A comprehensive evaluation of catalase-like activity of different classes of redox-active therapeutics

Abstract

The growing insight into the biological role of hydrogen peroxide (H₂O₂) under physiological and pathological condition and the role it presumably plays in the action of natural and synthetic redox-active drug imparts a need to accurately define the type and magnitude of reactions which may occur with this intriguing and key species of redoxome. Historically, and frequently incorrectly, the impact of catalase-like activity has been assigned to play a major role in the action of many redox-active drugs most so SOD mimics and peroxynitrite scavengers, and in particular MnTBAP³⁻ and Mn salen derivatives. The advantage of one redox-active compound over another has often been assigned to the differences in catalase-like activity. Our studies provide substantial evidence that Mn(III) N-alkylpyridylporphyrins couple with H₂O₂ in actions other than catalase-related. Herein we have assessed the catalase-like activities of different classes of compounds: Mn porphyrins (MnPs), Fe porphyrins (FePs), Mn(III) salen (EUK-8) and Mn(II) cyclic polyamines (SOD-active M40403 and SOD-inactive M40404). Nitroxide (Tempol), nitrone (NXY-059), ebselen and MnCl₂, which have not been reported as catalase-mimics, were used as negative controls, while catalase enzyme was a positive control. The dismutation of H₂O₂ to O₂ and H₂O was followed via measuring oxygen evolved with Clark oxygen electrode at 25°C. The catalase enzyme was found to have k_cat(H₂O₂) = 1.5 × 10⁶ M⁻¹ s⁻¹. The yield of dismutation, i.e. the amount of O₂ evolved, was assessed also. The magnitude of the yield reflects an interplay between the k_cat(H₂O₂) and the stability of compounds towards H₂O₂-driven oxidative degradation, and is thus an accurate measure of the efficacy of an catalyst. The k_cat(H₂O₂) values for 12 cationic Mn(III) N-substituted (alkyl and alkoxyalkyl) pyridylporphyrin-based SOD mimics and Mn(III) N,N’-dialkylimidazolium porphyrin, MnTDE-2-ImP⁵⁺ ranged from 23 to 88 M⁻¹ s⁻¹. The analogous Fe(III) N-alkylpyridylporphyrins showed ~10-fold higher activity than the corresponding MnPs, but the values of k_cat(H₂O₂) are still ~4 orders of magnitude lower than that of the enzyme. While the k_cat(H₂O₂) values for Fe ethyl and n-octyl analogs were 830 and 360 M¹ s⁻¹, respectively, the FePs are more prone to H₂O₂-driven oxidative degradation therefore allowing for similar yields in H₂O₂ dismutation as analogous MnPs. The k_cat(H₂O₂) values are dependent upon the electron deficiency of the metal site as it controls the peroxide binding in the 1^st step of dismutation process. SOD-like activities depend upon electron-deficiency of the metal site also, as it controls the 1^st step of O₂^.− dismutation. In turn, the k_cat(O₂^.−) parallels the k_cat(H₂O₂). Therefore, the electron-rich anionic non-SOD mimic MnTBAP³⁻ has essentially very low catalase-like activity, k_cat(H₂O₂) = 5.8 M⁻¹ s⁻¹. The catalase-like activity of Mn(III) and Fe(III) porphyrins are at most, 0.0004% and 0.05% of the enzyme activity, respectively. The k_cat(H₂O₂) of 8.2 and 6.5 M⁻¹ s⁻¹ were determined for electron-rich Mn(II) cyclic polyamine-based compounds, M40403 and M40404, respectively. The EUK-8, with modest SOD-like activity, has only slightly higher k_cat(H₂O₂) = 13 M⁻¹s⁻¹. The biological relevance of k_cat(H₂O₂) of MnTE-2-PyP⁵⁺, MnTDE-2-ImP⁵⁺, MnTBAP³⁻, FeTE-2-PyP⁵⁺, M40403, M40404 and Mn Salen was evaluated in wild type and peroxidase/catalase-deficient E. coli.

Graphical abstract

Introduction

H₂O₂ plays a critical role in metabolic processes under physiological and pathological conditions due to its longevity, neutrality and ability to cross membranes [1]. As a messenger it is in the forefront of cellular transcription [1, 2]. Its signaling effects have been extensively addressed in Methods in Enzymology 2013, vol 527. H₂O₂ also plays a therapeutic role; along with its progeny H₂O₂ is involved in cancer killing via chemo- and radiotherapy [3, 4]. It deserves mentioning that even H₂O₂, in its own right, was used as a treatment in stroke therapy supposedly inducing adaptive response [5]. Nature has developed multiple redundant systems to maintain H₂O₂ at nM intracellular levels which are sufficient enough for its role in cellular signaling. Such are families of glutathione peroxidases (GPx), glutathione reductases, catalases, peroxiredoxins, thioredoxin reductases, glutathione S-transferases, etc. Our first line of defense against oxygen toxicity and an ultimate requirement of all aerobic life is a family of superoxide dismutases, SODs. During the catalysis of O₂^.− dismutation by enzyme or its mimic (eq (1)) O₂^.− is oxidized one-electronically into O₂ and reduced into H₂O₂. In order for SOD enzymes to be considered antioxidative systems, the H₂O₂ production needs to be coupled to its elimination. H₂O₂ is either reduced 2-electronically into H₂O via peroxidases, or dismuted, i.e. oxidized 2-electronically to O₂ and reduced to H₂O via catalases (eq (2)).

$$\underset{\_}{\text{SOD activity}}-\text{catalysis of}{{\mathrm{O}}_2}^{.-}\text{dismutation}:2{{\mathrm{O}}_2}^{.-}+2{\mathrm{H}}^{+}<=====>{\mathrm{H}}_2{\mathrm{O}}_2+{\mathrm{O}}_2$$ (1)

$$\underset{\_}{\text{Catalase activity}}-\text{catalysis of}{\mathrm{H}}_2{\mathrm{O}}_2\text{dismutation}:2{\mathrm{H}}_2{\mathrm{O}}_2<=====>{\mathrm{O}}_2+2{\mathrm{H}}_2\mathrm{O}$$ (2)

A fine coupling of SODs and H₂O₂-removing systems exists under physiological conditions but gets perturbed under pathological conditions and frequently in cancer resulting in increased H₂O₂ levels. The state of oxidative stress and its implications on tumor biology and treatment have been addressed in numerous reports [6–14].

Redox-active drugs, either of synthetic nature such as metalloporphyrins, Mn(III) salens, Mn(III) cyclic polyamines, nitroxides, nitrones, MitoQ, or of natural sources (e.g. flavonoids, catechols) reportedly interfere either directly or indirectly with components of the cellular redox environment, redoxome [15]. Over the last 20 years, our knowledge on redox-active drugs, in particular SOD mimics, has increased and has been summarized in several reviews [8, 16–22]. The small molecular structure of SOD mimics, unlike that of the enzymes, allows them to interact rapidly with many other targets. Mn porphyrin-based SOD mimics are powerful antioxidants, reducing small molecules such as O₂, O₂^.−, ONOO⁻, CO₃^.− and ClO⁻. Yet they also act as pro-oxidants, oxidizing biological targets such as O₂^.−, thiols (both simple thiols such as glutathione and cysteine and protein thiols), tetrahydrobioterin and ascorbate [21]. Further, Mn porphyrins are able to employ H₂O₂ to catalyze S-glutathionylation of thiols of signaling proteins affecting in turn cellular transcription. The S-glutathionylation of p50 and p65 subunits of NF-κB [20, 21, 23], as well as complexes I, III and IV of mitochondrial respiration has been reported [21, 24].

Numerous reports have been published on catalase-like (eq (2)) and/or joint catalase- and SOD-like activities (eq (1)) of metalloporphyrins (in particular MnTBAP³⁻ and MnTM-4-PyP⁵⁺) and Mn salen compounds; such claims have been used to justify the involvement of O₂^.− or H₂O₂ in proposed pathways, while frequently compounds lacked either one or both activities [13, 25–41]. We have therefore decided to quantify the ability of different classes of redox-active compounds (some of whom are potent SOD mimics) to catalyze the H₂O₂ dismutation (eq (2)). Our goal has been to provide clear evidence that compounds currently used (or misused) as SOD mimics do not possess ability to catalyze H₂O₂ dismutation to any practical biologically-relevant extent. Yet, H₂O₂ seems to have major role in biology of metalloporphyrins and their therapeutic effects via pathways discussed elsewhere [21, 42–44].

Experimental

Materials and Methods

Redox-active compounds

SOD mimics and non-SOD mimics were analyzed. Mn-, Fe porphyrins, Mn Salen and EUK-8 were synthesized as reported [45–50]. Mn(II) cyclic polyamines, M40403 and M40404 were synthesized based on the literature procedures [51]. Nitrone NXY-059 was purchased from Selleckchem, Cat No. S6002, >99 % and nitroxide, Tempol, was obtained from Enzo, Cat No. ALX-430-081-G001, Lot No. L27338C. Catalase enzyme from bovine liver was bought from Sigma, Cat No. C1345-1G, Lot No. 010M7011V. Ebselen was purchased Cayman Chemical Company, Cat No. 70530, Lot No. 0406449-57. Figure 1 shows the structures of Mn porphyrins and Figure 2 the structures of other redox-active drugs studied here.

Figure 1. Structures of the Mn(III) porphyrins whose catalase-like activity was assessed herein.

Different Mn porphyrins are studied, and except MnTBAP³⁻ and MnTE-2-PyPhP⁵⁺, all are SOD mimics of similarly high k_cat(O₂^.−). Their SOD-like activity is controlled by the close vicinity (ortho position indicated with number 2) of either pyridyl (Py) or imidazolyl (Im) nitrogens to the metal site which controls kinetics and thermodynamics of the reactions of those Mn complexes with reactive species [13]. The data on those compounds are listed in Tables 1 and 3.

Figure 2. Structures of other redox-active drugs whose catalase-like activity was assessed herein.

Compounds include SOD mimics of different magnitudes of SOD-like activities, such as Fe porphyrins, as well as compounds such as nitrones and nitroxides which are not SOD mimics but can cycle with other reactive species whereby eventually removing O₂^.− also [19]. Nitrone can trap free radicals such as O₂^.− and form nitroxide and thus affect O₂^.− levels. Nitroxide in turn can be oxidized with CO₃^.− to oxoammonium cation which then rapidly oxidizes O₂^.− closing the catalytic cycle. Mn(II) cyclic polyamine, M40403 is a very potent SOD mimic, but its SOD-inactive analog, M40404, is not. If Mn complexes fall apart they would release Mn. Moreover Mn(II) low molecular weight complexes, and in particular Mn(II) lactate, are SOD mimics also [19, 21]. The kinetic and thermodynamic data on these compounds are listed in Tables 2 and 3. The data for Fe corrole are taken from literature [19, 21, 52, 53].

Assessing the ability of redox-active drugs to catalyze H₂O₂ dismutation

The experiments were carried out at (25±1)°C in 0.05 M tris buffer, pH 7.8, 0.1 mM EDTA. Prior to measurements, the solutions were purged with air (~21% oxygen). The O₂-sensitive Clark electrode (0.1 M KCl as filling solution) connected to a potentiostat was used. A potential of E = −0.8 V vs Ag/AgCl was applied to the electrode and once the initial current was stabilized (I_max, corresponding to the [O₂] = 0.255 mM [54] in air-saturated solution), the solution was purged with N₂ until the current was stabilized again (I_min, corresponding to the [O₂] ≈ 0 mM). This allows the calculation of O₂ concentration in the solution (in µM) as [O₂]_obs = ((I_obs − I_min) × 255) / (I_max − I_min), where I_obs is the current observed at any given time after purging. This calibration procedure is followed by the addition of H₂O₂ up to 1 mM concentration, whereas the redox-active drugs or catalase were then added at 2–20 µM or 0.1–1 pM, respectively, to start the dismutation reaction (eq (2)). The increase in current, I_obs, corresponding to the increase of O₂ concentration, Δ[O₂], as a function of time was followed for at least 300 s. The initial reaction rate, ν₀ = k_cat [catalyst]₀ [H₂O₂]₀, was calculated from the slope of the initial linear kinetics as ν₀ = ([O₂]_obs(t₂) − [O₂]_obs(t₁)) / (t₂-t₁). Under pseudo first-order conditions, k_cat [H₂O₂]₀ = k_obs, and the ν₀ = k_obs [catalyst]₀. When the values of ν₀ were plotted against the catalyst concentrations, the slope, k_obs, is obtained. Finally, since Δ[H₂O₂] = 2 Δ[O₂], then the second order rate constant in M⁻¹ s⁻¹, could be calculated as k_cat = 2 k_obs / [H₂O₂]₀. The data obtained for Mn porphyrin, MnTnBuOE-2-PyP⁵⁺ and for catalase itself are presented in Figure 3. Under identical conditions, when 0.05 M tris buffer was replaced with 0.05 M potassium phosphate buffer, no significant difference in kinetic parameters was observed for interaction of ortho, meta, and para isomers of MnTEPyP with H₂O₂.

Figure 3. Assessment of the catalase-like activity of redox-active therapeutics.

Experimental set-up for the determination of catalase-like activity (A); The determination of the k_obs from the dependence of initial rates, ν₀ on the concentration of either catalase enzyme (B) or MnTnBuOE-2-PyP⁵⁺ (C). Experiments were carried out at (25±1)°C in 0.05 M tris buffer, pH 7.8 and 0.1 mM EDTA.

The other parameters that describe the catalysis of H₂O₂ dismutation by redox-active compounds are the turnover number (TON), which describes the maximal yield of O₂ evolution (thus yield of H₂O₂ dismutation, in %), and turnover frequency (TOF). Briefly, the reaction run under same experimental conditions as described above with 20 µM catalyst and 1 mM H₂O₂, was followed until no further evolution of O₂ was registered. This maximal amount of O₂ evolved (in µM), [O₂]_max, was calculated as: [O₂]_max = ((I_obs − I_min) × 255) / (I_max − I_min), where I_obs is the current observed at maximal point of O₂ evolution, and I_max and I_min currents corresponding to air-saturated and N₂-purged solutions, respectively. The maximal yield of O₂ production (in %) was further calculated as % yield (O₂) = ([O₂]_max / 500) × 100%, where 500 corresponds to [H₂O₂]₀ / 2 = 500 µM, i.e. the maximal possible O₂ concentration. TON was calculated as maximal number of O₂ moles produced per mole of a catalyst. Initial rates for the reactions, calculated from the slopes of initial linear segments of kinetic curves, were expressed in µM s⁻¹. The TOF values (s⁻¹), which present the ratios of initial rates and concentrations of catalysts, are given in Table 3.

Table 3. The reduction potential E_1/2 vs NHE, k_cat(H₂O₂) for the catalysis of H₂O₂ dismutation and log k_cat(O₂^.−) for the catalysis of O₂^.− dismutation of various redox-active compounds and catalase.

If not indicated, the data are taken from [19, 21, 22, 49, 50]. The catalysis of H₂O₂ dismutation was followed in 0.05 M tris buffer at (25±1) °C with Clark oxygen electrode. In case of Fe(III) porphyrins, Mn(III) salen and Mn(II) cyclic polyamines the E_1/2 relates to the M^III/M^II redox couple, while it relates to M^IV/M^III redox couple in case of Fe(III) corrole. The catalysis of O₂^.− dismutation was followed in 0.05 M potassium phosphate buffer, pH 7.8 at (25±1) °C as described in references [49, 50, 83]. The k_cat(H₂O₂) value for Fe corrole was determined at 37 °C [53], while k_cat(O₂^.−) was measured at (25±1) °C [52].

Compound	E1/2, mV vs NHE	kcat (H2O2), M−1 s−1	log kcat (O2.−)
FeTE-2-PyP5+, a	+211	803.5	8.05
FeTnOct-2-PyP5+, a	+261	368.4	7.09
FeTrF5Ph-β(SO3)2-corrole2−	+1050b	6400c	6.48d
MnCl2	+850e	none	6.11–6.95
Mn Salen, EUK-8	+130	13.5	6.36f
M40403	+525(ACN)	8.2	7.08
M40404	+452(ACN)	6.5	none
Nitroxide, Tempol	+810g	none	< 3 at pH 7.8
Nitrone, NXY-059		none
Ebselen		none
Catalase enzyme		1.5 × 106

^aAt physiological pH the FeTE(or nOct)-2-PyP⁵⁺ has one hydroxo ligand bound axially and thus carries 4+ total charge;

^brefs [19, 21];

^cthe k_cat(H₂O₂) was determined at 37°C [53];

^dref [52];

^eoxidation potential only;

^fcorrected value;

^gone-electron reduction potential relates to RNO⁺/RNO redox couple.

Degradation of MnPs with H₂O₂

The degradation of MnTE-2-PyP⁵⁺, MnTnBuOE-2-PyP⁵⁺, MnTnHex-2-PyP⁵⁺, MnTnOct-2-PyP⁵⁺, MnTnHexOE-2-PyP⁵⁺, MnTDE-2-ImP⁵⁺ and MnTBAP³⁻ was followed in 0.05 M tris buffer, pH 7.8, 0.1 mM EDTA at (25±1)°C. 10 µM MnPs were mixed with 0.5 mM H₂O₂ and the absorbance at Soret band was followed for 60 s.

Evaluation of the redox-active drugs in an E. coli model of H₂O₂-induced damage

Ability of selected compounds to protect cells against damage imposed by H₂O₂ was tested on the following E. coli strains: GC4468 (F⁻Δlac U169 rpsL), provided by Dr. D. Touati (Institute Jacques Monod, CNRS, Paris, France); AB1157 [F⁻ thr-1 leuB6 proA2 his-4 thi-1 argE2 lacY1 galK2 rpsL surE44 ara-14 xyl-15 mtl-1 tsx-33]; MG1655, F− lambda- ilvG-rfb-50 rph-1 (parental to LC106), LC106 [same as MG1655 plus Δ(ahpC-ahpF′) kan::′ahpF Δ(katG17::Tn10)1 Δ(katE12::Tn10)1] [55] provided by Dr. J. Imlay (University of Illinois at Urbana-Champaign, Urbana, IL). Representative compounds were analyzed: MnTE-2-PyP⁵⁺, MnTDE-2-ImP⁵⁺, MnTBAP³⁻, FeTE-2-PyP⁵⁺, M40403, M40404, Mn Salen and Tempol. Overnight cultures were grown in Luria-Bertoni (LB) medium. They were then diluted to OD₆₀₀ ~ 0.01 with the same medium and grown to early log phase. Cultures were then diluted with LB medium to OD₇₀₀ 0.08 and were either not incubated or incubated for 60 min with or without 20 µM of the tested compounds; OD at 700 was used to avoid interference with the absorbance of the compounds. Since the effect of H₂O₂ depends on the number of cells, cultures were diluted after incubation to an OD₇₀₀ value of exactly 0.08. Then 100 µl aliquots were placed in a 96-well plate. After incubation with compounds cells were washed with PBS. To half of the wells H₂O₂ was added at a final concentration of 5.0 mM for the parental and 0.5 mM for the mutant cells. Plates were incubated 15 minutes at 37°C on a shaking (200 rpm) platform. At the end of the incubation 1,000 units/ml of catalase were added to each well to stop the action of H₂O₂. Controls incubated without any additions, with compounds only, with H₂O₂ only, and with catalase added before H₂O₂, were run in parallel. Varying the length of pre-incubation of cultures with MnPs (up to 3 hours) or using E. coli with different genetic backgrounds (GC4468 and AB1157) did not change the outcome of study. The same results were obtained when viability was assessed by plating and counting colonies.

Toxicity of H₂O₂ (and hence possible protection by compounds of interest) was assessed by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) test and by plating and counting colonies. The MTT test was carried out as described previously [56]. Formazan crystals were solubilized with 10% SDS in 10 mM HCl. At the end of the incubation 10 µl of MTT reagent (25 mg MTT in 5.0 ml PBS) were added to all wells. The plates were incubated in dark for 30 min on a shaker at 37°C. Afterwards, the 100 µl aliquots of SDS solution (10% SDS in 10 mM HCl) were added to each well and plates were incubated for 1 h at room temperature. The absorbance of each well was measured at 570 nm and 700 nm (background) using a microplate reader. For plating and counting colonies, after treatment samples were diluted in sterile PBS and plated on LB plates solidified with 1.5% agar. Colonies were counted 24 and 48 hours later. Student t-test was used to determine statistical significance. Results are presented as mean ± S.E.

Accumulation of MnTE-2-PyP⁵⁺ and FeTE-2-PyP⁵⁺ in E. coli

Mn porphyrins were incubated with LC106, catalase/peroxidase mutant in LB medium for 1 hour with 20 µM of either MnTE-2-PyP⁵⁺ or FeTE-2-PyP⁵⁺. The cells were then washed, centrifuged and the pellet suspended in 2% sodium dodecyl sulfate. The uv/vis spectral analysis was performed as described in [57].

Results and discussion

The following thermodynamic and kinetic data on metalloporphyrins are summarized in Table 1: rate constants for the catalysis of H₂O₂ dismutation, k_cat(H₂O₂) in M⁻¹ s⁻¹; log values of the rate constants for the catalysis of O₂^.− dismutation, log k_cat(O₂^.−); proton dissociation constants for the first axial water, pK_a(ax); proton dissociation constants for the 3^rd basic pyrrolic nitrogen proton of the corresponding metal-free porphyrin, pK_a3; and the metal-centered reduction potentials, E_1/2 in mV vs NHE, for Mn^IIIP/Mn^IIP redox couple. The data for other redox active therapeutics on the catalysis of dismutation of O₂^.− and H₂O₂, along with relevant reduction potentials are provided in Table 2. The k_cat(H₂O₂) for catalase enzyme was determined in parallel under same experimental conditions to allow for comparison. Some of the data are determined in this work, while the others are taken from literature for the sake of discussion. Among Mn porphyrins, only ortho cationic Mn(III) and Fe(III) N-substituted pyridylporphyrins and di-ortho Mn(III) N,N’-diethylimidazolylporphyrin, MnTDE-2-ImP⁵⁺ have minor catalase-like activity, described by k_cat(H₂O₂), ranging between 0.0004 and 0.05 % of enzyme activity. MnTBAP³⁻ has often been incorrectly described as SOD and catalase-mimic and was frequently misused in studies with a goal to provide the evidence for the involvement of O₂^.− and H₂O₂ in pathways explored. It has no SOD-like activity and 10-fold lower k_cat(H₂O₂) than the cationic MnPs such as MnTE-2-PyP⁵⁺ [28, 58]. MnTBAP³⁻ possesses modest ability to reduce ONOO⁻ [17, 58–61]. Recently Doctorovich’s group reported that MnTBAP³⁻ reacts rapidly with HNO [62]. Future research will hopefully address the origin of the therapeutic effects continuously demonstrated with MnTBAP³⁻; for details on possible mechanisms involved in its therapeutic efficacy see also reference [58].

Table 1. The metal-centered reduction potential E_1/2 vs NHE for Mn^IIIP/Mn^IIP, k_cat(H₂O₂) for the catalysis of H₂O₂ dismutation in M⁻¹ s⁻¹, log k_cat(O₂^.−) for the catalysis of O₂^.− dismutation, the proton dissociation constant for 1^st axial water of metalloporphyrins, pK_a(ax) and the proton dissociation constant for the 3^rd basic pyrrolic nitrogen proton of the corresponding metal-free porphyrin, pK_a3.

The catalysis of H₂O₂ dismutation by MnPs was followed in 0.05 M tris buffer, pH 7.8 at (25±1)°C by Clark oxygen electrode. The values for the catalysis of O₂^.− dismutation, determined by cyt c assay in 0.05 M potassium phosphate buffer, pH 7.8 at 25±1°C, are taken from [19, 21, 49, 50]. All E_1/2 values for Mn^IIIP/Mn^IIP redox couple are reported vs the normal hydrogen electrode (NHE), using the potential of MnTE-2-PyP⁵⁺, E_1/2 = + 228 mV vs NHE at pH 7.8, as a reference.

N	Compound	E1/2 (MnIIIP/MnIIP), mV vs NHE	log kcat (O2.−)a	kcat (H2O2), M−1 s−1	pKa(ax)	pKa3
		pH 7.8	pH 7.8	pH 7.8
1	MnTBAP3−	−194	3.16	5.84	12.6	5.5
2	MnTE-2-PyPhP5+	−65	5.55	21.10	12.1
3	MnTE-3-PyP5+	54	6.65	63.25	11.5	~1.8
4	MnTE-4-PyP5+	70	6.86	52.08	~11.6	~1.4
5	MnTnHex-3-PyP5+	66	6.64	46.21	~11.5
6	MnTnHex-4-PyP5+	68	6.75	59.05	~11.6
7	MnTnOct-3-PyP5+	74	6.53	67.44	~11.5
8	MnTE-2-PyP5+	228	7.76	63.32	10.7	−0.9
9	MnTPhE-2-PyP5+	259	7.66	23.54	~10.8
10	MnTnBuOE-2-PyP5+	277	7.83	88.47	10.7
11	MnTnHexOE-2-PyP5+	313	7.92	34.66	10.6
12	MnTnHex-2-PyP5+	314	7.48	28.31	11.0
13	MnTnOct-2-PyP5+	340	7.71	27.62	10.7
14	MnTDE-2-ImP5+	346	7.83	27.59

^ain the absence of SOD enzyme, O₂^.− self-dismutes at pH 7.8 with rate constant of log k(O₂^.−)_{self-dismutation} ~5.7. Therefore, the compounds cannot be functional SOD mimics, if they disproportionate O₂^.− with a log value of rate constant equal to or lower than 5.7.

^bWe have estimated several values of pK_a(ax)^. based on established relationships [25, 50, 68]. The lengthening of methyl chains (in MnTM-2 (or 3 or 4)-PyP⁵⁺) to ethyl chains (in MnTE-2(or 3 or 4)-PyP⁵⁺) has essentially no impact on log k_cat(O₂^.−) and E_1/2 vs NHE for Mn^IIIP/Mn^IIP, assuming therefore safely that it also has no significant impact on pK_a(ax).

Table 2.

The initial rates of O₂ evolution (v_o, nM s⁻¹), maximal amount of O₂ produced (max µM O₂), yield in O₂ (expressed as %), turnover number (TON), and turnover frequency (TOF, s⁻¹).

N	Compound	vo (nM s−1)	Endpoint (max µM O2)	TON	O2 yield (%)	TOF (s−1, based on initial rate)
1	MnTBAP3−	126.1	249.75	12.49	49.95	0.0063
2	MnTE-2-PyPhP5+	133.7	203.35	10.17	40.67	0.0067
3	MnTE-3-PyP5+	700.1	132.72	6.64	26.54	0.0350
4	MnTnHex-3-PyP5+	416.0	78.73	3.94	15.75	0.0208
5	MnTnHex-4-PyP5+	556.3	116.49	5.82	23.30	0.0278
6	MnTE-4-PyP5+	585.0	126.31	6.32	25.26	0.0293
7	MnTnOct-3-PyP5+	563.4	111.06	5.55	22.21	0.0282
8	MnTE-2-PyP5+	617.0	223.10	11.15	44.62	0.0309
9	MnTPhE-2-PyP5+	307.4	124.52	6.23	24.90	0.0154
10	MnTnBuOE-2-PyP5+	913.7	180.33	9.02	36.07	0.0457
11	MnTnHexOE-2-PyP5+	400.5	159.72	7.99	31.94	0.0200
12	MnTnHex-2-PyP5+	367.2	212.42	10.62	42.48	0.0184
13	MnTnOct-2-PyP5+	289.9	197.55	9.88	39.51	0.0145
14	MnTDE-2-ImP5+	291.8	250.85	12.54	50.17	0.0146
15	M40403	29.2	19.51	0.98	3.90	0.0015
16	Mn Salen, EUK-8	137.0	21.4	1.07	4.28	0.0068
17	MnCl2	22.2	7.28	0.36	1.46	0.0011

Catalysis of H₂O₂ dismutation by Mn porphyrins

In our aqueous system, the catalysis of H₂O₂ dismutation by metal complex presumably occurred via Mn^IIIP/(O)₂Mn^VP redox couple. The involvement of di-oxo species has been suggested for imidazolyl analog, MnTDE-2-ImP⁵⁺, whose chemistry is similar to that of MnP pyridyl analogs [63]. The dismutation involves 2-electron transfer as described by equations (3a) and (3b). (Note that axially ligated water molecules are indicated in equations (as they are involved in electron transfer) but are omitted throughout the text, Tables and Figures).

$${({\mathrm{H}}_2\mathrm{O})}_2{\text{Mn}}^{\text{III}}{\mathrm{P}}^{5+}+{\mathrm{H}}_2{\mathrm{O}}_2<=======>{(\mathrm{O})}_2{\text{Mn}}^{\mathrm{V}}{\mathrm{P}}^{3+}+2{\mathrm{H}}_2\mathrm{O}+2{\mathrm{H}}^{+}$$ (3a)

$${(\mathrm{O})}_2{\text{Mn}}^{\mathrm{V}}{\mathrm{P}}^{3+}+{\mathrm{H}}_2{\mathrm{O}}_2+2{\mathrm{H}}^{+}<=======>{({\mathrm{H}}_2\mathrm{O})}_2{\text{Mn}}^{\text{III}}{\mathrm{P}}^{5+}+{\mathrm{O}}_2$$ (3b)

In vivo, the cellular reductants (ascorbate or simple thiols or tetrahidrobopterin or thiol-bearing proteins) would rapidly reduce Fe^IIIPs and Mn^IIIPs to their corresponding M^IIPs [20, 21, 43, 64, 65]. In such scenario the dismutation would most likely involve the Mn^IIP/O=Mn^IVP redox couple as indicated in equations [4a] and [4b].

$$({\mathrm{H}}_2\mathrm{O}){\text{Mn}}^{\text{II}}{\mathrm{P}}^{4+}+{\mathrm{H}}_2{\mathrm{O}}_2<=======>({\mathrm{H}}_2\mathrm{O})(\mathrm{O}){\text{Mn}}^{\text{IV}}{\mathrm{P}}^{4+}+{\mathrm{H}}_2\mathrm{O}$$ (4a)

$$({\mathrm{H}}_2\mathrm{O})(\mathrm{O}){\text{Mn}}^{\text{IV}}{\mathrm{P}}^{4+}+{\mathrm{H}}_2{\mathrm{O}}_2<=======>({\mathrm{H}}_2\mathrm{O}){\text{Mn}}^{\text{II}}{\mathrm{P}}^{4+}+{\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}$$ (4b)

Further, a very likely in vivo scenario involving MnP and H₂O₂, given the low concentrations of H₂O₂ and high concentrations of cellular reductants, would be as follows: MnP gets oxidized with H₂O₂, acting as H₂O₂ reductase (eqs (3a) and (4a)) and the high-valent oxo complex gets reduced back with cellular reductants instead with H₂O₂ (eqs (3b) and (4b)) [21, 43, 44]. Such scenario was proposed for the peroxynitrite reductase activity of MnP; when coupled with cellular reductants the action of MnP upon ONOO⁻ becomes catalytic [66–68]. Similar scenario has been also proposed for O₂^.− dismutation: instead of oxidizing and reducing O₂^.−, MnP would act as superoxide reductase closing a catalytic cycle with cellular reductants [19, 21].

The E_1/2 values reported for O=Mn^IVP/Mn^IIIP as well as for (O)₂Mn^VP/Mn^IIIP are essentially identical for different Mn and Fe porphyrins [66, 69–76]. For example the E_1/2 values for the Mn^IIIP/Mn^IIP redox couple of MnTE-2-PyP⁵⁺, MnTE-3-PyP⁵⁺, MnTnBu-2-PyP⁵⁺, and MnTnBuOE-2-PyP⁵⁺ differ by as much as 223 mV (Table 1). Yet, the E_1/2 values for their O=Mn^IVP/Mn^IIIP redox couple differ by 32 mV only and are +509, +529, and +509, and +541 mV vs SHE at pH 11, respectively (values vs NHE and vs SHE differ by ~5.7 mV) [77]. Further, E_1/2 values for the O=Mn^IVP/Mn^IIIP redox couple of MnTM-2-PyP⁵⁺, MnTM-3-PyP⁵⁺ and MnTM-4-PyP⁵⁺ differ by 14 mV only and are +540, +526 and +532 mV vs NHE at pH 11, while E_1/2 values for their Mn^IIIP/Mn^IIP redox couples differ by up to 168 mV (Table 1) [66]. Finally, E_1/2 for (O)₂Mn^VP/Mn^IIIP redox couple for MnTM-4-PyP⁵⁺, MnTM-2-PyP⁵⁺ and MnTDM-2-ImP⁵⁺ at pH 11 are all around +800 mV vs NHE, while the E_1/2 for their Mn^IIIP/Mn^IIP redox couples differ by up to 286 mV (Table 1) [78]. The E_1/2 for two-electron transfer (from Mn +2 to +4) involving Mn^IIP/O=Mn^IVP redox couple is somewhat different for each Mn porphyrin as it reflects the differences in E_1/2 for their Mn^IIIP/Mn^IIP redox couples.

Impact of electron-density of the metal center of Mn porphyrins on the catalysis of H₂O₂ dismutation

The H₂O₂ dismutation involves oxidation of Mn^IIIP to (O)₂Mn^VP (eq 3a) and subsequent reduction of (O)₂Mn^VP back to Mn^IIIP (eq 3b). Oxidation of Mn^IIIP to (O)₂Mn^VP involves H₂O₂ binding to the metal site of metalloporphyrins in a 1^st step; thus the reduction of H₂O₂ occurs via inner-sphere electron transfer. The two-electron transfer in a 2^nd step of catalysis, resulting in (O)₂Mn^VP (eq (3a)), appears to be similar for all MnPs due to nearly identical E_1/2 values for (O)₂Mn^VP/Mn^IIIP (see above under Catalysis of H₂O₂ dismutation by Mn porphyrins). The 1^st step is dependent upon the electron density of the metal site; the more electron-deficient the metal site is, the higher its affinity for H₂O₂ binding. The electron-deficiency of the metal site is best described with proton-dissociation constant of the 1^st axial water, pK_a(ax), which refers to the following equilibrium: (H₂O)₂MnP⁵⁺ ↔ (HO)(H₂O)MnP⁴⁺ + H⁺. With more electron-deficient Mn site, the axial binding of water oxygen is stronger. That results in weakening of the H—OH bond. Consequently, proton dissociates more readily from the axial water giving rise to (HO)(H₂O)MnP⁴⁺. In turn, MnPs with more electron-deficient metal center, such as MnTE-2-PyP⁵⁺, have lower value of pK_a(ax) than those with electron-rich Mn, such as MnTBAP³⁻ (Table 1). The pK_a(ax) was previously shown to relate linearly but inversely to E_1/2 of Mn^IIIP/Mn^IIP redox couple; such relationship is here confirmed on new MnPs (Figure 4C) [50, 68]. Consequently, plots A and B in Figure 4 are essentially of identical shape. In other words, both pK_a(ax) and E_1/2 of Mn^IIIP/Mn^IIP redox couple describe the propensity of MnP to react with any species which involves its binding to the Mn site, such as ONOO⁻, H₂O₂ and ClO⁻. In summary, high electron-deficiency of Mn site supports: (i) low pK_a(ax); (ii) high E_1/2; (iii) high propensity of Mn to acquire electrons via axial binding of either water or any other species; and therefore (iv) high k_cat(H₂O₂). Similar impact of electron-density of Mn site was demonstrated in the catalysis of O₂^.− dismutation [25]; in such case, though, the electron hopping (outer-sphere electron transfer), and not O₂^.− binding to the metal site, is predominantly involved in the catalysis of O₂^.− dismutation. However, higher electron-deficiency is necessary for Mn site to be reduced from +3 to + 2 in a 1^st step of dismutation process. In turn, the higher the E_1/2 of Mn^IIIP/Mn^IIP redox couple the higher is the log k_cat(O₂^.−) (Figure 4A). Consequently, the catalase-like activity, log k_cat(H₂O₂) parallels the SOD-like activity, log k_cat(O₂^.−) (Figure 4C); in other words the more potent the SOD mimic is, the more able it is to catalyze H₂O₂ dismutation.

Figure 4. The relation between the log k_cat(H₂O₂) and log k_cat(O₂^.−) each with E_1/2 of Mn^IIIP/Mn^IIIP redox couple in mV vs NHE (A) and with proton dissociation constant of axial water, pK_a(ax) (B); the relation between the pK_a(ax) and E_1/2 for Mn^IIIP/Mn^IIP redox couple (C); and the relationship between the log k_cat(H₂O₂) and log k_cat(O₂^.−) (D).

The ability of MnP to catalyze H₂O₂ dismutation depends on H₂O₂ binding in a 1^st step of dismutation process and thus on the electron-deficiency of the metal site which is described with pK_a(ax) (B). The E_1/2 for any couple that involves species in high-oxidation states (eqs [3] and [4]) are similar for all Mn porphyrins (see under Catalysis of H₂O₂ dismutation by Mn porphyrins); thus the 2^nd step, electron transfer, has no impact on the magnitude of k_cat(H₂O₂). In turn, the pK_a(ax) was reported and confirmed here with new MnPs [49] to parallel the E_1/2 of Mn^IIIP/Mn^IIP redox couple (C) [50, 68]. Consequently, the log k_cat(H₂O₂) paralells the E_1/2 of Mn^IIIP/Mn^IIP redox couple (A) though this couple is not involved in H₂O₂ dismutation. Since the E_1/2 of Mn^IIIP/Mn^IIP redox couple controls the catalysis of O₂^.− dismutation also, the log k_cat(H₂O₂) is proportional to log k_cat(O₂^.−) (D). Please note that in plots A and B blue rhombs relate to O₂^.− dismutation while red squares relate to H₂O₂ dismuation. To help identify MnPs on plots please consult also Table 1. We have drawn dashed lines to indicate different trends in the catalysis of O₂^.− (blue) and H₂O₂ dismutations by MnPs (red). Please note that with a predominantly outer-sphere electron transfer in catalysis of O₂^.− dismutation (which involves electron hoping and not O₂^.− binding) we have a different behavior when compared to the catalysis of H₂O₂ dismutation by MnPs; the latter involves binding of H₂O₂ to the metal site (plots A and B). In such scenario, steric hindrance, imposed by longer alkyl chains with and without polar oxygen atoms within chains, plays significant role. Encircled are the MnPs with long and bulky ortho N-pyridyl substituents: MnTnOct-2-PyP⁵⁺, MnTnHex-2-PyP⁵⁺, MnTnHexOE-2-PyP⁵⁺, MnTPhE-2-PyP⁵⁺.

Those Mn porphyrins that bear sterically hindered moieties, such as long or bulky ortho N-pyridyl substituents (MnTnOct-2-PyP⁵⁺, MnTnHex-2-PyP⁵⁺, MnTnHexOE-2-PyP⁵⁺, MnTPhE-2-PyP⁵⁺ (encircled in Figure 4, plots A, B and D), while having similar E_1/2 value for Mn^IIIP/Mn^IIP redox couple, have the k_cat(H₂O₂) values 2–3-fold lower than that of MnTE-2-PyP⁵⁺. Sterics plays larger role with inner- (H₂O₂ dismutation) than with outer-sphere electron transfer (O₂^.−dismutation). Thus during H₂O₂ dismutation longer pyridyl substituents hinder the approach of H₂O₂ to the metal site. Besides sterics, the electronic factors contribute to the magnitude of k_cat(H₂O₂) also. MnTnBuOE-2-PyP⁵⁺ has long butoxyethyl pyridyl substituents each of which contains one oxygen atom. The oxygen atoms are exposed to the immediate environment of porphyrin and facilitate the approach of H₂O₂ via hydrogen bonding (#10 red square in Figure 4). Same dramatic impact of hydrogen bonding was demonstrated on the metalation rate of H₂TnBuOE-2-PyP⁴⁺ [47]. With MnTnHexOE-2-PyP⁵⁺, however, oxygen atoms are buried deeper into the hexoxyethyl chains resulting in lesser impact; in turn the log k_cat(H₂O₂) is closer to the value for alkyl analogs, MnTnHex(or Oct)-2-PyP⁵⁺ [49]. The hindrance of the oxygen atoms from the solvent is also reflected in the lipophilicity of MnTnHexOE-2-PyP⁵⁺ which is closer to the one of the alkyl analog of same number of carbon atoms in pyridyl substituents, MnTnNon-2-PyP⁵⁺.

Impact of stability of Mn porphyrins towards H₂O₂ – driven oxidative degradation on the catalysis of H₂O₂ dismutation

The interplay between the k_cat(H₂O₂) and the stability of MnPs towards degradation determines the yield of H₂O₂ dismutation – i.e. the yield in O₂ evolution. The O₂ yield describes best the quality of an catalyst. In addition to O₂ yield (%), the maximal amount of O₂ (µM) produced during catalysis, the turnover number (TON), and turn-over frequency (TOF, s⁻¹) were determined (Table 2). The two-electron process during H₂O₂ dismutation involves the formation of species in high-oxidation states of Mn with high oxidizing power. Such species can either oxidize H₂O₂ closing the catalase-like cycle, or oxidize itself or another catalyst with its subsequent degradation [79]. Therefore the stability of Mn porphyrins towards oxidative degradation was assessed.

Stability of MnP towards oxidative degradation is influenced by: (i) the protonation equilibrium at inner pyrrolic nitrogens of the corresponding metal-free ligand, pK_a3; and (ii) steric and electronic factors. The first two pyrrolic nitrogen protons of the metal-free porphyrin are very acidic and their dissociation constants are not accessible experimentally. The pK_a3 of 3^rd (basic) inner pyrrolic nitrogen proton (H₃P⁺ <==> H₂P + H⁺) has been determined for ortho, meta and para isomers of MnTMPyP⁵⁺ and MnTBAP³⁻ (Table 1) [25]. Based on established relationships [19, 21, 25] the pK_a3 of isomeric ethyl analogs, MnTEPyP⁵⁺, are presumably very similar. The higher the pK_a3 is, the less prone the protons are to leave pyrrolic nitrogens. The affinity of pyrrolic nitrogens to protons reflects in turn their affinity to manganese in Mn complex: the higher the pK_a3, the more stable is the metal complex. The pK_a3 is inversely proportional to E_1/2 of Mn^IIIP/Mn^IIP redox couple, i.e., the higher the E_1/2 the less stable are the Mn porphyrins. MnTBAP³⁻ with pK_a3 of 5.5 is much more stable complex than MnTE-2-PyP⁵⁺ which has pK_a3 of −0.9 [25]. Such stability towards the loss of Mn contributes also to the stability of MnPs towards oxidative degradation. This is also exemplified with Fe corrole; its high metal/ligand stability contributes to its stability as a catalyst of H₂O₂ dismutation (see below under Fe corrole). With pK_a3 =5.5, the anionic electron-rich MnTBAP³⁻ is a stable metal complex; during 30 min it does not undergo any degradation with H₂O₂ (Figure 5A and D). However, cationic, electron-deficient MnTE-2-PyP⁵⁺, with ~ 6 log units lower pK_a3 = −0.9, undergoes much faster degradation (Figure 5B and D). When left in the presence of H₂O₂ for ~200 s, MnTE-2-PyP⁵⁺ was degraded to a point of no return; in turn the addition of GSH fails to restore it (Figure 5C) [66, 67].

Figure 5. Degradation of Mn porphyrins with H₂O₂.

Spectral changes were measured within first 30 minutes for Mn porphyrins MnTBAP³⁻ (A) and MnTE-2-PyP⁵⁺ (B)). Time-dependent reduction in the absorbance at the Soret band for each MnP is shown in inset. (C) Time-dependent degradation of MnTE-2-PyP⁵⁺ with H₂O₂. The degradation of MnTE-2-PyP⁵⁺ was terminated with the addition of catalase. Subsequent addition of GSH did not cause any change in the absorbance indicating irreversible degradation of MnP to non-porphyrin species over the course of ~200 s of reaction. The drop in absorbance is due to a sample dilution. Experiments were carried out at (25±1)°C in 0.05 M tris buffer, pH 7.8 and 0.1 mM EDTA with 10 µM MnP and 0.5 mM H₂O₂. (D) The H₂O₂-driven degradation of several MnPs whose structures impose different steric and electronic effects upon H₂O₂ approach.

To address the impact of sterics and electronics, the H₂O₂-driven oxidative degradation of several other MnPs were explored and shown in Figure 5: two alkoxyalkyl derivatives, MnTnBuOE-2-PyP⁵⁺ (#10) and MnTnHexOE-2-PyP⁵⁺ (#11), and N,N’-diethylimidazolyl derivative, MnTDE-2-ImP⁵⁺ (#14) (Figure 5D). The steric and electronic factors of those MnPs are different from each other and from those of ortho Mn(III) N-alkylpyridylporphyrins. MnTnBuOE-2-PyP⁵⁺ and MnTnHexOE-2-PyP⁵⁺ have oxygen atoms in pyridyl substituents, while MnTDE-2-ImP⁵⁺ is di-ortho N,N’-diethylimidazolyl analog with alkyl chains placed vertically above and below the porphyrin plane.

All have similar values of thermodynamic parameters as ortho Mn(III) N-alkylpyridylporphyrins (such as E_1/2 and pK_a(ax)). They also have similar values for k_cat(O₂^.−) as catalysis of O₂^.− dismutation has predominantly outer-sphere character and does not involve O₂^.− binding to the Mn site. Yet their structural differences greatly impact their ability to catalyze H₂O₂ dismutation which process involves H₂O₂ binding to the metal site. The higher stability of MnTDE-2-imP⁵⁺ towards H₂O₂, than that of ortho Mn(III) N-substituted pyridylporphyrins, is afforded by the protection of Mn site with alkyl chains placed both above and below the porphyrin plane (Figure 5D). In the case of alkoxyalkyl MnTnBuOE-2-PyP⁵⁺, oxygen atoms within N-substituents facilitate the catalysis of H₂O₂ dismutation via hydrogen bonding, but destabilize the complex towards oxidative degradation (Figures 5D). Another alkoxyalkyl analog, MnTnHexOE-2-PyP⁵⁺, is similarly lipophilic to MnP with identical number of carbon atoms in pyridyl substituents, MnTnNon-2-PyP⁵⁺. This indicates that oxygen atoms in MnTnHexOE-2-PyP⁵⁺ have less impact on molecule-solvent interactions than in MnTnBuOE-2-PyP⁵⁺ and thus are only marginally involved in facilitation of H₂O₂ dismutation. For that same reason MnTnHexOE-2-PyP⁵⁺ is more stable towards degradation (Figure 5D). In turn the k_cat(H₂O₂) of MnTnHexOE-2-PyP⁵⁺ is 2.3-fold lower than MnTnBuOE-2-PyP⁵⁺ and similar to its alkyl analogs, MnTnHex(or Oct)-2-PyP⁵⁺ (Table 1). The interplay between their ability to catalyze H₂O₂ dismutation and their stability towards oxidative degradation gave rise to the similar O₂ production yields of MnTnBuOE-2-PyP⁵⁺ (#10) and MnTnHexOE-2-PyP⁵⁺ (#11) (Table 2).

Correlation between the rate of H₂O₂ dismutation and its yield

The impact of steric and electronic effects as well as of the Mn site electron density on the catalysis of H₂O₂ dismutation are best demonstrated in Figure 6 where the O₂ yield and the log k_cat(H₂O₂) are correlated with E_1/2 of Mn^III/Mn^II redox couple. The bell-shaped relationships indicate that there is an interplay between the log k_cat(H₂O₂) and stability of the compounds which controls the yield of H₂O₂ dismutation. Mn(III) N-alkylpyridylporphyrins, whose metal sites are electron- deficient (higher E_1/2 and lower values of pK_a(ax)) and favor H₂O₂ binding, such as MnTE-2-PyP⁵⁺(#8), are faster in catalyzing H₂O₂ dismutation than anionic electron-rich porphyrins such as MnTBAP³⁻ (#1). While MnTBAP³⁻ catalyzes H₂O₂ dismutation with a very low rate constant (due to vastly unfavorable E_1/2), the higher metal/ligand stability allows it to resist oxidative degradation and thus be able to give rise to a yield even slightly higher than that of a less stable MnTE-2-PyP⁵⁺. Differences in electronic and steric factors of alkoxyalkyl derivatives MnTnHexOE-2-PyP⁵⁺ (#11) and MnTnBuOE-2-PyP⁵⁺ (#10) result in different log k_cat(H₂O₂) and stabilities to oxidative degradation. Yet both afford similar O₂ yield which is an ultimate measure of their catalytic potency. The increased stability of MnTDE-2-ImP⁵⁺ (#14) (see above) affords highest yield among cationic Mn porphyrins.

Figure 6. The relationships between the kinetic (log k_cat(H₂O₂)) and thermodynamic parameter (E_1/2 for M^III/M^II redox couple) each with thermodynamic parameter, yield of O₂.

While of similar E_1/2, the presence of relatively approachable oxygen atoms in alkoxyalkyl chains in MnTnBuOE-2-PyP⁵⁺ (#10) relative to MnTnHexOE-2-PyP⁵⁺ (#11) resulted in 2.3-fold higher k_cat(H₂O₂); yet, yields in O₂ are identical, due to larger stability of latter compound (Figure 5). MnTDE-2-ImP⁵⁺ (#14), with identical k_cat(H₂O₂) but higher stability than MnTnBuOE-2-PyP⁵⁺, has higher yield which is also the highest among cationic MnPs. The interplay between the ability of MnPs to catalyze H₂O₂ dismutation (red line) and their stability to oxidative degradation determines the yield of H₂O₂ dismutation and the efficacy of catalyst (blue line). Dashed lines are meant to indicate the trends.

H₂O₂ dismutation by other metal complexes

Mn(II) cyclic polyamines

Mn(II) cyclic polyamine M40403, a potent SOD mimic, has essentially no catalase-like activity (Table 3). When compared to metalloporphyrins, M40403 with Mn in its +2 oxidation state has sufficient electron density not to favor H₂O₂ binding; its k_cat(H₂O₂) = 8.18 M⁻¹ s⁻¹. The same reasoning may apply to MnCl₂ with Mn in +2 oxidation state. The metal/ligand stability of Mn(II) cyclic polyamine M40403 is only log K =13.6 [80], which would favor catalyst decomposition. Its turnover number is therefore extremely low, TON = 0.98 (Table 2). No differences in any of the parameters were found between SOD-active M40403 and SOD-inactive M40404 analog. The vastly stronger metal complex MnTBAP³⁻, with essentially identically low k_cat(H₂O₂) = 5.84 M⁻¹ s⁻¹ has ~10-fold higher TON of 12.49, equal to the one of MnTE-2-PyP⁵⁺(TON = 11.15). The latter however is ~10-fold stronger catalase mimic k_cat(H₂O₂) = 63.32 M⁻¹ s⁻¹, yet less stable metal complex (Tables 1, 2 and 3).

Mn salen, EUK-8

Mn salen complexes were claimed to be advantageous over Mn porphyrins due to their combined SOD- and catalase-like activities. The reported data show no catalase-like activity which was confirmed in this work on EUK-8 analog [41, 81].

Fe porphyrins

FePs are more sensitive to oxidative degradation with H₂O₂ than are MnPs. In turn, despite an order of magnitude higher k_cat(H₂O₂), FePs are not superior catalysts to MnPs (Tables 2 and 3).

Fe corrole

Among Fe corroles, anionic disulfonato Fe corrole, FeTrF₅Ph-β(SO₃)₂-corrole²⁻ has the highest catalase-like activity, while cationic Mn corroles have no catalase-like activity [82]. When normalized to same conditions, Fe corrole has only a few fold-higher catalase-like activity than Fe porphyrins: log k_cat(H₂O₂) for Fe corrole = 3.81 at 37 °C [53], while log k_cat(H₂O₂) = 3.28 at 37°C for FeTE-2-PyP⁵⁺ [calculated using Arrhenius equation and log k_cat(H₂O₂) = 2.90 at 25 °C]. Yet, FeTrF₅Ph-β(SO₃)₂-corrole²⁻ is 37-fold less potent SOD mimic. The reason for higher catalase-like activity is the tri-anionic nature of a corrole ligand which gives rise to much stronger metal/ligand interactions and in turn higher stability of Fe/corrole complex than that observed with Fe/porphyrin complex. The higher stability of Fe/corrole complex allows it to survive longer redox-cycling with H₂O₂ and in turn gives rise to larger turnover number. The oxidizing potential of high-valent Fe corrole and its cycling with cellular reductants has only vaguely been discussed in the literature but is likely playing a role in vitro and in vivo due to the high levels of endogenous reductants. Its excellent catalase-activity has been discussed as the “main superiority of FeTrF₅Ph-β(SO₃)₂-corrole²⁻ relative to other synthetic metal complexes”. Yet, a Mn(III) meso bis(N-methylpyridium-4-yl)-mono-pentafluorophenylcorrole, with no catalase-like activity, was superior to FeTrF₅Ph-β(SO₃)₂-corrole²⁻ in several in vitro and in vivo experiments [82].

Evaluation of the redox-active drugs in an E. coli model of H₂O₂-induced damage

Hydrogen peroxide is unavoidable product of aerobic metabolism and organisms have developed efficient systems for its removal. Such systems and the mechanisms of their regulation are particularly well studied in E. coli. As in mammalian cells, the protection against H₂O₂ in E. coli is so efficient that peroxide concentration is normally kept below 10 nM [84, 85]. On such background it would be practically impossible to notice protection by compounds which display only a tiny fraction of the activity of natural enzymes. Therefore E. coli LC106 mutant, lacking both catalase and peroxidase activities [55], was used in our study along with its parental strain MG1655. Two different experimental scenarios were performed. In 1^st scenario, we have added compounds simultaneously with H₂O₂ to E. coli culture so that they interact with each other in the medium. In 2^nd scenario, E. coli culture was incubated with compounds for 1 hour, which allowed them to accumulate within E. coli. The culture was then exposed to H₂O₂.

1. Under 1^st experimental conditions 20 µM compounds were added in parallel with 0.5 mM H₂O₂ to the medium containing catalase/peroxidase-deficient LC106 strain (Figure 7A). Among tested Mn complexes, only MnTE-2-PyP⁵⁺, MnTDE-2-ImP⁵⁺ and MnTBAP³⁻ were protective to catalase/peroxidase deficient LC106 strain (Figure 7). The higher stability of MnTBAP³⁻ compensated for the lower k_cat(H₂O₂) relative to MnTE-2-PyP⁵⁺ and MnTDE-2-ImP⁵⁺ (see Discussion and Tables 1–2). In turn, these MnPs were able to protect E. coli against H₂O₂ toxicity to a similar extent (Figure 7A). Mn(II) cyclic polyamines, M40403 and M40404 contain Mn in +2 oxidation state and are thus very unstable complexes [19, 43]. In turn, they undergo fast degradation with H₂O₂ affording low yield of H₂O₂ dismutation (Table 2). Consequently no protection of E. coli against exogenous H₂O₂ was demonstrated with M40403 and M40404 (Figure 7A). FeTE-2-PyP⁵⁺ was protective to a similar extent as MnTE-2-PyP⁵⁺, MnTDE-2-ImP⁵⁺ and MnTBAP³⁻. While FeTE-2-PyP⁵⁺ has a higher k_cat(H₂O₂), it undergoes much faster oxidative degradation, releasing “free” Fe which is schematically depicted in Figure 9. The accumulation of “free” Fe within cell (its transport from the medium into the cell) is tightly controlled by E. coli and is in turn precluded to any dangerous level [50]. In turn under such conditions FeTE-2-PyP⁵⁺ was beneficial: in mutual interactions both FeTE-2-PyP⁵⁺ and H₂O₂ were removed from the culture which prevented significant cellular accumulation of FeTE-2-PyP⁵⁺ to toxic levels as seen in 2^nd experiment (for details on the impact of Fe porphyrin on the growth of E. coli see [50]) (Figure 8 and Figure 9). Data are in agreement with the lack of toxicity of extracellular Fe reported by others [86, 87]. As anticipated, Mn salen (EUK-8) and tempol showed no effect in such scenario (data not shown). EUK-8 has low k_cat(H₂O₂) and low yield (Table 1 and 2), while no catalase-like activity (no k_cat(H₂O₂)) was demonstrated with Tempol (Table 3). None of the tested compounds protected the parental strain MG1655 against 10-fold higher 5 mM H₂O₂ (Figure 7B). When compared to mutant LC106, 10-fold higher concentration of H₂O₂ (5 mM H₂O₂) is required to cause comparable viability loss in the parental strain MG1655 under same experimental conditions. The low catalase-like activity and low stability of the compounds precludes dismutation of 5 mM H₂O₂ to significant extent (Figure 7B).

Figure 7. Evaluation of redox-active drugs in an E. coli model of H₂O₂-induced damage - drugs were added to E coli–containing medium in parallel with H₂O₂.

20 µM compounds were exposed to H₂O₂ for 15 min. The parental (MG1655) strain (A) and catalase/peroxidase-deficient mutant (LC106) (B) were studied; with parental MG1655 5 mM H₂O₂, and with catalase/peroxidase LC106 0.5 mM H₂O₂ was used. The viability was measured via MTT test and expressed as a percentage of the MTT reduction by non-treated cells. At 5 mM H₂O₂, no compound was efficacious in protecting parental strain. At 0.5 mM H₂O₂, the Mn complexes that are relatively stable, or have modest k_cat(H₂O₂), that results in relatively high O₂ production yield, were able to dismute and remove H₂O₂, protecting cell against it. Such interaction, in the case of Fe porphyrin, eliminated both the toxicity of FeP and H₂O₂. Student t-test was used to determine the statistical significance. Mean ± S.E is presented. ^# statistical significance (p<0.05) compared to untreated cells; ^$ statistical significance (p<0.05) compared to cells treated with H₂O₂ only.

Figure 9. Proposed mechanism of FeP/H₂O₂ interactions impacting E. coli survival.

In plot (A), FeP was added to E.coli-containing medium for 1 hour prior to 15 min-exposure of culture to H₂O₂. At 15 min, all H₂O₂ was removed by the addition of catalase. In plot (B), FeP was added in parallel with H₂O₂. Under such conditions cycling of FeP with H₂O₂ releases ”free” Fe from the porphyrin ring, whose uptake is tightly controlled by E.coli; in turn FeP toxicity, seen when cells were preincubated with FeP (Figure 8), was avoided (Figure 7). In pre-incubation scenario (A), the toxicity of FeP may arise from: (i) Fenton chemistry of either “free” Fe released from Fe porphyrin or of Fe site of Fe^IIP⁴⁺; (ii) high-valent oxo FeP species of high oxidizing power; and (iii) direct interaction of FeP with specific cellular proteins/targets. In scenario where FeP catalyst and H₂O₂ were added simultaneously (plot B), their mutual interaction dismuted (removed) H₂O₂ only when 0.5 mM H₂O₂ (LC106) was applied to the cells but not 5 mM (MG1655). Their interaction also degrades FeP and in turn cells do not accumulate it within cell, where, otherwise, it might have caused toxicity as shown in (A). The accumulation of “free” Fe is tightly controlled by cell to avoid Fenton-based toxicity. MP – metalloporphyrin, either MnP or FeP.

Figure 8. Evaluation of metal complexes in an E. coli model (catalase/peroxidase-deficient LC106 strain) of H₂O₂-induced damage – drugs were pre-incubated with E. coli prior to H₂O₂ addition.

Except Mn salen, EUK-8, all other Mn complexes studied are the same as those in Figure 7. Two new abbreviations are introduced, MnE2 being MnTE-2-PyP⁵⁺ and MnDE2 being MnTDE-2-ImP⁵⁺. After 1 hour pre-incubation of E. coli with 20 µM of the drugs, the catalase/peroxidase mutant (LC106) cells were washed with PBS and exposed to 0.5 mM H₂O₂. After 15 min of incubation, H₂O₂ was decomposed by adding 1,000 units/ml of catalase (A). The viability was measured via MTT test and expressed as a percentage of the MTT reduction by non-treated cells. None of the compounds were toxic, but were also not able to suppress H₂O₂ toxicity. Fe porphyrin, FeTE-2-PyP⁵⁺, was toxic under given conditions. To verify the presence of pentacationic porphyrins in cells, the accumulation of MnTE-2-PyP⁵⁺ vs FeTE-2-PyP⁵⁺ during 1 hour of incubation was determined and appeared similar (plot B). Student t-test was used to determine the statistical significance. Mean ± S.E is presented. ^$statistical significance (p<0.05) compared to untreated cells; ^# statistical significance (p<0.05) compared to cells treated with H₂O₂ only.

2. Catalase/peroxidase-deficient E.coli strain was incubated with metal complexes for 1 hour, which allows them to accumulate within the cell and act intracellularly (Figure 8) [57]. We have reported that cationic Mn(III) N-substituted Mn and Fe pyridylporphyrins have different chemistry and in turn different biology. Neither MnTE-2-PyP⁵⁺ nor FeTE-2-PyP⁵⁺ at 20 µM concentration protected E.coli (catalase/peroxidase-deficient strain) against 0.5 mM H₂O₂ (Figure 8A) when pre-incubated for 1 hour prior addition of H₂O₂. As with 1^st experiment, addition of catalase to the medium, prior to the addition of H₂O₂, completely prevented the H₂O₂-induced loss of viability (data not shown). The same conclusion is valid for other Mn complexes.

Relative to the 1^st study, the lack of effect is presumably due to the low k_cat(H₂O₂) and low intracellular levels of compounds and H₂O₂. Therefore, neither MnTE-2-PyP⁵⁺ nor FeTE-2-PyP⁵⁺ no any other metal complex studied can be considered catalase mimic under biologically relevant conditions. None of the Mn compounds, including MnTE-2-PyP⁵⁺, were toxic in contrast to FeTE-2-PyP⁵⁺ which exerted toxicity to both strains (Figure 8A, data not shown for MG1655); the data are in agreement with our earlier observations [50].

To provide evidence that the efficacy in protecting cell against H₂O₂ toxicity is due to the ability of pentacationic porphyrins to cross E.coli membrane, we have measured the accumulation of MnTE-2-PyP⁵⁺ and FeTE-2-PyP⁵⁺ in catalase/peroxidase deficient LC106 strain (Figure 8B). The data obtained on LC106 strain are in agreement with our earlier observations on the accumulation of Mn porphyrins in cell wall and cytosol of SOD-deficient JI132strain and wild type AB1157 E.coli strains [50, 57]. The accumulation of Mn porphyrins increases with their lipophilicity, which is in turn controlled by the length of N-alkylpyridyl chains [57, 88, 89]. Accumulation of Mn porphyrins in yeast, Saccharomyces cerevisiae [90], and in mammalian cells [91] was reported also. Importantly, several Mn porphyrins were found to accumulate in mouse heart and brain mitochondria [20, 92–94]. Efficacy studies, where different Mn porphyrins mimicked MnSOD, support their mitochondrial localization [21, 43].

Fe porphyrin accumulates in E. coli to a similar level as MnTE-2-PyP⁵⁺, supposedly via heme uptake mechanism. The toxicity of FeP is not yet fully explored. Based on aqueous chemistry and in vitro studies [19, 21, 25, 50] various factors may contribute to the toxicity of FePs, such as: (i) Fenton chemistry of either “free” Fe released from Fe porphyrin or of Fe site of Fe^IIP⁴⁺; (ii) high-valent oxo FeP species of high oxidizing power; and (iii) direct interaction of FeP with specific cellular proteins/targets (Figure 9).

Our results provide substantial evidence that even the most efficient SOD mimics (like MnTE-2-PyP⁵⁺ and M40403) do not have sufficient catalase-like activity in order to protect organisms from H₂O₂ under biological nano- and submicromolar levels of H₂O₂ in catalase-like fashion.

Conclusions

Mn- and Fe(III) porphyrins possess at most 0.0004% and 0.05% of the activity of catalase, while the values are orders of magnitude lower for other redox-active compounds often studied in various oxidative stress in vitro and in vivo models. In vivo efficacy in suppressing H₂O₂-mediated loss of viability has been evaluated in wild type and catalase/peroxidase-deficient E. coli strains. When the compounds were allowed to accumulate within the cell to act there as catalase mimics, the protection against extracellular H₂O₂ was not observed. Protection against H₂O₂ has only been demonstrated when E. coli was incubated simultaneously with high medium concentrations of FeP/MnP and H₂O₂. Under such conditions the interaction of FeP/MnP with H₂O₂ happens (dismutation with subsequent degradation), resulting in a release of “free” Mn/Fe. Since the uptake of “free” Fe is tightly controlled by E. coli, we did not observe any toxicity that might have been otherwise seen with intracellular “free” Fe.

Removal of H₂O₂ with Mn and Fe porphyrins in a catalase-like fashion, given low k_cat(H₂O₂), and low intracellular H₂O₂ concentrations, seems not to be biologically relevant.

Emerging data point out to the involvement of H₂O₂ in the oxidation of thiols catalyzed by Mn porphyrins whereby signaling proteins are inhibited; in other words rather than simply removing H₂O₂, during its reduction, Mn porphyrins use it for therapeutic advantage affecting favorably cellular transcription [23, 24, 43, 44, 95]. Such scenario predicts that MnP mechanism of action is most likely of peroxidase-or thiol-oxidase- rather than catalase-type [21, 43, 65]. Studies are in progress to gain further insight into H₂O₂-related mechanism(s) of action(s) of Mn porphyrins.

Highlights.

• -Catalase-like activity of different classes of redox active compounds, described by k_cat(H₂O₂) and yield of O₂ evolution, was assessed
• -Mn(III) and Fe(III) porphyrins have catalase-like activity in the range of 0.0004% to 0.05% of the enzyme activity, respectively
• -Mn salen EUK-8 has 0.00135% of enzyme activity, while M40403, M40404 and Tempol have no detectable catalase-like activity
• -Low catalase-like activities of compounds studied must be taken into account when drawing conclusions on their mechanism(s) of action(s) in vitro an in vivo

Abbreviations

O₂^•−

superoxide

ONOO⁻

peroxynitrite

ClO⁻

hypochlorite

MnTE-2-PyP⁵⁺

Mn(III) meso-tetrakis(N-ethylpyridinium-2-yl)porphyrin (also known as AEOL10113)

MnTE-3(or 4)-PyP⁵⁺

Mn(III) meso-tetrakis(N-ethylpyridinium-3(or 4)-yl)porphyrin

MnTnHex-2(3 or 4)-PyP⁵⁺

Mn(III) meso-tetrakis(N-n-hexylpyridinium-2(3 or 4)-yl)porphyrin

MnTnOct-2(or 3)-PyP⁵⁺

Mn(III) meso-tetrakis(N-n-octylpyridinium-2(or 3)-yl)porphyrin

MnTPhE-2-PyP⁵⁺

Mn(III) meso-tetrakis (N-(2’-phenylethyl)pyridinium-2-yl)porphyrin

MnTnHexOE-2-PyP⁵⁺

Mn(III) meso-tetrakis(N-(2’-n-hexoxyethyl)pyridinium-2-yl)porphyrin

MnTE-2-PyPhP⁵⁺

Mn(III) meso-tetrakis(phenyl-4’-(N-ethylpyridinium-2-yl))porphyrin

MnTnBuOE-2-PyP⁵⁺

Mn(III) meso-tetrakis(N-(n-butoxyethyl)pyridinium-2-yl)porphyrin

MnTDE-2-ImP⁵⁺

Mn(III) meso-tetrakis[N,N’-diethylimidazolium-2-yl)porphyrin, AEOL10150

Mn^IIITBAP³⁻

Mn(III) meso-tetrakis(4-carboxyphenyl)porphyrin (also abbreviated as MnTBAP)

FeTE-2-PyP⁵⁺

Fe(III) meso-tetrakis(N-ethylpyridinium-2-yl)porphyrin

FeTnOct-2-PyP⁵⁺

Fe(III) meso-tetrakis(N-n-octylpyridinium-2-yl)porphyrin

FeTrF₅Ph-β(SO₃)₂-corrole²⁻

Fe(III) meso-tris(pentafluorophenyl)-β-bis(sulfonato)corrole

EUK-8

Mn(III) salen

tempol

4-OH-2,2,6,6,-tetramethylpiperidine-1-oxyl

NXY-059

2,4-disulfophenyl-N-tert-butylnitrone (also known as OKN -007)

E_1/2

half-wave reduction potential

SOD

superoxide dismutase

NHE

normal hydrogen electrode

TON

turnover number

TOF

turnover frequency; charges and axial ligands of MnPs and FePs are omitted in text and some Figures for simplicity