Exp. Eye Res. (1990) 50. 513-520

Non-selenium Glutathione S-Transferase Activity HITOSHI

Peroxidase without Glutathione from Bovine Ciliary Body

SHICHI”

AND

JAMES

C. DEMAR

Kresge Eye Institute, Department of Ophthalmology, Wayne State University School of Medicine, Detroit, Ml 48201, and Eye Research Institute of Oakland University, Rochester, Ml 48063, U.S.A. (Received

24 July

1989 and accepted

in revised form 13 October

1989)

A glutathione peroxidase was purified from bovine ciliary body by ammonium sulfate fractionation. Sephacryl S-300 gel filtration, diethylaminoethyl (DEAE)-cellulose chromatography and hydroxyapatite chromatography. The purified enzyme has an apparent mw of 112 lcDa by gel filtration and 29 kDa by SDS-polyacrylamide gel electrophoresis. The enzyme therefore is composed of four identical subunits. The ciliary enzyme is active with H,O, (25), cumene hydroperoxide (170), t-butyl hydroperoxide (22). triphenylcarbinyl hydroperoxide (12). linoleic hydroproxide (34) and 5-phenylpentenyl hydroperoxide (22): the numbers after substrates are K’, in ,UM. Glutathione is essential for the reaction: L-cysteine. dithiothreitol and 2-mercaptoethanol are inactive. Mercaptosuccinate (10 ,UM) inhibits the enzyme competitively (K, = 7 ,uM) when cumene hydroperoxide is substrate, and uncompetitively (K, = 10 /(MI when H,O, is substrate. No selenium was found in the enzyme by the fluorometric assay with 2.3diaminonaphthalene. The enzyme demonstrates no glutathione S-transferase activity when tested with l-chloro-2,4dinitrobenzene, and several other compounds. A partial sequence of the enzyme shows some similarities both to Se-glutathione peroxidases and a glutathione S-transferase isozyme. Key words: bovine: ciliary body : glutathione peroxidase : selenium-independent : purification : properties. 1. Introduction

2. Materials

In the eye the ciliary body possesses high drug metabolizing and detoxifying activities (Das and Shichi, 1981). The tissue also contains a significant level of palmityl CoA oxidase activity, a peroxisomal enzyme (Ng and Shichi, 1989). These enzymatic activities of the ciliary body generate hydrogen peroxide which is implicated in oxidative damage to neighboring tissues such as the lens (Bhuyan and Bhuyan, 1984: Varma et al., 1984). We have previously shown that glutathione peroxidase (GSH Px) in the ciliary body plays an important role in the detoxification of peroxides (Kishida et al., 1985 : Ng, Susan and Shichi, 1988). In this study, we purified GSH Px from bovine ciliary bodies, and investigated its properties. Two types of GSH Px are generally recognized ; selenium-dependent enzymes and selenium-independent enzymes (Tappel, 1978 : Wendel, 1981) which are distinguished on the basis of their substrate specificity. The seleniumindependent activity is attributed to an isozyme of GSH S-transferase. Bovine retina contains no seleniumdependent GSH Px activity (Saneto, Awasthi and Srivastave, 1982) : the GSH Px activity of this tissue is accounted for by an anionic isozyme of glutathione Stransferase. Bovine ciliary body GSH Px described in this work shows no GSH S-transferase activity and does not contain selenium. Therefore, the enzyme is distinct from previously described GSH peroxidases. *

For

correspondence.

0014-4835/90/050513+08

SO3.00/0

and Methods

Materials Fresh bovine eyes were obtained from a local slaughterhouse and ciliary bodies dissected and stored at - 70°C until use. Leupeptin, pepstatin, aprotinin, soy trypsin inhibitor, phenylmethylsulfonyl fluoride, DEAE cellulose, NADPH, polyethylene glycol pellets (15-20 KDa), polyethylene glycol-400, glutathione peroxidase (bovine erythrocyte), glutathione reductase (bakers yeast Type II), GSH, cytochrome c (bovine heart) and t-butyl hydroperoxide (HP) were purchased from Sigma Chemicals (St Louis, MO). Hydroxyapatite and reagents for SDS-PAGE were from Bio-Rad Labs. (Richmond, CA). Sephacryl S-300 was obtained from Pharmacia Fine Chemicals (Uppsala, Sweden). lChloro-2,4-dinitrobenzene, p-nitrobenzyl chloride, 1,2-epoxy-3-(p-nitrophenyl) propane, mercaptosuccinic acid, cumene HP and hydrogen peroxide were from Eastman Kodak (Rochester, NY). 1,2-Dinitrobenzene, 3.4~dlnitrobenzoic acid, 1,2-dichloro-4-n&obenzene, 4-nitropyridine-N-oxide, and triphenylcarbinyl HP were purchased from Aldrich Chemicals (Milwaukee, WI). Linoleic HP and 5-phenyl-3-pentenyl HP were gifts from Dr Lawrence J. Marnett. Spherogel TSK 3000 SW columns (7.5 x 300 mm) and a Spherogel TSK SW precolumn (7.5 x 100 mm) for HPLC were purchased from Beckman Instruments (Berkeley, CA). Selenious acid and 2,3-diaminonaphthalene were obtained from Aldrich Chemicals (Milwaukee, WI). 0 1990 Academic Press Limited

514

H SHICHI

--

(A) ‘\. Cl \

670 kDa

320 kDa

v

V

Fraction

I I2 kDa 158 kDo 80 kDo vu

number

J

C

DEMAR

40 kDa 28 kDa

u

v

v

number

IO Fmction

AND

12

14

I6

Fraction

18

20

22

24

26

number

FIG. 1. Purification of bovine ciliary body glutathione peroxidase. (A) Sephacryl S-300, (B) DBAEcellulose. and (C) hydroxyapatite. (0) absorbance at 280 nm; (0) enzyme activity. The numbers and arrows in the upper part of chromatogram (A) indicate molecular weights and positions of elution of polymers used for calibration. Lines in chromatograms (B) and (C) indicate linear gradients of NaCl and potassium phosphate, respectively. Bovine erythrocyte glutathione peroxidase activity is shown with stars in chromatograms (A) and (B). Enzyme activity was assayed with H,O,.

Methods Purifhtion of enzyme. Three hundred bovine ciliary bodies were homogenized at 3°C with a Tekmar Tissumizer (Cincinnati, OH) in 700 ml tris buffer (i.e. 50 mM tris buffer, pH 7.5, containing 002% NaN,, 1 mM phenylmethylsulfonyl fluoride, 2 mM ethylenediamine tetraacetic acid (EDTA), 2 mM ethyleneglycol bis(/?-aminoethylether)-N,N,N’, N’-tetraacetic acid (EGTA) and the following proteinase inhibitors at 10 pg ml-’ : leupeptin, pepstatin, aprotinin and soy trypsin inhibitor). The homogenate was filtered through two layers of cheesecloth and centrifuged at 4000 g for 15 min. The supernatant was saturated with (NH,),SO, to 70% at pH 8.0 at 3°C and centrifuged at 15 000 g for 45 min. The pellet was dissolved in tris buffer, dialyzed against tris buffer overnight and centrifuged at 15 000 g for 15 min. The supernatant was concentrated on polyethylene glycol pellets (1 S-20 kDa), loaded on a Sephacryl S-300

column (94 x 2.5 cm) and eluted with tris buffer at 8 ml hr-’ (8 ml per fraction). The fractions containing 112 kDa protein with enzymatic activity [i.e. fractions 54-64 in Fig. l(A)] were pooled, concentrated, dialyzed, filtered through a 0.22~pm filter, and loaded on a DEAE cellulose column (22.5 x 1.5 cm). The enzyme was eluted by linear gradient of 40-190 mM NaCl in tris buffer (500 ml total volume) at a rate of 4.5 ml hr’ (4-5 ml per fraction). The major fractions containing activity [i.e. fractions 8-12 in Fig. l(B)] were dialyzed overnight against phosphate buffer (i.e. 10 mM potassium phosphate buffer, pH 7.5, containing 0.1% polyethylene glycol-400 and 0.02 % NaN,), filtered through a 0.22~,zm filter and placed on a hydroxyapatite column (14 x 1.5 cm). Elution was effected with a linear gradient of 10-l 50 mM phosphate buffer at a rate of 4.5 ml hr-’ (4.5 ml per fraction). Active fractions were pooled, dialyzed and purified again on the same hydroxyapatite column. HPLC of the purified enzyme was performed on a

BOVINE

CILIARY-BODY

GLUTATHIONE

515

PEROXIDASE

Beckman 342 system using two Spherogel TSK 3000 SW columns (7.5 x 300 mm) connected in series with a Spherogel TSK-SW precolumn (7.5 x 100 mm). Elution was effected with 001 M potassium phosphate, pH 7.1. SDS-polyacrylamide gel (7.5 %) electrophoresis was performed as described previously (Shichi and O’Meara, 1986). Protein was determined using bicinchonic acid by the method of Smith et al. (1985). Selenium analysis. Selenium was analyzed with 2,3diaminonaphthalene on duplicate samples, according to the method of Bayfield and Romalis (1985). Briefly, a sample (025 ml) was heated with HNO,-perchloric acid (2 : 1) at 190 “C for 2.5 hr, mixed with 2 ml of the masking agent (EDTA/NH,OH/NH,OH/methyl orange), treated with 1 ml 0.1% 2,3-diaminonaphthalene in the dark, and heated at 50°C for 30 min. The selenium complex formed was extracted with 5 ml cyclohexane and determined fluorometrically (excitation at 364 nm and emitting at 52 5 nm) with a spectrofluorometer (Amer. Instrument Co., Silver Spring, MD). HPLC-grade water was used for washing glassware and for preparation of the reagents used. Enzyme assays. GSH Px activity was assayed by a modification of the method of Stone and Dratz (1982). The reaction mixture (2.45 ml) contained 0.36 pmol GSH, 0.67 units of glutathione reductase, 0.1 prnol EDTA, 2 rmol NADPH. 2.5 /cm01 H,O, (or 3.3 pmol cumene HP or organic HP), 4.5 pmol NaN,, and 50 pmol tris-HCl. pH 8.0. The reaction was started by addition of 0.0 5 ml enzyme and followed by measuring a decrease in the absorbance at 340 nm at 2 5°C using a Cary 17 spectrophotometer (Varian, Monrovia, CA). The absorbance decrease due to enzyme activity was corrected for absorbance changes due to non-enzymatic oxidation of NADPH. Specific activity was expressed as prnol NADPH oxidized per min per mg protein. GSH S-transferase activity was assayed by the method of Habig, Pabst and Jacoby (1974). Enzyme was added to a mixture of 2 mM l-chloro-2,4dinitrobenzene and 2 mM GSH in 1 ml of 0.1 M potassium phosphate, pH 7.0, and the increase of absorbance at 340 nm was followed at 25°C. For

cytochrome c peroxidase assay, 0.05 ml enzyme was added to 50 nmol reduced cytochrome c (prepared by reduction with Na,S,O, and aeration to remove excess Na,S,O,) and 20 nmol H,O, in 1 ml of 0.1 M potassium phosphate buffer, pH 6.7, and the decrease of absorbance at 550 nm was followed at 25°C. Partial Sequencingof Enzyme

Enzyme samples from the second hydroxyapatite chromatography were analyzed by the automated Edman degradation. Analyses were performed by Dr Ronald L. Niece of the University of Wisconsin Biotechnology Center.

3. Results Purification of GSH Px from Bovine Ciliury Body

More than 70% of the total activity (i.e. the activity of homogenate) was collected in the supernatant after centrifugation at 4000 g. Following (NH&SO, fractionation, gel filtration of the extracted enzyme on a precalibrated Sephacryl S300 column produced a major activity peak of apparent mw of 112 kDa, a shoulder of apparent mw of 80 kDa, and a minor peak of apparent mw of 29 kDa Fig. l(A)]. Since erythrocyte GSH Px was collected as a 80-kDa protein from the column in a separate experiment [see Fig. l(A)], the 80 kDa shoulder probably consisted largely of erythrocyte enzyme and a ciliary body Se enzyme of Mr of about 80 kDa ; dissected ciliary bodies contained a substantial amount of blood. This &action is henceforce designated 80 kDa GSH Px. About 40% of the total activity collected from the column was recovered in the 112 kDa peak. When the 112 k.Da peak (fractions 54-64) was pooled and purified on DEAEcellulose, the major activity peak was eluted by NaCl at 45-55 mM and lower levels of activity (broad shoulder) were eluted continuously at higher NaCl concentrations [Fig. l(B)]. Erythrocyte GSH Px was eluted in the shoulder fractions from the DEAEcellulose column in a separate experiment [see Fig.

TABLE I

Purification of glututhione peroxidusefrom bovine ciliury body

Total volume Purification

steps

(ml)

Supernatant from homogenate 70% (NH,),SO, Saturation Sephacryl S- 300 DEAE cellulose 1st Hydroxyapatite 2nd Hydroxyapatite

950

Total protein (mg) --- .-___--~ 4.845

393 657 102 365 182

Enzyme activity was assayed with H,O, as substrate.

1.769 425 52 17.5 5.3

Total activity (mol NADPH min.‘)

Specific activity (mol NADPH min-’ mg--I)

8.19 x 10-J

1.69 x 1 O-;

5.80 x 1OeJ

3.28 x lo-:

2.24 9.41 5.30 2.69

5.27 1.74 3-03 5.07

x x x x

1O-4 10-j 10-j 10-j

x x x x

lo-: 10-G 10-C 10-6

516

H

SHICHI

AND

J

C

DEMAR

A)

29.2

kDa

w

92.5

kDa

W

66.2

kDa

0

45.0

kDa

0

21.5

()

14.4 kDa

II

, ,A

kDa

l

60

80 Time (min)

100

120

FIG. 2. Molecuiar weight determinations of bovine ciiiary glutathione peroxidase. (A) Separation of three molecular species (112, 59 and 29 kDa) by HPLC. (B) SDS-PAGE.

l(B)]. Further purification of the major activity peak (fractions 8-12) on hydroxyapatite twice resulted in a single peak in which enzymatic activity and absorbance at 280 nm paralleled closely [Fig. l(C)]. Erythrocyte GSH Px was eluted from the hydroxyapatite column at SO-60 mM potassium phosphate buffer (data not shown). Typical results of enzyme purification are summarized in Table I. After several steps of purification the specific activity of enzyme increased only 30-fold over that of the supernatant from tissue homogenate. This is partly explained by the difference in specific activity between ciliary body 112 kDa enzyme and contaminating GSH Px activities (mainly erythrocyte enzymej. The specific activity of purified bovine erythrocyte enzyme was reported to be about 100 units per mg protein (Flohe, Eisele and Wendel 19 71), while the specific activity of purified ciliary body enzyme is 5.1 [lrnol NADPH min-’ mg-’ protein or about 13 units mg-’ protein (for definition of a unit of activity. see Wendel, 1980). Enzyme preparations of lesser purity contained significant amounts of erythrocyte GSH Px and also a ciliary body Se enzyme. Therefore, removal of these enzymes had a counter effect on the improvement of specific activity, thus resulting in a moderate increase in the specific activity of purified enzyme. Molecular Weight When the purified enzyme from the hydroxyapatite column was analyzed by HPLC. three peaks were separated which corresponded to 112, 59 and 29 kDa [Fig. 2(A)]. However. the purified enzyme migrated as

a protein of 29 kDa on SDS-PAGE [Fig 2(B)]. As described above, three peaks of 112 kDa, 80 kDa and 29 kDa were detected by gel filtration of crude extracts on Sephacryl S300. A small peak of 59 kDa might have been present in the chromatogram but was overshadowed by the 80 kDa peak. These results indicate that the enzyme has an apparent mw of 112 kDa and is composed of four identical subunits of about 29 kDa. The individual fractions containing the three molecular species are all enzymatically active. However. it is not known at present whether the subunit is truly active, because the subunit may selfassemble, producing the three species in equilibrium in aqueous buffer. The enzyme was stable during storage at - 70°C for several weeks. Repeated freezing and thawing caused aggregation of the enzyme and gradual loss of activity. Polyethylene glycol-400 at 0.1 % stabilized the enzyme at Iow temperatures. Kinetic Parameters The enzyme catalyzed decomposition of hydrogen peroxide as well as organic hydroperoxides by GSH. The apparent K, values (in pM) were 25 for H,O,, and 170, 22, 12, 34 and 22 for cumene HP, t-butyl HP, triphenylcarbinyl HP, linoleic HP. and 5-phenyl- 3pentenyl HP, respectively (Table II). Except cumene HP, the four other organic hydroperoxides served as good substrates ; IS,,, values for these substrates were comparable with that for H,O,. Although different hydroperoxides were reactive with the enzyme, GSH was essential and could not be replaced by dithiothreitol, r,-cysteine and 2-mercaptoethanol. The en-

BOVINE

CILIARY-BODY

GLUTATHIONE

517

PEROXIDASE

TABLE

II

Properties of ciliary body glutathione peroxidase Inhibition by 10 ,uM mercaptosuccinate

Kinetic parameters Molecular weight Hydroperoxides (HP) (kDa) _-~~~. .-__-~___ Gel Filtration 112 H,O, SDS-PAGE HPLC

29

cumene HP

112 t-butyl HP 59 triphenylcarbinyl 29 linoleic HP 5-phenyl- 3pentenyl HP

HP

K’, (l(M)

Vmax (pmol NADPH min-1 mg-I)

25

5.07

1 70

8.56

22 12 34 22

9.18 2.57 9.88 7.34

H,O, 9.6 (uncompetitive) cumene HP 7.0 (competitive)

Vmnr (ymol NADPH min-’ mg-‘) _-. -2,52 8.56

The K’, value for H,O, is several fold smaller than that for cumene hydroperoxide, indicating a higher affinity of the enzyme for H,O,. The difference in the mode of inhibition between the two substrates may be attributed to a greater chance for the inhibitor to bind to free enzyme when cumene HP is the substrate than when H,O, is the substrate. The kinetic parameters of enzyme, together with molecular weight, are summarized in Table II.

zyme did not show saturation with GSH, i.e the V,,, increased steadily with an increase in GSH concentration. Therefore, the apparent K,,, for GSH was not determined. The enzyme was inhibited by mercaptosuccinate, one of the most potent mercaptocarboxytic acid inhibitors of GSH Px (Chaudiere, Wilhelmsen and Tappel. 1984). The mode of inhibition was uncompetitive with H,O, as substrate (K’i = 2.52 ,UM) and competitive with cumene HP as substrate (K’, = 8.56 pM) (Fig. 3). An inhibitor and a substrate compete for free enzyme in competitive inhibition, while in uncompetitive inhibition, an inhibitor binds to the enzyme-substrate complex but not to free enzyme.

Selenium Content

The selenium content of purified ciliary body enzyme was determined fluorometricaily with 2, 3-diamino-

(B)

5 Se’ (mM-9

K’, (fcM)

Cumene

hydroperoxide

5 S-’ (mf4-I)

IO

15

FIG. 3. Inhibition of bovine ciliary body glutathione peroxidase by 10 ,UM mercaptosuccinic acid. (0) with inhibitor: (0) without inhibitor. (A) Assay with H,O,. (B) Assay with cumene hydroperoxide. 3h

BER 50

518

naphthalene on three different preparations of purified enzyme. Fluorescence intensity at 525 nm increased linearly at least up to 200 ng selenium in calibration experiments. In sample 1 (11 jig protein), 0.38 ng selenium was determined (the mean of duplicates). Sample 2 (14 /Lg protein) was found to contain 0.30 ng selenium (the mean of duplicates). Sample 3 (100 jcg protein) contained 0.36 ng selenium (the mean of duplicates). Blanks, i.e a mixture containing all reagents except sample, gave fluorescence readings equivalent to 030-O 3 6 ng selenium. Enzyme samples therefore contained no selenium. Erythrocyte enzyme samples (80 !lg protein) partially purified on the DEAEcellulose column were found to contain about 130 ng selenium or 1.65 mol selenium per mol of enzyme (80 kDa) by the fluorometric method. SDS-PAGE of the samples showed a band of 21 kDa polypeptide and about SOo/, contaminant bands. This indicates that the erythrocyte enzyme, if purified to homogeneity. would contain 3-4 mol selenium per mol. Incubation of purified ciliary body enzyme with OS mM SeO, in the presence of 2 mM reduced glutathione did not increase the activity of enzyme. From these results we concluded that selenium is not an essential element for the enzyme. Other Enzyme Activities The enzyme did not catalyze oxidation of reduced cytochrome c by H,O,. Neither did the enzyme show glutathione S-transferase activity when incubated with GSH and the following compounds: 1-chloro-2, 4-dtnitrobenzene, p-nitrobenzyl chloride, 1, 2-epoxy3-(p-nitro-phenyl) propane, 1, 2-dinitrobenzene, 3, 4dinitrobenzoic acid, 1, 2-dichloro4-nitrobenzene, and 4-nitropyridine-N-oxide. Partial Amino Acid Sequence A partial amino acid sequence of the purified enzyme was determined by the automated Edman degradation. The following sequence was obtained in about 200/O yield : Pro-Gly-Gly-Leu-Leu-Leu-Gly-AspGlu-Ala-Pro- Asn-Phe-Glu- Ala- Asn-Thr -Thr-Ile-Gly-(Gly)-Ile-(Ser)-Phe-I?-UTry-Leu-Gly. The parentheses indicate less confidence, and residues marked U are unidentified. Since the yield was somewhat low we could not conclude with certainty that the sequence represents the amino terminus of the enzyme. It is possible that the enzyme protein has a blocked amino terminus and a short polypeptide fragment was cleaved from the N-terminus by an endogenous proteinase during the purification procedure, even though several proteinase inhibitors were included in purification buffers. If this was the case, the determined sequence represents an internal peptide sequence of the enzyme.

H

SHICHI

AND

.J C

DEMAR

4. Discussion GSH PX activity is immunohistochemically localized in epithelial cells such as the cornea1 epithelium. ciliary epithelium and retinal pigmented epithelium (Atalla, Sevanian and Rao. 1988) and is shown by enzymatic assays to be highest in the ciliary body among different ocular tissues (Armstrong, Santangelo and Connole, 1981). Only the lens GSH Px has been purified extensively and identified as a selenoenzyme (Bergad, Rathburn and Linder, 1982). Selenium GSH Px and non-selenium GSH Px are distinguished on the basis of different substrate specificity : Se-enzyme reacts with both H,O, and organic hydroperoxides such as cumene HP, while non-Se enzyme is active only with organic hydroperoxides (Wendel. 19 8 0 ). The non-Se activity is attributed to a GSH S-transferase isozyme (Wendel, 1981). GSH Px activity in various tissues is often characterized as Se- or non-Se enzyme, depending on its substrate specificity without purification for selenium analysis. The ciliary body enzyme purified in this study does not seem to belong to either type of GSH peroxidases. The enzyme catalyzes reduction of both H,O, and organic hydroperoxides by GSH. Yet, the enzyme does not contain selenium and shows no GSH S-transferase activity. It awaits further studies to determine whether the novel GSH Px is widely distributed in different tissues. A simple interpretation of properties of this enzyme is that the enzyme is an artifact produced by removal of Se from a selenium-containing GSH Px during the extraction and purification procedure. Several lines of evidence argue against this possibility. First, when erythrocyte selenium GSH Px was purified similarly. the enzyme behaved as a protein of mw about 80 kDa on the Sephacryl S3OO column. suggesting that the 80 kDa shoulder we observed in Fig. l(A) probably contained a Se enzyme which is distinct from the 120 kDa enzyme. Chromatographic behavior on the DEAE-cellulose column was also different between the 120 kDa enzyme and the erythrocyte enzyme. This is not to say that GSH Px activity other than the activity of the 120 kDa enzyme is entirely accounted for by erythrocyte enzyme. The ciliary body may contain a Se-dependent GSH Px that is similar to, but distinct from, erythrocyte enzyme. Secondly the erythrocyte enzyme partially purified by DEAE-cellulose chromatography did not seem to lose selenium during purification. Loss of selenium from the enzyme on the hydroxyapatite is also unlikely, because seleniumcontaining erythrocyte enzyme is purified by hydroxyapatide chromatography without loss of enzymatic activity (Wendel. 1981). If ciliary body enzyme was a Se enzyme similar to erythrocyte enzyme, the enzyme would have retained selenium throughout the purification procedure. Thirdly, partial amino acid sequence of ciliary body enzyme we have determined is not found in the sequences of several selenium enzymes including bovine erythrocyte GSH Px (Table

BOVINE

CILIARY-BODY

GLUTATHIONE

519

PEROXIDASE

TABLE III

Comparison of partial sequences of GSH peroxidases and GSH S-transferase Glutathione peroxidase Bovine erythrocyte Rat liver Human liver Bovine ciliary body Glutathione s-transferase Rat liver (Ya) Bovine ciliary body

97 100 P G G - - - G F E - P N F ML 102 99 P G G - - - G F E - P N F T L 102 99 p G G - - G F E - P N F M L PGGLLLGDEAPNF-

F E K C E V N C E K A H P L F Gunzler et al. (1984) F E K C E V N G E K A H P L F Yoshimura et al. (1988) F E K C E V N G A G A H P L F Ishida et al. (198i’) - EANTTI GGI SFThiswork

163 LLLYVEEPDSALLT LLLGDEAPNFEANT

Pickett et al. ( 1984) This work

The numbers above residues shown the location of residues. The minus sign indicates no residue at the location.

III). Certain sequence similarities are noted between ciliary enzyme and enzymes from erythrocytes (Gunzler et al., 1984) and liver (Ishida et al., 1987; Yoshimura et al., 1988). However, in these enzymes, the sequence starting with PGG is located in the middle of the primary structure away from the Sebinding cysteine (residue 47) which is located closer to the N-terminus. If the PGG sequence represents the Nterminus of ciliary body enzyme, a comparable cysteine residue for Se binding is absent in the enzyme. The ciliary enzyme contains an LLLXXEXPXXXXXT sequence which is not present in selenium GSH peroxidases. Interestingly, the sequence is found in rat liver glutathione S-transferase Ya (Pickett et al., 1984) (Table III). X refers to non-identical residues. Although further eludidation of the primary sequence is required for more rigorous comparison, it is likely that the ciliary enzyme has certain structural similarities both to Se-containing GSH peroxidases and GSH S-transferase isozymes. The specific activity of purified ciliary enzyme is somewhat lower than those of many Se-dependent GSH peroxidases previously reported (Wendel, 1980 ; Takahashi et al., 1987) but is comparable with the specific activity of phospholipid HP GSH Px (Urisini, Maiorino and Gregolin, 1985 : Maiorino et al., 1986). In fact, the affinity of the ciliary body enzyme for triphenylcarbinyl HP is greater than for H,O,. The phospholipid hydroperoxide GSH Px was suggested to protect biomembranes by reducing hydroperoxides produced in membrane phospholipids. Phospholipid hydroperoxide GSH Px contains selenium and is different from the ciliary enzyme. The enzyme is also different in its molecular weight. Nevertheless, in view of the high affinity of the ciliary body enzyme for organic hydroperoxides including Iinoleic HP, it is tempting to suggest that a function of this enzyme may also be protection of membrane phospholipids. Acknowledgements This work wassupportedby a researchgrant (EY04694) from the National Eye Institute, N.I.H., U.S. Public Health

Service. We thank Mr Victor R. Leverenz for technical assistance,and MS Debra A. Jones for her help with preparation of the manuscript.

References Armstrong, D., Santangelo,G. and Connole,E. (198 1). The distribution of peroxide regulating enzymes in the canineeye. Curr. Eye Res. 1, 22542. Atalla, L. R., Sevanian.A. and Rao. N. A. (1988). Immunohistochemicallocalization of glutathione peroxidasein occular tissue.Cut-r. Eye Res. 7, 1023-7. Bayfield, R. F. and RomalisL. F. (1985). pH control in the fluorometric assay for selenium with 2,3-diaminonaphthaleneAnal. Biochem. 144, 569-76. Bergad. P. L.. Rathburn, W. B. and Linder, W. (1982). Glutathione peroxidasefrom bovine lens: a selenoenzyme. Exp. Eye Res. 34, 13144. Bhuyan, K. C. and Bhuyan D. K. (1984). Molecular mechanism of cataractogenesis:III. Toxic metaboiitesof oxygen as initiators of lipid peroxidation of cataract. Cur. Eye Res. 3. 67-81. Chaudiere,J., Wilhelmsen,E. C. and Tappel.A. L. (1984). Mechanismof selenium-glutathioneperoxidaseand its inhibition by mercaptocarboxylic acids and other mercaptans.I. Biol. Chem. 259, 1043-50. Das,N. D. and Shichi. H. (1981). Gamma-glutamyltranspeptidaseof bovineciliary body:purification andproperties. Exp. Eye Res. 29, 109-21. Flohe, L.. Eisele,B. and Wendel, A. (1971). Glutathionperoxidase.1. Hoppe-SeyIer’s Z. Physiol. Chem. 352, 151-8. Gunzler,W. A., Steffens.G. J.. Grossmann,A., Kim, S.-M. A., Otting, F.. Wendel,A. and Flohe. L. (1984). The amino acid sequenceof bovine glutathioneperoxidase.Hopprseiler’s Z. Physiol. Chem. 365. 195-212.

Habig, W. H., Pabst, M. J. and Jacoby, W. B. (1974). Glutathione S-transferases. The first enzymatic stepin mercapturicacidformation. 1. Biol Chem. 249. 7130-9. Ishida, K.. Morino, T., Takagi, K. and Sukenaga.Y. (1987). Nucleotidesequenceof a human genefor glutathione peroxidase.Nucl. Acids Res. 15. 10051. Kishida. K.. Kodama. T.. O’Meara, P. D. and Shichi, H. (1985). Glutathione depletion and oxidative stress: study with perfusedbovine eye. 1. Ocular Pharmacol. 1. 85-99. Maiorino. M.. Roveri, A.. Gregolin,C. and Ursini,F. (1986). Different effectsof Triton X-100 deoxycholateand fatty acids on the kinetics of glutathione peroxidaseand phospholipid hydroperoxide glutathione peroxidase. Arch. Biochem. Biophys. 251, 600-S.

520 Ng. M. C. and Shichi. H. (1989). PeroxisomaJ palmityi CoA oxidase activity in ocular tissues and cultured ciliary epithelial cells. I. Ocular Phnrmacol. 5, 65-70. Ng. M. C., Susan, S. K. and Shichi, H. (1988). Bovine nonpigmented and pigmented ciliary epithelial cells in culture : Comparison of catalase. superoxide dismutase and glutathione peroxidase activities. E;rp. Eye Res. 46. 919-28. Pickett, C. B.. Telakowski-Hopkins. C. A., Ding, G. J.-F., Argenbright L. and Lu, A. Y. H. (1984). Rat liver glutathione S-transferases. 1. Biol. Chem. 259, 5 182-8. Saneto. R. P.. Awasthi, Y. C. and Srivastava, S. K. (1982). Glutathione S-transferases of the bovine retina. Biochem. I. 205, 213-17. Shichi, H. and O’Meara P. D. (1986). Purification and properties of anionic glutathione S-transferanse from bovine ciliary body. Biochem. 1. 237, 365-71. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner. F. H., Provenzano. M. D.. Fujimoto. E. K., Goeke, N. M.. Olson, B. J. and Klenk. D. C. (1985). Measurement of protein using bicinchonic acid. Anal. Biochem. 150. 76-8 5. Stone, W. L. and Dratz, E. A. (1982). Selenium and nonselenium glutathione peroxidase activities in selected ocular and non-ocular rat tissues. Exp. Eye Res. 35, 405-12.

H

SHICHI

AND

J

C

DEMAR

Takahashi. K., Avissar. N.. Whitin, J. and Cohen. H. I 19871. Purification and characterization of human plasma glutathione peroxidase ; a selenoglycoprotein distinct from the known cellular enzyme. Arch. Riochem. Biophys. 256: 677-86. Tappel. A. I,. (1978). Glutathione peroxidase and hydroperoxides. Methods Enzymol. 53. 506-l 3. Ursini, F.. Maiorino, M. and Gregolin. C. (1985). The selenoenzyme phospholipid hydroperoxide glutathione proxidase. Biochim. Biophys. Acta 839, 62-70. Wendel, A. (1980). Glutathione peroxidase. In Enzymatic Basis of DetoxiJication (Ed. Jakoby. W. B.). Vol. 1. Pp. 333-53. Academic Press: New York. Wendel. A. (1981). Glutathione peroxidase. Methods Enzymol. 77, 325-33. Varma, S. D.. Chand. D., Sharma, Y. R., Kuck. J. F. and Richards, R. D. (1984). Oxidative stress on lens and cataract formation: role of light and oxygen. Curr. Eye Res. 3. 35-57. Yoshimura, S., Takekoshi, S., Watanabe, K. and FujiiKuriyama, Y. (1988). Determination of nucleotide sequence of cDNA coding rat glutathione peroxidase and diminished expression of the mRNA in selenium deficient rat liver. Biochem. Biophys. Res. Commun. 154, 1024-8.