Research in

Res Exp Med (2000) 199:243–251

© Springer-Verlag 2000

O R I G I N A L PA P E R
Effect of α-lipoic acid supplementation
on oxidative protein damage
in the streptozotocin-diabetic rat
U. Çakatay1, A. Telci1,3, R. Kayali2, A. Sivas1, T. Akçay2
.
1 Central Laboratory of Clinical Biochemistry, Istanbul Faculty of Medicine,
. .
University of Istanbul, Çapa 34390, Istanbul, Turkey .
2 Department of Biochemistry, Cerrahpaşa Faculty of Medicine, University of Istanbul,
.
Kocamustafapaşa 34303, Istanbul, Turkey .
3 Ataköy, 9. Kısım, S2 Blok, A Kapisi, Daire:7, Istanbul

Received: 23 April 1999 / Accepted: 5 October 1999

Abstract. An increase in oxidative stress may contribute to the development

of oxidative protein damage in streptozotocin diabetic rats. In the present
study, the influence of α-lipoic acid supplementation on plasma protein car-
bonyl, plasma thiol, and plasma lipid hydroperoxide levels was examined in
order to characterize the relationship between the oxidative stress and the
oxidative protein damage. Rats were randomly divided into three groups of
equal body weight. Chronic hyperglycemia was induced by intravenous
streptozotocin injection in both the group of male Wistar rats to be supple-
mented with α-lipoic acid and the group that was not to receive α-lipoic ac-
id. Nondiabetic rats formed the control group and received a saline injec-
tion. In streptozotocin diabetic rats with and without α-lipoic acid supple-
mentation, plasma carbonyl levels were significantly increased, while plas-
ma thiol levels were significantly decreased compared with those of the
control group. Plasma lipid hydroperoxide levels were significantly in-
creased in diabetic rats without α-lipoic acid supplementation compared
with those of the controls, but the lipid hydroperoxide levels in the α-lipoic
acid supplemented group were no different from those of the controls. In
streptozotocin-diabetic rats, oxidative stress was significantly decreased in
the α-lipoic acid-supplemented group. The results of this study suggest that
α-lipoic acid, by decreasing oxidative stress, may be effective in preventing
oxidative protein damage, which may contribute to the development of dia-
betic complications.

Key words: Diabetes – Protein oxidation – α-Lipoic acid – Protein

Carbonyl – Thiol

Correspondence to: A. Telci; Tel.: +90-212-635-1172, Fax: +90-212-533-2083

244

Introduction

Glycation and free radical reactions may play an important role in diabetic
complications [3 ,11]. Free radicals have previously been shown to be capa-
ble of damaging many cellular components such as DNA [6], proteins [9,
10, 15, 17], and lipids [11, 16]. The pathogenesis of atherosclerosis may in-
volve both free radicals and peroxidation products [16]. Indeed, damage to
DNA and proteins may often be more important than damage to lipids in ox-
idative stress situations in vivo [15]. Proteins can also be damaged by oxy-
gen radicals, leading to a loss of enzymatic activity and the conversion of
amino acids to carbonyl derivatives [17]. This type of alteration was charac-
terized as metal-catalyzed oxidation of proteins. Fe2+ or Cu2+ can bind to
cation binding sites on proteins, and further attack by H2O2 or O–2 can trans-
form side-chain amine groups of many amino acids into carbonyls [15].
The oxidative inactivation of enzymes by free radicals and the accumula-
tion of oxidized proteins may play a critical role in the alteration of cellular
function and cell death [18]. Attack by free radicals upon proteins can dam-
age several amino acid residues. Oxidative damage to amino acid residues
of proteins can generate carbonyl products. Indeed, measurement of protein
carbonyls has been used as a sensitive assay for oxidative protein damage
[15]. Liggins and Furth [9] reported that the carbonyls detected under in vit-
ro conditions were clearly the result of glycoxidation or sugar autoxidation,
while some carbonyls may be of purely amino acid origin. Miyata et al. [10]
hypothesized that both glycoxidation and lipoxidation reactions contribute
to the formation of protein carbonyls. To test this hypothesis, they detected
protein carbonyls in human diabetic tissues. It has already been shown that
free radicals cause oxidation of protein thiol groups in plasma. Essentially
all of the plasma thiol groups are associated with proteins. Albumin is the
most abundant plasma protein and is a powerful extracellular antioxidant
[8]. Protein thiols may preemptively scavenge oxidants that initiate peroxi-
dation, thus sparing vitamin E and/or lipids from attack [19].
α-Lipoic acid (ALA) is readily absorbed from the diet, transported to the
tissues, and taken up by cells, where a large proportion is rapidly converted
to dihydrolipoic acid (DHLA). The effects of both ALA and DHLA must
therefore be considered when proposing a mechanism for therapeutic ef-
fects. DHLA appears to be able to regenerate other antioxidants, such as as-
corbate and vitamin E, from their radical forms. On the other hand, ALA
also causes an increase in intracellular glutathione (GSH) levels [7]. GSH
may be a primary agent involved in redox regulation of protein thiols
[11–13]. The aim of this study is to elucidate oxidative protein damage in
diabetes by determining carbonyl and thiol levels in plasma proteins and the
efficacy of ALA supplementation in preventing this damage.

Materials and methods

Inbred male Wistar rats weighing 210–226 g, 8 weeks old, and supplied by the Center for
Experimental and Applied Medical Research, University of Istanbul, Turkey, were used
in this experiment. Animals were housed in conventional wire-mesh cages, four rats per
cage, in a room with the temparature regulated at 21±1°C, humidity 45%–50%, and
Table 1. General data on Wistar rats

Control group α-Lipoic acid-supplemented Diabetic group without

(n=15) diabetic group (n=12) supplementation (n=12)

Median Range Median Range Median Range

Initial weight (g) 216 210–220 219 215–226 224 220–226***

Final weight (g) 240 236–245 186 l82–190 180 176–184***
Red blood cell (µl) 8.7.106 7.34.106–9.45.106 6.7.106 5.39.106–9.77.106 5.37.106 4.16.106–7.83.106
Hemoglobin (mM) 141 122–165 132 108–166 138 102–145
Hematocrit (%) 48.5 42.9–52.4 38.5 32.6–50.1 29.2 21.8–42.8***

***P<0.001 vs. control

245
246

light-dark cycles (12 h). All animals were given ad libitum access to natural food and
water by drinking bottle throughout the course of the experiment. They were fed a stan-
dard laboratory diet not containing ALA for 5 weeks. Rats were randomly divided into
three groups of equal body weight. Chronic hyperglycemia was induced by intravenous
injection of 60 mg streptozotocin per kg body weight (Sigma, St Louis, MO, USA) in
Wistar rats that were to be supplemented with ALA (n=12) and in those that were not to
receive ALA (n=12). Non-diabetic rats formed the control group and received a saline
injection (n=15). Diabetes was monitored consecutively by means of body weight and
blood glucose determinations. ALA was dissolved in water and administered per gavage
at a dose of 49.5 mg per day throughout the feeding period of 5 weeks to rats that were to
be supplemented with ALA. This dosage was effective to show the potent antioxidant ac-
tivity of ALA in vivo in a study by Podda et al. [14]. After 5 weeks of feeding, rats were
fasted and blood was collected from the heart aorta under deep anesthesia with intraperi-
toneally administered sodium pentothal (Abbott, Campoverde di Aprilia, Italy, 100 mg/
kg). Blood was collected in plain tubes and tubes containing ethylenediaminetetraacetate
(EDTA) or lithium heparin, depending on the analysis. Following coagulation, blood in
plain tubes was centrifuged for 15 min at 4°C. General data on Wistar rats are given in
Table 1. Protein carbonyls were measured spectrophotometrically using the method de-
scribed by Reznick and Packer [15]. Protein contents were determined on the HCl blank
pellets using a BSA (bovine serum albumin) standard curve in guanidine-HCl. Plasma
thiol concentration was determined by using 5,5’-dithio-bis(2-nitrobenzoic acid)
(DTNB) as described by Hu [8]. Absorbances were measured at 412 nm against blank
samples without DTNB. Plasma lipid hydroperoxide levels were determined spectropho-
tometrically according to the method of ferrous oxidation with xylenol orange (FOX 1)
[20]. FOX 1 was suitable for determining water-soluble hydroperoxides such as butyl
and cumyl hydroperoxides. Whole blood was assayed immediately for GSH by the meth-
od described by Beutler et al. [2]. Hemoglobin content of whole blood was measured by
the H2 autoanalyzer (Technicon, Bayer Diagnostics, Dublin, Ireland). Blood samples
collected in tubes with EDTA were analyzed for erythrocyte GSH peroxidase (E.C.
1.11.1.9) and glucose-6-phosphate dehydrogenase (E.C. 1.1.1.49) activities with enzy-
matic colorimetric kits (Randox Lab. Ltd., Antrim, N. Ireland). Serum glucose, total pro-
tein, albumin, uric acid, total cholesterol, and triglyceride levels were determined colori-
metrically on the DAX-72 discrete autoanalyzer (Technicon, Bayer Diagnostics, New
York). Serum direct high-density lipoprotein (HDL) cholesterol (Randox Lab. Ltd.,
Antrim, N. Ireland), iron, unsaturated iron binding capacity (UIBC), and fructosamine
levels were determined colorimetrically on the RA-XT discrete autoanalyzer (Technicon,
Bayer Diagnostics, Dublin). All data were not normally distributed, and the data were
therefore expressed as median and range. The plasma protein carbonyl, plasma thiol,
plasma lipid hydroperoxide, serum lipid levels, erythrocyte GSH, erythrocyte GSH per-
oxidase, and glucose-6-phosphate dehydrogenase activities in the control and diabetic
groups were compared using the Kruskal-Wallis nonparametric ANOVA (analysis of
variance) Test. Unless specified otherwise, P<0.05 was considered significant.

Results

There was a significant weight reduction in Wistar rats following a period of

diabetes of 35 days (Table 1). Diabetic rats were observed to be hyperglyce-
mic, hyperphagic, polydipsic, and polyuric. Although the hematocrit values
of polyuric diabetic rats were expected to be higher because of dehydration,
the lower hematocrit values found in our study might be the result of effec-
tive rehydration. Serum glucose, fructosamine, total protein, albumin, cho-
lesterol, HDL cholesterol, triglyceride, uric acid, iron, and UIBC levels in
diabetic rats with and without ALA supplementation and in the control rats
are given in Table 2. The results of the statistical analyses of these parame-
Table 2. Biochemical parameters of diabetic rats with and without α-lipoic acid supplementation and the control rats

Control group α-Lipoic acid-supplemented Diabetic group without

(n=15) diabetic group (n=12) supplementation (n=12)

Median Range Median Range Median Range

Glucose (mM) 9.5 6.9–10.9 33.8 31.4–37.3*** 26.9 13.0–39.6***

Fructosamine (mM) 1.0 0.9–1.2 2.3 2.0–2.4** 2.3 2.1–2.4***
Total protein (g/l) 70 61–76 64 54–74 67 60–69
Albumin (g/l) 36 32–39 28 20–33*** 34 24–35
Cholesterol (mM) 1.27 0.80–1.63 1.40 0.72–1.58 1.42 1.14–1.63
HDL cholesterol (mM) 1.06 0.75–1.29 1.06 0.75–1.16 1.09 0.75–1.19
Triglyceride (mM) 0.79 0.29–1.37 0.90 0.42–1.35 0.89 0.56–1.29
Uric acid (mM) 0.053 0.024–0.100 0.044 0.029–0.083) 0.071 0.017–0.183
Iron (µM) 27 9.97–44.21 21 11.42–24.10 26 19.93–32.98
UIBC (µM) 74 27–85 41 36–49* 30 24–33**

HDL, high-density lipoprotein; UIBC, unsaturated iron binding capacity; *P<0.05, **P<0.01, ***P<0.001 vs. control
247
 248

Table 3. Parameters used to determine oxidative protein damage in diabetic rats with and without α-lipoic acid supplementation and in controls

Control group α-Lipoic acid-supplemented Diabetic group without

(n=15) diabetic group (n=12) supplementation (n=12)

Median Range Median Range Median Range

Carbonyl (nM/mg protein) 0.68 0.62–0.72 0.83 0.72–0.99** 0.95 0.79–1.21***

Thiol (µM) 317 256–370 222 131–278** 219 190–226**
Lipid hydroperoxides (µM) 1.18 0.70–1.25 1.17 0.94–1.66 1.70 1.42–1.96**
GSH (mg GSH/dl whole blood) 64.20 51.20–79.20 75.3 49.90–81.00** 40.00 20.80–71.70
GPX (U/dl whole blood) 454.7 368.9–554.5 464.1 394.4–534.3 537.2 457.8–596.5
G-6-P-DH (mU/RBC per ml blood) 942 437–1413 774 606–1009 673 471–908

GSH, glutathione; RBC, red blood cell; GPX, glutathione peroxidase; G-6-P-DH, glucose-6-phosphate dehydrogenase; *P<0.05, **P<0.01, ***P<0.001
 249

ters are also given in Table 2. Among these parameters, the levels of albu-
min and UIBC (markers of antioxidant activity) were decreased, while the
levels of uric acid (a peroxyl radical trapping antioxidant) were not changed.
Specific markers used to determine oxidative protein damage and oxida-
tive stress in the diabetic rats with and without ALA supplementation and in
the controls are given in Table 3. Plasma protein carbonyl levels were
increased and plasma thiol levels were decreased in the diabetic rats with
and without ALA supplementation compared with those of the controls
(Table 3). Plasma protein carbonyl levels were decreased in the diabetic rats
with ALA supplementation compared with those of the rats without ALA
supplementation (P<0.05). On the other hand, there was no significant dif-
ference between the plasma thiol levels in the two groups of diabetic rats.
Plasma lipid hydroperoxide levels were significantly increased in the diabet-
ic rats without ALA supplementation compared with those of the controls
(P<0.01) and the diabetic rats with ALA supplementation (P<0.05) (Table
3). Whole blood GSH levels in the diabetic rats with ALA supplementation
were significantly increased compared with those of the diabetic rats with-
out ALA supplementation (P<0.01). There was no significant difference be-
tween GSH peroxidase activities in the three groups of rats (Table 3). On the
other hand, glucose-6-phosphate dehydrogenase activity in both groups of
diabetic rats was not changed compared with that of the controls.

Discussion

After oral administration, ALA is present in the bloodstream [4, 13]. Eryth-
rocytes, platelets, lymphocytes, and other plasma constituents might there-
fore be targets for its actions [4, 12, 13]. The metabolic role of ALA has
been known for more than 40 years, but its effects when supplied exoge-
nously have become known only recently [12]. Exogenous ALA is reduced
intracellularly by at least two and possibly three enzymes, and it influences
a number of cell processes through the actions of its reduced form. These in-
clude direct radical scavenging, recycling of other antioxidants, and acceler-
ating GSH synthesis [12]. The dithiol nature of lipoate renders this com-
pound highly reactive against a number of reactive oxygen species, and it
also has the ability to regenerate oxidized antioxidants [4, 12, 13]. Both
ALA and DHLA have been shown to scavenge hydroxyl radicals in a metal-
catalysis system. During the course of lipid peroxidation, peroxyl radicals
that propagate the reaction are formed. DHLA can scavenge these radicals,
formed from both lipophilic and hydrophilic peroxyl radical generators. The
lipoate couple therefore represents a potent radical scavenging unit [4, 11,
13]. In our study, plasma lipid hydroperoxide levels of ALA-supplemented
diabetic rats were decreased compared with those of the unsupplemented di-
abetic rats. Hagen et al. [5] found that ALA supplementation for 2 weeks
decreased oxidative damage in old rats. Similarly, we found decreased oxi-
dative damage in streptozotocin-diabetic rats in our study.
The role of increased oxidative stress in the development of oxidative
protein damage in diabetes is currently a subject of great interest [9, 10].
Among the various oxidative modifications of amino acids in proteins, car-
250

bonyl formation may be an early marker for protein oxidation [15]. Transi-
tion metals such as iron are believed to be of crucial importance in cascade
of reactions, as they catalyse the autoxidation of glucose [11]. Asayama et
al. [1] reported that the decrease in antioxidant activity in diabetic rats was
due to the decrease in UIBC and albumin. While serum iron levels were not
different, decreased UIBC in both groups of diabetic rats in our study shows
the decreased metal-binding capacity which, in turn, contributes to the in-
creased oxidative stress and oxidative protein damage. However, we found
that plasma protein carbonyl levels of ALA-supplemented diabetic rats were
decreased compared with those of unsupplemented diabetic rats. This find-
ing may support our hypothesis that ALA supplementation prevents metal-
catalyzed protein oxidation. In a study with diabetic rats, thiol groups
proved to be useful as the total peroxyl radical-trapping antioxidant parame-
ter [1]. Asayama et al. [1] reported that plasma thiol levels of diabetic rats
were decreased compared with those of controls. Our findings are in concor-
dance with these findings with respect to plasma thiol levels. To our knowl-
edge, no investigation has previously been conducted in ALA-supplemented
diabetic rats. Both ALA and DHLA have high hydrophobicity, which allow
them to permeate biological membranes at a high rate. On the other hand,
the low negative redox potential makes ALA/DHLA a strong reductant [4,
13]. In accordance with the lower redox potential of ALA/DHLA compared
with GSH/GSH disulfide (GSSG), DHLA is a powerful reductant for
GSSG. There is a consensus that the GSH to GSSG ratio in diabetes is low-
er [3, 11]. The increased concentration of erythrocyte GSH in ALA-supple-
mented diabetic rats compared with nonsupplemented diabetic rats may sug-
gest that GSSG reduction is increased in ALA-supplemented diabetic rats.
Due to its biochemical characteristics, ALA can be considered a power-
ful effector of cellular metabolism at different levels and a potential thera-
peutic agent for the treatment of energy-impaired and redox-unbalanced dis-
eases [4]. Strategies to prevent metal-catalyzed protein oxidation in diabet-
ics, such as ALA supplementation, should be considered in future studies.
Acknowledgements. This work was supported by the Research Fund of the University of
Istanbul (Project number 1159/070998). The authors warrant that the experiments com-
ply with current national laws.

References
1. Asayama K, Nakane T, Uchida N, Hayashibe H, Dobashi K, Nakazawa S (1994)
Serum antioxidant status in streptozotocin-induced diabetic rats. Horm Metab Res
26:313–315
2. Beutler E, Duron O, Kelly BM (1963) Improved method for the determination of
blood glutathione. J Lab Clin Med 61:882–888
3. Brownlee M (l994) Glycation and diabetic complications. Diabetes 43:836–841
4. Bustamante J, Lodge JK, Marcocci L, Tritschler HJ, Packer L, Rihn BH (l998)
α-Lipoic acid in liver metabolism and disease. Free Radic Biol Med 24:1023–1039
5. Hagen TM, Ingersoll RT, Lykkesfeldt J, Liu J, Wehr CM, Vinarsky V, Bartholomew
JC, Ames BN (1999) (R)-α-Lipoic acid-supplemented old rats have improved mito-
chondrial function, decreased oxidative damage, and increased metabolic rate.
FASEB J 13:411–418
 251

6. Halliwell B, Aruoma OI (1991) DNA damage by oxygen-derived species. Its mecha-

nism and measurement in mammalian systems. FEBS Lett 281:9-l9
7. Han D, Tritschler HJ, Packer L (l995) α-Lipoic acid increases intracellular glutathi-
one in a human T-lymphocyte jurkat cell line. Biochem Biophys Res Commun
207:258–264
8. Hu ML (1994) Measurement of protein thiol groups and glutathione in plasma.
Methods Enzymol 233: 381–385
9. Liggins J, Furth AJ (l997) Role of protein bound carbonyl groups in the formation of
advanced glycation endproducts. Biochim Biophys Acta 1361:123–130
10. Miyata T, Inagi R, Asahi K, Yamada Y, Horie K, Sakai H, Uchida K, Kurokawa K
(1998) Generation of protein carbonyls by glycoxidation and lipoxidation reactions
with autooxidation products of ascorbic acid and polyunsaturated fatty acids. FEBS
Lett 437:24–28
11. Packer L (1993) The role of anti-oxidative treatment in diabetes mellitus. Dia-
betologia 36:1212–1213
12. Packer L (1998) Alpha-lipoic acid: a metabolic antioxidant which regulates NF-
kappa B signal transduction and protects against oxidative injury. Drug Metab Rev
30:245–275
13. Packer L, Witt EH, Tritschler HJ (l995) Alpha-lipoic acid as a biological antioxidant.
19:227–250
14. Podda M, Tritschler HJ, Ulrich H, Packer L (1994) α-Lipoic acid supplementation
prevents symptoms of vitamin E deficiency. Biochem Biophys Res Commun 204:
98–104
15. Reznick AZ, Packer L (l994) Oxidative damage to proteins:spectrophotometric
method for carbonyl assay. Methods Enzymol 233:357–363
16. Rubanyi GM (l993) The role of endothelium in cardiovascular homeostasis and dis-
eases. J Cardiovasc Pharmacol 22 [Suppl 4]:S1–S14
17. Stadtman ER (1990) Metal ion-catalyzed oxidation of proteins: biochemical mecha-
nism and biological consequences. Free Radic Biol Med 9:315–325
18. Sundari PN, Wilfred G, Ramakrishna B (l997) Does oxidative protein damage play a
role in the pathogenesis of carbon tetrachloride-induced liver injury in the rat?
Biochim Biophys Acta 1362:169–176
19. Takenaka Y, Miki M, Yasuda H, Mino M (1991) The effect of α-tocopherol as an an-
tioxidant on the oxidation of membran protein thiols induced by free radicals gener-
ated in different sites. Arch Biochem Biophys 285:344–350
20. Wolff SP (1994) Ferrous ion oxidation in presence of ferric ion indicator xylenol or-
ange for measurement of hydroperoxides. Methods Enzymol 233:182–189