Chem. Res. Toxicol. 1988,1, 97-100
97
Articles Reactivity of Perhydroxyl (HOO') with 1,4-Cyclohexadiene (Model for Allylic Groups in Biomembranes) Donald T. Sawyer,* M. Steven McDowell, and Kenneth S. Yamaguchi Department of Chemistry, Texas A&M University, College Station, Texas 77843 Received July 2, 1987
T h e electron-transfer reduction of molecular oxygen yields superoxide ion (02*-), which reacts with proton sources t o form H02'. In water the latter species disproportionates via reaction with 02'-(Itbi, 108 M-I 8-l) and itself (Itd, loe M-' s-l). T h e rate constants (kd) for the homolytic disproportionation process (H02' H02' H202 02),which have been determined from stopped-flow spectrophotometric decay data for H02' a t 25 "C,are (1.7 f 0.5) X lo4 M-' s-' in dimethyl sulfoxide (Me2SO), (5.3 f 0.5) X lo4 M-' s-l in dimethylformamide (DMF), and approximately lo7 M-' s-l in acetonitrile. With limiting fluxes of protons t o control the rate of formation of HOz' from 02'-,the rate of decay of H02' is enhanced by reaction with the allylic hydrogens of excess 1,4cyclohexadiene (RH). On the basis of such data the apparent second-order process (H02' RH R' H202)has a rate constant (h0J of (1.6 f 0.6) X lo2 M-' s-l. T h e reactivity of H02' decreases as its solvation energy increases.
+
-
+
-
+
The biogeneration of superoxide ion (02.-,from reduction of dioxygen in aerobic systems) appears to have led to the evolution of a family of metalloproteins, the superoxide dismutases (SOD), that catalyze ita disproportionation and thereby protect the organism from 02'toxicity (1-4). These proteins remove 02'-and limit its concentration t o less than lo* M (th uncatalyzed proton-induced disproportionation of 02'-at pH 7 in water has an observed second-order rate constant of lo6 M-' s-', which limits the concentration of 02.-to about lo4 M) (5). The absence of chemically authenticated cytotoxic reactions for 02'-(6, 7) and an enzymatic effect of only 2 orders of magnitude (with respect to lowering the 02.concentration) have prompted us to seek evidence for a more unique protective function for SOD enzymes (8). The perhydroxyl radical (H02*)abstracts hydrogen atoms from allylic groups in substrates such as 1,4-cyclohexadiene (1,CCHD) (9) and thereby initiates their autooxidation and peroxidation. This hydrogen atom abstraction represents a model for the H02'-initiated autooxidation of lipid (linoleic and arachidonic acid esters) membranes in biology. Although the reactivity of superoxide (02*-) and of its conjugate acid, perhydroxyl radical (HO,'), has been extensively studied in aqueous solution (10);the chemistry of HOz' in dipolar aprotic solvents (model matrices for biomembranes) has not been characterized. With aqueous media pulse radiolysis provides a convenient means to generate superoxide ion (11, 12) O2
+ e-(aq)
-
02'-
(1)
which reacts with available protons to form HOP 02'-
+ H30+
.-+
HO2'
+ H2O
+
PK,, 4.8
(2)
This species is unstable and rapidly decomposes via
heterolytic and homolytic disproportionation. HO2'
+ 02'kbi,
Hz0
H202
+ 02 + -OH
-
1 X lo8 M-' s - ~
HO2' + HO2' H202 + 0 2 8.5 X 10' M-' s-l (25 "C)
(3)
(4)
kd,
Another route for the formation of 02'-and H02' is UV irradiation of HOOH in aqueous (13) and alcohol-water (14,15) solutions. An earlier investigation (16) has demonstrated that the rate of protonation of 02*by Br~nstedacids in acetonitrile (MeCN), dimethylformamide (DMF), and dimethyl sulfoxide (Me2SO) is proportional to their acidity.
+ HA
02.-
+
HO2'
+ A-
k5m
KHA
(5)
Another route to the formation of H02' is the electrolytic oxidation of H202in acetonitrile (17).
-
H z 0 ~
HO;
+ H + + e-
An early kinetic study (9) of the radical-initiated autoxidation of 1,4-cyclohexadiene (1,4-CHD) in dipolar solvents had a chain-termination step via the homolytic disproportionationof H02'. The results demonstrated that the rate constants for the propagation step (H02' + 1,4CHD) and for the termination step (H02' + H02') are strongly dependent on the solvent. The present paper summarizes the results for the direct measurement of the kinetics for the reaction of H02' with itself and with 1,4-CHD in Me2S0, DMF, and MeCN so-
0893-228x/88/2701-0097$01.50/0@ 1988 American Chemical Society
98 Chem. Res. Tonicol., Vol. 1, No. 2, 1988
Sawyer et al.
Table I. Reactivity of HOz' in Various Solvents at 25 O C solvent W+I, mM 10-4kObed, M-1 8-1 kd,' M-' S-' notes A. Rate Constants for the Second-Order Disproportionation (kd) of HOz' [3 mM 02'Plus H+ (HClO, or NH,ClO,)]. 1. MezSO 16 3.0 f 0.6 23 2.6 f 0.6 35 4.4 f 0.4 41 4.0 f 0.8 88 (NH,') 2.8 f 0.2 94 3.8 f 1.0 4.2 f 0.4 140 234 4.6 f 0.4 avg 3.4 f 1.0 (1.7f 0.5) X lo4 this work 23 10.8 f 0.6 2. DMF 41 10.2 f 1.6 91 10.8 f 0.2 avg 10.6 f 1.0 (5.3 f 0.5) x 104 this work >1000 >5 x 106 this work 3. MeCN (1.0f 0.5) x 107 ref 17 (5 f 1) x 106 ref 16 4.3 X lo6 (30 "C) ref 9 6.3 X lo8 ref 9 4. PhCl (8.3 f 0.7) X lo6 ref 10 5. H20 solvent 1. MeoSO
B. Rate Constants for the Second-Order Reaction of HO,' (3 mM) with 1,4-Cyclohexadiene(k,J [H+l,mM [1,4-CHD],mM lo4 Ykobdn,M-' s-l k,,, M-l.9-l
56
212
a4
108
a4
212
4.0 f 0.4 4.2 f 0.6 4.4 f 0.6
232 148
avg (1.6 f 0.6) X lo2 3.5 x 102
2. MeCN
notes
89
1.4 x 103
3. PhCl
this work ref 9 ref 9
# k o =~2 k,; from the kinetic analysis and monitoring of [HO,'] and the integrated rate equation.
lutions. Product analysis for the 1,4-CHD substrate over a range of experimental conditions confiis that the initial step is rate limiting and involves abstraction of a n allylic hydrogen atom.
Experlmentai Section Instrumentation. The rates of decay of HO; in MeCN, DMF, and Me2S0were measured by recording the decrease in absorbance at 250-320 nm on a Durrum Model 216 stopped-flow spectrophotometer, which was thermostatically controlled at 25 f 1 OC. All reactions, including those with excess 1,4-CHD, followed second-order kinetics. The data were collected and processed with a Nicolet Model 1047 2000-channel signal averager interfad to an Apple IIe microprocessor. The data were analyzed by the use of a nonlinear least-squares computer program to evaluate the second-order rate constant, k o w , from the relation At = (A,, t A [H02*]&,,~t)/(l+ [HO2*]&,~t), where A is the absorbance. The observed and least-squares-fit values were in congruence for at least four half-lives. Solution spectra were recorded with a Varian Model 219 UV-vis spectrophotometer and a Hewlett-Packard Model 8451A diode array UV-vis spectrophotometer. The concentrations of 02'-were determined by measurement of the voltammetric peak-current at glassy-carbon electrodes with a BioAnalytical Systems Model CV-27 voltammograph. The reaction efficienciesand product distributions for the combination of H02' and 1,4-CHD were determined by analysis with a Hewlett-Packard Model 5880 gas chromatograph that was equipped with a 12.5-m capillary column. Reagents and Solvents. Acetonitrile (MeCN), dimethyl formamide (DMF),and dimethyl sulfoxide (MeaO) were obtained from Burdick and Jackson ("distilled in glass" spectra grade) and were used without further drying or purification. The 1,4cyclohexadiene (1,CCHD) (Aldrich Chemical)was assayed by gas chromatography to be 96% pure with less than 2% total of 1,3cyclohexadieneand benzene and was used as received. NH4C104 (GFS Chemical) was dried overnight on a vacuum line and then used without further purification. Doubly distilled water was further purified by passage through a Milli-Q filtration system. Stock solutions of perchloric acid were prepared by the dilution of 5.8 M HC104 (1:l HC104 (70%)/H,O) in the appropriate
solvent. Warning: Addition of concentrated HC104 (70%)to MezSO can lead t o violent explosions. Stock solutions of 0,'- were prepared either by controlled potential electrolytic reduction of O2at a platinum-mesh electrode or from weighed amounts of pure tetramethyl ammonium superoxide [Me4N(02)](18). Lifetimes of the 0;- solutions were significantlyenhanced by pretreatment of the glassware by soaking in 1M tetramethylammonium or tetrabutylammonium hydroxide solutions for 24 h and then rinsing with the appropriate solvent. Methods. The reaction efficiencies for the HOz'/ 1,4-CHD reactions were determined from the slow combination of a 10-mM solution of 0,'- to an MezSO solution that contained 1,QCHD and the proton source (HC104or H20). Alternatively, the proton source was added slowly to the 02'-solution that contained the substrate (blank experimentsdemonstrated that benzene was not produced in a solution of 0,'- and 1,4-CHD). Identical results were achieved regardless of the order of addition of reagents. The kinetic experiments involved the rapid mixing of deoxygenated solutions of 6 mM 0;- and the proton source (and excess 1,4-CHD for the substrate studies). Experiments also were conducted with 1,4-CHDpresent in both solutions prior to mixing; identical results were achieved.
Results In dipolar aprotic solvents proton sources induce the rapid disproportionation of superoxide ion via initial formation of HOP' (16). The combination of excess HC104 in Me2S0, DMF, or MeCN with a second solution of 02'results in the latter's rapid and complete conversion to H202and 02.Table IA summarizes the observed kinetic data from the measured decay of HOz' via stopped-flow spectrophotometry. The kobsdvalues, which are best-fit values of a least-squares nonlinear regression method, are the average of three or four runs for each combination of reactants. The kobsdvalues also are independent of the monitoring wavelength and the proton source (HC104or NH4C104). When limiting fluxes of protons are provided to solutions of 02'-and excess 1,4-cyclohexadiene (1,4-CHD), the buildup of HOz' is controlled. Because its disproportion-
Chem. Res. Toxicol., Vol. 1, No. 2, 1988 99
Reactivity of Perhydroxyl with 1,4-Cyclohexadiene
Table 11. Oxidation of 1,4-Cyclohexadiene (1,4-CHD) by HOz' (02'-P l u s HA)i n Dimethyl Sulfoxidea product' distribution, proton % source [HA11 [oz'-19 [1,4-CHD], reaction*" (HA) mM mM mM efficiency, % 1,3-CHD PhH 25 74 26 100 5.3 10.6 H2O 90 92 8 1000 3.3 6.6 H20 100 70 30 H20d 100 4.0 8.1 100 79 21 HCIOI 1.6 3.2 6.4 69 31 3.2 3.2 6.4 92 HC104 54 46 6.4 3.2 6.4 31 HC104 75 25 8.4 8.4 17.0 40 HC104
a A 10 mM Oz'- solution (Me2SO) was added slowly to an MezSO solution that contained 1,4-CHD and the proton source (HA). The indicated concentrations represent the initial values after mixing. 100% represents the reaction of one 1,4-CHD molecule per 02.- added. 'After the product solution (10 mL) was cooled in an ice-water bath, 100 mL of ice-cold water was added, and the mixture was extracted with diethyl ether. A 0.5-mL portion of the extract was analyzed by capillary-column gas chromatography. The reaction efficiencies and product distributions were calculated on the basis of the integrated peak areas and by adding a known amount of toluene or benzene as an internal standard. 5 mL of a Me2S0 solution that contained 200 mM HzO was mixed slowly with 5 mL of a MezSO solution that contained 8 mM 02'-and 16.2 mM 1,4-CHD.
ation is a second-order process, such limiting conditions can favor hydrogen atom abstraction from 1,4-CHD by HOz'. This is especially so for Me2S0, in which HOz' has its slowest rate of disproporationation. Table I1 summarizes the reaction efficiencies and product distributions that result from the slow addition of a 10 mM OZcsolution (MezSO) to MezSO solutions that contain 1,4-CHD and a proton source (HzO or HC10$. For some conditions, all of the 02'-, after protonation, is consumed via oxidation of 1,CCHD. The product distribution favors 1,3-CHD for conditions that enhance the reaction efficiency. The primary product, (C6H7)' will disproportionate to give equal amounts of 1,3-CHD and benzene, but the presence of excess 1,CCHD favors an exchange reaction [l,CCHD (C,&)' 1,3-CHD + (C,H,)'] and accounts for the high ratios of 1,3-CHD in Table 11. The rate of decomposition for HOz+ is enhanced by the presence of a large excess of 1,4-CHD. Table IB summarizes the observed rate constants for the apparent second-order disproportion of the HOP' that results when 6 mM 02'-is mixed with a solution that contains various amounts of protons and 1,4-CHD.
Table 111. Rate Constants for t h e Second-Order Reaction of HO,' with Organic Substrates (k".) solvent k,,, M-'s-l substrate notes 1. 1,4-CHD MezSO 1.6 X lo2 this work 3.5 x 102 ref 9 MeCN 1.4 X lo9 ref 9 PhCl 1.2 x 103 ref 19 2. linoleic acid 85:15 EtOH/H20 3. arachidonic acid 85:15 EtOH/H20 3.0 x 103 ref 19 4. a-tocopherol 85:15 EtOH/H20 2.0 x 106 ref 20 1.8 x 104 ref 21 5. cysteine HZO 1.6 x 104 ref 22 6. ascorbic acid HzO
Discussion and Conclusions
The presence of excess 1,4-CHD enhances the rate of disappearance of HO; because of its parallel oxidation of the allylic hydrogen to give a 'C6H7 radical that disproportionates to 1,3-CHD and benzene (Table 11). HO2' + C6Hs 'CsH7 + HzOz (9)
+
-
An earlier study (16) has established that the addition of one-half of an equivalent of protons to 02'-in aprotic media results in the rapid disproportionation of one-half of the 02'The observed rapid disappearance of HOz' that results from the combination of Oz'- with excess strong acid in MezSO or DMF (Table IB) obeys a second-order rate law over a t least four half-lives
Because the rate of disappearance for H02' is the observed quantity, there is a stoichiometric factor of two. The disproportionation rate constants (kd) from these measurements are summarized in Table IA. The rate in MeCN is so rapid that only a lower limit can be determined from the stopped-flow measurements (16). Independent evaluations for k d in MeCN (9,16,17), PhCl (91,and H 2 0 (10) also are included in the tabulation. On the basis of these data the relative rates for the homolytic disproportionation of HOz' in the various solvents follow the order PhCl > MeCN > H 2 0 > DMF > Me2S0. Thus, the rate is diminished by increased solvent viscosity and/or solvation energy of HOz'.
Scheme I 1/2 H 2 0 2
o;.+HA
&A
+
+ 1/2 O2
HOi
1/2 1 . 3 - C H D i 1/2PhH
I
L " ox
/
1.4-CHD
1.3-CHD + ICgH71' C6H700H
+ iC6H7]
+
1'4-CHD
'
1
C6H703,
%
C6H7 OOH
PhH
+ HO2'
-
When such systems are subjected to kinetic study in MezSO by stopped-flow spectrophotometry, the observed decay rate for H02' obeys the second-order rate law of eq 8 to give the k o b d values summarized in Table IB. If the decay of HOz' results from the parallel processes of homolytic disproportionation (eq 7) and the second-order oxidation of allylic hydrogen (eq 9), then the rate law should be the s u m of those for the independent processes.
2kd[H02']2 + ko,[H02'][1,4-CHD] (10) This relation can be arranged to give
Treatment of the observed rate constants in Table IB with this relation provides evaluations of kOx.Table I11 summarizes values for k,, in three solvents (MezSO, MeCN, and PhC1) for the oxidation of 1,CCHD by H02'. A reasonable reaction sequence (Scheme I) involves the initial formation of H02' (at diffusion-controlled rates for
100 Chem. Res. Toxicol., Vol. 1, No. 2, 1988
strong acids), with its subsequent disproportionation ( k d ) or attack of 1,4-CHD (kox). As with the rates of disproportionation, the relative rate of oxidation of 1,4-CHD by HO; in the three solvents is PhCl > MeCN > Me2S0. For Me2S0 the value of k,, is slightly greater than kd’/2, but with MeCN and PhCl the values for k,, are an order of magnitude lower than their kd1i2values. Hence, the propensity of H02’ to initiate autoxidation of allylic groups is greater in Me2S0 and occurs at much lower concentrations of substrate. The data of Table I1 confirm that with limiting proton fluxes essentially all of the 02’-generated in an aprotic matrix can react with allylic groups. Hence, protection by more reactive antioxidants appears to be essential in biological membranes. Rate constants for the oxidation of the allylic hydrogens of linoleic acid and arachidonic acid and of the heteroatom hydrogens of a-tocopherol, cysteine, and ascorbic acid in protic media also are summarized in Table I11 (9, 19-22). Again, the rate constants for allylic hydrogen oxidation of linoleic acid and arachidonic acid in protic media are slightly greater than (kd’/’)H,o (eq 4). The large rate constants (h0J for the reaction of H02’ with the antioxidants (a-tocopherol, cysteine, and ascorbic acid) is consistent with their protective role in biology. In summary, a likely biological function for the SOD enzymes is to remove 0;- and thereby preclude formation of H02’. By their removal of the 02’-they prevent formation of HO; and its initiation of lipid peroxidation and autooxidation. 1,4-Cyclohexadiene (with two allylic groups) is a convenient model substrate for the allylic groups of the linoleic and arachidonic acid esters in lipid membranes. Recent results (8) confirm that an SOD model complex can block the formation of HOP’ and thereby prevent reaction with 1,4-CHD.
Acknowledgment. This work was supported by the National Science Foundation under Grant CHE-8516247 and the Welch Foundation under Grant A-1042. Registry NO.1,4-CHD, 628-41-1; 1,3-CHD, 592-57-4; DMSO, 67-68-5; DMF, 68-12-2; HOO’, 3170-83-0; CHSCN, 75-05-8; CsH6, 71-43-2.
References (1) McCord, J. M., and Fridovich, I. (1968) “The reduction of cytochrome c by milk xanthine oxidase”. J . Biol. Chem. 243, 5753-5760. (2) McCord, J. M., and Fridovich, I. (1969) “Superoxide Dismutase; An Enzymic Function for Erythrocuprein (Hemocuprein)”. J . Biol. Chem. 244,6049-6055. (3) Michelson, A. M., McCord, J. M., and Fridovich, I., Eds. (1977) Superoxide and Superoxide Dismutases, Academic: New York. (4) Bannister, J. V., and Hill, H. A. 0. (1980) “Chemical and biochemical aspects of superoxide and superoxide dismutase”. In
Sawyer et al. Developments in Biochemistry, Vol. IIA, Elsevier, Amsterdam/ New York. (5) Sawyer, D. T., and Valentine, J. S. (1981) “How super is superoxide?”. Acc. Chem. Res. 14, 393-400. (6) Fridovich, I. (1982) “How innocuous is superoxide?”. Acc. Chem. Res. 15, 200. Sawyer, D. T., and Valentine, J. S. (1982) “Rebuttal”. Acc. Chem. Res. 15, 200. (7) Fee, J. A. (1980) “Superoxide, superoxide dismutase and oxygen toxicity”. In Metal Zon Activation of Dioxygen (Spiro, T. G., Ed.) pp 209-237, Wiley, New York. (8) Yamaguchi, K. S., Spencer, L., and Sawyer, D. T. (1986) “Tris(picolinato)manganese(II): a chemical model for the mechanism and function of mitochondrial superoxide dismutase“. FEBS Lett. 197, 249-252. (9) Howard, J. A., and Ingold, K. U. (1967) “Absolute rate constants for hydrocarbon autoxidation. V. The hydroperoxy radical in chain propagation and termination”. Can. J . Chem. 45,785-792. (10) Bielski, B. H. J., Cabelli, D. E., Arudi, R. L., and Ross, A. B. (1985) “Reactivity of H02/02- in aqueous solution”. J. Phys. Chem. Ref. Data 14, 1041-1060. (11) Draganic, I. G., and Draganic, Z. D. (1971) The Radiation Chemistry of Water, Academic, New York. (12) Bielski, B. H. J., and Arudi, R. L. (1983) “Preparation and stabilization of aqueous/ethanolic superoxide solutions”. Anal. Biochem. 133, 170-178. (13) Nadezhdin, A., and Dunford, H. B. (1979) “Oxidation of nicotinamide adenine dinucleotide by hydroperoxide radical”. J. Phys. Chem. 83, 1957-1961. (14) Bielski, B. H. J., and Gebicki, J. M. (1982) “Generation of superoxide radicals by photolysis of oxygenated ethanol solutions”. J . Am. Chem. SOC.104, 796-798. (15) McDowell, M. S., Bakac, A., and Espenson, J. H. (1983) “A convenient route to superoxide ion in aqueous solution”. Znorg. Chem. 22, 847-850. (16) Chin, D.-H., Chiericato, G., Jr., Nanni, E. J., Jr., and Sawyer, D. T. (1982) “Proton-induced disproportionation of superoxide ion in aprotic media”. J . Am. Chem. SOC.104, 1296-1299. (17) CofrB, P., and Sawyer, D. T. (1986) “Redox chemistry of hydrogen peroxide in anhydrous acetonitrile”. Inorg. Chem. 25, 2089-2092. (18) Yamaguchi, K., Calderwood, T. S., and Sawyer, D. T. (1986) “Corrections and additional insights to the synthesis and characterization of tetramethylammonium superoxide [(Me4N)O2]”. Inorg. Chem. 25, 1289-1290. (19) Bielski, B. H. J., Arudi, R. L., and Sutherland, M. W. (1983) “A study of the reactivity of HOz/OF with unsaturated fatty acids”. J . Biol. Chem. 258, 4759-4761. (20) Arudi, R. L., Sutherland, M. W., and Bielski, B. H. J. (1983) “Reactions of H02/0; with a-tocopherol in ethanolic solutions”. In O x y Radicals and Their Scavenger Systems (Cohen, G., and Greenwald, R. A., Eds.) Vol. 1, pp 26-31, Elsevier Biomedical, New York. (21) Al-Thannon, A. A., Barton, J. P., Packer, J. E., Sims, R. J., Trumbore, C. N., and Winchester, R. V. (1974) “The radiolysis of aqueous solutions of cysteine in the presence of oxygen”. Znt. J. Radiat. Phys. Chem. 6 , 233-248. (22) Cabelli, D. E., and Bielski, B. H. J. (1983) “Kinetics and mechanism for the oxidation of ascorbic acid/ascorbate by H02/O; radicals. A pulse radiolysis and stopped-flow photolysis study”. J . Phys. Chem. 87, 1809-1812.
