Manganese superoxide dismutase, MnSOD and its mimics

Abstract

Increased understanding of the role of mitochondria under physiological and pathological conditions parallels increased exploration of synthetic and natural compounds able to mimic MnSOD - endogenous mitochondrial antioxidant defense essential for the existence of virtually all aerobic organisms from bacteria to humans. This review describes most successful mitochondrially-targeted redox-active compounds, Mn porphyrins and MitoQ(10) in detail, and briefly addresses several other compounds that are either catalysts of O(2)(-) dismutation, or its non-catalytic scavengers, and that reportedly attenuate mitochondrial dysfunction. While not a true catalyst (SOD mimic) of O(2)(-) dismutation, MitoQ(10) oxidizes O(2)(-) to O(2) with a high rate constant. In vivo it is readily reduced to quinol, MitoQH(2), which in turn reduces ONOO(-) to NO(2), producing semiquinone radical that subsequently dismutes to MitoQ(10) and MitoQH(2), completing the "catalytic" cycle. In MitoQ(10), the redox-active unit was coupled via 10-carbon atom alkyl chain to monocationic triphenylphosphonium ion in order to reach the mitochondria. Mn porphyrin-based SOD mimics, however, were designed so that their multiple cationic charge and alkyl chains determine both their remarkable SOD potency and carry them into the mitochondria. Several animal efficacy studies such as skin carcinogenesis and UVB-mediated mtDNA damage, and subcellular distribution studies of Saccharomyces cerevisiae and mouse heart provided unambiguous evidence that Mn porphyrins mimic the site and action of MnSOD, which in turn contributes to their efficacy in numerous in vitro and in vivo models of oxidative stress. Within a class of Mn porphyrins, lipophilic analogs are particularly effective for treating central nervous system injuries where mitochondria play key role. This article is part of a Special Issue entitled: Antioxidants and Antioxidant Treatment in Disease.

Figure 1

A simplified presentation of the role of MnSOD under physiological and pathological conditions. The possible scenario presented here aims at reconciliation of the dichotomous role of MnSOD as tumor suppressor (antioxidant) or oncogene (pro-oxidant). The differences in these two opposing roles are likely related to the different redox-status of the cell, primarily the ratio of the endogenous antioxidants that controls O₂^·−/H₂O₂ ratio. The common understanding is that cells which have intrinsically lower levels of MnSOD are under oxidative stress and may eventually transform into cancer cells. Further, when exposure to either single or multiple oxidative insults transforms cells prior to their becoming malignant, MnSOD levels are low and the cells are consequently under oxidative stress. The impaired redox status would in turn result in higher oxidative damage of biological targets, nucleic acids included, which would amplify mutations and enforce tumorigenesis. Once the process starts, the oxidative stress is perpetuated and the cell fights it by upregulation of MnSOD. The reportedly reduced ability of a malignant cell to remove H₂O₂ [–36] further perpetuates the oxidative stress, and MnSOD would appear as an oncogene. Tumor utilizes increased levels of peroxide to signal the activation of transcription factors and upregulation of those proteins (such as HIF-1α, VEGF, NADPH oxidases), which would maintain its oxidative stress and facilitate its progression and metastasis [1]. The complex role of MnSOD in maintaing the cellular redox status, via both its traditional role and by modulating cellular production of H₂O₂, has been elaborated by Buetner et al [16], while St Clair group has recently [38] pointed to the critical role of Sp1 and p53 in early and late stages of tumorigenesis on levels of MnSOD expression. - normal cell; - transformed cell; - cancer cell.

Figure 2

Chemical structures of mitochondrially-targeted metal based SOD mimics.

Figure 3

Structures of mitochondrially-targeted non-metal based SOD mimics.

Figure 4

The mechanism of action of MitoQ with respect to scavenging O₂^·−. These redox reactions are similar to those of ubiquinone in mitochondrial electron transport chain, and are responsible for the production of low levels of superoxide.

Figure 5

The design of SOD mimics was based on the same thermodynamics and electrostatics which play critical roles in enzyme catalysis. Potent SOD mimics are those that have E_1/2 for M^III/M^II redox couple in the vicinity of the E_1/2 of SOD enzymes, ~ +300 mV vs NHE. All SOD enzymes regardless of the metal site, whether Mn, Fe, Cu or Ni, redox cycle at same potentials [107, 111]. Modified from ref [45].

Figure 6

Structure-activity relationship between the metal centered reduction potential, E_1/2 for Mn^III/Mn^IIP redox couple for cationic, neutral and anionic Mn porphyrins, and k_cat(O₂^·−). At around +200 mV vs NHE, the k_cat is ≥2 higher for cationic than for neutral and anionic Mn porphyrins, indicating a vast contribution by electrostatics in the catalysis of O₂^·− dismutation. Modified from ref [82].

Figure 7

The timeline for the optimization of the Mn porphyrin-based cellular redox modulators. Phase I studies were directed primarily toward generating compounds with high k_cat(O₂^·−), and were successfully accomplished. Pentacationic MnTE-2-PyP⁵⁺ has been identified as our lead compound. As the research progressed, the clinical relevance of such an excessively charged drug was questioned. To address this issue, analytical tools were developed to assess the pharmacokinetics of MnPs and their subcellular distribution. The first data indicate that even the fairly hydrophilic MnTE-2-PyP⁵⁺ targets mitochondria and crosses the BBB. In Phase II the MnP structure was modified to enhance its bioavailability, primarily lipophilicity, in order to increase its transport across the BBB and mitochondrial accumulation. Lipophilicity was enhanced 10-fold: (1) by moving ethyl groups from ortho to meta positions; and (2) by lengthening the alkyl chains by each additional carbon atom. Presently, Phase III efforts are directed toward reducing toxicity of MnPs, while maintaining high redox activity and lipophilicity. Longer-alkyl chain analogues possess surfactant-based toxicity. Such toxicity was suppressed by disrupting the hydrophobicity of alkyl chains via introduction of oxygen atoms: (1) at the alkyl chain periphery, and (2) closer to the pyridyl nitrogens. In the first case, with MnTMOHex-2-PyP⁵⁺ an unfavorable drop in hydrophobicity was observed relative to MnTnHex-2-PyP⁵⁺ [109, 117], while in the second case, with MnTnBuOE-2-PyP⁵⁺, not only was a high lipopohilicity preserved, but also a slight gain in catalytic potency was achieved when compared to MnTnHex-2-PyP⁵⁺ [118].

Figure 8

(A) Accumulation of MnTnHex-2-PyP⁵⁺ in C57BL/6 mouse heart cytosol and mitochondria at 6 hours after single ip injection of 2 mg/kg. Data obtained using the LC/ESI-MS/MS method [121]. (B) Comparison of the accumulation of hydrophilic MnTE-2-PyP⁵⁺ and lipophilic MnTnHex-2-PyP⁵⁺ in C57BL/6 mouse heart mitochondria at 6 hours after single ip injection. (C) The accumulation of MnTE-2-PyP⁵⁺ and MnTnHex-2-PyP⁵⁺ in wild type AB1157 E. coli membranes/cell wall (envelope) and cytosol after E. coli was incubated for 1 hour in the presence of 5 μM Mn porphyrins in M9CA medium [91]. One of the present evolutionary hypotheses is that eukaryotic mitochondrial membranes and matrix are derived from E. coli envelope and cytosol, respectively [83]. Thus, the biodistribution within E. coli resembles the biodistribution within heart mitochondria.

Figure 9

The distribution of MnTE-2-PyP⁵⁺ and MnTnHex-2-PyP⁵⁺ in murine brain at 24 hours after single ip injection [112]. Brain levels of lipophilic MnTnHex-2-PyP⁵⁺ are 9-fold higher than of hydrophilic MnTE-2-PyP⁵⁺.

Figure 10

The reactivity of Mn porphyrins toward reactive species (A) and transcription factors (B). The reactivity of Mn(III) substituted pyridyl(or imidazolyl)porphyrins toward several oxygen and nitrogen species, such as O₂^·−, ONOO⁻, ClO⁻, ^·NO₂, CO₃^·−, would produce antioxidative effects [–46]. Under physiological conditions, H₂O₂ formed during dismutation is removed by abundant peroxide-removing systems. The ONOO⁻ reacts with MnTE-2-PyP⁵⁺ which has Mn in +3 (k_red = 3.4 × 10⁷ M⁻¹ s⁻¹ at 37°C), and with MnTE-2-PyP⁴⁺ that contains Mn in +2 oxidation states (k ≫10⁷ M⁻¹ s⁻¹ at 37°C) [135, 136]. MnTE-2-PyP⁵⁺ reacts also with ^·NO (k ~ 10⁶ M⁻¹ s⁻¹) [137]. Based on the preliminary data, MnTE-2-PyP⁵⁺ reacts rapidly with ClO⁻ with a rate constant of k ≫10⁶ M⁻¹s⁻¹ [Ferrer-Sueta et al., unpublished]. Based on the thermodynamic and kinetic data thus far published on ortho isomers, those MnPs that are potent SOD mimics would likely favor reaction with ^•NO₂ too.

Figure 11

The generation of the reactive oxygen species by Mn porphyrins coupled to ascorbate redox cycling. With abundant peroxide-removing systems, such action would generate antioxidative effects. Yet, if peroxide-removing enzymes are reduced, the H₂O₂ levels may increase and activate cellular transcription, which would in turn perpetuate oxidative stress. Under such conditions, the ability of MnPs to catalyze ascorbate oxidation would contribute to the progression of oxidative stress. The metal site of Mn porphyrin redox cycles between Mn +3 and +2 oxidation states while transferring electron to ascorbate and producing ascorbyl radical, A^·−. The reoxidation of Mn^IIP to Mn^IIIP may occur with O₂ or O₂^·−, and in either case H₂O₂ would be eventually produced. The same is valid for the self-dismutation of A^·− and O₂^·−, the reaction of A^·−, HA⁻ and A²⁻ with O₂^·−, and the reaction of A^·− with O₂ [–159]. The type of outcome, antioxidant or pro-oxidative, would depend upon the cellular redox status, the levels of cellular endogenous defenses, SODs and peroxide-removing enzymes, levels of reactive species that MnP would encounter, levels of MnP and their site of accumulation. Modified from ref [142].

Figure 12

Tumor growth/suppression by MnP is a function of oxidative stress/levels of reactive species (RS). Depending on the level of RS, different cellular events happen: at lower, physiological RS levels, signaling events predominate; while at very high levels oxidative events prevail. These scenarios are depicted by two bell shape curves for the normal and cancer cells. A tumor is frequently under oxidative stress, a “physiological” status that is visualized here by the maximum of its bell shape curve shifted towards higher RS levels. The increased level of RS is a signal for the upregulation of the genes needed to support tumor angiogenesis and progression [184, 185]. However, a tumor is vulnerable, and any further increase in oxidative burden would force tumor cells to undergo death. The strategy to treat tumors may be either to remove RS or to vastly increase their levels [27, 95]. The latter strategy has already been used in clinic; one example is the combination of ascorbate and the redox cycling agent menadione which results in increased peroxide levels. Parthenolide also has such a pro-oxidative effect, which is exerted by the selective upregulation of NADPH oxidases [–190]. HIF-1α inhibition, and subsequent suppression of angiogenesis, is a part of the first strategy where MnP scavenges RS, and thereby affects cell signaling [143, 149].

Figure 13

MnTE-2-PyP⁵⁺ mimics MnSOD in suppressing skin carcinogenesis. The reduction in incidence and multiplicity of tumors (A and B) was achieved by inhibiting AP-1 activation (C), presumably via removing signaling species, which in turn suppresses oxidative stress as observed by reduction of protein carbonyl formation (D). Tumor was induced with 7,12-dimethylbenz (a)-anthracene (DMBA). TPA is a tumor promoter 12-O-tetradecanoylphorbol-13-acetate. MnTE-2-PyP⁵⁺ (MnP) was injected in two ways, before both apoptosis and proliferation (MnP/TPA), and after apoptosis and before proliferation (TPA/MnP). In the latter case, MnP did not attenuate apoptosis, but reduced proliferation, and therefore the effect was much more pronounced than in the former case (A). With MnSOD-overexpressor mice, MnSOD attenuated both apoptotic and proliferative pathways (B), which diminished the enzyme effect upon cancer incidence relative to the effect of timely administered MnSOD mimic. Modified from refs [20, 178].

Figure 14

The mouse model of UVB radiation-induced oxidative damage. MnSOD^+/+ wild type and heterozygous MnSOD knockout mice, MnSOD^+/− were studied. The UVB radiation caused nitration and subsequent inactivation of polyγ, a polymerase enzyme responsible for the replication and repair of mtDNA. With MnSOD^+/+ mice, the oxidative damage was significantly reduce, and was fully suppressed when mice were treated with 5 mg/kg of MnTE-2-PyP⁵⁺ ip twice daily for 2 days before radiation. Quantification of polyγ co-immunoprecipitation by anti-3-nitrotyrosine antibody in mice skin lysates in terms of arbitrary units, is shown. **P < 0.01, ***P < 0.001 compared with control; ^###P < 0.001 compared between 1 hour and 24 hours after UVB treatment, ^#P < 0.05 compared between 1 hour and 24 hours after UVB treatment; ^P < 0.05 compared between MnSOD^+/+ and MnSOD^+/− mice; ^^P < 0.01 compared between Mn^IIITE-2-PyP⁵⁺ and saline pre-treatment. Modified from ref [198].

Figure 15

The mouse model of chronic morphine tolerance. Mn porphyrins were able to either substitute for or protect MnSOD from inactivation in a mouse whole brain homogenate when injected ip at 3 mg/kg (MnTE-2-PyP⁵⁺) and 0.1 mg/kg (MnTnHex-2-PyP⁵⁺) for 4 days along with morphine. No effect was seen on the expression of MnSOD protein. T describes tolerant group (which develops chronic morphine tolerance), and V is a vehicle group. The 30-fold increased efficacy of MnTnHex-2-PyP⁵⁺ resembles its 20-fold increased mitochondrial accumulation relative to MnTE-2-PyP⁵⁺. The higher ability of MnTnHex-2-PyP⁵⁺ to cross BBB further potentiates its effect in comparison with MnTE-2-PyP⁵⁺. Modified from ref [201].

Figure 16

The rat renal ischemia/reperfusion model. MnTnHex-2-PyP⁵⁺ (MnP) preserved MnSOD activity but did not affect protein expression. It was given iv into penil vein, at 0.05 mg/kg 30 minutes and 24 hours before ischemia/reperfusion. SH describes sham, and I/R ischemia/reperfusion. Modified from ref [195].