Cellular reactive oxygen species (ROS) are produced in the mitochondria, cell membrane, endoplasmic reticulum, cytoplasm and peroxisome and are essential for important cellular pathways [1]. Mitochondria are the major contributors to ROS production (about 90%) at the electron transport chain during the oxidative phosphorylation process [2]. Oxidative stress results when endogenous cellular antioxidants cannot manage excess concentration of ROS. This results in modifications to cellular pathways and oxidation of lipids, proteins, and DNA [3]. Oxidative stress is a precursor to many diseases and is also increased in aging [4]. An important ROS is the superoxide radical anion (O₂^•−) that has a significant biological role. It is involved in oxygen consumption for ATP formation and storage in the mitochondria. About 0.2–2.0% of the molecular oxygen consumed by mitochondria is reduced to superoxide anions, which are converted to other ROS, such as H₂O₂ and hydroxyl radical [2]. Leaking of the superoxide radical from the mitochondria membrane is strongly suspected to be responsible for the development of diabetes, cancer, neurodegenerative and other severe health conditions, including COVID-19 [5,6,7].

Endogenous antioxidant defense metalloproteins, such as the superoxide dismutases (SODs) protect cells by lowering excess superoxide concentration through catalyzing its chemical degradation to give O₂ and H₂O₂. This is performed by Cu and Zn SODs in the intermembrane space and Mn SOD in the matrix [8]. The H₂O₂ formed is degraded by specific enzymes, including, for instance, catalase [9].

Assisting in the decrease of superfluous superoxide leakage from mitochondrial membranes are plant polyphenols, naturally occurring compounds found in fruits and vegetables. The mechanistic role that these polyphenols play is the subject of many studies. For instance, a recent electrochemical study among flavonoids (AH) and superoxide itemized most of the cyclic voltammetry interactions using the following Equations (1)–(4) [10].

Specifically, Equation (1) indicates that superoxide is obtained after reduction of oxygen in the voltaic cell, containing an aprotic solvent, such as DMSO or DMF; Equation (2) involves the AH flavonoid donating an H atom to superoxide (or superoxide abstracting a proton from the AH); in Equation (3), another proton is donated to the previously obtained HO_2⁻ to form H₂O₂. As a result, it can be concluded that from AH antioxidant interaction with superoxide, H₂O₂ is produced.

2. 4-Methyl-7,8-di-hydroxy Coumarin

4-Methyl-7,8-di-hydroxy coumarin, as shown in Figure 1A, is an antioxidant [11,12], which can be associated with a standard superoxide scavenging mechanism through polyphenol hydrogen atom abstraction. Computational results describe the interaction between one polyphenol H(hydroxyl) and the radical; Figure 1B shows the initial van der Waals interaction between the two species, which we call type σ. Upon DFT geometry optimization, this system evolves toward formation of an HO₂- species, separated 1.529 Å from the remaining radical semiquinone, having the unpaired electron located in the reacted ring, Figure 1C. Next, the Figure 1C DFT minimized structure has HO₂⁻ anion posed at van der Waals distance to a proton (2.60 Å) in Figure 1D. Finally, upon geometry minimization, this evolves toward H₂O₂ formation, Figure 1E, which is the reaction described in Equation (3). After elimination of H₂O₂ from Figure 1E the semiquinone is minimized. Hence, this polyphenol can interact with an additional superoxide, van der Waals posed π-π to the centroid ring (3.50 Å), Figure 1F. When this initial configuration is DFT geometrically optimized, Figure 1G, the result differs markedly from the σ attack by the superoxide in Figure 1C. In fact, the unpaired superoxide electron gets transferred into the ring and the formed molecule of O₂ is displaced 5.989 Å, Figure 1G. This polyphenol anion reacts easily with an additional proton to reform the coumarin, Figure 1H. All these reactions proceed without any energy barrier. Therefore, the overall reaction involves 2 superoxide anions (Figure 1B,F) plus 2 protons (Figure 1D,H), and results in formation of H₂O₂ (Figure 1E) and O₂ (Figure 1G), as well as coumarin reformation (Figure 1H), ready for a further cycle of scavenging. The overall sequence is represented in Equation (5), which is the same shown by the SOD enzymes.

3. Galangin

A recent study by us has shown the scavenging of superoxide by the flavonoid galangin [13]. As seen with the coumarin earlier described, a straightforward SOD action is observed when the first attack of superoxide occurs on galangin H3 and is shown in Scheme 1 below. As described earlier with 4-methyl-7,8-di-hydroxy coumarin, Scheme 1 shows galangin mimics SOD action, where the reactants are 2 superoxide radicals, one attacking type σ [reaction involving compound (a)] and the second reacting π-π [involving compound (d)] plus two protons, one used to form H₂O₂ [involving compound (c)] and the second proton regenerating galangin [involving compound (f)]; the products are H₂O₂ [involving compound (c)] plus O₂ involving compound (e). An alternative initial π-π approach involving one aromatic ring of galangin and superoxide was also explored and is shown in Scheme 2 The original study [13] has a typographical error whereby Scheme 1 and Scheme 2 are identical. The correct Scheme 1 is shown in this study. PlosOne was informed and a corrigendum should appear accordingly [13]. A molecule of H₂O₂ results from the superoxide release of its unpaired electron to the galangin ring [steps (d) → (e)], and it is still linked to the ring. Figure 2 completes this study to show the mechanism of scavenging superoxide by galangin. Thus, the overall process also involves the formation and release of H₂O₂ [steps (f)–(g)], plus a molecule of oxygen separated 3.106 Å, Figure 2 right. With this alternative mechanism, galangin also mimics the SOD enzyme action, Equation (5). However, this alternative galangin procedure, following first a π-π and later a σ interaction with superoxide, differs from the 4-methyl coumarin mechanism described earlier and galangin Scheme 1. In fact, the reformed galangin includes a molecule of O₂ and one of H₂O₂, π-π bound to a galangin pyrone ring, e.g., η-(H₂O₂)-galangin-η-O₂, Figure 2 left. Also, this alternative procedure enables galangin to dismutate two superoxide radicals producing H₂O₂, Scheme 2g, and O₂ (Figure 2 right) and galangin reformation, Figure 2 right.

4. Butein

A study of butein and other related chalcone antioxidants was recently described by us [14]. RRDE cyclic voltammetry and X-ray diffraction data, followed by DFT and docking calculations show antioxidant and potential antimalarial properties. As shown by other antioxidants described in this work, butein (Figure 3A) is able to scavenge superoxide. The sequence of superoxide interaction starts with the butein H(hydroxyl) group in position 4 (ring B), Figure 3B. The whole sequence is shown in Figure 3A–F and is closely related to what has been above described for 4-methyl-7,8-di-hydroxy coumarin and galangin (Scheme 1). The overall reaction sequence is again represented by the superoxide dismutase enzyme (SOD) Equation (5). The involved molecules are seen in Figure 3B,E (2 reacting O₂•⁻) and Figure 3D (H₂O₂ formation), whereas O₂ remains trapped within the reformed butein (Figure 3F). We envision the reformed butein containing a π-π molecule of O₂ bound to ring B, as the potential catalyst, Figure 3F, and differing structurally from the reformed three previously discussed cases: reformed 4-methyl-7,8-di-hydroxy coumarin, galangin (Scheme 1) and galangin stemming from Scheme 2. It is well known that the specific structure and arrangements of hydroxyls in the aromatic rings of a polyphenol can determine important mechanistic differences, as observed in this study.

5. Additional Scavengers

In this section, we study only the initial superoxide π-π interaction on other polyphenols. That is, this action precludes the most common superoxide σ attack on an H(hydroxyl). Figure 4 shows the DFT results after posing a superoxide (having O–O bond length of 1.373 Å) at 3.50 Å from a ring centroid. In Figure 4A, the energy-minimized 4-methyl-7,8-dimethoxy-coumarin structure [12] shows that the superoxide penetrates the ring environment, with centroid separation of 3.042 Å, while the superoxide O–O bond length becomes shorter, 1.323 Å, and so a coumarin-η-superoxide complex is formed. As seen in Figure 4B, a similar complex is reached by DFT minimization of the dimethoxy coumarin fraxidin [15], with even shorter separation between centroids, 2.927 Å. We were also interested in investigating potential superoxide interaction with the non-pyrone ring of fraxidin, Figure 4C. This interaction between both species is weak, as shown by separation between centroids of 3.348 Å, closer to the initial van der Waals separation. Hence, we studied the interaction between the pyrone ring of another coumarin, bergamottin [15]. The related bergamottin-η-superoxide complex shows a strong interaction with superoxide, as the distance between them is 2.955 Å, Figure 4D. Additional π-π complexes with superoxide are formed by the antioxidants galangin, Figure 4E, and 4-methyl,3,6,8-triacetate-coumarin [12], Figure 4F. These interactions (3.190 Å and 3.162 Å, respectively) are somewhat weaker than that of bergamottin and more similar to the superoxide-η-4-methyl-7,8-dimethoxy-coumarin complex. Butein here is an exception to compounds shown in Figure 4, where the π-π attack of superoxide is directed towards H(hydroxyl) in position 4, and so it is equivalent to the σ mechanism of scavenging shown in Figure 1B and Figure 3B. The related sequence is shown as a video in manuscript Supplementary Material, Video S1.

6. RRDE Cyclovoltammetry

Most of the studied compounds described in this work have also been studied using the Rotating Ring Disk Electrode method. This was developed in our lab [16] and resulted in a quantitative description for polyphenol scavenging of the superoxide radical. In an RRDE voltammetry experiment, the generation of the superoxide radicals occurs at the disk electrode, while the oxidation of the residual superoxide radicals (that have not been scavenged by the antioxidant) occurs at the ring electrode. Reaction 1: Reduction of molecular oxygen at disk electrode Reaction 2: Oxidation of superoxide radicals at the ring electrode Thus, the rate at which increasing concentrations of antioxidants scavenge the generated superoxide radicals during the electrolytic reaction is determined by obtaining the percent value of the ring current/disk current at each concentration. These percent values are denoted as the collection efficiency of each antioxidant at different concentrations. Collective efficiency values are plotted against the total concentrations of antioxidants to produce a graph illustrating the effect of increasing concentrations of antioxidants on the scavenging of superoxide radicals in the electrolytic solution. It is observed that for low concentration of antioxidant (μM range), there is a linear trend. Ultimately, the slope of the curves serves as a quantitative measure of the antioxidant activity of superoxide scavengers; the steeper the slope, the stronger the scavenger.

Table 1 shows results according to this method: It is observed that catechol polyphenols are more effective than those having isolated single hydroxyl groups.

Table 1. Slopes of Collection Efficiency (indicators of scavenging superoxide) [16] for several polyphenols using the RRDE method.

BHT	Chrysin	Eriodictyol	DHDM	Butein	Clovamide	Quercetin **	Galangin
-0.16x104 [17]	-1.10x104 [16]	-2.20x104 [16]	−8.0x104 [14]	−11.2x104 [14]	−12.0x104 [18]	−5.30x104 [16] −15.4x104 [17] −15.5x104 [18]	-19.0x104 [13]

(**) Ref [16] has the earliest RRDE determination of quercetin, a polyphenol used as a standard for comparison with each RRDE scavenger analyzed by us. Quercetin in DMSO solutions tends to slightly decrease its yellowish color and its scavenging activity with time. Consequently, all successive RRDE studies were done using fresh solutions. Quercetin measured in this manner [17,18] shows the correct slope, steeper than in [16].

7. Conclusions

Polyphenols are valuable natural products present in our diet that may contribute to decrease aging effects and some diseases. However, because of their poor absorption in the gut and consequent low concentration in biological fluids, reservations about antioxidant polyphenol efficiency have been raised. In this study, we focus our attention on the potential reformation of polyphenols after scavenging superoxide and we find that polyphenol mimicking SOD enzyme activity is feasible for several polyphenolic compounds. Thus, after scavenging superoxide, 4-methyl-7,8-di-hydroxy-coumarin and galangin (Chart 1) exactly reform themselves; butein incorporates a molecule of O₂ in the polyphenol system, e.g., a η-O₂-butein complex is formed; galangin alternative mechanism (Chart 2) includes an additional H₂O₂ ligand and forms the η-(H₂O₂)-galangin-η-O₂ complex.

We demonstrate using DFT methods that the π-π interaction between superoxide and one polyphenol ring is associated with oxidation of the superoxide radical, due to transfer of its unpaired electron to an aromatic ring of the polyphenol. This is also responsible for O₂ release as part of SOD activity here proposed. However, it is very difficult to establish if this π-π reaction

O₂•^- + polyphenol-OH → O₂ + polyphenol-OH•^-

proceeds before or after the most common mechanism of polyphenol scavenging (abstraction of H(hydroxyl), σ scavenging)

O₂•^- + polyphenol-OH → HO₂^- + polyphenol-O•

In addition, the initial direct π-π interaction between superoxide and one polyphenol aromatic ring was also studied for a variety of polyphenols. In this specific initial π-π interaction, all polyphenols studied are shown through DFT methods to have minimum energy radical complexes of the type polyphenol-η-O₂ except for butein. By circumventing the π-π interaction butein resulted in behaving as a typical σ scavenger of superoxide, e.g. sequestering a H(hydroxyl). It is suggested that these polyphenols are able to mimic SOD enzyme action in the disproportionation reaction of O₂•^- [equation (5)] and, thus refute the claim that low polyphenol concentration in biological fluids as a limiting factor in their effective scavenging of superoxide.