Theoretical insights into the antiradical activity and copper-catalysed oxidative damage of mexidol in the physiological environment

Abstract

Mexidol (MD, 2-ethyl-6-methyl-3-hydroxypyridine) is a registered therapeutic agent for the treatment of anxiety disorders. The chemical structure suggests that MD may also act as an antioxidant. In this study, the hydroperoxyl radical scavenging activity of MD was studied to establish baseline antioxidant activity, followed by an investigation of the effect of MD on the copper-catalysed oxidative damage in biological systems, using computational methods. It was found that MD exhibits moderate radical scavenging activity against HOO^• in water and pentyl ethanoate solvents following the single electron transfer and formal hydrogen transfer mechanisms, respectively. MD can chelate Cu(II), forming complexes that are much harder to reduce than free Cu(II): MD chelation completely quenches the Cu(II) reduction by ascorbic acid and suppresses the rate of reduction reaction by ${\mathrm{O}}_2^{\cdot -}$ that are the main reductants of Cu(II) in biological environments. Therefore, MD exerts its anti-HO^• activity primarily as an OIL-1 inhibitor.

1. Introduction

Mexidol (MD, emoxypine, 2-ethyl-6-methyl-3-hydroxypyridine) is a drug used primarily for the treatment of anxiety. It is known to have antiischemic, antihypoxic, neuroprotective, antistress, nootropic and geroprotective properties [1,2]. MD is also used as an antioxidant for reducing tissue damage by reactive oxygen species [1,3]. For this reason, the radical scavenging activity of MD was assessed in experimental studies [3,4]. The rate constant of MD reaction with peroxyl radical in methyl oleate was measured as k = 2.8 × 10⁴ M⁻¹ s⁻¹ [4], while that for 1,4-dioxane solvent was k = 2.1 ± 0.3 × 10⁴ M⁻¹ s⁻¹ [3]. However, organic solutions such as methyl oleate and 1,4-dioxane are not suitable model environments to assess in vivo activity; in the physiologically relevant aqueous environment MD may exhibit very different behaviour due to dissociation of the OH moiety. Thus there is an impetus to study the antiradical activity of MD in physiological media.

In evaluating the antioxidant activity of MD, one has to also consider that pyridines are very good coordinating ligands and therefore it is highly likely that MD forms chelates with trace metals [5,6]. Previous studies showed that the presence of transition metals such as Cu(I) can produce hydroxyl radicals via the Fenton-like reaction (1.1) [7–11]:

$${\mathrm{Cu}}^{+}+\hspace{0.17em}{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{Cu}}^{2+}+{\mathrm{HO}}^{\cdot }.$$ 1.1

However, Cu(I) is not the most stable or abundant form of copper in physiological environments. The concentration of Cu(I) is defined by the amount of Cu(II) ions undergoing reduction by hyperoxide radical (${\mathrm{O}}_2^{\cdot -}$) or deprotonated ascobic acid (AA⁻) with the rate constants k = 4.46 × 10⁹ M⁻¹ s⁻¹ and 1.33 × 10⁸ M⁻¹ s⁻¹, respectively [10]:

$${\mathrm{Cu}}^{2+}+{\mathrm{O}}_2^{\cdot -}\to {\mathrm{Cu}}^{+}+{}^3{\mathrm{O}}_2$$ 1.2

and

$${\mathrm{Cu}}^{2+}+{\mathrm{AA}}^{-}\to {\mathrm{Cu}}^{+}+{\mathrm{AA}}^{\cdot }.$$ 1.3

Chelation of Cu(II) has the potential to inhibit or suppress reactions (1.2) and (1.3) and therefore indirectly inhibit reaction (1.1). Thus, the capacity of MD to inhibit Cu(II) reduction by chelation is also of interest (figure 1).

Figure 1.

Molecular structure and atomic numbering of mexidol (MD).

Previous studies showed that computational method offers the most convenient path for studying structure–activity relationships in radical reactions to guide the design of novel antioxidants with enhanced activity [12–19]. In several prior studies, the radical scavenging activity of organic compounds in physiological environments was successfully evaluated by quantum chemistry calculations [20–23]. Thus, in this study, the HOO^• radical scavenging activity of MD was assessed in lipid and polar media using the quantum mechanics-based test for overall free radical scavenging activity (QM-ORSA) protocol [13,20]. Cu(II) chelation ability was also assessed and the ability of MD to act as an OIL-1 inhibitor of the copper-catalysed oxidative damage in biological systems was investigated.

2. Computational details

All density functional theory (DFT) calculations were carried out with Gaussian 09 suite of programs [24]. M06-2X functional [25] and 6-311++G(d,p) basis set were used for all calculations. The M06-2X functional is one of the most reliable methods to study thermodynamics and kinetics of radical reactions [20,25–30].

The stability of Cu(II) chelates was compared by calculating the Gibbs free energy of formation for all possible chelates that were first constructed with MD based on [Cu(H₂O)₄]²⁺ geometry, optimized by molecular mechanics calculations using the Spartan software [31], then energy-minimized with DFT as per above.

The kinetic calculations were performed following the QM-ORSA protocol [13,20] and following the literature [29,30,32–37]. The details of the method are shown in electronic supplementary material, table S1. Atom-in-molecule (AIM) analysis [38] was performed by using the AIM2000 software [39].

3. Results and discussion

3.1. The HOO^• radical scavenging activity of MD

3.1.1. Gas phase evaluation

To reduce computing time, the radical scavenging activity of MD was first evaluated in the gas phase following the liturature [16,40], according to the three main reaction pathways: formal hydrogen transfer (FHT), single electron transfer followed by proton transfer (SETPT) and sequential proton loss electron transfer (SPLET) [41,42]. Radical adduct formation (RAF) is another common pathway that however requires a localized C=C bond; the aromatic ring of MD cannot support this mechanism [21,43,44]. The probabilities of the three feasible antioxidant mechanisms (FHT, SETPT and SPLET) were first evaluated by computing the main thermodynamic parameters associated with these mechanisms: bond dissociation enthalpy (BDE), ionization energy (IE) and proton affinity (PA), respectively. The calculated BDE, IE and PA values of MD are shown in electronic supplementary material, table S3.

The lowest BDE value was calculated for O3−H at 84.4 kcal mol⁻¹. This value is higher than that of natural antioxidants such as resveratrol (83.9 kcal mol⁻¹) [45], piceatannol (73.1 [21] or 75.1 [46] kcal mol⁻¹), Trolox (73.0 kcal mol⁻¹) [47] and ascorbic acid (73.9 kcal mol⁻¹) [47]. The BDE values of the C−H bonds are even higher by about 2.0–16.0 kcal mol⁻¹. At the same time, the IE and PA values are about 2.3 and 4.1 times higher than the lowest BDE value. Thus, based on the computed data, the antioxidant activity of MD in apolar and low-dielectric environments is projected to proceed via the FHT pathway. Calculations of ΔG° (Gibbs free energy change) values of the defining step of the MD + HOO^• reaction (electronic supplementary material, table S3) also confirmed that the FHT mechanism is the main path of the HOO^• radical scavenging activity of MD in the gas phase as well as in non-polar media.

Based on the thermodynamic results, the kinetics of the antiradical activity of MD was evaluated for the thermodynamically favourable positions and mechanisms according to the QM-ORSA protocol [20], and the data are shown in electronic supplementary material, table S4, and figure 2. The results suggest that the O3−H bond (ΔG^≠ = 13.0 kcal mol⁻¹; k_Eck = 2.95 × 10⁵ M⁻¹ s⁻¹; Γ = 100.0%) is the only feasible site for H-abstraction during the MD + HOO^• reaction. The C−H bonds do not make any contribution (Γ = 0%) in the overall rate constant of the antiradical activity of MD. That is consistent with the lowest BDE(O−H) values as calculated in the thermodynamic section. Thus the results suggest that the HOO^• radical scavenging activity of MD is dominated by the FHT mechanism at the O3−H bond; therefore, this reaction is further analysed in lipid medium.

Figure 2.

Optimized structures of the transition states (TS) along the FHT pathway of the HOO^• radical scavenging activity of MD: (G) in the gas phase, (P) in pentyl ethanoate and (W) in water.

3.1.2. The HOO^• radical scavenging activity of MD in physiological environments

In aqueous environment, the antiradical activity of acidic species is typically dominated by the ionic form [26,30,48]. Therefore, the protonation state of MD was first evaluated at physiological pH to find the most likely radical scavenging reactions. The MD structure allows protonation at the N1−H and O3−H bonds (figure 3); thus the pK_a values of MD were calculated based on the literature [48,49] and are shown in figure 3.

Figure 3.

The deprotonation of MD at pH = 7.4.

The calculated pK_a values for the amine were pK_a1 = 7.17 (for N−H bond of cation form) and pK_a2 = 9.79 (for O3−H bond). Therefore, in pH = 7.4 aqueous solution, MD exists in three states: the cation (H₂A⁺, 37.0%), neutral (HA, 62.7%) and anion (A⁻, 0.3%) states. Therefore, all three states were used in the kinetic evaluation of HOO^• radical removal of MD in water at physiological pH = 7.4.

The preferred radical scavenging pathways of the neutral and anionic states have been established; the cationic state requires initial evaluation. The computing of ΔG° of the H₂A⁺ + HOO^• reaction for the possible pathways (electronic supplementary material, table S5) showed that the H₂A⁺ + HOO^• reaction is only clearly spontaneous for FHT at the C7−H bond (ΔG° = −4.4 kcal mol⁻¹). Thus this reaction was used to calculate the rate constant of the HOO^• radical scavenging of H₂A⁺ in the aqueous solution. The overall rate constant of HOO^• + MD reaction was computed following equations (3.1) and (3.2); the results are presented in table 1 and figure 2.

Table 1.

Calculated ΔG^≠ (kcal mol⁻¹), κ, rate constants (k_app, k_f and k_overall (M⁻¹ s⁻¹)), and Γ (%) in the MD + HOO^• reaction in the studied media.

mechanisms		pentyl ethanoate				water
		ΔG≠	κ	kapp	Γ	ΔG≠	κ	kapp	f	kf	Γ
SET	HA					36.8	18.1a	4.80 × 10−17	0.627	3.01 × 10−17	0.0
	A−					5.1	16.0a	7.70 × 106	0.003	2.31 × 104	86.2
FHT	H2A+					24.4	483.1	3.90 × 10−3	0.37	1.44 × 10−3	0.0
	HA	15.9	328.1	4.80 × 103	100.0	16.6	1466.0	5.90 × 103	0.627	3.70 × 103	13.8
koverall				4.80 × 103						2.68 × 104

^aThe nuclear reorganization energy (λ, kcal mol⁻¹); f = %A⁻/100; k_f = f · k_app; Γ = k_f ·100/k_overall.

In the lipid medium

$$k_{\mathrm{overall}}=k_{\mathrm{app}}\hspace{0.17em}(\mathrm{FHT}(\mathrm{O}3-\mathrm{H})-\mathrm{neutral}).$$ 3.1

In the aqueous solution

$$\begin{array}{r}{k}_{\mathrm{overall}}=k_{\mathrm{f}}\hspace{0.17em}(\mathrm{SET}-\mathrm{HA})+k_{\mathrm{f}}\hspace{0.17em}(\mathrm{SET}-{\mathrm{A}}^{-})+k_{\mathrm{f}}\hspace{0.17em}(\mathrm{FHT}-{\mathrm{H}}_2{\mathrm{A}}^{+}(\mathrm{C}7-\mathrm{H}))+k_{\mathrm{f}}\hspace{0.17em}(\mathrm{FHT}-\mathrm{HA}(\mathrm{O}3-\mathrm{H})).\hspace{0.17em}\end{array}$$ 3.2

As per table 1, the HOO^• + MD reaction in pentyl ethanoate is moderate with k_overall = 4.40 × 10³ M⁻¹ s⁻¹ by the FHT mechanism at the O3−H bond (Γ = 100%). By contrast, MD exhibits good antiradical activity in water with k_overall = 2.68 × 10⁴ M⁻¹ s⁻¹. This reaction was defined (Γ ∼ 100%) by the SET pathway of the dianion A⁻. The rate constant for the HA + HOO^• reaction following the FHT mechanism at the O3−H bond is k_f = 3.70 × 10³ M⁻¹ s⁻¹ (Γ = 13.8%), whereas that for the C7−H bond (H₂A⁺) is only k_f = 1.44 × 10⁻³ M⁻¹ s⁻¹ and this reaction makes a negligible contribution (approx. 0%) to the total antiradical activity of MD. Compared with the reference antioxidant Trolox (k = 1.30 × 10⁵ and 1.00 × 10⁵ M⁻¹ s⁻¹ in polar and nonpolar media, respectively) [40], the HOO^• radical scavenging activity of MD is lower in all of the studied environments. However, according to the empirical test for radical scavenging (activity should exceed 10³) [20,50] MD can be considered an antioxidant, albeit a weak one.

3.2. OIL-1 inhibition of copper-catalysed oxidative damage in biological systems

It was first calculated whether MD in any form can reduce free Cu(II), i.e. whether the chelation itself is reductive or not. Results are shown in table 2. Only the anionic form yields a negative Gibbs free energy change and thus it may reduce Cu(II). However, in the catalytic cycle, it competes with O^•−₂ and AA⁻ that have an order of magnitude higher rate constants (4.46 × 10⁹ M⁻¹ s⁻¹ and 1.33 × 10⁸ M⁻¹ s⁻¹, respectively) [10]; hence MD does not contribute substantially to the reduction of free Cu(II).

Table 2.

The calculated ΔG°, ΔG^≠, λ (kcal mol⁻¹), and rate constants (k_app, k_f and k_overall (M⁻¹ s⁻¹)) at 298.15 K in the MD + Cu(II) reactions in water at pH = 7.40.

reactions	ΔG°	ΔG#	λ	kapp	f	kf	koverall
H2A+ + Cu2+	39.6	94.7	5.3	2.20 × 10−57	0.37	8.14 × 10−58	2.19 × 107
HA + Cu2+	20.5	28.7	6.2	6.00 × 10−9	0.627	3.76 × 10−9
A− + Cu2+	−10.9	2.8	4.1	7.30 × 109	0.003	2.19 × 107

To find the most stable complexes of the Cu(II) + MD reaction in water at pH = 7.4, the Cu(II) chelating ability of MD was evaluated considering the possible chelation sites of MD involving N and O atoms (electronic supplementary material, table S2). It was found that all of MD states (H₂A ⁺ , HA and A⁻) present in water could react with Cu(II) following 23 reactions and forming a range of complexes (electronic supplementary material, figure S1). The lowest Gibbs free energy of formation (ΔG°, kcal mol⁻¹) was predicted for the complex [Cu(HA)₂(H₂O)₂]²⁺ (reaction 14, N site, electronic supplementary material, table S2; figure 4) at −82.5 kcal mol⁻¹ that is more than two times lower than the second lowest value (reaction 8, N site, electronic supplementary material, table S2). The equilibrium constant for reaction 14 is K = 3.01 × 10⁶⁰ M⁻¹ and this chelate will make up 100% of the possible complexes, based on Maxwell–Boltzmann calculations. The mono-, tri- and tetra-coordinate Cu–MD complexes were also investigated; however, these reactions were not favoured for the MD + Cu(II) in water as indicated by the higher values of ΔG° of formation, compared with the [Cu(HA)₂(H₂O)₂]²⁺. Thus, the antioxidant activity calculations were performed for this chelate.

Figure 4.

Optimized structures and AIM topological shapes of [Cu(HA)₂(H₂O)₂]²⁺ and [Cu(HA)₂(H₂O)₂]⁺ in aqueous solution, at 298.15 K. Small red spheres indicate bond critical points.

For a deeper understanding of the structure and stability of the complexes, AIM analysis was applied to [Cu(HA)₂(H₂O)₂]²⁺ and [Cu(HA)₂(H₂O)₂]⁺ in water. The topological shape and selected parameters at the bond critical points (BCPs) at metal–molecule contacts of the complexes are presented in figure 4 and table 3.

Table 3.

Selected parameters at the BCPs at intermolecular contacts for [Cu(HA)₂(H₂O)₂]²⁺ and [Cu(HA)₂(H₂O)₂]⁺ in aqueous solution, at 298.15 K.

contacts	ρ(r) (au)	∇2ρ(r) (au)	G(r)a) (au)	V(r)b) (au)	G(r)/|V(r)|	H(r)c) (au)	Ed)HB (kcal mol−1)
[Cu(HA)2(H2O)2]2+
N2 ··· Cu	0.0695	0.3124	0.0872	−0.0962	0.9057	−0.0091	−30.2
N30 ··· Cu	0.0745	0.3426	0.0959	−0.1061	0.9035	−0.0102	−33.3
O23 ··· Cu	0.0597	0.3479	0.0878	−0.0886	0.9908	−0.0008	−27.8
O25 ··· Cu	0.0603	0.3514	0.0887	−0.0895	0.9907	−0.0008	−28.1
[Cu(HA)2(H2O)2]+
N2 ··· Cu	0.0753	0.3492	0.0999	−0.1125	0.8881	−0.0126	−35.3
N30 ··· Cu	0.0753	0.3494	0.0999	−0.1124	0.8886	−0.0125	−35.3
O23 ··· Cu	0.0112	0.0441	0.0101	−0.0092	1.0991	0.0009	−2.9

It can be observed that [Cu(HA)₂(H₂O)₂]²⁺ is mainly stabilized by Cu(II) ··· O and Cu(II) ··· N contacts. That was affirmed by the existence of BCPs (red spheres) in the Cu(II)–O23, Cu(II)–O25, Cu(II)–N2 and Cu(II)–N30 electron densities. It can be inferred from table 3 that all of the Cu(II) ··· X (X = O, N) electron densities are partly covalent in nature as indicated by ∇²ρ(r) > 0, G(r)/|V(r)| ≤ 1 and H(r) ≤ 0 [51,52]. These contacts play a decisive role in the stability of the complex [Cu(HA)₂(H₂O)₂]²⁺ with the significant negative values of E_HD (−27.8 to −33.3 kcal mol⁻¹).

On the other hand, the stability of Cu(I) complex ([Cu(HA)₂(H₂O)₂]⁺) is defined by the Cu(I) ··· N2(30) electron fluxes (∇²ρ(r) = 0.3492(0.3494), G(r)/|V(r)| = 0.8881(0.8886), H(r) = −0.0126(−0.0125) and E_HD = −35.3 kcal mol⁻¹; table 3). The contact of Cu(I) ··· O23 (H₂O) is a weak interaction with ∇²ρ(r) = 0.0441, G(r)/|V(r)| = 1.0991, H(r) = 0.0009 and E_HD = −2.9 kcalmol⁻¹. Thus two H₂O molecules in the [Cu(HA)₂(H₂O)₂]⁺ are only solvating the system. That is consistent with previous studies of the coordination numbers and geometries of Cu(II) and Cu(I) complexes [9,53].

To evaluate the capacity of MD to reduce the copper-induced oxidative stress following the OIL-1 process in water at pH = 7.4, reduction reactions of the most stable complex ([Cu(HA)₂(H₂O)₂]²⁺) and free Cu(II) were evaluated against typical copper-reducing species (${\mathrm{O}}_2^{\cdot -}$ and ascorbic acid anion (AA⁻)) and the results are shown in table 4. It was found that the rate constant of the reaction of ${\mathrm{O}}_2^{\cdot -}$ with [Cu(HA)₂(H₂O)₂]²⁺ is 2.10 × 10⁹ M⁻¹ s⁻¹ that is about 2.1 times lower than that of the reduction of free Cu(II) in water (k(Cu²⁺ + ${\mathrm{O}}_2^{\cdot -}$) = 4.46 × 10⁹ M⁻¹ s⁻¹) [10]. However, complexation suppressed the rate constant of AA⁻-driven reduction of Cu(II) by about 10⁸ times compared to free Cu(II). Based on the calculated data, MD is a good OIL-1 inhibitor of Cu(II)-catalysed oxidative stress.

Table 4.

Calculated ΔG°, λ, ΔG^≠ (kcal mol⁻¹), and rate constants (k_app, k_f and k_overall (M⁻¹ s⁻¹)) for the reduction of Cu(II) and the most likely Cu(II) chelates by ${\mathrm{O}}_2^{\bullet -}$ and AA⁻ in aqueous solution, at 298.15 K. k(Cu²⁺ + ${\mathrm{O}}_2^{\bullet -}$) = 4.46 × 10⁹ M⁻¹ s⁻¹; k(Cu²⁺+ AA⁻) = 1.33 × 10⁸ M⁻¹ s⁻¹ [9,10].

reactions	ΔG°	ΔG≠	λ	kapp	f	kf	koverall
[Cu(HA)2(H2O)2]2+ + ${\mathrm{O}}_2^{\bullet -}$	−10.5	4.6	36.2	2.10 × 109	0.9975a	2.10 × 109	2.10 × 109
[Cu(HA)2(H2O)2]2+ + AA−	14.6	16.9	31.5	2.80	0.9993b	2.80	2.80

^af(${\mathrm{O}}_2^{\bullet -}$).

^bf(AA⁻).

4. Conclusion

The antioxidant activity of MD was investigated by evaluating the radical scavenging activity and the OIL-1 suppression of copper-catalysed oxidative damage in biological systems using computer calculations. It was found that MD exhibits moderate hydroperoxyl radical scavenging activity in both lipid and polar media. The antiradical activity in non-polar environments follows the FHT mechanism at the O3−H bond, whereas in aqueous solution, it follows the SET pathway of the anion state. Chelation with MD could suppress the Cu(II) reduction by ${\mathrm{O}}_2^{\cdot -}$ and AA⁻ in aqueous environment. Thus MD is predominantly an OIL-1 antioxidant.

Data accessibility

All relevant necessary data to reproduce all results in the paper are within the main text, electronic supplementary material and the Dryad Digital Repository: https://doi.org/10.5061/dryad.bzkh18997.

Authors' contributions

N.T.H.: data curation, formal analysis, investigation, validation, visualization, writing—original draft; M.V.B.: data curation, formal analysis, investigation, validation, writing—original draft; A.M.: conceptualization, project administration, software, supervision, writing—review & editing; Q.V.V.: conceptualization, formal analysis, methodology, project administration, resources, software, supervision, validation, writing—original draft, writing—review & editing. All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Competing interests

We declare we have no competing interests.

Funding

This research is funded by the Vietnamese Ministry of Education and Training under project no. B2021-DNA-16.