH 7.8) containing 200 mM NaCI. The dialysed material was re-centrifuged (30 min, 20000x g, 4°C) and applied to the S-hexylglutathione affinity columns (1.6×20.0 cm). After extensive washing, the columns were developed in two stages, using two separate gradients of S-hexylglutathione (see Ref. 21 for details). Panels a and b show the profiles obtained form bovine adrenal cortex and liver, respectively. Fractions (5.0 ml) were collected and GST activity with CDNB (•) and absorbance at 280 nm (o) measured. The S-hexylglutathione gradient is represented by a solid line. !.n each panel the numbered horizontal bars indicate the fractions which were combined for SDS-PAGE (Fig. 4).
1.00-1.20
Discussion This study was undertaJ~ea to identify the GST present in bovine organs before starting to investigate the
biological control of these transferases in primary cell cultures derived from different organs. Previous work has shown the existence of only two bovine GST, isolated from livec cytosol, catalytically similar and possessing identical N-terminal amino acid sequences [19]; subsequently, both have been designated # class GST [7]. Although quantitatively bovine organs are a
28 b) Liver
a) Adrenal cortex 1
2
/.
5
1
2
3
l,
5
Fig, 4. SDS-PAGE of bovine proteins resolved by gradient elution of $.hexylglutathione-Sepharose 6B. Fractions from chromatograms of adrenal cortex and liver proteins eluted from the affinity matrix were combined into 5 pools (shown in Fig. 3). Each of these pools analysed by SDS-PAGE in 12~ (w/v) polyacrylamide gels comprising 0.32~ (w/v) N,N-methylenebisacrylamide. The gels were all loaded with about 4/~g of protein. Panel a shows pools 1-5 from the adrenal cortex, and panel b shows pools 1-5 from the liver.
good source of cells, they clearly could not provide a good model to study GST expression if only one of the three GST gene families is represented in this animal. The data described in this paper show that bovine organs contain all three classes of cytosolic GST (a, and ~r) that have been reported in rodents and in man [7,11,21]. The Ya- and Yc-type subunits (a class GST), the Yb- and Yn-type subunits (/~ class GST) and the Yf-type subunits (vr class GST) have all been identified, thereby demonstrating the existence of at least five distinct basic subunit types. The bovine GST system is thus more complex than previously supposed and primary cell cultures of bovine organs, such as the adrenal cortex [17,18], should allow a wide cross-section of GST enzymes to be studied. Small differences in the electrophoretic properties of the Ya-type subunits in adrenal cortex and fiver were observed. During SDS-PAGE in gels comprising C Bis 2.6~ (w/w), the Ya-type subunit in adrenal cortex was found to migrate slightly faster than the Ya-type subunit in bovine liver; this difference was not observed in gels comprising C ais 0.6~ (w/w). In the rat, both fetal fiver and adult kidney have been shown to express a "slow Ya-type subunit', related immunochemically to the "fast Ya-type subunit' [21,29,30]. In rat kidney, this slower migrating polypeptide has been referred to as Y1 [21]; in the rat, the Ya and YI subunits have apparent
molecular masses of 25.5 and 25.7 kDa, respectively, in gels comprising C his 2.6% (w/w), but in gels comprising C nis 0.6~ Ya and YI have identical electrophoretic mobilities. It is possible that the 'slow Ya-type subunit' in bovine liver is equivalent to the YI subunit in rat kidney and fetal liver. The rat YI subunit (25.7 kDa) has higher peroxidase activity with cumene hydroperoxide as substrate than the Ya subunit (25.5 kDa). During the present study, the affinity-purified hepatic GST pool exhibited a significantly higher peroxidase activity than the GST pool from the adrenal cortex. It is clear that further structural data are required to establish the relationship between the bovine hepatic Ya-type subunit and the rat YI subunit. Sheehan and Mantle [31] have described two distinct Ya subunits in the rat and it remains to be determined whether either of these corresponds to YI. As in other species, bovine GST are expressed in an organ-specific fashion. The spectrum of subunits that the various bovine organs contain is similar to that noted in the rat [10-14], especially in the adrenals, liver and lung. Of all the subunits, the pattern of expression of bovine Ya shows closest similarity to the rat. However, differences in expression of the other subunits between these two species are apparent. For example, bovine kidney lacks the Yf subunit, which represents a major renal polypeptide in the rat. Similarly, the Yc subunit is absent from bovine spleen but is readily detected in rat spleen. In addition to similarities in their immunochemical properties, bovine GST subunits appear to share certain other characteristics with rat GST. For example, the Ya subunit possesses a mobility during SDS-PAGE that is dependent on the amount of cross-linker incorporated into the resolving gel (cf. Ref. 28). When gels comprising CBi s 0.6t~ are employed, the Ya subunit migrates faster towards the anode than Yf, whereas the converse is true when gels comprising Cais 2.6% are used to resolve Ya and Yf. The order of elution of bovine GST from the $-hexylglutathione affinity matrix is closely similar to that observed in the rat (cf. Ref. 21). Bovine GST, like their rat counterparts, appear to be able to form both homodimers and heterodimers (cf. Refs. 32 and 33). Chromatofocusing of pool 1 from the adrenal cortex (Fig. 3) yielded a purified peak of the bovine YaYc heterodimer at pH 8.6, and chromatofocusing of pool 3 yielded a purified peak of "~'bYn heterodimer at pH 8.1 (Milner, S.W., Walker, S.W. and Hayes, J.D., unpublished data). Several differences between bovine and rat GST were observed during the present study. Bovine GST have little activity towards 1,2-dichloro-4-nitrobenzene or trans-4-phenyl-3-buten-2-one, whereas rat Ybl and Yh2 subunits are active with both these substrates [4,7,33]. Also, the bovine Yc subunit was found to possess a significantly faster anodal mobility during SDS-PAGE
29 than its rat homologue. The reason for this latter difference is not known, but it is recognised that the rats Yc subunit has a much slower electrophoretic mobility than would be predicted from its primary structure; the rat Ya subunit comprises 222 amino acid residues and the Yc subunit 221 residues [34,35] but their apparent molecular masses, as determined by SDS-PAGE, are 25.5 and 27.5 kDa, respectively [28]. It is possible that the slower mobility of rat Yc is the result of an Ala -~ Pro substitution at position 171 (for a similar example see Refs. 36 and 37). If this hypothesis is correct, it might be expected that a different substitution has occurred at position i71 in the bovine Yc subuniL The purification strategy devised by Asaoka [19] for bovine hepatic GST resulted in the successful purification of mu class enzymes but not of the a class enzymes. The loss of a class GST can be ascribed to the use of the triazine dye agarose, Orange A-agarose, as an affinity purification step. It can therefore be concluded that/t class but not a class GST bind to this triazine dye matrix. Substituted S-chlorotriazines, atrazine and simazine, are used to control annual weeds and a major detoxication pathway for these herbicides is by conjugation to GSH [3]. It would be interesting to know whether the bovine ~t class GST can metabolise the S-chlorotriazine herbicides since they bind to triazine dye agarose. Acknowledgements We thank Professor L.G. Whitby for critically reading this script. The financial support of the MRC (G8622978 CA) and The Wellcome Trust is gratefully acknowledged. We thank Mrs. E. Ward and Miss D. Fisher for help with the preparation of this manuscript. References 1 Meister, A. and Anderson, M.G. (1983) Annu. Rev. Biochem. 52, 711-760. 2 Mannervik, B. (1980) in Enzymatic Basis of Detoxication, Vol. 2. (Jakoby, W.B., ed.), pp. 263-273, Academic Press, New York. 3 Hayes, J.D. and Wolf, C.R. (1988) in Glutathione Conjugation: Its Mechanism and Biological Significance (Sies, H. and Ketterer, B., eds.), pp. 315-355, Academic Press, New York. 4 Mannervik, B. (1985), Adv. Enzymol. Relat. Areas Mol. Biol. 57, 357-417. 5 Ketterer, B. (1986) Xenobiotica 126, 957-973. 6 Chasseaud, L.F. (1979) Adv. Cancer Res. 29, 175-274.
7 Mannervik, B., Afin, P., Guthenberg, C., Jennson, H., Tahir, M.K., Warholm, M. and Jornvall, H. (1985) Proc. Natl. Acad. Sci. USA 82, 7202-7206. 8 Wendel, A. (1980) in Enzymatic Basis of Detoxication, Voi. 1 (Jakoby, W.B., ed.), pp. 333-353, Academic Press, New York. 9 Haifiwell, B. and Gutteridge, J.M.C. (1985) Free Radicals in Biology and Medicine, Clarendon Press, Oxford. 10 Tu, C.-P.D., Weiss, M.J., Li, N. and Reddy, C.C. (1983) J. Biol. Chem. 258, 4659-4662. 11 Hayes, J.D. and Mantle, T.J. (1986) Biochem. J. 233, 779-788. 12 Pemble, S.E., Taylor, J.B. and Ketterer, B. (1986) Biochem. J. 240, 885-889. 13 Scully, N.C. and Mantle, T.J. (1981) Biochem. J. 193, 367-370. 14 Li, N.-Q., Reddanna, P., Thyagaraju, K., Reddy, C.C. and Tu, C.-P.D. (1986) J. Biol. Chem. 261, 7596-7599. 15 Hanlgan, M.J. and Pitot, H.C. (1985) Carcinogenesis 6, 167-172. 16 Tate, S.S. (1980) in Enzymatic Basis of Detoxication, Vol. 2 (Jakoby, W.B., ed.), pp. 95-120, Academic Press, New York. 17 Williams, B.C., Lightly, E.R.T., Ross, A.R., Bird, I.M. and Walker, S.W. (1988) J. Endocrinol., in press. 18 Walker, S.W., Lightly, E.R.T., Milner, S.W. and Williams, B.C. (1988) Mol. Ceil. Endocrinol. 57, 139-147. 19 Asaoka, K. (1984) J. Biochem. 95, 685-696. 20 Mannervik, B. and Guthenberg, C. (1981) Methods Enzymol. 77, 231-235. 21 Hayes, J.D. (1988) Biochem. J. 225, 913-922. 22 Racker, E. (1951) J. Biol. Chem. 190, 685-696. 23 Reddy, C.C., Tu, C.-P.D., Burgess, J.R., Ho, C.Y., Scholz, R.W. and Massaro, E.J. (1981) Biochem. Biophys. Res. Commun. 101, 970-978. 24 Habig, W.H. and Jakoby, W.B. (1981) Methods Enzymol. 77, 398 -405. 25 Bradford, M. (1976) Anal. Biochem. 72, 248-254. 26 Hayes, J.D. and Clarkson, G.H.D. (1982) Biochem. J. 207, 459-470. 27 McLellan, L.I. and Hayes, J.D. (1987) Biochem. J. 245, 399-406. 28 Hayes, J.D. and Mantle, T.J. (1986) Biochem. J. 237, 731-740. 29 Meyer, D.J., Tan, K.H., Christodoulides, L.G. and Ketterer, B. (1985) in Free Radicals in Liver Injury (Poll, G., Cheeseman, K.H., Dianzani, M.U. and Slater, T.F., eds.), pp. 221-224, IRL Press, Oxford. 30 Scott, T.R. and Kirsch, R.E. (1987) Biochim. Biophys. Acta 926, 264-269. 31 Sheehan, D. and Mantle, T.J. (1984) Biochem. J. 218, 893-897. 32 Hayes, J.D., Strange, R.C. and Percy-Robb, I.W. (1981) Biochem. J. 197, 491-502. 33 Hayes, J.D. (1984) Biochem. J. 224, 839-852. 34 Telakowski-Hopkins, C.A., Rodkey, J.A., Bennett, C.D., Lu, A.Y.H. and Pickett, C.B. (1985) J. Biol. Chem. 260, 5820-5825. 35 Tu, C.-P.D. Lai, H.-C. and Reddy, C.C. (1987) in Glutathione S-transferases and Carcinogenesis (Mantle, T.J., Pickett, C.B. and Hayes, J.D., eds.), pp. 87-~ 10, Taylor and Francis, London. 36 De Jong, W.W., Zweers, A. and Cohen, L.H. (1978) Biochem. Biophys. Res. Commun. 82, 532-539. 37 Carstens, E.B., Krebs, A. a~d Gallerneault, C.E. (1986) J. Virol. 58, 684-688.