Hydroxyl free radical reactions with amino acids and proteins studied by electron spin resonance spectroscopy and spin-trapping

238

Biochimica et Biophysica Acta, 790 (1984) 238-250

Elsevier BBA 32027

HYDROXYL FREE RADICAL REACTIONS WITH AMINO ACIDS AND PROTEINS S T U D I E D BY ELECTRON S P I N R E S O N A N C E S P E C T R O S C O P Y A N D S P I N - T R A P P I N G IMRE ZS.-NAGY * and ROBERT A. FLOYD Oklahoma Medical Research Foundation, Biomembrane Research Program, 825 N.E. 13th Street, Oklahoma City, OK 73104 (U.S.A.)

(Received March 2nd, 1984) (Revised manuscript received July 20th, 1984)

Key words: Free-radical scavenger," Amino acid," ESR," Ferrozine," Fenton reaction

It has recently been shown that Fe(II) complexes of ADP or ATP generate OH radicals with H202 in a Fenton-type reaction. The OH radicals can be detected by using 5,5-dimethyl-l-pyrroline-N-oxide (DMPO) as a spin trap in electron spin resonance spectroscopy. All the biologically occurring amino acids, some related compounds and several proteins (histone, bovine serum albumin, collagen) were tested as OH radical scavengers against DMPO. The tested compounds competed with D M P O in trapping OH radicals to various extents as shown by the decrease of signal intensity of DMPO-OH spin-adduct. The tested compounds did not oxidize Fe(II) itself, with the only exception being tyrosine, as revealed by properly designed ferrozine reaction. Some of the amino acids reacted also with theDMPO-OH spin-adduct to a certain extent, whereas others did not. The formation of carbon centered organic radicals of the amino acids could be detected under the influence of OH radicals by using the spin traps phenyl-tert-butylnitrone (PBN) and a-pyridyl-l-oxideN-tert-butylnitrone (4-POBN). The proteins, however, did not react with these spin traps. One can conclnde that the amino acids and proteins can be targets of OH radical damage even in vivo, and such phenomena may be of importance in the deterioration of the conformation of proteins, e.g., during aging or in some pathological processes.

Introduction There is no d o u b t that oxygen free radicals occur in biological systems [1-4]. Free radicals have been implicated as etiological factors in a n u m b e r of biological p h e n o m e n a such as aging, mutagenesis, inflammation and other pathological * Guest scientist from: F. Verzar International Laboratory for Experimental Gerontology (VILEG), Hungarian Section, University Medical School, H-4012 Debrecen, Hungary, to whom reprint requests should be sent at this address. Abbreviations: ADP, adenosine 5'-diphosphate; ATP, adenosine 5'-triphosphate; DMPO, 5,5'-dimethyl-l-pyrroline-l-oxide; 4-POBN, a-pyridyl-l-oxide-N-tert-butylnitrone; TMPN, 4-hydroxy-2,2,6,6-tetramethylpiperinooxy. 0167-4838/84/$03.00 © 1984 Elsevier Science Publishers B.V.

conditions [5-13]. The effect of a small a m o u n t of oxidative damage in enzymatic function and intracellular proteolytic degradation of glutamine synthetase has recently been demonstrated [14,15] being highly suggestive of the biological significance of free radical reactions in vivo. Therefore, it is essential to learn more about the reactivity and molecular mechanisms of biologically important c o m p o u n d s like amino acids and proteins with oxygen free radicals. Spin-trapping [16,17] has recently been utilized to demonstrate O H radical formation from hydrogen peroxide as catalyzed by ferrous iron-nucleotide complexes [18]. It was demonstrated that A D P or A T P form complexes with ferrous iron even at

239 neutral pH, which in the presence of hydrogen peroxide produces hydroxyl free radicals in a Fenton type reaction. These O H radicals were trapped by DMPO, resulting in a unique ESR signal which can be under certain circumstances used to estimate the amount of radicals trapped. It has also been demonstrated that known hydroxyl free radical scavengers such as ethanol and thiourea were able to considerably reduce the amount of trapped radicals [18]. These observations have been extended to di- and triphosphate nucleotides other than adenine as well [19]. These findings take on biological importance in light of the observations showing that Fe-nucleotide complexes do exist intracellularly [20] and that they take part also in iron transport processes [21,22]. On the other hand, it was demonstrated long ago that Fe complexes of A D P and ATP strongly increased lipid peroxidation of rat liver microsomes [23]. ESR studies of spin-trapped radicals of amino acids and related compounds have been carried out under various experimental conditions [24-32]. However, the authors listed above used either gamma-radiolysis or ultraviolet-photolysis for generating free radicals; therefore, no direct comparison is possible with the Fe(II)-ADP-H202 model. Nevertheless, it should be noted that quite recently radicals were separated from gamma-irradiated polycrystalline amino acids by chromatographic methods and some well-defined products could be identified by spin-trapping [32]. For example, deamination was observed for all amino acids and H-abstraction was shown to take place in some of them [32]. All these data suggested that it is desirable to test whether amino acids and proteins are able to react with O H radicals generated in the ADP-Fe(II)-H202 system. The present paper summarizes results of the first of a series of experiments where we demonstrate that amino acids and proteins are targets of free radical attack in Fenton-type reactions. It is quite possible that similar types of reaction occur intracellularly. The biological implications of these reactions appear to be of considerable importance. Material and Methods

The preparation and maintenance of the stock solutions used for the ADP-Fe-H202 reaction

mixture and spin-trapping of O H radicals were essentially the same as described by Floyd and Lewis [18]. The following stock solutions were used routinely: 20 m M ADP; 1.0 m M FeC12 in 0.0012 M HCI; 200 m M D M P O and 0.3% H202. The composition as well as the sequence of addition of the components to the reaction mixture was slightly modified according to the purposes of the particular experiments. In the basic experimental set-up, we added the compounds in a volume of stock solution as follows: 10 /zl ADP; 10 /~1 FeC12; 4 - 2 0 /tl test substance (or buffer for controis); buffer (100 m M N a C I / 2 5 m M N a H C O 3 (pH 7.1)) in various volumes in order to obtain a total volume of 100 #1; 5-25 #1 spin trap (DMPO); and 10 #1 H202. We label this type of experiment 'pre-addition' to differentiate from the 'post-addition' experiments where the test substance was added only after the H202. In order to assure uniform timing, 30 s elapsed between the addition of each component, while the tubes were kept at 37°C. After the addition of H202 and incubation for 30 s at 37°C, the sample was transferred to a heat-sealed Pasteur pipet in order to obtain the ESR spectrum. The recording of the spectrum started within 30-40 s after the end of the incubation. The ESR spectra were obtained on a Varian-E109 X-band spectrometer operating in the first derivative mode. Typical instrumental parameters were as follows for the D M P O spin-trapping experiments: Field set 324 mT; scan range 10 mT; 4 min scan time; time constant 0.3 s; modulation frequency 100 kHz; modulation amplitude 0.2 mT; receiver gain 1.25. 103; incident microwave frequency was 9.14 G H z with a power level of 25 mW. The ESR spectra were obtained while the samples were at room temperature (25°C). The following amino acids and related compounds were tested (all products were obtained from Sigma Chemical Co. and were of the L type); alanine, glycine, isoleucine, leucine, serine, threonine, valine, arginine, N-e-acetyllysine, yaminobutyric acid, guanidine, histidine, hydroxylysine, lysine, polylysine, cysteine, cystine, methionine, phenylalanine, tryptophan, aspartic acid, asparagine, glutamic acid and glutamine. The following proteins were tested: bovine serum albumin (2-10 m g / m l ) ; histone VI (0.2-2.0 m g /

240 ml); collagen (from rat tail tendon, acid soluble) (0.5-2.5 m g / m l ) . All proteins were products of Sigma, and the concentrations presented are final ones. All the tested compounds were dissolved in the N a C I / N a H C O 3 buffer used throughout the experiments. Whenever the pH of the stock solution increased above 7.5, it was adjusted to pH 7.0-7.5 by the addition of HC1. The dilution effect was negligible. At this point, one has to summarize briefly the available knowledge regarding the reactions taking place in the Fenton reagent. We shall use the data of Walling [59] in order to point out the competitive reactions which may be of importance in our experimental model. Although the experiments of Walling [59] were performed at pH < 2, the basic concepts are certainly valid in our system, too. The starting reaction can be written as follows: kl

H202 + Fo(II) -o Fe(III) +HO- + HO

(1)

where k 1 = 76, i.e. the reaction is not extremely fast [59]. All the further reactions in which the O H radicals participate are considerably faster: OH + Fe(II) -,k2Fe(III) + HO-

(2)

The best estimate for k 2 is 3.108 M -~ • s -~ [59]. This means that the O H radicals generated by Eqn. 1 will react extremely quickly with the rest of Fe(II). However, if an organic compound (RH) is present: k3

HO+RH ---,R +H20

(3)

may take place, where k 3 is in the range of 107-101° M -1- s -1 [59]. Eqn. 3 is written for simplicity as hydrogen abstraction; however, in the case of unsaturated bonds, kinetically equivalent additions may replace this reaction. Such an addition is, e.g.: k4

DMPO + OH ~ DMPO-OH

(4)

where k 4 = 3.4-109 M -x- s - t [18]. This implies that the O H radicals formed in reaction 1 can be eliminated in a competitive way. If R H is missing, a given portion of O H radicals will be trapped by

DMPO. This portion is determined by the value of whereas, in the presence of RH, the value of k 3 / k z will also contribute to the formation of the products according to the Eqns. 2, 3 and 4. However, the situation is further complicated by the fact that R may undergo further reactions. Namely, R can be oxidized by Fe(III), i.e., it will regenerate Fe(II) and thus propagate a redox chain through reaction 1; it may dimerize or may even be reduced to R H by Fe(II) [59]. Even the DMPOOH may undergo decomposition as shown by the decrease of its ESR signal height in time. It is also possible that in some cases R , DMPO-OH and the ADP-Fe(II) complex can react with each other or with some yet unidentified intermediary products. A further point to be considered is that in our system the molar ratio of H202 to Fe(II) is almost 90, i.e., hydrogen peroxide is available in great abundance. Therefore, we have to calculate the Fe(III) catalyzed decomposition of H202, too. Walling [59] gives the equations of such reactions and concludes that the redox chain mechanism is firmly established for such systems. This implies that OH radical formation will proceed even after the first cycle of reaction (1) is finished by means of regeneration of Fe(II) in the system. However, the rate of OH radical formation will be much lower in this period of the reaction. This assumption is supported by the experimental results of Floyd and Lewis [18]; that is, when starting with the ADP-Fe(III) complex, the amount of OH radicals trapped by DMPO was to about 22% of that observed in the case of ADP-Fe(II) complex. It is obvious from the values of k 2, k 3 and k 4 that any amount of regenerated Fe(II) will be oxidized again very rapidly into Fe(III) until H202 is available, i.e., we cannot detect the presence of Fe(II) in our experimental system. Considering the above, regarding the reactions taking place in our experimental model, one can conclude that the preaddition and post-addition models are similar in qualitative terms, since in both cases OH radicals are formed in the system; however, they are different in quantitative aspects. That is, in post-addition experiments most probably many fewer O H radicals are available for the test compounds. On the other hand, one has "to stress that some interactions between R and D M P O - O H are taking place in the preaddition

k4/k2;

241 model but may have different rates in the post-addition experiments. The present paper cannot, for obvious reasons, clarify all the reactions which may be involved in the interaction of our test substances with the F e n t o n - D M P O reagent. Nevertheless, it was essential to prove that the added substances do not interfere with the state of iron in the ADP-Fe(II) complex. For this purpose, the following controls have been performed. We determined whether the test substances oxidized ferrous iron in the ADP-Fe(II) complex by utilizing a colorimetric determination of Fe(II) with Ferrozine [33]. Ferrozine reacts only with Fe(II), yielding a pink-colored complex absorbing visible light with a maximum at 562 nm and a molar absorptivity of 28 000 [34]. The complex is fairly stable over a wide p H range. The Ferrozine reagent was prepared by dissolving 75 mg Ferrozinc (Aldrich) in 25 ml water containing one drop of concentrated HC1 as suggested by Carter [34]. This resulted in a stock solution of 5.88 m M which was diluted 10-fold when added to the spin-trapping reaction mixture described above in place of H202. Since Ferrozine reacts with Fe(II) in a molar ratio of 3 : 1 [33], its final concentration was nearly twice as high as necessary to insure the 3 : 1 complex with all possible ferrous iron. The Ferrozinc reaction was completed at 37°C for 2 min. The absorbance was measured immediately thereafter at 562 nm in microcuvettes of 210/xl volume. Proper blanks and standards were prepared in all cases. The Ferrozine reaction was also carried out in the absence of ADP, in order to determine whether the test substance had any influence on ferrous iron if not protected by ADP. In addition, the Ferrozine reaction was also performed before and after the addition of H202, in order to determine whether ferrous iron was oxidized in the presence of the test substances.

to a certain extent and also reacts with other radicals formed in the system. 4-POBN is quite soluble in water; therefore, a stock solution of 0.5 M was prepared in buffer and a final concentration of 100 m M was used in the experiments. (B) PBN (phenyl-tert-butyl-nitrone) [16]. This trap has been widely used by numerous authors in various systems [36-42]. A stock solution of 150 m M was prepared in buffer and the final concentration used was 50 mM. (C) M N P (2-methyl-2-nitrosopropane). This compound was one of the first spin traps ever used and up until now has been one of the most popular ones [24,31,32,43-52]. This compound is only sparingly soluble in water; therefore, a stock solution of 5 m g / m l was prepared in buffer by stirring in the dark at 37°C. The final concentration of this spin trap was diluted three times. The final concentration was about 8 # M [45]. The ESR spectra when using these spin traps were obtained essentially as described for D M P O with the only difference that higher receiver gains were used with a correspondingly increased time constant and scan time. The amount of the radicals spin-trapped by 4-POBN and PBN was estimated using a calibration method [54] with the stable free radical T M P N (Aldrich) as a standard for this purpose in a concentration range from 4 to 58 /xM. The linewidths of most spin-trapped free radicals are near that of T M P N , and thus the peak height of this stable free radical was measured under identical instrumental conditions as those used in the 4POBN and PBN experiments. A standard curve was established in order to approximate the amount of free radicals spin-trapped. Estimates were calculated only for spin-trapped radicals yielding the strongest signals.

The use of other spin traps In order to determine whether the test compounds formed reactive radicals in the ADPFe(II)-H202 system, spin traps other than D M P O were also used, namely: (A) 4 - P O B N ( a - p y r i d y l - l - o x i d e - N - t e r t - b u t y l nitrone). This trap has been described by Janzen et al. [35]. It is able to react with O H radicals directly

The effect of amino acids and proteins on the hydroxyl free radical spin-adduct of DMPO A typical ESR spectrum obtained in the ADPFe(II)-H202 system when D M P O is used as the spin trap is shown in Fig. 1A. This spectrum is characteristic for the O H radical spin-adduct of the D M P O molecule in water having coupling constants of A N = A f t = 1.492 m T [18,53,55].

Results

242 1 mT

B

G

H

Fig. 1. ESR spectra of the DMPO-OH spin-adduct obtained in the ADP-Fe(II)-H202 reaction mixture. The concentration of DMPO was 50 mM in each case, and the instrumental parameters were described in the experimental procedure, except that the receiver gain was 5.102 and time constant 0.1 s: (A) The control system; (B) and (C), the same as (A) except that H202 or Fe(II) were omitted, respectively; (D)-(I), the same as (A) except in the presence of 50 mM of the followingamino acids: isoleucine, valine, phenylalanine, methionine, histidine and lysine, respectively. (J) The same as (A) in presence of 2 mg/ml historic.

Utilizing similar concentrations of D M P O , quantitatively the same amount of spin-trapped O H was obtained as that by Floyd and Lewis [18]. Properly designed control experiments, i.e., exclusion of Fe(II), ADP, D M P O or H202 from the reaction mixture, clearly demonstrated that the trapped O H radicals were derived from Fe(II) catalysis of H202 (Figs. 1B and 1C). If the experimental conditions and time interval of addition of the components were standardized as described in the experimental procedure, the signal height of D M P O - O H - a d d u c t could be reproduced within _5%, allowing us to make relative quantitative

estimations of the amount of OH radicals trapped. The presence of amino acids and proteins in most cases decreased the amount of DMPO-OHadduct present. Figs. 1 D - 1 J show some examples. In these experiments, the concentration of the test substance was equimolar with that of DMPO, unless the substance was of low water solubility which then forced us to lower the concentration. It is apparent that methionine behaved quite differently from the others. It decreased the signal height of the DMPO-OH-adduct, but at the same time, another trapped free radical signal appeared in the spectrum (Fig. 1G). These additional peaks were not present, if D M P O was omitted from the reaction, i.e., they represent the reaction of some radical other than OH with DMPO. This observation deserves further study. A summary of the results obtained with various groups of amino acids, some other related compounds and proteins is presented in Figs. 2 - 6 and Table I for a concentration range of D M P O from 0 to 50 mM. It can be seen from these figures that many of the amino acids, if present during the generation of the spin-adduct, decrease the height of the hydroxyl free radical D M P O spin-adduct signal. Some of them do not, however. Of the neutral amino acids, leucine and isoleucine as well as, to a smaller extent, valine and serine decrease the amount of D M P O hydroxyl free radical spin-adduct observed. As the data in Table I show, this apparently is in part due to these amino acids acting upon and causing a decrease in the nitroxyl free radical itself. We do not know the chemical mechanism as to how the nitroxyl free radical is rendered diamagnetic; presumably, it is either reduced or oxidized [58]. In the group of basic amino acids, we have included two other related compounds, guanidine and "t-aminobutyric acid. It is interesting to note that arginine is a much better scavenger of OH radicals than guanidine, i.e., the guanidino group of arginine in itself is not responsible. On the other hand, we compared lysine with some of its derivatives: acetyllysine, polylysine and hydroxylysine and found that acetyllysine and polylysine were effective scavengers of OH, whereas lysine and hydroxylysine were less effective but exerted most of their effect also in post-addition. "y-Aminobutyric acid was less active in scavenging OH radi

243

/ 100

~

ALA ~ ' ~ TGHLY E

~ 50-

VAL • / // /

~

)

~ I S O L LEU÷

,

10

2'o

i

30

20

P

sb mM

Fig. 2. Summary of the effect of neutral amino acids on the signal height of the D M P O - O H spin-adduct. Control means those values obtained with 5 - 5 0 m M DMPO, whereas the other curves show the signal heights recorded in the presence of amino acids at concentrations equimolar to that of DMPO. Each symbol represents one measurement. Leucine could not be dissolved to the level of 250 m M stock solution; therefore, it was used as a 100 m M stock solution in buffer, and the volumes added are changed accordingly. The reproducibility of the control curve is ± 5.0% at each point, whereas for the tested substances it is ± 10%. The designations for each experimental curve from top to bottom represent the amino acids alanine, glycine, threonine, serine, valine, isoleucine and leucine, respectively.

100~

I

Control GUA

~

5 "~ 50t = i

cals. Histidine was extremely effective in scavenging ()H. This amino acid did not exert an effect in post-addition. We found that all of the sulfur containing amino acids considerably decreased the amount of O H spin-trapped (Fig. 4). But,. as Table I shows, both cysteine and cystine act in post-addition; but methionine does not. The low solubility of cystine did not allow a proper comparison of this compound; however, it reduced the signal height considerably even at 5 m M concentration. Methionine is a very effective OH scavenger even at low concentrations. As was noted previously, D M P O traps a radical of methionine induced by the reaction of this amino acid with OH. Aspartic acid as well as asparagine enhanced the amount of OH spin-trapped, even though at 50 m M they were acting upon the spin-adduct per se (Fig. 5). The enhancing effect of these amino acids was more at 30 to 40 mM. Glutamic acid and glutamine had very little effect on the amount of O H spin-trapped, but the aromatic amino acids even at low concentrations were very effective OH scavengers. It should be noted that tyrosine was tested only at very low concentrations due to its direct oxidizing effect on the ADP-Fe(II) complex (see below). Tryptophan at all concentrations al-

100

~

Control

; I

/ "I'/

~.__~....A3HOLYS

14L~S

1'0

20

3'0

4'0

LYS PLYS

~"

|

"

HI8

5~) n;M

Fig. 3. Summary of the effect of the basic amino acids and some related c o m p o u n d s on the signal height of the D M P O - O H spin-adduct. Explanation for the meaning of the curves is the same as in Fig. 2. Polylysine (PLYS) concentration was calculated as if it was lysine (LYS). The designations for each experiment from top to bottom are as follows: guanidine, y-aminobutyric acid, hydroxylysine, arginine, lysine, acetyllysine, polylysine and histidine, respectively.

1'0

2'0

3'0

4"0

6T0 mM

Fig. 4. Sum_maryof the effects of the su]fur-contalmng amino acids on the signal height of the D M P O - O H spin-adduct. Curves as in Fig. 2. Cystine (CYSTI + ) was added to only 1 - 5 m M concentration due its low solubility at neutral pH. CYSTE' refers to a concentration range from 2 to 10 mM. The designations M E T H and CYSTE mean methionine and cysteine, respectively.

244 TABLE I COMPARISON OF THE RADICAL SCAVENGER PROPERTIES OF THE AMINO ACIDS AND RELATED COMPOUNDS IN THE PRE- AND POST-ADDITION EXPERIMENTS IN PERCENTAGE OF PEAK HEIGHT OF THE DMPO-OH-ADDUCT MEASURED AT 50 mM DMPO Compound

Final concn. (mM)

Neutral amino acids Alanine 50 Glycine 50 Isoleucine 50 Leucine 50 Serine 50 Threonine 50 Valine 50 Basic amino acids and related compounds Arginine 50 N-e-Acetyllysine 50 y-Aminobutyric acid 50 Guanidine-HCl 50 Histidine 50 Hydroxylysine 50 Lysine 50 Polylysine 50 Sulfur-containing amino acids Cysteine 50 Cysteine 10 Cystine 5 Methionine 50 Aromatic amino acids Phenylalanine 50 Tryp tophan 50 Tyrosine 5 Dicarboxylic amino acids and their amines Aspartic acid 50 Asparagine 50 Glutamic acid 50 Glutamine 50 Proteins Bovine serum albumin 10.0 a Collagen 2.5 a Histone 2.0 a

Preaddition (%)

Post-addition (%)

122.5 102.5 21.3 16.3 63.8 97.5 51.3

102.5 77.5 46.3 67.5 58.8 67.5 61.3

41.3 20.0 78.8 92.5 13.8 48.8 31.3 20.0

71.3 61.3 100.0 78.8 100.0 72.5 41.3 85.0

0.0 48.8 70.0 20.0

0.0 51.3 100.0 95.0

5.0 2.5 25.0

82.5 68.8 41.3

107.5 97.5 85.0 87.5

67.5 68.8 62.5 93.8

68.8 102.5 53.7

87.5 67.5 91.3

a mg/ml.

most completely eliminated DMPO-OH-adduct formation. T h r e e p r o t e i n s w e r e t e s t e d (Fig. 6) in t h e p r e sent experiments. Histones showed excellent OH s c a v e n g e r ability, w h e r e a s b o v i n e s e r u m a l b u m i n w a s m u c h less e f f i c i e n t (see t h e c o n c e n t r a t i o n differences). T h e c o l l a g e n u t i l i z e d w a s p o o r l y s o l u b l e at n e u t r a l p H ; t h e r e f o r e , w e c o u l d n o t test h i g h e r c o n c e n t r a t i o n s o f it. U n d e r t h e s e c o n d i t i o n s , c o l -

l a g e n was n o t c o m p e t i t i v e w i t h D M P O w i t h res p e c t to t r a p p i n g O H radicals. T h e results o b t a i n e d w i t h p r o l i n e a n d h y d r o x y p r o l i n e are o m i t t e d here. T h e s e a m i n o a c i d s n o t o n l y r e d u c e d t h e signal h e i g h t o f t h e D M P O - O H a d d u c t , b u t a l s o g a v e rise to l o n g - l i v e d free r a d i cals t h e m s e l v e s w h e n r e a c t e d w i t h O H . T h e properties of these radicals were analyzed in detail a n d will b e p u b l i s h e d s e p a r a t e l y [56].

245

/•'•

ASPAC

10(

# J~

x 60-

a q) o.

~ ~/"

T _

~-o

Y

R

~

÷

~

~ -~PHEA ~ ~ T R Y P T .

~o

~o

go

5'o mu

"

Fig. 5. Summary of the effect of the aromatic and dicarboxylic amino acids and some amines on the signal height of DMPOOH spin-adduct. Explanation of the curves is the same as in Fig. 2. Tyrosine was added in concentrations from 1 to 5 mM, because it oxidized Fe(II) in concentratoins of 4 and 5 mM or higher. The designations for each experimental curve from top to bottom represents the amino acids, aspartic acid, asparagine, glutamine, glutamic acid, tyrosine, phenylalanine and tryptophan, respectively. A s m e n t i o n e d previously, all of the c o m p o u n d s e x a m i n e d were tested to d e t e r m i n e the effect of their a d d i t i o n after the D M P O - O H - s p i n - a d d u c t was a l r e a d y formed. T h e results are shown in

lOO

f

C

O

L

-~ 50-

L

~

HIST

#.

1()

2'0

3'0

4'0

I50 mM

Fig, 6. Summaryof the effects of someproteins on the signal height of DMPO-OH spin-adduct. The control curve is explained as in Fig. 2. The proteins were present in the following concentrations: bovine serum albumin (BSA), 2-10 mg/ml; collagen (COLL), 0.5-2.5 mg/ml; histone (HIST), 0.2-2.0 mg/ml; whereas the DMPO concentrations ranged from 10 to 50 mM as in the previous figures.

T a b l e I. T h e m a i n conclusion f r o m these experim e n t s is that s o m e a m i n o acids react with the n i t r o x y l radical o f the D M P O - s p i n - a d d u c t ; however, the pre- a n d p o s t - a d d i t i o n a l e x p e r i m e n t s are c o n s p i c u o u s l y different in some cases: it is striking that histidine, p h e n y l a l a n i n e , t r y p t o p h a n a n d m e t h i o n i n e a n d to s o m e w h a t less extent leucine, acetyllysine a n d p o l y l y s i n e p r o v e d to b e very efficient scavengers in the p r e a d d i t i o n model, whereas they were a l m o s t c o m p l e t e l y ineffective in the p o s t - a d d i t i o n system ( T a b l e I). It is also interesti n g that lysine was an almost equally efficient scavenger in pre- a n d p o s t - a d d i t i o n s ; whereas s o m e derivatives of it, acetyllysine, p o l y l y s i n e a n d hydroxyllysine, p r o v e d to b e m u c h less effective in p o s t - a d d i t i o n s . H i s t o n e s a n d b o v i n e s e r u m alb u m i n were less efficient scavengers in p o s t - a d d i tion, whereas collagen showed the reverse effect ( T a b l e I).

Measurements regarding the state of Fe(II) in the reaction mixtures F e r r o u s iron c a n be k e p t in dilute HCI for an e x t e n d e d p e r i o d in the d i v a l e n t state; however, at n e u t r a l p H it is oxidized quickly. This can be d e m o n s t r a t e d b y the ferrozine reaction. If A D P c o m p l e x e s iron, it is p r o t e c t e d against o x i d a t i o n at n e u t r a l p H for a relatively long time. If ferrozine is a d d e d to the r e a c t i o n m i x t u r e in place o f H 2 0 2 , i.e., 2 min after A D P a n d F e C l 2, were mixed, we f o u n d 85% of the i r o n in the ferrous state. If A D P was omitted, n o ferrous iron was d e t e c t e d at this time. U s i n g this model, we e x a m i n e d the effect of all the c o m p o u n d s s t u d i e d on the o x i d a t i o n state o f iron. W e f o u n d that if the stock solution o f the test c o m p o u n d was n e u t r a l i z e d to p H 7.0-7.5, n o n e o f the c o m p o u n d s c h a n g e d the o x i d a t i o n state of i r o n except for tyrosine at 4 m M conc e n t r a t i o n a n d higher. These o b s e r v a t i o n s allowed us to c o n c l u d e that the a m o u n t of available F e ( I I ) for catalysis of H202 was n o t affected b y the test c o m p o u n d s except for tyrosine. This is of i m p o r tance, since the limiting factor for s p i n - a d d u c t f o r m a t i o n in the A D P - F e ( I I ) - H 2 0 2 system is the a m o u n t of F e ( I I ) u n d e r the c o n d i t i o n s utilized [18]. O m i s s i o n of A D P f r o m the r e a c t i o n m i x t u r e allowed us to d e t e r m i n e w h e t h e r the test c o m p o u n d s were c o m p l e x i n g with i r o n a n d thus pre-

246 v e n t i n g it from b e i n g oxidized. E x p e r i m e n t s to test this idea revealed that the majority of the comp o u n d s tested did n o t protect iron from oxidation in the absence of A D P ; however, some of the c o m p o u n d s did. F o r instance, cysteine a n d g l u t a m i n e protected Fe(II) completely, whereas m e t h i o n i n e , serine, threonine, p h e n y l a l a n i n e a n d aspartic acid did so only partially (to 30-60%). Nevertheless, even in these cases Fe(II) r e m a i n e d i n a state such that it was oxidized b y H202 as s h o w n by the ferrozine reactions performed after the a d d i t i o n of H202. Irrespective of the presence or absence of A D P , n o Fe(II) was detected after the a d d i t i o n of H202. It should be n o t e d that the ferrozine-Fe(II) complex is not oxidized b y H202. T h a t is, one can be sure that the a d d i t i o n of H202 really oxidizes all of the iron present to Fe(II) before ferrozine is added.

The results with other spin traps As m e n t i o n e d in the description of the experim e n t a l procedures, 4-POBN, PBN a n d M N P were used for t r a p p i n g organic radicals formed d u r i n g radiolytic or photolytic reactions by n u m e r o u s authors. This allows us to answer the question whether the O H radicals formed in the A D P F e ( I I ) - H 2 0 2 system are able to transform the a m i n o acids (or other c o m p o u n d s ) into other radicals which are then able to react with the spin trap. F u r t h e r m o r e , a q u a n t i t a t i v e estimation of the a m o u n t of t r a p p e d radicals is also possible. Some typical E S R spectra are shown in Figs. 7 a n d 8. W i t h regards to spin trap 4-POBN, it is necessary to realize in each case that the ESR spectra of its reaction with O H radicals is also present in the spectrum (Fig. 7B). The spectra o b t a i n e d with 4 - P O B N a n d PBN show the formation of c a r b o n centered radicals. I n all cases, we can distinguish the t r a p p i n g of a d o m i n a n t radical giving rise to the most characteristic part of the E S R spectra a n d other radicals present in smaller a m o u n t in the spectrum. W e estimated the a m o u n t of the d o m i n a n t radical present a n d indicated also the presence of minor, or weak radicals in the spectrum. T h e results are presented in T a b l e II.

D [ F

Fig. 7. Some charactedstic results from the experiments with 4-POBN and PBN. Instrumental parameters were somewhat different from those used in the DMPO experiments. Namely, scan time was 8 rain, time constant 1.0 s and the receiver gain 1.25.104, except in (A), where it was 8.0.103, others were unchanged. (A) ESR spectrum obtained with 4-POBN when arginine was present in the ADP-Fe(II)-H202 system. (B) The same as (A) except that arginine was omitted. This curve indicates the direct reaction of 4-POBN with OH radicals. (C) Spectrum obtained when y-aminobutyric acid and 4-POBN were present. Note that the peaks are only slightly higher than in (B). (D) "t-Aminobutyricacid-PBN reaction where only very low peaks can be observed. (E) Guanidine-4-POBN reaction. The peak heights cannot be distinguished from those obtained in (B), i.e., guanidine is not reacting significantly. (F) PBNguanidine reaction, in which no peaks can be observed.

C Fig. 8. Further examples obtained in the experiments when 4-POBN and PBN were used as spin-traps. Instrumental parameters were as described in Fig. 7 except for the receiver gain, 8.0.103. (A) Alanine-4-POBN reaction. Note the similarity of this spectrum and that of Fig. 7A. (B) Alanine-PBN reaction. This spectrum is more complex than the usual ones obtained with PBN and other amino acids. (C) Reaction of alanine with the ADP-Fe(II)-H202 system in the absence of any spin trap. No peak can be observed.

247 The data included in Table II were obtained as follows: the largest dominant peaks were measured in each spectrum and normalized to a uniform receiver gain (1.25-10 4 ) and compared to the standard peaks of TMPN at identical receiver gain. In the case of 4-POBN the direct reaction w i t h O H r a d i c a l s w a s c a l c u l a t e d as a n a v e r a g e v a l u e o f 1 . 8 / ~ M r a d i c a l (S.E., s t a n d a r d e r r o r o f t h e m e a n , w a s 0 . 1 4 /~M f o r five m e a s u r e m e n t s ) , a n d this value was subtracted from the values obtained.

T h e d a t a i n T a b l e II s h o w t h a t r a d i c a l f o r m a t i o n w a s o b s e r v e d w i t h 4 - P O B N a n d P B N w i t h all test compounds except guanidine, polylysine and t h e p r o t e i n s . T h e a m o u n t o f t r a p p e d r a d i c a l s as calculated from the TMPN calibration experim e n t s w e r e d i f f e r e n t f o r t h e t w o t r a p s ( T a b l e II). T h e s p i n t r a p 4 - P O B N w a s u s e d a t 100 m M c o n c e n t r a t i o n , w h e r e a s P B N w a s u s e d a t 50 m M . Although 4-POBN usually spin-trapped more radicals, i n s o m e c a s e s P B N t r a p p e d m o r e . T h i s d e m onstrates that the chemical structure of the spin

TABLE 1I ESTIMATED AMOUNTS OF THE DOMINANT RADICALS TRAPPED BY 4-POBN AND PBN ON THE BASIS OF CALIBRATION WITH TMPN WITH INDICATION OF THE PRESENCE OF SOME WEAK RADICALS Tyrosine was measured at 3 mM and extrapolated to 10 mM; cysteine and cystine were measured at 10 and 5 raM, respectively, and extrapolated to 50 mM. All other compounds were measured in 50 mM concentration. The proteins studied did not give any significant peak with either of the spin traps. The intensity of the weak radicals was usually at least 5-10-times less than that of the dominant ones. ? indicates especially weak radicals. The values for 4-POBN do not contain the amount of directly trapped OH radicals (in average 1.8 #M). Compound

Radicals with 4-POBN Dominant (gM)

Neutral amino acids Alanine 3.1 Glycine 4.8 Isoleucine 7.2 Leucine 13.7 Serine 15.6 Threonine 15.8 Valine 6.1 Basic amino acids and related compounds Arginine 5.2 N-e-Acetyllysine 17.1 y-Aminobutyric acid 0.4 Guanidine-HCl 0.0 Histidine 2.4 Hydroxylysine 10.0 Lysine 5.3 Polylysine 0.0 Sulfur-containing amino acids Cysteine 11.2 Cystine 9.5 Methionine 8.4 Aromatic amino acids Phenylalanine 4.0 Tryptophan 2.0 Tyrosine 19.6 Dicarboxylic amino acids and their amines Aspartic acid 9.2 Asparagine 5.9 Glutamic acid 3.0 Glutamine 5.6

Radicals with PBN Weak (number)

Dominant (gM)

Weak (number)

1 1 1? 1? 1 1 1

8.3+3.7 3.1 2.5 3.0 13.2 7.9 5.9

1 1 1 1 1 1

1

2.6

1

-

5.0

1

-

0.7

-

-

0.0

-

1.4 5.3 3.9 0.8

1 1

1

1 1

4.8 6.0 32.4 1 1 1

2.6 6.6 12.8

1 1 1 1?

8.8 4.2 2.2 1.1

1 1 + 1? 1 1 1?

248 trap as well as the structure of the free radical dictates the efficiency of spin-trapping. The comparison of the spin-trapping results obtained with D M P O and the other spin traps m a y be performed on the following basis. As it has been shown by Floyd and Lewis [18], 50 m M D M P O was able to trap about 15 # M O H radicals when air was present in the ADP-Fe(II)-H202 system. Obviously, this is only a fraction of the total amount of O H radicals generated, since all of them would certainly not be trapped. Theoretically, we would expect 100 # M O H radicals, if each Fe(II) atom yields one O H radical when oxidized. The data in Table II show reasonable agreement with the D M P O experiments with regard to the amount of trapped radicals. It is interesting that the highest trapping rate was observed with PBN on OH-induced methionine free radicals (32.4 #M), whereas all other values were below 20 p M. It should be stressed that, although in some cases there is a clear correlation between the O H radical scavenging property of the amino acids, as determined in the D M P O model, with the result observed with the other spin-traps (e.g., leucine, valine, acetyllysine, GABA, guanidine, methionine, etc.), in other cases, no correlation was observed, i.e., histidine, phenylalanine, tryptophan and polylysine were very good scavengers in the D M P O system, but these proved to be much less efficient in yielding radicals trapped with 4-POBN and PBN (Tables I and II). These latter amino acids all have a cyclic structure, except for polylysine, which is a protein-like macromolecule, which may explain its weaker interaction with 4-POBN and PBN. It is very interesting that none of the compounds studied reacted even to the slightest extent with M N P under our experimental conditions. The ESR spectra were recorded within the usual time (1 rain) subsequent to the addition of the H202. However, if the reaction mixture was exposed to light for 1 - 3 h before the spectrum was recorded, characteristic peaks appeared with lysine and arginine, whereas alanine, glycine, histidine and m a n y other compounds did not show any peaks even after exposure to light. The reaction of lysine and arginine with M N P after long illumination demonstrates that the spin trap was present in

sufficient concentration in spite of its low solubility in water; however, the radicals generated by the ADP-Fe(II)-H202 system were not sufficient to activate this spin trap at all. Discussion The following facts must be kept in mind regarding the interpretation of the results presented in this paper. Assignment of the ESR spectra to the D M P O O H spin-adduct is well established and accepted [4,17,18,52-55]. The use of the ADP-Fe(II)-H202 system for generating O H radicals and their trapping by D M P O yields highly reproducible results which allowed us to determine the amount of O H radicals trapped if the experimental conditions and timing schedule was adhered to rigorously. Certain compounds, the hydroxyl free radicals scavenger properties of which has been established, decrease the signal height of the D M P O - O H adduct [18]. (2) The experimental results obtained by using the Ferrozine reaction to determine the oxidation state of iron in the reaction mixture confirms that with only one exception, tyrosine, the compounds involved in the present experiments did not oxidize Fe(II) in the presence of ADP, i.e., the amount of available Fe(II) for oxidation was not decreased. On the other hand, we also obtained evidence that all the Fe(II) becomes oxidized at neutral p H in the absence of A D P and in the presence of most of the compounds studied. The exceptions were cysteine, glutamine, methionine, serine, threonine, phenylalanine and aspartic acid, which protected Fe(II) from oxidation either completely or partially. Furthermore, it should also be kept in mind that Fe(II) in the ADP-iron complex remained oxidizable in all cases, i.e., none of the compounds tested formed a strong complex with Fe(II) which prevented it from reacting with H202. With these points in mind, the effect of the compounds studied on the amount of D M P O - O H adduct present can be considered as a complex result of two major processes. Namely, there is a direct interaction of the O H radicals formed by the ADP-Fe(II)-H202 with the test compounds resulting in competition with DMPO. In addition, the test compounds directly interact with the

249

nitroxyl radical of the D M P O - O H adduct itself. A confirmation of the reaction of OH with the test substances is supported by the results obtained with the two other spin-traps, 4-POBN and PBN. The radicals detected by these spin traps cannot be explained other than by assuming that the compounds were attacked by the oxygen radicals of the ADP-Fe(II)-H202 system, and the radicals thus formed reacted thereafter with the spintraps. ADP does not play any role in these reactions directly with the exception that it stabilizes Fe(II) at neutral pH. This was demonstrated by the fact that the very same spectra can be obtained both with DMPO and the other spin traps if ADP is omitted and Fe(II) is added to the system as the last component. The reactivity of the compounds tested with the nitroxyl radical of DMPO is supported by the results showing that many of them decreased the signal height in post-addition experiments. It is necessary to realize, however, that the formation of the D M P O - O H complex is not entirely finished by 30 s after the addition of H202, and therefore, OH-radical trapping by the added compounds may take place even in the post-addition experiments. Nevertheless, there is good evidence for supporting the assumption that the reactions are different in the pre- and post-addition experiments. Namely, some of the compounds tested displayed a strong scavenger capacity in the preaddition, whereas they proved to be much less efficient in post-additions, or not at all (e.g., histidine, phenylalanine, tryptophan, histone, acetyllysine, polylysine, etc.), whereas others behaved exactly opposite (e.g., glycine, threonine, aspartic acid, collagen, etc.). Finally, it should be mentioned that both reactions may exist simultaneously, with the molecular structure of the compounds dictating whether one or the other reaction occurs to a different extent, thus resulting in the seemingly confusing pattern shown by Table I. The lack of radical formation with M N P deserves comments. Although it has not been explored deeply enough during the present experiments, it is interesting to note that in previous experiments with MNP, to the best of our knowledge, it was always used in irradiation systems. Therefore, it is possible that its reaction with amino-acid radicals requires some of the radicals

that are produced by radiolysis (e.g., solvated electrons) which are not present in the ADP-Fe(II)H202 system. Our observations with some compounds i.e., lysine and arginine, after long exposure to light supports this assumption. The possible biological significance of our findings is important. We believe that ADP-Fe(II) complexes may well be the source of OH radical formation in vivo, since such complexes do exist [20] and they are involved in physiological transport processes of iron [21,22], and also mitochondrial respiration produces H202 and superoxide [2-4,57]. Amino acids and proteins may be the target of OH radical attack in living cells. This may result in changes of protein conformation, as has been shown in the case of glutamine synthetase [14,15], or may even destroy some part of the amino acid pool. It is interesting to stress that histidine proved to be one of the best scavengers in preaddition experiments, supporting the possibility of oxidative damage on the histidine groups of glutamate synthetase [14,15]. Furthermore, the involvement of free-radical damage in cellular aging [6,7,12,13] is supported by the present experimental results. Another consequence of the results presented here is that the radial scavenger properties of the proteins may be an important aspect of their function. For example, it is interesting that histones containing high amounts of arginine and lysine proved to be much better scavengers than bovine serum albumin, which is low in these amino acids. Histones therefore may be protecting DNA against free radical attack. Acknowledgements The work presented here was supported partially by money provided by an N I H grant AG02599. The authors express their appreciation to Anita Hill, Lisa Gropf, Kay Wallace, Sandra Nank and other scientists on the Oklahoma Medical Research Foundation staff for help in the conduction of the experiments and preparation of the manuscript. References 1 Fridovich, I. (1978) Science 201,875-880 2 Nohl, H. and Hegner, D. (1978) Eur. J. Biochem. 82, 563-567

250 3 Nohl, H. and Hegner, D. (1978) Eur. J. Biochem. 111, 203-210 4 Nohl, H., Jordan, W. and Hegner, D. (1981) FEBS Lett. 123, 241-244 5 McCord, J.M. (1974) Science 185, 529-531 6 Zs.-Nagy, I. (1978) J. Theor. Biol. 75, 189-195 7 Zs.-Nagy, I. (1979) Mech. Ageing Dev. 9, 237-246 8 McCord, J.M., Wong, K., Stokes, S.H., Petrone, W.F. and English, D. (1980) Acta Physiol. Scand. Suppl. 492, 25-30 9 Petkau, A. (1980) Acta Physiol. Scand. Suppl. 492, 81-90 10 Siesjo, B.K., Rehncrona, S. and Smith, D. (1980) Acta Physiol. Scand. Suppl. 492, 121-128 11 Tolmasoff, J.M., Ono, T. and Cutler, R.G. (1980) Proc. Natl. Acad. Sci. USA 77, 2777-2781 12 Zs.-Nagy, I. and Nagy, K. (1980) Mech. Ageing Dev. 14, 245-251 13 Zs.-Nagy, I. and Semsei, I. (1984) Exp. Gerontol. 19, 171-178 14 Levine, R.L. (1983) J. Biol. Chem. 258, 11823-11827 15 Levine, R.L. (1983) J. Biol. Chem. 258, 11828-11833 16 Janzen, E.G. (1971) Acc. Chem. Res. 4, 31-40 17 Janzen, E.G. (1980) in Free Radicals in Biology (Pryor, W., ed.), Vol. IV, pp. 115-154, Academic Press, New York 18 Floyd, R.A. and Lewis, A. (1983) Biochemistry 22, 2645-2649 19 Floyd, R.A. (1983) Arch. Biochem. Biophys. 225, 263-270 20 Hochstein, P. (1981) Isr. J. Chem. 21, 52-53 21 Konopka, K. (1978) FEBS Lett. 92, 308-312 22 Konopka, K. and Romslo, I. (1980) Eur. J. Biochem. 107, 433-439 23 Hochstein, P., Nordenbrand, K. and Ernster, L. (1964) Biochem. Biophys. Res. Commun. 14, 323-328 24 Lagercrantz, C. and Forshult, S. (1968) Nature 218, 1247 - 1248 25 Taniguchi, H. and Hatano, H. (1974) Chem. Lett. 531-534 26 Taniguchi, H. and Hatano, H. (1975) Chem. Lett. 9-12 27 Rustgi, S., Joshi, A., Moss, H. and Riesz, P. (1977) Int. J. Radiat. Biol. 31,415-440 28 Rustgi, S., Joshi, A. and Riesz, P. (1977b) Int. J. Radiat. Biol. 32, 533-552 29 Rustgi, S. and Riesz, P. (1978) Int. J. Radiat. Biol. 34, 301-316 30 Rustgi, S. and Riesz, P. (1978) Int. J. Radiat. Biol. 34, 449-460 31 Moriya, F., Makino, K., Suzuki, N., Rokushika, S. and Hatano, H. (1982) J. Am. Chem. Soc. 104, 830-836 32 Makino, K. and Riesz, P. (1982) Int. J. Radiat. Biol. 41, 615-624

33 Stookey, L. (1970) Anal. Chem. 42, 779-781 34 Carter, P. (1971) Anal. Biochem. 40, 450-458 35 Janzen, E.D., Wang, Y.Y. and Shetty, R.V. (1978) J. Am. Chem. Soc. 100, 2923-2925 36 Janzen, E.G. and Blackburn, B.J. (1969) J. Am. Chem. Soc. 91, 4481-4490 37 lwahashi, H., Ishikawa, Y., Sato, S. and Koyano, K. (1977) Bull. Chem. Soc. Japan 50, 1278-1281 38 Uda, T., Kazusaka, A., Howe, R.F. and Keulks, G.W. (1979) J. Am. Chem. Soc. 101, 2758-2759 39 Kremers, W. and Singh, A. (1980) Can. J. Chem. 58, 1592-1595 40 Janzen, E.G., Stronks, H.J., Nutter, D.E., Jr., Davis, E.R., Blount, H.N., Poyer, J.L. and MacCay, P.B. (1980) Can. J. Chem. 58, 1596-1598 41 Augusto, O., Ortiz de Montellano, P. and Quintanilha, A. (1981) Biochem. Biophys. Res. Commun. 101, 1324-1330 42 Niki, E., Yokoi, S., Tsuchiya, J. and Kamiya, Y. (1983) J. Am. Chem. Soc. 105, 1498-1503 43 Torsell, K. (1970) Tetrahedron 26, 2759-2773 44 Lagercrantz, C. (1971) J. Phys. Chem. 75, 3466-3475 45 Lagercrantz, C. (1973) J. Am. Chem. Soc. 95, 220-225 46 Sargent, F.P. and Gardy, E.M. (1974) Can. J. Chem. 52, 3645-3650 47 Sargent, F.P. and Gardy, E.M. (1976) J. Phys. Chem. 80, 854-856 48 Bolton, J.R., Chen, K.S., Lawrence, A.H. and De Mayo, P. (1975) J. Am. Chem. Soc. 97, 1832-1837 49 Joshi, A., Rustgi, S. and Riesz, P. (1976) Int. J. Radiat. Biol. 30, 151-170 50 Sargent, F.P. (1977) J. Phys. Chem. 81, 89-90 51 Rustgi, S. and Riesz, P. (1978) Radiat. Res. 75, 1-17 52 Janzen, E.G. and Davis, E.R. (1982) in Free Radicals and Cancer (Floyd, R.A., ed.), pp. 397-422, Marcel Dekker, New York 53 Harbour, J.R., Chow, V. and Bolton, J.R. (1974) Can. J. Chem. 52, 3549-3553 54 Floyd, R.A., Soong, L.M., Stuart, M.A. and Reigh, D.L. (1978a) Arch. Biochem. Biophys. 185, 450-457 55 Floyd, R.A., Soong, L.M., Stuart, M.A. and Reigh, D.L. (1978b) Photochem. Photobiol. 28, 857-862 56 Floyd, R.A. and Zs.-Nagy, I. (1984) Biochim. Biophys. Acta 790, 94-97 57 Loschen, G., Flohe, L. and Chance, B. (1971) FEBS Lett. 18, 261-264 58 Floyd, R.A. (1983) Biochim. Biophys. Acta 756, 204-216 59 Walling, C. (1975) Acc. Chem. Res. 8, 125-131