Content uploaded by Elias S J Arnér
Author content
All content in this area was uploaded by Elias S J Arnér on Nov 28, 2017
Content may be subject to copyright.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
225, 268–274 (1996)
ARTICLE NO
. 1165
Efficient Reduction of Lipoamide and Lipoic Acid
by Mammalian Thioredoxin Reductase
Elias S. J. Arne
´r,
1
Jonas Nordberg, and Arne Holmgren
Medical Nobel Institute for Biochemistry I, Department of Medical Biochemistry and Biophysics,
Karolinska Institutet, S-171 77 Stockholm, Sweden
Received June 27, 1996
Reduction of the antioxidant lipoic acid has been proposed to be catalyzed in vivo by lipoamide dehydroge-
nase (LipDH) or glutathione reductase (GR). We have found that thioredoxin reductase (TR) from calf
thymus, calf liver, human placenta, and rat liver efficiently reduced both lipoic acid and lipoamide with
Michaelis-Menten type kinetics in NADPH-dependent reactions. In contrast to LipDH, lipoic acid was
reduced almost as efficiently as lipoamide. Under equivalent conditions at 20
7
C, pH 8.0, mammalian TR
reduced lipoic acid by NADPH 15 times more efficiently than the corresponding NADH dependent reduction
catalyzed by LipDH (297 min
01
for TR vs. 20.3 min
01
for LipDH). Moreover, TR was 2.5 times faster in
reducing lipoic acid with NADPH than in catalyzing the reverse reaction (oxidation of dihydrolipoic acid
with NADP
/
). In contrast, LipDH was only 0.048 times as efficient in the forward reaction as compared
to the reverse reaction (using NADH and NAD
/
). We conclude that all or part of the previously described
NADPH-dependent lipoamide dehydrogenase (diaphorase) activities in mammalian systems should be attrib-
uted to TR. Our results suggest that in mammalian cells a significant part of the therapeutically important
reduction of lipoic acid is catalyzed by thioredoxin reductase.
q
1996 Academic Press, Inc.
Thioredoxin reductase (TR), lipoamide dehydrogenase (LipDH) and glutathione reductase
(GR) belong to a family of pyridine nucleotide-disulfide oxidoreductases containing FAD and
a redox active disulfide (1). The two latter enzymes are highly similar homo-dimeric enzymes
with 50 kDa subunits conserved between all species. In contrast, the properties of thioredoxin
reductase (EC 1.6.4.5) differs between Escherichia coli and mammalian cells (2). The reaction
catalyzed by TR from all species is:
NADPH
/
H
/
/
Trx-S
2
S
NADP
/
/
Trx-(SH)
2
where Trx-S
2
is oxidized thioredoxin and Trx-(SH)
2
reduced thioredoxin. Thioredoxin (12
kDa) in reduced form is an efficient protein disulfide reductase with numerous functions,
including function as hydrogen donor of ribonucleotide reductase, regulation of chroloplast
enzyme activities, redox control of transcription factors and cytokine effects on normal and
malignant cells. For reviews on thioredoxin, see (3-5).
Thioredoxin reductase from E. coli with a subunit M
r
of 35 kDa has been studied in depth,
with cloning, sequencing and a high resolution X-ray structure of the enzyme (6, 7). The three-
dimensional structure including the amino acid sequence around the redox active disulfide in
1
Corresponding author. Fax: 46-8-31 15 51. E-mail: Elias.Arner@mbb.ki.se.
Abbreviations used are: TR, thioredoxin reductase; LipDH, lipoamide dehydrogenase; GR, glutathione reductase;
Trx, thioredoxin; Trx-S
2
, oxidized Trx; Trx-(SH)
2
, reduced Trx; DTNB, 5,5
*
-dithiobis-(2-nitrobenzoic acid); lipoamide,
DL-6,8-thioctic acid amide; lipoic acid, oxidized DL-6,8-thioctic acid; dihydrolipoic acid, reduced DL-6,8-thioctic
acid; TE buffer, 50 mM Tris-Cl, 2 mM EDTA, pH 8.0.
0006-291X/96 $18.00
Copyright
q
1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.
268
AID BBRC 5144 / 6905$$$861
07-17-96 07:41:18 bbrcas AP: BBRC
Vol. 225, No. 1, 1996 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
E. coli TR is strikingly different from the corresponding consensus structures of GR and
LipDH and it was concluded that the disulfide reductase activities of E coli TR and GR must
have evolved convergently (7). Mammalian TR was originally purified from calf liver and
thymus (8), and to homogeneity from rat liver (9). The rat liver enzyme showed subunits with
M
r
of 58,000 and the human placenta enzyme had similar properties with even higher subunit
M
r
(10). Mammalian TR, in contrast to the E. coli enzyme, shows a broad substrate specificity
and reduces not only thioredoxins from many species (8) but also several low molecular weight
substrates including 5,5
*
-dithiobis(2-nitrobenzoic acid) (DTNB), vitamin K, selenodiglutathi-
one, selenite, lipid hydroperoxides, and alloxan (reviewed in (4). Inhibitors of mammalian TR
include antitumour quinones (11, 12), nitrosoureas (13), 13-cis-retinoic acid (14), and 1-chloro-
2,4-dinitrobenzene (DNCB) (15). DNCB also induces NADPH oxidase activity after alkylation
of mammalian TR (15), which resembles the NADPH oxidase (or diaphorase) activity of both
glutathione reductase (16) and lipoamide dehydrogenase (1).
We have sequenced a large number of internal peptides of bovine TR and cloned and
sequenced the rat enzyme (manuscript submitted) which largely confirms the cDNA sequence
of a putative human TR recently reported by Powis and coworkers (17). The sequences revealed
a close structural relationship between mammalian TR and GR or LipDH. We have therefore
reinvestigated the activity of mammalian TR with the principal substrates for GR and LipDH,
and found that lipoamide and lipoic acid were efficient substrates for the enzyme.
Lipoic acid bound in amide linkage to transacetylase or transsuccinylase is an essential
cofactor in the mitochondrial oxoacid dehydrogenase complexes (1, 18, 19). Recently lipoic
acid has gained interest as an antioxidant, with potential therapeutic applications in a range
of conditions, including ischemia-reperfusion injuries, cataract formation, HIV activation, neu-
rodegeneration and radiation injury (19).
Both lipoic acid and dihydrolipoic acid are present extracellularly and human plasma levels
have been reported to be 1-25 ng/ml and 33-145 ng/ml, respectively (20). Exogenously adminis-
tered lipoic acid is rapidly and to a high extent reduced to dihydrolipoic acid by mammalian
cells (19, 21, 22). LipDH cannot be the sole enzyme catalyzing this reduction, since cells
which lack LipDH, like erythrocytes, still reduce lipoic acid to dihydrolipoic acid by an
NADPH dependent reaction (22). Glutathione reductase is an NADPH dependent lipoic acid
reducing enzyme, but the efficiency in this reaction was very low compared to reduction of
GSSG (22, 23).
MATERIALS AND METHODS
Material. 2
*
,5
*
-ADP-Sepharose and Q Sepharose were from Pharmacia Inc. Recombinant human Trx was prepared
as described (24) while E. coli Trx came from IMCO (Sweden). Porcine LipDH was purchased from Sigma, as was
lipoamide (oxidized DL-6,8-thioctic acid amide), lipoic acid (oxidized DL-6,8-thioctic acid) and dihydrolipoic acid
(reduced DL-6,8-thioctic acid). Dihydrolipoic acid was used within two hours upon opening of the sealed ampule.
All other chemicals used were of analytical grade or better. TR from bovine liver and thymus, rat liver or human
placenta was purified based upon the method described earlier (9). The purifications typically resulted in 2.5 mg of
TR from 1 kg of tissue, homogenous as judged by SDS PAGE and with a specific activity in DTNB reduction (8)
of more than 1000 A
412
/min/mg.
Ion exchange chromatography. Aliquots of TR (6
m
g in 200
m
l TE buffer) were loaded on a Mono Q PC 1.6/5
column (SMART system, Pharmacia Biotech Inc., Sweden) equilibrated with TE buffer. TE buffer was allowed to
run through until UV baselines were steady (approximately 5 min at 100
m
l/min) whereupon protein was eluted with
a linear NaCl gradient 0 -0.3 M in TE buffer in 20 min. Fractions of 100
m
l were continuously collected and stored
at
0
20
7
C until analyzed for enzyme activity. Protein content in each fraction was determined using the SMART
manager software for calculation of the integrated absorbance at 280 nm.
Enzyme activity determination. Enzyme activity was measured spectrophotometrically at 20
7
C in TE buffer as
described earlier (8, 9), with the following modifications. Screening of activity in fractions of the MonoQ chromatogra-
phy (Figure 2) was carried out in 96-well Cel-Cult plates (Sterilin Ltd., England) and consumption of NADPH was
measured through decrease of absorbance at 340 nm determined using the THERMOmax microplate reader (Molecular
269
AID BBRC 5144 / 6905$$$862
07-17-96 07:41:18 bbrcas AP: BBRC
Vol. 225, No. 1, 1996 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Devices Corp., CA) with the kinetic application of the accompanying SOFTmax software. Samples (10
m
l) were
added to wells containing 150
m
l TE buffer with 200
m
M NADPH and substrate, i.e. either SeO
20
3
(200
m
M), lipoamide
(400
m
M) or insulin (100
m
M), the latter supplemented with either human (5
m
M) or E coli (5
m
M) Trx. In each case,
reference wells contained only TE buffer and 200
m
M NADPH. Reduction of DTNB (500
m
M) in the fractions was
performed in the same manner, but in the presence of 300
m
M NADPH and the reduction of DTNB determined by
increase of absorbance at 405 nm (instead of 412 nm). A well containing TE buffer, 300
m
M NADPH and 500
m
M
DTNB but no enzyme was used as reference.
When lipoamide and lipoic acid as substrates were further analyzed (Tables I and II, Figure 2), enzyme activity
was determined at 20
7
C using a Shimadzu double-beam spectrophotometer with semi-micro cuvettes containing 500
m
l TE buffer (pH 8.0) with 200
m
M pyridine nucleotide, enzyme and lipoamide or lipoic acid (0-2500
m
M), with
everything included except lipoamide or lipoic acid in the reference cuvette. In each case, 5
m
l of pure enzyme was
added, which gave final TR concentrations of 1.2 nM (human placenta TR), 6.9 nM (calf thymus TR), 12.4 nM (rat
liver TR) and 13.8 nM (calf liver TR), as determined in separate experiments by DTNB reduction calculating with
an M
r
for the holoenzyme of 116000 and assuming 1200 A
412
units/min/mg for the pure enzyme (4). Porcine LipDH
from Sigma (10.4 mg/ml) was diluted in TE buffer at the time for experiments and an M
r
of 100 000 was used in
calculations of concentrations (3
m
l of 100 times diluted LipDH in 500
m
l was considered to give 6.24 nM LipDH
in final concentration). Conversion of pyridine nucleotide was determined by decrease (oxidation of NAD(P)H) or
increase (reduction of NAD(P)
/
) of absorbance at 340 nm and using a molar extinction coefficient of 6200 M
01
cm
01
.
Lipoamide and lipoic acid were dissolved in EtOH as 50 mM stock solutions and EtOH was therefore added to an
equal final concentration of 4% in all cuvettes. In each case, enzyme and pyridine nucleotide was added first, pre-
incubated 5 min to reduce the enzyme, whereafter substrates were added and activity was measured.
RESULTS
At the onset of this study, we screened several of our preparations of purified mammalian
TR with the primary substrates of GR and LipDH. GSSG was confirmed not to be a substrate,
in agreement with previous results (9). However, we found a significant NADPH dependent
FIG. 1. Ion exchange chromatography of mammalian TR. Purified calf thymus TR was analyzed on a column of
MonoQ anion exchanger with an isocratic NaCl gradient as shown in A (
l
; protein profile, solid line; salt gradient).
B (inset) shows the enzymatic activity in those fractions containing protein, using DTNB (
l
, dashed line), insulin
(coupled to human Trx,
l
,orE coli Trx,
m
), selenite (
j
), or lipoamide (
l
) as substrates.
270
AID BBRC 5144 / 6905$$$862
07-17-96 07:41:18 bbrcas AP: BBRC
Vol. 225, No. 1, 1996 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TABLE I
Kinetic Parameters of Mammalian Thioredoxin Reductase in
Reduction of Lipoamide, Reduction of Lipoic Acid,
and Oxidation of Dihydrolipoic Acid
a
K
m
k
cat
k
cat
/K
m
Source of TR (mM) (min
01
) (mM
01
min
01
)
Calf thymus 0.49 503 1035
Calf liver 1.00 918 913
Human placenta 0.51 1102 2182
Rat liver 0.86 1296 1498
Kinetic parameters of rat liver TR for
Substrate lipoic acid and dihydrolipoic acid
Lipoic acid 0.71 368 518
Dihydrolipoic acid
b
0.88 173 196
a
Consumption of pyridine nucleotide was determined by following
the change of absorbance at 340 nm using a molar extinction coefficient
of 6200 M
01
cm
01
. Assays were performed at 20
7
C,pH 8.0, in 50 mM
Tris, 2 mM EDTA, and 200
m
M NADPH with 3–15 nM TR (depending
on preparation) and 0–2500
m
M lipoamide or lipoic acid, as described
in further detail in Materials and Methods. The kinetic parameters given
in the table are those derived if Michaelis-Menten kinetics are fitted to
the experimental data. The fitted Michaelis-Menten plots in reduction
of lipamide and lipoic acid and oxidation of dihydrolipoic acid by rat
liver TR are shown in Fig. 2.
b
NADP
/
used instead of NADPH.
lipoamide reduction in all TR preparations that we analyzed. To exclude contamination of
lipoamide dehydrogenase, one of the pure calf thymus TR preparations (showing one single
distinct band on SDS gel) was further analyzed with MonoQ ion exchange chromatography.
It was clear that previously known TR-catalyzed reactions (Trx-dependent insulin reduction
as well as reduction of DTNB or selenite) co-eluted in one single peak together with reduction
of lipoamide (Figure 1). The same result was repeated with two additional TR preparations;
one from calf thymus and one from rat liver (data not shown). We also tested TR from E.
coli for lipoamide reduction, but this enzyme did not reduce lipoamide to any detectable level
(data not shown).
Kinetic parameters in the NADPH dependent lipoamide reduction by human, rat and calf
TR were then determined (Table I). For the rat liver TR preparation, that had the highest
specific activity, reduction of lipoic acid as well as the corresponding reverse reaction (NADP
/
coupled dihydrolipoic acid oxidation) was also analyzed (Table I). Also lipoic acid and dihy-
drolipoic acid were good substrates, with K
m
values in the same range as those for lipoamide
(Table I). In none of the reactions (reduction of lipoamide or lipoic acid with NADPH or
oxidation of dihydrolipoic acid with NADP
/
) could any cooperativity be observed and the
experimental data were easily fitted to classic Michaeli-Menten type kinetics (Figure 2).
To compare the activities of mammalian TR with those of mammalian LipDH, both
NADP(H) and NAD(H) were used as pyridine nucleotides under equivalent conditions (20
7
C,
50 mM Tris, 1 mM EDTA, pH 8.0, 200
m
M pyridine nucleotide and 1 mM lipoic acid
derivative) with both LipDH and TR. The results of these assays are given in Table II. The
reduction of lipoamide was ten times more efficient by LipDH using NADH than by TR using
271
AID BBRC 5144 / 6905$$$862
07-17-96 07:41:18 bbrcas AP: BBRC
Vol. 225, No. 1, 1996 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Reduction of lipoamide and lipoic acid and oxidation of dihydrolipoic acid by rat liver TR. TR from rat
liver (7.4 nM) was analyzed for reduction of lipoamide (
l
) or lipoic acid (
j
)at20
7
C, pH 8.0 in 50 mM Tris-Cl, 2
mM EDTA, and 200
m
M NADPH or oxidation of dihydrolipoic acid (
h
) using 200
m
M NADP
/
instead of NADPH.
The curves fitted to the data are Michaelis-Menten plots and the derived kinetic parameters are given in Table I.
NADPH (9215 min
01
vs. 889 min
01
). TR, on the other hand, was fifteen times more efficient
in reducing lipoic acid than was LipDH (297 min
01
vs. 20.3 min
01
). It was also clear that TR
was more efficient in the forward reaction, i.e. reducing lipoic acid, as compared to the reverse
reaction, i.e. catalyzing oxidation of dihydrolipoic acid (ratio
Å
2.49). The opposite, however,
was true for LipDH, that was 20 times more efficient catalyzing the reverse reaction (ratio
between forward and reverse reactions
Å
0.048).
DISCUSSION
Our results show that mammalian thioredoxin reductase efficiently catalyzes NADPH-depen-
dent lipoamide and lipoic acid reduction. This reduction of lipoamide and lipoic acid illustrates
the close relationship of the mammalian TR with other enzymes of the pyridine nucleotide
disulfide oxidoreductase family, such as LipDH and GR, which has interesting implications
since TR of the E. coli type appears not to have evolved from a common ancestral protein
(7). It should be noted that we did not see evidence of cooperativity in the reduction of
lipoamide or lipoic acid by TR, which is different from the pronounced cooperativity seen in
the case of LipDH (1).
The efficient reduction of lipoic acid by mammalian TR adds to its known wide substrate
specificity (1-4). The strong 20-fold preference of LipDH for dihydrolipoic acid oxidation
compared to lipoic acid reduction under our assay conditions (Table II) can probably in part
be explained by a lack of NAD
/
in the assay, since NAD
/
is a strong positive effector (1).
TR readily reduced lipoic acid under without NADP
/
, demonstrating differences in enzyme
mechanism between these two enzymes. The K
m
-values of mammalian TR for lipoic acid and
its derivatives of 0.49 - 1.0 mM should be compared with K
m
values of 2.5 - 5
m
M for
mammalian Trx (8, 9), illustrating the higher affinity of the enzyme for its natural substrate
thioredoxin. However, the k
cat
values of 503 - 1296 min
01
for lipoamide demonstrate an
272
AID BBRC 5144 / 6905$$$862
07-17-96 07:41:18 bbrcas AP: BBRC
Vol. 225, No. 1, 1996 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TABLE II
Activity of Mammalian TR and LipDH in Reduction of Lipoamide and Lipoic Acid
and in the Reverse Reaction Oxidizing Dihydrolipoic Acid,
Using Either NADP or NAD as Pyridine Nucleotides
Enzyme activity (mol pyridine
nucleotide/mol enzyme/min)
a
Pyridine
Substrate nucleotide TR LipDH
Lipoamide NADPH 889
õ
1
NADH 43.4 9215
Lipoic acid NADPH 297
õ
1
NADH
õ
1 20.3
Dihydrolipoic acid NADP
/
119 12.9
NAD
/
23.8 424
Ratio in activity between
Substrates compared substrates compared
Lipoamide/lipoic acid 2.99
b
454
c
Lipoic acid/dihydrolipoic acid 2.49
d
0.048
e
a
All assays were performed at 20
7
C in 50 mM Tris, pH 8.0, and 2 mM EDTA with 1 mM lipoic
acid derivative and 200
m
M pyridine nucleotide. The concentration used of purified rat liver TR
was 7.4 nM and that of porcine heart LipDH was 6.2 nM. Activity was determined as oxidation or
reduction of the pyridine nucleotides by following the change in absorbance at 340 nm using a
molar extinction coefficient of 6200 M
01
cm
01
.
b
With NADPH.
c
With NADH.
d
Using NADPH with lipoic acid and NADP
/
with dihydrolipoic acid.
e
Using NADH with lipoic acid and NAD
/
with dihydrolipoic acid.
efficient reaction, when compared to the corresponding values of 3000 - 4200 min
01
for
different mammalian thioredoxins (9, 24).
Several NADPH-dependent diaphorases or NADPH oxidases, with varying degree of over-
lapping lipoamide dehydrogenase activity, have been described (25-29), many for which the
true nature has not yet been fully characterized. Also, NADPH-dependent diaphorase activity
has been associated with NO synthase expression in the nervous system, but additional neuronal
NADPH-dependent diaphorase activity which is not dependent on NO synthase has been
clearly demonstrated (30, 31). It will be important to determine how TR contributes to these
less characterized diaphorase activities. It is known that the enzyme is present in the nervous
system (32, 33) and the NADPH oxidase activity of TR (15) also readily reduces nitro blue
tetrazolium (NBT), the classic substrate in determination of neuronal NADPH dependent
diaphorase activity (Ane
´r, Nordberg and Holmgren, unpublished observations).
It should be noted that mammalian lipoamide dehydrogenase is a mitochondrial enzyme as
part of several
a
-keto acid dehydrogenase complexes (1, 18) whereas TR is a cytosolic enzyme
with some association with membrane structures (33, 34). The association of TR with the
plasma membrane is worth emphasizing, since lipoamide dehydrogenase activity has also been
detected in plasma membrane fractions (35).
Dihydrolipoic acid and lipoic acid have different antioxidant properties and reduction of
lipoic acid is a necessary prerequisite for several of its antioxidant effects (19). The lipoic acid
reducing activity of mammalian TR should therefore play a significant role in the metabolism of
this compound and for activation of its many antioxidant and redox regulatory qualities.
273
AID BBRC 5144 / 6905$$$862
07-17-96 07:41:18 bbrcas AP: BBRC
Vol. 225, No. 1, 1996 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
ACKNOWLEDGMENTS
This study was supported by the Swedish Cancer Society (961), the Swedish Medical Research Council (13X-
3529), Inga-Britt and Arne Lundbergs Stiftelse, Knut and Alice Wallenbergs Stiftelse, Makarna Agnes och Gustaf
Backlunds fond fo
¨r cancerforskning, Stiftelsen Sigurd och Elsa Goljes minne, and the Karolinska Institute.
REFERENCES
1. Williams, C. H., Jr. (1992) in Chemistry and Biochemistry of Flavoenzymes. (Mu
¨ller, F., Ed.) pp. 121–211, CRC
Press, Boca Raton, FL.
2. Holmgren, A. (1989) J. Biol. Chem. 264, 13963–13966.
3. Holmgren, A. (1985) Annu. Rev. Biochem. 54, 237–271.
4. Holmgren, A., and Bjo
¨rnstedt, M. (1995) Methods Enzymol. 252, 199–208.
5. Yodoi, J., and Tursz, T. (1991) Adv. Cancer Res. 57, 381–411.
6. Waksman, G., Krishna, T. S., Williams, C. H., Jr., and Kuriyan, J. (1994) J. Mol. Biol. 236, 800–816.
7. Kuriyan, J., Krishna, T. S. R., Wong, L., Guenther, B., Pahler, A., Williams, C. H., Jr., and Model, P. (1991)
Nature 352, 172–174.
8. Holmgren, A. (1977) J. Biol. Chem. 252, 4600–4606.
9. Luthman, M., and Holmgren, A. (1982) Biochemistry 21, 6628–6633.
10. Oblong, J. E., Gasdaska, P. Y., Sherrill, K., and Powis, G. (1993) Biochemistry 32, 7271–7277.
11. Mau, B.-L., and Powis, G. (1992) Biochem. Pharm. 43, 1613–1620.
12. Mau, B.-L., and Powis, G. (1992) Biochem. Pharm. 43, 1621–1627.
13. Schallreuter, K. U., Gleason, F. K., and Wood, J. M. (1990) Biochim. Biophys. Acta 1054, 14–20.
14. Schallreuter, K. U., and Wood, J. M. (1989) Biochem. Biophys. Res. Commun. 160, 573–9.
15. Arne
´r, E. S. J., Bjo
¨rnstedt, M., and Holmgren, A. (1995) J. Biol. Chem. 270, 3479–3482.
16. Carlberg, I., and Mannervik, B. (1980) FEBS Lett. 115, 265–268.
17. Gasdaska, P. Y., Gasdaska, J. R., Cochran, S., and Powis, G. (1995) FEBS Lett. 373, 5–9.
18. Perham, R. N. (1991) Biochemistry 30, 8501–8512.
19. Packer, L., Witt, E. H., and Tritschler, H. J. (1995) Free Rad. Biol. Med. 19, 227–250.
20. Teichert, J., and Preiss, R. (1992) Int. J. Clin. Pharmacol. Ther. Toxicol. 30, 511–512.
21. Handelman, G. J., Han, D., Tritschler, H., and Packer, L. (1994) Biochem. Pharmacol. 47, 1725–1730.
22. Constantinescu, A., Pick, U., Handelman, G. J., Haramaki, N., Han, D., Podda, M., Tritschler, H. J., and Packer,
L. (1995) Biochem. Pharmacol. 50, 253–261.
23. Pick, U., Harmaki, N., Constantinescu, A., Handelman, G. J., Tritschler, H. J., and Packer, L. (1995) Biochem.
Biophys. Res. Commun. 206, 724–730.
24. Ren, X., Bjo
¨rnstedt, M., Shen, B., Ericson, M. L., and Holmgren, A. (1993) Biochemistry 32, 9701–8.
25. Cross, A. R., Yarchover, J. L., and Curnutte, J. T. (1994) J. Biol. Chem. 269, 21448–21454.
26. Gasdaska, P. Y., Fisher, H., and Powis, G. (1995) Cancer Res. 55, 2542–2547.
27. Carlberg, I., Sahlman, L., and Mannervik, B. (1985) FEBS Lett. 180, 102–106.
28. Otsuka-Murakami, H., and Nisimoto, Y. (1995) FEBS Lett. 361, 206–210.
29. Sreider, C. M., Grinblat, L., and Stoppani, A. O. M. (1990) Biochem. Pharmacol. 40, 1849–1857.
30. Afework, M., Tomlinson, A., and Burnstock, G. (1994) Cell Tissue Res. 276, 133–141.
31. Hope, B. T., Michael, G. J., Knigge, K. M., and Vincent, S. R. (1991) Proc. Natl. Acad. Sci. U.S.A. 88, 2811–
2814.
32. Hansson, H.-A., Holmgren, A., Norstedt, G., and Rozell, B. (1989) Exp. Eye Res. 48, 411–420.
33. Rozell, B., Hansson, H. A., Luthman, M., and Holmgren, A. (1985) Eur. J. Cell Biol. 38, 79–86.
34. Schallreuter, K. U., Ja
¨nner, M., Mensing, H., Breitbart, E. W., Berger, J., and Wood, J. M. (1991) Int. J. Cancer
48, 15–9.
35. Danson, M. J. (1988) Biochem. Soc. Trans. 16, 87–89.
274
AID BBRC 5144 / 6905$$$862
07-17-96 07:41:18 bbrcas AP: BBRC