Thermodynamic and kinetic considerations for the reaction of semiquinone radicals to form superoxide and hydrogen peroxide

Abstract

The quinone/semiquinone/hydroquinone triad (Q/SQ(*-)/H(2)Q) represents a class of compounds that has great importance in a wide range of biological processes. The half-cell reduction potentials of these redox couples in aqueous solutions at neutral pH, E degrees ', provide a window to understanding the thermodynamic and kinetic characteristics of this triad and their associated chemistry and biochemistry in vivo. Substituents on the quinone ring can significantly influence the electron density "on the ring" and thus modify E degrees' dramatically. E degrees' of the quinone governs the reaction of semiquinone with dioxygen to form superoxide. At near-neutral pH the pK(a)'s of the hydroquinone are outstanding indicators of the electron density in the aromatic ring of the members of these triads (electrophilicity) and thus are excellent tools to predict half-cell reduction potentials for both the one-electron and two-electron couples, which in turn allow estimates of rate constants for the reactions of these triads. For example, the higher the pK(a)'s of H(2)Q, the lower the reduction potentials and the higher the rate constants for the reaction of SQ(*-) with dioxygen to form superoxide. However, hydroquinone autoxidation is controlled by the concentration of di-ionized hydroquinone; thus, the lower the pK(a)'s the less stable H(2)Q to autoxidation. Catalysts, e.g., metals and quinone, can accelerate oxidation processes; by removing superoxide and increasing the rate of formation of quinone, superoxide dismutase can accelerate oxidation of hydroquinones and thereby increase the flux of hydrogen peroxide. The principal reactions of quinones are with nucleophiles via Michael addition, for example, with thiols and amines. The rate constants for these addition reactions are also related to E degrees'. Thus, pK(a)'s of a hydroquinone and E degrees ' are central to the chemistry of these triads.

Fig. 1

Shown are the basic structures of quinones, semiquinones, and corresponding hydroquinones. We show here the dominant species of each component of the triad that would be present in aqueous solution at pH 7. For example, for 1,4-hydroquinone the pK_a's are 9.8 and 11.4 for the first and second protons of the phenoxyl moieties, respectively [11,12]. The pK_a of the protonated 1,4-semiquinone radical (SQH^•) is 4.1 [13,14]; thus, it is shown as the radical anion.

Fig. 2

Chemical structures of quinones.

Fig. 3

The disproportionation (dismutation) reaction of semiquinone radicals is spontaneous and reversible at pH 7. Shown are the standard one-electron $\left(E_1^{°\prime }\right)$ and two-electron $\left(E_2^{°\prime }\right)$ reduction potentials for the duroquinone (D-Q), benzoquinone (B-Q), and 2,5-dichlorobenzoquinone (Cl₂-Q) triads. Under standard conditions, i.e., ≈ 1.0 M concentration for each species, the dismutation reaction will be spontaneous because the overall change in the Gibbs free energy is favorable. A positive change in ΔE will yield a negative value for the change in the Gibbs free energy (ΔG=−nFΔE), indicating a spontaneous process. However, comproportionation of Q and H₂Q to yield SQ^•− is also possible when the value of the mass action expression for the system is not equal to the equilibrium constant. For Eq. (6), the complete mass action expression will be Q_r=[Q][H₂Q]/[SQ^•−]²[H⁺]². However, E°′ is the reduction potential at pH 7; thus, at this fixed pH, the mass action expression can be simplified to Q_r=[Q][H₂Q]/[SQ^•−]², because [H⁺] is constant. The equilibrium constants shown were determined with this form of Q_r. The equilibrium constant for the disproportionation reaction (Eq. (6)) for a Q/SQ^•−/H₂Q triad can be determined using $\mathrm{\Delta }E=({(E_1^{°\prime })}_{\mathrm{Q}/{\text{SQ}}^{•-}}-{(E_1^{°\prime })}_{{\text{SQ}}^{•-}/\mathrm{H}2\mathrm{Q}})-(\mathit{\text{RT}}/n\mathrm{F})\text{ln}K$. At equilibrium ΔE = 0. If Q_r<K, the reaction will proceed to the right until equilibrium is achieved; if Q_r>K the reaction will proceed to the left; K is the equilibrium constant. (The quinones of each triad are shown with the same relative free energies. This was done only to conveniently show the relative changes in the free energy of each triad; it is not meant to imply they have the same Gibbs free energy.)

Fig. 4

Substituents influence E°′(Q/SQ^•−) and E°′(Q, 2H⁺/H₂Q) of quinones. (a) E°′(Q/SQ^•−) vs number of alkyl groups on the benzoquinone ring: (i) all three isomers of dimethyl-benzoquinone are included (2,3-dimethyl, 2,5-dimethyl, and 2,6-dimethyl); they are similar so the points overlap on this plot; (ii) 2-methyl-5-iso-propyl-benzoquinone; again it is similar and overlap; (iii) ethyl-1,4-benzoquinone; (iv) tert-butyl-1,4-benzoquinone. (b) E°′(Q/SQ^•−) vs number of methyl groups on 1,4-naphthoquinone; here we treat the second ring as benzoquinone with two alkyl groups. (c) E°′(Q/SQ^•−) vs number of chlorines on benzoquinone. (d) $E_2^{°\prime }(\mathrm{Q},2{\mathrm{H}}^{+}/{\mathrm{H}}_2\mathrm{Q})$ vs number of methyl groups on benzoquinone. Note these are two-electron half-cell reduction potentials.

Fig. 5

Second-order rate constants for formation of superoxide by SQ^•− (Eq. (7)), as well as the reverse reactions, are a function of E°′. (A) For the forward reaction, the rate constants fall into two categories: (i) when E°′<−150 mV the rate constants appear to be influenced by the diffusion limit, i.e., > 10⁸ M⁻¹ s⁻¹; and (ii) when E°′>−150 mV the rate constants appear to follow the Marcus theory for electron transfer, here approximated by a linear relationship. (B) For the reverse reaction of Eq. (7), the rate constants correlate well with the one-electron reduction potential of the quinones. The specific semiquinone radical for each point has as the parent quinone: a, 1,4-benzoquinone; b, methyl-1,4-benzoquinone; c, 2,3-dimethyl-1,4-benzoquinone; d, 2,5-dimethyl-1,4-benzoquinone; e, 2,6-dimethyl-1,4-benzoquinone; f, duroquinone; g, 2-methyl-1,4-naphthoquinone; h, 2,3-dimethyl-1,4-naphthoquinone; i, anthraquinone (estimated here for B); j, Mitomycin (estimated here for B); k, Adriamycin (estimated here for B); l, AZQ: 2,5-diaziridinyl-3,6-bis(carbethoxyamino)-1,4-benzoquinone.

Fig. 6

Experimental equilibrium constants match theoretical equilibrium constants closely. The dashed black line represents the theoretical values of K for Eq. (7) using the Nernst equation (Eq. (7)) with $E°\prime ({\mathrm{O}}_2/{\mathrm{O}}_2^{•-})=-180\text{mV}$; the slope of this line is 1/(59.1 mV) and the intercept is (−180 mV)/(59.1 mV). The variation in the experimental values represents the uncertainties of measurement. (Solid line, a fit of experimental data by linear regression.) However, if the very fast reactions of Eq. (7) are influenced by limitations from diffusion, then the largest forward rate constants are lower than theoretical rate constants, resulting in the corresponding Ks being underestimated. Likewise, when E°′(Q/SQ^•−) is very positive, the experimental, very fast rate constants for the reverse reactions of Eq. (7) are lower than theoretical rate constants and, thus K would be overestimated. These two extremes may account for the shallower slope and the underestimation of the intercept for experimental values.

Fig. 7

E°′(Q/SQ^•−) correlates with the pK_a's of the corresponding para-hydroquinones. The squares (, blue) are the first pK_a of a particular hydroquinone; the diamonds (◆, black) are the second pK_a of the hydroquinone; pK_a's are from [11,12,39]. The para-quinones/hydroquinones a–m show a linear relationship between the one-electron reduction potential of the quinone (E°′(Q/SQ^•−)) and the first and second pK_a's of the corresponding hydroquinone. The one-electron reduction potentials for quinones k and l have been estimated from Fig. 4; compounds A (catechol) and B (tetrachloro-1,4-hydroquinone) are not included in the linear regression of the two lines. We anticipate that ortho-hydroquinones would behave similarly to para-hydroquinones, but there would most likely be a systematic offset relative to the plots for para-hydroquinones. (A note on tetrachloro-1,4-hydroquinone: E°′(Q/SQ^•−) for tetrachloro-1,4-hydroquinone (B) has been estimated from the correlation of the reduction potentials determined in methyl cyanide [40,78]; knowing that pK_a1 for tetrachloro-1,4-hydroquinone is 5.6 [12] suggests that the one-electron reduction potential is actually about +800 mV in aqueous solution.) The compounds are: a. ubiquionol-1, coenzyme Q-1; b, tetramethyl-1,4-hydroquinone; c, 2,3,5-trimethyl-1,4-hydroquinone; d, plastoquinol-1; e, 2,6-dimethyl-1,4-hydroquinone; f, 2-methyl-5-isopropyl-1,4-hydroquinone; g, 2,3-dimethyl-1,4-hydroquinone; h, 2,5-dimethyl-1,4-hydroquinone; i, 2-methyl-1,4-hydroquinone; j, 1,4-hydroquinone; k, 2-chloro-1,4-hydroquinone; l, 2,6-dichloro-1,4-hydroquinone; m, 2,5-dichloro-1,4-hydroquinone; A, catechol; B, tetrachloro-1,4-hydroquinone.

Fig. 8

Second-order rate constants for formation of superoxide by SQ^•− (Eq. (7)) as well as the reverse reactions are a function of pK_a1. pK_a's are from Fig. 5. The para-hydroquinones a–f show a linear relationship between the rate constant of formation of semiquinone (Eq. (7)) as well as corresponding reverse Eq. (7) vs the first pK_a's of the corresponding hydroquinone. The compounds are: a, 1,4-benzoquinone; b, methyl-1,4-benzoquinone; c, 2,3-dimethyl-1,4-benzoquinone; d, 2,5-dimethyl-1,4-benzoquinone; e, 2,6-dimethyl-1,4-benzoquinone; f, duroquinone.

Fig. 9

There is a linear relationship of pK_a1 of H₂Q with pK_a of the corresponding SQH^•. (a) 1,4-Benzoquinone; (b) methyl-1,4-benzoquinone; (c) 2,3-dimethyl-1,4-benzoquinone; (d) 2,5-dimethyl-1,4-benzoquinone; (e) 2,6-dimethyl-1,4-benzoquinone; (f) 2,3,5-trimethyl-1,4-benzoquinone; (g) duroquinone.

Fig. 10

Time course of oxygen consumption during the autoxidation of PCB hydroquinones with different degrees of chlorination on the oxygenated phenyl ring. (a) 4′-Cl-2,5-H₂Q; (b) 4,4′-Cl-2,5-H₂Q; (c) 3,6,4′-Cl-2,5-H₂Q; and (d) 3,4,6-Cl-2,5-H₂Q. Solutions contained [H₂Q] = 2.0 mM in pH 7.4 phosphate buffer at 25 °C. Adapted from [56].

Fig. 11

The rate of O₂ consumption (µM s⁻¹) increases with the fraction of the di-ionized hydroquinone, which is determined by the number of chlorines on the ring. Rates were determined from the initial slopes for the uptake of oxygen in Fig. 10. (a) 4′-Cl-2,5-H₂Q; (b) 4,4′-Cl-2,5-H₂Q; (c) 3,6,4′-Cl-2,5-H₂Q; and (d) 3,4,6-Cl-2,5-H₂Q. A plot of log(rate) vs pK_a2 of the hydroquinone works well except for (d). However, pK_a1 for (d) is 7; thus there is not an approximate factor of 10 increase in the concentration of Q²⁻, compared to (c), because the solution pH is 7.4. However, when the log of fraction di-ionized (f_di-ionized) is plotted against log(rate), all points fit. The fraction of di-ionized (f_di-ionized) H₂Q was determined in pH 7.4 buffer, by using pK_a2 of the H₂Q's as 12, 11, 10, and 9 for H₂Q with 0, 1, 2, and 3 chlorines on the oxygenated ring, respectively [56]. The values are pK_a1 + 2, as indicated by Fig. 7. A slope of 1.0 indicates the reaction is first-order in [di-ionized hydroquinone].

Fig. 12

The rate constants for the Michael reaction of glutathione with various quinones are a function of E°′. The rate constants for reductive addition of GSH to benzoquinone and methylated benzoquinones decrease linearly with increasing degrees of methylation; methyl groups donate electrons into the quinone ring. (A) One-electron reduction potential for Q/SQ^•−; and (B) two-electron reduction potential of the quinones, Q,2H⁺/H₂Q. (a) 1,4-Benzoquinone; (b) methyl-1,4-benzoquinone; (c) 2,6-dimethyl-1,4-benzoquinone; (d) 2,5-dimethyl-1,4-benzoquinone; (e) 2,3,5-trimethyl-1,4-benzoquinone. Data from [70]. The kinetic data are at pH 6.0. Because it is actually the ionized thiol, here GS⁻, that dominates the reaction of Eq. (14), rate constants at pH 7 will be about 10 times larger.

Fig. 13

The rate constants for the Michael reaction of glutathione with various quinones are a function of pK_a1 of the corresponding hydroquinone. The rate constants for reductive addition of GSH to benzoquinone and methylated benzoquinones decrease linearly with increase of pK_a1. (a) 1,4-Benzoquinone; (b) methyl-1,4-benzoquinone; (c) 2,6-dimethyl-1,4-benzoquinone; (d) 2,5-dimethyl-1,4-benzoquinone; (e) 2,3,5-trimethyl-1,4-benzoquinone.