ANALYTICAL BIOCHtMISTR\r 153, 536-541 (1986)

A Sensitive Spectrophotometric Method for the Determination of Superoxide Dismutase Activity in Tissue Extracts

FRANCEXO PAOLETTI, DONATELLA ALDINUCCI, ALESSANDRA MOCALI, AND ANNA CAPARRINI

Received October 3. 1985

Superoxide dismutase (EC I. IS. I. I) has been assayed by a spectrophotometric method based on the inhibition of a superoxide-driven N.ADH oxidation. The assay consists ofa purely chemical reaction sequence which involves EDTA. Mn(II). mercaptoethanol. and molecular oxygen. re- quiring neither auxiliary enzymes nor sophisticated equipment. The method is very flexible and rapid and is applicable with high sensitivity to the determination of both pure and crude superoxide dismutase preparations. The decrease of the rate of NADH oxidation is a function of enzyme concentration. and saturation levels are attainable. Fifty percent inhibition. corresponding to one unit ofthe enzyme, is produced by approximately I5 ng of pure superoxide dismutase. Experiments on rat liver cytosol have shown the specificity of the method for superoxide dismutase. Moreover. common cellular components do not interfere with the measurement. except for hemoglobin when present at relatively high concentrations. The assay is performed at physiological pH and is unaffected by catalase. 15 1986 Academtc Press. Inc

KEY WORDS: superoxide dismutase; spectrophotometric determination: chemical assay: su- peroxide: NADH oxidation: metal complex.

Superoxide dismutase (SOD)’ (EC 1.15.1.1) is a family of metalloenzymes which is known to accelerate spontaneous dismutation of the superoxide radical to hydrogen peroxide and molecular oxygen (1). SOD is widely distrib- uted among aerobically living organisms and has been inferred to play an important role in controlling superoxide levels in cellular com- partments (2,3).

The direct measurement of SOD activity (4- 7) is possible but its application is hampered by the requirement of special apparatus not commonly available in the typical laboratory. The other methods employed for enzyme de- termination are indirect and rely on the ability of SOD to inhibit a superoxide-driven reac- tion. The extent to which SOD reduces the rate of that reaction is taken as a measure of enzymic activity. Either enzymatic or non- enzymatic systems are used for the generation

’ Abbreviations used: SOD. superoxide dismutdse: Tea- Dea. triethanolamine-diethanolamine; NBT. nitro blue tetrazolium.

of superoxide (see Refs. (8-9) for review): de- tection is then accomplished by luminometric ( 10). polarographic (1 l), or calorimetric ( 1,12,13) techniques, depending on the differ- ent approaches and experimental require- ments.

Notwithstanding the large number of methods available, the need still exists to in- crease the specificity. accuracy, and simplicity of the assay. In this paper we describe a spec- trophotometric method for SOD determina- tion based on a purely chemical reaction se- quence which eventually leads to NADH ox- idation. This procedure, involving stable and inexpensive reagents. allows a rapid and highly sensitive measurement of SOD activity in pure and crude enzyme preparations, with negli- gible interference by cellular components.

MATERIALS AND METHODS

Chemicals. Reduced adenine dinucleotide (P-NADH, disodium salt) was obtained from

0003-2697/86 $3.00 536 Copyright 2 1986 by Academic Press. Inc All rights of reproductmn in any form reserved.

SUPEROXIDE DISMUTASE DETERMINATION 537

Boehringer-Mannheim (West Germany), while MnClz . 2H20, ethylenedinitrilotetraac- etic acid (EDTA), and 2-mercaptoethanol were supplied by Merck. Darmstad (West Ger- many). Pure SOD preparations were obtained from Diagnostic Data Inc. (beef erythrocytes, 3300 U/mg protein) and from Sigma Chemical Company (human erythrocytes, ca. 2500 U/ mg protein). Catalase (beef liver. 350 mg/ml ammonium sulfate suspension) was provided by Boehringer-Mannheim. All other chemicals were analytical grade and used without further purification.

Eqrr;pmenr. Assays were performed with a Gilford spectrophotometer (Model 350) con- nected to a recorder.

Rcqynts l~inrl solutions. All solutions were made up with deionized or well-aerated dis- tilled water. according to the following scheme.

( 1) Triethanolamine-diethanolamine (Tea-Dea) buffer. 100 mM each, pH 7.4. Di- lute 14.9 g Tea. 10.5 g Dea. and ca. 13.8 ml of 37%, HCI up to 1 liter with water, taking care to maintain the pH around 7.4-7.5.

(3) NADH, 7.5 mM. For 100 assays, dis- solve 20 mg of reduced nucleotide (disodium salt) in 4 ml of water.

(3) EDTA/MnCI,, 100 mM/50 mM. Make a stock solution of EDTA. 0.3 M (i.e., dissolve 5.85 g of EDTA-acid in a final volume of 100 ml. adjusting the pH to around 7 with molar NaOH solution) and of MnCL2. 0.1 M (by dis-

solving 1.62 g of MnC12. 7Hz0 in 100 ml wa- ter). The mixture of equal volumes of these two stock solutions yields our third reagent (EDTA/MnCI:).

(4) Mercaptoethanol, 10 mrvr. Dilute 0.05 ml of concentrated thiol. 14.2 M, up to 7 I ml with water.

The NADH solution is stable for at least 3 days in the refrigerator. For longer storage. keep it at ~20°C. The solutions of EDTA. MnCl?, and mercaptoethanol are quite stable, even at room temperature. for months. Re- agent 3, once made up. can be used over a 2- week period (see further comments in l.he Re- sults section).

Preparation qf’ rat liver c~msoi. For the measurement of SOD in rat liver, the tissue extract is first prepared by homogenizing the liver in 3 vol of Tea-Dea buffer 25 mM, pH 7.4, and then cleared by centrifugation at 30,000 rpm for 60 min at 4°C. The superna- tant is dialyzed against cold homogenization buffer and then employed for enzymatic as- says. Protein determination was carried out by the L.owry method ( 14) on samples dialyzed against 0.9% NaCl buffered at pH 7.5-8 with molar NaHCOI solution.

RESULTS

Description oJ‘t/re method. The principle of this method is based on the oxidation of NADH, mediated by superoxide radical, in a purely chemical system recently developed in our laboratory. Coenzyme oxidation occurs in the presence of suitable concentrations of EDTA. Mr? and mercaptoethanol (see be- low) through a free-radical chain of reactions involving thiol oxidation and univalent 02 re- duction. A detailed explanation of the reaction mechanism is beyond the scope of the present paper and will be reported elsewhere (manu- script submitted). ‘The addition of SOD to the reaction mixture causes a proportionate in- hibition, of the rate of NADH oxidation, thus confirming the involvement of superoxide in the process and providing the basis for SOD activity determination.

To perform the assay sequentially add the following (see Materials and Methods) to a cuvette: Tea-Dea buffer. 0.800 ml: NADH solution, 0.040 ml; EDTA/MnC12 solution, 0.035 ml; sample (or water or buffer) 0.100 ml: mix thoroughly and read against air at 340 nm for a stable baseline: then add mercapto- ethanol solution. 0.100 ml.

Mix and monitor the decrease in absorbance using 0.5-I full scale deflection. The final vol- ume in the cuvette is 1.065 ml and the light path is IO mm.

A typical analysis of SOD activity is shown in Fig. 1 where the kinetics of NADH oxida- tion in the absence (control) and in the pres-

538 PAOLETTI ET AL

, SgAq

1.4 -.-*--

i

-----__ ---_ - - -SOD 3

- -. - . --_

\

-11 --__

’ ‘1

--__

E \ - - S O D 2

c

z 1.2

I

’ ‘\

' \ 0 \ \

z \ \

\ \

\ \ \

\ \ \ \ 'SOD,

\ \

L ‘\CONIROI

0

0 8 16 24

INCUBATION (mtnl

FIG. I. Effect of superoxide dismutase on the rate of NADH oxidation. Four assays are carried out simulta- neously in the absence (control) and in the presence of SOD (sample 1. IO ng; sample 2, 80 ng: sample 3, 380 ng of pure enzyme from bovine erythrocytes. Diagnostic Data Inc.). Assay mixtures are prepared as described in the text and decreases in absorbance. at 340 nm. are recorded for about 15 min after mercaptoethanol addition. The rate of NADH oxidation in the control is ca. 5 nmol per min. at room temperature.

ence of superoxide dismutase (SODrm3) are compared by simultaneous assay using a mul- ticell holder.

The reaction is started by mercaptoethanol and changes in absorbances are recorded for about 15 min. Rates of NADH oxidation are initially low, then increase progressively (usu- ally 2-4 min after mercaptoethanol addition) to yield good linear kinetics (5- 10 min) which are used for calculations. Under our conditions Ai1340 over an S-min interval is 0.250 for the control, while the presence of 10, 80. and 380 ng of SOD in the assay mixture. yields AA values of 0.150.0.038, and 0.008, respectively. For calculations the maximal rates obtained are expressed as a percentage of the control (ordinate) and plotted against a suitable ref- erence (abscissa). One unit of the enzyme is the amount of SOD capable of inhibiting by

50%’ the rate of NADH oxidation observed in the control.

Culihration cm’e \t’ii/l pure SOD. The de- termination of SOD activity in pure enzyme preparations from bovine erythrocytes is shown in Fig. 2. Relative rates of NADH ox- idation are reported as a function of the amount of enzyme in the assay mixture. The curve thus obtained shows that inhibition is not directly proportionate to SOD concentra- tion, but rather follows an exponential-like function. Almost complete saturation levels (99% inhibition) are obtained with 400 ng of the enzyme, while the same amount of heated SOD fails to inhibit the reaction. One unit of the enzyme corresponds to ca. 15 ng of pure superoxide dismutase from beef erythrocytes.

Least-square linear regression analysis was used to obtain a best-fitting curve over the range 4-40 ng, by transforming SOD values into logarithms. The equation is ?; = 1 16.638 - 55.619 s; the correlation coefficient Y = -0.9925; n = 22.

Deterrninution ofSOD in liver euxtructs. To test the applicability of this method to SOD determination in tissue extracts, experiments were carried out using rat liver cytosol as the sample (Fig. 3). The assay conditions are the same as described for pure SOD and the per- centage of NADH oxidation is reported as a function of the total proteins in the extract. The curve obtained is essentially identical to that shown in Fig. 2 for the purified enzyme and in this case too, the boiled sample fails to inhibit NADH oxidation at any rate. Fifty percent inhibition is produced by approxi- mately 10 pg of liver extract which means an average of 100 units of SOD per mg protein of cell cytosol. while the saturation level is reached with ca. 180-200 yg proteir.

Curve fitting to experimental data is com- parable to that described in the comments to Fig. 2. The linear regression equation, over a range of 4-40 pg protein, is c = 115.366 - 67.687 s; the correlation coefficient t = -0.9869; n = 28.

Precautions and optimal conditiom ,fi~r rnemuwnmt. EDTA or other chelators for Mn2+ (not EGTA) may alter the optimum

SUPEROXIDE DISMUTASE DETERMINATION 539

-‘-t-l--c-i-+- 20 40 400

SUPEROXIDE DISMlJiASE lngl

FIG. 2. Titration curve with pure superoxide dismutase. Increasing amounts of SOD (O-400 ng) from bovine erythrocytes (Diagnostic Data Inc.) are added to incubation mixtures and assayed for activity as previously reported. The rate of NADH oxidation (g-min linear kinetic) is expressed as a percentage of the control. which is always run in each set of assays. Values (0: n = 27) refer to separate determinations carned out individually by three of us using different stock solutions of the enzyme. Samples containing inactive SOD (0). heated at 100°C for 2 min. are shown for comparison.

EDTA/Mn’+ ratio and affect the rate of re- action. Alternatively, excess of Mn” ions in the sample could slow down the rate of NADH oxidation. Other divalent cations of the second transition series, at concentrations comparable to that of Mn”. do not start the reaction, but may compete for the chelator. In addition, when the sample contains free thiols, i.e., mercaptoethanol, cysteine and reduced glu- tathione. faster reaction rates, as compared to the control. are observed. To avoid all inter- ference by compounds mentioned above. samples must be dialyzed against suitable me-

dia before the analysis. However, owing to handle a large number of samples. dialysis could be conveniently replaced by a rapid de- salting through a small Sephadex G-25 (coarse) column.

For precise calculations, NADH-oxidase activity, if present in the sample, should be evaluated before mercaptoethanol addition and subtracted from the final rate. Moreover, because of the high sensitivity of the method. samples are usually diluted by such a large factor that NADH-oxidase or any other inter- fering activity is practically undetectable. In

0 L’+- ++- t~-+--+-~.+ .-.. ++ --+ - 0 8 I6 24 32 184

LIVER CYTOSOL ,,,g pratml

FIG. 3. Titration of SOD in rat liver cytosol. The liver extract 1’see Materials and Methods) is diluted with 100 mM Tea-Dea buffer. pH 7.4. to concentrations suitable for the assay. Measurements are carried out as

described in the text. Protein content of the sample is within a range of O-400 pg as determined by the Lowry method. Closed dots (0) represent the average of four separate determinations and bars (=) indicate the range of experimental values. Open squares (0) refer to assays with boiled samples.

540 PAOLETTI tT AL.

addition it is worth mentioning that NADPH reacts as well as NADH in the system (data not shown) without serving as a substrate for NADH-oxidase. This fact confirms the flexi- bility of our method and may turn out very useful when assaying for SOD in fractions containing high levels of NADH-oxidase.

Changes of temperature, pH and oxygen tension in the system may influence the ab- solute rate of NADH oxidation, but are with- out effect on the relative degree of inhibition observed. Each set of assays must refer to its own control and best measurements are ob- tained when the values of .&4+,,, of the control. over an S-min interval, are within the range of 0.150-0.400. Reactivity of reagent 3 is in- creased by storage: therefore aged solutions immediately yield maximal rates of NADH oxidation without the initial delay shown by fresh-made preparation of the complex.

Catalase does not interfere with the assay, while hydrogen peroxide inhibits it at a mil- limolar concentration level.

DISCUSSION

Any reaction inhibitable by superoxide dis- mutase could potentially provide the basis for an indirect assay of the enzyme and, according to that principle, several methods have been developed over the years. However. as also pointed out by Oberley (15). only a few pro- cedures permit the sensitive and reliable de- termination of enzymatic activity in tissue ex- tracts with low SOD levels.

Our chemical assay seems particularly suit- able to that purpose since it allows the mea- surement of minute amounts of SOD, such as 2 ng, which are far below the detection limit of most published methods. In addition. we have found that fifty percent inhibition, i.e., one unit of the enzyme, is produced by 15 ng of pure protein, while values of catalytic ac- tivity. as determined by the xanthine-oxidase/ cytochrome c ( 1 ), NADH-diaphorase/hy- droxylamine ( 13) and xanthine-oxidase/nitro blue tetrazolium (NBT) ( 12) systems, are ca. 200,626, and 630 ng, respectively. A substan- tial improvement of the NBT assay has been obtained by Buettner (16), whose procedure

yields exactly the same sensitivity reported here. However. with his method, saturation levels are not attainable and different values for half-maximal inhibition are obtained for pure and crude SOD preparations. These in- conveniences are frequently observed in assays involving the reduction of NBT. cytochrome (‘. or other suitable detector and ascribed to the action of aspecific electron donors on chromogenic substrates. Our method. on the contrary, relies on the oxidation of the detector (NADH in this case) and therefore will not be affected by the presence of reductants which are known to occur in tissue extracts ( 12). This lack of interference is clearly shown by the fact that calibration curves for pure SOD and rat liver cytosol are almost identical and sat- uration levels (99%) are attainable in both cases. The latter result implies that in our sys- tem the same value for catalytic activity is ob- tained by using either 50% or half-maximal inhibition for calculation. This avoids the ne- cessity of running a full calibration curve each time and is of valuable practical importance, especially when dealing with samples having low SOD levels. From our data. a specific ac- tivity of ca. 66.000 unit/mg protein and ca. 100 unit/mg protein can be calculated for SOD in pure beef erythrocytes preparations (Diag- nostic Data Inc.: see Materials and Methods) and rat liver cytosol, respectively.

In addition to the experiments with rat liver cytosol, the present method has been applied successfully to the measurement of SOD in a variety of other cell extracts and body fluids (data not shown). So far, the only major in- convenience encountered comes from he- moglobin; thus, for reliable determination of SOD in hemolysates. this molecule must be removed from the sample before assaying.

An important problem in SOD analysis is the discrimination between the cuprozinc- and manganese-form of the enzyme. Cyanide. at concentrations used for inhibiting the cupro- zinc enzyme, does not interfere with our assay and manganese-SOD (Mn-SOD) can be easily determined by differential measurement. In addition, this method has proved ofgreat value in determining traces of Mn-SOD separated

SUPEROXIDE DISMUTASE DETERMINATION 541

from rat liver cytosol by means of gel filtration (data not shown). In this regard it is worth recalling that our assay is carried out at pH 7.4 which favors the detection of Mn-SOD. In fact, physiological pH is the most suitable for optimal activity of Mn-SOD, which is not re- liably assayed at elevated pH values as required by other sensitive methods ( 17,18).

On the whole, the present procedure for SOD determination involves stable and in- expensive reagents and consists of a single spectrophotometric step, easily performed on a time scale of minutes. It appears particularly suitable for application in the field of bio- chemistry. plant physiology, and clinical chemistry.

ACKNOWLEDGMENTS

We would like to thank Professor A. Fonnesu. &airman of the Institute. and Dr. V. Boddi for his continuous sug- gestions and statistical analyses. This research was sup- ported by a grant from the Minister0 della Pubblica Istru- zione (6Og’).

REFERENCES

I. McCord. J. M., and Fridovich, 1. ( 1969) J. Viol C’11crn. 244,6049-6055.

2. McCord. J. M.. Keele. B. B.. Jr.. and Fridovich. I. (1971) Proi,. Nar. .-I(& SC,; C51 68, 1024-1027.

3. Fridovich. I. (1978) Sc~rc~nc~c 201, 875-880.

4. Ballou. D., Palmer. G.. and Massey. V. (1969) Biochrm. Bioqlzm R~.L Cornrnm 36, 898-904.

5. Rotilio. G.. Bray, R. C., and Fielden. E. M. (1972) Birxhirn. BiuphJ,.r. .1ctu 268. 605-609.

6. Klug. D.. Rabani, J.. and Fridovich, 1. (1972) J Bid C7rrm. 247, 4839-4442.

7. Rigo. A.. Viglino. P.. and Rotilio. G. (1975) A& Biodxw. 68, 1-X.

8. McCord. J. M.. Crapo. J. D., and Fridovich. I. (1977) in Superoxide and Superoxide Dismutases (Mi- chelson A. M.. McCord. J. M.. and Fridovich, I.. eds.), pp. I l- 17. Academic Press. New York.

9. Flohe. L.. and Gtting, F. ( 1984) in Methods in En- zymology (Colowick. S. P.. and Kaplan. N. 0.. eds.). Vol. 105, pp. 93-104. Academic Press. New York.

IO. Bensinger. R. E., and Johnson, C. M. ( I98 I ) .I& Biodwnz. 116. I4?- 145.

I I. Tyler. D. D. (1975) B~odw~~~. J. 147, 493-504.

12. Beuchamp. C.. and Fridovich. I. ( 197 I ) .,lr~ul. B,nchrtn. 44, 276-187.

13. Elstner, E. F.. Youngman. R. J.. and OBwald, W. ( 1’383) irz Methods of Enzymatic Analysis (Berg- mmeyer, H. U.. ed.). Vol. III. pp. 293-302, Verlag Chemie, Weinheim.

14. Lowry. 0. H.. Rosebrough. H. J.. Farr. A. L., and Randall. R. J. ( 195 I ) J. Blol. Ch~7 121, 404-420.

15. Oberley. L. W.. and Spitz. D. R. ( 1984) in Methods in Enzymology (Colowick. S. P.. and Kaplan. N. 0.. eds.). Vol. 105. pp. 457-464. Academic Press. New York.

16. Buettner. G. R.. Oberley, L. W.. and Leuthauser. S. W. H. C. ( 1978) Phr~toc~hcr,?. Photohid 28, 693- 695.

17. Marklund, S. (1976) J Bird C‘hn. 251, 7504-7507. 18. Misra. H. P.. and Fridovich, I. (1972) d Bid Chc177

247, 3170-3175.