Eur. J. Biochem.

47,469-474 (1974)

Involvement of the Superoxide Anion Radical in the Autoxidation

of Pyrogallol and a Convenient Assay for Superoxide Dismutase
Stefan MARKLUND and Gudrun MARKLUND
Department of Chemistry, Section of Physiological Chemistry, University of Umei

(Received March 8 /June 8, 1974)

The autoxidation of pyrogallol was investigated in the presence ofEDTA in the pH range7.9- 10.6.
The rate of autoxidation increases with increasing pH. At pH 7.9 the reaction is inhibited to 99
by superoxide dismutase, indicating an almost total dependence on the participation of the superoxide
anion radical, 02.-, in the reaction. Up to pH 9.1 the reaction is still inhibited to over 90% by
superoxide dismutase, but at higher alkalinity, O,.--independent mechanisms rapidly become
dominant.
Catalase has no effect on the autoxidation but decreases the oxygen consumption by half,
showing that H202is the stable product of oxygen and that H20, is not involved in the autoxidation
mechanism.
A simple and rapid method for the assay of superoxide dismutase is described, based on the
ability of the enzyme to inhibit the autoxidation of pyrogallol.
A plausible explanation is given for the non-competitive part of the inhibition of catechol
0-methyltransferase brought about by pyrogallol.

Pyrogallol (1,2,3-benzenetriol) has long been no saturation kinetics at the O,.- concentrations
known to autoxidize rapidly, especially in alkaline which have been feasible to produce [7] and hence
solution and the reaction has been employed for the induces a pseudo-first order dismutation of O,.-.
removal of oxygen from gases. The present paper describes studies of the autoxida-
Molecular oxygen, carrying two unpaired electrons tion of pyrogallol under various conditions. The role
with parallel spins, has a preference for univalent of 02.-in the reactions was investigated with the aid
reduction because spin restrictions arise when reduc- of superoxide dismutase. The data obtained allotted
tion with electron pairs is attempted [l]. The recently suitable conditions for a convenient assay of super-
discovered enzyme superoxide dismutase [2] extremely oxide dismutase.
rapidly dismutases univalently reduced oxygen 02.-,
+
the superoxide anion radical (2 02.- 2 H + -+ 0,
+ H,O,). The enzyme has proven to be a useful probe
MATERIALS AND METHODS
for studying
- - the participation of the radical in reac-
tions involving oxygen such as autoxidations. Thus All experiments were performed at 25 "C & 0.1.
02.-has been shown to be involved in the autoxida- Pyrogallol (p.a. Merck, A.G.) was purified by sublima-
tion of e.g. sulphite [3], adrenalin [4] and 6-hydroxy- tion. Stock solutions in 10 mM HCI are stable for
dopamine [ 5 ] . weeks. Cacodylic acid, was obtained from British
Bovine Cu-Zn superoxide dismutase (erythro- Drug Houses; Tris from Merck A.G. ; diethylenetri-
cuprein) was the dismutase used in the present in- aminepentaacetic acid from Sigma Chemical Compa-
vestigation. Its activity is essentially independent of ny; EDTA from Riedel-De HaZn A.-G.; NaCN (p.a.)
pH in the range 5.5 to 9.5 [6,7] and it also has con- from Merck, A.G. Water was double distilled in glass
siderable activity at higher pH values. Another proper- vessels and bovine Cu-Zn superoxide dismutase (ery-
ty of relevance in the present context is that it shows throcuprein) was prepared from erythrocytes by the
Enzymcv. Superoxide dismutase (EC 1.15.1.1); catalase method of McCord and Fridovich [21 as modified by
(EC 1.11.1.6); catechol 0-methyltransferase (EC 2.1.1.6). Weser et al. [8].

Eur. J. Bioc,hem.47 (1974)

470 Pyrogallol Autoxidation, Superoxide Dismutase

Crystalline bovine liver catalase was obtained from

Boehringer Mannheim GmbH. The crystals were
washed several times in distilled water before being
dissolved. E~~~ was taken as 324mM-' cm-' [9].
For spectrophotometry a Beckman Acta 111 was used.
-100-
Oxygen consumption was determined polarographical-
ly with a Clark electrode using a Yellow Springs - 90
%
Instruments model 53 biological oxygen monitor. c
-80 =
The different O2 saturations were obtained by 5
._
mixing solutions carefully equilibrated with air, argon - 70
.-U
or pure oxygen. The latter two solutions were cautious- -60
ly stratified under initially added air-equilibrated Q
01

solution before mixing. The procedure was checked -50 2

by the oxygen electrode and appeared to give the l o "
desired composition within 1 or 2 %.

y o
10 r"
RESULTS AND DISCUSSION
Pyrogallol autoxidizes rapidly in aqueous solution :
the faster the higher the pH, and several intermediate
products are apparently formed. Thus the solution
Fig. 1. Effects of p H on the autoxidation of 0.2 mM pyrogallol
first becomes yellow-brown with a spectrum showing in air-equilibrated buffer. Open symbols, 50 mM Tris-caco-
a shoulder between 400 and 425 nm. After a number dylic acid buffer with 1 mM EDTA; filled symbols, 50 mM
of minutes the colour begins to turn green and finally, sodium carbonate buffer with 1 mM EDTA. (A) Rate of
after a few hours, a yellow colour appears. In the autoxidation as determined from the increase in &,,. per
present investigation, the autoxidation was studied min; (0)oxygen consumption, pM x min-' ; (0) maximal
inhibition of autoxidation by superoxide dismutase (> 40 pg
essentially during the first step(s) and the rate was x ml-I); (0) requirement of superoxide dismutase for 50%
taken from the linear increase in absorbance at inhibition. (----) The part of the autoxidation which may
420 nm which is seen for a number of minutes after an be inhibited by superoxide dismutase
induction period of some 10 s. The validity of this
procedure is demonstrated in Fig. 1 by the constant
relationship between increase in absorbance at 420 nm
per min and oxygen consumption. However, this units around pH 8. When pH 9 is approached, the
relationship was not tested at the higher pH values slope of the curve is decreased and a minimum is
because of too slow a response of the oxygen electrode. found around pH 9.5. At higher pH values, the slope
again increases and finally corresponds to a doubling
of the rate per 0.3 pH units. At pH 7.9 the autoxida-
Effect of EDTA tion is inhibited to 99% by superoxide dismutase.
In the absence of EDTA the autoxidation is faster The sensitivity to superoxide dismutase decreases
than in its presence, and it tends to vary in rate and when the pH is increased, but still amounts to 93%
is less affected by superoxide dismutase. In the at pH 9. At higher pH values there is a strong decrease
presence of EDTA the rate is independent of the con- in the sensitivity to superoxide dismutase and at
centration of the chelator (tested in the range 0.1- pH 10.6 the autoxidation is inhibited to only 15%
2 mM), which indicates that the effect of EDTA is by superoxide dismutase. The decrease in inhibition
only due to its binding of traces of metal ions. Most as the pH is raised can hardly be attributed to inactiva-
experiments in the present report were performed in tion of the superoxide dismutase, as it was added in
the presence of 1 mM EDTA. a large excess and addition of more superoxide
dismutase had no further effect.
The results indicate that there are at least two
Effect of p H on the Autoxidation different mechanisms of autoxidation in the pH
Fig. 1 shows the effect of pH on a number of interval investigated. One mechanism, predominant
parameters of the autoxidation. The rate of autoxida- at pH < 9.5 involves 02.-as a chain-propagating
tion, taken as the initial rate of increase in absorbance species. At higher pH values, another mechanism,
at 420 nm, is increased by a factor of 2 per 0.3 pH not depending on 0 2 . -becomes
, predominant.

Eur. J. Biochem. 47 (1974)

S. Marklund and G. Marklund 47 1

The part of the autoxidation which may be in-

r A,lgo0 --
hibited by superoxide dismutase (dotted line in Fig. 1)
reaches a plateau above pH 10 at a rate about twice the - 80
5
rate at pH 9. This fact and the decrease in slope of E

the curve describing the rate of autoxidation (Fig. 1) -$ 6700 1

as pH 9 is approached, suggest that the first anion
of pyrogallol (pK,, z 9, [lo]) has a central role in the
02.--dependent autoxidation of pyrogallol.
._
2 50
TI
1500 g
Ln
x
The pH dependence of the involvement of 02.- 0

in the autoxidation of pyrogallol is opposite to that

3 40 4400

found for the autoxidation of adrenalin, where 02.--

dependent paths increase in importance with increasing
pH and predominate at pH > 9 [4].

Effects oj'Pyrogallo1 and Oxygen Concentration

on the Autoxidation
Fig. 2 shows that the rate of autoxidation increases [ Pyrogallol] (mM)
linearly with pyrogallol concentration, but the straight Fig. 2. Effects of pyrogallol concentration on the autoxidution.
line does not pass through the origin. The maximal Air equilibrated 50 mM Tris-cacodylic acid buffer, pH 8.20
inhibition of the autoxidation by superoxide dismutase
+ 1 mM EDTA. (A) Rate of autoxidation as determined
from the increase in AdZ0per min; (0)requirement of super-
remains unaltered at 97 % in the concentration interval. oxide dismutase for 50% inhibition of the autoxidation;
The effect of the other reactant, O,, is shown in (0) requirement of a different kind of superoxide dismutase
Fig. 3. The rate increases with increasing 0, concen- with high molecular weight, probably containing Mn 1181,
relative amount
tration, but not linearly. The maximal inhibition by
superoxide dismutase increases from 97 % at 20 % O2 I
to 99% at 100% 0,.

Effects of Transition-Metal Ions

Cuz' and Mn2+ do not significantly affect the
autoxidation in the presence of 1 mM EDTA, until
the chelating capacity is exceeded. Then both ions
are powerful catalysts of the autoxidation and the
reaction is not affected by superoxide dismutase.
Fez', on the other hand, is active in micromolar
concentrations in spite of the presence of EDTA as
shown in Fig. 4. The reaction is apparently not
dependent on 0 2 . -as, the increment in rate brought
about by Fez+ is essentially identical in the presence 02 saturation (%)
and absence of superoxide dismutase. Superoxide
Fig. 3. Effects of oxygen saturation on autoxidation of pyro-
dismutase is not inhibited by Fez+ [4]. gallol. 0.2 mM pyrogallol in 50 mM Tris-cacodylicacid buffer
The autoxidation of pyrogallol in the presence of pH 8.20 + 1 mM EDTA. (A) Rate of autoxidation as de-
EDTA may be initiated by, but is apparently not termined from the increase in A420 per min; (0)requirement
directly catalyzed by, traces of transition metal ions. of superoxide dismutase for 50 % inhibition of the autoxida-
tion
Effect of Catalase
Catalase (0.2 pM) has no effect on the rate of molecules in which one carbon has been oxidized to
autoxidation of pyrogallol (tested at pH 8.2 and 9.1) CO, [12]. CO evolution has also been found during
as determined spectrophotometrically, whereas the pyrogallol autoxidation [13]. However, the effect of
rate of oxygen consumption is halved by the enzyme. catalase shows that hydrogen peroxide is the only
Thus, hydrogen peroxide is not involved in the autoxi- significant product of oxygen during the initial part
dation. of the autoxidation.
A late product of pyrogallol autoxidation is pur- An old report of inhibition of the autoxidation
purogallin [ l l ] which is formed from two pyrogallol by catalase [14] may be explained by the fact that

Eur. J. Biochem. 47 (1974)

412 Pyrogallol Autoxidation, Superoxide Dismutase

100 1 0

1 4
'0 50 100 150 200 250 300 350 ZOO00
Superoxide dlsmutase (ng/ml)
Fig. 5. Inhibition 01'pyrogallol autoxidation by superoxide
dismutase. 0.2 mM pyrogallol in air equilibrated 50 mM
+
Tris-cacodylic acid buffer, pH 8.20 1 mM diethylenetri-
aminepentaacetic acid. The rate of autoxidation was taken
from the increase in A,,, per min

lo/
It'
i Most procedures for determining superoxide dis-
mutase are based on the ability of the enzyme to in-
1 1 1 1 1 1 , < I I I I L I I I I I I I I I I I I I hibit O,.--dependent reactions. A unit of the enzyme
10 20 30 is generally defined as the amount of the enzyme
[FeSO*] (W) which inhibits the reaction by 50%. In the present
Fig. 4. Ejfects o j Felt on the uuto.uidation of py?yrogallol. method one unit corresponds to 100 ng bovine (Cu-
0.2 mM pyrogallol in air equilibrated 50 mM Tris-cacodylic
acid buffer, pH 8.20 + 1 mM EDTA. (A) Rate of autoxida- Zn) superoxide dismutase in a total volume of 1 ml;
tion in the absence of superoxide dismutase; (A) rate of thus the method is about as sensitive as the method
autoxidation in the presence of 20 pg/ml superoxide dismutase based on the reduction of cytochrome c by xanthine
oxidase [2].
This sensitivity is for some purposes, e.g. monitor-
superoxide dismutase often contaminates catalase ing column eluates, inconveniently high. Higher con-
preparations [IS]. centrations of pyrogallol decrease the sensitivity,
cf Fig. 2. At 0.5 mM pyrogallol the sensitivity is
decreased by a factor of 3.3.
Determination of Superoxide-Dismutase Activity
The pyrogallol used in the present experiments
by Means of Inhibition of Pyrogallol Autoxidation
had been purified by sublimation. However, two
The inhibition of pyrogallol autoxidation brought reagent grade pyrogallol preparations (Merck AG ;
about by superoxide dismutase can be employed in a British Drug Houses) were found to give the same
rapid and convenient method for the determination result. HC1 can be used instead of cacodylic acid in
of the enzyme. After the above investigation we chose the buffer but with some 30% loss of sensitivity. As
to use 0.2 mM pyrogallol in air-equilibrated 50 mM can be deduced from Fig. 1 and 3, adequate pH control
Tris-cacodylic acid buffer pH 8.20, containing 1 mM and air equilibration of the buffer are essential for
diethylenetriaminepentaacetic acid. Iron, even in high reproducibility.
trace amounts, accelerates pyrogallol autoxidation in Cu-Zn-containing superoxide dismutases are ef-
spite of the presence of EDTA, Fig. 4. Diethylene- ficien tly inhibited by cyanide [7] whereas those contain-
triaminepentaacetic acid was found to prevent inter- ing Mn [16] and Fe [17] are unaffected. Hence cyanide
ference from Fez+ (as well as from Cu2+ and M n Z +) may be used as a means of differentiating between
and was therefore chosen as chelator in the assay different types of superoxide dismutase. 1 mM cyanide
medium. The rate of autoxidation is taken from the decreases the rate of pyrogallol autoxidation by 25 %
increase in absorbance at 420 nm, which is 0.020 min-' and the amount of superoxide dismutase (probably
in the absence of superoxide dismutase. Fig. 5 shows Mn containing [18]) required for 50% inhibition is
the inhibition of autoxidation as a function of added decreased by 20 %. The increased sensitivity is probably
superoxide dismutase. Superoxide dismutase may be due to a change in the characteristics of the autoxida-
added before or after the pyrogallol. The maximal tion and not to an activation of the superoxide
inhibition which can be obtained is 97.5 %. dismutase.

.Eur. J. Biochem. 47 (1974)

S. Marklund and G. Marklund 473

Reducing compounds interfere with pyrogallol Effects of Pyrogallol on Catechol 0-Methyltransferase

autoxidation [19] by acting as scavengers. Thus 10 pM Pyrogallol is a mixed competitive and noncompeti-
reduced glutathione and ascorbate inhibit the reaction tive inhibitor of catechol 0-methyltransferase [27,28].
by 20 ”/,. The noncompetitive inhibition increases with increas-
The method seems to be applicable for the assay ing purification of the enzyme and with increased
of crude preparations ; bovine Cu-Zn superoxide preincubation with pyrogallol [27]. The assay of
dismutase exerted its full activity as internal standard catechol 0-methyltransferase is performed at 37 “C
in a bovine liver homogenate. The maximal inhibition
in a buffer of pH 7.9 for at least 10 min [27]. The
that can be obtained with a liver homogenate is pyrogallol will be rapidly autoxidized under these
> 95 % and the presence of an amount of liver homo- conditions. During autoxidation of pyrogallol 02.-,
genate giving 80 % inhibition does not at all influence
H 2 0 2 and possibly OH. are formed, the latter by the
the maximal inhibition obtainable with pure super-
oxide dismutase.
+
Haber-Weiss mechanism [29], 02.- H 2 0 2 O H . --f

Peroxidases interfere by accelerating the autoxida- + OH- + 02.These reactive species may contribute
significantly to the noncompetitive effects of pyro-
tion. Catalase in a 10 to 20-fold molar excess totally
gallol on catechol 0-methyltransferase, which con-
eliminates the interference by competing with per-
tains an “essential” sulfhydryl group [30]. It can also
oxidase for H202.Thus, 0.2 pM catalase allows the
be questioned whether the competitive part of the
assay of superoxide dismutase in horseradish homo-
inhibition is due to the pyrogallol or to its oxidation
genates.
products.

We wish to thank Professor K.-G. Paul for valuable

Possibility of Involvement of Singlet Oxygen discussions. The study was supported by Statens Medicinska
Forskningsrcid, grant number B74-13X-4267-01A.
in Pyrogallol Autoxidation
In addition to being a superoxide dismutase,
erythrocuprein has .been stated to be an efficient
REFERENCES
quencher of singlet oxygen (‘dg) [20-221. Singlet
oxygen has been suggested as forming during non- 1. Taube, H. (1965) in Oxygen, Proc. Symp. Spons. New
enzymatic dismutation of 02.-[23] and the inhibitory York Heurt Association, pp. 29- 52, Little Brown
and Co.
effect of superoxide dismutase on pyrogallol autoxi- 2. Mc Cord, J. M. & Fridovich, I. (1969) J . Bid. Chem.
dation could consequently indicate a role of singlet 244,6049 - 6055.
oxygen in the reaction. However, there are several 3. Mc Cord, J. M. & Fridovich, I. (1969) J . Biol. Chem.
objections to this idea. 244,6056 - 6063.
The arguments for a role of erythrocuprein as a 4. Misra, H. P. & Fridovich, I. (1972) J . Bid. Chem. 247,
3170- 3175.
singlet oxygen quencher are questionable. In one 5. Heikkila, R. E. & Cohen, G. (1973) Science (Wash.
paper, the quenching was studied using luminol D. C.) 181,456-457.
chemiluminescence as a detector of singlet oxygen 6. Klug, D., Rabani, J. & Fridovich, I. (1972) J . Biol.
produced during perchromate decomposition [22]. Chem. 247,4839 - 4842.
However. luminol chemiluminescence may be brought 7. Rotilio, G., Bray, R. C. & Fielden, E. M. (1972) Biochim.
Biophys. Acta, 268, 605 - 609.
about by a variety of reagents, and the mechanism(s) 8. Weser, U., Bunnenberg, E., Cammack, R., Djerassi, C.,
is composite [24,25]. In several cases erythrocuprein Flohe, L., Thomas, G. & Volter, W. (1971) Biochim.
inhibits the luminescence and the inhibition could Biophys. Actu, 243, 203 - 213.
reasonably be ascribed to 02-- dismutation [25]. 9. Samejima, T. & Tsi Yang, J. (1963) J . Biol. Chem. 238,
3256-3261.
The lifetime of singlet oxygen (‘dg) in aqueous 10. Kortiim, G., Vogel, W. & Andrussow, K . (1961) Dissociu-
solution has been given as 2 ps [26]. The quenching tion Constants of Organic Acids in Aqueous Solution,
by erythrocuprein was stated to be observable already p. 443, Butterworths, London.
at 0.1 nM [22], requiring a rate constant for the 11. Nierenstein, M. & Spiers, C. W. (1913) Ber. Dtsch. Chcm.
Ges. 46, 3151 -3157.
quenching reaction of the order of 10l5 M-’ x sK1. 12. Barltrop, J. A. & Nicholson, J. S. (1948) J . Chem. Soc.
This is physically impossible, and if erythrocuprein 116- 120.
is a quencher, the lifetime of singlet oxygen must be 13. Miyahara, S. & Takahashi, H. (1971)J . Biochem. (Tokyo)
at least lo5 times longer than reported [26]. 69, 231 -233.
Finally, if there is any significant reaction between 14. Siegel, S . M. & Siegel, B. Z. (1958) Nature (Lond.) 181,
1153- 1154.
the 0.2 mM pyrogallol and singlet oxygen possibly 15. Halliwell, B. (1973) Biochem. J . 135, 379-381.
formed during autoxidation, an extraordinarily high 16. Weisiger, R. A. & Fridovich, I. (1973) J. Biol. Chem.
rate constant is required, 2 lo9 M-’ x sK’, 248,3582- 3592.

Eur. J. Biochem. 47 (1974)

474 S. Marklund and G. Marklund : Pyrogallol Autoxidation, Superoxide Dismutase

17. Yost, F. J. & Fridovich, I. (1974) Arch. Biochem. Bio- 24. Lee, K. D. & Seliger, H. H. (1972) Photochem. Photobiol.
phys. 161,395-401. 15,227-237.
18. Marklund, S. (1973) Acta Chem. Scand. 27, 1458- 1460. 25. Hodgson, E. K. &Fridovich, I. (1973) Photochem. Photo-
19. Ghosh, J. C. & Rakshit, P. C. (1937) Biochem. 2. 294, biol. 18,451-455.
330- 335. 26. Merkel, P. B. & Kearns, D. R. (1972) J . Am. Chem. SOC.
20. Finazzi Agro, A,, Giovagnoli, C., De Sole, P., Calabrese, 94, 7244-7253.
L., Rotilio, G. & Mondovi, B. (1972) FEBS Lett. 21, 27. Crout, J. R. (1961) Biochem. Pharmacol. 6,47-50.
183-185. 28. Baldessarini, R. J. & Greiner, E. (1973) Biochem. Phar-
21. Weser, U. & Paschen, W. (1972) FEBS Lett. 27,248 - 250. macol. 22,247- 256.
22. Paschen, W. & Weser, U. (1973) Biochim. Biophys.
. . Acta, 29. Haber, F. & Weiss, J. (1934) Proc. SOC.Lond. A Math.
327,217-222. Phys. Sci. 147,332-351.
23. Khan, A. U. (1970) Science (Wash. D. C.) 168,476-477. 30. Axelrod, J. & Tomchick, R. (1958) J. Biol. Chem. 233,
702 - 705.

S. Marklund and G. Marklund, Kemiska Institutionen, Avdelningen for Medicinsk Kemi, Umeg Universitet,
S-901 87 Ume%,Sweden

Eur. J. Biochem. 47 (1974)