Biochimica et Biophysics Elsevier
BBALIP
63
Acta, 1047 (1990) 63-69
53504
Diabetic heart and kidney exhibit increased resistance to lipid peroxidation Narasimham
L. Parinandi,
Ed W. Thompson
The Hormel Institute,
(Revised
Key words:
Diabetes;
Oxygen
University of Minnesota,
and Harald
H.O. Schmid
Austin, MN (U.S.A.)
(Received 27 February 1990) manuscript received 19 June 1990)
radical;
Lipid peroxidation;
Antioxidant;
(Heart);
(Kidney)
Alloxan-diabetic rats and age-matched controls were killed after 6 weeks of diabetes; heart and kidneys were removed and assayed for thiobarbituric acid-reactive substances (TRARS), lipid hydroperoxides, lipid phosphorus, total fatty acid composition and glutathione. Tissue homogenates from a second group of diabetic and control rats were incubated in oxygen-saturated buffer with and without the free radical generating system Fe2+/ascorbate (O.l/l.O mM) and were assayed for lipid peroxidation. Diabetic hearts contained markedly lower levels of TRARS and lipid hydroperoxides (40% and 18%, respectively) than control hearts, whereas differences in TRARS were less pronounced in kidneys (9%). Incubation of homogenates of both organs in the presence or absence of Fe2+/ascorbate for up to 2 h yielded significantly lower levels of TRARS and lipid hydroperoxides with diabetic tissue. Diabetic hearts and kidneys contained higher levels of glutathione (28% and 13% over controls) and both diabetic tissues showed much higher linoleate/arachidonate ratios than did the controls (9.86 vs. 2.56 for heart, 2.01 vs. 0.86 for kidney). We conclude that diabetic tissues develop enhanced defense systems against oxidative stress and we assume that the lower levels of arachidonate contribute to their resistance to lipid per-oxidation as well.
Introduction Degenerative changes in the heart and kidney are almost universal complications of insulin-dependent diabetes mellitus. Cardiomyopathy and nephropathy share a common etiology, i.e., the altered hormonal and biochemical milieu resulting from the lack of insulin. Many studies of these and other diabetic sequelae have thus been based on the hypothesis that similar mechanisms of cytopathology are involved even though the structural, biochemical, and functional alterations are quite dissimilar in different organs and tissues. One possible mechanism is cellular damage from cytotoxic oxygen free radical species. They are generated through various metabolic pathways in all cells and must be rapidly and efficiently scavenged if cellular damage is to be pre-
Abbreviations: EDTA, ethylenediaminetetraacetic acid; Hepes, 4-(2hydroxyethyl)-l-piperazineethanesulfonic acid; DTNB, 5,5’-dinitrobis(2nitrobenzoic acid); TBA, thiobarbituric acid; TBARS, thiobarbituric acid-reactive substances; GSH, reduced glutathione; TCA, trichloroacetic acid; GC, gas chromatography. Correspondence: H.H.O. Schmid, The Hormel Institute, University Minnesota, 801 16th Avenue NE, Austin, MN 55912, U.S.A.
0005-2760/90/$03.50
0 1990 Elsevier Science Publishers
of
B.V. (Biomedical
vented [l]. Indeed, there is evidence that diabetes alters free radical metabolism in blood [2,3] and tissues [4,5] but these alterations are quite heterogeneous, and a clear picture of how they may be involved in organspecific complications has yet to emerge. Insulin-dependent diabetes often leads to coronary atherosclerosis [6] related to hypertriglyceridemia and altered lipoprotein profiles [7], resulting in vascular insufficiency, ischemia and a markedly increased risk of myocardial infarction. A second pattern of diabetic heart disease, diabetic cardiomyopathy, is less well defined, presenting as contractile dysfunction in individuals with no evidence of macrovascular disease [8,9] and leading to cardiac pump failure. The structural pattern of progression for this cardiomyopathy in the alloxan diabetic rat shows that the cardiocytes, capillaries and extracellular matrix are all involved and that it can be almost completely reversed by insulin replacement [lo]. The diabetic kidney exhibits a characteristic pattern of changes in the glomerulus producing initial hyperfiltration, but eventually leading to renal insufficiency or complete kidney failure. The primary pathology involves marked thickening of the glomerular basement membrane [ll] and functional alteration of the mesangial cells in disposing of filtration residues [12]. Division)
64 Because membrane alterations caused by free radical-induced lipid peroxidation may be a common mechanism in the development of both diabetic cardiomyopathy and renal failure, we examined hearts and kidneys of diabetic rats for the presence of lipid peroxidation products and the capacity of these tissues to generate them. Materials and Methods Reagents Alloxan, DTNB, Hepes, thiobarbituric acid and reduced glutathione were all obtained from Sigma (St. Louis, MO). Malondialdehyde bis-methyl acetal was purchased from Aldrich (Milwaukee, WI). All other chemicals used were of AR grade. Induction of diabetes Male Sprague-Dawley rats were obtained from a commercial supplier (Biolab, St. Paul, MN) and housed individually in a temperature (70 o F) and humidity controlled facility with a 12 h light/l2 h dark cycle. All animals had free access to food and water throughout the study. Diabetes was induced in rats with body weights of 150-180 g by a single injection into a tail vein of 4% alloxan in sterile saline at a dose of 55 mg/kg body weight as previously described [lo]. Control animals received the same volume of saline. 3 days later and at biweekly intervals throughout the study, blood samples were taken from the tail and blood glucose levels were measured with an AccuChek II meter (Boehringer Mannheim, Indianapolis, IN). At these times, urine glucose, ketones and protein were also estimated by dipstick (Boehringer Mannheim, Indianapolis, IN). Each rat was weighed weekly and its rate of weight gain was calculated. Only those alloxan-injected animals which had blood glucose values above 300 mg/dl(l6.8 mM) and showed little or no weight gain were considered insulin-deficient. Saline-injected rats with nonfasting blood glucose levels below 120 mg/dl (6.7 mM) and gaining weight at normal rates were considered nondiabetic. Any animal not meeting these criteria, e.g., diabetic with blood glucose levels below 300 mg/dl or hyperglycemic but weight gaining, was excluded from the study. This effectively limited the study to the more severe end of the spectrum of diabetes, minimizing experimental variations due to differing degrees of the disease. Tissue isolation and sample preparation Diabetic and age-matched control animals were killed 6 weeks after induction of diabetes and the heart and right kidney of each were removed for analysis. Each rat was fully anesthetized by intraperitoneal injection of sodium pentobarbital (65 mg/kg body weight) and the abdomen and thorax were opened by midline incision.
The heart was quickly excised, rinsed three times in ice-cold saline and its vasculature was cleared of blood by perfusion through an aortic cannula with 25 mM potassium phosphate buffer (pH 7.3) containing 90 mM NaCl and 25 mM KC1 to arrest the heart in diastole and decrease its oxygen demand [lo]. Immediately after excision of the heart, a second cannula was placed in the distal stump of the aorta and advanced to its midthoracic descending section. The abdominal aorta was occluded immediately below the renal arteries and the kidneys were cleared of blood by in situ perfusion with lactated Ringer’s solution (100 mM NaCl, 20 mM sodium lactate, 4 mM KCl, 2 mM CaCl, (pH 7.3)). For both organs, perfusion began within 60 s of thoracotomy and was continued for 30 to 120 s. The atria and vessels were trimmed from the heart, the renal vessels and ureters were trimmed from the right kidney and the organs were frozen in liquid nitrogen. They were later thawed, minced well with scissors and a 10% (wt/vol) homogenate was prepared in ice-cold 125 mM Hepes/lSO mM NaCl buffer (pH 7.4) with a Polytron tissue homogenizer (Brinkmann Instruments, Westbury, NY) at a setting of 7 for 1 min. The homogenates were used for the chemical analyses described below. Lipid peroxidation assay Lipid peroxidation products in the tissue homogenates were determined as TBARS (thiobarbituric acid-reactive substances) and also as lipid hydroperoxides according to Buege and Aust [13]. To 1.0 ml of the homogenate, 2.0 ml of TCA-TBA-HCl (15%/ 0.375%/0.15 M) were added and mixed thoroughly in a screw-capped test tube. The mixture was heated for 15 min at 80°C in a hot water bath and cooled to room temperature, then centrifuged at 1000 X g for 10 min. The absorbance of the clear supernatant was measured at 535 nm in a Beckman DU-5 spectrophotometer against an appropriate blank. The amounts of TBARS were calculated using a standard curve prepared with malondialdehyde bis-methyl acetal and expressed as nmol TBARS/pmol lipid phosphorus. Total lipids from 1.0 ml tissue homogenate were extracted according to Folch et al. [14]. The lipid extract was dried under a stream of nitrogen at 40° C and redissolved in 0.5 ml of chloroform/methanol (2 : 1). An aliquot of 200 ~1 of the redissolved lipid extract was assayed for lipid hydroperoxides. This aliquot was dried under a stream of nitrogen at 40°C and, under nitrogen, 1.0 ml of acetic acid/chloroform (3 : 2, v/v) was added, mixed and was immediately followed by the addition of 50 ~1 KI solution (6.0 g KI/5.0 ml of water). The reactants were mixed quickly, capped and kept in the dark for exactly 5 min. To the mixture, 3.0 ml of cadmium acetate (0.5 g/100 ml of water) was added, mixed vigorously and centrifuged at 1000 X g
65
for 10 min. The absorbance of the upper layer was measured at 353 nm in a Beckman DU-5 spectrophotometer against a suitable blank. The amount of lipid hydropero~des was calculated using a molar extinction coefficient of 1.73. f04 M-i and expressed as nmol/pmol lipid phosphorus.
dialdehyde bis-methyl acetal and expressed as nmol/ ,umol lipid phosphorus. The amount of lipid hydroperoxides was calculated using the molar extinction coefficient of 1.73 . lo4 M-’ and expressed as nmol/~mol lipid phosphorus. Statistical analyses
Lipid extraction and fatty acid analysis
Total fatty acid composition of tissue lipids was determined as follows. An aliquot of the lipid extract was transesterified in methanolic NaOH (0.2 M) with appropriate amounts of methyl heptadecanoate as an internal standard. The methyl esters were analyzed by GC on a column packed with 10% SP-2330 on 100/200 Chromosorb WAW (Supelco) using a Packard model 428 gas c~omato~aph connected to a Spectra-Physics 4270 integrator. Lipid phosphorus was determined according to Bartlett [15], and protein was determined according to Lowry et al. [16]. ~ete~i~ation
of reduced gIutathione
Reduced glutathione (GSH) contents of the tissues were determined according to Beutler et al. [17]. To 1.0 ml of the tissue homogenate, 1.5 ml of precipitating solution (1.67 g metaphosphoric acid + 0.1 g sodium EDTA + 30 g sodium chloride per 100 ml of distilled water) was added, mixed and allowed to stand for 5 min at room temperature. The mixture was centrifuged at 2000 x g for 10 min. An aliquot of 1.0 ml of the supematant was mixed with 4.0 ml phosphate solution (0.3 M Na,HPO,) followed by the addition of 0.5 ml DTNB solution (0.04 g/100 ml of 1% sodium citrate). The reactants were quickly mixed and the absorbance at 412 nm was measured in a Beckman DU-5 spectrophotometer against an appropriate blank. The amount of GSH was calculated using a standard curve prepared with GSH and expressed as pg/g wet weight of the tissue. Induction of lipid peroxidation in vitro
The extent of lipid peroxidation in the tissue homogenates in vitro, in the absence or presence of the free radic~-generating system Fe’+-ascorbate (0.1 mM FeSO, and 1.0 mM ascorbate), was studied as follows. Hearts and kidneys were obtained from a second group of diabetic rats, 6 weeks after alloxan injection, and from age- and sex-matched controls. In a final volume of 1.0 ml, 0.3 ml of the 10% tissue homogenate was incubated with 0.1 mM Fe*’ as FeSO, and 1.0 mM ascorbate in 25 mM Hepes-150 mM NaCl buffer (pH 7.4) at 37 o C on a shaking water bath. After intervals of 0, 30,60 or 120 min, samples were removed and assayed for the TBARS and/or lipid hydroperoxides [13], as described above. The amounts of TBARS were calculated using a standard curve prepared with malon-
The standard deviations of the means were calculated and the data were subjected to Student’s t-test for statistical significance. Results General characteristics of the diabetic rats
More than 75% of the animals that received alloxan developed severe chronic diabetes. The remaining animals either died within a few days of alloxan injection or were euthanized. The body, heart and kidney weights and blood glucose values of the diabetic animals and of their age-matched controls are presented in Table I. As expected, diabetic animals were markedly hyperglycemic and gained or lost body weight at a fraction of a gram per day, while nondiabetic animals were normoglycemic and gained more than 5 g per day during the study. The weights of both hearts and kidneys were less in the diabetic animals than in the controls, but when expressed relative to body weight, both heart and kidney weights in the diabetic animals were markedly elevated, indicating that the effect of insulin deficiency upon these organs was less than upon overall growth. The loss of both subcutaneous and retroperitoneal fat was striking in all of the diabetic animals. All the diabetic rats developed cataracts in one or both eyes, not seen in any of the controls. Urine glucose in the diabetic rats, without exception, exceeded the maximum value on the dipstick, corresponding to 1 g/dl, and urine ketones typically fell between the readings corresponding to 18 and 23 mM acetoacetate.
TABLE
I
General features
of the experimentul
model
All values are X f S.D. Each group
Body weight (g) at sacrifice Weight gain per day (g) Blood glucose (mg/dl) Heart weight(g) Heart weight(g) per body weight (100 g) Kidney weight a (g) Kidney weight (g) per body weight (100 g)
consisted
of eight animals.
Normal
Diabetic
400.6 f39.40 5.07f 0.74 89.1 f13.3 1.341 0.14
162.8 0.01 413.0 0.68
k 20.8 b f 0.57 b f 31.2 b * 0.05 b
0.34* 1.79*
0.03 0.11
0.42* 1.565
0.45 f
0.03
0.98 f 0.16 b
a The right kidney was used for all determinations. b Significantly different (P < 0.01) from the control tailed Student’s t-test.
value
0.06P 0.16 b
by two-
66 TABLE
II
In vivo TBARS, lipid hydroperoxides, iota1 glutathione and All values are X f SD. Each value is an average
lipid phosphorus contents in the hearts and kidneys of normal and diabetic rats (n = I/group)
of four individual
determinations.
TBARS,
thiobarbituric
Heart normal TBARS (nmol/pmol lipid P) Decrease in TBARS from normal Lipid hydroperoxides (nmol/~mol lipid P) Decrease in lipid hydroperoxides from normal Glutathine (GSH) (pg/g wet wt. tissue) Increase in GHS from normal Lipid phosphorus (pmol/g wet wt. tissue) a Significantly b n=3.
different
3.78 rf:0.77 29.30 f 2.02 238.26 k 9.56 28.20 + 2.50
(P < 0.05) from the normal
value by Student’s
Neither glucose nor ketones were detected in the control animals. Trace amounts of urine protein were occasionally noted in all groups.
acid-reactive
substances.
Kidney diabetic
normal
2.08+ 0.51a 45% 24.16 k 1.03 a.h 18% 305.50 k 78.65 28% 24.30 + 0.83
diabetic
4.10*
0.20
6.37+
0.92
3.74f 0.57 9% 5.97+ 1.74 6% 287.72 f 72.73 13% 32.03* 1.73”
238.43 + 63.31 28.23 +
1.05
t-test.
Products of lipid peroxidation in vivo In vivo contents of TBARS, lipid hydroperoxides, GSH and lipid phosphorus were determined in the
80 -
A
NORMAL (t Fe 2+/ASC)
50 NORMAL (t Fe *+/AX)
OIABI ETIC I\rre I r_ 2+/ASC)
DIABETIC (tFe 2+/ASC)
NORMAL
DIABETIC
I
0’
O-
0
1 HOURS
2
I
0
1
2
HOURS
Fig. 1. Formation of TBARS in vitro in the homogenates of diabetic (0) and normal (0) rat heart (Panel A) and kidney (Panel B) in the presence medium contained 3.6-5.0 mg ) and absence (---) of 0.1 mM Fe2+ (FeSO,) and 1.0 mM ascorbate. The final 1.0 ml of incubation (protein and was incubated at 37 o C and pH 7.4. TBARS were determined at the end of each incubation period as described in Materials and Methods. Each point is an average f S.D. of three to four individual experiments.
67 hearts and kidneys after 6 weeks of diabetes and in the hearts and kidneys of age-matched control rats. The values are given in Table II. The amounts of TBARS in the diabetic hearts were 45% lower than in controls (P < 0.05). Even though the kidneys of the diabetic rats contained decreased amounts of TBARS (9’%), this difference was not quite statistically significant. Similarly, the lipid hydroperoxide content in the hearts of &week diabetic animals was significantly (18%) lower than that in the normal hearts (P < O.OS), whereas the diabetic kidneys showed only slightly lower amounts of lipid hydroperoxides (6%). The GSH contents of the diabetic hearts and kidneys were elevated (28% and 13% respectively), but these increases were not statistically significant. Diabetic hearts contained slightly lower amounts of lipid phosphorus on a wet weight basis, whereas diabetic kidneys had slightly higher amounts compared to controls. Lipid pero~~datio~ in vitro
In vitro lipid peroxidation assayed as the formation of TBARS and/or lipid hydroperoxides in the homogenates of hearts and kidneys of diabetic and control rats, was determined in the absence or presence of Fe’+-ascorbate. The incubations were carried out for up to 2 h and the data presented in Fig. 1 were obtained. After each time of incubation, the formation of TBARS in the homogenates of diabetic hearts, either in the absence or presence of Fe2+-ascorbate, was significantly lower than in those of control animals (Fig. 1A). In the presence of Fe2+-ascorbate, the heart homogenates showed an increase in the formation of TBARS with increased incubation times. The extent of formation of TBARS in the homogenates of diabetic kidney was significantly lower at 30 and 60 min of incubation
in the presence of Fe2+- ascorbate and at 2 h reached a level lower than, but not significantly different from, that of the corresponding normal kidney homogenates. In the absence of Fe2+-ascorbate there were measurable, but not significant, decreases in the formation of TBARS in the diabetic animals (Fig. 1B). The amount of lipid hydroperoxides formed in the heart homogenates of 6-week diabetic animals in either the absence or presence of Fe2+-ascorbate was also lower than in controls. After incubating homogenates of diabetic hearts for 1 h in the absence of Fe*+/ascorbate, lipid hydroperoxide levels were 23.10 + 3.56 nmol/pmol lipid P (n = 4) and in the presence of Fe’+/ascorbate (0.1 mM/l.O mM) 28.00 + 0.90 nmol/pmol lipid P (n = 3). Homogenates of control hearts exhibited hydroperoxide levels of 28.70 of:5.64 nmol,/,umol lipid P (n = 4) in the absence and 42.35 1_5.60 nmol/pmol lipid P (n = 4) in the presence of Fe2+/ascorbate. The difference in hydroperoxide levels obtained with homogenates from normal and diabetic hearts in the presence of Fe’+/ascorbate was sig~ficant at P -C0.025. Fatty acid composition of total lipids
The lipid extracts of heart and kidney of diabetic rats and of control animals were analyzed for fatty acid composition and the data are summarized in Table III. The levels of palmitic acid (16 : 0) in the diabetic heart were lower than in the controls, whereas in the diabetic kidney they were slightly elevated. More importantly, both tissues of the diabetic animals exhibited significantly elevated levels of linoleate (18 : 2) and reduced levels of arac~donate (20: 4) resulting in increased linoleate/arachidonate ratios. Little or no change was observed in other long-chain polyunsaturated fatty acids (20 : 5, 22 : 5, 22 : 6) as the result of diabetes.
TABLE III Fatly acid composition (mol%) of the total lipids in the heart and kidney of normal and diabetic rats Data represent means* SD. The right kidney was used for all determinations. a P -C0.001, b P < 0.005, ’ P i 0.01. FA (chain length : no. of double bonds)
Heart (n=4) normal
diabetic
normal
diabetic
16:0 16:l 18:O 18:l 18:2(n -6) 20:3(n-6) 20:4(n -6) 20:5(n-3) 22:5(n -6) 22:6(n-3) 18:2/20:4
20.4f 2.3 1.9*0.7 18.5 * 2.2 20.1 f 2.6 23.8 kO.8 0.2*0.05 9.2 f 2.0 0.2 f 0.1 1.0*0.1 4.5*1.1 2.59
13.7*1.1 0.4*0.1 2o.oi 1.1 19.1 k 1.6 37.5 + 2.0 a 0.4*0.1 3.8*o.3c 0.4+0.2 1.1 f0.3 3.4kO.65 9.87
23.0k0.8 1.2f0.5 22.5 * 1 .o 13.0*0.X 17.0*0.3 0.8 kO.04 19.7f0.8 0.7 f 0.1 0.3 * 0.05 1.s*o.2 0.86
26.6 f 3.0 1.5 f 0.9 16.7kO.8 18.5 i 1.0 23.0 f 0.8 a 0.8 50.1 11.4+2.0b 1.0*0.1 0.3*0.1 2.4k0.3 2.01
Kidney (n=4)
68 Discussion Based upon the results of the present study, we conclude that heart and kidney from diabetic rats exhibit increased resistance to free radical-induced lipid peroxidation. Endogenous lipid peroxide levels were measured as TBARS which includes end products of lipid peroxidation such as malondialdehyde and other carbonyls, and as lipid hydroperoxides. We observed that in vivo levels of TBARS and of lipid hydroperoxides in the heart were significantly lower in diabetic rats. In addition, lipid peroxidation induced in heart homogenates of diabetic rats in the absence or presence of Fe’+ -ascorbate indicated significantly lower rates of TBARS formation. Diabetic kidney, either in vivo or in vitro, exhibited somewhat lower levels of TBARS and decreased capacity to form lipid peroxidation products, but the decrease in the lipid peroxidation potential was not as dramatic as that exhibited by the diabetic heart. The decreased levels of TBARS and lipid peroxides in the hearts and kidneys of diabetic animals may be attributed to several factors related to increased resistance to oxidative damage. Increased levels of non-enzymatic antioxidants such as glutathione have been noted previously in the livers of streptozotocin-diabetic rats [4]. In human diabetic individuals, elevated activity of glutathione reductase was observed [18]. This would result in enhanced reduction of oxidized glutathione, which could play a major role in glutathione-dependent antioxidant processes [19]. In addition, increased activities of the protective enzymes superoxide dismutase, catalase, and glutathione peroxidase in the tissues of human diabetics as well as experimental models were reported [4]. Diabetic rats, 12 weeks after administration of streptozotocin, exhibited elevated activities of catalase and glutathione reductase in the heart and of glutathione reductase and Cu-Zn superoxide dismutase in pancreas [5]. The observed significant decreases in the levels of TBARS and lipid peroxides in the diabetic hearts in the current study might, therefore, be due to increased activities of myocardial catalase and glutathione reductase. Diabetic rat kidney, 12 weeks after streptozotocin administration showed only increased activities of glutathione peroxidase, whereas the activities of catalase and superoxide dismutase were decreased [5]. This may explain why diabetic kidney was less resistant than the diabetic heart to lipid peroxidation compared to normal controls. The fatty acid composition of tissue lipids also appears to be a contributing factor in their susceptibility to peroxidation [20,21]. Rat liver endoplasmic reticulum containing higher levels of eicosapentaenoic acid (20 : 5 (n - 3)) and docosahexaenoic acid (22 : 6 (n - 3)) as the result of dietary fish oil showed higher levels of TBARS compared to controls having lower levels of eicosa-
pentaenoic and docosahexaenoic acids among liver lipids [20]. Novikoff heptoma cells were found to resist in vitro oxidative stress, and this was partly attributed to the lower levels of arachidonic acid (20 : 4 (n - 6)) and docosahexaenoic acid (22 : 6 (n - 3)) in these cells compared to normal liver cells [21]. A general trend in the fatty acid composition in diabetic tissues is that linoleate (18 : 2 (n - 6)) increases and arachidonate (20 : 4 (n - 6)) decreases [22-261. The results of the current work agree with those of ethers regarding elevated linoleic acid and reduced arachidonic acid levels in the diabetic heart and kidney. Decreased unsaturation of tissue lipids is thus likely to have contributed to their resistance to lipid peroxidation. Recently, Tani and Neely [27] observed that diabetic rat hearts were more resistant to in vitro ischemia and showed a greater recovery upon reperfusion after ischemic stress than normal rat hearts. Because postischemic reperfusion injury in the myocardium is likely to be mediated by oxygen free radicals [28], this observation [27] supports our present results which indicate that diabetic rat heart and kidney develop a resistance to oxidative stress. Acknowledgements Ann Schmeichel and Barbara Weis provided excellent technical assistance. This work was supported in part by PHS Program Project Grants HL08214 and NS14304, Training Grant HL07311, the Minnesota Affiliate of the American Heart Association and the Hormel Foundation. References 1 Halliwell, B. and Gutteridge, J.M.C. (1984) B&hem. J. 219, 1-14. 2 Sato, Y., Hotta, N., Sakamoto, N., Matsuoka, S., Oshishi, N. and Yagi, K. (1979) B&hem. Med. 21, 104-107. 3 Karpen, C.W., Pritchard, Jr., K.A., Arnold, J.H., Cornell, D.G. and Panaganamala, R.V. (1982) Diabetes 31, 947-951. 4 Oberley, L.W. (1988) Free Radical Biol. Med. 5, 113-124. 5 Wohaieb, S.A. and Godin, D.V. (1987) Diabetes 36, 1014-1018. 6 Beach, K.W. and Strandness, D.E. (1980) Diabetes 29, 882-888. 7 Winocour, P.H., Durrington, P.N., Ishola, M. and Anderson, D.C. (1986) Lancet 1, 117661179. 8 Regan, T.J., Lyons, M.M., Ahmed, S.S., Levinson, G.E., Oldewurtel, H.A., Ahmad, M.R. and Haider, B. (1977) J. Clin. Invest. 60, 885-889. 9 Hamby, R.I., Zoneraich, S. and Sherman, L. (1974) J. Am. Med. Assoc. 229, 1749-1754. 10 Thompson, E.W. (1988) Am. J. Anat. 182, 270-282. 11 Cohen, A.J., McGill, P.D., Rossetti, R.G., Guberski, D.L. and Like, A.A. (1987) Diabetes 36, 944951. 12 Farquhar, M.G. (1975) Kidney Int. 8, 197-211. 13 Buege, J.A. and Aust, S.D. (1978) Methods Enzymol. LII, 302-310. 14 Folch, J., Lees, M. and Sloane Stanley, G.H. (1957) J. Biol. Chem. 226, 497-509. 15 Bartlett, G.R. (1959) J. Biol. Chem. 234, 466-468. 16 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275.
69 17 Beutler, E., Duron, 0. and Kelly, B.M. (1963) J. Lab. Clin. Med. 61, 882-888. 18 Long, W.K. and Carson, P.E. (1961) B&hem. Biophys. Res. Commun. 5, 394-399. 19 Sies, H. (1987) Am. Rev. Respir. Dis. 136, 478-480. 20 Hammer, C.T. and Wills, E.D. (1978) B&hem. J. 174, 584-593. 21 Cheesman, K.H., Collins, M., Proudfoot, K., Slater, T.F., Burton, G.W., Webb, A.C. and Ingold, K.U. (1986) Biochem. J. 235, 507-514. 22 Holman, R.T., Johnson, S.B., Gerrard, J.M., Mauer, S.M., Kupcho-Sandberg, S. and Brown, D.M. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 2375-2379.
23 Clark, D.L., Hamel, F.G. and Queener, S.F. (1983) Lipids 18, 696-705. 24 Huang, Y.S., Horrobin, D.F., Mar&u, M.S., Mitchell, J. and Ryan, M.A. (1984) Lipids 19, 367-370. 25 Lin, C.-J., Peterson, R. and Eichberg, J. (1985) Neurochem. Res. 10, 1453-1465. 26 Chattopadhyay, J., Thompson, E.W. and S&mid, H.H.O. (1990) Lipids 25, 307-310. 27 Tani, M. and Neely, J.R. (1988) Circ. Res. 62, 931-940. 28 McCord, J.M. (1985) N. Engl. J. Med. 312, 159-163.
BBALIP
63
Acta, 1047 (1990) 63-69
53504
Diabetic heart and kidney exhibit increased resistance to lipid peroxidation Narasimham
L. Parinandi,
Ed W. Thompson
The Hormel Institute,
(Revised
Key words:
Diabetes;
Oxygen
University of Minnesota,
and Harald
H.O. Schmid
Austin, MN (U.S.A.)
(Received 27 February 1990) manuscript received 19 June 1990)
radical;
Lipid peroxidation;
Antioxidant;
(Heart);
(Kidney)
Alloxan-diabetic rats and age-matched controls were killed after 6 weeks of diabetes; heart and kidneys were removed and assayed for thiobarbituric acid-reactive substances (TRARS), lipid hydroperoxides, lipid phosphorus, total fatty acid composition and glutathione. Tissue homogenates from a second group of diabetic and control rats were incubated in oxygen-saturated buffer with and without the free radical generating system Fe2+/ascorbate (O.l/l.O mM) and were assayed for lipid peroxidation. Diabetic hearts contained markedly lower levels of TRARS and lipid hydroperoxides (40% and 18%, respectively) than control hearts, whereas differences in TRARS were less pronounced in kidneys (9%). Incubation of homogenates of both organs in the presence or absence of Fe2+/ascorbate for up to 2 h yielded significantly lower levels of TRARS and lipid hydroperoxides with diabetic tissue. Diabetic hearts and kidneys contained higher levels of glutathione (28% and 13% over controls) and both diabetic tissues showed much higher linoleate/arachidonate ratios than did the controls (9.86 vs. 2.56 for heart, 2.01 vs. 0.86 for kidney). We conclude that diabetic tissues develop enhanced defense systems against oxidative stress and we assume that the lower levels of arachidonate contribute to their resistance to lipid per-oxidation as well.
Introduction Degenerative changes in the heart and kidney are almost universal complications of insulin-dependent diabetes mellitus. Cardiomyopathy and nephropathy share a common etiology, i.e., the altered hormonal and biochemical milieu resulting from the lack of insulin. Many studies of these and other diabetic sequelae have thus been based on the hypothesis that similar mechanisms of cytopathology are involved even though the structural, biochemical, and functional alterations are quite dissimilar in different organs and tissues. One possible mechanism is cellular damage from cytotoxic oxygen free radical species. They are generated through various metabolic pathways in all cells and must be rapidly and efficiently scavenged if cellular damage is to be pre-
Abbreviations: EDTA, ethylenediaminetetraacetic acid; Hepes, 4-(2hydroxyethyl)-l-piperazineethanesulfonic acid; DTNB, 5,5’-dinitrobis(2nitrobenzoic acid); TBA, thiobarbituric acid; TBARS, thiobarbituric acid-reactive substances; GSH, reduced glutathione; TCA, trichloroacetic acid; GC, gas chromatography. Correspondence: H.H.O. Schmid, The Hormel Institute, University Minnesota, 801 16th Avenue NE, Austin, MN 55912, U.S.A.
0005-2760/90/$03.50
0 1990 Elsevier Science Publishers
of
B.V. (Biomedical
vented [l]. Indeed, there is evidence that diabetes alters free radical metabolism in blood [2,3] and tissues [4,5] but these alterations are quite heterogeneous, and a clear picture of how they may be involved in organspecific complications has yet to emerge. Insulin-dependent diabetes often leads to coronary atherosclerosis [6] related to hypertriglyceridemia and altered lipoprotein profiles [7], resulting in vascular insufficiency, ischemia and a markedly increased risk of myocardial infarction. A second pattern of diabetic heart disease, diabetic cardiomyopathy, is less well defined, presenting as contractile dysfunction in individuals with no evidence of macrovascular disease [8,9] and leading to cardiac pump failure. The structural pattern of progression for this cardiomyopathy in the alloxan diabetic rat shows that the cardiocytes, capillaries and extracellular matrix are all involved and that it can be almost completely reversed by insulin replacement [lo]. The diabetic kidney exhibits a characteristic pattern of changes in the glomerulus producing initial hyperfiltration, but eventually leading to renal insufficiency or complete kidney failure. The primary pathology involves marked thickening of the glomerular basement membrane [ll] and functional alteration of the mesangial cells in disposing of filtration residues [12]. Division)
64 Because membrane alterations caused by free radical-induced lipid peroxidation may be a common mechanism in the development of both diabetic cardiomyopathy and renal failure, we examined hearts and kidneys of diabetic rats for the presence of lipid peroxidation products and the capacity of these tissues to generate them. Materials and Methods Reagents Alloxan, DTNB, Hepes, thiobarbituric acid and reduced glutathione were all obtained from Sigma (St. Louis, MO). Malondialdehyde bis-methyl acetal was purchased from Aldrich (Milwaukee, WI). All other chemicals used were of AR grade. Induction of diabetes Male Sprague-Dawley rats were obtained from a commercial supplier (Biolab, St. Paul, MN) and housed individually in a temperature (70 o F) and humidity controlled facility with a 12 h light/l2 h dark cycle. All animals had free access to food and water throughout the study. Diabetes was induced in rats with body weights of 150-180 g by a single injection into a tail vein of 4% alloxan in sterile saline at a dose of 55 mg/kg body weight as previously described [lo]. Control animals received the same volume of saline. 3 days later and at biweekly intervals throughout the study, blood samples were taken from the tail and blood glucose levels were measured with an AccuChek II meter (Boehringer Mannheim, Indianapolis, IN). At these times, urine glucose, ketones and protein were also estimated by dipstick (Boehringer Mannheim, Indianapolis, IN). Each rat was weighed weekly and its rate of weight gain was calculated. Only those alloxan-injected animals which had blood glucose values above 300 mg/dl(l6.8 mM) and showed little or no weight gain were considered insulin-deficient. Saline-injected rats with nonfasting blood glucose levels below 120 mg/dl (6.7 mM) and gaining weight at normal rates were considered nondiabetic. Any animal not meeting these criteria, e.g., diabetic with blood glucose levels below 300 mg/dl or hyperglycemic but weight gaining, was excluded from the study. This effectively limited the study to the more severe end of the spectrum of diabetes, minimizing experimental variations due to differing degrees of the disease. Tissue isolation and sample preparation Diabetic and age-matched control animals were killed 6 weeks after induction of diabetes and the heart and right kidney of each were removed for analysis. Each rat was fully anesthetized by intraperitoneal injection of sodium pentobarbital (65 mg/kg body weight) and the abdomen and thorax were opened by midline incision.
The heart was quickly excised, rinsed three times in ice-cold saline and its vasculature was cleared of blood by perfusion through an aortic cannula with 25 mM potassium phosphate buffer (pH 7.3) containing 90 mM NaCl and 25 mM KC1 to arrest the heart in diastole and decrease its oxygen demand [lo]. Immediately after excision of the heart, a second cannula was placed in the distal stump of the aorta and advanced to its midthoracic descending section. The abdominal aorta was occluded immediately below the renal arteries and the kidneys were cleared of blood by in situ perfusion with lactated Ringer’s solution (100 mM NaCl, 20 mM sodium lactate, 4 mM KCl, 2 mM CaCl, (pH 7.3)). For both organs, perfusion began within 60 s of thoracotomy and was continued for 30 to 120 s. The atria and vessels were trimmed from the heart, the renal vessels and ureters were trimmed from the right kidney and the organs were frozen in liquid nitrogen. They were later thawed, minced well with scissors and a 10% (wt/vol) homogenate was prepared in ice-cold 125 mM Hepes/lSO mM NaCl buffer (pH 7.4) with a Polytron tissue homogenizer (Brinkmann Instruments, Westbury, NY) at a setting of 7 for 1 min. The homogenates were used for the chemical analyses described below. Lipid peroxidation assay Lipid peroxidation products in the tissue homogenates were determined as TBARS (thiobarbituric acid-reactive substances) and also as lipid hydroperoxides according to Buege and Aust [13]. To 1.0 ml of the homogenate, 2.0 ml of TCA-TBA-HCl (15%/ 0.375%/0.15 M) were added and mixed thoroughly in a screw-capped test tube. The mixture was heated for 15 min at 80°C in a hot water bath and cooled to room temperature, then centrifuged at 1000 X g for 10 min. The absorbance of the clear supernatant was measured at 535 nm in a Beckman DU-5 spectrophotometer against an appropriate blank. The amounts of TBARS were calculated using a standard curve prepared with malondialdehyde bis-methyl acetal and expressed as nmol TBARS/pmol lipid phosphorus. Total lipids from 1.0 ml tissue homogenate were extracted according to Folch et al. [14]. The lipid extract was dried under a stream of nitrogen at 40° C and redissolved in 0.5 ml of chloroform/methanol (2 : 1). An aliquot of 200 ~1 of the redissolved lipid extract was assayed for lipid hydroperoxides. This aliquot was dried under a stream of nitrogen at 40°C and, under nitrogen, 1.0 ml of acetic acid/chloroform (3 : 2, v/v) was added, mixed and was immediately followed by the addition of 50 ~1 KI solution (6.0 g KI/5.0 ml of water). The reactants were mixed quickly, capped and kept in the dark for exactly 5 min. To the mixture, 3.0 ml of cadmium acetate (0.5 g/100 ml of water) was added, mixed vigorously and centrifuged at 1000 X g
65
for 10 min. The absorbance of the upper layer was measured at 353 nm in a Beckman DU-5 spectrophotometer against a suitable blank. The amount of lipid hydropero~des was calculated using a molar extinction coefficient of 1.73. f04 M-i and expressed as nmol/pmol lipid phosphorus.
dialdehyde bis-methyl acetal and expressed as nmol/ ,umol lipid phosphorus. The amount of lipid hydroperoxides was calculated using the molar extinction coefficient of 1.73 . lo4 M-’ and expressed as nmol/~mol lipid phosphorus. Statistical analyses
Lipid extraction and fatty acid analysis
Total fatty acid composition of tissue lipids was determined as follows. An aliquot of the lipid extract was transesterified in methanolic NaOH (0.2 M) with appropriate amounts of methyl heptadecanoate as an internal standard. The methyl esters were analyzed by GC on a column packed with 10% SP-2330 on 100/200 Chromosorb WAW (Supelco) using a Packard model 428 gas c~omato~aph connected to a Spectra-Physics 4270 integrator. Lipid phosphorus was determined according to Bartlett [15], and protein was determined according to Lowry et al. [16]. ~ete~i~ation
of reduced gIutathione
Reduced glutathione (GSH) contents of the tissues were determined according to Beutler et al. [17]. To 1.0 ml of the tissue homogenate, 1.5 ml of precipitating solution (1.67 g metaphosphoric acid + 0.1 g sodium EDTA + 30 g sodium chloride per 100 ml of distilled water) was added, mixed and allowed to stand for 5 min at room temperature. The mixture was centrifuged at 2000 x g for 10 min. An aliquot of 1.0 ml of the supematant was mixed with 4.0 ml phosphate solution (0.3 M Na,HPO,) followed by the addition of 0.5 ml DTNB solution (0.04 g/100 ml of 1% sodium citrate). The reactants were quickly mixed and the absorbance at 412 nm was measured in a Beckman DU-5 spectrophotometer against an appropriate blank. The amount of GSH was calculated using a standard curve prepared with GSH and expressed as pg/g wet weight of the tissue. Induction of lipid peroxidation in vitro
The extent of lipid peroxidation in the tissue homogenates in vitro, in the absence or presence of the free radic~-generating system Fe’+-ascorbate (0.1 mM FeSO, and 1.0 mM ascorbate), was studied as follows. Hearts and kidneys were obtained from a second group of diabetic rats, 6 weeks after alloxan injection, and from age- and sex-matched controls. In a final volume of 1.0 ml, 0.3 ml of the 10% tissue homogenate was incubated with 0.1 mM Fe*’ as FeSO, and 1.0 mM ascorbate in 25 mM Hepes-150 mM NaCl buffer (pH 7.4) at 37 o C on a shaking water bath. After intervals of 0, 30,60 or 120 min, samples were removed and assayed for the TBARS and/or lipid hydroperoxides [13], as described above. The amounts of TBARS were calculated using a standard curve prepared with malon-
The standard deviations of the means were calculated and the data were subjected to Student’s t-test for statistical significance. Results General characteristics of the diabetic rats
More than 75% of the animals that received alloxan developed severe chronic diabetes. The remaining animals either died within a few days of alloxan injection or were euthanized. The body, heart and kidney weights and blood glucose values of the diabetic animals and of their age-matched controls are presented in Table I. As expected, diabetic animals were markedly hyperglycemic and gained or lost body weight at a fraction of a gram per day, while nondiabetic animals were normoglycemic and gained more than 5 g per day during the study. The weights of both hearts and kidneys were less in the diabetic animals than in the controls, but when expressed relative to body weight, both heart and kidney weights in the diabetic animals were markedly elevated, indicating that the effect of insulin deficiency upon these organs was less than upon overall growth. The loss of both subcutaneous and retroperitoneal fat was striking in all of the diabetic animals. All the diabetic rats developed cataracts in one or both eyes, not seen in any of the controls. Urine glucose in the diabetic rats, without exception, exceeded the maximum value on the dipstick, corresponding to 1 g/dl, and urine ketones typically fell between the readings corresponding to 18 and 23 mM acetoacetate.
TABLE
I
General features
of the experimentul
model
All values are X f S.D. Each group
Body weight (g) at sacrifice Weight gain per day (g) Blood glucose (mg/dl) Heart weight(g) Heart weight(g) per body weight (100 g) Kidney weight a (g) Kidney weight (g) per body weight (100 g)
consisted
of eight animals.
Normal
Diabetic
400.6 f39.40 5.07f 0.74 89.1 f13.3 1.341 0.14
162.8 0.01 413.0 0.68
k 20.8 b f 0.57 b f 31.2 b * 0.05 b
0.34* 1.79*
0.03 0.11
0.42* 1.565
0.45 f
0.03
0.98 f 0.16 b
a The right kidney was used for all determinations. b Significantly different (P < 0.01) from the control tailed Student’s t-test.
value
0.06P 0.16 b
by two-
66 TABLE
II
In vivo TBARS, lipid hydroperoxides, iota1 glutathione and All values are X f SD. Each value is an average
lipid phosphorus contents in the hearts and kidneys of normal and diabetic rats (n = I/group)
of four individual
determinations.
TBARS,
thiobarbituric
Heart normal TBARS (nmol/pmol lipid P) Decrease in TBARS from normal Lipid hydroperoxides (nmol/~mol lipid P) Decrease in lipid hydroperoxides from normal Glutathine (GSH) (pg/g wet wt. tissue) Increase in GHS from normal Lipid phosphorus (pmol/g wet wt. tissue) a Significantly b n=3.
different
3.78 rf:0.77 29.30 f 2.02 238.26 k 9.56 28.20 + 2.50
(P < 0.05) from the normal
value by Student’s
Neither glucose nor ketones were detected in the control animals. Trace amounts of urine protein were occasionally noted in all groups.
acid-reactive
substances.
Kidney diabetic
normal
2.08+ 0.51a 45% 24.16 k 1.03 a.h 18% 305.50 k 78.65 28% 24.30 + 0.83
diabetic
4.10*
0.20
6.37+
0.92
3.74f 0.57 9% 5.97+ 1.74 6% 287.72 f 72.73 13% 32.03* 1.73”
238.43 + 63.31 28.23 +
1.05
t-test.
Products of lipid peroxidation in vivo In vivo contents of TBARS, lipid hydroperoxides, GSH and lipid phosphorus were determined in the
80 -
A
NORMAL (t Fe 2+/ASC)
50 NORMAL (t Fe *+/AX)
OIABI ETIC I\rre I r_ 2+/ASC)
DIABETIC (tFe 2+/ASC)
NORMAL
DIABETIC
I
0’
O-
0
1 HOURS
2
I
0
1
2
HOURS
Fig. 1. Formation of TBARS in vitro in the homogenates of diabetic (0) and normal (0) rat heart (Panel A) and kidney (Panel B) in the presence medium contained 3.6-5.0 mg ) and absence (---) of 0.1 mM Fe2+ (FeSO,) and 1.0 mM ascorbate. The final 1.0 ml of incubation (protein and was incubated at 37 o C and pH 7.4. TBARS were determined at the end of each incubation period as described in Materials and Methods. Each point is an average f S.D. of three to four individual experiments.
67 hearts and kidneys after 6 weeks of diabetes and in the hearts and kidneys of age-matched control rats. The values are given in Table II. The amounts of TBARS in the diabetic hearts were 45% lower than in controls (P < 0.05). Even though the kidneys of the diabetic rats contained decreased amounts of TBARS (9’%), this difference was not quite statistically significant. Similarly, the lipid hydroperoxide content in the hearts of &week diabetic animals was significantly (18%) lower than that in the normal hearts (P < O.OS), whereas the diabetic kidneys showed only slightly lower amounts of lipid hydroperoxides (6%). The GSH contents of the diabetic hearts and kidneys were elevated (28% and 13% respectively), but these increases were not statistically significant. Diabetic hearts contained slightly lower amounts of lipid phosphorus on a wet weight basis, whereas diabetic kidneys had slightly higher amounts compared to controls. Lipid pero~~datio~ in vitro
In vitro lipid peroxidation assayed as the formation of TBARS and/or lipid hydroperoxides in the homogenates of hearts and kidneys of diabetic and control rats, was determined in the absence or presence of Fe’+-ascorbate. The incubations were carried out for up to 2 h and the data presented in Fig. 1 were obtained. After each time of incubation, the formation of TBARS in the homogenates of diabetic hearts, either in the absence or presence of Fe2+-ascorbate, was significantly lower than in those of control animals (Fig. 1A). In the presence of Fe2+-ascorbate, the heart homogenates showed an increase in the formation of TBARS with increased incubation times. The extent of formation of TBARS in the homogenates of diabetic kidney was significantly lower at 30 and 60 min of incubation
in the presence of Fe2+- ascorbate and at 2 h reached a level lower than, but not significantly different from, that of the corresponding normal kidney homogenates. In the absence of Fe2+-ascorbate there were measurable, but not significant, decreases in the formation of TBARS in the diabetic animals (Fig. 1B). The amount of lipid hydroperoxides formed in the heart homogenates of 6-week diabetic animals in either the absence or presence of Fe2+-ascorbate was also lower than in controls. After incubating homogenates of diabetic hearts for 1 h in the absence of Fe*+/ascorbate, lipid hydroperoxide levels were 23.10 + 3.56 nmol/pmol lipid P (n = 4) and in the presence of Fe’+/ascorbate (0.1 mM/l.O mM) 28.00 + 0.90 nmol/pmol lipid P (n = 3). Homogenates of control hearts exhibited hydroperoxide levels of 28.70 of:5.64 nmol,/,umol lipid P (n = 4) in the absence and 42.35 1_5.60 nmol/pmol lipid P (n = 4) in the presence of Fe2+/ascorbate. The difference in hydroperoxide levels obtained with homogenates from normal and diabetic hearts in the presence of Fe’+/ascorbate was sig~ficant at P -C0.025. Fatty acid composition of total lipids
The lipid extracts of heart and kidney of diabetic rats and of control animals were analyzed for fatty acid composition and the data are summarized in Table III. The levels of palmitic acid (16 : 0) in the diabetic heart were lower than in the controls, whereas in the diabetic kidney they were slightly elevated. More importantly, both tissues of the diabetic animals exhibited significantly elevated levels of linoleate (18 : 2) and reduced levels of arac~donate (20: 4) resulting in increased linoleate/arachidonate ratios. Little or no change was observed in other long-chain polyunsaturated fatty acids (20 : 5, 22 : 5, 22 : 6) as the result of diabetes.
TABLE III Fatly acid composition (mol%) of the total lipids in the heart and kidney of normal and diabetic rats Data represent means* SD. The right kidney was used for all determinations. a P -C0.001, b P < 0.005, ’ P i 0.01. FA (chain length : no. of double bonds)
Heart (n=4) normal
diabetic
normal
diabetic
16:0 16:l 18:O 18:l 18:2(n -6) 20:3(n-6) 20:4(n -6) 20:5(n-3) 22:5(n -6) 22:6(n-3) 18:2/20:4
20.4f 2.3 1.9*0.7 18.5 * 2.2 20.1 f 2.6 23.8 kO.8 0.2*0.05 9.2 f 2.0 0.2 f 0.1 1.0*0.1 4.5*1.1 2.59
13.7*1.1 0.4*0.1 2o.oi 1.1 19.1 k 1.6 37.5 + 2.0 a 0.4*0.1 3.8*o.3c 0.4+0.2 1.1 f0.3 3.4kO.65 9.87
23.0k0.8 1.2f0.5 22.5 * 1 .o 13.0*0.X 17.0*0.3 0.8 kO.04 19.7f0.8 0.7 f 0.1 0.3 * 0.05 1.s*o.2 0.86
26.6 f 3.0 1.5 f 0.9 16.7kO.8 18.5 i 1.0 23.0 f 0.8 a 0.8 50.1 11.4+2.0b 1.0*0.1 0.3*0.1 2.4k0.3 2.01
Kidney (n=4)
68 Discussion Based upon the results of the present study, we conclude that heart and kidney from diabetic rats exhibit increased resistance to free radical-induced lipid peroxidation. Endogenous lipid peroxide levels were measured as TBARS which includes end products of lipid peroxidation such as malondialdehyde and other carbonyls, and as lipid hydroperoxides. We observed that in vivo levels of TBARS and of lipid hydroperoxides in the heart were significantly lower in diabetic rats. In addition, lipid peroxidation induced in heart homogenates of diabetic rats in the absence or presence of Fe’+ -ascorbate indicated significantly lower rates of TBARS formation. Diabetic kidney, either in vivo or in vitro, exhibited somewhat lower levels of TBARS and decreased capacity to form lipid peroxidation products, but the decrease in the lipid peroxidation potential was not as dramatic as that exhibited by the diabetic heart. The decreased levels of TBARS and lipid peroxides in the hearts and kidneys of diabetic animals may be attributed to several factors related to increased resistance to oxidative damage. Increased levels of non-enzymatic antioxidants such as glutathione have been noted previously in the livers of streptozotocin-diabetic rats [4]. In human diabetic individuals, elevated activity of glutathione reductase was observed [18]. This would result in enhanced reduction of oxidized glutathione, which could play a major role in glutathione-dependent antioxidant processes [19]. In addition, increased activities of the protective enzymes superoxide dismutase, catalase, and glutathione peroxidase in the tissues of human diabetics as well as experimental models were reported [4]. Diabetic rats, 12 weeks after administration of streptozotocin, exhibited elevated activities of catalase and glutathione reductase in the heart and of glutathione reductase and Cu-Zn superoxide dismutase in pancreas [5]. The observed significant decreases in the levels of TBARS and lipid peroxides in the diabetic hearts in the current study might, therefore, be due to increased activities of myocardial catalase and glutathione reductase. Diabetic rat kidney, 12 weeks after streptozotocin administration showed only increased activities of glutathione peroxidase, whereas the activities of catalase and superoxide dismutase were decreased [5]. This may explain why diabetic kidney was less resistant than the diabetic heart to lipid peroxidation compared to normal controls. The fatty acid composition of tissue lipids also appears to be a contributing factor in their susceptibility to peroxidation [20,21]. Rat liver endoplasmic reticulum containing higher levels of eicosapentaenoic acid (20 : 5 (n - 3)) and docosahexaenoic acid (22 : 6 (n - 3)) as the result of dietary fish oil showed higher levels of TBARS compared to controls having lower levels of eicosa-
pentaenoic and docosahexaenoic acids among liver lipids [20]. Novikoff heptoma cells were found to resist in vitro oxidative stress, and this was partly attributed to the lower levels of arachidonic acid (20 : 4 (n - 6)) and docosahexaenoic acid (22 : 6 (n - 3)) in these cells compared to normal liver cells [21]. A general trend in the fatty acid composition in diabetic tissues is that linoleate (18 : 2 (n - 6)) increases and arachidonate (20 : 4 (n - 6)) decreases [22-261. The results of the current work agree with those of ethers regarding elevated linoleic acid and reduced arachidonic acid levels in the diabetic heart and kidney. Decreased unsaturation of tissue lipids is thus likely to have contributed to their resistance to lipid peroxidation. Recently, Tani and Neely [27] observed that diabetic rat hearts were more resistant to in vitro ischemia and showed a greater recovery upon reperfusion after ischemic stress than normal rat hearts. Because postischemic reperfusion injury in the myocardium is likely to be mediated by oxygen free radicals [28], this observation [27] supports our present results which indicate that diabetic rat heart and kidney develop a resistance to oxidative stress. Acknowledgements Ann Schmeichel and Barbara Weis provided excellent technical assistance. This work was supported in part by PHS Program Project Grants HL08214 and NS14304, Training Grant HL07311, the Minnesota Affiliate of the American Heart Association and the Hormel Foundation. References 1 Halliwell, B. and Gutteridge, J.M.C. (1984) B&hem. J. 219, 1-14. 2 Sato, Y., Hotta, N., Sakamoto, N., Matsuoka, S., Oshishi, N. and Yagi, K. (1979) B&hem. Med. 21, 104-107. 3 Karpen, C.W., Pritchard, Jr., K.A., Arnold, J.H., Cornell, D.G. and Panaganamala, R.V. (1982) Diabetes 31, 947-951. 4 Oberley, L.W. (1988) Free Radical Biol. Med. 5, 113-124. 5 Wohaieb, S.A. and Godin, D.V. (1987) Diabetes 36, 1014-1018. 6 Beach, K.W. and Strandness, D.E. (1980) Diabetes 29, 882-888. 7 Winocour, P.H., Durrington, P.N., Ishola, M. and Anderson, D.C. (1986) Lancet 1, 117661179. 8 Regan, T.J., Lyons, M.M., Ahmed, S.S., Levinson, G.E., Oldewurtel, H.A., Ahmad, M.R. and Haider, B. (1977) J. Clin. Invest. 60, 885-889. 9 Hamby, R.I., Zoneraich, S. and Sherman, L. (1974) J. Am. Med. Assoc. 229, 1749-1754. 10 Thompson, E.W. (1988) Am. J. Anat. 182, 270-282. 11 Cohen, A.J., McGill, P.D., Rossetti, R.G., Guberski, D.L. and Like, A.A. (1987) Diabetes 36, 944951. 12 Farquhar, M.G. (1975) Kidney Int. 8, 197-211. 13 Buege, J.A. and Aust, S.D. (1978) Methods Enzymol. LII, 302-310. 14 Folch, J., Lees, M. and Sloane Stanley, G.H. (1957) J. Biol. Chem. 226, 497-509. 15 Bartlett, G.R. (1959) J. Biol. Chem. 234, 466-468. 16 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275.
69 17 Beutler, E., Duron, 0. and Kelly, B.M. (1963) J. Lab. Clin. Med. 61, 882-888. 18 Long, W.K. and Carson, P.E. (1961) B&hem. Biophys. Res. Commun. 5, 394-399. 19 Sies, H. (1987) Am. Rev. Respir. Dis. 136, 478-480. 20 Hammer, C.T. and Wills, E.D. (1978) B&hem. J. 174, 584-593. 21 Cheesman, K.H., Collins, M., Proudfoot, K., Slater, T.F., Burton, G.W., Webb, A.C. and Ingold, K.U. (1986) Biochem. J. 235, 507-514. 22 Holman, R.T., Johnson, S.B., Gerrard, J.M., Mauer, S.M., Kupcho-Sandberg, S. and Brown, D.M. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 2375-2379.
23 Clark, D.L., Hamel, F.G. and Queener, S.F. (1983) Lipids 18, 696-705. 24 Huang, Y.S., Horrobin, D.F., Mar&u, M.S., Mitchell, J. and Ryan, M.A. (1984) Lipids 19, 367-370. 25 Lin, C.-J., Peterson, R. and Eichberg, J. (1985) Neurochem. Res. 10, 1453-1465. 26 Chattopadhyay, J., Thompson, E.W. and S&mid, H.H.O. (1990) Lipids 25, 307-310. 27 Tani, M. and Neely, J.R. (1988) Circ. Res. 62, 931-940. 28 McCord, J.M. (1985) N. Engl. J. Med. 312, 159-163.
