Elevated

Lipid Peroxidation

Levels in Red Blood Cells of Streptozotocin-Treated Diabetic Rats

Sushi1 K. Jain, Steven

N. Levine, John Duett, and Becky Hollier

This study was performed to determine whether or not hyperglycemia in diabetes results in elevated levels of lipid peroxidation products in red blood cells (RBC). Diabetes was induced in rats by treatment with streptozotocin. The level of lipid peroxidation products was examined in fresh RBC by measuring their thiobarbituric acid (TBA) reactivity after 2 and 4 months of induction of diabetes. Hyperglycemia was assessed by measuring the level of glycosylated hemoglobin and blood glucose. Results show that lipid peroxidation levels were significantly higher (50% to 84%) in RBC of diabetic rats than in controls. The increase in the level of lipid peroxidation was blocked in diabetic rats in which hyperglycemia was controlled by insulin treatment. Among phospholipid classes, relative percentage of sphingomyelin (SM) was significantly reduced in RBC at both 2 and 4 months of diabetes; whereas phosphatidylethanolamine (PE) levels were higher in RBC at 4 months of diabetes only. The level of phosphatidylcholine (PC) did not differ significantly between RBC of control and diabetic rats. This study suggests a significantly altered lipid composition and an accumulation of lipid peroxidation products in RBC of streptozotocin-treated diabetic rats. 0 7990 by W.B. Saunders Comp8ny.

A

NUMBER OF STUDIES have suggested that altered properties of red blood cells (RBC), such as increased viscosity, reduced life span, and increased adhesivity, have a role in the rheological impairment and ischemia of certain tissues in diabetic patients.‘.” The biochemical mechanism by which hyperglycemia could result in altered RBC is not known. Recent studies in a cell-free system have suggested that glucose can enolize and reduce molecular oxygen, which can yield free radical intermediates such as oxygen radicals.““’ We have observed in our own in vitro investigations that treatment of RBC with elevated levels of glucose can result in membrane lipid peroxidation’4.‘5 and have suggested that hyperglycemia may cause peroxidative injury to membranes. The present study of diabetic rats was undertaken to examine whether or not hyperglycemia in vivo has any effect on RBC lipid peroxidation and total phospholipid and cholesterol levels. We also investigated the effect of insulin treatment on hyperglycemia and of duration of diabetes on lipid abnormalities in RBC. MATERIALS AND METHODS Female Sprague-Dawley rats weighing I80 to 200 g were divided into three groups: C, control; D, diabetic; D+I, insulin-treated diabetics. Rats in groups D and D+ I were made diabetic by a single intraperitoneal (IP) injection of 55 mg/kg streptozotocin dissolved in citrate buffer (pH 4.5). Control rats in group C were injected IP with buffer alone. Two days after administration of streptozotocin (or control buffer), tail vein blood glucose was measured in all animals. A streptozotocin-treated rat having a blood glucose less than 250 mg/dL was excluded from further analysis. Rats in group D+ I were injected subcutaneously daily with 1.4 U protamine zinc insulin/ 100 g body weight. All animals were provided with food and water ad libitum. Sixteen hours before sacrifice, food was withdrawn but animals were allowed free access to water. At the time of death, rats were weighed and anesthetized with 40 mg/kg sodium pentobarbital IP. Blood from the heart was collected after entering the abdominal and thoracic cavities into EDTA-tubes using IO-mL syringes and centrifuged at 2,000 rpm for 10 minutes in a refrigerated RC3B Sorvall centrifuge (Du Pont Instruments, Wilmington, DE). Plasma and huffy coat were discarded. Packed RBC were washed with normal saline two times after one to 10 dilutions to remove left-over leukocytes and plasma components. Preparation of washed RBC and biochemical analyses were performed immediately after blood collection. Metabolism,

Vol39,

No 9 (September), 1990: pp 97 l-975

Measurement

of Lipid Peroxidation

Membrane lipid peroxidation was determined by thiobarbituric acid (TBA) reactivity. Malonyldialdehyde (MDA), an end-product of fatty acid peroxidation, can react with TBA to form a colored complex that has a maximum absorbance at 532 nm.16 For this purpose, 0.2 mL of packed cells were suspended in 0.8 mL phosphatebuffered saline (made up of 8.1 g NaCl + 2.302 g Na,HPO, + 0.194 g NaH,PO, per liter, pH 7.4) and 0.025 mL of butylated hydroxytoluene (BHT, 88 mg/lO mL absolute alcohol). To this, 0.5 mL of 30% trichloroacetic acid was added. Tubes were vortexed and allowed to stand in ice for at least 2 hours. Tubes were centrifuged at 2,000 rpm for 15 minutes. One milliliter each of the supernatant was transferred into another tube. To this was added 0.075 mL of 0.1 mol/L EDTA and 0.25 mL of TBA (1% in 0.05N NaOH). Tubes were mixed and kept in a boiling water bath for 15 minutes. After tubes were cooled to room temperature, absorbance was read at 532 and 600 nm in a double-beam Lambda 3B Perkin Elmer (Norwalk, CT) Spectrophotometer. BHT, an antioxidant, was added to rule out any MDA formation during the assay procedure, which could result in falsely elevated TBA reactivity. Addition of BHT to standard MDA did not affect its color development with the TBA. Absorbance at 600 nm was substracted from absorbance at 532 nm. MDA values in nanomoles per milliliter packed cells were determined using the extinction coefficient of MDA-TBA complex at 532 nm = 1.56 x IO’ per cm per molar solution. The packed cell volume of RBC was determined by using an Autocrit Centrifuge (Becton Dickinson, Parsipanny, NJ). Measurement of MDA by the TBA reactivity is the most widely used method to assess lipid peroxidation.” RBC lipid extraction, drying, and washing of the lipid extract were performed as described by Rose and Oklander.18 Separation of various phospholipid classes in the lipid extract was accomplished by thin-layer chromatography (TLC) on silica gel H glass plates (silica 60, 0.25-mm thickness, Brinkman Instruments, Westbury, NY) using solvent system chloroform-methanol-glacial acetic acid-

From the Departments of Pediatrics and Medicine, Louisiana State University School of Medicine, Shreveport, LA. Supported by grants from the National American Diabetes Association, Inc and the National Institutes of Health (No. I ROl HL30247). Address reprint requests to Sushi1 K. Jain, PhD. Department of Pediatrics, LSU Medical Center, 1501 Kings Highway, Shreveport, LA 71130. o 1990 by W.B. Saunders Company. 0026-0495/90/3909-0016$03.00/0 971

972

JAIN ET AL

Table 1. Blood Glucose and GHb Levels in Control, Diabetic, and Insulin-Treated Diabetic Rats 2 months

4 months Fasting

Fasting Group

GlUXW3

GHb

GlUC0S.e

GHb

(mgfd)

(%)

(mg/dLl

(%)

113 i- 9

Control

5.07

(14) 376 k 28

Diabetic

10.73

(8) Diabetic + insulin

k 0.47 + 1.34 (8)

126 + 34

4.49

(11)

113 + 16

4.62

(14)

114)

402 k 79

10.91

(7)

r 0.58

93 * 48 17)

(11)

+ 0.44 (14) + 0.49 (7)

4.68

t 0.76 (7)

NOTE. Values are mean k SD. Number of rats in each group is given in parantheses. Rats were made diabetic by injecting streptozotocin. Details of treatment are given in Materials and Methods.

levels close to those of control rats. The extent of hyperglycemia was similar in the f-month and 4-month diabetic groups. Figure I illustrates TBA-reactivity of fresh, untreated RBC from control. diabetic, and insulin-treated diabetic groups. There was a significant (50% to 84%) increase in the TBA reactivity of RBC obtained from both 2- and 4-month diabetic rats compared with respective controls. The extent of lipid peroxidation in RBC of 2- and 4-month diabetic rats was similar. Insulin treatment prevented the RBC lipid peroxidation level in both groups below that of the diabetic group maintained without insulin treatment for 2- and 4-months; however, the lower TBA reactivity on insulin treatment was significant only in the 4-month group. Table 2 shows total phospholipid and cholesterol levels in RBC of control, diabetic, and insulin-treated diabetic rats after 2 and 4 months of treatment. There was a significant increase in phospholipid in RBC of rats with 4-month diabetes. There was a significant decrease in the cholesterol to phospholipid molar ratio in diabetic compared with control rats in both 2- and 4-month groups. Insulin treatment of diabetic rats attenuated the decreased cholesterol to phospholipid molar ratio. Figure 2 illustrates the percentage of major phospholipid classes in RBC of control, diabetic, and insulin-treated diabetic rats maintained for 2 and 4 months. There was a significant decrease in percentage of sphingomyelin (SM) in RBC of the diabetic groups, which remained at control levels in the insulin-treated groups. There was no significant difference in phosphatidylethanolamine (PE) levels in RBC of rats diabetic for 2 months, but a significant increase in PE was found in rats diabetic for 4 months. This increase in PE was blocked by insulin treatment. There was no significant relationship between increase in the relative amount of PE

50-25-8-4 (vol/vol). Different phospholipid spots on the TLC plate were visualized by exposing the plate to iodine vapors and were encircled with a fine needle. Localization of various phospholipids on the TLC plate was confirmed by using authentic standards and by using specific sprays on the plate.” The amount of phospholipid was determined by quantitating phospholipid-phosphorus from the silica gel after scraping it into Pyrex glass tubes.20 For this, the silica gel was directly digested with 10N H,SO, to liberate free inorganic phosphorus, and color was developed while the gel was still in the test tube; silica gel per se does not interfere in the color development. Absorbance in the supernatant was read after centrifuging down the silica gel or letting tubes sit overnight after the color development. Cholesterol was measured by the method of Zlatkis et al.” Blood glucose was measured using a glucose oxidase method (Boehringer Mannheim Diagnostics, Indianapolis, IN). water/

Measurement of Glycosylated Hemoglobin The human erythrocyte is freely permeable to glucose, and within each erythrocyte, glycosylated hemoglobin (GHb) is formed continuously from hemoglobin A at a rate dependent on the ambient glucose concentration. The formation of GHb is nonenzymatic, slow, and largely irreversible. It is now widely accepted that measurement of GHb in a single blood sample provides an index of the mean blood glucose level.“,” GHb was measured using Glyc-affinity columns and kit of Iso-Lab, Akron. OH. Data were analyzed statistically using nonpaired Student’s t test and regressional analyses with EPISTAT statistical software for IBM PC/XT. RESULTS

Table

1 gives

the levels of fasting

blood

glucose

and GHb

in control, diabetic, and insulin-treated diabetic rats. Both blood glucose and GHb were higher in the diabetic group. However, diabetic rats treated with insulin showed a significant correction of hyperglycemia and blood glucose and GHb 3.0

4

2.0l.O-

p-o.004

T

l-

’

(6)

0.0 SD

piO.006

P-O.0006 -I-

-I-_

T

(15)

(9)

C

(10) D+I

2 MODNTHS

l-

p-o.02 l-

D+I

C

4 M&THS

Fig 1. TBA of RBC of control (Cl. diabetic (D), and insulintreated diabetic (D+I) rats at 2 and 4 months of treatment. Number of rats in each group is given in parentheses. Note significant increase in the TBA-reactivity in RBC of diabetic rats.

RBC LIPID PEROXIDATION IN DIABETES

973

Table 2. Total Cholesterol, Phospholipid, and Cholesterol to Phospholipid Molar Ratio in RBC of Control, Diabetic, and Insulin-Treated Diabetic Bats P Value

Group

C

D

!J+t

CvD

DvDfI

CvD+I

1.38 c 0.20

1.28 z+ 0.06

1.42 f 0.12

NS

.Ol

NS

(14)

(9)

(11) NS

NS

NS

2 Months CHOL (mg/mL) PL

2.00 (mg/mL)

CHOL/PL molar ratio

+ 0.32

2.17

* 0.17

2.03

f 0.15

(14)

(9)

(11)

1.38 2 0.09

1.22 + 0.07

1.40 f 0.06

(14)

(9)

(11)

1.32 i 0.19

1.28 * 0.15

1.29 f 0.16

(141

(6)

(7)

.0004

.00004

NS

4 Months CHOL (mg/mL) PL

1.70 f 0.31 (mg/mL)

CHOL/PL molar ratio

2.09

f 0.27

1.89 c 0.19

(14)

(6)

(7)

1.60 k 0.32

1.24 + 0.14

1.36 f 0.09

(141

(6)

(7)

NS

NS

NS

.02

NS

NS

.02

NS

NS

NOTE. Values are mean + SD. Number of rats in each group is given in parentheses. Details of treatment to rats and lipid analyses are given in Materials and Methods. Abbreviations: CHDL, cholesterol; PL, phospholipid.

ever, there is no study that examines occurrence of membrane lipid peroxidation in fresh, untreated RBC of diabetic rats with and without insulin treatment. The present study documents a significant occurrence of lipid peroxidation in untreated, fresh RBC of diabetic rats, which was blocked by insulin treatment. Thus, hyperglycemia appears to induce lipid peroxidation in diabetic rats. The extent of increase in lipid peroxidation did not appear to be different at 2 or 4 months after induction of diabetes. We also found decreased percentages of SM in RBC in rats with 2 and 4 months of diabetes and increased levels of PE in those with 4 months of diabetes. The decreased percentage of SM and increased PE were prevented with control of hyperglycemia by insulin administration. Previous studies in diabetic patients have reported norma132,33 and increased34s35 SM+lysophosphatidylcholine (LPC) in RBC.

and the decrease in the relative amount of SM in RBC of 4-month diabetic rats. Phosphatidylcholine (PC) levels of RBC were not different between control and 2- and 4-month diabetic rats. DISCUSSION

Lipid peroxidation processes have been implicated in a variety of disease states. Sato et a124were the first to report increased levels of lipid peroxide in plasma of diabetic patients. Subsequent studies have confirmed this observation ‘.. dirbetic patients and animals.25-30 Previous studies have also ,+jrted that the RBC of diabetic patients are more susceptible to lipid peroxidation when treated with hydrogen peroxide in vitro.25.30 A recent study reported a significant relationship between the amount of RBC membrane lipid peroxidation and hyperglycemia in diabetic patients3’ How-

Fig 2. PC. PE. and SM levels of RBC of control (CL diabetic (D). and insulin-treated diabetic (D+I) rats at 2 and 4 months of treatments. Number of rats in each group is given in paranthasas. Note significant decrease in SM. increase in PE (only at 4 month& and similar levels of PC in REC of diabetic rats.

56540 a

SD

52;

504a46

(12)

(aI

C

D+I

2 MO\TliS

D+I

c

4 MODNTHS

JAIN ET AL

974

In the present study, SM was quantitated after its separation from LPC. The increase in PE in RBC of the diabetic group appears to be an effect of duration of diabetes, but its cause is not known. Studies on RBC of diabetic patients have reported normal 32.33.35 and increased34 PE levels. The absence of any significant differences in total cholesterol in RBC of the diabetic rats is consistent with earlier reports35’39; however, other studies on RBC of diabetic patients have reported an increase in cholesterol level.40.4’ The decrease in cholesterol to phospholipid molar ratio in RBC of the diabetic rats agrees with earlier reports on streptozotocin-treated diabetic rats,37.39 but contrasts with an increase38 reported in RBC of rats with 5 days of alloxan-induced diabetes.

RBC membrane lipid peroxidation is known to cause decreased cell survival,” altered membrane lipid asymmetry,“.46 hypercoagu1ability,J’~4’ and increased adhesivity to the endothelium.45 The present finding of increased membrane lipid peroxidation in RBC of rats exposed to hyperglycemia and its attenuation by the control of hyperglycemia with insulin administration suggests that membrane lipid peroxidation may havea role in the hyperviscosity, altered membrane lipid asymmetry, hypercoagulability, increased adhesivity. and reduced life span of RBC known to occur in diabetic patients.‘.‘” ACKNOWLEDGMENT

The authors arc grateful to Barbara

MacRoberts

for editing.

REFERENCES 1. Jones RL. Peterson CM: Hematologic alteration in diabetes mellitus. Am J Med 70:33-352, 1981 2. Lehrman ML: Reversible hematologic sequelae of diabetes mellitus. Ann Intern Med 86:425-429, 1977 3. Pescarmona GP, Bosia A, Ghigo D: Shortened red cell life span in diabetes: Mechanism of hemolysis, in Weatherall DJ, Fiorelli G, Gorini S (eds): Advances in Red Cell Biol. New York, NY, Raven 1982.~~391-397 4. Schmid-Schonbein H, Volger E: Red cell aggregation and red cell deformability in diabetes. Diabetes 25:897-902, 1976 (suppl2) 5. Satoh M, Imazumi K, Bessho T, et al: Increased erythrocyte aggregation in diabetes mellitus and its relationship to glycosylated haemoglobin and retinopathy. Diabetologia 27:517-521, 1984 6. Wali RK, Jafe S, Kumar D, et al: Alterations in organization of phospholipids in erythrocytes as factor in adherence to endothelial cells in diabetes mellitus. Diabetes 37:104-l 11, 1988 7. Wautier JL, Paton RC, Wautier MP, et al: Increased adhesion of erythrocytes to endothelial cells in diabetes mellitus and its relation to vascular complications. N Engl J Med 305:237-242, 198 1 8. Simpson LO: Intrinsic stiffening of red blood cells as the fundamental cause of diabetic nephropathy and microangiopathy. Nephron 39:344-351, 1985 9. McMillan DE: The blood viscosity problem in diabetes. Clin Diabetes 7:61-71, 1989 10. McMillan DE, Utterback NG, La Puma J: Reduced erythrocyte deformability in diabetes. Diabetes 27:895-901, 1978 I 1. Wolff SP, Dean RT: Glucose autoxidation and protein modification. The potential role of autoxidative glycosylation in diabetes. Biochem J 245:243-250,1987 12. Mashino T, Fridovich I: Mechanism of the cyanide-catalyzed oxidation of alpha-ketoalchohols. Arch Biochem Biophys 252: 163170.1987 13. Hunt JV, Dean RT, Wolff SP: Hydroxyl radical production and autoxidative glycosylation. Glucose autoxidation as the cause of protein damage in the experimental glycation model of diabetes mellitus and aging. Biochem J 256:205-212, 1988 14. Jain SK, McVie R: Hyperglycemia induces the membrane lipid peroxidation in human red blood cells. Blood 68:36A, 1986 15. Jain SK: Hyperglycemia can cause membrane lipid peroxidation and osmotic fragility in human red blood cells. J Biol Chem 264:21340-21345, 1989 16. Stocks J, Dormandy TL: The autoxidation of human red cell lipids induced by hydrogen peroxide. Br J Haematol20:95-111, 1971 17. Slater TF: Overview of methods used for detecting lipid peroxidation. Methods Enzymol 195:283-293, 1984 18. Rose HG, Oklander M: Improved procedure for the extraction of lipids from human erythrocytes. J Lipid Res 6:528-531, 1965

19. Jain SK, Subrahmanyam D: Two-dimensional thin-layer chromatography of polar lipids. Ital J Biochemistry 27: 11- 19, 1978 20. Fiske CH, Subbarow Y: The calorimetric phosphorus. J Biol Chem 66:375-400, 1925 21. Zlatkis determination 1953

determination

of

A, Bennie Z, Boyle AJ: A new method for the direct of serum cholesterol. J Lab Clin Med 41:486-492.

22. Bunn HF: Evaluation of glycosylated patients. Diabetes 30:613-617, 198 1

hemoglobin

in diabetic

23. Nathan DM, Singer DE, Hurxthal K, et al: The clinical information value of the glycosylated hemoglobin assay. N Engl J Med 310:341-346. 1984 24. Sato Y, Hotta N, Sakamoto N, et al: Lipid peroxide level in plasma of diabetic patients. Biochem Med 21:104-107. 1979 25. Matkovics B. Varga SI, Szabo L, et al: The effect of diabetes on the activities of the peroxide metabolism enzymes. Horm Metab Res 17:77-79, 1982 26. Karpen CW, Pritchard KA, Arnold JH, et al: Restoration of prostacyclin/thromboxane Az balance in the diabetic rat. Influence of dietary vitamin E. Diabetes 3 1:947-95 1, 1982 27. Kaji H, Kurasake M, lto K. et al: Increased lipoperoxide value and glutathione peroxidase activity in blood plasma of type 2 (non-insulin dependent) diabetic women. Klin Wochenschr 63:765768.1985 28. Tsuchida M, Miura T. Mizutani K, et al: Fluorescent substances in mouse and human sera as a parameter of in vivo fipid peroxidation. Biochim Biophys Acta 834:196-204, I985 29. Dohi T, Kawamura K. Morita K, et al: Alteration of the plasma selenium concentrations and the activities of tissue peroxide metabolism enzymes in streptozotocin-induced diabetic rats. Horm Metab Res 20:671-675, 1988 30. Uzel N, Sivas A, Uysal M, et al: Erythrocyte lipid peroxidation and glutathione peroxidase activities in patients with diabetes melfitus. Horm Metab Res 19:89-90, 1987 31. Jain SK, McVie R, Duett lipid peroxidation and glycosyfated 38:1539-1543, 1989

J, et al: Erythrocyte membrane hemoglobin in diabetes. Diabetes

32. Caimi G, Catania A, D’Asaro S, et al: Red cell phospholipids and membrane microviscosity in diabetics. Clin Hemorheol 8:939944,1988 33. Freyburger G, Gin H, Heape A, et al: Phospholipid and fatty acid composition of erythrocytes in type I and type II diabetes, Metabolism 38:673-678, 1989 34. Kamada T, Otsuji S: Lower levels of erythrocyte fluidity in diabetic patients. Diabetes 32:585-591, 1983 35. Prisco D, Paniccia

R, Coppo

membrane

M, et al: Red blood cell lipid

RBC LIPID PEROXIDATION

IN DIABETES

alterations in type II diabetes mellitus. Thromb Res 54:751-758, 1989 36. Hill MA, Court JM: Erythrocyte membrane fluidity in type 1 diabetes mellitus. Pathology 15:449-451, 1983 37. Chandramouli V, Carter JR: Cell membrane changes in chronically diabetic rats. Diabetes 24:257-262, 1975 38. Kapoor R, Chauhan VPS, Sarkar AK: Effect of acute diabetes and insulin treatment on erythrocyte lipid composition in rats. IRCS Med Sci 6:413, 1978 39. Petit-Thevenin JL, Nobili 0, Boyer J: Decreased acylation of phosphatidylcholine in diabetic rat erythrocytes. Diabetes 37:142146,1988 40. Bryszewska M, Watala C, Torzecka W: Changes in fluidity and composition of erythrocyte membranes and in composition of plasma lipids in type I diabetes. Br J Haematol62:ll l-l 16, 1986 41. Cignarelli M, Blonda M, Cospite MR, et al: Alterations of erythrocyte lipid pattern and of some membrane related functions as a consequence of plasma lipid disorder in diabetes mellitus. Diabetes Metab 91272-276, 1983 42. Jain SK, Mohandas N, Clark M, et al: The effect of

975

malonyldialdehyde, a product of lipid peroxidtion, on the deformability, dehydration, and 5 I -Cr-survival of erythrocytes. Br J Haematol 53:247-255, 1983 43. Jain SK: The accumulation of malonyldialdehyde, an end product of fatty acid peroxidation, can disturb aminophospholipid organization in the membrane bilayer of human erythrocytes. J Biol Chem 259:3391-3394,1984 44. Jain SK: In vivo externalization of phosphatidylserine and phosphatidylethanolamine in the membrane bilayer and hypercoagulability by the lipid peroxidation of erythrocytes in rats. J Clin Invest 76:281-286, 1985 45. Wali RK, Jafe S, Kumar D, et al: Increased adherence of oxidant-treated human and bovine erythrocytes to cultured endothelial cells. J Cell Physiol I 13:25-36, 1987 46. Arduini A, Stern A, Storto S, et al: Effect of oxidative stress on membrane phospholipid an protein organization in human erythrocytes. Arch Biochem Biophys 273:112-120, 1989 47. Cazana FJD, Marques MLD, Puyol DR, et al: Active role of plasma in blood hypercoagulability induced by phenylhydrazine. Thromb Res 53:215-220, 1989