L I P I D P E R O X I D A T I O N A N D R E T I N O P A T H Y IN
S T R E P T O Z O T O C I N - I N D U C E D D I A B E T E S
DONALD ARMSTRONG a n d FARIDA AL-AWADI
Department of Medical Laboratory Technology and Biochemistry, Kuwait University Health Science Center, Kuwait
(Received 16 January 1991; Revised 19 April 1991; Accepted 14 May 1991)
A b s t r a c t - - U s i n g the streptozotocin (STZ)-induced diabetic rat model, we have established a time-related curve for lipid hydro- peroxides (LHP) in plasma and have correlated the period corresponding to maximal increase with histologic changes in the outer retina. Measurement of thiobarbituric acid reacting substances (TBARS) provides a convenient assessment of LHP concentration in plasma. Our results demonstrate a seven-fold elevation of TBARS at 10 days post-induction which increased to fifteen times above normal at 22 days and then fell dramatically to below baseline values at 39 days. Structural damage to the retina consisted of a reduction in cell number throughout the inner and outer nuclear layers, disorganization and loss of photoreceptor segments, and dilation of the basal region of the retinal pigment epithelium. The present observations establish a correlation between LHP concentration and retinal structure and function. Taken together with other reports in the literature showing alterations of protec- tive enzymes and antioxidants, it appears that free radicals and lipid peroxidation are involved in the etiology of diabetic retinopa- thy in the STZ rat model. The TBARS assay is a simple, sensitive and inexpensive method to monitor changes in oxidative status and may prove useful in diagnosis and monitoring of patients with diabetes.
Lipid peroxidation has been implicated in the pathogen- esis of many degenerative disorders,l'2 including natu- rally occurring 3 and chemically induced diabetes mel- litus. 4"5 Consequently, mechanisms in the formation of lipid hydroperoxides and biologically active metabo- lites, together with their effect on cellular structure and function are becoming of increasing importance to the study of diabetogenesis. 6
Lipid hydroperoxides (LHP) produced from a vari- ety of long-chain polyunsaturated fatty acid precursors via intermediate radical reactions, involve oxygen and metal cations (iron and copper). The net result of these combined reactions is the generation of highly reactive and cytotoxic lipid radicals, which generate new LHP because of their close proximity in biomembranes to other lipids. Extracellularly, lipid hydroperoxides are transported in the systemic circulation by low- and high-density lipoproteins. 3 When released locally, LHP
This study was supported by grant MDH 183 and in part by grant MB 011 from The State of Kuwait.
Address correspondence to Donald Armstrong, Ph.D., Dr.Sc., Department of Medical Technology, SUNY at Buffalo, UB Clinical Center, 462 Grider St., Buffalo, NY 14215.
433
produce structural damage. 7'8 Peroxidative regulation occurs through intervention by lipid and water-soluble antioxidants, as well as by specific antioxidant en- zymes, i.e., dioxide (1-) dismutase, peroxidase and catalase.
A convenient and frequently used assay for quanti- fication of LHP is the thiobarbituric acid (TBA) test, which measures various aldehydes derived primarily from LHP. Since other compounds may form chro- mogenic adducts with similar spectra, it is more cor- rectly referred to as measuring thiobarbituric acid react- ing substances (TBARS). Using this approach, others have reported TBARS in normal human plasma to be in the range of 0.60+-0.13 Ixmol/1, with no sex dif- ferences noted. 9,t° In uncontrolled diabetics, TBARS are elevated 33% above controls and in patients with advanced stage disease where nephropathy, retinopa- thy, and microangiopathy are present, the elevation is increased to 44-50%. 11'12 Increases of the same mag- nitude are also noted in long-term (1 yr) versus short- term (1 mo) animals made diabetic with alloxan. 13 Retinal tissues in the rat contain increased levels of LHP after treatment with diabetogenic drugs. 14.15
The present study has calculated the amount of TBARS in plasma of the streptozotocin (STZ), non-in- sulin supplemented rat model as a function of time, and
demonstrates a correlation for retinal damage coinci- dent with peak concentration.
MATERIALS AND METHODS
Animals
Rats were fed a standard commercial diet. Blood was collected from 24 normal, adult, male Wistar rats to establish baseline fasting and nonfasting TBARS values. The same animals then received a single, in- traperitoneal injection of STZ (60 mg/kg/BW; Sigma Chem. Co.), and were given insulin subcutaneously (1 unit/200 g BW) for 3 days until stabilized. Thereafter, no additional insulin was administered. When the blood glucose (Glucose Enzymatique PAP 1200, BioMerieux) reached more than 6 mmol/1 and the Glucostix (Lilly & Co., Indianapolis, IN) was positive, samples were collected and TBARS measured at 10, 22, and 39 days post-injection, when all remaining animals were sacri- ficed. In separate experiments, eyes from three rats were removed and processed for light microscopy at 22 days post-injection as previously described. 16
was freshly prepared and filtered prior to use. Two hundred ixl of specimen or the standard, was added to 1 ml of TBA reagent, incubated for 60 min in a water bath at 100°C, cooled to 25°C for 15 min, extracted with 5 ml of n-butanal/pyridine (15:1), the upper phase collected and the TBA adducts measured in an Aminco- Bowman spectrophotofluorometer set at 510 nm for excitation, and 553 nm for recording of emission val- ues. Calculations were performed using a linear regres- sion program.
RESULTS
Clinical Data
The mean weight before STZ was 282 ± 23 g and remained essentially unchanged during the first 10 days. However, by 22 days post-induction, body weights had decreased 25-30% on the average, to 213 ___ 38 g. Three animals expired during the initial 10 days, four more died spontaneously over the next 12 days, and ten sur- vived until the end of the study.
Fasting blood glucose levels before induction of di- abetes ranged between 3-6 mmol/1. During the first week of experimentation, the average of the group was 18___3 mmol/1 and it remained elevated (above 10 mmol/1) thereafter. Glucosuria (2+) was present one day after STZ administration and remained present throughout the experimental period.
TBARS values
Collection and preparation of plasma
Animals feeding ad libitum were sampled initially and then one week later, food was withheld from these animals for 24 h prior to obtaining the comparison fasting specimen. Rats were placed in a warming cham- ber (47°C) chamber for 10-15 min, the tip of the tail surgically removed, and approximately 1 ml of whole blood collected into test tubes containing 5.25 mg of EDTA. The mixed sample was centrifuged for 5 min at 3,000 rpm and the plasma separated. Plasma was mixed with a one-half volume of 20% trichloracetic acid, centrifuged for 5 min at 12 000 rpm in an Ep- pendorf microfuge (Eppendorf GmbH, Hamberg, West Germany) and the final supernatant decanted.
Analytic conditions
TBARS were determined using a modification 17 of the fluorescence method of Yagi. 1~ Standards were prepared at concentrations from 0.156 to 2.5 Ixmol/L using 1,1,3,3-tetramethoxypropane (Fluka Chem. Co.). A stock solution of 0.53% TBA (Sigma Chem. Co.) dissolved in glacial acetic acid and adjusted to pH 3.5
The mean nonfasting baseline value was 2.34__+0.33 Ixmol/1 as compared to a fasting value of 0.39__+0.92 ~mol/1. The latter value increased to 2.90___0.92 Ixmol/1 at 10 days following the induction of diabetes (p<0.02) and continued to show a linear response, reaching a maximum value of 6.19 ± 1.28 Ixmol/1 at 22 days (p<0.001). A precipitous drop occurred at 39 days with values recorded below baseline (Table 1).
Retinal pathology
The main histologic abnormality was a consistent and focal damage to the photoreceptor layer of the ret- ina (Fig. 1) in all animals examined. This consisted of loss of inner (IS) and outer (OS) photoreceptor seg- ments and reduction in the number of photoreceptor cells (ONL). The basal region of the retinal pigment epithelium (RPE) showed marked dilation. Although the full thickness of the retina of STZ animals was es- sentially unchanged, the distance from the surface of the inner nuclear layer (INL) to the surface of the choroidea was 20% less. Disorganization of cells in the INL was also observed.
Lipid peroxidation in diabetes 435
Fig. 1. Comparison of normal (a) and 22 day STZ-treated rat (b). The various layers from the inner to the outer retina are indicated: GCL = ganglion cell, INL = inner nuclear, ONL = outer nuclear, IS = inner segment, OS = outer segment, RPE = retinal pigment epithelium, and C = choroid (original mag. = 400X). Optimal fo- cus in the IS--OS layer is difficult because segments are badly dis- torted and displaced from the normal planar orientation.
DISCUSSION
The fluorescence TBA assay is a simple and inex- pensive test that reflects circulating LHP levels in plasma. 11 Its application in the clinical laboratory has recently been proposed. 9"1° In the present study, we found a higher nonfasting value which is probably due to an increase in triglycerides as reported by Nomura e t a l . , 18 since they are the main source of fatty acids available to react in the TBA assay. The ability of the test to detect changing concentrations of TBARS is re- flected in the STZ-induced, uncontrolled diabetic rat model, where a steady increase in fasting plasma lev- els as the disease progressed over a span of 4-5 wks was demonstrated. Thus, at 10 days following the on- set of diabetes, the TBARS value was already seven times above normal and by 22 days, the level was maximal (fifteen times above normal). Additional evi- dence of test sensitivity was shown by the measure- ments taken at 39 days when animals were severely affected by the disease process. At this point, TBARS were extremely low, suggesting that substrates for the assay had been depleted from their source(s).
Pilot experiments which measured TBARS in reti- nal tissues from the STZ-treated rats showed increased LHP production at 10 days (45% above baseline) and 17 days (74% above baseline) when maximal concen- tration was observed. By 22 days, the TBARS began the decrease but were still above baseline (31%) and from 30 to 39 days post-injection, the levels were slightly below (12%) baseline. These changes, parallel the trend reported for serum TBARS.
The formation of LHP and their metabolites are im- portant in clinical medicine because they alter mem- brane structure and function, especially in the retinal portion of eye which is very sensitive to oxidative
stress. For example, a steady decline is observed in the electroretinogram not only in the STZ model, t9 but also when synthetic LHP is injected into the vitreous of experimental animals. 2° These changes are irrevers- ible. Electrical activity depends on the integrity of the OS and the interface between it and the RPE. 19
The present study confirms changes in the basal in- foldings of the RPE as previously described, 21 however our findings appear to be the first histologic evidence of damage produced at the outer retina (Fig. 1) that correlates with elevated plasma LHP in STZ-induced diabetes, suggesting a causal relationship. Indeed, a detailed histological study has been conducted subse- quent to this report on a large number of STZ-treated rats. That data is being reported and it completely sup- ports our present hypothesis because sequential and progressive pathological changes referable to the outer segment region were detected at 7-10 days post-in- jection which persisted throughout the experimental period.
From these data, we may assume that elevated plasma levels of LHP result from the effect of STZ which ap- parently increases oxidative activity. Radical initiation and LHP formation may originate from a variety of cellular membranes, especially in organs lacking appro- priate defense mechanisms. In the pancreas for exam- ple, peroxidase is normally quite low and therefore the islet is prone to oxidative injury. 22"23 In the retinal outer photoreceptor segments, peroxidase is absent, 24 so radical propagation may proceed rapidly. 25 Other tissues show increased or decreased alterations in en- zyme activity in response to LHP arising in STZ-in- duced diabetes. 26 During experimental STZ diabetes, both serum peroxidase t3 and ot-tocopherol 5 are also decreased.
The evidence to data in the literature suggests a role for radicals and lipid peroxidation in the etiology of diabetes. 27 Our results reported here represent another finding confirming the value of TBARS as a labora- tory test in the diagnosis and perhaps management of diabetes, and furthermore links increased peroxidation to the progressive, visual complication associated with the retinopathic process.
Acknowledgements -- The authors wish to thank Ehad Abdel-Rah- mann, Amami AI-Amal, (Department of Biochemistry) Alice Prakash, (EM Unit) Jim Hall (Department of Anatomy), and Milan Sieber (Medical Photography Unit) for expert technical assistance.
REFERENCES
1. Armstrong, D.; Sohal, R.; Cutler, R.; Slater, T. Free radicals in molecular biology and aging. New York: Raven Press Publ.; 1982.
2. Yagi, K. Lipid peroxides in biology and medicine. New York: Academic Press, Inc.; 1982.
3. Nishigaki, I.; Hagishara, M.; Tsumakawa, H.; Maseki, M.;
436 D. ARMSTRONC and F. AL-AWAD!
Yagi, K. Lipid peroxide levels of serum lipoprotein fractions of diabetic patients. Biochem. Med. 25:373-378; 1981.
4. Rerup, C. Drugs producing diabetes through damage of the in- sulin secreting cells. Phamacol. Rev. 22:485-518; 1970.
5. Higuchi, Y. Lipid peroxides and et-tocopberol in rat streptozoto- cin-induced diabetes mellitus. Acta Med. Okayama 36:165-175; 1982.
6. Crabbe, M. Diabetic complications: Scientific and clinical as- pects. New York: Churchill Livingston, Inc.; 1987.
7. Berdanier, C. Role of membrane lipids in metabolic regulation. Nutr. Rev. 46:145-149; 1988.
8. Hennig, B.; Chow, C. Lipid peroxidation and endothelial cell injury: implications in atherosclerosis. Free Radical Biol. Med. 4:99-106; 1988.
9. Knight, J.; Smith, S.; Kinder, V.; Austall, H. Reference values for plasma lipoperoxides: Age, sex, and specimen related varia- tions. Clin. Chem. 33:2289-2291; 1987.
10. Wong, S.; Knight, J.; Hopfer, S.; Zaharia, O.; Leach, C.; Sun- derman, F. Lipoperoxides in plasma as measured by liquid- chromatographic separation of malondialydehyde-thiobarbituric acid adduct. Clin. Chem. 33:214-220; 1987.
11. Yagi, K. Assay for serum lipid peroxide level and its clinical significance. In: Yagi, K., ed. Lipid peroxides in biology and medicine. New York: Academic Press, Inc.; 1982:223-242.
12. Jennings, P.; Jones, A.; Florkowski, C.; Lunec, J.; Barnett, A. Increased diene conjugates in diabetic subjects with microan- giopathy. Diabetic Med. 4:452-456; 1987.
13. Kadoya, K. Lipoperoxide level and scavenging enzyme activi- ties in the diabetic albino rabbit retina. Acta. Soc. Ophthalmol. Jpn 85:646-655; 1981.
14. Murata, T.; Nishida, T.; Eto, S.; Mukai, N. Lipid peroxidation in diabetic rat retina. Metabol. Pediatr. Ophthalmol. 5:83-87; 1981.
15. Nishimura, C.; Kuriyama, K. Alterations of lipid peroxide and endogenous antioxidant contents in retina of streptozotocin-in- duced diabetic rats: Effect of Vitamin A administration. Jpn J. Pharmacol. 37:365-372; 1985.
16. Armstrong, D.; Hiramitsu, T. Studies on experimentally induced retinal degeneration. 2. Early morphological changes produced by lipid peroxides in the albino rabbit. Jpn J. Ophthalmol. 34: 158-173; 1990.
17. Abdella, N.; AI-Awadi, F.; Salman, A.; Armstrong, D. The TBA assay as a measure of lipid peroxidation in Arab patients with NIDDM. Clin. Diabetes Res.: Clin. Exp.; [1991].
18. Normura, T.; Hagino, Y.; Gotch, M.; Iguchi, A.; Sakamoto, N. The effect of streptozotocin diabetes on tissue specific lipase in the rat. Lipids 19:594-599; 1984.
19. Pautler, E.; Ennis, S. The effect of induced diabetes on the
electroretinogram components of pigmented rat. Invest. Ophthal- mol. Vis. Sci. 19:702-705; 1980.
20. Armstrong, D.; Hiramitsu, T.; Gutteridge, J.; Nilsson, S. Stud- ies on experimentally induced retinal degeneration. 1. Effect of lipid peroxides on electroretinographic activity in the albino rab- bit. Exp. Eye Res. 35:157-171; 1982.
21. Grimes, P.; Laties, A. Early Morphological alteration of the pigment epithelium in streptozotocin-induced diabetes: Increased surface area of the basal celt membranes. Exp. Eye Res. 30:631- 639; 1980.
22. Gankvist, K.; Marldund, S.; Taljedal, I. CuZn-superoxide dis- mutase, Mn-superoxide dismutase, catalase and glutathione per- oxidases in pancreatic islets and other tissues in the mouse. Biochem. J. 199:393-398; 1981.
23. Malaisse, W.; Malaisse-Lagae, F.; Sener, A.; Pipeleers, D. De- terminants of the selective toxicity of alloxan to the pancreatic B cell. Proc. Nat'l Acad. Sci. USA 79:927-930; 1982.
24. Armstrong, D.; Santangelo, G.; Connole, E. The distribution of peroxide regulating enzymes in the canine eye. Curr. Eye Res. 1:225-242;1981.
25. Handelman, G.; Datz, E. The role of antioxidants in the retina and retinal pigment epithelium and the nature of prooxidant-in- duced damage. In: Pryor, W.; ed. Advances in free radicals in biology and medicine. Vol. 2. Elmsford, NY: Pergamon Press; 1986.
26. Wohaieb, S.; Godin, D. Alterations in free radical tissue-defense mechanisms in streptozotocin-induced diabetes in rat. Effects of insulin treatment. Diabetes 36:1014-1018; 1987.
27. Oberly, L. Free radicals and diabetes. Free Radical Biol. Med. 5:113-124; 1988.
