Electrostatic Contribution in the Catalysis of O 2 . - Dismutation by Superoxide Dismutase Mimics . MnIIITE - 2 - PyP 5 + vs MnIIIBr 8 T - 2 - PyP

The Mn(III) meso -tetrakis(N-ethylpyridinium-2-yl)porphyrin, MnIIITE-2PyP5+, is a potent SOD mimic in vitro, and was beneficial in rodent models of oxidative stress pathologies. Its high activity has been ascribed to both the favorable redox potential of its metal center and to the electrostatic facilitation assured by the four positive charges encircling the metal center. Its comparison with the non-alkylated, singly-charged analogue Mn(III) beta-octabromo meso-tetrakis(2-pyridyl)porphyrin, MnIIIBr8T-2-PyP+, enabled us to evaluate the electrostatic contribution to the catalysis of O2dismutation. Both compounds exhibit nearly identical metal-centered redox potential for MnIII/MnII redox couple: +228 mV for MnIIITE-2-PyP5+ and +219 mV vs NHE for MnIIIBr8T-2-PyP+. The eight electron-withdrawing beta pyrrolic bromines contribute equally to the redox properties of the parent MnIIIT-2-PyP+ as do four quaternized cationic meso ortho pyridyl nitrogens. However, the SOD-like activity of the highly charged MnIIITE-2-PyP5+ is 200-fold higher (log kcat = 7.76) than that of the singly-charged MnIIIBr8T-2-PyP+ (log kcat = 5.63). The kinetic salt effect showed that the catalytic rate constants of the MnIIITE-2-PyP5+ and of its methyl by Spasojevic et al. Electrostatics in Superoxide Dismutation 2 by gest on O cber 1, 2017 hp://w w w .jb.org/ D ow nladed from analogue, MnIIITM-2-PyP5+, are exactly 5-fold more sensitive to ionic strength than is the kcat of MnIIIBr8T-2-PyP+, which parallels the charge ratio of these compounds. Interestingly, only a small effect of ionic strength on the rate constant was found in the case of penta-charged para (MnIIITM-4-PyP5+) and meta isomers (MnIIITM-3PyP5+), indicating that the placement of the positive charges in the close proximity of the metal center (ortho position) is essential for the electrostatic facilitation of O2.dismutation. Introduction The thermodynamic (1,2-8) and electrostatic effects (9,10-15) in the catalysis of O2.dismutation by superoxide dismutases have been extensively studied. The redox potential of all superoxide dismutases was found to be similar, independently of the type of the metal in the active site. It is around midway (+360 mV vs NHE) (16) between the potential for the oxidation (-160 mV vs NHE) and for the reduction of O2.(+890 mV vs NHE). Thus it allows equal driving force (hence kox = kred = ~2 x 109 M-1 s-1) for both half-reactions of the catalytic cycles (eqs [1] and [2]) (17-19). For E. coli FeSOD MnIIISOD + O2.<====> MnIISOD + O2 kox [1] MnIISOD + O2.+ 2 H+ <====> MnIIISOD + H2O2 kred [2] by Spasojevic et al. Electrostatics in Superoxide Dismutation 3 by gest on O cber 1, 2017 hp://w w w .jb.org/ D ow nladed from the E1/2 is +223 mV vs NHE at pH 7.4, and for MnSOD from the B. stearothermophilus and E. coli the E1/2 were +260 and +310 mV vs NHE at pH 7 (1,20,21). However, when manganese was replaced by iron in the active site of MnSOD the enzymatic activity was lost (17), which has been attributed to the decrease of the redox potential below that required for the oxidation of superoxide ion (3,17). Such metal ion specificity (2) has been recently explained by the higher affinity of Fe3+ than Mn3+ for hydroxide (6,8). The crystal structures of different SODs reveal a highly conserved electrostatic “funnel” (12,13,15) that is believed to guide the negatively charged superoxide towards the active site of the enzyme. In the past there have been considerable efforts to evaluate the extent of electrostatic facilitation, but a major difficulty lies in the inability to specifically modify the positively charged residues without affecting the structural integrity of the active site (9). The Mn(III) porphyrin, MnIIITE-2-PyP5+ (AEOL-10113) has been shown (2226) to possess high SOD-like activity in vitro with log kcat = 7.76. The compound has further proven effective in protection of SOD-deficient E. coli (25) and in stroke (27,28), spinal cord injury (29), diabetes (30,31), sickle cell disease (32), and radiation/cancer (33,34) rodent models of oxidative stress injuries. Much like SOD (1,9,20,21) its high catalytic potency has been ascribed both to the favorable redox properties of the metal center and to the effect of the positively charged ortho N-ethylpyridyl nitrogens that provide electrostatic facilitation for the approach of the negatively charged superoxide (22). by Spasojevic et al. Electrostatics in Superoxide Dismutation 4 by gest on O cber 1, 2017 hp://w w w .jb.org/ D ow nladed from The MnIIITE-2-PyP5+ (Scheme I) exists as a mixture of rotational isomers (35). Expectations that the four positive charges in αααα isomer will guide the superoxide anion towards the metal center in a cooperative fashion making it the most powerful SOD mimics among the isomers, proved to be groundless; all four isomers were found to be of equal catalytic potency. Recently, the synthesis and characterization of the β-brominated non-Nalkylated analogue of MnIIIT-2-PyP+ has been reported (36). The electronwithdrawing effect of the β-pyrrolic bromines on the redox properties of the metal center of the porphyrins (50-70 mV/bromine) has been previously established (37-43). The effect of eight β-pyrrolic bromines was expected to be similar in magnitude to the effect of the four quaternized pyridyl nitrogens on the redox properties of the starting unsubstituted MnIIIT-2-PyP+ porphyrin molecule. Herein, we show that the redox properties of the brominated non-N-alkylated and the N-alkylated Mn(III) ortho pyridylporphyrins are indeed nearly identical, allowing us to evaluate the electrostatic contribution in the catalysis of O2.dismutation. Similar studies would be hard to conduct on the enzymes themselves. Thus our findings give us the unique opportunity to understand the relative importance of thermodynamics and kinetics in the O2.dismutation by the superoxide dismutases (17-19). Experimental Materials by Spasojevic et al. Electrostatics in Superoxide Dismutation 5 by gest on O cber 1, 2017 hp://w w w .jb.org/ D ow nladed from General. Xanthine and ferricytochrome c were from Sigma, and NaCl, KOH, KH2PO4, methanol and EDTA from Mallinckrodt. Xanthine oxidase was prepared by R. Wiley and was supplied by K. V. Rajagopalan (44). Catalase was from Boehringer, ultrapure argon from National Welders Supply Co., and tris buffer (ultrapure) was from ICN Biomedicals, Inc. Mn(III) porphyrins. The H2T-2-PyP+ and MnIIITM-3(4)-PyP5+ were obtained from MidCentury Chemicals (Chicago, IL). MnIIITE(M)-2-PyP5+ (25,26) and MnIIIBr8T-2-PyP+ (36) were prepared as previously described. The molar absorptivities of the Soret bands of MnTM-2-PyP5+ (log ε453.4=5.11), MnTM-3-PyP5+ (log ε459.8=5.14), MnTM-4-PyP5+ (log ε462.2=5.11), MnTE-2-PyP5+ (log ε454 = 5.14) (25,26) all in water and of MnIIIBr8T-2-PyP+ (log ε 482 = 4.66) in acetonitrile were used for quantitation. Due to the low water-solubility, a 2mM stock solution of MnIIIBr8T-2PyP+ in methanol was used throughout this study. Methods Electrochemistry. Measurements were performed on a CH Instruments Model 600 Voltammetric Analyzer. A three-electrode system was utilized with a glassy carbon (3 mm) or gold (2mm) button working electrode (Bioanalytical Systems), a Ag/AgCl reference and a Pt wire as auxiliary electrode. Due to the low water-solubility of the by Spasojevic et al. Electrostatics in Superoxide Dismutation 6 by gest on O cber 1, 2017 hp://w w w .jb.org/ D ow nladed from MnIIIBr8T-2-PyP+, electrochemical studies of both compounds were performed in 9/1 (v/v) methanol/aqueous solutions as previously reported (45a). The 9/1 (v/v) methanol/aqueous solutions contained 0.05 M tris, pH 7.8, 0.1 M NaCl, and 0.3 mM metalloporphyrin. Tris buffer was used instead of phosphate buffer because the latter precipitates in methanol. The potentials were standardized against potassium ferrocyanide/ferricyanide (46) and MnIIITE-2-PyP5+. The redox potential of the MnIII/MnIV redox couple which was previously found to be proton-dependent (47) was determined at pH 12.3. The scan rates were 0.0110 V/s. E1/2 for MnII/MnIII and MnIII/MnIV redox couples obtained in 9/1 (v/v) methanol/aqueous solutions were extrapolated to aqueous medium values as previously described (45a). Catalysis of O2.dismutation. We have previously shown that the convenient cytochrome c assay gives the same catalytic rate constants as does pulse radiolysis in the case of MnIIITE-2-PyP5+, {MnIIIBVDME}2, {MnIIIBV2-}2 and MnIICl2 (45a) and it was therefore utilized in this study. The xanthine/xanthine oxidase reaction was the source of O2.and ferricytochrome c was used as the indicating scavenger for O2.(48). The reduction of cytochrome c was followed at 550 nm. Assays were conducted at (25+1) oC, in 0.05 M phosphate buffer, pH 7.8, 0.1 mM by Spasojevic et al. Electrostatics in Superoxide Dismutation 7 by gest on O cber 1, 2017 hp://w w w .jb.org/ D ow nladed from EDTA, 10 μM cytochrome c, 40 μM xanthine, + 15 μg/mL of catalase. Aqueous stock solutions of MnIIITE(M)-2-PyP5+ and MnTM-3(4)-PyP5+ and the methanolic stock solution of MnIIIBr8T-2-PyP+ were diluted into the assay mixture. Rate constants for the reaction of metalloporphyrins with O2.were based upon competition with 10 μM cytochrome c as described elsewhere (45a). The kcyt c = 2.6 x 105 M-1 s-1 obtained under the same experimental conditions (pH 7.8, 21 oC, 0.05 M phosphate buffer, 0.1 mM EDTA) (45b) was used to calculate kcat. The O2.was produced at the rate of 1.2 μΜ per minute. Any possible interference through inhibition of the xanthine/ xanthine oxidase reaction by the test compounds was examined by following the rate of urate accumulation at 295 nm in the absence of cytochrome c. No reoxidation of ferrocytochrome c by metalloporphyrins was observed. No effect of catalase was detected implying sufficient stability of the compounds towards H2O2. We have previously determined the rate constants for the degradation of MnIIITE-2-PyP5+ and of MnIIITM-2-PyP5+ by H2O2 to be 1.3 M-1 s-1 and for MnIIITM-3-PyP5+ and MnIIITM-4-PyP5+ 4.9 and 4.6 M-1 s-1, respectively. In this work we found that the MnIIIBr8T-2-PyP+ proved to be at least two orders of magnitude more stable. Kinetic salt effect. The dependence of the catalytic rate constant for the O2.dismutation upon ionic strength was determined in 0.05 M phosphate buffer, pH 7.8 with by Spasojevic et al. Electrostatics in Superoxide Dismutation 8 by gest on O cber 1, 2017 hp://w w w .jb.org/ D ow nladed from NaCl ranging from 0 to 0.4 M. Results Electrochemistry. The MnII/MnIII redox couple. Reversible cyclic voltammograms of the MnII/MnIII redox were obtained for both compounds, MnIIITE2-PyP5+ and MnIIIBr8T-2-PyP+, at scan rates of 0.1 V/s (Figure 1). Thus it was possible to determine the half-wave potentials, E1/2, given in Table 1. The two compounds have almost identical MnII/MnIII metal centered E1/2 values at pH 7.8, as predicted from the number and the nature of the electron-withdrawing substituents on the meso positions of the porphyrin ring (37-43). Moreover, their voltammetric behavior in terms of electrochemical reversibility (peak-to peak potential separation, ∆Epp, Figure 2A, and current response to a change in scan-rate, Figure 2C) as well as the chemical reversibility (ratio between the reduction and oxidation peak currents, Figure 2B) is strikingly similar which leaves us to believe that the difference in the reactivity towards superoxide is indeed due to the difference in the overall positive charge (electrostatic attraction of superoxide anion) and not due to a difference in the rates of electron transfer (electronic and structural differences). The MnIII/MnIV redox couple. Redox properties of the Mn sites were further explored by studying MnIII/MnIV redox couple. Because the hydroxo-MnIII and oxoMnIV species are involved (47), this redox process is accessible only in basic solution by Spasojevic et al. Electrostatics in Superoxide Dismutation 9 by gest on O cber 1, 2017 hp://w w w .jb.org/ D ow nladed from and is proton-dependent. At pH 12.3, reversible cyclic voltammograms were obtained in the case of both compounds, MnIIITE-2-PyP5+ and MnIIIBr8T-2-PyP+, with essentially equal MnIII/MnIV redox couple potentials of +381 mV and +372 mV vs NHE, respectively (Figure 1B, Table 1). The MnIII/MnIV and MnII/MnIII redox processes are independent as demonstrated on Figure 1B (dashed traces) whereas reversible voltammetric waves were obtained even when the cycling was done only in a narrow potential range around the E1/2 of the corresponding redox couple. Because at pH 12.3 a deprotonation of the axially ligated water on Mn(III) porphyrins occurs, the MnII/MnIII redox potential shifts negatively (47). Therefore, we ascribe the 145 mV difference in the shift between the two redox couples (Figure 1B, Table 1) to a difference in the pKa,axs of their axially ligated water. We have previously found that E1/2 reflects the electron density of the porphyrin ring and the metal center in such a way that there is a linear relationship between the pKa of the pyrrolic nitrogen protons of the porphyrin ligand and the metal-centered E1/2 for a series of differently substituted Mn porphyrins (25). We have further found that the axial ligation also is influenced by the electron density of the metal center; in the case of ortho, meta and para MnTM-2-PyP5+ more positive E1/2 correlates with lower pKa,ax (47). However, in the case of a series of Mn(III) ortho N-alkylpyridylporphyrins (alkyl = methyl through octyl) we saw that hydrophobic effects may reverse the trend (49). With more hydrophobic members of the series, despite a more positive E1//2, the creation of charge is hindered by Spasojevic et al. Electrostatics in Superoxide Dismutation 10 by gest on O cber 1, 2017 hp://w w w .jb.org/ D ow nladed from resulting in higher pKas of the pyrrolic nitrogens of the parent ligands (49). In the present work at pH 12.3, the hydrophobic MnIIIBr8T-2-PyP+ with presumably higher pKa,ax (thus resisting deprotonation), exhibits a lower shift of the MnII/MnIII couple than highly hydrophilic MnIIITE-2-PyP5+. In the case of MnIII/MnIV couple which is also proton-dependent (47), there was practically no difference in E1/2 between the two compounds which is in line with findings reported by us (23,45-47) and others (50-56) that MnIII/MnIV redox potential is fairly insensitive to the porphyrin structure. Catalysis of O2dismutation. The catalytic rate constants were calculated from the linear plots of v0/vi-1 vs concentration obtained from the spectrophotometric cytochrome c assay measurements, as described elsewhere (45a). The kcat values determined in 0.05 M phosphate buffer, pH 7.8 are given in Table 1. The MnIIIBr8T-2PyP+ is ~200-fold less efficient a catalyst than MnIIITE-2-PyP5+. Kinetic salt effect. The effect of the ionic strength (μ) on the catalytic rate constant was assessed using eq [3] which is based on Debye-Huckel relation (57) for the effect of the ionic strength of the solution on the activity coefficient of an ion. log k = log kref + 2 A zAzB (μ1/2/1 + μ1/2) [3] by Spasojevic et al. Electrostatics in Superoxide Dismutation 11 by gest on O cber 1, 2017 hp://w w w .jb.org/ D ow nladed from The k is the rate constant at any given ionic strength, while kref is the rate constant at μ = 0. The A is a collection of physical constants with a value of 0.509 and zA and zB are the charges of the reacting species. The equation predicts a linear plot of log k vs (μ1/2/1 + μ1/2). Eq [3] assumes a coefficient of 1.0 (βai) for μ1/2 in the denominator, i.e., the distance of the closest approach, ai to be 3 Α and β is a physical constant, 0.33 x 1010m-1. It is doubtful whether great significance can be attributed to the ai, thus to the product zAzB (35), especially so in the light of the bulkiness, high charge and solvation shell of the metalloporphyrins. Accounting for the monoand diprotonated phosphates as the major species at pH 7.8 (pKa = 7.2), and the concentration of the NaCl, the ionic strength was calculated using equation [4], where mi is the molality and zi the charge of the given ion.