THE JOURNAL OF BIOLOCKXL CHEMISTRY 0 1990 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 265, No. 13, Issue of May 5, pp. 7150-7157.1990 Printed ,n U.S. A.

Purification and Characterization of Glutathione Transferases with an Activity toward Nitroglycerin from Human Aorta and Heart MULTIPLICITY OF THE HUMAN CLASS Mu FORMS*

(Received for publication, October 2, 1989)

Shigeki Tsuchida, Takako Maki, and Kiyomi Sato$ From the Second Department of Biochemistry, Hirosaki University School of Medicine, Hirosaki 036, Japan

Although recent studies suggest involvement of glu- tathione transferase (GST) of blood vessels in vasodi- lation by nitroglycerin, GST forms in blood vessels remain to be studied. In this study, three GST forms (p1 values 8.3,6.6, and 4.8) were purified from human aorta and four (PI values 6.0, 5.6, 5.3, and 4.6) from the heart by affinity chromatography followed by chromatofocusing. The major form of both aorta (p1 4.8) and heart (p1 4.6) was identified as GST-?r, and the other five forms were immunologically related to GST-p, suggesting that the five belong to the Mu class. Among nine human GST forms, including three in the Alpha class purified from the liver, GST-p, aorta p1 8.3 form, and GST-I (a form of the Alpha class, corre- sponding to GST-c (BIBI)) showed high activities to- ward nitroglycerin, 1.08,0.85, and 0.78 units/mg pro- tein, respectively. GST-r did not exhibit the activity. The K, values of the aorta form (~18.3) for glutathione (GSH) and nitroglycerin were calculated as 0.12 and 1.1 mM, respectively. The K,,, values of GST-r and GST-I for GSH were 0.29 and 0.09 mM, and those for nitroglycerin were 2.5 and 0.3 mM, respectively. The activity of the p1 8.3 form as well as GST-p toward nitroglycerin was inhibited by bromosulfophthalein, which is known to inhibit the relaxation of rabbit aorta induced by nitroglycerin, at the lower concentration (I&,, 2 PM) than was GST-I (I&,, 32 PM).

Two-dimensional gel electrophoresis and N-terminal amino acid sequence analysis revealed that five forms in the Mu class are homo- or heterodimers of five different subunits named M1 (p1 7.0/M, 27,000), Mz (6.6/27,000), MB (6.0/27,000), N1 (6.5/26,500), and N2 (5.9/26,500). The subunit structures of the five forms are as follows: ~18.3 form, M1M2; 6.6 form, M2N1; 6.0 form, M3M3; 5.6 form, M3Nz; and 5.3 form, N2N2. M3 and Nz seem to correspond to the subunits of GST-W, and -4 (Board, P. G., Suzuki, T., and Shaw, D. C. (1988) Biochim. Biophys. Acta 953, 214-217), re- spectively. These subunits except N1 are different from each other at two or three positions in the first 20 residues of N-terminal amino acid sequence. These results indicate the presence of five different subunits in the human Mu class and also suggest that GST-M1M2 and -M2N1 found in the aorta are involved in the

* This work was supported in part by a Grant-in-Aid for Cancer Research from the Ministry of Education, Science and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be heieby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence should be addressed: Second Depart- ’ The abbreviations used are: GST, glutathione transferase; CDNB, ment of Biochemistry, Hirosaki University School of Medicine, Zaifu- 1-chloro-2,4-dinitrobenzene; SDS, sodium dodecyl sulfate; BSP, bro- cho 5, Hirosaki 036, Japan. mosulfophthalein.

expression of the pharmacologic effect of nitro- glycerin.

Glutathione transferases (EC 2.5.1.18, GSTs)’ are a family of multifunctional dimeric proteins, and multimolecular forms are known to exist in the various organs of different species (for recent reviews, see Refs. l-5). These forms are involved in metabolism of not only exogenous but also endogenous substances, including lipid hydroperoxides, prostaglandins, and leukotriene Aq (l-5). They are grouped into three classes, Alpha, Mu, and Pi, in species-independent classification (6). In the rat, six subunits have been reported in the Mu class: subunits 3, 4, 6 (Ynl (7, 8) or Yb, (9)), 9 (Yn2 (8)), 11 (lo), and Yb, (11). As human class Mu forms, only two forms, GST-p (12,13) and GST-4 (14), have been well characterized.

Nitroglycerin has still been one of the most commonly used organic nitrates for the treatment of angina pectoris. Habig et al. (15) showed that some GST forms catalyze the formation of equivalent amounts of nitrous acid, glyceryl dinitrate, and oxidized glutathione from nitroglycerin and GSH. This reac- tion has been considered as the detoxication process of nitro- glycerin in the liver (16), and Keen et al. (17) reported that rat GST-C (3-4 in the new nomenclature (18)) and GST-AA (2-2), and human GST-b present in the liver have high activities toward nitroglycerin. However, recent studies (19- 21) suggest that nitrous acid and glyceryl dinitrate thus formed play an important role in vasodilation by nitro- glycerin. Yeates et al. (22) have recently shown that BSP, an inhibitor of GST, inhibits the relaxation of rabbit aorta in- duced by nitroglycerin. Pharmacologic effects of nitroglycerin have been examined in the aorta (22-24) and other blood vessels (19). Distribution of GST forms has been studied in many organs (l-5), but the forms in blood vessels remain to be studied. We, therefore, studied GST forms in human aorta and heart, and examined their activities toward nitroglycerin. Two affinity matrices, S-hexylglutathione- and glutathione- Sepharose were employed to purify the forms, since a rat form, 8-8, is reported to show a preferential affinity for glutathione-Sepharose (25). In this paper we report new forms purified from the aorta, which have a high activity toward nitroglycerin, and also present a line of evidence to indicate that five different subunits are present in the human Mu class, and their binary combinations make five at least homo- and heterodimers.

7150

by guest on July 1, 2018http://w

ww

.jbc.org/D

ownloaded from

Human Aorta and Heart GSTs with Activity to Nitroglycerin 7151

EXPERIMENTAL PROCEDURES

Materials-Epoxy-activated Sepharose 6B, Polybuffer exchangers (PBE 118 and 94), Pharmalyte S-10.5, and polybuffer 74 were ob- tained from Pharmacia LKB Biotechnology Inc.; nitrocellulose mem- brane and anti-rabbit IgG conjugated with horseradish peroxidase from Bio-Rad; Ampholine (pH 3.5-10) from LKB; yeast glutathione reductase from Oriental Yeast Co., Ltd., Tokyo, Japan. Nitroglycerin was kindly supplied at a 2.2 mM concentration from Nippon Kayaku Co. Ltd., Tokyo, Japan. All other chemicals were of analytical grade.

Human organs, aorta and heart, were obtained by autopsy within 6 h after the death of patients. These organs were stored at -80 “C and used for the purification of GSTs within 4 weeks.

Purification of GSTs from Human Aorta-Aorta (70 g) minced by a meat grinder was homogenized with four volumes of 10 mM Tris- HCI, pH 7.8, containing 0.2 M NaCl (homogenizing buffer) by a Polytron homogenizer (Ystral GmbH, type X-1020, Goettingen, West Germany). The supernatant obtained by centrifugation at 105,000 x g for 45 min was applied to an S-hexylglutathione-Sepharose column (1.2 x 8 cm) equilibrated with the homogenizing buffer. S-Hexylglu- tathione-Sepharose was prepared according to Guthenberg and Man- nervik (26). After washing the column with the same buffer, the adsorbed GSTs were eluted with the buffer containing 5 mM S- hexylglutathione (26). These GST forms were resolved by chroma- tofocusing at pH 7.4-4 as described previously (7, 8, 27). The flow- through fractions from the column were then applied to a glutathione- Sepharose column (2 x 10 cm) equilibrated with the homogenizing buffer. Glutathione-Sepharose was prepared according to Simons and Vander Jagt (28). GST forms bound to this affinity column were eluted with 25 mM GSH in 50 mM Tris-HCl, pH 7.8, containing 0.2 M NaCl, and resolved by chromatofocusing as described above.

Purification of Human Heart and Liver GSTs-Seventy g of human heart (mainly heart muscle) was homogenized with four volumes of 10 mM Tris-HCl, pH 7.8, containing 0.2 M NaCl. The supernatant obtained by centrifugation at 105,000 x g for 45 min was subjected to S-hexylglutathione-Sepharose affinity chromatography followed by chromatofocusing at pH 7.4-4.

Human GST-I, -II, -IV, and -p (III) were purified from the liver and GST-r from the placenta as described previously (27). GST-I, -11, and -IV correspond to GST-BiBi (0, -BIB2 (a), and -B2B2 (a-y), respectively (29, 30), and all these forms belong to the Alpha class

and have a same subunit M, on SDS-polyacrylamide gel electropho- resis (27).

Assay for GST Activity-GST activity was assayed with I-chloro- 2,4-dinitrobenzene (CDNB) and other substrates according to Habig et al. (31) or Keen and Jakoby (32). One unit of GST activity is the amount of enzyme catalyzing the conjugation of 1 pmol of substrate/ min at 25 “C. Glutathione peroxidase activity was determined accord- ing to the method of Paglia and Valentine (33) as modified by Lawrence and Burk (34). One unit of peroxidase activity is the amount of enzyme catalyzing the conversion of 1 pmol of NADPH to NADP/ min at 25 “C.

Since GSTs catalyze the formation of equivalent amounts of ni- trous acid, glyceryl dinitrate, and oxidized glutathione from nitro- glycerin and GSH (15, 16), an activity toward nitroglycerin was assayed by quantitating oxidized glutathione essentially according to Keen et al. (17). Oxidized glutathione formed by GST was coupled with a conversion reaction of NADPH to NADP by glutathione reductase, and the activity was determined from the decreased amount of NADPH/min as glutathione peroxidase activity. In this method the decrease of NADPH was dependent on enzyme amount, and the nonenzymatic decrease was not detected at 1 mM nitro- glycerin and 1 mM GSH.

Protein Content-Protein content was measured by the methods of Lowry et al. (35) or Bradford (36) using bovine serum albumin as a standard.

Electrophoresis-SDS-polyacrylamide gel electrophoresis was per- formed according to the Laemmli’s method (37). Two-dimensional gel electrophoresis was carried out as reported earlier (7). Protein was stained with Coomassie Brilliant Blue R-250. Immunoblot was per- formed according to Towbin et al. (38) using anti-GST-p or anti- GST-* antibodies.

Antibody Preparations-Antibodies to GST-F, GST-*, and the aorta form (~18.3) were raised in rabbits as described previously (27). Double immunodiffusion was carried out according to Ouchterlony (39).

N-terminal Amino Acid Sequence Analysis-Edman degradations were performed with a gas-phase protein sequencer (Applied Biosys- terns model 470A, Foster, CA), The amino acid phenylthiohydantoin derivatives were identified and quantified by high performance liquid chromatography with a Spectra Physics model SP-8100 (San Jose, CA).

TABLE I

Purification of GST forms from human aorta The starting supernatant prepared from homogenate of 76 g of human aorta by centrifugation at 105,000 X g,

for 45 min. was applied to a column of S-hexvlglutathione-Sepharose. GST forms bound to the column were eluted with 5 mM S-hexylglutathione and then resol&d by chromatofocusing (A). GST forms recovered in flow-through fractions of the affinity chromatography were further applied to a column of glutathione-Sepharose and eluted with 25 mM glutathione (B). GST forms were also resolved by chromatofocusing at pH 7.4-4 and then at pH ll- 8. The data shown are from one preparation and are similar to results obtained in two other experiments. The details of procedures are described under “Experimental Procedures.” Activity is thaF toward l-chloro-2,4- dinitrobenzene.

Step A

Volume

ml

Total activity

units

Total protein

m&T

Specific activity

unitsfmg

105,000 X g, supernatant 320 451 4,860 0.09 S-Hexvl-GSH-Sepharose 31 304 3.10 98.2 Chromatofocusing pH 7.4-4

p1 over 7.4 18 3.03 p1 6.2 13 1.11 p1 5.8 18 2.35 p1 5.6 18 2.37 p1 4.8 65 239 2.02 119

Step B 105,000 X g, supernatant S-Hexyl-GSH-Sepharose unbound GSH-Sepharose bound Chromatofocusina RH 7.4-4

- - p1 over 7.4 p1 7.2 PI 6.6 pI 4.8

Chromatofocusing pH 11-8 n1 8.3

320 451 4,860 0.09 300 141 4,350 0.03

36 125 2.48 50.3

44 56.4 1.72 32.8 8 3.5

23 13.7 0.30 46.2 26 31.7

60 32.7 1.02 32.4

by guest on July 1, 2018http://w

ww

.jbc.org/D

ownloaded from

7152 Human Aorta and Heart GSTs with Activity to Nitroglycerin

RESULTS

Separation of GSTs from Human Aorta-GST activity to- ward CDNB of human aorta cytosol was 6.4 units/g of aorta tissue. About 67% GST activity of cytosol was bound to an S-hexylglutathione-Sepharose affinity column. These GST forms were resolved into five peaks by chromatofocusing at pH 7.4-4 (Table IA). Among them p1 4.8 form was the dominant (about 79% activity of bound fractions to the col- umn) and identified as GST-R by two-dimensional gel elec- trophoresis and immunoblot with anti-GST-* antibody (data not shown). Other minor peaks were identified as GST-I (PI, over 7.4, Ref. 27), GST-p (p1 6.2), and GST-IV (p1 5.8). A peak with ~15.6 was not characterized because of only a trace amount of protein. Unbound GSTs to S-hexylglutathione- Sepharose were further purified using a glutathione-sepha- rose column (Table IB). GST forms bound to this column were separated into four peaks by chromatofocusing at pH 7.4-4 (Fig. 1). The p1 values of three peaks were 7.2, 6.6, and 4.8. A major peak, about 46% of GST activity bound to glutathione-Sepharose, had a p1 value of over 7.4, and then was subjected to the second chromatofocusing at pH 11-8. It was eluted at pH 8.3 (Table IB).

These GST forms were subjected to SDS-polyacrylamide gel electrophoresis and their M, were compared with those of marker GST subunits; GST-p (27,000), GST-I (26,000, ex- pressed as a in Fig. 2A), and GST-?r (24,500, Ref. 27). The p1 8.3 form as well as a minor form (p1 7.2) gave a subunit (Mr 27,000) which had a similar M, to the subunit of GST-k (lanes 2 and 3, Fig. 2.4). The p1 6.6 form (lane 4) gave two subunits; one had a similar M, to that of the GST-p subunit, and another had M, of 26,500, slightly larger than cy subunit. The ~14.8 form gave a subunit (Mr 24,500) similar to that of GST- x (lane 5). Although the pIs 8.3, 7.2, and 6.6 forms had more basic pIs in chromatofocusing than GST-p (p1 6.0, Ref. 27), immunoblot analysis revealed that the subunits of these three forms were immunologically related to the subunit of GST-p (Fig. 2B), suggesting that these belong to the Mu class. The ~14.8 form of glutathione-Sepharose bound fractions was also identified as GST-r by immunoblotting with anti-GST-a antibody (Fig. 2C, lane 5).

Separation of GSTs from Human Heart-About 92% GST

4

FIG. 1. Separation of human aorta GSTs bound to glutathi- one-Sepharose by chromatofocusing at pH 7.4-4. GSTs un- bound to an S-hexylglutathione-Sepharose column were applied to a glutathione-Sepharose column (2 x 10 cm) equilibrated with 10 mM Tris-HCl, pH 7.8, containing 0.2 M NaCl. The GSTs eluted with 25 mM GSH in 50 mM Tris-HCl, pH 7.8, containing 0.2 M NaCl were dialyzed against 25 mM imidazole-HCl, pH 7.4, and then applied to a column (1 x 16 cm) of chromatofocusing gel PBE 94 equilibrated with the same buffer. Elution was carried out as described previously (27). Flow rate was 15 ml/h and fractions of 2.6 ml were collected. The breakthrough fractions (No. 6-20) were further applied to a column of chromatofocusing gel PBE 118 (27), and its p1 (8.3) was determined. l , GST activity toward 1-chloro-2,4-dinitrobenzene (CDNB); - - -, pH.

A B 1234

C 5 1 2 3 4 1 5

,:- ., ,;; ..,. ;..:;

FIG. 2. SDS-polyacrylamide gel electrophoresis of aorta GST forms resolved by chromatofocusing. SDS-polyacrylamide gel electrophoresis was performed in a 12.5% acrylamide gel. A, protein staining; B and C, immunoblots with anti-GST-p and anti- GST-r antibodies, respectively. Lane I, 3 pg each of homogeneous preparations of GST-I (Alpha class), -cc, and -n; lane 2, pI 8.3 form (10 pg); lane 3, pI 7.2 form (fraction 30 in Fig. 1, 2 pg); lane 4, p1 6.6 form (fraction 62, 5 rg); and lane 5, pI 4.8 form (fraction 112, 5 pg).

activity of heart cytosol was bound to an S-hexylglutathione- Sepharose column. These GST forms were separated into four peaks by chromatofocusing at pH 7.4-4. The p1 values of these peaks were 6.0, 5.6, 5.3, and 4.6, respectively (Table II), and the purification steps are summarized in Table II. The p1 4.6 form occupied 56% of the total GST activity bound to S- hexylglutathione-Sepharose. SDS-polyacrylamide gel electro- phoresis revealed that the subunits of p1 6.0 (lane 1, Fig. 3A) and 4.6 (lane 4) forms had similar M, to those of GST-p (lane 5) and GST-7r (lane 7), respectively. Although the ~14.6 form was slightly more acidic than the aorta GST-a (p1 4.8 in chromatofocusing, Table IA and B), both the heart and aorta forms gave the same spot on two-dimensional gel and the heart form was also identified as GST-* by the immunoblot- ting (data not shown). The ~15.3 form (lane 3) gave a subunit with M, 26,500. The p1 5.6 form (lane 2) consisted of two subunits; one was the same as that of GST-p and another was as that of the p1 5.3 form. Immunoblot with anti-GST-p antibody showed that subunits of p1 values 6.0, 5.6, and 5.3 forms were immunologically related to GST-p (Fig. 3B). Al- though unbound fractions of heart cytosol to S-hexylgluta- thione-Sepharose were also applied to a glutathione-sepha- rose column, no forms differing from those bound to S- hexylglutathione-Sepharose were detected.

Activities of Human GST Forms toward Nitroglycerin- Activities toward nitroglycerin of human GST forms, includ- ing those from the aorta and heart, were summarized in Table III. GST-p had the highest activity, 1.08 units/mg protein, among nine forms. An aorta form (p1 8.3) and GST-I, a form of the Alpha class in the liver (27), had a high activity, 0.78 and 0.85 units/mg protein, respectively. GST-7r, the major form of human aorta and heart (Tables I and II), or a heart form with a p1 5.3 did not exhibit the activity. The p1 values 6.6 and 5.6 forms showed lower activities than GST-p or p1 8.3 form. GST-II, a heterodimer consisting of subunits of GST-I and -IV (27), showed the intermediate activity between those of GST-I and -IV and its value was similar to that of GST-6 reported by Keen et al. (17).

From the Lineweaver-Burk plot (Fig. 4), the K, values of the aorta form (p1 8.3) for GSH and nitroglycerin were cal- culated as 0.12 and 1.1 mM, respectively. The V,,, value was calculated as 1.54 units/mg protein from Fig. 4B, about 2-fold higher than the value obtained at 1 mM nitroglycerin (Table III). The K, values of GST-I and -p for GSH were 0.09 and 0.29 mM, and those for nitroglycerin were 0.3 and 2.5 mM,

respectively. The K, value of GST-I for nitroglycerin obtained in this study is similar to the value of GST-6 (0.5 mM) reported by Keen et al. (17). The V,,, values of GST-I and -p were

by guest on July 1, 2018http://w

ww

.jbc.org/D

ownloaded from

Human Aorta and Heart GSTs with Activity to Nitroglycerin 7153

TABLE II

Purification of (;S’l’/orms from human heart The starting supernatant prepared from homogenate of 30 g of human heart by centrifugation at 105,000 X R.

for 45 mm, was used for the purification of GST forms as described in Table I.

Step Volume rota1 act1v1tv Total protein Specific act1v1t>

105,000 X 6, supernatant S-Hexvl-(;SH-Senharose

ml urllt\ mg unrt\/mg 230 340 .5,750 0.06

30 321 3.98 80.7 Chromatofocusini pH 7.4-4

p1 6.0 12 pI 5.6 12 pI 5.3 15 p1 4.6 45

A B 12 3 4 5 6 7 12 34

FIG. 3. SDS-polyacrylamide gel electrophoresis of heart GSTs. A, protein staining. Lane I, p1 6.0 form in Table II (5 pg); lnnr 2, p1 5.6 form (6 tip); lane 3, p1 5.3 form (5 pg); lane 4, p1 4.6 form (5 pg); lane .5, GST-p (10 ~.rg); lane 6, GST-I (IO pg); and lane 7, GST-x (6 pg). H, immunohlot with anti-GST-p antibody. Lane I, GST-p (10 pg): lane 2, p1 6.0 form (6 pg); lane 3, p1 5.6 form (6 fig); and lane 4, p1 5.3 form (5 pg).

TABLE III Activif,y of human GST forms to~rd nitroglycerin

The reaction mixture contained 50 mM potassium phosphate (pH 7.0): 1 mM EDTA, 0.2 mM NADPH, 1 unit of glutathione reductase, 1 mM GSH, 1 mM nitroglycerin, and GST in a total volume of 1 ml. Oxidized glutathione formation from GSH and nitroglycerin by GST was coupled with conversion of NADPH to NADP in the presence of glutathione reductase, and activity toward nitroglycerin was assayed by measuring the decreased amount of NADPH/min from absorbance at 340 nm. GST-I, -11, -IV, and -p (~1 6.0) were purified from human liver. and GST-T from the placenta, as reported previously (27).

Form Activity toward nitroelvcerin

GST-I GST-II GST-IV

p1 8.3 p1 6.6 PI 6.6 (rc) p1 5.6 p1 5.3

GST-x

unlt.\/mgproteln 0.85 0.48 0.17 0.78 0.43 1.08 0.43

co.02 co.01

also calculated from Fig. 4B as 1.18 and 3.33 units/mgprotein, respectively.

Inhibition of GST Activity toward Nitroglycerin by Bromo- sulfophthalein (BSP)-BSP, an inhibitor of GST, especially for the Mu class forms (l), has been used to examine the involvement of GST in vasodilation by nitroglycerin (21, 22). We examined, therefore, the inhibitory effect of BSP on activities toward nitroglycerin of GST-I, -F, and the p1 8.3 form (Fig. 5). The activity of the p1 8.3 form as well as GST- p was inhibited by BSP at the lower concentration (IC& value, 2 PM) than was GST-I of the Alpha class (I&, 32 FM).

Two-Dimensional Gel Electrophoresis of Human Class Mu

4.5 19.2 0.23 83.5 15.3 0.21 72.9

180 1.50 120

:GSH] -’ mM -’ :NlVOQlyCerl": -' ",M -'

FIG. 4. Lineweaver-Burk plot of GST-8.3, -/*, and -I. A, the enzyme assay was performed with various concentrations of glutathi- one (GSH) at 1 mM nitroglycerin in 50 mM potassium phosphate, pH 7.0. B, GSH was 1 mM and nitroglycerin concentration was changed. Other assay conditions were the same as those in A. 0, GST-P.3; 0, GST-p; and A, GST-I.

BSP. flP”l

FIG. 5. Inhibition of GST activity toward nitroglycerin by bromosulfophthalein (BSP). The activity was assayed at pH 7.0, 25 “C, with 1 mM nitroglycerin and 1 mM GSH as substrates. 0, GST- 8.3; 0, GST-F; and A, GST-I.

Forms-To characterize subunit structures of p1 values 8.3, 6.6, 6.0, 5.6, and 5.3 forms from human aorta and heart, these were subjected to two-dimensional gel electrophoresis (Fig. 6). As the first dimension, gel isoelectric focusing in the presence of 9 M urea was used for the dissociation of their composing subunits. The p1 8.3 form, which showed a single band on SDS-polyacrylamide gel (Fig. 2, lane 2), gave two major spots with a minor spot on two-dimensional gel, which had the same M, (27,000) but differed in p1 (Fig. 6A). The two major spots were named M, (subunit ~17.0) and M.’ (6.6). The minor spot seemed to be a charge isomer of M,, as reported on the other GST forms (40). Another aorta form (~16.6) was also separated into two spots (panel B); one was identical to Me of the p1 8.3 form in M, and p1, and another spot (M, 26,5OO/pI 6.5) was named N,. The p1 6.0 form from the heart gave a spot named M1 (panel D) which had a similar M, to that of Mi or MS (27,000) (see also Fig. 3, lane 1) but was more acidic (Ml, p1 6.0) than these subunits. This spot, Ml, was identical to the subunit of GST-c( purified from the

by guest on July 1, 2018http://w

ww

.jbc.org/D

ownloaded from

7154

A

Human Aorta and Heart GSTs with Activity to Nitroglycerin

II

Ml 4 M3 _e

B E

4 . - Nt

4 -. N2

C

M3 N2 e

8 76 5 8 76 5 PH

FIG. 6. Subunit composition of human class Mu GST forms revealed by two-dimensional gel electrophoresis. As the first dimension, isoelectric focusing in polyacrylamide gel using 2% Am- pholine, pH 3.5-10, was performed under denaturing conditions for the dissociation of their composing subunits as reported previously (7). SDS-polyacrylamide gel electrophoresis as the second dimension was performed in 11% acrylamide gel. A, aorta pI 8.3 form (8 pg); B, aorta pI 6.6 form (4 pg); C, liver GST-p (15 pg); D, heart p1 6.0 form (4 c(g); E, heart pI 5.6 form (8 c(g); and F, heart p1 5.3 form (5 rg).

liver (panel C) in M, and p1. The p1 5.3 form gave a spot named NP (panel F) which had the similar M, as N, (26,500) but separable from it in p1 (N2, 5.9). The p1 5.6 form was separated into two spots; one was identical to Ms and another to NP (panel E), as is also evident in Fig. 3, lane 2. These results suggested that the five forms purified from the aorta and heart were homodimers (p1 values 6.0 and 5.3 forms) or heterodimers (8.3,6.6, and 5.6 forms) of five different subunits which were immunologically related to GST-p (Figs. 2B and 3B).

N-terminal Amino Acid Sequence Analysis of the Class Mu Forms-In order to clarify the difference in primary structure between the forms, the N-terminal amino acid sequences were examined. It was found that p1 values 8.3, 6.6, and 5.3 forms had a high degree of similarity in the first 20 residues as GST- p (Table IV). The sequence of GST-p determined in this study was consistent with that reported by Mannervik et al. (6) or Board et al. (14), and also consistent with the deduced se- quence from GST-p cDNA reported by Seidegard et al. (41). The ~18.3 form contained equivalent amounts of two different amino acids at three positions: isoleucine and threonine at the third position, glutamine and glutamic acid at the eighth, and valine and isoleucine at the ninth. This result supports that it is a heterodimer. Comparison of the sequences between this form and GST-p showed at least one difference at the eighth position; GST-11 had aspartic acid. The p1 6.6 form which shares the less basic subunit of the ~18.3 form (M2 in Fig. 6, A and B) contained threonine at the third position, glutamic acid at the eighth, and isoleucine at the ninth. These results suggest that another subunit of the ~18.3 form (Mi in Fig. 6A) contains isoleucine at the third, glutamine at the eighth, and valine at the ninth. Although the ~16.6 form gave

TABLE IV

Comparison of N-terminal amino acid sequences of human aorta GST-pI 8.3 and -6.6 forms, heart GST-pI 5.3 form, and GST-p

N-terminal amino acid sequence analyses of p1 8.3 form (10 pg), 6.6 (15 Fg), 5.3 (8 pg), and GST-p from the liver (p1 6.0, 50 rg) were performed twice with a gas-phase protein sequencer as described under “Experimental Procedures.” The residues in boxes indicate differences from those of GST-p.

Position pI 8.3 (M,M,)

GST form

6.6 (MzNd 5.3 (NSNJ 6.0 (P)

1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Pro Met Ile +JKJ Leu GUY Tyr Tw

w a e

-0 GUY Leu Ala

pm01 70 88

46+ 36 88 58

150 99

53 +60 82 +67

- 131 121

89 -

Ala 55 Ile 41

- Leu Leu Leu Glu Sr Thr ASP

49 64 52 42 40 20 25

Pro Met

m Leu Gb Tyr Trp

FF

GUY Leu Ala

- Ala Ile

- Leu Leu Leu Glu Trr Thr

pm01 199 128 61

179 180 214 130 207 122 -

107 193 153

153 105 -

126 159 142

74 52 29

Pro Met

pEJ Leu QY Tyr

Gl; Leu Ala

F

41 34 46

-

38 73

Leu 43 Leu 52 Leu 30

- -

pm01 140

72 43

110 51 56 61 27 44

Pro Met Ile Leu GUY Tyr Trp Asp Ile

pm01 514 842 436 648 517 540 652 239 309

GUY Leu Ala

- 366 494 508

- -

Ala 469 Ile 218

- - Leu 389 Leu 751 Leu 678 Glu 188 Tyr 300 Thr 475 Asu 143

’ -, no assignment.

by guest on July 1, 2018http://w

ww

.jbc.org/D

ownloaded from

Human Aorta and Heart GSTs with Activity to Nitroglycersn 7155

two spots on two-dimensional gel electrophoresis (Fig. 6B and see also Fig. 2, lane 4) indicating that it is a heterodimer, it gave a single amino acid at each position of the first 20 residues. N-terminal amino acid sequence of the p1 5.3 form in the first 20 residues was not different from that of GST-4 reported by Board et al. (14). Although the subunit of this form (NJ had a very similar M, as a subunit of the p1 6.6 form (N1 in Fig. 6B), comparison of the sequences between the two forms revealed that two subunits, N1 and NP, differed in two positions; N, had glutamic acid at the eighth and alanine at the 15th, and N, did asparagine and serine, respec- tively. These results indicate that both subunits are different in the primary structure as well as in p1 (Fig. 6, B and F).

Double Immunodiffusion-The p1 8.3 form (well 1 in Fig. 7) and GST-11 (well 2) formed precipitin lines with anti-GST- p antibody (well A), but the line between GST-p and the antibody formed a spur with the line between the ~18.3 form and the antibody, indicating that the subunits of these forms share a common antigenicity but are not identical. However, GST-c( did not form a clear precipitin line with anti-p1 8.3 form antibody (well B). Since it has been reported that rat

t

I I FIG. 7. Immunological relationship of the aorta GST (p1

8.3) and GST-p. Double immunodiffusion was performed according to Ouchterlony (39). Wells 1, 2, A, and B contain the aorta GST (p1 8.3,4 pg), GST-k (4 pg), anti-GST-F antibody (90 fig), and anti-aorta GST (PI 8.3) antibody (90 gg), respectively.

GST 2-2 sometimes does not form a clear precipitin line with an antibody against GST l-l which belongs to the same Alpha class as GST 2-2 (42), immunological relationship of GST-p and the p1 8.3 form was also examined by immuno- blotting. The subunit of GST-p was stained by immunoblot- ting with anti-p1 8.3 form antibody (data not shown), sup- porting that the subunits of two forms are immunologically related but are not identical.

Substrate Specificity of the Class Mu Forms-Specific ac- tivities of these forms for a number of substrates are shown in Table V. The p1 8.3 form and GST-p did not show a conjugation activity toward BSP, which, however, inhibited the activity of these forms toward nitroglycerin (Fig. 5). The ~18.3 form showed a lower activity toward CDNB and higher activities toward trans-4-phenyl-3-buten-2-one or ethacrynic acid than the other four forms. GST-p purified from the liver had the highest activity toward CDNB among them, and its activities with several substrates were similar to those re- ported by Warholm et al. (13). The ~16.0 form from the heart showed the similar activities as GST-11. The p1 5.3 form had the highest activity with 1,2-dichloro-4-nitrobenzene, and its activities to other substrates were similar to those of GST-4 reported as GST-5.1 by Singh et al. (43). The p1 5.6 form showed the intermediate activities toward CDNB, 1,2-di- chloro-4-nitrobenzene, p-nitrophenyl acetate, and trans-4- phenyl-3-buten-2-one between those of GST-p and the ~15.3 form. Since the activity of GST heterodimers has been de- scribed to be the sum of the activities of the component monomers (l-5), this result is consistent with the finding that the ~15.6 form is a heterodimer of the subunits of GST-p and the p1 5.3 form (Fig. 6, D-F).

DISCUSSION

So far two forms, GST-p (12,13) and GST-4 (14,43), have been well characterized as the human class Mu forms. Al-

TABLE V

Specific activities of human class Mu forms toward various substrates The assay was performed according to Habig et al. (31) or Keen and Jakoby (32), as described under

“Experimental Procedures.” GST-p (~16.0) was purified from the liver (27). GST form

UI 8.3 6.6 6.0 b.4 5.6 5.3

1-Chloro-2,4-dinitrobenzene 1,2-Dichloro-4-nitrobenzene p-Nitrophenyl acetate trans-4-Phenyl-3-buten-2-one Bromosulfophthalein Cumene hydroperoxide 1,2-Epoxy-3-@-nitrophenoxy)propane Ethacrynic acid

32.6 46.5 0.94 0.92

co.02 <0.02 0.64 0.53

co.02 <0.02 0.88 <0.02

co.02 co.02 0.70 0.40

unitslmg protein 105

0.08 0.34 0.42

co.01 0.60

co.01 <O.Ol

83.5 1.05 0.21 0.17

<0.02 co.02 <0.02 <0.02

72.9 7.38 0.04

<0.02 co.02 <0.02 co.02 <0.02

TABLE VI

Properties of human class Mu forms

GST form UI 8.3 6.6 6.0 5.6 5.3

Nomenclature by others Subunit structure

M, (x lo-“) PI

Distribution

Affinity column a Warholm et al. (12, 13). ’ Board et al. (14).

? ? M,M, MJ’JI 27127 27126.5

7.016.6 6.6/6.5 Aorta Aorta

Glutathione-Sepharose

27127 6.0/6.0 Liver Heart Brain

? 4* MINP NJ% 27126.5 26.5 126.5

6.0/5.9 5.9/5.9 Heart Heart Brain Brain

S-hexyl-GSH-Sepharose -

by guest on July 1, 2018http://w

ww

.jbc.org/D

ownloaded from

7156 Human Aorta and Heart GSTs with Activity to Nitroglycerin

though three additional forms, acidic p (44), GST-$ (45, 46), and GST-5 (47) have been suggested to be members of the class, their properties and relationships to GST-p and -4 remain to be established. DeJong et al. (48) reported cDNA sequence of the class Mu subunit designated as subunit 4. However, it differs from the subunit of GST-4 by Board et al. (14) in N-terminal amino acid sequence. The deduced amino acid sequence (217 amino acids) of the subunit (48) was identical with that of GST-p reported by Seidegard et al. (41) except only in a residue, the 172nd amino acid, lysine (48) or asparagine (41). Both of them also reported that there are at least three, probably more, class Mu genes in the human genome from Southern blot analysis utilizing GST+ cDNA (41, 48). However, RNA blot hybridization revealed that only a GST-p mRNA is expressed in human liver (41, 48). In this study five forms belonging to the class Mu have been isolated from human aorta (Table I) and heart (Table II). Two- dimensional gel electrophoresis (Fig. 6) and N-terminal amino acid sequence analysis (Table IV) revealed the presence of five different subunits in the class. We propose to name them M1, M2, MS, N,, and NZ, as tentatively used in Fig. 6, for convenience of expression of subunit structure, and to name the multiple subunits systematically. M and N are subunits with M, of 27,000 and 26,500, respectively, and subscripts are numbered in decreasing order of their p1 values (Table VI). Since several properties of the p1 value 6.0 and 5.3 forms including subunit M, (Figs. 3 and 6) and p1, substrate speci- ficity (Table V), immunological properties (Figs. 3 and 7), and N-terminal amino acid sequence (Table IV) were consistent with those of GST-p (6, 12, 13, 41, 48) and GST-4 (14, 43), their subunits, MB and N?, seem to correspond to the subunits of GST-p and -4 (14), respectively. GST-M,N, (p1 6.6) gave a single amino acid at each position of the first 20 N-terminal residues (Table IV). Some subunits of the class Alpha have been reported to be N terminally blocked (4, 5). A possibility that a subunit of GST-M2N1, especially the Ni subunit, is blocked is unlikely, since average yields of phenylthiohydan- toin-amino acids in sequence analysis were not different be- tween GST-M,Ni and -M1M2 or -M3M3. Subunit composition and properties of the five forms are summarized in Table VI. Hirrell et al. (49) also reported the form named GST 4 from the heart and diaphragm. Since the subunit M, of it is larger than that of GST-II, it seems to be different from GST-4 by Board et al. (14). The expression of the class Mu forms is different between the aorta and heart. GST-p is known to be expressed in about 50% of individuals (12, 41). Although the aorta and heart used in this study were obtained from separate individuals, GST-M1M2 and -M2N1 were observed in all aortas from three individuals. So far, GST-M,M, and -M2N1 were detected only in the aorta, whereas GST-M,N, and -NZN2 were also detected in the brain (data not shown), suggesting that the expression of class Mu forms is tissue-specific.

GST-M,M2 (~18.3 form) of human aorta had a high activity toward nitroglycerin (Table III) and the K, values for nitro- glycerin and GSH were 1.1 and 0.12 mM, respectively (Fig. 4, A and B). Although Bennett et al. (50) have reported that formation of glyceryl-1,2-dinitrate from nitroglycerin is partly dependent on deoxyhemoglobin or deoxymyoglobin, Yeates et al. (22) have very recently reported that BSP, an inhibitor of GST, inhibits the relaxation of rabbit aorta induced by nitro- glycerin but S-hexylglutathione does not. The activity of GST-M1M2 was inhibited by BSP at very low concentration (I&,, 2 jtM, Fig. 5). GST-M1M2 as well as -M2N1 did not bind to S-hexylglutathione-Sepharose but bound to glutathione- Sepharose (Table IB). These properties of GST-M1M2 are consistent with those of GST form(s) in rabbit aorta involved

in pharmacologic activation of nitroglycerin (22). In addition to GST-MIM2, GST-p (M3M3) and GST-I had

a high activity toward nitroglycerin (Table III). Among the rat liver GST forms, GST 3-4 shows the higher activity than GST 3-3 (17), suggesting that the rat subunit 4 is the most active to nitroglycerin (5). Although DeJong et al. (48) sug- gested that human GST-p is more closely related to rat Ybs (9), which may be identical with GST-YniYn, (newly named GST 6-6) (7, lo), than to GST 4-4 from homology in nucleo- tide and amino acid sequences, double immunodiffusion re- vealed that GST-p was more closely related to rat 4-4 than to GST-Yn,Yn, and 3-3.’ Seidegard et al. (41) also described that GST-p and rat GST 4-4 share a high activity toward trans-stilbene oxide. High activity of GST-p for nitroglycerin seems to be consistent with the findings obtained in the rat class Mu forms. On the other hand, human GST-N2N2, which seems to be identical with GST-4 (14), was closely related to rat GST-Yn,Yn, rather than to 4-4 (YbzYbs) or 3-3 in double immunodiffusion.’ A high activity toward 1,2-dichloro-4-ni- trobenzene is a common property of human GST-N2N2 (Table IV) and rat GST-YniYni (7). Thus, GST-NZN2 seems to be an equivalent form to the rat GST-YniYn,, and GST-M~Nz in human heart may correspond to the GST-Yb2Ynl detected in rat heart and testis (8). It seems that human GST-MiM, or -M,N, do not correspond to any of the rat class Mu forms so far reported (l-5).

Previous studies (15-17, 21) revealed that metabolic con- version from nitroglycerin to glyceryl dinitrate and nitrous acid is catalyzed by GST. However, GSTs from the liver were utilized in these studies. GSTs occur in multiple forms, and the expression of these forms is tissue-specific (l-5). In fact, GST forms in the aorta are quite different from those in the liver (27). Although pharmacological significance of GST- M1M2 and -M2N1 in vasodilation by nitroglycerin remains to be clarified, isolation of these forms will provide an enzymatic basis for activation and metabolism of nitroglycerin in blood vessels.

1.

2. 3. 4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

REFERENCES

Mannervik, B. (1985) Ada Enzymol. Relat. Areas Mol. Biol. 57, 357-417

Pickett, C. B. (1987) Essays Biochem. 23, 116-143 Sato. K. (1989) Adv. Cancer Res. 52, 205-25s Mannervik, B:, and Danielson, LJ.. H. (1988) CRC Crit. Rev.

Biochem. 23.283-337 Ketterer, B., Meyer, D. J., and Clark, A. G. (1988) in Glutathione

Conjugation (Sies, H., and Ketterer, B., eds) DD. 73-135. Aca- demic Press, London

Mannervik, B., Alin, P., Guthenberg, C., Jensson, H., Tahir, M. K., Warholm, M., and Jornvall, H. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 7202-7206

Tsuchida, S., Izumi, T., Shim& T., Ishikawa, T., Hatayama, I., Satoh, K., and Sato, K. (1987) Eur. J. Biochem. 170,159-164

Ishikawa, T., Tsuchida, S., Satoh, K., and Sato, K. (1988) Eur. J. Biochem. 176,551-557

Abramovitz, M., and Listowsky, I. (1987) J. Biol. Chem. 262, 7770-7773

Kispert, A., Meyer, D. J., Lalor, E., Coles, B., and Ketterer, B. (1989) Biochem. J. 260, 789-793

Lai, H.-C. J., Qian, B., Grove, G., and Tu, C-P. D. (1988) J. Biol. Chem. 263,11389-11395

Warholm, M., Guthenberg, C., Mannervik, B., and von Bahr, C. (1981) Biochem. Biophys. Res. Comma. 98, 512-519

Warholm, M., Guthenberg, C., and Mannervik, B. (1983) Bio- chemistry 22, 3610-3617

Board, P. G., Suzuki, T., and Shaw, D. C. (1988) Biochim. Biophys. Acta 953,214-217

Habig, W. H., Keen, J. H., and Jakoby, W. B. (1975) Biochem. Biophys. Res. Commun. 64, 501-506

’ S. Tsuchida, and K. Sato, unpublished results.

by guest on July 1, 2018http://w

ww

.jbc.org/D

ownloaded from

Human Aorta and Heart GSTs with Activity to Nitroglycerin 7157

16.

17.

18.

19.

20.

Jakoby, W. B., and Habig, W. H. (1980) in Enzymatic Basis of Detonication Vol. II (Jakoby, W. B., ed) pp. 63-94, Academic Press, New York

Keen, J. H., Habig, W. H., and Jakoby, W. B. (1976) J. Biol. Chem. 251,6183-6188

Jakoby, W. B., Ketterer, B., and Mannervik, B. (1984) Biochem. PharmacoL. 33,2539-2540

Ignarro, L. I., Lipton, H., Edwards, J. C., Baricos, W. H., Hyman, A. L., Kadowitz, P. J., and Gruetter, C. A. (1981) J. Pharmacol. Exp. Ther. 218, 739-749

Kukovetz, W. R., and Holzmann, S. (1985) 2. Kordiol. 74, Suppl. 1,39-44

21. Lau, D. T-W., and Benet, L. Z. (1989) Biochem. Pharmacol. 38, 543-546

515 “’

22. Yeates, R. A., Schmid, M., and Leitold, M. (1989) Biochem. Pharmacol. 38, 1749-1753

40. Satoh, K., Kitahara, A., and Sato, K. (1985) Arch. Biochem. Biophys. 242, 104-111

23. Needleman, P., and Johnson, E. M., Jr. (1973) J. Pharmacol. Exp. Ther. 184, 709-715

24. Brien, J. F., McLaughlin, B. E., Kobus, S. M., Kawamoto, J. H., Nakatsu, K., and Marks, G. S. (1988) J. Pharmacol. Exp. Ther. 244,322-327

41. Seidegard, J., Vorachek, W. R., Pero, R. W., and Pearson, W. R. (1988) Proc. N&l. Acad. Sci. U. S. A. 85, 7293-7297

42. Kitahara, A., and Sato, K. (1981) Biochem. Biophys. Res. Com- nun. 103.943-950

43.

44.

Singh, S. V.,.Ahmad, H., Kurosky, A., and Awasthi, Y. C. (1988) Arch. Biochem. Biophys. 264, 13-22

Hussev. A. J.. Stockman. P. K.. Beckett. G. J.. and Haves. J. D. 25. Hayes, J. D. (1986) Biochem. J. 233, 789-798 26. Guthenberg, C., and Mannervik, B. (1979) Biochem. Biophys.

Res. Commun. 86, 1304-1310 27. Soma, Y., Satoh, K., and Sato, K. (1986) Biochim. Biophys. Acta

869,247-258

(1986) Biochim. Biophis. AC& 874, l-12 ’ ” ’ 45. Awasthi. Y. C.. Dao. D. D.. and Saneto. R. P. (1980) B&hem. J.

191, i-10 46. Singh, S. V., Kurosky, A., and Awasthi, Y. C. (1987) Biochem. J.

218,893-897 28. Simons, P. C., and Vander Jagt, D. L. (1977) Anal. Biochem. 82,

334-341

47. Suzuki, T., Coggan, M., Shaw, D. C., and Board, P. G. (1987) Ann. Hum. Genet. 51.95-106

29. Kamisaka, K., Habig, W. H., Ketley, J. N., Arias, I. M., and Jakoby, W. B. (1975) Eur. J. B&hem. 60,153-161

30. Stockman, P. K., Beckett, G. J., and Hayes, J. D. (1985) Biochem. J. 227, 457-465

48. DeJong, J. L., Chang, C-M., Whang-Peng, J., Knutsen, T., and Tu, C-P. D. (1988) Nucleic Acids Res. 16,8541-8554

49. Hirrell, P. A., Hume, R., Fryer, A. A., Collins, M. F., Drew, R., Bradwell, A. R., and Strange, R. C. (1987) Biochim. Biophys. Acta 915,371-377

31. Habig, W. H., Pabst, M. J., and Jakoby, W. B. (1974) J. Biol. 50. Bennett, B. M., Kobus, S. M., Brien, J. F., Nakatsu, K., and Chem. 249, 7130-7139 Marks, G. S. (1986) J. Pharmacol. Exp. Ther. 237,629-635

32.

33.

34. Lawrence, R. A., and Burk, R. F. (1976) Biochem. Biophys. Res. Commun. 71,952-958

35.

36. 37. 38.

Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275

Bradford, M. M. (1976) Anal. B&hem. 72, 248-254 Laemmli, U. K. (1970) Nature 227, 680-685 Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad.

Sci. U. S. A. 76, 4350-4354 39. Ouchterlonv. 0. (1949) Acta Pathol. Microbial. &and. 26. 507-

Keen, J. H., and Jakoby, W. B. (1978) J. Biol. Chem. 253,5654- 5657

Paglia, D. E., and Valentine, W. N. (1967) J. Lab. Clin. Med. 70, 158-169

by guest on July 1, 2018http://w

ww

.jbc.org/D

ownloaded from

S Tsuchida, T Maki and K SatoMu forms.

classtoward nitroglycerin from human aorta and heart. Multiplicity of the human Purification and characterization of glutathione transferases with an activity

1990, 265:7150-7157.J. Biol. Chem. 

  http://www.jbc.org/content/265/13/7150Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/265/13/7150.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on July 1, 2018http://w

ww

.jbc.org/D

ownloaded from