JOURNAL OF BIOLUMINESCENCE AND CHEMILUMINESCENCE VOL 7 123-132 (1992)

Effect of Antioxidants on Chemiluminescence Produced by Reactive Oxygen Species Carlos Pascual* and Cheyla Romay Centro Nacional de lnvestigaciones Cientificas, Apartado Postal 6990, Cubanachn, Playa, La Habana, Cuba

Luminol chemiluminescence was used t o evaluate the scavenging of superoxide, hydroxyl and alkoxy radicals by four antioxidants: dipyridamole, diethyldithiocarbamic acid, ( + )catechin, and ascorbic acid. Different concentrations o f these compounds were compared w i t h well-known oxygen radical scavengers i n their capacity t o inhibit the chemiluminescence produced in the reaction between luminol and specific oxygen radicals. Hydroxyl radicals were generated using the Fenton reaction and these produced chemiluminescence which was inhibited by diethyldithiocarbamate. Alkoxy radicals were generated using the reaction o f tert-butyl hydroperoxide and ferrous ion and produced chemiluminescence which was inhibited equally by a l l o f the compounds tested. For the determination o f superoxide scavengers w e describe a new. simple, economic, and rapid chemiluminescence method consisting o f the reaction between luminol and horseradish peroxidase (HRP). With this method it was found that 40 nmol/l dipyridamole, 0.18 pmol/l ascorbic acid, 0.23 pmol/l (+)catechin, and 3 pmol/l diethyldithiocarbamic acid are equivalent t o 3.9 ng/ml superoxide dismutase (specific scavenger of superoxide) i n causing the same degree o f chemiluminescence inhibition. These results not only indicated that the antioxidative properties of these compounds showed different degrees o f effectiveness against a particular radical but also that they may exert their action against more than one radical. Keywords: Chemiluminescence; oxygen radicals; reactive oxygen species; ascorbic acid; dipyridamole; diethyldithiocarbamate; catechin; vitamin E

INTRO DUCTlON

Superoxide, hydroxyl radicals and alkoxy radicals can be considered to be the most important reactive oxygen species produced in biological systems by diverse metabolic pathways including microsomes, mitochondria1electron transport, and lipooxygenasesand cycloxygenasesin eicosanoidmetabolism. Other important sources or reactive oxygen species are the reaction catalysed by xantine oxidase in the conversion of hypoxanthine to xanthine and uric acid, as well as the activation of NADPH

oxidase of polymorphonuclear cells during superoxide production and the process of phagocytosis (Colepicolo et al., 1990; Uyama et al., 1990). Singlet oxygen appears to be involved in the non-enzymatic destruction of catecholamines (Kruck et al., 1989) as well as in other oxidative processes (Patterson et al., 1990). increased generation of oxygen radicals in biological systems has been observed in tissue injuries due to infection, inflammation, ulcers, cancers, radiation, rheumatoid arthritis, mechanical trauma and burns, as well as ischaemia and reperfusion

* Author for correspondence 0884-3996/92/020123- 10W5.00 0 1992 by John Wiley SL Sons, Ltd.

Received I2 July 1991 Revised 12 October 1991

124

CARLOS PASCUAL AND CHEYLA ROMAY

(Basaga, 1990). This usually leads to peroxidation of unsaturated fatty acids in cell membranes causing decreased fluidity, inactivation of membranebound receptors and enzymes, and consequently alteration of cell permeability and integrity (Gutteridge and Halliwell, 1990). There is increasing interest in determining the radical scavenging properties of different agents (Kanofsky, 1990; Roussean-Richard et al., 1990; Tosaki et al., 1990) and various methods for evaluating compounds for the ability to suppress radical formation or destroy radicals have been proposed (Rao et al., 1988; Salaris and Babbs, 1988). Lurninol chemiluminescence may constitute a simple and general method for the determination of different oxygen radicals (Rao et al., 1988). This method is based upon the principle that oxygen radicals react with luminol, giving rise to chemiluminescence. In the present work we used the Fenton reaction to generate hydroxyl radicals (HO * ) and the reaction of tert-butyl hydroperoxide (t-BOOH) with ferrous ion was used for the formation of alkoxy radicals (RO * ). Superoxide radical was formed by the hypoxanthine-xanthine oxidase reaction and also by a very simple, economic, and rapid reaction between luminol and horseradish peroxidase (HRP). Identification of the radical produced can be easily determined by addition of specific scavengers, and this served as a control for confirming that only one main radical species was formed. Therefore the inhibition of chemiluminescence produced from such a reaction is the rationale used for evaluating, in a simple manner, the in vitro scavenging capacity of compounds of interest. These methods were used to evaluate substances with antioxidant effects based upon different chemical pathways: ascorbic acid as a reductant, diethyldithiocarbamic acid as a chelator, and ( + )catechin, and dipyridamole as superoxide sequestering agents. Results obtained provide useful information concerning the extent to which a given antioxidant may act as a scavenger of one or more radicals.

tained from Sigma Chemical Co. (St Louis, MO). Xanthine oxidase (1 U/mg) from cow milk and horseradish peroxidase (HRP) (1 85 U/mg) were obtained from Boehringer Mannheim GmbH (Indianapolis, IN), p-iodophenol was obtained from Aldrich Chemical Co. (Milwaukee, WI). Superoxide dismutase (SOD) (3333 U/mg) from bovine erythrocytes and hypoxanthine were obtained from Serva Feinbiochimica GmBH. Dimethylsulphoxide (DMSO), hydrogen peroxide (H202), ethylenediaminetetracetic acid (EDTA) disodium salt, sodium dodecylsulphate (SDS), sodium carbonate, and potassium dihydrogen orthophosphate were from BDH Ltd (Poole, UK). Ferrous sulphate and ascorbic acid were from Merck (Rahway, NJ). Ethanol and Vitamin E (1.1 units/mg) were from Fluka Biochemica (Hauppauge, NY). All other reagents were of analytical grade.

MATERIALS AND METHODS

It was found that addition of HRP to a buffered solution of luminol produces a chemiluminescence signal which is abolished by small amounts of superoxide radical scavengers. The reaction mixture consisted of: 0.1 mol/l glycine-NaOH buffer, pH 9.4, containing 1 mmol/l

Reagents

Luminol, catalase (2890 U/mg), glycine, tert-butyl hydroperoxide, and selenomethionine were ob-

Test compounds

Dipyridamole was obtained from Sigma. (+)Catechin and diethyldithiocarbamic acid Na salt were from Serva. These compounds were dissolved in distilled water, except vitamin E and dipyridamole which were dissolved in absolute ethanol and ethanol 50 % (w/v), respectively. ChemiIum inescence measurements

The chemiluminescence intensity was measured in a LKB Wallac 1250 luminometer coupled to a potentiometric recorder and the chemiluminescence signal was measured in millivolts. All the reactions were performed at room temperature. Chemiluminescence produced from the reaction of luminol with horseradish peroxidase as a simple method for generation of superoxide anion and evaluation of superoxide radical scavengers

125

EFFECT OF ANTIOXIDANTS ON CHEMILUMINESCENCE

mV

A

I

mv

30.

B

,

mv

C

61

Figure 1. Time course of chemiluminescence produced by reactions generating oxygen free radicals and its inhibition by specific scavengers. (A)Lurninol-HRP reaction with formation of superoxide radical: without (left) and with addition (right) of SOD (1 5 ng/rnl). (B) Fenton reaction with formation of hydroxyl radicals: without (left) and with addition (right) of DMSO (10.5 mmol/l). ( C ) Reaction of t-BOOH with ferrous ion for evaluation of alkoxy radical scavengers. Without (left) and with addition (right) of vitamin E

EDTA and 62.5 pmol/l luminol. 10 pl of distilled water was added to 0.8 ml of this solution in the cuvette. The reaction was started by the addition of 10 pl HRP (4.6 U/ml) dissolved in 50 mmol/l phosphate buffer pH 7.6 and the peak value of the chemiluminescence signal (mV) was registered (Fig. 1A) immediately after rapid mixing. The same procedure was used to determine the scavenging action of an agent, except that 10 pl of the compound to be tested is substituted in the step where distilled water is added. SOD was used in this system as a control of the method (Fig. 1A). Chemiluminescenceintensity was plotted against different concentrations of the compound examined and the concentration which caused 50 % inhibition of chemiluminescenceintensity was compared with the superoxide scavenging activity of these substances. Results are expressed as percentage of chemiluminescence inhibition, I , caused by a specified concentration of the compound. I

= (CL,/CLo - 1 )

x 100

Chemiluminescence produced by superoxide generated from the hypoxanthine-xanthine oxidase reaction for evaluation of superoxide radical scavengers

This method was described in detail previously (Pascual et al., 1991). In brief, the reaction mixture consisted of 68 mmol/l glycine buffer pH 8.6, 10 pmol/l luminol and 5 pmol/l p-iodophenol. 10 p1 of distilled water or a solution of the compound to be tested was added. Then 2 p1 of xanthine oxidase (2 U/ml) was added and the reaction initiated with 10 pl 1 mmol/l hypoxanthine. The intensity of the chemiluminescence was recorded and the results were calculated as indicated in the previous method. Chemilurninescence produced in the Fenton reaction for evaluating hydroxyl radical scavengers

Hydroxyl radicals were generated from the Fenton reaction:

where CL, and CL, are the peak chemiluminesFe2+ + H,O, -+Fe3+ + OH- HO cence value in the presence or absence of the test compound. The same procedure for expressing results was The method used was similar to the one described by Rao et al. (1988) with minor modifications. used for all the methods employed.

+

-

126

CARLOS PASCUAL AND CHEYLA ROMAY

Reactions of luminol with hydroxyl radicals produces a chemiluminescence signal which is inhibited by specific radical scavengers. The reaction mixture in the cuvette consisted of 0.8ml 50 mmol/l K-phosphate buffer pH 7.8; 2 mmol/l EDTA and 0.1 mmol/l luminol. To this 10 pl of distilled water was added and mixed. Then 5 p1 of 6 mmol/l H,O, was also added and mixed, The reaction was started with 10 p1 of 20 mmol/l Fe,SO, and immediately, after rapid mixing, the chemiluminescence signal (mV) was registered (Fig. 1B). The same procedure was used to determine the scavenging action of an agent except that lop1 of a solution of the test compound was substituted for the distilled water. Ethanol and DMSO were used as specific scavengers of hydroxyl radicals as a control of the method. The results are calculated using the same procedure described for the luminol/HRP system.

Chemiluminescence produced by the reaction of tert-butyl hydroperoxide with ferrous ion for evaluation of alkoxy radical scavengers

RESULTS

In previous work (Pascual et aE. 1991) we found that dipyridamole, ( + )catechin, and diethyldithiocarbamate inhibit superoxide chemiluminescence produced by the hypoxanthine-xanthine oxidase reaction. In the present work, we included ascorbate because of its well-known antioxidant activity and the results show its inhibitory action in this system (Fig. 2). Using these compounds we have evaluated a simple method for the production of superoxide with the finding that a reaction between luminol and horseradish peroxidase produces a significant chemiluminescent signal which can be effectively suppressed by superoxide dismutase and in much less degree by catalase (Fig. 3). Addition of several known sequestering agents which exert their action against radicals other than superoxide did not significantAy reduce chemiluminescence (Table 1). When the effects of different compounds were tested (Fig. 4) and compared with SOD as a superoxide radical scavenger, it was determined that

mV

The alkoxy radical (RO * ) was generated from the reaction of tert-butyl hydroperoxide with ferrous ion as described by Fukuzawa et al., (1988): R O O H + F e 2 + - + R O - + F e 3 ++ O H Reaction of luminol with alkoxy radicals produces a chemiluminescent signal which can be inhibited by specific radical scavengers. The reaction mixture consisted of: 0.8 ml of 50 mmol/l glycine buffer pH 8.6, 50 mmol/l SDS, 0.025 mmol/l luminol. 10 p1 of distilled water was added and mixed prior to the addition and mixing of 5 pl.of 1.5mmol/l t-BOOH. The reaction was started with 10 pl of 0.4 mmol/l Fe,SO, and immediately after rapid mixing the chemiluminescence signal (mV) was recorded (Fig. 1C ) . In order to determine the scavenging action of a compound, the same procedure was used, except that 10 pl of a solution of the agent was substituted for distilled water. Vitamin E which is a well-known scavenger of peroxyl radicals was used in this system as a positive control of the method. The results are calculated as described for the luminol/ HRP test systems.

c 0

1

2

pmol/ 1 Figure 2. Ascorbate inhibition of chemiluminescence pro duced by the hypoxanthine xanthine oxidase reaction The peak value of chemilurninescence intensity is indicated on the ordinate. on the abscissa the concentration of ascorbate Each point represents a mean of three determinations Verti cal bars represent 1 standard deviation

127

EFFECT OF ANTlOXl DANTS ON CH EM I LUM I NESCENCE

. I

I

2 .o

4,O

I

I

0.05 urn[-‘

0,O25

0

0

U mbl

Figure 3. Effect of SOD and catalase on chemiluminescence produced by products of the reaction of luminol with HRP. The peak value of the chemiluminescence intensity is indicated on the ordinate. Curve on the left: SOD. Curve on the right: catalase Abscissa represents enzyme units in the reaction mixture. Each point represent a mean of four determinations. Vertical bars represent 1 standard deviation

Table 1. Effect of scavengers on the chemiluminescence produced in different oxygen radical generating systems Lurninol-sensitized chemiluminescence signal from

Scavenger None SOD (15 ng/ml) DMSO Vitamin E Ethanol Selenomethionine Histidine (0.35rnmol/l)

Scavenger substrate -

superoxide hydroxyl radicals alkoxy radicals hydroxyl radicals alkoxy radicals singlet oxygen

Superoxide

Hydroxyl radicals

Alkoxy radicals

30.7f 2.95 4.6f 0.25

36.2k 1.40 34.3f 1.07

57.8 2.61 57.3f 7.05

4.21k 1.27

55.52 3.0 (1 75 mmol/l) 9.3f 0.62 (7.5pmol/l) 47.3k 2.85 (200mrnol/l) 21.8f 1.31 (31.3prnol/l) 47.0 1.51

30.5f 0.24 (1 75 mmol/l) 24.4f 2.7 (7.5pmol/l) 30.9f 0.56 (200mmol/l) 32.3f 1.52 (62.5pnol/l) 32.7& 0.52

(1 7.5 mmol/l) 35.8& 0.69

(75pmol/l)” 16.0& 3.0 (50mmol/l) 37.9f 3.9 (62.5pmol/l) 33.2 1.77

Values are expressed in millivolts; X f 1 standard deviation ( n = 4) and represent peak chemiluminescence intensity. Concentrations of scavengers are in parentheses. Superoxide was generated with the luminol-peroxidase system. Hydroxyl radicals were generated with the Fenton reaction. Alkoxy radicals were generated from the reaction of tea-butyl hydroperoxide with ferrous ion. a Since Vitamin E was dissolved in ethanol, a correction was performed by subtracting inhibitory effect caused by ethanol.

128

CARLOS PASCUAL AND CHEYLA ROMAY

*" I

A

c

I

1

€3

I

1

Figure 4. Effect of test compounds on chemiluminescence intensity produced by products of the reaction of luminol with HRP for evaluation of superoxide radical scavengers. (A) ascorbic acid; (B) dipyridamole; (C) ( + ) catechin and (D) diethyldithiocarbamic acid. Peak values of chemiluminescence intensity are represented on the ordinate. Abscissa indicates the concentration of these compounds. Each point represents mean of four determinations; vertical bars represent 1 standard deviation

0.013 U/ml(3.9 ng/ml) of SOD, which inhibits 50% of the chemiluminescence in this system, corresponds to approximately 40 nmol/l dipyridamole, 0.18 pmol/l ascorbic acid, 0.23 pmol/l ( + )catechin and 3 pmol/l diethyldithiocarbamic acid. In order to determine whether or not these compounds also have scavenging capacity for other

oxygen radicals, we studied the effect on hydroxyl and alkoxy radical generating reactions. Hydroxyl radicals generated with the Fenton reaction produced chemiluminescence proportional to the concentration of added hydrogen peroxide (Fig. 5). When the antioxidant compounds were tested in this system it was found that 50 'X inhibi-

129

EFFECT OF ANTIOXIDANTS ON CHEMILUMINESCENCE mv 0

20

40 I

60 prnol/l I

H,O,

i

mv 100

so

0 0

12

24 pmol/l

Figure 6. Effect of compounds on chemiluminescence produced by the hydroxyl radical of the Fenton reaction. The value of the chemiluminescence intensity i s indicated on the 0 9 18 rnmol/l DMSO ordinate. The final concentration of the following comFigure 5. Chernilurninescence produced with the Fenton pounds is indicated on the abscissa. ( 0 )(+)catechin; (0) reaction for evaluation of hydroxyl radical scavenging. (0)ascorbate: ( A ) diethyldithiocarbamate; (0) dipyridamole. Effect of H,O,. The abscissa in the upper part represents the The abscissa represented on the upper part of the figure concentration of H,O, and the ordinate at the right repre- indicates the final concentration of (A)ethanol. Each point sents the corresponding values of chemiluminescence inten- represents the mean of three determinations. Vertical bars sity. ( 0 )Effect of DMSO. The lower abscissa represents the represent 1 standard deviation concentration of DMSO and left ordinate represents the corresponding chemiluminescence intensity. 0

Selenomethionine was also tested since some organoselenium compounds have been reported to tion of chemiluminescence was achieved with 14 be scavengers for this type of radical (Narayanaspmol/l diethyldithiocarbamic acid (Fig. 6), equiva- wami and Sies, 1990). Inhibition of chemilumineslent to 35 mmol/l ethanol or 3.9 mmol/l DMSO cence by selenomethionine is shown in Table 1. which are specific scavengers for hydroxyl radicals. In order to know if the scavenger has action on No other compounds produced such a significant more than one type of radical, or if the reaction action in this system. Since the apparent inhibition produced more than one type of radical, each of caused by dipyridamole is mainly due to the eth- these scavengers was tested in each system. Results anol in which it was dissolved, correction revealed in Table 1 indicate that these scavengers are selecthat even at the highest concentration of dipyrida- tive against one type of free radical generating mole used (24 pmol/l) chemiluminescence could reaction and this serves as a control for the specifionly be suppressed by 23 %; ascorbate produced a city of the system. Therefore, the method could be similar inhibition as dipyridamole alone, while used for the in vitro determination of scavengers against free radicals. (+)catechin did not cause any inhibition. Chemiluminescence produced with the alkoxy radical depended on t-BOOH concentration and could be inhibited effectively by vitamin E which DISCUSSION was used as a scavenger for this type of radical (Fig. 7). From results of the inhibition curves, it can These luminol chemiluminescence methods reprebe determined that 0.48 pmol/l dipyridamole, 0.48 sent a useful tool for the in vitro evaluation of pmol/l diethyldithiocarbamic acid, 0.6 pmol/l oxygen radical scavengers. The difficulties due to (+)catechin and 7.2 pmol/l ascorbic acid caused interferences that can arise through the use of the same effect as 0.15 prnol/l vitamin E with biological samples do not occur in our system, since respect to inhibiting chemiluminescence by 50 % these determinations are performed under welldefined in vitro conditions with relatively pure (Fig. 8).

CARLOS PASCUAL AND CHEYLA ROMAY

130

mV

0

I

8 pmol/L t-BOOH

4

1

1

0

0,1

0.3

0.5

47 ,urnol/l VITAMlN E

Figure 7 . Chernilurninescence produced by alkoxy radicals generated by the reaction of tert-butyl hydroperoxide with ferrous ions (0) effect of text-butyl hydroperoxide The abscissa on the upper part represents the final concentration of tert-butyl hydroperoxide The abscissa on the lower part represents the final concentration of vitamin E ( 0 ) Chernilurninescence intensity is indicated on the ordinate Each point represents the mean of four determinations. vertical bars represent 1 standard deviatiori

substances. By selecting the appropriate reaction controlled with specific well-known scavengers, it was demonstrated (Table 1) that each reaction generated the desired radical which reacts with luminol. Particularly interesting is the luminol peroxidase chemiluminescence reaction which represents a simple, rapid, and economic method for evaluating superoxide scavengers. Superoxide and hydrogen peroxide are the two species produced since the chemiluminescence could be effectively inhibited by superoxide dismutase and catalase but not by specific scavengers of hydroxyl radicals, alkoxy radicals, or singlet oxygen. The chemiluminescence produced by this reaction can be explained by the mechanism postulated by Merenyi et al. (1990) in which HRP reacts with luminol yielding free luminol radicals. Reaction between 0, and the luminol radical generates superoxide (Lock et al.,

1988) and the luminol radical is oxidized to a diazoquinone. The efficiency of an agent as a scavenger was easily determined using the luminol-sensitized chemiluminescence method from the degree of chemiluminescence inhibition caused by a specific concentration of the compound with respect to a control. Also, the different compounds tested were compared with the well-known oxygen radical scavengers; SOD, vitamin E, and DMSO. These results not only indicated that the antioxidant properties of these compounds showed different degrees of effectiveness against a particular radical, but also that they may exert their action against more than one radical. For example, dipyridamole and (+)catechin, which are known exclusively, as superoxide scavengers, were found to have significant effects, against alkoxy radicals. Ascorbic acid was also found to be very effective

131

EFFECT OF ANTIOXIDANTS ON CHEMILUMINESCENCE

0

1.2

2,4 pmol/L

Figure 8. Effect of compounds on the cherniluminescence produced by alkoxy radicals generated by the reaction of tert-butyl : (+)catechin (a):dipyridamole (0);and diethyldithiocarbamic hydroperoxide with ferrous ion. Effect of ascorbic acid (0) acid (m). The abscissa represented on the upper part indicates the final concentration of ascorbic acid in the test. Chemilurninescence intensity is indicated on the ordinate. Each point represents the mean of three determinations. Vertical bars represent 1 standard deviation

against superoxide but to a lesser degree than others against alkoxy radicals. Diethyldithiocarbamate exerted its action against superoxide and alkoxy radical, and was the only substance able to cause significant inhibition of hydroxyl radicals. The antioxidant properties of the substances evaluated are based upon different chemical pathways in their mechanism of action. For example, the inhibitory action of diethyldithiocarbamate on both alkoxy and hydroxyl radicals may be due to its known effect as an iron chelator (Spath and Tempei, 1987). Many flavonoids with structures similar to (+)catechin have been reported to scavenge superoxide (Hart et al., 1990). Vitamin E is regarded as the major hydroperoxyl or hydroperoxide radical trapping lipid soluble chain breaking antioxidant (Farris, 1990) preventing oxidative attack on membrane lipids. Ascorbate has been implicated as an antioxidant by scavenging peroxyl

radicals in the aqueous phase of cells (Richards et al., 1990), b.y scavenging superoxide (Hayase et al., 1989), or by restoring reduced vitamin E. However, the antioxidant activity of redox compounds such as ascorbic acid must be evaluated with caution. A paradoxical response as an oxidant has been reported in microsomal lipid peroxidation (Wefers and Sies, 1988) and in the presence of transition metal ions (Miller and Aust, 1989), and more recently, in the haemoglobjn-mediated oxidative damage to the central nervous system (Prat and Turrens, 1990). We suggest that even though the luminol-sensitized chemiluminescence method may have limitations, owing to the formation of small amounts of other radical species, it constitutes a simple, rapid, and economic procedure. It could serve as a screening test for the antioxidant properties of substances and in identifying new compounds which act as

132

CARLOS PASCUAL AND CHEYLA ROMAY

biological radical scavengers. These results could Merenyi, G., Lind, J. and Eriksen, T. E. (1990). Luminol chemiluminescence: chemistry, excitation, emitter. J. Biolube of importance, since great effort is being made in min. Chemilumin., 5, 53-56. evaluating radical scavengers (Parnham et a!., Miller, D. M. and Aust, S . D. (1989). Studies of ascorbate1991; Sjoquist et al., 1991) such as SOD for the dependent iron catalysed lipid peroxidation. Arch. Biochem. treatment of certain diseases. Compounds such as Biophys., 271, 113-119. dipyridamolecould act as an efficient substitute for Narayanaswami, V. and Sies, H. (1990). Antioxidant activity of Ebselen and related selenorganic compounds in microsomal biological radical scavengers in the treatment of lipid peroxidation. Free Radic. Res. Commun., 10,234-237. diseases where oxygen radicals are involved. Parnham, M. J. Leyck, G., Graf, E., Dowling, E. J. and Blake, D. Acknowledgements We are grateful to Julio Armesto for typing this paper, and t o Professor Dr J.R.J. Sorenson of the University of Arkansas College of Pharmacy for his useful suggestions and t o Professor Dr Juan Josi? Arag6n of the University Autonoma d e Madrid for his discussion a n d criticism of the manuscript.

REFERENCES Basaga, H. S. (1990). Biochemical aspects of free radicals. Biochem. Cell Biol., 68,989-998. Colepicolo, P., Camarero, V. C. P. C., Nicolas, M. T., Bassot, J. M., Kamovsky, M. L. and Hastings, J. W. (1990). A sensitive and specific assay superoxide anion released by neutrophils or macrophages based on bioluminescence of polynoidin. Anal. Biochem., 184, 369-374. Farris, M. W. (1990). Oxygen toxicity: unique cytoprotective properties of vitamin E succinate in hepatocytes. Free Radic. Biol. Med., 9, 333-334. Fukuzawd, K., Yoshimura, T., Fuju, T., Takauchi, K., Miki, M., Tamai, H. and Mino, M. (1988).Theoretical analysis ofa sitespecific chemiluminescence reaction and its application to quantitation of lipid hydroperoxides. Chem. Phys. Lipids, 48, 197-203. Gutteridge, J. M. C. and Halliwell, B. (1990). The measurement and mechanism of lipid peroxidation in biological systems. TZBS, 15, 129-135. Hart, B. A. T., Theo, R. A. M., Ching, I. V., Van Dijk, H. and Labadie, R. P. (1990). How flavonoids inhibit the generation of luminol-dependent chemiluminescence by activated human neutrophils. Chem. Biol. Interactions., 13, 323-325. Hayase, F., Hirashima, S., Okamoto, G. and Kato, H. (1989). Scavenging of active oxygens by melanoidins. Agric. BioL Chem., 53,3383-3385. Kanofsky, J. R. (1990). Quenching of singlet oxygen by human plasma. Photochem. Photobiol., 51,299-303. Kruck. I., Lechszteld, K., Michalska, T. and Bounias, M. (1989). Evidence for the generation of singlet molecular oxygen during Dopa and Dopamine peroxidation. 2. Naturforsch, 44c, 39-44. Lock, R., Johanson, A., Orsilius, K. and Dahlgren, C. (1988). Analysis of horseradish peroxidase-amplified chemiluminescence produced by human neutrophils reveals a role for the superoxide anion in the light-emitting reaction. Anal. Biochem., 173,450-455.

R. (1991). The pharmacology of ebselen. Agents Actions, 32, 4-9. Pascual, C., Gonzalez, R. and Romay, Ch. (1991). Drugs effects on superoxide generation and chemiluminescence response of human leukocytes. Agents Actions, 32,277-282. Patterson, M. S., Madsen, S. G. and Wilson, B. C. (1990). Experimental tests on the feasibility of singlet oxygen luminescence monitoring in oivo during photodynamic therapy. J. Photochem. Photobiol. B: Biol. 5, 69-84. Prat, A. G. and Turrens, J. F. (1990). Ascorbate and haemoglobin dependent brain chemiluminescence. Free Radic. Biol. Med., 8, 319-325. Rao, P. S., Luber, J. M. Jr., Milinowicz, J., Lolezari, P. and Mueller, H. S. (1988). Specificity of oxygen radical scavengers and assessment of free radical scavenger efficiency using luminol-enhanced chemiluminescence. Biochem. Biophys. Res. Commun., 150, 39-44. Richards, G. A., Theron, A. J., Van Rensburg, E. J., Van Rensburg, A. J., Van der Merive, C. A., Kuyl, J. M. and Anderson, R. (1990). Investigation of the effects of oral administration of vitamin E and beta-carotene on the chemiluminescence responses and the frequency of sister chromatid exchanges in circulating leukocytes from cigarette smokers. Am. Rev. Respir. Dis., 142, 648-654. Rousseau-Richard, C., Auchair, C., Richard, C. and Martin, R. (1990). Free radical scavenging and cytotoxic properties in the ellipticine series. Free Radic. Bid. Med., 8, 223-230. Salaris, S. C. and Babbs, C. F. (1988). A rapid, widely applicable screen for drugs that suppress free radical formation in ischemia/reperfusion. J. Pharmacol. Methods, 20, 335-345. Sjoquist, P. O., Carlsson, L.,Jonason, G., Marklund, S. L. and Abrahamsson, T. (1991). Cardioprotective effects of recombinant human extracellular superoxide dismutase type-C in rat isolated heart subjected to ischemia and reperfusion. J. Cardiooasc. PharmacoL, 17,678-683. Spath, A. and Tempel, K. (1987). Diethyldithiocarbmate inhibits scheduled and unscheduled DNA synthesis of rat thymocytes in uitro and in oioo: dose-effect relationship and mechanism of action. Chem. Biol. Interactions, 64, 151--166. Tosaki, A., Blasig, I. E., Pali, T. and Ebert, B. (1990). Heart protection and radical trapping by DMSO during reperfusion in isolated working rat hearts. Free Radic. Biol. Med., 8, 363-372. Uyama, O., Shiratsuki, N. Matsuyama, T., Nakanishi, T., Matsumoto, Y., Yamada, T., Narita, M. and Sugita, M. (1990). Protective effects of superoxide dismutase on acute reperfusion injury of gerbil brain. Free Radic. Biol. Med., 8,265-268. Wefers, H. and Sies, H. (1988). The protection by ascorbate and glutathione against microsomal lipid peroxidation is dependent on vitamin E. Eur. J. Biochem., 174, 353-357.