Vol. 151, No. 2, 1988

March 15, 1988

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages 649-655

CATALYSIS OF LIPID PEROXIDATION BY GLUCOSE AND GLYCOSYLATED

COLLAGEN

Mark Hicks, Leigh Delbridge, Dennis K. Yue and Tom S. Reeve

Department of Surgery, The University of Sydney, Royal North Shore

Hospital, St Leonards N.S.W. 2065, AUSTRALIA

Received January 25, 1988

Peroxidation of membranes of linoleic/arachadonic acid vesicles at 40°C was catalysed by glucose or glycosylated collagen. Kinetics of hydroperoxide production were similar in both cases. There was a kinetic lag, the duration of which was inversely proportional to the amount of glucose or glycosylated collagen used to initiate the peroxidation. This was followed by a rapid increase in the rate of oxidation. The final rate was independent of the amount of catalyst added. Vesicle lysis accompanied the increases in peroxidation. Addition of desferrioxamine, an iron chelator, only partially inhibited the oxidative process. © 1 9 8 8 A o ~ d . . . . P . . . . . Inc.

Free radical induced lipid peroxidation has been associated with a

number of disease processes (i, 2), including diabetes mellitus

(3). Increased amounts of malondialdehyde, a decomposition product

of lipid hydroperoxides, have been found in the serum and high

density lipoprotein fractions of patients with uncontrolled

diabetes (4, 5). Evidence of lipid peroxidation has also been

observed in a number of complications accompanying diabetes such as

retinopathy, cataract formation and atherosclerosis (6, 7, 8). To

date, no overall mechanism has been presented to account for the

induction of lipid peroxidation in the diabetic state.

Nonenzymatic glycosylation of long-lived proteins occurs at a slow

rate in nondiabetic subjects as a function of aging and at an

increased rate in diabetes mellitus as a result of hyperglycemia

(9). Recently, free radical intermediates have been postulated to

occur in carbohydrates and glycosylated proteins during their

autoxidation under physiological conditions (i0, ii, 12). Such

free radicals could potentially induce peroxidation of lipids under

appropriate circumstances.

6 4 9

0006-291X/88 $1.50 Copyright © 1988 by Academic Press, Inc.

All rights of reproduction in any form reserved.

Vol. 151, No. 2, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

The present study describes catalysis of peroxidation of membranes

of vesicles prepared from a linoleic/arachadonic acid mixture by

free glucose or glycosylated collagen.

MATERIALS AND METHODS

Linoleic and arachadonic acids were obtained from Sedary Research Laboratories (London, Ontario) and tris(hydroxymethyl)aminomethane from Sigma Chemical Co. (St Louis, Mo°). Methanol (GR grade) was purchased from Merck (Darmstadt) and desferrioxamine from Ciba-Geigy (Australia). All other chemicals were of A.R. grade and were bought from BDH (Australia). Water used for solution preparation and glassware washing was passed through a Milli-Q water purification unit (Millipore, Pty. Ltd., Australia).

Rat tail tendons were used as a source of collagen. Male Wistars (8-10 weeks old) were obtained from the Atomic Energy Commission (Lucas Heights). After sacrifice, tail tendons were removed, washed thoroughly in physiological saline then distilled water, dried and stored at -70°C until required.

Non-enzymatic glycosylation of tendon collagen was carried out as follows. Tendon fibres (i0 cm in length, weighing l-2mg) were incubated at 40°C for 48 hours in 200mM glucose in phosphate buffer, pH 7.4 which contained 0.2% azide. The extent of glycosylation was measured by thiobarbituric acid chromophore absorbance at 443nm (13) after solubilisation of the tendon collagen with collagenase (14). Hydroxymethyl furfural (HMF) was used as a standard. Collagen was determined by the method of Stegmann and Stalder (15). The relative extent of glycosylation was expressed as nmol HMF/mg collagen.

Fatty acid vesicles were prepared by the method of Gebicki and Hicks (16). Linoleic acid (40mg), arachadonic acid (0.4mg) and 4ml 0.1M Tris/HCl, pH8 were vortexed together in a test tube then diluted to 20ml with the same buffer (final concentration of fatty acid, 7.2mM). Vesicles were always freshly prepared. Peroxidation of lipids was measured as an increase in 234nm absorbance due to conjugated d~ne _~ormation, using an extinction coefficient of 28,000 l.mole .cm ~ (17).

RESULTS

Catalysis of thermal peroxidation of vesicle membrane fatty acids

by glucose from 2.0 to 19mM, (physiological concentrations up to

those seen in poorly controlled diabetes), is shown in Figure i. No

significant increase in hydroperoxide content was observed in the

absence of glucose over the duration of bhe experiment. When

glucose was added at t=0, there was a kinetic lag period before

rapid peroxidation commenced. The duration of this lag period was

inversely proportional to the amount of glucose added. The maximum

rate of peroxidation was independent of the amount of glucose used

to initiate the reaction.

650

Vol. 151, No. 2, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

E ul Q

8 n- LU Q_ o ~r ~3 >- -r

®

2"0,

1.C

0

2.0.

1.0 o

I

20 4() 0 2'5 5b 75

TIME (hr) ( ~ TIME (hr)

Figure i. Kinetics of glucose-induced hydroperoxide production in linoleic/arachadonic acid membranes at 40°C. Bulk fatty acid concentration, 7.2mM. (~) no, (O) 2mM, (O) 5mM, ([]) 13mM, (A) 19mM glucose added at t=0.

Figure 2. Kinetics of hydroperoxide production in linoleic/ arachadonic acid membranes induced by collagen glycosylated to varying extents. Temperature, 40°C. Bulk fatty acid concentration, 7.2mM. (A) no collagen, (O) 5.3mg control collagen (3.8mg HMF/mg collagen), (~ 2.3mg and (•) 5.3mg nonenzymically glycosylated collagen (7.4 nmol HMF/mg collagen) added at t=0.

Similar peroxidation kinetics are seen with vesicles incubated with

rat tail tendon collagen glycosylated to varying extents (Figure

2). The duration of the lag period is inversely related to the

extent of glycosylation rather than the absolute amount of tendon

present. The lag period for the vesicle preparation incubated with

5.3mg of nonenzymically glycosylated tendon (containing 7.4nmol

HMF/mg collagen), was 36 hours, whilst that of the vesicles

peroxidised with 5.3mg of control tendon (containing 3.8nmol HMF/mg

collagen), was 50 hours (figure 2). However, the final rate of

peroxidation was independent of the extent of glycosylation.

In order to test the potential role of iron in the catalysis of the

observed peroxidation, imM desferrioxamine was added to vesicle

preparations at t=0, before the addition of glucose or glycosylated

tendon (Figure 3). The inclusion of the chelating agent slowed the

formation of hydroperoxides in both cases, but did not completely

inhibit the process.

When the time course of the experiment was increased, hydroperoxide

decomposition was observed in all treatments after about 1.8-2.0mM

hydroperoxide was accumulated. An example of the kinetics of this

change are shown for vesicles treated with 5mM glucose (Figure 4).

651

VoI. 151, No. 2, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

£_ W Q

LU a_ O rr E3 >- T

®

1-5. 2.0

1.0-

i 1.0 0.5-

I 0

0 20 40 OF - , , , O 20 40

TIME {hr) ~ TIME (hr)

Figure 3. Effect of desferrioxamine on kinetics of hydroperoxide formation in linoleic/arachadonic membranes induced by glucose or nonenzymaticall~ glycosylated collagen (i0 nmol HMF/mg collagen). Temperature, 40 C. Bulk fatty acid concentration, 7.2mM. (0) 19mM glucose, (O) 19mM glucose + 1.0mM desferrioxamine; (.) 5.6mg collagen, ([~ 5.6mg collagen + imM desferrioxamine added at t=0.

Figure 4. Changes in hydroperoxide content and turbidity of linoleic/arachadonic acid vesicles after addition of 5mM glucose at t=0. Temperature, 40°C. Bulk fatty acid concentration, 7.2mM. Hydroperoxide content after no additions (A), and glucose (0). Absorbance at 450nm after no additions (~-0, and glucose (O).

.1-O

0.5 o

<

There was a decrease in the turbidity of the vesicle preparation,

measured as 450nm absorbance, indicative of vesicle lysis were

associated with these peroxidative changes (Figure 4). This lysis

was also observed in vesicles treated with glycosylated collagen.

No such decrease was observed in the absence of peroxidation.

DISCUSSION

There is a large body of evidence to suggest that oxygen free

radicals and the products of lipid peroxidation are associated with

natural aging (18). Many diabetic complications are similar to

senescent changes and may well have common etiologies. It has been

proposed that the extent and duration of hyperglycemia in diabetes

are two important contributors to the severity of diabetic

complications (19), although the underlying biochemical mechanisms

by which these complications develop remain unresolved.

An increase in non-enzymatic glycosylation in long-lived proteins,

with the attendant formation of advanced glycosylation endproducts

resulting in peptide crosslinking and fluorescent changes has been

proposed as the principal mechanism for many diabetic complications

652

VoI. 151, No. 2, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

(20). However, free radical induced oxidative damage has also been

proposed to account for similar fluorescent changes in short

half-life proteins in diabetic microangiopathy (21).

The data presented here demonstrate a potential source of free

radical damage during hyperglycemia: the catalysis of peroxidation

of polyunsaturated lipids by glucose and glycosylated proteins

(Figures 1 and 2).

Free radical mechanisms involving co-oxidation of carbohydrate or

glycoprotein and lipid hydroperoxide, analagous to the mechanisms

for carbohydrate autoxidation (i0), or glycoprotein oxidation (12),

could account for the observed peroxidative damage. Lipid

hydroperoxides accumulate at a slow rate,in the absence of glucose

or glycosylated collagen (Figures 1 and 2). In the presence of

glucose, these hydroperoxides are able to react with the enol form

of glucose, when it is in the open chain configuration:

OH OH OH O" I I I I

H - C = C - R + LOOH ~ H - C = C - R +

Glucose (G) Hydro- (G') peroxide

This process is dependent on the

(I0, 22).

LO" + H20

rate of enolisation of glucose

If the catalyst is glycosylated collagen, the lipid hydroperoxide

could react with the eneaminol form of the glycosylated residue:

Peptide - NH Peptide - NH I i CH CH

C-OH + LOOH > C-O" + LO" + H20 I I

(CHOH) 3 (CHOH) 3 I I CH20H CH20H

Glycosylated Hydro- (GP') Peptide (GP) peroxide

Where G" and GP" are glucose and glycosylated peptide radicals and

LO" is an alkoxy radical. These species can initiate further

peroxidative chain reactions which result in the rapid increases in

peroxidation rates seen in Figures 1 and 2. Mabrouk and Dugan (23)

found that free d%rbohydrates accelerated autocatalysis of lipid

peroxidation by activating hydroperoxide decomposition. The present

653

Vol. 151, No. 2, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

study demonstrates that lipid peroxidation can also be induced by

nonenzymatically glycosylated collagen, a fact which has not been

previously been reported.

Adventitious iron, has been postulated to form a loose complex with

carbohydrates (24). Figure 3 shows that this iron also plays a role

in peroxidation catalysed by both glucose and glycosylated

collagen. However, the addition of 1.0mM desferrioxamine, which

binds the iron in a form in which it is unable to catalyse either

lipid peroxidation or hydroperoxide decomposition (25), only

partially inhibits the overall rate of peroxidation.

When about 0.2mM hydroperoxide (3% of the total amount of lipid in

the vesicle preparation) had accumulated, vesicles began to lyse

(Figure 4). Similar alterations in the physical properties of

linoleic acid membranes have been observed when peroxidation was

initiated by gamma radiolysis (26) or lipoxygenase (27). Physical

changes have also been observed in a number of peroxidised natural

membranes (28). Similar increases in hydroperoxides in the

membranes of endothelial cells may be the cause of increases in

capillary permeability observed in diabetic microangiopathy.

A generalised increase in the amount of in vivo lipid oxidation

initiated by sustained hyperglycemia would be expected to

compromise natural antioxidant mechanisms. In keeping with this,

it has been found that dietary deficiency of, or supplementation

with vitamin E, potentiates or ameliorates a number of pathological

changes associated with diabetes (29, 30).

ACKNOWLEDGEMENTS:This work was supported by the Surgical Trust Fund of the Royal North Shore Hospital, The James N. Kirby Fund, The Baxter Perpetual Charitable Trust and FAI Insurances Ltd.

REFERENCES

i. Slater, T. F. (1987) Br. J. Cancer, 55, (Suppl. 8), 5-10. 2. Pryor, W. A. (1985) Am. Chem. Soc. Symp. Ser. 277, 77-96. 3. Loven, D. P., and Oberley, L° W. (1985) in Superoxide Dismutase

L. W. Oberley, Ed.) Vol III, pp. 151-189. CRC Press, Boca Raton, Florida.

4. Sato,Y., Sakamoto, N., Matsuoka, S., Ohishi, N., and Yagi, K. (1979) Biochem. Med. 21, 104-107.

5. Nishigaki, I., Hagihara, M., Tsunekawa, H., Maseki, M., and Yagi, K. (1981) Biochem. Med. 25, 373-378.

654

Vol. 151, No. 2, 1988 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

6. Murata, T. Nishada, T., Eto, S., and Mukai, N. (1981) Metabol. Pediat. Opthalmol. 5, 83-87.

7. Zigler, J. S., and Hess, H. H. (1985) Exp. Eye Res. 41, 67-76. 8. Mowri, H., Chinen, K., Ohkuma, S., and Takano, T. (1986)

Biochem. Int. 12, 347-352. 9. Cerami, A., Vlassara, H., and Brownlee, M. (1987) Sci. Am. 256,

82-88. i0. Wolff, S.P., Crabbe, M. J. C., and Thornalley, P. J. (1984)

Experientia 40, 244-246. ii. Jones, A. F., Winkles, J., Lunec, J., Jennings, P. E., and

Barnett, A. H. (1986) Clin. Sci. 71, (Suppl. 15), 33P. 12. Jones, A. F., and Lunec, J. (1987) Br. J. Cancer 55,

(Suppl. 8), 60-65. 13. Pecoraro, R. E., Graf, R. J. Halter, J. B., Beiter, H., and

Porte, D. (1979) Diabetes 28, 1120-1125. 14. Kohn, R. R., Cerami, A., and Monnier, V. M. (1984) Diabetes 33,

57-59. 15. Stegemann, H., and Stalder, K. (1967) Clin. Chim. Acta 18, 267-

273. 16. Gebicki, J. M. and Hicks, M. (1976) Chem. Phys. Lipids 16, 142-

160. 17. Pitt, G. A. J., and Morton, R. A. (1957) Progr. Chem. Fats

Lipids 4, 228-278. 18. Harman, D. (1984) Age 7, 111-131. 19. Tchobroutsky, G. (1978) Diabetologia 15, 143-152. 20. Brownlee, M., Vlassara, H., and Cerami, A. (1984) Ann. Internal

Med. i01, 527-537. 21. Jones, A. F., Jennings, P. E., Wakefield, A., Winkles, J.,

Lunec, J., and Barnett, A. H. (1986) New England J. Med. 315, 323-324.

22. Isbell, H. S., Frush, H. L., Wade, C. W. R., and Hunter, C. E. (1969) Carbohyd. Res. 9, 163-175.

23. Mabrouk, F., and Dugan, L. R. (1961) J. Am. Oil Chemists' Soc. 38, 692-695.

24. Halliwell, B., and Gutteridge, J. M. C. (1986) Arch. Biochem. Biophys. 246, 501-514.

25. Frankel, E. N. (1982) Prog. Lipid Res. 22, 1-33. 26. Hicks, M., and Gebicki, J. M. (1983) Proc. Aust. Biochem. Soc.

15, 5. 27. Hicks, M., and Gebicki, J. M. (1978) Biochem. Biophys. Res.

Commun. 80, 704-708. 28. Benedetti, A., Casani, A. F., Ferrali, M., and Comporti, M.

(1979) Biochem. J. 180, 303-312. 29. Slonim, A. E., Surber, M. L., Page, D. L., Sharp, R. A., and

Burr, I. M. (1983) J. Clin. Invest. 71, 1282-1288. 30. Bierenbaum, M. L., Noonan, F. J., Machlin, L. J., Machlin,

S., Stier, A., Watson,P. B., Naso, A. M., and Fleischman, A. I. (1985) Nutr. Rep. Int. 31, 1171-1180.

655