Acta Physiol Scund 1992, 145, 151-157
Pro-oxidant effects of normobaric hyperoxia in rat tissues M. AHOTUPA, E. MANTYLA, V. PELTOLA, A. P U N T A L A and H. T O I V O N E N " Department of Physiology, University of Turku, Finland and Department of Anaesthesia, Helsinki University Hospital, Finland
"
AHOTUPA,M., MANTYLA,E., PELTOLA, V., PUNTALA, A. & TOIVONEN, H. 1992. Prooxidant effects of normobaric hyperoxia in rat tissues. Acta Physiol Scund 145, 151-157. Received 2 October 1991, accepted 24 January 1992. ISSN 0001-6772. Department of Physiology, University of Turku, and Department of Anaesthesia, Helsinki University Hospital, Finland. Rats were exposed to 1 0 0 ~ 0, o atmosphere for 12, 36 or 48 h, and their lungs, brain, liver and kidneys were studied for signs of oxidative damage. Oxidative damage at molecular level was estimated by: (1) the appearance of conjugated diene double bonds and (2) the amount of fluorescent chromolipids in lipids extracted from tissues. As important intracellular regulators of oxidative stress, the response of enzymes detoxifying reactive oxygen species was also studied. Macroscopically, the brain and the lungs were most susceptible to oxygen-induced effects. As an indication of oxidative tissue damage, hyperoxia caused accumulation of fluorescent chromolipids in brain and lung tissues, whereas diene conjugation did not reveal any signs of lipid peroxidation. Accumulation of fluorescent chromolipids was most prominent in the brain, where 99 and 138% increases over the control were detected after 36 and 48 h hyperoxia, respectively. Fluorescent chromolipids appeared in urine already before their concentrations were elevated in tissues. The activity of superoxide disrnutase in the brain was initially decreased, followed then by a slight induction of activity at the later time-points. Pulmonary and hepatic catalase activities were markedly decreased after prolonged (36 and 48 h) hyperoxia. In conclusion, fluorescent chromolipid formation seems to be a sensitive indicator of hyperoxia-induced oxidative damage in rat tissues. The lipid peroxidation-derived fluorescent chromolipids are eliminated from the body via urinary excretion. Moreover, impaired detoxication of reactive oxygen may be implicated in tissue damage due to hyperoxia. Key words : antioxidant enzymes, brain, hyperoxia, kidney, lipid peroxidation, liver, lung, rat. Exposure to a hyperoxic environment leads to elevated partial pressure of oxygen, especially in the lungs but also in other tissues. As a function of Po,, oxygen consumption is augmented and the fraction of oxygen metabolism that produces partly reduced oxygen species (superoxide anion, hydrogen peroxide, hydroxyl radical) is increased (Fridovich 1970, Freeman & Crapo 1981, Crapo Correspondence : Markku Ahotupa, Department of Physiology, University of Turku, Kiinamyllynkatu 10, SF-20520 Turku, Finland.
1986). As initially proposed by Gerschmann et al. (1954), reactive oxygen species play a decisive role in development of oxygen toxicity. This idea is supported e.g. by the role of antioxidant enzyme induction in development of tolerance to oxygen toxicity (see Fridovich & Freeman 1986), by effects of various non-enzymatic antioxidants (see Haugaard 1968, Klein 1990) and also by the direct evidence on superoxide and hydroxyl radical formation in pulmonary endothelial cells exposed to oxygen (Zweier et al. 1989). Against this background it seems somewhat
151
152
.21. .4hotzipn et al.
surprising that only small increases in pulmonary lipid peroxide levels are found in hyperoxic lung (Nishiki er ul. 1976, Freeman et al. 1982). YIoreover, lipid peroxide levels are increased in the brains of animals exposed to elevated pressures of oxygen, b u t n o difference can b e seen in lipid peroxide levels between the convulsing and non-convulsing animals (Becker S. Galvin 1962). I n most studies concerning lipid peroxidation in hyperoxic tissues, the amount of lipid peroxidation has been estimated b!- the classical thiobarbituric acid-method. Yet, tissue lipid peroxidation is a complex network of reactions with varying substrates, pathwa!-s, and e n d products (Porter 1985) and, hence, estimation of the actual level of 'lipid peroxidation' in a particular tissue has been a matter of contro\-ers!-. T h e existing methods to measure lipid peroxidation frequently give apparently contradictory results (e.g. S m i t h et nl. 1982, Kostrucha &- K a p p u s 1986) and, as a n important methodological aspect, the concentration of oxygen in the respired air of experimental animals ma!profoundly influence the lipid peroxidation pathways tn i'ii'o (.\luliawan &. K a p p u s 1983). I n the present stud!- we have investigated the pro-oxidative effects of hyperoxia in lung, liver, kidney and brain tissue of the rat, estimating the le\-el of oxidative damage: (1) by the a m o u n t of diene conjugation a n d (2) by the lipid peroxidation-induced formation of fluorescent chromolipids in tissues. Since the cellular levels of reactive ou!-gen species are regulated by the antioxidant enzymes, superoxide dismutase and catalase, the response of these enzymes t o hyperoxia was also studied.
II.%TERI.4I.S AND I I E T H O D S Ckumriuls. .-ldrenaline, hydrogen peroxide and the en7ymes superoxide dismutase (Cu/Zn-form, bovine erythrocytes) and catalase (bo\-ine h e r ) \\ere purchased from Sigma Chemical Co. (St Louis, 110, LS.1). E.tpo.iwr to /i,,,perc/.xid. T h e animal experiments described here Kere performed with the permission of the Ethics Committee for Animal Experiments of the L~nil-ersir!of Turliu. r\dult male \i-istar rats weighing ZX(b.120 g were exposed to a hyperoxic environment for 12, 36 or 48 h in perspex chambers as described earlier (Valimalii et al. 1974). The chamber (21x 33 x 40 cm) \ u s airtight except for one small ( 2 mm in diameter) hole on each side for the exit of gas. The top of the chamber a a s transparent, with
12 h light/dark c!-cle in the room. Three to five animals at a time were in the chamber where they had free access to water and commercial food pellets. The chamber was continuously flushed with 1000; 0, at 1 I x min-' per animal, which was known to maintain hyperoxic environment and prevent CO, accumulation. The temperature was 23-25 "C which was ahout 2 "C above the room temperature. The chambers were cleaned once a day, causing an interruption of hyperoxia for about 2 min. For every time-point measured, two non-exposed control animals were analysed together with the oxygen exposed animals and these rats were combined together to represent the control group. .4fter the exposure the rats were killed with 1009, carbon dioxide ; after reaching unconsciousness their neck was cut to exsanguinate the animals. Tissur preparation. The lungs, brain, liyer and kidneys were dissected out, rinsed in saline and weighed. A 20°, (w/v) homogenate m-as prepared in a 0.25 M sucrose solution (0 "C) with a PotterElvehjem glass-Teflon homogenizer, except the lungs that were homogenized with Ultra-Turrax homogenizer. The homogenates were stored at -50 "C and the samples of two rats from each time-point were alm!-s analysed at the same time to minimize day-today variations in analyses. Dime conjugation. Lipids that were extracted from tissue homogenates by chloroform-methanol, dried under nitrogen atmosphere and then redissolved in cyclohexane, w-ereanalysed spectrophotometrically (at 232 nm) as described (Corongiu et al. 1983). Fluorrscrnt chromolipids. Fluorescent chromolipids were assayed in tissue homogenates essentially as described by Esterbauer et al. (1986). T h e samples were eluted with chloroform-methanol (2: 1) and centrifuged at 2000 g max for 10 min. The organic phase was evaporated to dryness and redissolved in chloroform-methanol (10: l), whereafter the fluorescence was measured at 360 nm (excitation)/430 nm (emission). For technical reasons, we were not able to collect rat urine during the oxygen exposure. Urinary excretion of fluorescent chromolipids was measured in another group of rats after 1, 6 or 12 h of oxygen exposure. Immediately after the challenge the rats were put into a metabolic cage and their urine was collected for the next 10 h. The urine was analysed for fluorescent chromolipids as described for tissue samples aboi-e. .4ntiosidant enzymes. Superoxide dismutase (Cu/Zn-form) and catalase activities were determined in tissue homogenates by the following methods. The actij-itl-of superoxide dismutase was assayed spectrophotometrically by inhibition of adrenaline autoxidation (Zlisra & Fridovich 1972). Catalase activity was determined by measuring thc rate of disappearance of 1-i mb%hk-drogen peroxide at 240 nm (Beers & Sizer
Oxidative damage in rat tissues 1952). In the analysis of superoxide dismutase activity, 1 pg of the purified enzyme preparation (from bovine erythrocytes; Sigma) corresponds to 3.6 U, and 1 pg of catalase (bovine liver catalase ; Sigma) corresponds to 2.5 U. Protein. Protein content was measured by the biuret method (Layne 1957) with bovine serum albumin as the reference protein. Statistics. Dunnett's test was used in the statistical evaluation of the results. A P-value less than 0.05 was selected as the limit of statistical significance.
153
4 400
i
.E
r
(b) Liver
2?1 I
I
I
I
36
48
I
(b) Liver 300
250
I
1
I
1
r
t
I
I
b
0.6
O5
'.O
0.9 0.8
0.7
-
0
1 t
(d) Kidney
12
.
24
Hyperoxia (h)
Fig. 2. Tissue protein concentrations of rats exposed to 1 0 0 ~ ooxygen. Each point represents the mean fSEM from six animals. * P < 0.05. I
I
r
1
I
I
(d) Kidney
1k-?-----i\i 1
I
I
0
12
24
I
I
36
48
Hyperoxia (h)
Fig. 1. Organ/body wt ratios of rats exposed to 100yooxygen. Each point represents the mean SEM from six animals. ** P < 0.01.
RESULTS T h e first sign of oxygen toxicity was the increase of brain wet tissue wt that was evident already after 12 h, and increased thereafter reaching a striking increase of 58% after 48 h of oxygen exposure (Fig. 1). T h e protein content of brain tissue remained unchanged (Fig. 2 ) suggesting proteinaceous oedema formation in the brains. Damage to rat pulmonary tissue was evident somewhat later, after 36 h in an oxygen atmosphere (Fig. l), when lung/body wt ratio was increased by 31%. In contrast to the brain, the
154
'l4..4hotupn e t al.
.*
8.0 70.
f-,-.'
,'"
50-
/
t.
20 10-
0
12
24
36
48
Hyperoxia (h) Fig. 3. \tct/dr! at ratio of rat lungs after exposure to lOO",, oskgen. Each point represents the mean iSESI
from six animals. # * P < 0.01. neight gain of thc lungs was caused b!- the increased tissue 1% atcr, as tissue protein content was decreased (Fig. 2) and the \\-et/dr!- wt ratio of the tissue increased from the control of 2..i to 6.0 (Fig. 3). r i v e r and kidney were more resistant to macroscopicall!- detectable changes, as neither the organ heights nor the protein contents were changed during the first 36 h. After 48 h hyperoxia, hou.ever, the h e r s had shrunk as the liver/hod!- wt ratio h a s decreased by 24(',,, with 11 simultaneous increase in protein concentration of liver tissue (Figs 1 &- 2 ) . T h e amount of tluorescent chromolipids increased in the kidney by 3 j 0 , already after 12 h of h!-pcrouic exposure, and remained elekated at the later time-points (Fig. 4). Similar increases were detected after 36 and 48 h in the brain (99 and 138",, increases, respecti\-el!-) and the lungs (27 and 32'j0 increases, respectively) whereas no changes \\ere detected in liver tissue (Fig. 4). T h e early- increase of fluorescent chromolipids in the kidney ma!- be explained blurinar!- excretion of the chromolipid, as it \\as detectable in the urine aiready after 6 h of oxygen exposure, and the amount excreted correlated with the length of the hyperosic period (Fig. 5 ) . Yo efkcts due to hpperoaia could be detected in the amount of d i m e conjugation of the lung,
r
(d) Kidney
lOOr
80C
..i
,/ u . 0 12 24 36 48
Hyperoxia (h)
Fig. 4. The amount of fluorescent chromolipids in rat tissues after exposure to
IOOO,, oxygen. Each point represents the mean fSEW1 from six animals.
P < 0.01. brain or kidney. Only in the case of the liver was there an effect, an increase by 3 9 O 4 after 48 h hpperoxia (data not shown). I n brain tissue, a decline of superoxide dismutase activity was detected at the first timepoint studied. When measured after 36 or 48 h h!-peroxia, however, slightly increased levels of the actil-ity (30°,, P < 0.05, and 359;, I-' < 0.01, respectively) were seen (data not shown). I n the case of the lung, quantitation of the effects of h!-peroxia was hampered by methodological difficulties. T h e level of superoxide dismutase
Oxidative damage in rat tissues
A
IlOC
900 a,
w
155
LUNG
' 3 700
m
t-
8 500 9 300 L
I
I
I
I
0
12
24
36
48
Time (h) 40
T
L 0
1
6
12
02 exposure (h)
2o
k
LIVER
t
0 12 24 38 46 Fig. 5. Chloroform-methanol extractable fluorescent material excreted in rat urine after exposure to lOOyo Time (h) oxygen. Urine was collected during the 10 h period Fig. 6. Catalase (CAT) activity in lung and liver immediately after oxygen exposure. Each point tissues of rats exposed to 100% oxygen. Each point represents the mean SEM from four animals. represents the mean fSEM from six animals.
* P < 0.05; **P < 0.01. activity in control lung was 92.8k10.5 (mean & SEM) p g SOD g-l tissue, and after 12 h hyperoxia this activity was decreased by 31 yo (data not shown). However, when analyses were performed from lung homogenates of rats exposed for 36 or 48 h to hyperoxia, no scavenging of the superoxide anion could be detected but, on the contrary, the oxidation of adrenaline was increased. This effect was probably due to the appearance in lung tissue of reactive substances capable of oxidizing adrenaline. Regardless of this, the fact remains that the enzymatic superoxide anion detoxifying capacity in lungs of rats was severely impaired after prolonged hyperoxia. Hyperoxia did not affect superoxide dismutase activity in the liver (data not shown). Hyperoxia brought about a drastic decline in pulmonary and hepatic catalase activities, and the activities remained decreased to the end of the hyperoxic exposure period (Fig. 6). The
effect was more pronounced in the lung than in the liver. DISCUSSION The objective of the present study was to obtain more detailed information on mechanisms of hyperoxia-induced oxidative damage to rat tissues. The hyperoxia period was sufficient to produce lung damage tb rats, as demonstrated by increased wet wt/dry wt ratio of the tissue. Another tissue profoundly affected by hyperoxia was the brain. The brain was the first tissue to show oedema formation, already after 12 h exposure, followed by pulmonary oedema formation after 36 h of exposure. Contrary to the lungs, the weight gain of the brain was not due to accumulation of fluid, as the protein content of the brain did not decrease with increasing organ weight.
156
M. Ahotupa et al.
With the exception of the liver, hyperoxia did not affect the level of diene conjugation in rat tissues. This result is in agreement with earlier reports where lipid peroxidation was measured by the thiobarbituric acid method (Nishiki et a / . 1976, Freeman et a / . 1982). T h e apparent lack of lipid peroxidation would suggest that the reactive oxygen species generated by hyperoxia would exert their effects without involvement of lipids. This conclusion is, however, contradicted by results of the present study showing hyperoxiadependent accumulation of fluorescent chromolipids in rat tissues and, moreover, by increased urinary excretion of fluorescent chromolipids due to hyperoxia. OH-ing to the high sensiti\-ity, fluorescent chromolipid assay has been one of the commonly used methods to quantitate lipid peroxidation in tissues (Tappel 1980, Tsuchida et ai. 1985). Fluorescent compounds with similar spectral characteristics have been detected in lipid extracts of age pigment (lipofuscin) (Tappel 1980). high concentrations of which have been found e.g. in the brains of patients with -4lzheimer’s disease or Down’s syndrome (see Cohen 1985). In the present study the hyperoxiacaused accumulation of fluorescent chromolipids was strikingly prominent in the brain, but was also evident in the lung and kidney. T h e tendency of the brain to accumulate fluorescent chromolipids may indicate higher susceptibilitl- to oxidative stress-caused effects, or alternativell- a poor ability to eliminate these lipid peroxidation products. Although the kidney was macroscopicall!resistant to oxygen damage, the concentration of fluorescent chromolipids in the kidney increased, and even more rapidly than in the brain or lungs. This was most probably due to excretion of these chromolipids from the body via the kidneys, as shown by the urinary excretion of these chromolipids. T h e fact that the urinary excretion of fluorescent chromolipids increased after hvperoxia treatment, and was proportional to the length of hyperoxic challenge, indeed, indicates the existence of physiological mechanism(s) for the elimination of these substances from the body. A most prominent effect caused by hyperoxia was the inactivation of the enzymes detoxifying oxygen-free radicals. In accordance, it has been demonstrated earlier that normobaric hyperoxia (6+100°, 0, atmosphere) inactivates the pul-
monary Cu/Zn-superoxide dismutase activity (Januszkiewicz et al. 1986), whereas prolonged exposure to sublethal hyperoxia results in induction of the above and also other antioxidant enzymes (see Fridovich & Freeman 1986). Hyperoxia-caused inactivation of a multitude of enzymes with important metabolic functions is well documented and has been suggested to play a crucial role in the mechanism of oxygen toxicity (Haugaard 1968). Oxygen-free radicals are implicated not only in hyperoxia-induced tissue damage, but also in radiation injury (Gerschman et al. 1954), and are also the ultimate damaging species in toxic processes caused by several different chemicals (see Smith 1986). Similarly to hyperoxia, we have shown that enzymatic antioxidant functions are impaired also upon exposure to other causes of oxidative stress, such as UV-irradiation (Punnonen et al. 1991) or carcinogenic chemicals (Ahotupa et al. 1987). Impaired defence against reactive oxygen favours the formation of cellular pro-oxidant states, and may thus be implicated in the further development of oxidative damage. In short, the present study clearly demonstrated that exposure to hyperoxia leads to accumulation of fluorescent chromolipids in rat tissues. As the fluorescent chromolipids are formed in reactions of peroxidized lipids with phospholipids, these observations confirm the involvement of lipid peroxidation in tissue damage due to hyperoxia. Lung, brain, liver and kidney differ from each other in their susceptibility to hyperoxia-induced oxidative damage and, moreover, there seem to exist physiological mechanisms for the elimination of fluorescent chromolipids from the body via urinary excretion. T h e decreased superoxide dismutase and catalase activities most probably contribute to the overall pro-oxidative effects caused by hyperoxia. The technical assistance of Mrs Leena Soderholm and hlrs Raija Soderholm is gratefully acknowledged.The stud! was financially supported by the Academy of Finland, the Finnish Cancer Foundation and the Juho 1-ainio Foundation. Finland.
REFERENCES
M.,BLSS4CCHINI-GRIOT, v., BEREZIAT, J -C., C 4 u q .4.51., & BARTSCH, H. 1987. Rapid oxidathe stress induced by N-nitrosoamines. Brochem Biophys Res Commun 146, 1047-10.54
.\HOTLP4,
Oxidative damage in rat tissues BECKER, N.H. & GALVIN, J.F. 1962. Effect of oxygenrich atmospheres on cerebral lipid peroxides. Aerospace Med 33, 985-987. BEERS,B. & SIZER, W. 1952. A spectrophotometric method for measuring the breakdown of hydrogen peroxide by catalase. 3 Biol Chem 195, 133-139. COHEN,G. 1985. Oxidative stress in the nervous system. In: H. Sies (ed.), Oxidative Stress, pp. 383402. Academic Press, New York. CORONGIU, F., LAI,M. & MILIA,A. 1983. Carbon tetrachloride, bromotrichloromethane and ethanol acute intoxication. Biochem 3 212, 625-631. CRAPO, J.D. 1986. Morphologic changes in pulmonary oxygen toxicity. Ann Rev Physiol48, 721-731. ESTERBAUER, H., KOLLER,E., SLEE, R.G. & KOSTER, J.F. 1986. Possible involvement of the lipid peroxidation product 4-hydroxynonenal in the formation of fluorescent chromolipids. Biochem 3 239, 405409. FREEMAN, B.A. & CRAPO,J.D. 1981. Hyperoxia increases oxygen radical production in rat lungs and lung mitochondria. 3Biol Chem 256, 10986-10992. FREEMAN, B.A., TOPOLSKY, M.K. & CRAPO,J.D. 1982. Hyperoxia increases oxygen radical production in rat lung homogenates. Arch Biochem Biophys 216, 477484. FRIDOVICH, I. 1970. Quantitative aspects of the production of superoxide anion radical by milk xanthine oxidase. 3 Biol Chem 245, 40534057. FRIDOVICH, I. & FREEMAN, B. 1986. Antioxidant defenses in the lung. Ann Rev Physiol48, 693-702. GERSCHMAN, R., GILBERT,D.L., NYE,S., DWYER, W. & FENN,W.O. 1954. Oxygen poisoning and xirradiation: a mechanism in common. Science 119, 623-626. HAUGAARD, N. 1968. Cellular mechanisms of oxygen toxicity. Physiol Rev 48, 31 1-373. JANUSZKIEWICZ, A.J., HUNTRAKOON, M., WILSON,P.K. & FAIMAN, M.D. 1986. Isolated perfused lung histamine release, lipid peroxidation, and tissue superoxide dismutase from rats exposed to normobaric hyperoxia. Toxicology 39, 3746. KLEIN, J. 1990. Normobaric pulmonary oxygen toxicity. Anest Analg 70, 195-207. KOSTRUCHA, J. & KAPPUS,H. 1986. Inverse re-
157
lationship of ethane or n-pentane and malondialdehyde formed during lipid peroxidation in rat liver microsomes with different oxygen concentrations. Biochim Biophys Acta 879, 120-125. LAYNE, E.K. 1957. Spectrophotometric and turbidimetric methods for measuring protein. Methods Enzymol 3, 447454. I. 1972. The role of MISRA,H.P. & FRIDOVICH, superoxide anion in the autoxidation of epinephrine and a simple assay for superoxide dismutase. 3 Biol Chem 247, 3170-3175. H. & KAPPUS,H. 1983. Ferrous ionMULIAWAN, stimulated alkane expiration in rats treated with carbon tetrachloride. Toxicology 28, 29-36. N. & CHANCE, B. NISHIKI, K., JAMIESON, D., OSHINO, 1976. Oxygen toxicity in the perfused rat liver and lung under hyperbaric conditions. Biochem 3 160, 343-3 5 5. PORTER,N.A. 1984. Chemistry of lipid peroxidation. Methods Enzymol 105, 273-282. PUNNONEN,K., PUNTALA,A., JANSEN, C.T. & AHOTUPA, M. 1991. UVB Irradiation induces lipid peroxidation and decreases antioxidant enzyme activities in human keratinocytes in vitro. Acta Dermato- Venereol71, 239-273. SMITH,L.L. 1986. The response of the lung to foreign compounds that produce free radicals. Ann Rev Physiol48, 681-692. SMITH,M.T., THOR, H., HARTZELL, P. & ORRENIUS, S. 1982. The measurement of lipid peroxidation in isolated hepatocytes. Biochem Pharmacol31, 19-26. TAPPEL, A.L. 1980. Measurement and protection from in vivo lipid peroxidation. In: W.A. Pryor (ed.), Free Radicals in Biology vol. 4, pp. 1117. Academic Press, New York. TSUCHIDA, M., MIURA,T., MIZUTANI, K. & AIBARA, K. 1985. Fluorescent substances in mouse and human sera as a parameter of in vivo lipid peroxidation. Biochim Biophys Acta 834, 196-204. J. 1974. VALIMAKI,M., KIVISAARI, J. & NIINIKOSKI, Permeability of alveolar capillary membrane in oxygen poisoning. Aerospace Med 45, 370-374. P., SYLVESTER, ZWEIER, J.L., DUKE,S.S., KUPPUSAMY, J.T. & GABRIELSON, E.W. 1989. Electron paramagnetic resonance evidence that cellular oxygen toxicity is caused by the generation of superoxide and hydroxyl free radicals. FEBS Lett 252, 12-16.
Pro-oxidant effects of normobaric hyperoxia in rat tissues M. AHOTUPA, E. MANTYLA, V. PELTOLA, A. P U N T A L A and H. T O I V O N E N " Department of Physiology, University of Turku, Finland and Department of Anaesthesia, Helsinki University Hospital, Finland
"
AHOTUPA,M., MANTYLA,E., PELTOLA, V., PUNTALA, A. & TOIVONEN, H. 1992. Prooxidant effects of normobaric hyperoxia in rat tissues. Acta Physiol Scund 145, 151-157. Received 2 October 1991, accepted 24 January 1992. ISSN 0001-6772. Department of Physiology, University of Turku, and Department of Anaesthesia, Helsinki University Hospital, Finland. Rats were exposed to 1 0 0 ~ 0, o atmosphere for 12, 36 or 48 h, and their lungs, brain, liver and kidneys were studied for signs of oxidative damage. Oxidative damage at molecular level was estimated by: (1) the appearance of conjugated diene double bonds and (2) the amount of fluorescent chromolipids in lipids extracted from tissues. As important intracellular regulators of oxidative stress, the response of enzymes detoxifying reactive oxygen species was also studied. Macroscopically, the brain and the lungs were most susceptible to oxygen-induced effects. As an indication of oxidative tissue damage, hyperoxia caused accumulation of fluorescent chromolipids in brain and lung tissues, whereas diene conjugation did not reveal any signs of lipid peroxidation. Accumulation of fluorescent chromolipids was most prominent in the brain, where 99 and 138% increases over the control were detected after 36 and 48 h hyperoxia, respectively. Fluorescent chromolipids appeared in urine already before their concentrations were elevated in tissues. The activity of superoxide disrnutase in the brain was initially decreased, followed then by a slight induction of activity at the later time-points. Pulmonary and hepatic catalase activities were markedly decreased after prolonged (36 and 48 h) hyperoxia. In conclusion, fluorescent chromolipid formation seems to be a sensitive indicator of hyperoxia-induced oxidative damage in rat tissues. The lipid peroxidation-derived fluorescent chromolipids are eliminated from the body via urinary excretion. Moreover, impaired detoxication of reactive oxygen may be implicated in tissue damage due to hyperoxia. Key words : antioxidant enzymes, brain, hyperoxia, kidney, lipid peroxidation, liver, lung, rat. Exposure to a hyperoxic environment leads to elevated partial pressure of oxygen, especially in the lungs but also in other tissues. As a function of Po,, oxygen consumption is augmented and the fraction of oxygen metabolism that produces partly reduced oxygen species (superoxide anion, hydrogen peroxide, hydroxyl radical) is increased (Fridovich 1970, Freeman & Crapo 1981, Crapo Correspondence : Markku Ahotupa, Department of Physiology, University of Turku, Kiinamyllynkatu 10, SF-20520 Turku, Finland.
1986). As initially proposed by Gerschmann et al. (1954), reactive oxygen species play a decisive role in development of oxygen toxicity. This idea is supported e.g. by the role of antioxidant enzyme induction in development of tolerance to oxygen toxicity (see Fridovich & Freeman 1986), by effects of various non-enzymatic antioxidants (see Haugaard 1968, Klein 1990) and also by the direct evidence on superoxide and hydroxyl radical formation in pulmonary endothelial cells exposed to oxygen (Zweier et al. 1989). Against this background it seems somewhat
151
152
.21. .4hotzipn et al.
surprising that only small increases in pulmonary lipid peroxide levels are found in hyperoxic lung (Nishiki er ul. 1976, Freeman et al. 1982). YIoreover, lipid peroxide levels are increased in the brains of animals exposed to elevated pressures of oxygen, b u t n o difference can b e seen in lipid peroxide levels between the convulsing and non-convulsing animals (Becker S. Galvin 1962). I n most studies concerning lipid peroxidation in hyperoxic tissues, the amount of lipid peroxidation has been estimated b!- the classical thiobarbituric acid-method. Yet, tissue lipid peroxidation is a complex network of reactions with varying substrates, pathwa!-s, and e n d products (Porter 1985) and, hence, estimation of the actual level of 'lipid peroxidation' in a particular tissue has been a matter of contro\-ers!-. T h e existing methods to measure lipid peroxidation frequently give apparently contradictory results (e.g. S m i t h et nl. 1982, Kostrucha &- K a p p u s 1986) and, as a n important methodological aspect, the concentration of oxygen in the respired air of experimental animals ma!profoundly influence the lipid peroxidation pathways tn i'ii'o (.\luliawan &. K a p p u s 1983). I n the present stud!- we have investigated the pro-oxidative effects of hyperoxia in lung, liver, kidney and brain tissue of the rat, estimating the le\-el of oxidative damage: (1) by the a m o u n t of diene conjugation a n d (2) by the lipid peroxidation-induced formation of fluorescent chromolipids in tissues. Since the cellular levels of reactive ou!-gen species are regulated by the antioxidant enzymes, superoxide dismutase and catalase, the response of these enzymes t o hyperoxia was also studied.
II.%TERI.4I.S AND I I E T H O D S Ckumriuls. .-ldrenaline, hydrogen peroxide and the en7ymes superoxide dismutase (Cu/Zn-form, bovine erythrocytes) and catalase (bo\-ine h e r ) \\ere purchased from Sigma Chemical Co. (St Louis, 110, LS.1). E.tpo.iwr to /i,,,perc/.xid. T h e animal experiments described here Kere performed with the permission of the Ethics Committee for Animal Experiments of the L~nil-ersir!of Turliu. r\dult male \i-istar rats weighing ZX(b.120 g were exposed to a hyperoxic environment for 12, 36 or 48 h in perspex chambers as described earlier (Valimalii et al. 1974). The chamber (21x 33 x 40 cm) \ u s airtight except for one small ( 2 mm in diameter) hole on each side for the exit of gas. The top of the chamber a a s transparent, with
12 h light/dark c!-cle in the room. Three to five animals at a time were in the chamber where they had free access to water and commercial food pellets. The chamber was continuously flushed with 1000; 0, at 1 I x min-' per animal, which was known to maintain hyperoxic environment and prevent CO, accumulation. The temperature was 23-25 "C which was ahout 2 "C above the room temperature. The chambers were cleaned once a day, causing an interruption of hyperoxia for about 2 min. For every time-point measured, two non-exposed control animals were analysed together with the oxygen exposed animals and these rats were combined together to represent the control group. .4fter the exposure the rats were killed with 1009, carbon dioxide ; after reaching unconsciousness their neck was cut to exsanguinate the animals. Tissur preparation. The lungs, brain, liyer and kidneys were dissected out, rinsed in saline and weighed. A 20°, (w/v) homogenate m-as prepared in a 0.25 M sucrose solution (0 "C) with a PotterElvehjem glass-Teflon homogenizer, except the lungs that were homogenized with Ultra-Turrax homogenizer. The homogenates were stored at -50 "C and the samples of two rats from each time-point were alm!-s analysed at the same time to minimize day-today variations in analyses. Dime conjugation. Lipids that were extracted from tissue homogenates by chloroform-methanol, dried under nitrogen atmosphere and then redissolved in cyclohexane, w-ereanalysed spectrophotometrically (at 232 nm) as described (Corongiu et al. 1983). Fluorrscrnt chromolipids. Fluorescent chromolipids were assayed in tissue homogenates essentially as described by Esterbauer et al. (1986). T h e samples were eluted with chloroform-methanol (2: 1) and centrifuged at 2000 g max for 10 min. The organic phase was evaporated to dryness and redissolved in chloroform-methanol (10: l), whereafter the fluorescence was measured at 360 nm (excitation)/430 nm (emission). For technical reasons, we were not able to collect rat urine during the oxygen exposure. Urinary excretion of fluorescent chromolipids was measured in another group of rats after 1, 6 or 12 h of oxygen exposure. Immediately after the challenge the rats were put into a metabolic cage and their urine was collected for the next 10 h. The urine was analysed for fluorescent chromolipids as described for tissue samples aboi-e. .4ntiosidant enzymes. Superoxide dismutase (Cu/Zn-form) and catalase activities were determined in tissue homogenates by the following methods. The actij-itl-of superoxide dismutase was assayed spectrophotometrically by inhibition of adrenaline autoxidation (Zlisra & Fridovich 1972). Catalase activity was determined by measuring thc rate of disappearance of 1-i mb%hk-drogen peroxide at 240 nm (Beers & Sizer
Oxidative damage in rat tissues 1952). In the analysis of superoxide dismutase activity, 1 pg of the purified enzyme preparation (from bovine erythrocytes; Sigma) corresponds to 3.6 U, and 1 pg of catalase (bovine liver catalase ; Sigma) corresponds to 2.5 U. Protein. Protein content was measured by the biuret method (Layne 1957) with bovine serum albumin as the reference protein. Statistics. Dunnett's test was used in the statistical evaluation of the results. A P-value less than 0.05 was selected as the limit of statistical significance.
153
4 400
i
.E
r
(b) Liver
2?1 I
I
I
I
36
48
I
(b) Liver 300
250
I
1
I
1
r
t
I
I
b
0.6
O5
'.O
0.9 0.8
0.7
-
0
1 t
(d) Kidney
12
.
24
Hyperoxia (h)
Fig. 2. Tissue protein concentrations of rats exposed to 1 0 0 ~ ooxygen. Each point represents the mean fSEM from six animals. * P < 0.05. I
I
r
1
I
I
(d) Kidney
1k-?-----i\i 1
I
I
0
12
24
I
I
36
48
Hyperoxia (h)
Fig. 1. Organ/body wt ratios of rats exposed to 100yooxygen. Each point represents the mean SEM from six animals. ** P < 0.01.
RESULTS T h e first sign of oxygen toxicity was the increase of brain wet tissue wt that was evident already after 12 h, and increased thereafter reaching a striking increase of 58% after 48 h of oxygen exposure (Fig. 1). T h e protein content of brain tissue remained unchanged (Fig. 2 ) suggesting proteinaceous oedema formation in the brains. Damage to rat pulmonary tissue was evident somewhat later, after 36 h in an oxygen atmosphere (Fig. l), when lung/body wt ratio was increased by 31%. In contrast to the brain, the
154
'l4..4hotupn e t al.
.*
8.0 70.
f-,-.'
,'"
50-
/
t.
20 10-
0
12
24
36
48
Hyperoxia (h) Fig. 3. \tct/dr! at ratio of rat lungs after exposure to lOO",, oskgen. Each point represents the mean iSESI
from six animals. # * P < 0.01. neight gain of thc lungs was caused b!- the increased tissue 1% atcr, as tissue protein content was decreased (Fig. 2) and the \\-et/dr!- wt ratio of the tissue increased from the control of 2..i to 6.0 (Fig. 3). r i v e r and kidney were more resistant to macroscopicall!- detectable changes, as neither the organ heights nor the protein contents were changed during the first 36 h. After 48 h hyperoxia, hou.ever, the h e r s had shrunk as the liver/hod!- wt ratio h a s decreased by 24(',,, with 11 simultaneous increase in protein concentration of liver tissue (Figs 1 &- 2 ) . T h e amount of tluorescent chromolipids increased in the kidney by 3 j 0 , already after 12 h of h!-pcrouic exposure, and remained elekated at the later time-points (Fig. 4). Similar increases were detected after 36 and 48 h in the brain (99 and 138",, increases, respecti\-el!-) and the lungs (27 and 32'j0 increases, respectively) whereas no changes \\ere detected in liver tissue (Fig. 4). T h e early- increase of fluorescent chromolipids in the kidney ma!- be explained blurinar!- excretion of the chromolipid, as it \\as detectable in the urine aiready after 6 h of oxygen exposure, and the amount excreted correlated with the length of the hyperosic period (Fig. 5 ) . Yo efkcts due to hpperoaia could be detected in the amount of d i m e conjugation of the lung,
r
(d) Kidney
lOOr
80C
..i
,/ u . 0 12 24 36 48
Hyperoxia (h)
Fig. 4. The amount of fluorescent chromolipids in rat tissues after exposure to
IOOO,, oxygen. Each point represents the mean fSEW1 from six animals.
P < 0.01. brain or kidney. Only in the case of the liver was there an effect, an increase by 3 9 O 4 after 48 h hpperoxia (data not shown). I n brain tissue, a decline of superoxide dismutase activity was detected at the first timepoint studied. When measured after 36 or 48 h h!-peroxia, however, slightly increased levels of the actil-ity (30°,, P < 0.05, and 359;, I-' < 0.01, respectively) were seen (data not shown). I n the case of the lung, quantitation of the effects of h!-peroxia was hampered by methodological difficulties. T h e level of superoxide dismutase
Oxidative damage in rat tissues
A
IlOC
900 a,
w
155
LUNG
' 3 700
m
t-
8 500 9 300 L
I
I
I
I
0
12
24
36
48
Time (h) 40
T
L 0
1
6
12
02 exposure (h)
2o
k
LIVER
t
0 12 24 38 46 Fig. 5. Chloroform-methanol extractable fluorescent material excreted in rat urine after exposure to lOOyo Time (h) oxygen. Urine was collected during the 10 h period Fig. 6. Catalase (CAT) activity in lung and liver immediately after oxygen exposure. Each point tissues of rats exposed to 100% oxygen. Each point represents the mean SEM from four animals. represents the mean fSEM from six animals.
* P < 0.05; **P < 0.01. activity in control lung was 92.8k10.5 (mean & SEM) p g SOD g-l tissue, and after 12 h hyperoxia this activity was decreased by 31 yo (data not shown). However, when analyses were performed from lung homogenates of rats exposed for 36 or 48 h to hyperoxia, no scavenging of the superoxide anion could be detected but, on the contrary, the oxidation of adrenaline was increased. This effect was probably due to the appearance in lung tissue of reactive substances capable of oxidizing adrenaline. Regardless of this, the fact remains that the enzymatic superoxide anion detoxifying capacity in lungs of rats was severely impaired after prolonged hyperoxia. Hyperoxia did not affect superoxide dismutase activity in the liver (data not shown). Hyperoxia brought about a drastic decline in pulmonary and hepatic catalase activities, and the activities remained decreased to the end of the hyperoxic exposure period (Fig. 6). The
effect was more pronounced in the lung than in the liver. DISCUSSION The objective of the present study was to obtain more detailed information on mechanisms of hyperoxia-induced oxidative damage to rat tissues. The hyperoxia period was sufficient to produce lung damage tb rats, as demonstrated by increased wet wt/dry wt ratio of the tissue. Another tissue profoundly affected by hyperoxia was the brain. The brain was the first tissue to show oedema formation, already after 12 h exposure, followed by pulmonary oedema formation after 36 h of exposure. Contrary to the lungs, the weight gain of the brain was not due to accumulation of fluid, as the protein content of the brain did not decrease with increasing organ weight.
156
M. Ahotupa et al.
With the exception of the liver, hyperoxia did not affect the level of diene conjugation in rat tissues. This result is in agreement with earlier reports where lipid peroxidation was measured by the thiobarbituric acid method (Nishiki et a / . 1976, Freeman et a / . 1982). T h e apparent lack of lipid peroxidation would suggest that the reactive oxygen species generated by hyperoxia would exert their effects without involvement of lipids. This conclusion is, however, contradicted by results of the present study showing hyperoxiadependent accumulation of fluorescent chromolipids in rat tissues and, moreover, by increased urinary excretion of fluorescent chromolipids due to hyperoxia. OH-ing to the high sensiti\-ity, fluorescent chromolipid assay has been one of the commonly used methods to quantitate lipid peroxidation in tissues (Tappel 1980, Tsuchida et ai. 1985). Fluorescent compounds with similar spectral characteristics have been detected in lipid extracts of age pigment (lipofuscin) (Tappel 1980). high concentrations of which have been found e.g. in the brains of patients with -4lzheimer’s disease or Down’s syndrome (see Cohen 1985). In the present study the hyperoxiacaused accumulation of fluorescent chromolipids was strikingly prominent in the brain, but was also evident in the lung and kidney. T h e tendency of the brain to accumulate fluorescent chromolipids may indicate higher susceptibilitl- to oxidative stress-caused effects, or alternativell- a poor ability to eliminate these lipid peroxidation products. Although the kidney was macroscopicall!resistant to oxygen damage, the concentration of fluorescent chromolipids in the kidney increased, and even more rapidly than in the brain or lungs. This was most probably due to excretion of these chromolipids from the body via the kidneys, as shown by the urinary excretion of these chromolipids. T h e fact that the urinary excretion of fluorescent chromolipids increased after hvperoxia treatment, and was proportional to the length of hyperoxic challenge, indeed, indicates the existence of physiological mechanism(s) for the elimination of these substances from the body. A most prominent effect caused by hyperoxia was the inactivation of the enzymes detoxifying oxygen-free radicals. In accordance, it has been demonstrated earlier that normobaric hyperoxia (6+100°, 0, atmosphere) inactivates the pul-
monary Cu/Zn-superoxide dismutase activity (Januszkiewicz et al. 1986), whereas prolonged exposure to sublethal hyperoxia results in induction of the above and also other antioxidant enzymes (see Fridovich & Freeman 1986). Hyperoxia-caused inactivation of a multitude of enzymes with important metabolic functions is well documented and has been suggested to play a crucial role in the mechanism of oxygen toxicity (Haugaard 1968). Oxygen-free radicals are implicated not only in hyperoxia-induced tissue damage, but also in radiation injury (Gerschman et al. 1954), and are also the ultimate damaging species in toxic processes caused by several different chemicals (see Smith 1986). Similarly to hyperoxia, we have shown that enzymatic antioxidant functions are impaired also upon exposure to other causes of oxidative stress, such as UV-irradiation (Punnonen et al. 1991) or carcinogenic chemicals (Ahotupa et al. 1987). Impaired defence against reactive oxygen favours the formation of cellular pro-oxidant states, and may thus be implicated in the further development of oxidative damage. In short, the present study clearly demonstrated that exposure to hyperoxia leads to accumulation of fluorescent chromolipids in rat tissues. As the fluorescent chromolipids are formed in reactions of peroxidized lipids with phospholipids, these observations confirm the involvement of lipid peroxidation in tissue damage due to hyperoxia. Lung, brain, liver and kidney differ from each other in their susceptibility to hyperoxia-induced oxidative damage and, moreover, there seem to exist physiological mechanisms for the elimination of fluorescent chromolipids from the body via urinary excretion. T h e decreased superoxide dismutase and catalase activities most probably contribute to the overall pro-oxidative effects caused by hyperoxia. The technical assistance of Mrs Leena Soderholm and hlrs Raija Soderholm is gratefully acknowledged.The stud! was financially supported by the Academy of Finland, the Finnish Cancer Foundation and the Juho 1-ainio Foundation. Finland.
REFERENCES
M.,BLSS4CCHINI-GRIOT, v., BEREZIAT, J -C., C 4 u q .4.51., & BARTSCH, H. 1987. Rapid oxidathe stress induced by N-nitrosoamines. Brochem Biophys Res Commun 146, 1047-10.54
.\HOTLP4,
Oxidative damage in rat tissues BECKER, N.H. & GALVIN, J.F. 1962. Effect of oxygenrich atmospheres on cerebral lipid peroxides. Aerospace Med 33, 985-987. BEERS,B. & SIZER, W. 1952. A spectrophotometric method for measuring the breakdown of hydrogen peroxide by catalase. 3 Biol Chem 195, 133-139. COHEN,G. 1985. Oxidative stress in the nervous system. In: H. Sies (ed.), Oxidative Stress, pp. 383402. Academic Press, New York. CORONGIU, F., LAI,M. & MILIA,A. 1983. Carbon tetrachloride, bromotrichloromethane and ethanol acute intoxication. Biochem 3 212, 625-631. CRAPO, J.D. 1986. Morphologic changes in pulmonary oxygen toxicity. Ann Rev Physiol48, 721-731. ESTERBAUER, H., KOLLER,E., SLEE, R.G. & KOSTER, J.F. 1986. Possible involvement of the lipid peroxidation product 4-hydroxynonenal in the formation of fluorescent chromolipids. Biochem 3 239, 405409. FREEMAN, B.A. & CRAPO,J.D. 1981. Hyperoxia increases oxygen radical production in rat lungs and lung mitochondria. 3Biol Chem 256, 10986-10992. FREEMAN, B.A., TOPOLSKY, M.K. & CRAPO,J.D. 1982. Hyperoxia increases oxygen radical production in rat lung homogenates. Arch Biochem Biophys 216, 477484. FRIDOVICH, I. 1970. Quantitative aspects of the production of superoxide anion radical by milk xanthine oxidase. 3 Biol Chem 245, 40534057. FRIDOVICH, I. & FREEMAN, B. 1986. Antioxidant defenses in the lung. Ann Rev Physiol48, 693-702. GERSCHMAN, R., GILBERT,D.L., NYE,S., DWYER, W. & FENN,W.O. 1954. Oxygen poisoning and xirradiation: a mechanism in common. Science 119, 623-626. HAUGAARD, N. 1968. Cellular mechanisms of oxygen toxicity. Physiol Rev 48, 31 1-373. JANUSZKIEWICZ, A.J., HUNTRAKOON, M., WILSON,P.K. & FAIMAN, M.D. 1986. Isolated perfused lung histamine release, lipid peroxidation, and tissue superoxide dismutase from rats exposed to normobaric hyperoxia. Toxicology 39, 3746. KLEIN, J. 1990. Normobaric pulmonary oxygen toxicity. Anest Analg 70, 195-207. KOSTRUCHA, J. & KAPPUS,H. 1986. Inverse re-
157
lationship of ethane or n-pentane and malondialdehyde formed during lipid peroxidation in rat liver microsomes with different oxygen concentrations. Biochim Biophys Acta 879, 120-125. LAYNE, E.K. 1957. Spectrophotometric and turbidimetric methods for measuring protein. Methods Enzymol 3, 447454. I. 1972. The role of MISRA,H.P. & FRIDOVICH, superoxide anion in the autoxidation of epinephrine and a simple assay for superoxide dismutase. 3 Biol Chem 247, 3170-3175. H. & KAPPUS,H. 1983. Ferrous ionMULIAWAN, stimulated alkane expiration in rats treated with carbon tetrachloride. Toxicology 28, 29-36. N. & CHANCE, B. NISHIKI, K., JAMIESON, D., OSHINO, 1976. Oxygen toxicity in the perfused rat liver and lung under hyperbaric conditions. Biochem 3 160, 343-3 5 5. PORTER,N.A. 1984. Chemistry of lipid peroxidation. Methods Enzymol 105, 273-282. PUNNONEN,K., PUNTALA,A., JANSEN, C.T. & AHOTUPA, M. 1991. UVB Irradiation induces lipid peroxidation and decreases antioxidant enzyme activities in human keratinocytes in vitro. Acta Dermato- Venereol71, 239-273. SMITH,L.L. 1986. The response of the lung to foreign compounds that produce free radicals. Ann Rev Physiol48, 681-692. SMITH,M.T., THOR, H., HARTZELL, P. & ORRENIUS, S. 1982. The measurement of lipid peroxidation in isolated hepatocytes. Biochem Pharmacol31, 19-26. TAPPEL, A.L. 1980. Measurement and protection from in vivo lipid peroxidation. In: W.A. Pryor (ed.), Free Radicals in Biology vol. 4, pp. 1117. Academic Press, New York. TSUCHIDA, M., MIURA,T., MIZUTANI, K. & AIBARA, K. 1985. Fluorescent substances in mouse and human sera as a parameter of in vivo lipid peroxidation. Biochim Biophys Acta 834, 196-204. J. 1974. VALIMAKI,M., KIVISAARI, J. & NIINIKOSKI, Permeability of alveolar capillary membrane in oxygen poisoning. Aerospace Med 45, 370-374. P., SYLVESTER, ZWEIER, J.L., DUKE,S.S., KUPPUSAMY, J.T. & GABRIELSON, E.W. 1989. Electron paramagnetic resonance evidence that cellular oxygen toxicity is caused by the generation of superoxide and hydroxyl free radicals. FEBS Lett 252, 12-16.
