Inhibition of Mitochondrial Hydrogen Peroxide Production by Lipophilic Metalloporphyrins

Abstract

Many studies have established a role for oxidative stress and mitochondrial dysfunction as an important mechanism in the pathogenesis of neuronal disorders. Metalloporphyrins are a class of catalytic antioxidants that are capable of detoxifying a wide range of reactive oxygen species. The AEOL112 series of glyoxylate metalloporphyrins were designed with increased lipid solubility for better oral bioavailability and penetration of the blood-brain barrier. The goal of this study was to develop an in vitro assay using rat brain mitochondria to reliably detect endogenously released hydrogen peroxide (H₂O₂) and identify glyoxylate metalloporphyrins based on rank order of potency for removal of physiologically relevant H₂O₂. A polarographic method was established for the sensitive, accurate, and reproducible detection of low levels of H₂O₂. The assay identified several potent glyoxylate metalloporphyrins with H₂O₂ scavenging potencies (IC₅₀) in the nanomolar range. These results provide a simplified in vitro model system to detect physiologically generated mitochondrial H₂O₂ as a screening tool to predict the biological efficacy of potential therapeutic entities.

Oxidative stress is strongly implicated as a mediator of neuronal damage in diverse acute and chronic neuronal disorders (Lin and Beal, 2006). Metalloporphyrins are a class of catalytic antioxidants that are capable of detoxifying a wide range of reactive oxygen species (ROS), such as superoxide ( ${\mathrm{O}}_2^{\overline{.}}$), H₂O₂, peroxynitrite, and lipid peroxyl radicals (Patel and Day, 1999). Several water-soluble metalloporphyrin compounds, including manganese (III) meso-tetrakis (4-carboxyphenyl or benzoic acid) porphyrin (MnTBAP), AEOL10150, and AEOL10113, have been shown to be efficacious in animal models of central nervous system disorders, including status epilepticus, (Liang et al., 2000), stroke (Mackensen et al., 2001), and amyotrophic lateral sclerosis (Crow et al., 2005). Previous work has also demonstrated that manganic porphyrins can protect mature neuronal cultures from excitotoxic injury by scavenging intracellular ${\mathrm{O}}_2^{\overline{.}}$ (Patel et al., 1996; Li et al., 2001). These compounds contain a manganese center that catalytically dismutes both ${\mathrm{O}}_2^{\overline{.}}$ and H₂O₂ (Pasternack and Skowronek, 1979; Day et al., 1997) Previous meso-substituted porphyrin rings contained positively charged pyridyl (AEOL10113) or imidazole (AEOL10150) groups to electrostatically facilitate reaction with negatively charged ${\mathrm{O}}_2^{\overline{.}}$ (Batinic-Haberle et al., 1998; Kachadourian et al., 2004). However, the charged nature of the water-soluble pyridine- and imidazole-substituted metalloporphyrins makes them less efficient in crossing lipid membranes. To overcome these issues, a series of novel glyoxylate metalloporphyrins (AEOL112 series) with improved lipid solubility have been developed and chemically characterized (Trova et al., 2003) to improve the potential for in vivo therapeutic use in neurological disorders characterized by increased ROS levels and oxidative stress. This newly developed glyoxylate series of metalloporphyrins (Fig. 1) have been shown to dismute H₂O₂ in a catalase-like reaction to generate O₂ and inhibit lipid peroxidation in cell-free systems (Kachadourian et al., 2003, 2004; Trova et al., 2003; Liang et al., 2007).

Fig. 1.

Chemical structure of AEOL compounds. The structure of hydrophobic glyoxylate manganoporphyrins with the indicated side groups (R1–R4) is shown. a, AEOL11249, Mn²⁺ was substituted by Zn²⁺; b, AEOL11250, Mn²⁺ was substituted by Fe²⁺; c, AEOL11251, Mn²⁺ was substituted by Co²⁺.

A major initial step toward determining the in vivo efficacy of glyoxylate (AEOL112 series) metalloporphyrins is the pre-selection of lead compounds in a simple yet physiologically relevant in vitro system. Cell-free antioxidant assays can be used for this purpose and have the advantage of allowing accurate assessment of antioxidant potencies without interference from cellular components. However, the lack of endogenous factors renders these systems less predictive of in vivo efficacy. These issues may be overcome by using simplified in vitro model systems that recapitulate more physiologically relevant conditions and therefore serve as better screening tools to predict the biological efficacy of potential therapeutic entities. We have recently demonstrated the mechanism of net ROS production from purified rat brain mitochondria by the redox-cycling agent paraquat (PQ²⁺) using a polarographic assay (Castello et al., 2007). The goals of this study were to 1) develop an in vitro assay that generated physiologically relevant H₂O₂ levels and 2) identify lead metalloporphyrin compounds based on rank order of potency for scavenging endogenously generated H₂O₂.

Materials and Methods

Materials

Metalloporphyrins with >97% purity were provided by Aeolus Pharmaceuticals (Laguna Niguel, CA). With the exceptions indicated, all the other drugs used in these studies were obtained from Sigma-Aldrich (St. Louis, MO).

Isolation of Purified Rat Brain Mitochondria

Animal housing was conducted in compliance with University of Colorado at Denver and Health Sciences Center (Denver, CO) procedures. Mitochondria were isolated from adult male Sprague-Dawley rats using Percoll gradient density centrifugation as described previously (Anderson and Sims, 2000) with minor modifications (Castello et al., 2007). The purity of mitochondrial fractions was assessed using Western blotting techniques. In brief, denatured protein fractions of cytosol, mitochondria, and whole-cell homogenate were separated by electrophoresis on a 10% polyacrylamide gel (Bio-Rad, Hercules, CA) and transferred to nitrocellulose membrane. Membrane blots were incubated with primary antibodies against lactate dehydrogenase (LDH) (1:1000; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) or cytochrome c oxidase subunit IV (COX), (1:1000; Molecular Probes, Eugene, OR). LDH and COX membranes were incubated with horse-radish peroxidase-conjugated anti-goat or anti-mouse secondary antibodies, respectively. Membranes were developed using an ECL Western blotting detection reagent (GE Healthcare, Buckinghamshire, UK). Figure 2 shows that COX was undetectable in cytosolic fractions and robustly expressed in mitochondrial fractions. In contrast, LDH was undetectable in mitochondrial samples indicating highly purified isolated fractions (Fig. 2).

Fig. 2.

Purity of brain mitochondria fraction. Cytosolic and mitochondrial fractions of rat brain were isolated as described under Materials and Methods and subjected to Western blot analysis for cytosolic and mitochondrial protein markers, LDH and COX, respectively. Mitochondria samples robustly expressed COX, whereas LDH was undetectable, indicating highly purified mitochondrial fractions.

Polarographic Measurement of Net H₂O₂ Production

H₂O₂ net production by isolated brain mitochondria was measured using an Apollo 4000 Free Radical Analyzer (WPI, Sarasota, FL) equipped with a 100-μm H₂O₂ sensor. The measurements were conducted in a thermostatted open chamber at 30°C with a final reaction volume of 2 ml. Each measurement was started with the addition of reaction buffer (100 mM KCl, 75 mM mannitol, 25 mM sucrose, 10 mM Tris-Cl, and 10 mM KH₂PO₄, pH 7.4) to the chamber. Once the output signal stabilized, the following were consecutively added: respiration substrate (2.5 mM malate + 5 mM pyruvate) and mitochondrial protein (200 μg of protein). The output signal was allowed to stabilize subsequent to each addition, followed by the addition of 250 μM PQ²⁺ to the chamber, and the trace was recorded. After 2 to 3 min of recording, vehicle [1 μl of dimethyl sulfoxide (DMSO)] or tested compound (dissolved in 1 μl of DMSO) was added. The anti-oxidant activity of the metalloporphyrins was evaluated by measuring the change in the rate of H₂O₂ net production after the addition of the tested compound. The activity was expressed by the percentage of decrease in the rate of H₂O₂ net increase by mitochondria after the addition of the compound.

Fluorometric Detection of H₂O₂

H₂O₂ was measured using the horseradish peroxidase (HRP)-linked fluorometric assay (Amplex Ultra Red; Invitrogen, Carlsbad, CA). Brain mitochondria (10 μg) were added to a 96-well plate containing 100 μl of reaction buffer containing 0.1 U/ml HRP, 50 μM Amplex UltraRed and 2.5 mM malate + 5 mM pyruvate, 250 μM PQ²⁺, and 1 μl of DMSO control or tested compound dissolved in 1 μl of DMSO. Resorufin fluorescence was followed by a Gemini fluorescence microplate reader (Molecular Devices, Sunnyvale, CA). Superoxide dismutase (SOD) and catalase were added as controls at concentrations of 500 and 40 U/ml, respectively. The antioxidant activity of the compounds was expressed by the percentage of inhibition of the rate of H₂O₂ net increase by mitochondria in the presence of the compound.

SOD Screening Assay

The SOD-like activities were measured using the xanthine/xanthine oxidase system as a source of ${\mathrm{O}}_2^{\overline{.}}$ and ferricytochrome c as its indicating scavenger. ${\mathrm{O}}_2^{\overline{.}}$ was produced at the rate of 1.2 μM/min, and reduction of ferricytochrome c was followed at 550 nm. Assays were conducted in the presence of 0.1 mM EDTA in 0.05 M phosphate buffer, pH 7.8, at 25°C. Some metalloporphyrins that were analyzed interfered with the activity of xanthine oxidase, as checked by following urate production at 295 nm in the absence of cytochrome c, or they reoxidized cytochrome c at concentrations necessary to measure SOD activity.

Lipid Peroxidation Screening Assay

The concentration of thiobarbituric acid (TBA) reactive species (TBARS) in rat brain homogenates was used as an index of lipid peroxidation (Bernheim et al., 1948). Malondialdehyde (MDA) standards were obtained by adding 8.2 μl of 1,1,3,3-tetramethoxypropane in 10 ml of 0.01 M HCl and mixing for 10 min at room temperature. This stock was further diluted in water to give standards that ranged from 0.25 to 25 μM. Samples or standards (200 μl) were acidified with 200 μl of 0.2 M phosphoric acid in 1.5 ml of locking microfuge tubes. The color reaction was initiated by the addition of 25 μl of a 0.11 M thiobarbituric acid solution, and samples were placed in a 90°C heating block for 45 min. TBARS were extracted with 0.5 ml of n-butanol by vortexing samples for 3 min and chilling on ice for 1 min. The samples were then centrifuged at 12,000g for 3 min, and 150 μl of aliquots of the n-butanol phase were placed in each well of a 96-well plate and read at 535 nm in a plate-reader (Spectramax 340PC; Molecular Devices) at 25°C. Sample absorbances were converted to MDA equivalencies (micromolar) by extrapolation from the MDA standard curve. None of the antioxidants at concentrations used in these studies affected the reaction of MDA standards with TBA, and reactions without TBA were used as subtraction blanks.

Statistical Analysis

All experimental data shown were derived from at least two or three independent experiments. Nonlinear regression was performed with GraphPad Prism 4.0 (GraphPad Software Inc., San Diego, CA) adjusting a sigmoidal dose (concentration)-response equation (variable slope).

Results

A Polarographic Assay for Evaluation of H₂O₂-Scavenging Activity

To develop a biologically relevant in vitro assay for detection of H₂O₂-scavenging activities, a polarographic method of H₂O₂ detection was used. Unlike other methods of H₂O₂ detection based on endpoint readings, the use of polarographic sensors allows for the real-time measurement of steady-state H₂O₂ concentrations that takes into account the contributions of both production and consumption of H₂O₂ in the system under study. We have previously shown that the addition of PQ²⁺ to isolated rat brain mitochondria supplemented with respiration substrates results in a rapid and robust net production/increase of H₂O₂ (Castello et al., 2007). The consistent and stable increase in H₂O₂ allows for the analysis of antioxidant activity of added compounds over a short period of time. To validate this method as a tool for identifying antioxidant molecules, we analyzed the effect of externally added endogenous antioxidant enzymes, catalase, and SOD. SOD had no effect on net H₂O₂ production at a saturating concentration of 500 U/ml, whereas catalase produced a concentration-dependent inhibition of PQ²⁺-induced mitochondrial H₂O₂ (Figs. 3, A and B, and 4).

Fig. 3.

Change in the steady-state PQ²⁺-induced mitochondrial H₂O₂ and effect of catalase. A, polarographic detection of H₂O₂ net production from isolated mitochondria with the addition of 250 μM PQ²⁺. Rates of H₂O₂ net production before and after the addition of catalase (steady state 1 and 2, respectively) were compared to obtain the percentage of inhibition under each condition. B, concentration-response curve demonstrating inhibition of PQ²⁺-induced H₂O₂ by catalase, indicating an IC₅₀ value of 5.5 nM.

Fig. 4.

Inhibition of PQ²⁺-induced mitochondrial H₂O₂ by a representative glyoxylate metalloporphyrin. After the addition of PQ²⁺, H₂O₂ net production in mitochondria was monitored with addition of AEOL11207 (100 and 300 nM). AEOL11207 demonstrates a dose-dependent inhibition of mitochondrial net H₂O₂ production. Although exogenously added SOD at a saturating concentration (500 U/ml) had no effect on H₂O₂ net production, catalase is capable of scavenging H₂O₂, thus serving as a positive control. Similar results were obtained with other glyoxylate metalloporphyrins.

Because polarographic measurement of H₂O₂ is a novel, electrode-based methodology, a series of experiments were performed with the aim of identifying possible interferences. Control studies with the following treatments showed no changes over baseline H₂O₂ signal: 1) reaction buffer, 2) reaction buffer + mitochondria, 3) reaction buffer + mitochondria + PQ²⁺, 4) reaction buffer + mitochondria + malate + pyruvate, and 5) reaction buffer + PQ²⁺ + malate + pyruvate. Upon addition of PQ²⁺ to reaction buffer + mitochondria + malate + pyruvate, H₂O₂ was generated at the rate of 0.75 to 1 nmol/mg/min. Overall, these controls indicate that the polarographic measurement using an H₂O₂ electrode is not affected by the majority of the components used in the assays.

Inhibition of PQ²⁺-Induced Mitochondrial H₂O₂ by AEOL112 Compounds

Using the polarographic method described above, we were able to measure the H₂O₂-scavenging activities of metalloporphyrin compounds. Figure 4 shows an example of the results using AEOL11207. After the addition of PQ²⁺, H₂O₂ levels increased at a steady rate for several minutes at which point compounds were added to determine inhibitory effects. Catalase at a concentration of 12 nM was chosen as a positive control and was considered to represent 100% inhibition of the H₂O₂ net increase. This concentration of catalase caused no significant change in levels of H₂O₂, indicating a balance between the production and removal of H₂O₂ in the system. Increasing concentrations of catalase (125 nM) led to a decrease in H₂O₂ signal corresponding with the enzymatic removal of endogenously generated H₂O₂ in the system. Addition of SOD at saturating concentrations had no effect on the rate of H₂O₂ net production. A full range of inhibition curves was determined for different concentrations of AEOL112 compounds (Fig. 5). Figure 5 represents the percentage of inhibition of steady state of PQ²⁺-induced mitochondrial H₂O₂ as a function of the logarithm concentration of each compound. Using non-linear regression, the best fit for each concentration-response curve was obtained to determine IC₅₀ values for each compound. Compounds were divided into two groups based on their observed H₂O₂-scavenging activities: 1) those compounds exhibiting a strong concentration-response relationship (Fig. 5; Table 1) and 2) those compounds exhibiting a less optimal concentration-response relationship (Table 1). The former group included compounds showing canonical concentration-response relationship with IC₅₀ values obtained after nonlinear regression of the activity data. The latter contains compounds that do not demonstrate any inhibition of H₂O₂ net increase and/or have an IC₅₀ value higher than 3 μM. Table 1 shows the IC₅₀ values for the compounds that exhibited a strong concentration-response relationship. A notable exception is AEOL10150, which interfered with the polarographic detection. To ascertain whether the observed inhibition of H₂O₂ net production by metalloporphyrins was not due to a change in the sensitivity of the electrode for the H₂O₂ after addition of the compound, the effect of exogenously added H₂O₂ was determined. No change in signal was observed after the addition of the AEOL112 series compounds listed in Table 1 to the reaction buffer + mitochondria + malate + pyruvate + 2 μM H₂O₂. The most potent metalloporphyrins identified with IC₅₀ < 1 μM are as follows: AEOL11209 (IC₅₀ = 17 nM) > AEOL11216 (IC₅₀ = 93 nM) ≥ AEOL11207 (IC₅₀ = 104 nM) > AEOL11215 (IC₅₀ = 206 nM) > AEOL11223 (IC₅₀ = 408) > AEOL11210 (IC₅₀ = 725 nM) > AEOL11202 (IC₅₀ = 1642 nM).

Fig. 5.

Concentration-response curves of the inhibition of PQ²⁺-induced mitochondrial H₂O₂ by glyoxylate metalloporphyrins. The ability of glyoxylate metalloporphyrins to inhibit PQ²⁺-induced mitochondrial H₂O₂ net production was evaluated. Compounds that demonstrated a strong concentration-response relationship are shown. Compounds that exhibited a less optimal concentration-response relationship are indicated in Table 1. Each point represents the mean values from independent duplicate experiments.

TABLE 1.

Inhibition of PQ²⁺-induced H₂O₂ and catalase activity of selected glyoxylate metalloporphyrins

Compound	IC50	% Catalase Activity
	nM
SOD	N.A.	N.A.
Catalase	5.5	100
AEOL10150	(3000)*	(0.2)*
AEOL11209a	17	32.4
AEOL11216	93	5.9
AEOL11207	104 (30)*	5.3 (18.3)*
AEOL11215	206	2.2
AEOL11223	408	1.3
AEOL11210	725	0.8
AEOL11202b	1642	0.3
AEOL11203	>3000	>0.2
AEOL11204	>3000	>0.2
AEOL11206	>3000	>0.2
AEOL11219	>3000	>0.2
AEOL11227	>3000	>0.2
AEOL11238	>3000	>0.2
AEOL11239	>3000	>0.2
AEOL11243	>3000	>0.2
AEOL11244	>3000	>0.2
AEOL11249	>3000	>0.2
AEOL11250	>3000	>0.2
AEOL11251	>3000	>0.2

N.A., not applicable.

^*Activity determined using Amplex Red assay.

^aAEOL112 compounds in italics exhibit a strong concentration-response relationship.

^bAEOL112 compounds in regular font exhibit a less optimal concentration-response relationship.

Comparison of the Inhibition of PQ²⁺-Induced Mitochondrial H₂O₂ by AEOL10150 and AEOL112 Series

To validate the polarographic method and overcome its interference with AEOL10150, a HRP-linked fluorometric method (Amplex Red assay) was used. The values obtained were used to construct the concentration-response curve shown in Fig. 6. To compare the IC₅₀ values obtained using the fluorometric assay with the polarographic method, AEOL11207 was used as a positive control. AEOL11207 showed a concentration-response relationship comparable to that obtained using the polarographic method. The IC₅₀ values of AEOL11207 using the fluorometric and the polarographic assays were 30 and 104 nM, respectively. AEOL10150 showed a concentration-response relationship with an IC₅₀ value of 3 μM, using the fluorometric assay.

Fig. 6.

Concentration-response curves of inhibition of PQ²⁺-induced mitochondrial H₂O₂ by AEOL11207 and AEOL10150. H₂O₂ was fluorometrically detected using the HRP-linked Amplex Red assay. Rates of H₂O₂ net production were measured in the presence of different final concentrations of AEOL11207 and AEOL10150. The percentage of inhibition of H₂O₂ net production was determined for each concentration and used to calculate IC₅₀ values for each compound. Each point is the mean of values from duplicate experiments.

Structure Activity Relationships of the AEOL112 Series with Their Ability to Inhibit Net Production of H₂O₂

The most potent inhibitors of H₂O₂ net production in the AEOL112 series were meso-substituted with electron-withdrawing groups such as aldehydes, as seen in the bis-substituted AEOL11209 compound and trifluoromethyl groups in the bis-substituted AEOL11207 and AEOL11216 compounds. It was interesting to note that n-alkyl-substituted ester groups also correlated with the ability of the compounds to inhibit the net production of H₂O₂ as seen with AEOL11215 that has a meso-tetrakis propyl ester substitution, but activity drops off substantially as one lengthens the alkyl chain as seen in the meso-tetrakis hexyl ester-substituted compound AEOL11238.

To determine whether inhibition of PQ²⁺-induced H₂O₂ by metalloporphyrins showing an optimal concentration-response relationship was due to the manganese moiety, the effects of metal-substituted analogs of AEOL11215 (AEOL11249, Zn²⁺ analog; AEOL11250, Fe²⁺ analog; AEOL11251, Co²⁺ analog) were evaluated. Figure 7 shows that none of these analogs was able to inhibit the net production of H₂O₂ in a broad range of concentrations. These studies suggest that manganese is the optimal metal to support H₂O₂-scavenging activity.

Fig. 7.

Concentration-response curves by metal-substituted analogs of AEOL11215. The ability of metal-substituted analogs of AEOL11215 to inhibit PQ²⁺-induced mitochondrial H₂O₂ net production was evaluated. These data show that none of the metal-substituted analogs was able to inhibit the net production of H₂O₂ over a broad range of concentrations, suggesting that manganese is critical in H₂O₂-scavenging activity. AEOL11215, Mn⁺; AEOL11249, Zn²⁺ analog; AEOL11250, Fe²⁺ analog; AEOL11251, Co²⁺ analog.

Antioxidant Properties of AEOL112 Compounds

Table 2 shows the antioxidant activity of hydrophobic metalloporphyrins evaluated as SOD-like activity and inhibition of lipid peroxidation (TBARS). Overall, the data indicate that AEOL112 compounds have very low SOD-like activity and very high lipid peroxidation inhibition activity. These properties, together with the high capacity to dismute H₂O₂, suggest that these compounds have an alternative antioxidant mechanism compared with that observed in known SOD mimetics.

TABLE 2.

Structures and activities of selected glyoxylate metalloporphyrins Each column contains the structure of the side chains (R1–R4) indicated in Fig 1.

AEOL	Product No.a	SOD Activity	TBARS (IC50)
		U/mg	μM
10150	N.L.	13,658	0.1
11202	30	N.A.	0.1
11203	26	N.A.	0.725
11204	27	N.A.	0.259
11206	28	5.62	3.35
11207	37	N.A.	8.2
11209	N.L.	183	0.2
11210	N.L.	81	9
11215	32	129.5	6.65
11216	39	98	4
11219	33	Interferes	8.2
11221	45	Red cytochrome c	8.4
11223	47	Interferes	4.4
11225	43	Interferes	4.7
11227	36	4.5	30
11231	54	67.4	7.1
11238	35	Red cytochrome c	306
11239	52	6.5	0.91
11243	N.L.	Interferes	N.A.
11244	N.L.	84	36
11249	N.L.	Inactive	Inactive
11250	N.L.	Inactive	Inactive
11251	N.L.	inactive	Inactive

N.L., not listed; N.A., not applicable.

^aFormer product number taken from (Gauuan et al., 2002; Trova et al., 2003).

Discussion

In this study, we developed a novel in vitro screening assay using polarographic detection of H₂O₂ endogenously generated in isolated rat brain mitochondria treated with PQ²⁺. Using this assay, we identified several potent antioxidant compounds belonging to a novel class of lipophilic glyoxylate metalloporphyrins.

A polarographic method was established for the sensitive, accurate, and reproducible detection of H₂O₂ scavenging by AEOL compounds in a physiologically relevant in vitro model involving rat brain mitochondrial H₂O₂. The method is rapid and sensitive, with low level of interferences, and has the potential for high-throughput analysis. The assay uses respiring mitochondria and PQ²⁺ to generate endogenous H₂O₂. The system is specific for H₂O₂ as shown by the high sensitivity to the addition of catalase but not SOD (Fig. 4). The ability of catalase to inhibit the H₂O₂ signal is due to the ability of intramitochondrially generated H₂O₂ to cross mitochondrial membranes, which then can be readily dismuted by catalase. The lack of effect with SOD is probably due to the inability of this large protein to penetrate mitochondrial membranes to dismute intramitochondrial ${\mathrm{O}}_2^{\overline{.}}$, which is short-lived and not very permeable to biological membranes. Once the steady-state net increase of H₂O₂ was induced by PQ²⁺ in mitochondria, the addition of either several metalloporphyrins or catalase (but not SOD) changed the steady-state net production of H₂O₂. The velocity of H₂O₂ net production in the lower steady state was used as an indicator of compound potency. This approach has the advantage of using the initial steady state as a control, eliminating the random differences between different measurements and preparations in which the initial steady state may vary.

It is worth mentioning that in vitro H₂O₂ concentrations previously used in the measurement of catalase activity of AEOL compounds was 1 mM (Day et al., 1997). This concentration is orders of magnitude higher than H₂O₂ steady-state concentrations in physiological systems that are in the nanomolar range (Chance et al., 1979). Because catalase activity is assumed to follow pseudo-first-order kinetics, the use of such high amounts of H₂O₂ can result in an overestimation of the pseudo-first-order rate constant and, therefore, of catalase activity. Moreover, it has been reported that high amounts of H₂O₂ can inactivate the metalloporphyrins (Day et al., 1997). This new assay presents several advantages over the established ones. First, it uses concentrations of H₂O₂ (~1–100 nM) that may be achieved physiologically. Steady-state concentrations of H₂O₂ are estimated to 10⁻⁹ to 10⁻⁷ M (Chance et al., 1979; Gardner et al., 2006). Second, it uses H₂O₂ produced by brain mitochondria, which are an important cellular source of ROS contributing to neurodegeneration and aging. Third, it is based in an in vitro system using the redox-cycling agent PQ²⁺, an environmental toxin implicated in the etiology of Parkinson’s disease (Di Monte, 2003).

The results presented in Table 1 reveal the following order of potencies of the metalloporphyrins tested in this study that showed an IC₅₀ < 1 μM: AEOL11209 (IC₅₀ = 17 nM) > AEOL11216 (IC₅₀ = 93 nM) > AEOL11207 (IC₅₀ = 104 nM) > AEOL11215 (IC₅₀ = 206 nM) > AEOL11223 (IC₅₀ = 408) > AEOL11210 (IC₅₀ = 725 nM) > AEOL11202 (IC₅₀ = 1642 nM). The potencies of the compounds obtained in our in vitro assay have been validated in the in vivo setting by the demonstration that orally administered AEOL11207 achieving brain concentrations of ~200 nM inhibited oxidative stress indices and neuronal damage in a mouse model of mitochondrial oxidative stress (Liang et al., 2007).

The ability of manganese-substituted, but not zinc-, cobalt-, and iron-substituted, metalloporphyrins to inhibit the net production of H₂O₂ illustrates the importance of manganese as the optimal metal in the H₂O₂-scavenging effects of the compounds. Although the glyoxylate metalloporphyrins are lipid-soluble, the control studies described under Results suggest that most AEOL compounds do not inhibit the net production of H₂O₂ by interfering with the redox-cycling mechanism of PQ²⁺ in the mitochondria. Their ability to remove H₂O₂ is probably not caused by scavenging of intramitochondrial ${\mathrm{O}}_2^{\overline{.}}$. This observation is based on their low SOD activity in cell-free assays (Trova et al., 2003), which renders the compounds less suitable as SOD mimetics.

Comparison of the data obtained in this study (Table 1) with previously published values for the AEOL112 series (Gauuan et al., 2002; Trova et al., 2003) demonstrates that compounds exhibiting a strong concentration-response relationship (Fig. 5; Table 1) have an average catalase activity ~145% higher than the average activity of the compounds exhibiting a less optimal concentration-response relationship (Table 1). On the other hand, the compounds demonstrating a strong concentration-response relationship (Fig. 5; Table 1) display average TBARS levels that are eight times lower than the average of the compounds exhibiting a less optimal concentration-response relationship, indicating a greater ability to remove lipid peroxides (Table 2). Together, these results conclude that the grouping of compounds according to their antioxidant properties was reflected not only by the current method but also by other previous screening methods.

One interesting finding that has emerged from several in vitro models of neuronal injury is a discrepancy between anti-oxidant potency and neuroprotective efficacy of metalloporphyrins. For example, although the water-soluble metalloporphyrin MnTE-2-PyP (AEOL10113) has at least 20 times more SOD activity in cell-free assays compared with MnTBAP, it was only two to three times more potent in its efficacy as a neuroprotective agent in Sod2^−/− cultures (Patel, 2003). This paradoxical difference between antioxidant activities and neuroprotective efficacy has also been observed in other in vitro models involving glutamate excitotoxicity and oxygen-glucose deprivation injury (Li et al., 2001). These observations suggest that high SOD activity in a cell-free assay per se may not be sufficient to predict neuroprotection in vivo (cells or animals) and provides the rationale for the development of metalloporphyrins with broad antioxidant properties other than antioxidant activities derived from cell-free assays. The glyoxylate (AEOL112) series of metalloporphyrins have modest SOD activity but show high potencies as inhibitors of lipid peroxidation and cell and tissue injury (Choudhary et al., 2001; Kachadourian et al., 2003, 2004; Trova et al., 2003; Liang et al., 2007).

The ability of metalloporphyrins to scavenge mitochondrially generated H₂O₂ is highly significant based on previous demonstrations that mitochondrial overexpression of catalase provides protection against menadione toxicity, a chemical agent that preferentially generates ${\mathrm{O}}_2^{\overline{.}}$ intramitochondrially (Gurgul et al., 2004). Moreover, median and maximal life spans were increased the most in animals overexpressing mitochondrial catalase due to a reduction in mitochondrial generated oxidative stress (Schriner et al., 2005). Although the overexpression of mitochondrial catalase is difficult to achieve, the particular scavenging of mitochondrial H₂O₂ by AEOL compounds opens a new paradigm in the therapeutic treatment of neuronal diseases in which mitochondrial oxidative stress is a major contributor.

ABBREVIATIONS

ROS

reactive oxygen species

MnTBAP

manganese (III) meso-tetrakis (4-carboxyphenyl or benzoic acid) porphyrin

AEOL10150

manganese (III) meso-tetrakis (N,N′-diethylimidazolium-2-yl) porphyrin

AEOL10113

manganese (III) meso-tetrakis (N-ethyl pyridinium-2-yl) porphyrin

PQ²⁺

paraquat

LDH

lactate dehydrogenase

COX

cytochrome c oxidase

DMSO

dimethyl sulfoxide

HRP

horseradish peroxidase

SOD

superoxide dismutase

TBA

thiobarbituric acid

TBARS

thiobarbituric acid reactive species

MDA

malondialdehyde

MnTE-2-PyP

manganese tetrakis-(N-ethyl-2-pyridyl) porphyrin