Mouse Liver Glutathione S-Transferase Lsoenzyme Activity

8/10/2019 Mouse Liver Glutathione S-Transferase Lsoenzyme Activity

1/10

TOXIC OLO GY AND APPLIED PHARMACOLOGY 105,2 16-225 (1990)

Mouse Liver Glutathione S-Transferase lsoenzyme Activity toward

Aflatoxin B,-8,9-Epoxide and Benzo[a]pyrene-

7,8-dihydrodiol-9,10-epoxide

HOWARD S. RAMSDELL AND DAVID L. EATON~

Department ofEnvironmental Health and Institutefor Environmental&dies,

Universi ty of Washington, Seattle, Washington 98195

Received October 1 I, 1989; accepted May 21, 1990

Mouse Liver Glutathione STransferase Isoenzyme Act ivi ty toward Aflatoxin B,-8,9-Epoxide

and Benzo[a]pyrene-7,8-dihydrodiol-9, IO-epoxide. RAMSDELL, H. S., AND EAT ON, D. L.

(1990). Toxicol. Appl. Pharmacol. 105, 216-225. As part of the studies of the biochemical

basis for species differences in biotransformation of the carcinogen aflatoxin B, (AFB, ) and its

modulation by phenohc antioxidants, we have investigated the role of mouse liver glutathione

Stransferase (GST) isoenzymes in the conjugation of AFB,-8,9-epoxide. lsoenzymes of GST

were purified to electrophoretic homogeneity from Swiss-Webster mouse liver cytosol by aff in-

ity chromatography and chromatofocusing. The isoenzyme fractions were characterized in

terms of act ivi ty toward surrogate substrates and immunologic cross-react ivity with antisera to

rat GSTs. The major isoenzymes were identified as SW 4-4, SW 3-3, and SW I-1. The specific

act ivi ty of SW 4-4 toward AFBr-8,9-epoxide was at least 50- and 150-fold greater than that of

SW 3-3 and SW l-l, respectively. Relatively high activity toward another epoxide carcinogen,

benzo[a]pyrene-7,8-dihydrodiol-9, IO-epoxide, was observed with both SW 4-4 and SW 3-3. SW

1-l had the highest activi ty toward I-chloro-2,4-dinitrobenzene (CDNB) whereas SW 4-4 had

relat ively low CDNB activity. Following pretreatment with 0.75% butylated hydroxyanisole in

the diet, the fraction of total GST contributed by SW I-1 appeared to increase dramatically,

whereas in control mice SW 3-3 constituted the predominant isoenzyme. The high GST activi ty

of mouse liver cytosol toward AFB,-8,9-epoxide is apparently due to an isoenzyme that contrib-

utes little to the overall cytosohc CDNB activity.

o 1990 Academ ic press, hc.

Previous research from this laboratory has

demonstrated that mouse liver cytosol con-

tains very high glutathione S-transferase

(GST) activity toward aflatoxin B,-8,9-epox-

ide, the putative activated form of the potent

hepatocarcinogen aflatoxin B, (AFBi ) (Mon-

roe and Eaton, 1987). In contrast, the specific

activity of rat liver cytosol for the conjugation

This research was presented in part at the 28th An-

nual Meeting of the Society of Toxicology, Atlanta, GA

1989.

To whom correspondence should be addressed at De-

partment o f Environmental Health, SC-34, Universi ty of

Washington. Seattle, WA 98 195.

of AFB,-8,9-epoxide is approximately 1/50th

that of mouse hepatic cytosol even though ac-

tivity toward 1 chloro-2,4-dinitrobenzene

(CDNB) was comparable (Monroe and Ea-

ton, 1987). As part of our investigations of the

biochemical basis for this difference, the stud-

ies reported here address the question of

whether the relatively high Al+-8,9-epoxide

GST activity of mouse liver cytosol is due to

the presence of a GST isoenzyme with partic-

ularly high specific activity.

Three major GST isoenzymes have been

purified from mouse liver (Hatayama et al.,

1986; Warholm et al., 1986) and there is evi-

dence for the existence of additional minor

0041-008X/90 $3.00

Copyright 0 1990 by Academic Press, Inc.

All righ ts of reproduction in any for m reserved.

216


2/10

MOUSE GST CONJUGATION OF AFB AND BPDE

117

forms (Pearson et

al.

1983, 1988; Benson

et

al..

1989). There is little information avail-

able regarding the role of individual mouse

liver GST isoenzymes in the conjugation of

epoxide proximate carcinogens. A previous

study examined the conjugation of AFB,-8,9-

epoxide by cytosol fractions prepared by iso-

electric focusing (Neal

et al.

1987). The high-

est activity was found in a high pI fraction.

suggesting the involvement of a basic GST

isoenzyme. However, GST isoenzymes were

not purified or characterized, so data for spe-

cific activity toward AFB, -8.9-epoxide is not

available.

The selectivity of GST isoenzymes from

the human (Robertson

ec al.

1986) and rat

(Jernstrom

et al.

1985) in the conjugation of

benzo[a]pyrene - 7,8 - dihydrodiol - 9,10 - ep -

oxide (BPDE) has been reported. Thus we

were interested in measuring the conjugation

of BPDE by mouse GST isoenzymes.

Pretreatment of mice with the dietary anti-

oxidant butylated hydroxyanisole (BHA) did

not substantially increase AFB,-8,9-epoxide

conjugation on a per milligram cytosolic pro-

tein basis (Monroe and Eaton, 1988)

whereas cytosolic GST specific activity with

CDNB as the substrate was increased 3- to

1O-fold (Benson et al. 1978; Monroe and Ea-

ton, 1987). We were therefore also interested

in examining the effects of BHA on the rela-

tive content of mouse liver cytosol GST iso-

enzymes with AFB,-8,9-epoxide conjugating

activity. A previous study, using immunoti-

nation, indicated that levels of one of the ma-

jor mouse liver GST isoenzymes increased

following BHA treatment (Pearson et al.

1983) and induction of mRNA coding for

two mouse liver GSTs was subsequently

demonstrated (Pearson et

al.

1988).

We report here the purification and charac-

terization of GST isoenzymes from mouse

liver cytosol and the determination of their

activities toward the epoxide carcinogens

AFB,-8,9-epoxide and BPDE.

MATERIALS AND METHODS

Chemiculs. Organic buf fers. substrates for spectropho-

tometric GST assays. 7(3)-ter/-butyl-4-hydroxyanisole

(mixed isomers), bovine serum albumin. Shexylgluta-

thione agarose, Sephadex G-75, aflatoxin 8,. afla-

toxin G, , NADPH, glucose 6-phosphate, glucose-h-

phosphate dehydrogenase, GSH. and 2-methoxynaph-

thalene were obtained from Sigma Chemical Co. (St.

Louis. MO). HPLC-grade solvents from J. T. Baker Inc.

(Phillipsburg. NJ) were used. All other solvents and salts

were analytical reagent grade from commercial sources.

Benzo[a]pyrene - 4,5 - epoxide

( BaPO) and (+) - u?z//-

benzo[a]pyrene-7.8dihydrodiol-9.10-epoxide were oh-

tained from Midwest Research Institute (Kansas City.

MO). SHexylglutathione was synthesized by a published

method (Mannervik and Guthenberg, 198 1). Antisera lo

rat liver GST isoenzymes were purchased from Medlahs

(Dublin, Ireland) and peroxidase-conjugated anti-rabbit

IgG was from Kirkegaard and Perry Laboratories (&I-

thersburg, MD).

.drzimu/s

trr~d I~~u[I~cII~.\.Male Swiss-Webster mice.

obtained f rom Tyler Laboratories (Bellevue, WA). wcrc

housed on corncob bedding and fed Wayne Rodent Blox

ad lihitum. The animals were divided into two groups,

one of which was fed BHA (0.75% w/w) incorporated

into the ground chow for I2 days before they were killed

for liver fraction preparation.

Animals were killed by cervical dislocation and the Ii\-

ers were excised. All subsequent steps were carried out at

4C. Livers from animals within each group were pooled.

weighed, and homogenized in 2 vol of buffer: 0.25 M su-

crose. 10 mM Tris, 0.2 mM dithioerythritol (DTE), I TTIM

ethylenediaminetetraacetic acid (EDTA). pH 7.4. The

homogenate was centrifuged at I O,OOOg,ollowed by cen-

trifugation of the supernatant at 15.OOOg.Cytosol was

prepared by centrifugation at 105.OOOg and filtered

through cotton gauze to remove lipid.

Glutathiorw S-transftirusc~spurificatiorl. Liver cytosolic

GSTs were purified by modifications of published meth-

ods (Jensson YI (I(., 1982: Alin cf ul.. 1985). Activity of

fractions toward CDNB was determined using a standard

method (Habig et u/. . 1974) adapted to the use of 96-well

microtiter plates and a UV,,, plate reader (Molecular

Devices Corp.. Menlo Park. CA) in the kinetic mode at

a wavelength of 340 nm. The cytosol was applied to a

Sephadex G-75 column (5 X 55 cm) previously equili-

brated with Buffer A (IO mM Tris. 0.2 mM DTE. 1 mM

EDTA. pH 7.8). Following elution with Buffer A. frac-

tions with GST activi ty were pooled and a portion was

applied to a column containing 25 ml of.S-hexylglutathi-

one-agarose (also equilibrated with Buffer A). This

af fini ty column was washed with Buffer A containing 0.2

M NaCl until the effluent was free of-material absorbing

at 280 nm. The GSTs were then eluted with Buffer A

containing 0.2 M NaCI. 2.5 mM GSH, and 5 mM S-

hexyl-GSH. Following concentration by ultrafiltration.

the effluent was dialyzed overnight against 10 mM Tris

containing 0.2 mM DTE (pH 7.8). These steps were com-

pleted within 48 hr after the livers were obtained.

An aliquot of the af fini ty .purified GST was immedi-

ately fractionated by chromatofocusing. The remainder


3/10

218

RAMSDELL AND EATON

TABLE 1

CHROMATOFOCUSING BUFFER

Component

Group 10

Triethylamine

y-Aminobutyric acid

3-(Cyclohexylamino)- 1 propanesulfonic acid

&Alanine

Group 9

Glycine

3-(Cyclohexylamino)-2-hydroxy-I-propanesulfonic acid

2-(N-Cyclohexylamino)ethanesulfonic acid

Histidine

Tamine

Group 8 b

Asparagine

Tris(hydroxymethyl)methylamino propanesulfonic acid

N,N-Bis(2-hydroxyethyl)glycine

Glycylglycine

Tris(hydroxymethyl)aminomethane

N-Tris(hydroxymethyl)methylglycine

Group 7 b

Piperazine-N,N-bis(2-hydroxypropanesulfonic acid)

3-(N-Tris(hydroxymethyl)methylamino)-2-hydroxypropanesulfonic acid

N-2-Hydroxyethylpiperazine-W-(2-ethanesulfonic acid)

N-Tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid

3-(N-Morpholino)propanesulfonic acid

N,N-Bis(2-hydroxyethyl)-2-aminoethanesulfonic acid

Abbreviation

Et,N

GABA

CAPS

P-Ala

GUY

CAPS0

CHES

His

Tau

Asn

Taps

Bicine

GlyGly

Tris

Tricine

POPS0

TAPS0

Hepes

Tes

Mops

Bes

In the buf fer used for chromatofocusing, the final concentration of Group 10 components was 5 mM and those

for the remaining groups were 2 mrvt. Each group was prepared as a 50X stock solution in HPLC-grade HzO. All

components used were free acid and/or free base form except for His (hydrochloride).

* 2-Amino-2-methyl- 1 propanol ( 1M) was added to effectsolubiIi ty of all components.

was stored in aliquots at -80C. A PBE 118 column (1.5

X 28 cm) (Pharmacia-LKB, Piscataway, NJ) was equili-

brated with 25 mM triethylamine-HC1 (pH 11 O) prior to

the application of the GST sample. The column was

eluted using a mixture of low molecular weight buf fers

(Table I), adapted from a focusing buf fer described by

Hutchens and co-workers (Hutchens et al., 1986). An-

ionic GSTs retained on the column at the end of the pH

gradient were eluted with 1 M NaCI. The absorbance of

the effluent was monitored at 280 nm and the CDNB

conjugating act ivi ty of fractions (e2.5 ml) was deter-

mined. The pH of fractions was checked (at room tem-

perature) following the run. Peaks with act ivi ty were

combined, dialyzed against 10 mM Tris with 0.2 mM

DTE (pH 7.8), concentrated, and frozen in aliquots.

Assays. Protein concentrations were determined using

the bicinchoninic acid method with bovine serum albu-

min as the standard (Smith et al., 1985). Sodium dodecyl

sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)

was done using 12.5% separating gels (Laemmli, 1970).

Spectrophotometric assaysof GST act ivi ty using 1 chlo-

ro2.4dinitrobenzene, 2,4dichloronitrobenzene (DCNB),

and ethactynic acid were carried out at 30C by published

procedures (Habig et al., 1974). Peroxidase activity with

cumene hydroperoxide as the substrate was measured by

the method of Igarashi (Igarashi et al., 1983). Glyoxylase

I act ivi ty with methylglyoxal as the substrate was deter-

mined by a published procedure (Mannervik

et al.,

198 I). GST activit ies with BaPO (Eaton and Stapleton,

1989) and AFB,-8,Pepoxide (Monroe and Eaton, 1987)

were determined at 37C by HPLC analysis as described

previously.

GST acti vity with BPDE was determined in incuba-

tion mixtures containing the following: 165 pl 110 mM

Tris-HCL, 0.1 mM EDTA, (pH 7.4), 50 ~19.8 mM GSH

and 25 rd enzyme. Following preincubation for 5 min at


4/10


219

37C the reaction was initiated by adding 10 ~1 BPDE

solution (1 mg/ml in dimethyl sulfoxide). After incuba-

tion for 1 min at 37C, the reaction was stopped by the

addition of 250 ~1 of cold acetonitrile containing 2-meth-

oxynaphthalene as an internal standard. Following stor-

age overnight at -20C 500 ~1 of HZ0 was added and

the sample was centrifuged. Aliquots were analyzed by

reversed-phase HPLC (Econosphere C 18, 150 mm X 4.6

mm column. Alltech) using absorbance detection at 248

nm. The solvents used were A: 0.1% H3P04 adjusted to

pH 3.5 with NH,OH, B: CH&YN, and C: HPLC-grade

H,O. The following gradient program was utilized (flow

rate 1.5 ml/min): 85% A. 15% B for 0.5 min followed by

a 0.5 min linear gradient to 15% B, 85% C. The concen-

tration of B was increased to 30% in 1 min and then to

35% over the next 7 min. The column was then hushed

with 100% B for 2 min before reequilibration. Peak areas

were determined with an integrator and product forma-

tion was quantified relative to the internal standard (cor-

rected for nonenzymatic conjugation) assuming that the

BPDE-GSH conjugate had the same extinction coeffi-

cient as BPDE. Baseline resolution of the GSH conjugate

(retention time = 5.5 mm) from the hydrolysis product

of BPDE (retention time = 7.7 min) was achieved with

this method.

The reactiv ity of antibodies against rat GST isoen-

zymes with the mouse isoenzyme fractions was mea-

sured using an enzyme-linked immunosorbent assay

(ELISA) (Clark et al., 1986). Mouse isoenzyme fractions

were coated overnight at 4C in lmmulon 196-well plates

(Dynatech Laboratories, Chantilly. VA). Plates were

washed after the coating and antibody conjugation steps

with I50 mM NaCI. 10 mM sodium phosphate (pH 7.4)

containing0.05% Tween 20. The plates were treated with

rabbit anti-rat GST sera for 2 hr at room temperature,

washed, and then treated with peroxidase-linked goat

anti-rabbit IgG for I hr. Peroxidase act ivi ty was deter-

mined calorimetrically at 490 nm using a 1,2-diamino-

benzene reagent.

RESULTS

Typical chromatofocusing chromatograms

of affinity-purified GSTs from untreated and

BHA-treated mice are shown in Fig. 1. Three

major ( 1,2, and 4) and four minor (3, 5, and

7) peaks with CDNB activity were identified.

The fraction from untreated mouse liver dis-

played Peak 2 as the major GST isoenzyme

while Peak 4 represented the major isoen-

zyme from BHA-treated mouse liver. The

three major peaks were obtained in appar-

ently homogeneous form as determined by

SDS-PAGE analysis, and apparent M, values

1.0-

; i f : 00 180

. lOOO

E - ____.

o 0.8.

;: ; ;

-800:: :

;

ii $

:: i j

; 0.6- -600 o

0 20 40 60 a0 100 120

Fraction Numbers

0 BHA

0 20 40 60 60 100 120

Fraction Numbers

FIG. I Chromatofocusing of mouse liver glutathione

S-transferases. Atfinity-purified mouse liver cytosol GST

samples from (A) control (chow-fed) and (B) BHA (fed

chow + 0.75% BHA) mice were separated by chromato-

focusing on PBE I I8 ( 1.5 X 28 cm). The samples (5.2 and

6.2 mg protein for A and B. respectively) were applied

following equilibration with 25 mM triethylamine-HCI

(pH 11.0). The pH gradient was generated by elution

with a mixture oflow molecular weight buf fers (Table 1).

The acidic fraction was eluted with 1 M NaCl as noted.

Absorbance at 280 nm was recorded and CDNB act ivi ty

of collected fractions (-2.5 ml) was determined. CDNB

act ivi ty units are equal to the values (in mOD/min) ob-

tained from assaysat room temperature conducted in 96-

well plates with a UV,, plate reader. High ac tiv ity frac-

tions giving off-scale readings are arbitrarily assigned a

value of 1000. The pH of every fourth fraction was mea-

sured at room temperature following completion of the

run.


5/10

220 RAMSDELL AND EATON

for corresponding isoenzymes from control

and BHA-treated mice were the same (data

not shown).

SDS-PAGE analysis of Peak 3 indicated

an apparent subunit M, of 24.5 kDa, the same

as that of Peak 2 (data not shown). Because

Peak 3 was not fully resolved (Fig. I), it was

difficult to obtain sufficient material in pure

form for extensive characterization. Peak 3

appeared to have relatively low activity to-

ward AFBi-8,9-epoxide and was thus omitted

from further studies. Peak 5, eluting at about

fraction 80, may arise from Peak 4 (Mr = 27

kDa). Peak 5 also had an apparent subunit iV&

of 27 kDa and a similar pattern of reactivity

with antisera to rat GST isoenzymes (see be-

low). Upon chromatofocusing of an aliquot

of affinity-purified GST stored for 4 weeks at

-20C Peak 5 was larger by approximately

the same amount that Peak 4 diminished

(data not shown). Apparently an artifact,

Peak 5 provided very limited amounts of pro-

tein and extensive analysis of its properties

was not pursued. The acidic fractions of the

affinity-purified GSTs, giving rise to Peak 7,

apparently contain a small amount of glyoxy-

lase I activity (0.1-0.5 gmol/mg protein/min

with methylglyoxal as the substrate), an en-

zyme which has been shown to be retained by

S-hexyl GSH agarose (Hayes, 1988). Acidic

GST isoenzymes have not been previously

identified in mouse liver cytosol. SDS-PAGE

analysis of Peak 7 protein revealed the pres-

ence of three subunits with apparent M, val-

ues of 24.5, 26, and 27 kDa, indicating that

multiple GSTs were present. Since these were

equal to the sizes observed for the isoenzymes

eluting earlier from the PBE 118 column, it

is possible that the acidic fraction may repre-

sent deamidation products of constitutive

isoenzymes. The protein contained in Peak 7

elutes in a narrower peak than that which is

apparent from the absorbance plot, suggest-

ing that the size of Peak 7 is due in part to a

refractive index effect of the 1 M NaCl eluent.

The GST activities of PBE 118 peaks are

shown in Table 2. Samples from both un-

treated and BHA-treated mice had similar

specific activities toward CDNB. The specific

activities of Peaks 1, 2, and 4 were measured

with an additional series of surrogate sub-

strates (Table 2). The activity pattern for

these three GST forms corresponds to that re-

ported previously for the major GST isoen-

zymes of mouse liver (Warholm

et

al., 1986;

Hayes

et al.,

1987), both in terms of relative

isoenzyme activity toward different sub-

strates and relative activity toward substrates

with a given isoenzyme. Table 2 lists the ap-

parent subunit molecular weights for the iso-

enzyme fractions as determined by SDS-

PAGE, which are also in good agreement

with previously published data (Warholm

et

al., 1986; Hayes et al., 1987). Based on the

properties of the isoenzymes and using no-

menclature suggested for mouse GST isoen-

zymes (Warholm et al., 1986), we have iden-

tified Peaks 1, 2, and 4 as SW 4-4, SW 3-3,

and SW I- 1, respectively (SW denoting

Swiss-Webster). These isoenzymes belong to

classes (Y, 7r, and p, respectively (Mannervik

et al., 1985).

Analysis of the reactivity of the mouse liver

GST fractions with antisera against rat GST

isoenzymes was conducted by ELISA. Be-

cause of the lack of standards and varying ti-

ters among the sera, the results in Table 3 are

normalized to the value of the mouse fraction

with the highest cross-reactivity. The high

cross-reactivity of SW 3-3 with anti-rat sub-

unit 7 (?Fclass) and SW 1- 1 with anti-rat sub-

unit 3 (CLclass) is consistent with previous

studies that used double-diffusion techniques

(Warholm

et al.,

1986; Hayes

et al.,

1987).

The specificity of anti-rat subunit 1 (CX lass)

for SW 4-4 is also similar to an earlier result

(Hayes

et

al., 1987) whereas Warholm and

co-workers ( 1986) observed cross-reactivity

of anti-rat subunit 2 with isoenzyme N 4-4.

The anti-rat subunit 2 sera we used had a

broader range of cross-reactivity. The differ-

ences between our results and previous stud-

ies may be due to the different mouse strains

used in each case or to variations in the speci-

ficity of the antibodies.


6/10


TABLE 2

SPECIFIC ACT IVITIES AND SUBUNIT SIZE OF

MOUSE

LIVER

GSTs

221

Activity

Peak: 1

2

4 6

I

sw4-4 sw3-3

SW l-l

Class O( Class ?r

Class 1

CDNB

Untreated

BHA-treated

DCNB

ECA

CHP

BaPO

Approximate subunit

M, (kDa)

4.00 19.3 114 38.1 14.2

3.23 17.8 156 73.3 22.6


7/10

222

RAMSDELL AND EATON

TABLE 4

MOUSE LIVERGLUTATH~ONES-TRANSFERASEA~IVITYTOWARDAFLATOXINB,-~,~-EPOXIDEAND

BENZO[A]PYRENE-7,&DrHYDRODIOL-9,IO-EPOXIDE

Substrate: AFBr-8,9-epoxide BPDE Act ivi ty ratios (X 1000)

(nmol/mg protein/min) (nmol/mg protein/min)b

AFB,-8,9-epoxide BPDE

Untreated BHA BHA CDNB CDNB

Fraction

Cytosol

Affinity-purified

Peak

1 (SW 4-4, a)

2 (SW 3-3,7r)

4(SW 1-1,/A)

6

7

10.3 +- 0.4 10.2 * 0.3 129a5 1.7 7d

54i3 36+4 480 f 20 0.66 6

198&2 260 f 30 440 + 30 49 136

3.1 kO.9 4.7 * 0.3 540 k 10 0.18 30

1.3 f 0.4 0.40 + 0.02 39 k 5 0.016 0.25

1.1 1.1 43 f 3 0.029 1.8

8.7 7.8 38 +4 0.61 1.7

Values for Peaks 1.2, and 4 are means (*SE) obtained in two separate experiments (N = 5). Differences between

control and BHA preparations were not statistically significant (p > 0.05, two-sided t test). Cytosol and affinity-

purified data are means (*SE) of triplicate assaysand the others are averages from duplicate determinations.

b Values are means (*SE) of triplicate determinations.

cAct ivi ty of AFB,-8,9-epoxide conjugation (in nmol/mg proteimmin) divided by CDNB activ ity (in Fmol/mg

protein/min). Values are derived from untreated mouse data.

d Act ivi ty o f BPDE conjugation (in nmol/mg protein/mm) divided by CDNB activ ity (in pmol/mg protein/min).

Values are derived from BHA mouse data.

with (-)-BPDE and in some cases the latter

may inhibit conjugation of the former (Rob-

ertson et al., 1986). Because racemic BPDE

was used in our experiments, the GST spe-

cific activities toward the more carcinogenic

(+)-BPDE may be higher than those shown

in Table 4.

The ratios of isoenzyme specific GST activ-

ity with AFB,-8,9-epoxide and BPDE to the

values obtained with CDNB as the substrate

are also shown in Table 4. These ratios indi-

cate the poor predictive value of CDNB as a

surrogate GST substrate for these epoxide

carcinogens. Because SW 4-4 has the lowest

activity with CDNB but the highest with

AFB,-8,9-epoxide, CDNB is a particularly

poor surrogate substrate for AFB,-8,9-epox-

ide. Likewise, CDNB is not a good surrogate

substrate for BPDE with mouse liver GST

isoenzymes.

The effect of BHA pretreatment on the rel-

ative amounts of GST isoenzymes in mouse

liver is indicated qualitatively in Figs. 1A and

1B. Similar amounts of protein were applied

to the column in each case. Assuming that re-

covery of GST protein was similar for corre-

sponding peaks from control and BHA-

treated mouse preparations, the most striking

change is in the contributions of Peaks 2 and

4 (SW 3-3 and SW l-l, respectively). It ap-

pears that SW l-l is induced by BHA pre-

treatment, becoming the principal isoenzyme

found in the affinity-purified fraction. It also

appears that Peak 1 (SW 4-4) is larger follow-

ing BHA pretreatment. It should be noted

that the patterns observed in Figs. 1A and 1B

may not precisely reflect the isoenzyme con-

tent of cytosol, as there may be selective loss

in the isolation procedure through the affinity

purification step. A lower recovery of SW 4-

4 than the others is suggested by an average

recovery of total CDNB activity present in

mouse cytosol of 54% in the affinity-purified

material compared with an average recovery


8/10


223

of AFB,-8,9-epoxide conjugation activity of

34% (data not shown).

DISCUSSION

The major GST isoenzymes we have iso-

lated from mouse liver, SW 4-4, SW 3-3, and

SW l-l (Peaks 1,2, and 4, respectively), corre-

spond in their subunit it& and substrate speci-

ficities (Table 2) to those prepared by similar

methods (MI, MB, and MIII, respectively) as

reported previously for another mouse strain

(Warholm et al., 1986). In another study

(Hayes et al., 1987). mouse liver GSTs were

separated by hydroxyapatite chromatogra-

phy and two fractions, P4 and P5, had prop-

erties very similar to those of SW 4-4 and SW

1 1, respectively. Three peaks that differed lit-

tle in substrate specificity were partially sepa-

rated by hydroxyapatite chromatography

and were all very similar in their properties to

those ofSW 3-3.

Glutathione conjugation of the activated

epoxides of two carcinogens, aflatoxin B, and

benzo[a]pyrene, was measured for the isoen-

zymes isolated in this study. The Class (Y so-

enzyme SW 4-4 had the highest AFB,-8,9-

epoxide-conjugating activity (Table 4). The

activity of SW 4-4 toward CDNB (Table 2) is

relatively low compared with that of SW 1 I,

an isoenzyme with little activity toward

AFB,-8,9-epoxide. These observations and

the apparent increase in SW 1-l content fol-

lowing BHA treatment are consistent with

previous observations in our laboratory that

BHA pretreatment of mice failed to substan-

tially affect total cytosolic GST specific activ-

ity toward AFB,-8,9-epoxide, whereas CDNB

activity was induced severalfold (Monroe and

Eaton, 1987).

The isoenzyme peak that appeared to in-

crease the most following BHA treatment

(SW l-l) had the lowest activity toward

BPDE (Table 4). This is consistent with previ-

ous results showing that BHA increased the

apparent K, of mouse liver cytosol for BPDE

conjugation, suggesting that isoenzyme(s)

were induced that had relatively low affinity

for BPDE (Dock et al., 1984). The BPDE re-

sults also illustrate the limitations of the use

of surrogate substrates (Table 4). CDNB is a

poor surrogate substrate for BPDE in cyto-

solic assays due to the relatively high rate of

BPDE conjugation by SW 4-4 and SW 3-3

(Table 4), isoenzymes with relatively low

CDNB activity. Even the 4,5-epoxide deriva-

tive of benzo[a]pyrene (BaPO) is a poor indi-

cator of BPDE activity because SW 1- 1 has

the highest activity for BaPO conjugation

(Table 2) but has relatively low activity with

BPDE (Table 4).

Pretreatment of mice with BHA appeared

to increase the relative content of two of the

major GST isoenzymes, SW 4-4 and SW I- 1

(Fig. 1). A GST isoenzyme from CD- 1 mouse

liver (GT 8.7 or C l-l), apparently homolo-

gous to SW l-l (Warholm et al., 1986) was

induced 12-fold by BHA pretreatment as in-

dicated by immunotitration (Pearson et ai..

1983). Induction of a Class acmRNA by BHA

was also demonstrated (Pearson et al., 1988).

However, a recent study suggested that.

rather than causing an increase in C 4-4 (GT

10.6) BHA induces the Class (Y isoenzyme

GT 10.3 (Benson et al., 1989). It is unlikely

that our purification method would have sep-

arated a GT 10.3 homologue from SW 4-4.

Therefore, we cannot rule out the possibilit

that such an isoenzyme may have activity to-

ward AFB,-8,9-epoxide.

We did not observe an isoenzyme homolo-

gous to C 2-2 (GT 9.3). Although such a form

would be expected to elute before SW 1- 1,

Peak 3 is not likely to represent a homologue

because its subunit M, was equal to that of

SW 3-3 rather than to that of SW l-l. The

same subunit M, was observed for C 1- 1 and

C 2-2 (Pearson et al., 1983).

There appear to be differences in the levels

of cytosolic GST activity among mouse

strains (Wheldrake et al., 198 1: McLellan

and Hayes. 1987; Makary et al., 1988).

Differences were observed in the levels of cu-

mene hydroperoxidase activity, apparently

representative of mouse 4-4 isoenzyme activ-


9/10

224 RAMSDELL AND EATON

ity (Table 2; Warholm et al., 1986), in cyto-

sols from the different strains (McLellan and

Hayes, 1987). Thus, it is possible that there

are also differences in the ability of different

mouse strains to conjugate AFT+-8,9-epoxide.

SW 4-4 appears to correspond to a Class (Y

mouse liver GST (C 4-4) which, by compari-

son of cDNA clone sequence data, has 95%

amino acid sequence identity with a rat GST

l-1 (Pearson et al., 1988). Relative to mice,

rats have much lower cytosolic GST activity

toward AFB,-8,9-epoxide (Monroe and Ea-

ton, 1987) even though rat GST I- 1 com-

prises a substantial proportion of total liver

GSTs (Alin

et al.,

1985). This suggests that

the high activity of SW 4-4 toward AFB,-8,9-

epoxide is due to a relatively small difference

in the amino acid sequence compared to that

of the rat isoenzyme.

The high activity of SW 4-4 appears to be

responsible for the high rate of detoxification

of AFB,-8,Pepoxide by mouse liver frac-

tions. This GST isoenzyme is thus an impor-

tant factor in the resistance of mice to the car-

cinogenic effects of AFBl .

ACKNOWLEDGMENTS

We thank Dennis Slone for excellent technical assis-

tance and Dr. Zhi-Ying Chen for advice on ELISA meth-

ods. This work was supported by NIH Grants T32 ES-

07032, ES-03933, ES-04696, and CA-4756 I.

REFERENCES

ALIN, P., JENSSON,H., GUTHENBERG, C., DANIELSON,

U. H., TAHIR, M. K., AND MANNERVIK, B. (1985).

Purification of major basic glutathione transferase iso-

enzymes from rat liver by use of affinity chromatogra-

phy and fast protein liquid chromatofocusing. Anal.

Biochem. 146,3 13-320.

BENSON, A. M., BATZINGER, R. P., Ou, S.-Y. L., BUE-

DING, E., CHA, Y.-N., AND TALALAY, P. (1978). Ele-

vation of hepatic glutathione S-transferase activi ties

and protection against mutagenic metabolites of

benzo[a]pyrene by dietary antioxidants. Cancer Res.

38,4486-4495.

BENSON, A. M., HUNKELER, M. J., AND YORK, J. L.

(1989). Mouse hepatic glutathione transferase isoen-

zymes and their differential induction by anticarcino-

gens. Biochem J. 261, 1023-1029.

CLARK, M. F., LISTER, R. M., AND BAR-JOSEPH, M.

(1986). ELISA techniques. In Methods in Enzymology

(A. Weissbach and H. Weissbach. Eds.). Vol. 118, pp.

742-766. Academic Press, San Diego.

DOCK, L., MARTINEZ, M., AND JERNSTROM, B. (1984).

Induction of hepatic glutathione S-transferase act ivi ty

by butylated hydroxyanisole and conjugation of

benzo[a]pyrene diol-epoxide. Curcinogenesis 5, 84 l-

844.

EATON, D. L., AND STAPLETON, P. L. (1989). Simulta-

neous determination of cytosolic glutathione S-trans-

ferase and microsomal epoxide hydrolase activity to-

ward benzo[a]pyrene-4,5-oxide by high-performance

liquid chromatography. Anal. Biochem. 178,153-l 58.

HABIG, W. H., PABST, M. J., AND JAKOBY, W. B. (1974).

Glutathione S-transferases. The f irs t enzymatic step in

mercapturic acid formation. J. Biol. Chem. 249,7 130-

7139.

HATAYAMA, I., SATOH, K., AND SATO, K. (1986). Devel-

opmental and hormonal regulation of the major form

of hepatic glutathione S-transferase in male mice. Bio-

them. Biophys. Rex Commun. 140,58 l-588.

HAYES, J. D. (1988). Selective elution of rodent glutathi-

one S-transferases and glyoxylase I from the Shexyl-

glutathione-Sepharose af fin ity matrix. Biochem. J.

255,9 13-922.

HAYES,

J.D..

COULTHWAITE,

R.E..

STOCKMAN,

P.K.,

HUSSEY, A. J., MANTLE, T. J., AND WOLF, C. R.

(1987). Glutathione S-transferase subunits in the

mouse and their catalytic activities towards reactive

electrophiles. Arch. Toxicol. Suppl. 10, 136-146.

HUTCHENS, T. W., LI, C. M., AND BESCH, P. K. (1986).

Development of focusing buffer systems for generation

of wide-range pH gradients during high-performance

chromatofocusing. J. Chromatogr. 359, 157- 168.

IGARASHI, T., SATOH, T., UENO, K., AND KITAGAWA,

H. (1983). Sex related difference in the hepatic gluta-

thione level and related enzyme activit ies in rat. J. Bio-

them. 93,33-36.

JENSSON, H., ALIN, P.. AND MANNERVIK, B. (1982).

Separation of glutathione S-transferases from rat liver

cytosol by chromatofocusing. Actu Chem. Stand. Ser.

B. 36,205-206.

JERNSTROM, B., MARTINEZ, M., MEYER, D. J.. AND

KETTERER, B. ( 1985). Glutathione conjugation of the

carcinogenic and mutagenic electrophile (-t)-7&8cr-

dihydroxy-9ru, IOa-oxy-7,8,9,1 O-tetrahydrobenzo[u]py-

rene catalyzed by purified rat liver glutathione trans-

ferases. Curcinogenesis 6,85-89.

LAEMMLI, U. K. (1970). Cleavage of structural proteins

during the assembly of the head of bacteriophage T4.

Nature (London) 227,680-685.

MAKARY, M., KIM, H. L. , SAFE, S., WOMACK, J ., AND

IVIE, G. W. (1988). Constitutive and Arochlor 1254-

induced hepatic glutathione S-transferase, peroxidase


10/10


225

and reductase activi ties in genetically inbred mice.

Camp. Biochem. Physiol. 91C, 425-429.

MANNERVIK. B., ALIN, P., GUTHENBERG, C., JENSSON.

H.. TAHIR, M. K.. WARHOLM, M., AND JORNVALL,

H. (I 985). Identification of three classes of cytosolic

glutathione transferase common to several mamma-

lian species: Correlation between structural data and

enzymatic properties. Pro

Documents

Mouse Liver Glutathione S-Transferase Lsoenzyme Activity