Concentration Dependence of Anti- and Pro-Oxidant Activity of Polyphenols as Evaluated with a Light-Emitting Fe2+-Egta-H2O2 System

Hydroxyl radical (•OH) scavenging and the regeneration of Fe2+ may inhibit or enhance peroxidative damage induced by a Fenton system, respectively. Plant polyphenols reveal the afore-mentioned activities, and their cumulative net effect may determine anti- or pro-oxidant actions. We investigated the influence of 17 phenolics on ultra-weak photon emission (UPE) from a modified Fenton system (92.6 µmol/L Fe2+, 185.2 µmol/L EGTA (ethylene glycol-bis(β-aminoethyl-ether)-N,N,N′,N,-tetraacetic acid) and 2.6 mmol/L H2O2 pH = 7.4). A total of 8 compounds inhibited (antioxidant effect), and 5 enhanced (pro-oxidant effect) UPE at all studied concentrations (5 to 50 µmol/L). A total of 4 compounds altered their activity from pro- to antioxidant (or vice versa) along with increasing concentrations. A total of 3 the most active of those (ferulic acid, chlorogenic acid and cyanidin 3-O-glucoside; mean UPE enhancement by 63%, 5% and 445% at 5 µmol/L; mean UPE inhibition by 28%, 94% and 24% at 50 µmol/L, respectively) contained catechol or methoxyphenol structures that are associated with effective •OH scavenging and Fe2+ regeneration. Most likely, these structures can determine the bidirectional, concentration-dependent activity of some phenolics under stable in vitro conditions. This is because the concentrations of the studied compounds are close to those occurring in human fluids, and this phenomenon should be considered in the case of dietary supplementation with isolated phenolics.


Introduction
Numerous phytochemicals, including phenolics, are recognized as potent antioxidants that may contribute to the health benefits of diets rich in fruits and vegetables [1,2]. The antioxidant effect of ingested polyphenols is multidirectional; they can stimulate the biosynthesis of antioxidant enzymes [3], suppress processes leading to the production of reactive oxygen species (ROS) in vivo [3,4] or directly inhibit or decompose ROS in chemical reactions [5]. Direct antioxidant activity may involve reactions with ROS and reactive nitrogen species (e.g., hydroxyl radicals, superoxide radicals, nitric oxide or peroxynitrite) to form unreactive products and the chelation of transition metals ions (Fe 2+ , Fe 3+ , Cu 2+ ) to prevent their reactions with H 2 O 2 , leading to the generation of hydroxyl Table 1 shows the antioxidant activity of eight phenolics evaluated with the Fe 2+ -EGTA-H 2 O 2 test in descending order. 3,4-dihydroxyphenylacetic acid and orthocresol, already at concentrations of 5 µmol/L, almost completely inhibited the UPE of the Fe 2+ -EGTA-H 2 O 2 system. Moderate suppression (mean inhibition ranged from 46% to 30% at concentration of 5 µmol/L) of the UPE was noted for homovanillic acid, vanillic acid and caffeic acid. The lowest antioxidant activity was found for 4-hydroxy phenyl acetic acid, 3-hydroxybenzoic acid and hippuric acid. In addition, the last two compounds, even at concentrations of 50 µmol/L, did not inhibit more than one third of light emission from the Fe 2+ -EGTA-H 2 O 2 system. Of six compounds causing moderate or weak UPE suppression at concentrations of 5 µmol/L, only caffeic acid and 4-hydroxy phenyl acetic acid revealed a positive, almost linear relationship between compound concentration and antioxidant effects (Table 1).    Fe 2+ -EGTA-H 2 O 2 system was found for gallic acid (1689 ± 358%), and the lowest one (but still significant) was observed for pelargonidin-3-O-rutinoside (75 ± 7%). A distinct positive linear relationship between compound concentration and the % enhancement of UPE was observed only for phloroglucinol and ellagic acid ( Table 2).                         (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 25 µmol/L and 50 µmol/L were antioxidant. Trolox, which served as a positive control, behaved in the same manner. Resorcinol was antioxidant at a concentration of 5 µmol/L, whereas at a higher concentration, it revealed a pro-oxidant effect. It should be mentioned that, for ferulic acid and resorcinol, a linear relationship between compound concentration and effect on light emission was noted (Table 3). Three phenolics, ferulic acid, chlorogenic acid and cyanidin 3-O-glucoside at concentrations of 5 µmol/L, enhanced the UPE of the Fe 2+ -EGTA-H2O2 system. Both ferulic acid and cyanidin 3-O-glucoside at concentrations of 25 µmol/L still enhanced UPE (but the effect was significantly lower), and at concentrations of 50 µmol/L, they revealed an antioxidant effect. A decrease in light emission from the modified Fenton system was noted (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 2. Three phenolics, ferulic acid, chlorogenic acid and cyanidin 3-O-glucoside at concentrations of 5 µmol/L, enhanced the UPE of the Fe 2+ -EGTA-H2O2 system. Both ferulic acid and cyanidin 3-O-glucoside at concentrations of 25 µmol/L still enhanced UPE (but the effect was significantly lower), and at concentrations of 50 µmol/L, they revealed an antioxidant effect. A decrease in light emission from the modified Fenton system was noted (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 2.   (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 2.   (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 2.   (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 2.   (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 2.

Polyphenols which Altered their Antioxidant Activity into Pro-Oxidant Activity (or Vice Versa) within the Concentration Range of 5 µmol/L to 50 µmol/L
Three phenolics, ferulic acid, chlorogenic acid and cyanidin 3-O-glucoside at concentrations of 5 µmol/L, enhanced the UPE of the Fe 2+ -EGTA-H2O2 system. Both ferulic acid and cyanidin 3-O-glucoside at concentrations of 25 µmol/L still enhanced UPE (but the effect was significantly lower), and at concentrations of 50 µmol/L, they revealed an antioxidant effect. A decrease in light emission from the modified Fenton system was noted (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 2.

Polyphenols which Altered their Antioxidant Activity into Pro-Oxidant Activity (or Vice Versa) within the Concentration Range of 5 µmol/L to 50 µmol/L
Three phenolics, ferulic acid, chlorogenic acid and cyanidin 3-O-glucoside at concentrations of 5 µmol/L, enhanced the UPE of the Fe 2+ -EGTA-H2O2 system. Both ferulic acid and cyanidin 3-O-glucoside at concentrations of 25 µmol/L still enhanced UPE (but the effect was significantly lower), and at concentrations of 50 µmol/L, they revealed an antioxidant effect. A decrease in light emission from the modified Fenton system was noted (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 2.  Resorcinol Three phenolics, ferulic acid, chlorogenic acid and cyanidin 3-O-glucoside at concentrations of 5 µmol/L, enhanced the UPE of the Fe 2+ -EGTA-H2O2 system. Both ferulic acid and cyanidin 3-O-glucoside at concentrations of 25 µmol/L still enhanced UPE (but the effect was significantly lower), and at concentrations of 50 µmol/L, they revealed an antioxidant effect. A decrease in light emission from the modified Fenton system was noted (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 2.

Versa) within the Concentration Range of 5 µmol/L to 50 µmol/L
Three phenolics, ferulic acid, chlorogenic acid and cyanidin 3-O-glucoside at concentrations of 5 µmol/L, enhanced the UPE of the Fe 2+ -EGTA-H2O2 system. Both ferulic acid and cyanidin 3-O-glucoside at concentrations of 25 µmol/L still enhanced UPE (but the effect was significantly lower), and at concentrations of 50 µmol/L, they revealed an antioxidant effect. A decrease in light emission from the modified Fenton system was noted (Table 3). However, for chlorogenic acid, "the break point" occurred earlier; concentrations of 2.

Discussion
Plant polyphenols can act as antioxidants or pro-oxidants depending on the kind of studied compound and are used in Fenton systems in vitro [9,18]. For instance, myricetin inhibited deoxyribose degradation by the Fe 3+ -EDTA-H 2 O 2 -ascorbic acid system and enhanced this process by Fe 3+ -EDTA. On the other hand, tricetin (which differs from myricetin in the lack of -OH group on ring C at C 3 ) also protected deoxyribose from oxidative damage induced by Fe 3+ -EDTA-H 2 O 2 -ascorbic acid, whereas it had a very weak pro-oxidant effect in the case of Fe 3+ -EDTA [18]. In another study, the effect of 13 phenolics on Fe 2+ -EDTA-H 2 O 2 -induced deoxyribose degradation was evaluated; 4 of them significantly protected deoxyribose (antioxidant effect), and 7 revealed significant pro-oxidant activity. In addition, it is interesting that phenolics having remarkably similar structure revealed the opposite activity. 3,4-dihydroxycinnamic acid has a longer aliphatic substitute (just one carbon atom) on the catechol ring than 3,4-dihydroxyphenylacetic acid. However, 3,4-dihydroxycinnamic acid inhibited deoxyribose degradation, whereas the latter one enhanced the oxidative damage to this monosaccharide [9]. Our results confirm these observations but additionally show that a given phenolic tested under fixed conditions of a modified Fenton redox system can function as pro-or antioxidant, depending on its concentration. It should be pointed out that studied phenolic concentrations ranging from 5 µmol/L to 50 µmol/L can occur in systemic circulation or in blood in the portal vein after the ingestion of meals rich in plant phenolics [15]. Moreover, this phenomenon seems not to be unique because 4 of the 17 tested phenolics behaved in such a manner. There are numerous redox systems in the human body (e.g., cysteine/cystine (Cys/CySS), reduced glutathione/oxidized glutathione (GSH/GSSG), reduced nicotinamide adenine dinucleotide phosphate/oxidized nicotinamide adenine dinucleotide phosphate (NADPH/NADP + ), Fe 2+ /Fe 3+ and Cu + /Cu 2+ ) [19][20][21][22][23], and many plant phenolics can interact with them at concentrations that can change depending on the type of consumed food [15]. Some compounds may inhibit, and other phenolics can augment peroxidative processes in the human body. Moreover, it is suggested that antioxidants, including plant polyphenols especially ingested in supraphysiological doses, may act as pro-oxidants; can augment oxidative damage to a variety of biomolecules (DNA, proteins and lipids); and can switch on the intracellular signaling pathways leading to the stimulation of an inflammatory response [24]. These clearly show that the issue of dietary supplementation with plant phenolics, in order to suppress oxidative processes in the human body, is very complicated and still requires controlled clinical trials. On the other hand, diets rich in fruits and vegetables have beneficial effects on human health, and the antioxidant and anti-inflammatory effects of ingested polyphenols is one of mechanisms explaining these epidemiological data [24]. Therefore, so far, the optimal prophylactic action is to increase the ingestion of polyphenols as well as vitamins, minerals and dietary fibers with a natural food matrix rather than the supplementation of the diet with one or more purified polyphenols.

Plausible Mechanism by Which Polyphenols May Affect the UPE of the Fe 2+ -EGTA-H 2 O 2 System
The proposed mechanism of light emission by the Fe 2+ -EGTA-H 2 O 2 system was described in detail in our previous study [16]. A summary of the chemical reaction leading to • OH radical generation in our Fenton system is given below: • OH radicals generated in the Fenton reaction can attack one of the ether bonds in the backbone structure of EGTA, leading to its cleavage and to the generation of peroxyl radicals. They can react with each other, with subsequent formations of product with triplet excited carbonyl groups emitting light [16]. However, the yield of triplet excited carbonyl group formation is very low, and thus, the intensity of chemiluminescence is very small and is named UPE (ultraweak photon emission). Free Fe 3+ ions have very low solubility at a pH of  3 . In our system, an excess of EGTA prevents this process, and Fe 3+-EGTA can undergo reduction into Fe 2+ -EGTA by any reducing agent added to the Fenton system. Table 4 shows the main plausible mechanisms by which plant polyphenols may alter the UPE of the 92.6 µmol/L Fe 2+ -185.2 µmol/L EGTA-2.6 mmol/L H 2 O 2 system. The highest tested polyphenol concentration (50 µmol/L) was 52-times lower than the concentration of H 2 O 2 in the modified Fenton system. Therefore, the direct reaction of the tested compound with H 2 O 2 could not significantly decrease the H 2 O 2 pool reacting with Fe 2+ . Moreover, the incubation of each tested compound with Fe 2+ -H 2 O 2 did not result in light emission. EGTA is an effective chelator of divalent cations, and in our experiments, Fe 2+ ions were added to the reaction mixture after polyphenol and EGTA. Furthermore, the concentration of EGTA was 2-times higher than Fe 2+ ions and was 3.7-to 37-times higher than that of the studied polyphenols. Taking the above into consideration, one may conclude that the decomposition of H 2 O 2 and the chelation of Fe 2+ ions by the studied polyphenols, as well as light emissions from polyphenols undergoing oxidation by • OH radicals and H 2 O 2 (mechanisms 1, 3 and 5 shown in Table 4), are not responsible for changes in UPE from the Fe 2+ -EGTA-H 2 O 2 system. There was a 28-fold molar excess of H 2 O 2 in comparison to the Fe 2+ ions in the modified Fenton system. Therefore, the regeneration of Fe 2+ -EGTA by the reduction of Fe 3+ into Fe 2+ could substantially enhance • OH radical generation and, by consequence, the UPE. Conversely, the reaction of the tested compound with • OH radicals decreases light emission by protecting ether bonds of EGTA from oxidative attack. Superoxide radicals (O 2 − ) are also generated in the Fenton system [16]. They can regenerate Fe 2+ -EGTA in the following reaction: Superoxide dismutase (a highly effective O 2 − radical scavenger) inhibited the UPE of the Fe 2+ -EGTA-H 2 O 2 system [16]. Therefore, these three aforementioned processes (mechanisms 2, 4 and 6, shown in Table 4) seem to be responsible for changes in the UPE of the Fe 2+ -EGTA-H 2 O 2 system caused by the evaluated polyphenols. − radicals that protected EGTA from peroxidative attack and that inhibited the regeneration of the Fe 2+ -EGTA complex, respectively. This is in line with the observation that mannitol and DMSO ( • OH radical scavengers), as well as superoxide dismutase, decreased light emission from the Fe 2+ -EGTA-H 2 O 2 system [16]. Polyphenols can react with superoxide radicals, and this includes proton-transfer (acidbase) and/or radical-transfer pathways [25].
Proton transfer mechanism: Radical transfer mechanism: Phenoxyl radicals can dimerize or oligomerize to form nonradical products. Transformation into quinone or semiquinone can be another way of phenoxyl radical elimination. Both phenolic acids (benzoic acid and cinnamic acid derivatives) and polyphenols can react with superoxide, whereas the highest activity was observed for compounds containing a o-diphenol ring (e.g., flavonoids) [25]. All compounds that inhibited the UPE of Fe 2+ -EGTA-H 2 O 2 were phenolic acids (monophenols), and one may suppose that proton transfer is a leading pathway of O 2 − inactivation that was responsible for this phenomenon. However, the pH of the chemical reaction environment was stabilized with phosphate buffer (pH = 7.4), and therefore, the radical transfer mechanism of O 2 − inactivation seems dominant under the conditions of our experiments. Phenolic acids with one hydroxy group in the ortho or para position to the carboxyl group were reported to be effective O 2 − scavengers [26]. Moreover, compounds that possess more than one hydroxy group in their aromatic ring (such as gallic acid, caffeic acid), especially those with -OH groups in the ortho and para position to the carboxyl group, showed antioxidant properties against O 2 − radicals [26,27].
Hydrogen abstraction and -OH addition to double bonds are chemical reactions involved in the inactivation of • OH radicals by polyphenols [28]. • OH radicals can abstract H• from -OH or -OCH 3 substituents of a given polyphenol and can form H 2 O and phenolic radicals [28,29]. In the second type of reaction, • OH radical can be added to the double bond of the phenolic ring of a polyphenol or to a double bond of an aliphatic substitute at the benzene ring. This results in the formation of phenolic hydroxy derivative radicals [28]. The presence of two adjacent (position ortho) -OH substituents (catechol structure) or one -OH and one -OCH 3 on a benzene ring (methoxyphenol structure) facilitates reactivity with • OH radicals by lowering the activation barrier for -OH addition [30]. Since the phenolic radicals did not emit light (Table 5, control experiments, sample no 4, 6 and 7), the resulting • OH radicals inactivation may be responsible for the suppression of the UPE of the Fe 2+ -EGTA-H 2 O 2 system. It should be pointed out that phenolics that revealed the highest antioxidant activity (3,4-dihydroxyphenyl-acetic acid, homovanillic acid, vanillic acid and caffeic acid) had catechol or methoxyphenol structures inside its molecule. The only exception was orthocresol, having a -CH 3 substituent instead of -OCH 3 . Moreover, the aliphatic substitute of caffeic acid has one double bond. The remaining three phenolics (4-hydroxy phenyl acetic acid, 3-hydroxybenzoic acid and hippuric acid) had no aforementioned structures, and their inhibition of the UPE of Fe 2+ -EGTA-H 2 O 2 did not exceed 50% at concentrations of 50 µmol/L and may result from the scavenging of superoxide radicals [26,27,31].

Phenolics That Enhanced Light Emission from the Fe 2+ -EGTA-H 2 O 2 System within the Concentration Range of 5 µmol/L to 50 µmol/L
The regeneration of the Fe 2+ -EGTA complex by the reduction of Fe 3+ into Fe 2+ seems to be the most likely mechanism responsible for the polyphenol-induced enhancement of light emission from the Fe 2+ -EGTA-H 2 O 2 system. However, the ability to reduce ferric ions is recognized as a measure of the antioxidant potential of a given compound [32], but in our experimental model, the high efficacy of Fe 2+ regeneration leads to increased • OH production being the pro-oxidant mechanism. Numerous studies on plant polyphenolinduced Fe 3+ reduction that involved a dozen to several dozens of various compounds have been executed [8,33,34]. They showed that polyphenols can reduce Fe 3+ to Fe 2+ , and some of them had a ferric reducing ability (FRAP), similar to that of equimolar concentrations of ascorbic acid [8]. The FRAP of polyphenols is positively correlated with the presence of catechol or a methoxyphenol ring in their molecules [8,33]. Moreover, three hydroxy groups being located in the ortho or para position to each other results in the increased ability to reduce Fe 3+ ions [34]. The presence of -OH substituents at position 1 and other substituents at position 3 (meta position) enhances the FRAP of phenolic acids [33]. Moreover, the presence of -OH substituents at position 3 on ring C increases the FRAP of flavonoids [33]. On the other hand, phenolic acids that contain only one hydroxy group, as well as those with two hydroxy groups in the meta position, have lower activity than those containing catechol structures [34].
The reaction between the polyphenol -OH substituent (A-OH) and Fe 3+ -EGTA most likely involves the formation of complexes and hydrogen atom transfers, leading to the formation of Fe 2+ -EGTA and phenoxyl radicals (AO • ) [34].
The amount of Fe 2+ -EGTA produced depends on the Fe 3+ to polyphenol ratio and on the type of compound [35,36]. Of five compounds that enhanced the UPE of the Fe 2+ -EGTA-H 2 O 2 system (Table 2), gallic acid, ellagic acid and pelargonidin contained the aforementioned structures responsible for the high ability to reduce Fe 3+ into Fe 2+ . On the other hand, pelargonidin-3-O-rutinoside, at a concentration of 5µmol/L, had lower pro-oxidant activity than pelargonidin, most likely due to the presence of a rutinosyl group at the position 3 [37]. Phloroglucinol, containing three hydroxy groups in the meta position to each other, was the second after gallic acid for the enhancing of light emission from the Fe 2+ -EGTA-H 2 O 2 system. This contrasts with the aforementioned relationships between the chemical structures of phenolics and their ability to reduce Fe 3+ ions. It should be noted that FRAP is measured with tripyridyltriazine under conditions of acidic milieu (pH = 3.6) [32]. In our study, chemical reactions were conducted at a pH of 7.4, and this may explain this discrepancy.
3.4. Polyphenols Which Altered Their Antioxidant Activity into Pro-Oxidant Activity (or Vice Versa) within the Concentration Range of 5 µmol/L to 50 µmol/L As discussed above, the presence of catechol or methoxyphenol structures in polyphenol molecules is associated with the efficient decomposition of • OH radicals and the regeneration of the Fe 2+ -EGTA complex. These processes have the opposite effect on the UPE of the Fe 2+ -EGTA-H 2 O 2 system. The scavenging of • OH radicals decreases light emission, and Fe 3+ reduction into Fe 2+ enhances light emission (Table 4). Three phenolics (ferulic acid, chlorogenic acid and cyanidin 3-O-glucoside), which altered their activity from pro-oxidant into antioxidant along with increasing concentrations, contained a catechol or methoxyphenol ring. It is possible that these phenolics can simultaneously scavenge • OH radicals and reduce Fe 3+ ions, and at higher concentrations, the first process prevails, being responsible for the antioxidant effect reflected by the inhibition of light emission from the Fe 2+ -EGTA-H 2 O 2 system (Figure 1). This assumption can be generalized to other studied phenolics containing the aforementioned rings and can the explain non-linear relationship between compound concentrations and alterations in light emissions revealed by vanillic acid, homovanillic acid (Table 1), gallic acid, pelargonidin and pelargonidin-3-Orutinoside (Table 2). However, the scavenging of O 2 − radicals can be an additional factor responsible for this non-linearity in the case of gallic acid, pelargonidin and pelargonidin-3-O-rutinoside [27,38]. Resorcinol (1,3-dihydroxybenzene, two hydroxy groups at the benzene ring in a meta position) confirms this explanation. At concentrations of 5 and 25 µmol/L, this compound did not alter significantly, whereas at a concentration of 50 µmol/L, it enhanced the UPE of the Fe 2+ -EGTA-H 2 O 2 system. Trolox serves as a positive control because it reveals antioxidant or pro-oxidant activities, depending on its concentrations in experiments with cell cultures [11]. Trolox efficiently reduces Fe 3+ into Fe 2+ ions [39] and scavenges • OH radicals [40]. Moreover, the addition of Trolox into an aerated solution of Cu 2+ ions results in the generation of • OH radicals due to the reduction of Cu 2+ into Cu + [41]. Therefore, the effect of increasing concentrations of Trolox on light emission from the Fe 2+ -EGTA-H 2 O 2 system was bidirectional and depended on the net outcome of the direct scavenging of • OH radicals and on their enhanced formation due to Fe 3+ reduction.

Strengths and Limitations of the Study
The concentrations of reactants used in our experimental model are close to possibly occurring in human body fluids. The plasma levels of H2O2 and iron, complexed with low molecular weight compounds, can reach 50 µmol/L and 10 µmol/L in certain diseases [42,43]. It is possible that H2O2 concentrations can be even higher in activated inflammatory cells. Moreover, 17 phenolics and Trolox were studied at 3 concentrations that can occur in the plasma of systemic circulation, portal circulation or intracellular fluid of enterocytes after eating meals rich in plant phytochemicals [15,44]. In addition, the usage of an undeaerated medium with pH = 7.4, a temperature of 37 °C and not being exposed to sunlight resembles in vivo conditions. These are the advantages of our study, which indicate that the observed phenomena can occur in the human body. Numerous organic buffers, including Hepes (N-(2-hydroxyethyl), piperazine-N′-(2-ethanesulfonic acid)), Tricine (N-[tris (hydroxymethyl)methyl] glycine) and Tris (tris(hydroxymethyl) aminomethane) can effectively scavenge • OH radicals [45]. The phosphate buffer system is important in buffering intracellular fluid. These were the reasons for choosing the phosphate buffer in our experiments. However, phosphate buffer solution is able to bind Fe 2+ and Fe 3+ to form complexes with low solubility and reactivity, thus inhibiting • OH generation by the Fenton system [46]. On the other hand, this inhibitory effect was completely abolished by the addition of EDTA, most likely by the formation of ternary complexes, chelator-Fe 2+ -phosphate [46] or chelator-Fe 3+ -phosphate. Since the EDTA chemical structure is similar, to some extent, to that of EGTA, it can be assumed that the latter has the same effect on • OH formation in a medium containing phosphate buffer. • OH radicals are the major ROS generated in reactions of Fe 2+ with H2O2 only in acidic conditions, whereas high-valent oxoiron (IV) species are mainly generated when a reaction is carried out in a neutral or alkaline environment. However, in the presence of a phosphate buffer, oxoiron (IV) species are efficiently converted to • OH radicals [47]. Therefore, phosphate buffers seem to facilitate • OH radical generation in the Fenton reaction under conditions of physiological pH. Phosphoric acid, in contrast, can inactivate • OH radicals through hydrogen atom transfer reactions [47]. These clearly show the difficulties with the selection of effective and

Strengths and Limitations of the Study
The concentrations of reactants used in our experimental model are close to possibly occurring in human body fluids. The plasma levels of H 2 O 2 and iron, complexed with low molecular weight compounds, can reach 50 µmol/L and 10 µmol/L in certain diseases [42,43]. It is possible that H 2 O 2 concentrations can be even higher in activated inflammatory cells. Moreover, 17 phenolics and Trolox were studied at 3 concentrations that can occur in the plasma of systemic circulation, portal circulation or intracellular fluid of enterocytes after eating meals rich in plant phytochemicals [15,44]. In addition, the usage of an undeaerated medium with pH = 7.4, a temperature of 37 • C and not being exposed to sunlight resembles in vivo conditions. These are the advantages of our study, which indicate that the observed phenomena can occur in the human body. Numerous organic buffers, including Hepes (N-(2-hydroxyethyl), piperazine-N -(2-ethanesulfonic acid)), Tricine (N-[tris (hydroxymethyl)methyl] glycine) and Tris (tris(hydroxymethyl) aminomethane) can effectively scavenge • OH radicals [45]. The phosphate buffer system is important in buffering intracellular fluid. These were the reasons for choosing the phosphate buffer in our experiments. However, phosphate buffer solution is able to bind Fe 2+ and Fe 3+ to form complexes with low solubility and reactivity, thus inhibiting • OH generation by the Fenton system [46]. On the other hand, this inhibitory effect was completely abolished by the addition of EDTA, most likely by the formation of ternary complexes, chelator-Fe 2+ -phosphate [46] or chelator-Fe 3+ -phosphate. Since the EDTA chemical structure is similar, to some extent, to that of EGTA, it can be assumed that the latter has the same effect on • OH formation in a medium containing phosphate buffer. • OH radicals are the major ROS generated in reactions of Fe 2+ with H 2 O 2 only in acidic conditions, whereas high-valent oxoiron (IV) species are mainly generated when a reaction is carried out in a neutral or alkaline environment. However, in the presence of a phosphate buffer, oxoiron (IV) species are efficiently converted to • OH radicals [47]. Therefore, phosphate buffers seem to facilitate • OH radical generation in the Fenton reaction under conditions of physiological pH. Phosphoric acid, in contrast, can inactivate • OH radicals through hydrogen atom transfer reactions [47]. These clearly show the difficulties with the selection of effective and physiologically relevant buffers to conduct the Fenton reaction in vitro. Reactions of polyphenols with O 2 − , • OH radicals and Fe 3+ result in the formation of various polyphenol radicals. The exposition of EGTA to • OH radicals causes the cleavage of the ether bond in the backbone structure of this chelator, with further formation of peroxyl radicals that may convert into products with triplet excited carbonyl groups emitting light [16]. It cannot be excluded that polyphenols and EGTA-derived radicals can react with each other and thus decrease the UPE of Fe 2+ -EGTA-H 2 O 2 . We suggest that the presence of catechol or methoxyphenol as a key structure is responsible for polyphenol reactivity with • OH radicals and Fe 3+ ions. However, we were not able to solve two important questions: (1) What is responsible for the opposite activities (anti-and pro-oxidant) of the two phenolics (e.g., 3,4-dihydroxyphenyl-acetic acid and gallic acid) containing catechol or methoxyphenol structures?; (2) Why can the same phenolic-containing catechol structure (e.g., ferulic acid) have pro-oxidant activity at a concentration of 5 µmol/L and have antioxidant activity at a concentration of 50 µmol/L? The spatial structure of a given phenolic; its molecular mass; the presence and properties of substituents on benzene rings; and the ability to form complexes with Fe 2+ -EGTA or Fe 3+ -EGTA may be the possible determinants of these phenomena. These unresolved questions can be recognized as the limitations of our study. On the other hand, they can be an inspiration for further studies with sophisticated research protocols and equipment to precisely characterize the influence of plant phenolics on redox processes in the human body and the consequences of the dietary supplementation of these phytochemicals.
working solution was added to the tube (Lumi Vial Tube, 5 mL, 12 × 75 mm, Berthold Technologies, Bad Wildbad, Germany) containing 940 µL of PBS, and then 20 µL of 5 mmol/L working solution of FeSO 4 was added. After gentle mixing, the tube was placed in the luminometer chain and was incubated for 10 min in the dark at 37 • C. Then, 100 µL of 28 mmol H 2 O 2 solution was injected by an automatic dispenser, and the total light emission (expressed in RLU-relative light units) was measured for 120 s.
The incomplete system of Fe 2+ -H 2 O 2 as well as Fe 2+ -EGTA without H 2 O 2 served as controls [16,17] To assess the effect of polyphenols on the UPE of 92.6 µmol/L Fe 2+ -185.2 µmol/L EGTA-2.6 mmol/L H 2 O 2 system, 30 µL of the working solution of the studied compounds in PBS or their appropriate dilutions was added to the luminometer tube containing EGTA and FeSO 4 in PBS and was incubated for 10 min at 37 • C in the dark. Then, 100 µL of H 2 O 2 solution was injected, and the total light emission was measured for 2 min. The final concentrations of a given polyphenol in the reaction mixture were 5, 25 and 50 µmol/L.
The control reagent set included: Fe 2+ -EGTA-H 2 O 2 in PBS without polyphenol; an incomplete system of Fe 2+-H 2 O 2 with and without polyphenol; and polyphenol alone in PBS. The final concentration of polyphenol in the controls was 50 µmol/L. Three concentrations of the given polyphenol were tested in one series of experiments repeated at least four times. Table 5 shows the designs of these experiments. The inhibitory effect of the tested polyphenols on UPE is expressed as the percent inhibition (%I) calculated according to the formula: %I = [(A−B)/(A−C)] × 100%, where A, B and C are the total light emissions from Fe 2+ -EGTA-H 2 O 2 ; Fe 2+ -EGTA-studied polyphenol -H 2 O 2 ; and polyphenol alone in the medium, respectively. When UPE was augmented, the percent enhancement (%E) was calculated as follows: %E = [(B−A)/(A−C)] × 100%. We omitted measurement of the background UPE signal from the medium alone (H 2 O injected into PBS) because it was the same as those revealed by the evaluated antioxidants (including polyphenols) alone in the medium [16,17]. This shortened the experiment time and therefore reduced the risk of unexpected errors related to the extended incubation of samples in the luminometer chain.

Statistical Analysis
The results (total light emission, % inhibition or % enhancement of light emission) are expressed as the means (standard deviation), medians and interquartile ranges (IQR). Comparisons between the UPE of the Fe 2+ -EGTA-H 2 O 2 system and the light emission from corresponding samples of a modified system (e.g., an incomplete system, a system with the addition of polyphenol or polyphenol in the medium alone) were analyzed with the independent-samples (unpaired) t-test or the Mann-Whitney U test, depending on the data distribution, which was tested with the Kolmogorov-Smirnov-Lilliefors test. The Brown-Forsythe test, used for analysis of the equality of the group variances, was used prior to the application of the unpaired t-test, and if variances were unequal, then the Welch's t-test was used instead of the standard t-test. Comparisons of the % inhibition or % enhancement of UPE caused by the evaluated polyphenols were analyzed in the same way. A p-value < 0.05 was considered significant.

Conclusions
Plant polyphenols can act as anti-or pro-oxidants in a medium containing a modified Fenton system (Fe 2+ -EGTA-H 2 O 2 ) in vitro. Distinct antioxidant activity was found for 3,4-dihydroxyphenylacetic acid, homovanillic acid, vanillic acid and caffeic acid, whereas gallic acid, phloroglucinol, pelargonidin and ellagic acid were pro-oxidant. Moreover, some of them (e.g., ferulic acid, chlorogenic acid and cyanidin 3-O-glucoside) can alter their activity from pro-oxidant to antioxidant (or vice versa), along with increasing concentrations from 5 µmol/L to 50 µmol/L. The presence of catechol or methoxyphenol in the backbone structures of phenolics seems to be responsible for this phenomena because they are predisposed to efficient • OH radical scavenging (antioxidant activity) and to the regeneration of Fe 2+ by the reduction of Fe 3+ (pro-oxidant activity). The resultant effect of these processes determines the pro-or antioxidant activity of a given plant phenolic. It is assumed that this bidirectional effect of plant phenolics can occur in vivo because the concentrations of the studied phytochemicals and reagents of the Fenton system (except EGTA) are close to those of human fluids. These should be considered in the case of long-term diet supplementation with one or two polyphenols. Our results also indirectly suggest the use of a mixture of numerous plant polyphenols rather than single compounds for prophylactic dietary supplementation.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.