Light Emission from the Fe2+-EGTA-H2O2 System: Possible Application for the Determination of Antioxidant Activity of Plant Phenolics

Oxidative reactions can result in the formation of electronically excited species that undergo radiative decay depending on electronic transition from the excited state to the ground state with subsequent ultra-weak photon emission (UPE). We investigated the UPE from the Fe2+-EGTA (ethylene glycol-bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid)–H2O2 system with a multitube luminometer (Peltier-cooled photon counter, spectral range 380 to 630 nm). The UPE of 92.6 µmol/L Fe2+—185.2 µmol/L EGTA—2.6 mmol/L H2O2 reached 4319 ± 755 relative light units during 2 min measurement and was about seven times higher (p < 0.001) than the UPE of incomplete systems (Fe2+-H2O2, EGTA-H2O2) and medium alone. Substitution of Fe2+ with Cr2+, Co2+, Mn2+ or Cu2+ as well as of EGTA with EDTA (ethylenediaminetetraacetic acid) or citrate completely abolished UPE. Experiments with ROS scavengers revealed the dependence of UPE on hydroxyl radicals suggesting occurrence of oxidative attack and cleavage of the ether bond in EGTA backbone structure and formation of triplet excited carbonyl groups with subsequent light emission. Plant phenolics (ferulic, chlorogenic and caffec acids) at concentration 87 µmol/L and ascorbate at 0.46 mmol/L inhibited UPE by 90 ± 4%, 90 ± 5%, 97 ± 2% and 92 ± 1%, respectively. Quenching of UPE from Fe2+-EGTA-H2O2 system can be used for evaluation of antioxidant activity of phytochemicals.


Introduction
Oxidative metabolic reactions in living cells can result in the formation of electronically excited species. They undergo radiative decay depending on electronic transition from the singlet or triplet excited state to the singlet ground state what is accompanied by ultra-weak photon emission (UPE) [1]. In vitro oxidation of pure lipids or cell cultures and organ homogenates containing lipids (especially brain homogenates) resulted in UPE along with accumulation of various products of lipid peroxidation [2][3][4][5][6]. Free radical scavengers (mannitol, butylated hydroxytoluene, D-α-tocopherol) quenched UPE related to lipid peroxidation in cell culture and organ homogenates [3,4]. Moreover, rats fed a tocopherol-free diet for seven months revealed higher ex vivo UPE from brain, liver and heart homogenates than animals on normal feed [4]. Fenton's reagent (solution of H 2 O 2 with Fe 2+ ) and chelate-modified Fenton's reagent are used to study the hydroxyl radical ( • OH)-induced peroxidative damage to various organic compounds and biomolecules and in some cases this can be accompanied by light emission including UPE [7][8][9]. Because, the Fenton reaction involves the creation of reactive oxygen species (ROS) by chemicals that are present in vivo these experimental models have importance in studies on free radicals related pathology in humans [10,11]. Moreover, measurement of UPE as well as other forms of chemiluminescence related to peroxidative damage to biomolecules can be used for monitoring the effectiveness of various compounds as potential ROS scavengers and pharmacological interventions leading to suppression of oxidative stress in vivo [12,13]. In addition, chemiluminescence can reflect the intensity of oxidative processes and current balance between generated ROS and antioxidant capacity in various cells and tissues under normal conditions [1,14].
A recent study showed that chelating agents frequently used to modify Fenton's reagent such as EDTA and EGTA can react with various oxidants including hypochlorite, peroxyl radicals and peroxynitrite [15]. It cannot be excluded that these reactions lead to generation of electronically excited chemical groups in chelating compounds with subsequent light emission. Although, Fenton's reagent alone was reported to generate UPE [7,8] no data exist (to the best of our knowledge) on photon emission from Fe 2+ -EGTA-H 2 O 2 system. Therefore, in this study we investigated the chemilumiescence of Fe 2+ -EGTA-H 2 O 2 system with special attention to elucidate what ROS are involved in this phenomenon and its possible application as a tool for evaluation of antioxidant activity of selected phenolic acids.

Ultra Weak Photon Emission from Fe 2+ -EGTA-H 2 O 2 System
The complete system 9.3 µmol/L Fe 2+ -18.5 µmol/L EGTA-0.26 mmol/L H 2 O 2 emitted 1351 ± 178 relative light units (RLU) during 2 min of measurement. This was about 2.2-times higher (p < 0.001) than UPE from incomplete systems (Fe 2+ -H 2 O 2 or EGTA-H 2 O 2 ) and 2.5-times higher (p < 0.001) than that of medium alone (Table 1).  (Table 1). This increase was not linear and the difference between mean UPE of Fe 2+ -EGTA-H 2 O 2 and mean background photon emission of medium alone (H 2 O injected to PBS) was 821 RLU, 1061 RLU, 3718 RLU and 5781 RLU for the baseline and 5-, 10-and 20-times higher concentrations of the modified Fenton system, respectively. It should be pointed out that light emission from Fe 2+ -H 2 O 2 (incomplete system I, Fenton reagent) in the case of two highest concentrations (experiment C and D, Table 1) was higher (p < 0.05) than those from corresponding control systems. The UPE of 92.6 µmol/L Fe 2+ -185.2 µmol/L EGTA-2.6 mmol/L H 2 O 2 reached 4319 ± 755 RLU and was about seven times higher than UPE of incomplete systems and medium alone ( Table 1). This concentration of Fenton system was used for further experiments.

Effect of Iron and EGTA Replacement by other Divalent Cations and Metal Chelators on Light Emission from Fe 2+ -EGTA-H 2 O 2 System
Replacement of Fe 2+ with other divalent cations (Cr 2+ , Co 2+ , Mn 2+ , Cu 2+ ) almost completely abolished any light emission from the Fenton system ( Figure 1). Mean UPE ranged from 513 ± 41 RLU for Cu 2+ -EGTA-H 2 O 2 to 625 ± 62 RLU for Cr 2+ -EGTA-H 2 O 2 and was comparable to that observed for medium alone (508 ± 28 RLU). Similar results were observed for Fe 2+ -EDTA-H 2 O 2 and Fe 2+ -citric acid-H 2 O 2 , both systems emitted 6.5-and 6-times less photons ( Figure 1) than Fe 2+ -EGTA-H 2 O 2 over 2 min of counting, respectively. Mean photon emission from all corresponding control systems did not exceed 690 RLU (data not shown).

Light Emission from Fe 2+ -EGTA-H 2 O 2 System
We found that Fe 2+ -EGTA-H 2 O 2 system emitted light in a concentration dependent (but not linear) manner under conditions of a stable ratio of molar concentrations of its constituents. Injection of H 2 O 2 to Fe 2+ or EGTA alone did not result in the significant increase in UPE. In these cases the light emission was similar or slightly higher than that observed for medium alone. Substitution of Fe 2+ with other divalent cations (Cu 2+ , Mn 2+ , Co 2+ and Cr 2+ ) almost completely abolished UPE. The same effect was observed for substitution of EGTA with EDTA or citric acid. These indicate that UPE is specific for Fe 2+ -EGTA-H 2 O 2 system and could not be obtained from the combination of other divalent cations and chelating agents with H 2 O 2 .

Plausible Mechanism of Light Generation from
It is well known that Fenton reagent (Fe 2+ -H 2 O 2 ) generates UPE via O 2 ( 1 ∆ g ) formation and its decay [7,8] with a three characteristic bands emission at 1270 nm (monomolecular decay from its first excited state to ground state), and at 634 nm and 703 nm (bimolecular transition) [17]. Two of these bands (1270 nm and 703 nm) were far away and one (634 nm) was at the border of the spectral range (from 380 nm to 630 nm) of detection of our luminometer. The width of the 634 nm band is about 35 nm, therefore this may be at least in part responsible for UPE of Fe 2+ -EGTA-H 2 O 2 under conditions of our experiments. However, NaN 3 a scavenger of O 2 ( 1 ∆ g ) did not significantly decrease UPE which suggests that O 2 ( 1 ∆ g ) was not involved in UPE of Fe 2+ -EGTA-H 2 O 2 system. There are two possible explanations of this observation: firstly, the intensity of photons emission related to bimolecular decay of O 2 ( 1 ∆ g ) was too low to significantly contribute to UPE of Fe 2+ -EGTA-H 2 O 2 system and was not detected; secondly, almost all O 2 · − (generated in reaction 5) was consumed for reduction of Fe 3+ -EGTA complex (reaction 6) and thus formation of O 2 ( 1 ∆ g ) and subsequent photons emission was inhibited. Another source of UPE are triplet excited carbonyl groups ( 3 (R=C)*) [2,18,19] emitting photons with spectral range of 350 nm to 550 nm [12]. 3 (R=C)* can be formed by • OH-induced oxidation of various low molecular weight compounds (e.g., uric acid, vitamin B 12 , tryptophan) as well as lipids and DNA [12]. • OH can attack the ether group in the backbone chain of various molecules resulting in the cleavage of the ether bond [20][21][22] with its further degradation and formation of another radical and carbonyl group [20,23]. EGTA contains two ether bonds in the middle of the backbone chain ( Figure 3). It is possible that • OH can react with these bonds and form 3 (R=C)* with subsequent photons emission. Figure 4 shows the proposed mechanism of these reactions. This suggested elucidation of light emission from Fe 2+ -EGTA-H 2 O 2 system is supported by three observations:

Inhibitory Effect of Phenolic Acids on Light Emission from Fe 2+ -EGTA-H 2 O 2 System
According to the proposed mechanism of UPE from Fe 2+ -EGTA-H 2 O 2 any given compound would inhibit light emission if: (A) it effectively scavenges at least one of the following ROS: H 2 O 2 , • OH and O 2 − ; (B) is a stronger Fe 2+ chelating agent than EGTA. The first action would result in the decreased activity of • OH and protection of EGTA ether bonds from oxidative attack. The second one would involve abstraction of Fe 2+ ions from Fe 2+ -EGTA complex and formation of another complex to be less effective in reaction with H 2 O 2 leading to • OH formation. Moreover, part of • OH radicals formed in this complex can oxidize other molecules (e.g., added compound with chelating properties or bicarbonate ions derived from dissolved in water atmospheric CO 2 [24]) before their reaction with ether bonds of EGTA and subsequent light emission. All three tested phenolic acids inhibited UPE from the Fe 2+ -EGTA-H 2 O 2 system. They have hydroxyl substituents in the backbone aromatic ring: ferulic acid one, caffeic and chlorogenic acids have two. It is possible that • OH can grab a hydrogen atom from one of the hydroxyl groups at the phenolic ring to form water and a less reactive and more stable radical. Thus less • OH was available for photon emitting reactions with ether bonds of EGTA. The observation that ferulic acid was a weaker inhibitor of UPE than caffeic acid is in line with this explanation. Moreover, this plausible mechanism of UPE inhibition from Fe 2+ -EGTA-H 2 O 2 system by phenolic acids is in agreement with previous reports demonstrating an intensification of the • OH scavenging activity of flavonoids with an increased number of -OH substituents in an aromatic ring [25]. Furthermore, hydroxyl groups and catechol group, at position 3, 5, 7 and 40 are critical for the effective scavenging of peroxynitrite by flavonoids [26] as well as the inhibition of total ROS generation in kidney homogenates by flavonoids intensifies as the number of total -OH groups in their structure increases [27]. Moreover, the protective effect of polyphenols against • OH-induced degradation of deoxyribose correlated with the number of -OH substitutions in the backbone structure [28]. Our results correspond well with a studies showing distinct • OH and O 2 · − scavenging activity of chlorogenic and caffeic acids in vitro [29] and the inhibitory effect of ferulic acid on • OH-induced damage to synaptosomes and neuronal cells [30,31]. In another study ferulic acid scavenged O 2 · − as proved by using electron spin resonance spectroscopy [32]. Therefore, decomposition of O 2 · − apart from direct scavenging of • OH may additionally be responsible for inhibitory effect of studied phenolic acids on UPE from Fe 2+ -EGTA-H 2 O 2 system. These phenolic acids can also from complexes with divalent cations including Fe 2+ [33][34][35] which can decrease Fe 2+ reactivity with H 2 O 2 [33,34]. However, two-fold molar excess of EGTA compared to Fe 2+ ions in the reaction mixture seems to prevent formation of Fe 2+ -phenolic acid complexes. It is in line with previous studies showing negligible binding of Fe 2+ and Fe 3+ to polyphenols in the presence of excess of EDTA another strong chelating agent [34,36,37]. Therefore, formation of phenolic acid-iron complexes had insignificant contribution to phenolic acid-induced suppression of UPE from Fe 2+ -EGTA-H 2 O 2 . Caffeic, chlorogenic and ferulic acids were able to reduce Fe 3+ to Fe 2+ [38]. Thus they can replace O 2 − as an Fe 3+ reducing agent (reaction 6) and enhance • OH generation in the Fe 2+ -EGTA-H 2 O 2 system. Therefore although all phenolics revealed about 90% inhibition of UPE at concentration of 87 µmol/L we tested 2-and 10-times higher concentrations to exclude any possible pro-oxidant action of these compounds. Ascorbic acid is also a powerful Fe 3+ reducing agent [38] and is frequently used as a component of modified Fenton systems (e.g., Fe 3+ -H 2 O 2 -ascorbate and Fe 3+ -EDTA-H 2 O 2 -ascorbate) to enhance • OH generation in in vitro studies on antioxidant properties of various phytochemicals [28]. However, ascorbic acid itself can scavenge various ROS including • OH [39,40] and is recognized as an efficient antioxidant vitamin in vivo [40,41]. Ascorbic acid almost entirely inhibited UPE from Fe 2+ -EGTA-H 2 O 2 system which shows that scavenging of • OH by this vitamin definitely prevailed over potential pro-oxidant action under conditions of our experiments.

Strengths and Weaknesses of the Study
The photon counter of luminometer used for measurement of light emission had narrow spectral range from 380 nm to 630 nm and was cooled with Peltier module only to 8 • C. This precluded the measurement of photons derived from decay of O 2 ( 1 ∆ g ) formed in Fenton system (reactions 7 and 8). Application of photomultiplier device with much wider spectral range sensitivity and lower working temperature of photomultiplier (e.g., 300 nm to 900 nm and −40 • C, 160 nm to 710 nm and −30 • C) provides the opportunity to measure signals from all possible photon emitters ( 3 (R=C)* and O 2 ( 1 ∆ g )) and ensures low background and high signal-to-noise ratio [2,3]. Therefore, it cannot be excluded that the real UPE from Fe 2+ -EGTA-H 2 O 2 system is higher than we observed under conditions of our experiments. On the other hand, the AutoLumat Plus is a commercially available instrument for chemiluminescent determination of various compounds and enzymes in one batch in a quasi-parallel mode and our experiments could be easily repeated and extended to various Fenton systems by other researches. Moreover, we proposed the quenching of UPE from Fe 2+ -EGTA-H 2 O 2 system as a simple tool for evaluation of antioxidant activity of various phytochemicals. Because the system simplicity and short time of UPE recording the assay is inexpensive, suitable for automation and the obtained results could be easy for interpretation. Therefore, a proven possibility to execute such tests with AutoLumat Plus seems to be the advantage of our study. We did not determine spectrum of light emitted from Fe 2+ -EGTA-H 2 O 2 system and it could also be recognized as the second weakness of our study. However, with the use of various ROS scavengers and by substitution of EGTA with EDTA we were able to conclude that the photons emitters are 3 (R=C)*. Moreover, by substitution of Fe 2+ with other divalent cations we proved that UPE emission within the range from 380 nm to 630 nm is specific for Fe 2+ -EGTA-H 2 O 2 combination. The ratio of molar concentrations of FeSO 4 to EGTA to H 2 O 2 in Fenton reaction system was 1:2:28.1. In our previous studies this reaction mixture generated large amounts of • OH as reflected by damage to deoxyribose [28,42] and the cytotoxicity against cell suspensions in vitro [43]. Therefore, we analyzed the relationship between UPE and increasing concentrations of Fenton system under conditions of stable ratio of molar concentrations of its components. The maximal photon emission from Fenton system occurred during the first several dozen seconds after addition of H 2 O 2 to Fe 2+ solution and then was terminated [8]. Therefore, we measured UPE from Fe 2+ -EGTA-H 2 O 2 system for 120 s and did not analyze the kinetics of this phenomenon.

Reagents
All chemicals were of analytical grade. DMSO, D-mannitol, sodium azide (NaN 3  (concentrations of 0.5, 2.5, 5, and 10 mmol/L) and 5 mmol/L solutions of CuSO 4 , MnSO 4 , CoSO 4 were prepared before the assay. To minimize oxidation of Cr 2+ ions, aqueous solution (5 mmol/L) of CrCl 2 was prepared with deaerated water within 1 min before addition to a luminescent reaction mixture.
Working solutions of H 2 O 2 (2.8, 14, 28 and 56 mmol/L) were prepared by dilution of 30% H 2 O 2 solution and the concentration was confirmed by the measurement of absorbance at 240 nm using a molar extinction coefficient of 43.6/mol cm [44]. Stock solution of EGTA (100 mmol/L) was prepared in PBS with pH adjusted to 8.0 with 5 mol/L NaOH and stored at room temperature in the dark for no longer than 3 months. EGTA working solutions (concentrations of 1, 5, 10 and 20 mmol/L) were obtained by appropriate dilution of EGTA stock solution with water. Catalase and SOD were dissolved in PBS to an activity of 10 U/µL. Phenolic acids (ferulic, chlorogenic and caffeic acids) were dissolved in PBS to concentrations of 0.1, 0.2 and 1 mmol/L. Solutions of DMSO (concentrations of 20, 40, 80 and 120 mmol/L), NaN 3 (20 mmol/L), mannitol (20 mmol/L), EDTA (10 mmol/L), citric acid (10 mmol/L) and sodium ascorbate (25 mmol/L) were prepared in PBS freshly before the assay.

Light Emission from Fe 2+ -EGTA-H 2 O 2 System
The chemiluminescence was measured with a multitube luminometer (AutoLumat Plus LB 953, Berthold, Germany) equipped with a Peltier-cooled photon counter (spectral range from 380 to 630 nm) to ensure high sensitivity and low and stable background noise signal. Twenty µL of 10 mmol/L EGTA solution was added to the tube (Lumi Vial Tube, 5 mL, 12 × 75 mm, Berthold Technologies, Bad Wildbad, Germany) containing 940 µL of PBS. Then 20 µL of 5 mmol/L solution of FeSO 4 was added and after gentle mixing the tube was placed in the luminometer chain and incubated for 10 min in the dark at 37 • C. Then 100 µL of 28 mmol H 2 O 2 solution was added by an automatic dispenser and the total light emission (expressed in RLU) was measured for 120 s.  (Table 3). These experiments were also performed with 2-and 10-times lower and 2-times higher concentrations of FeSO 4 , EGTA and H 2 O 2 (the ratio of molar concentrations of compounds was always the same).

Effect of Iron and EGTA Replacement by other Divalent Cations and Metal Chelators on Light Emission from Fe 2+ -EGTA-H 2 O 2 System
In these experiments we checked whether replacement of Fe 2+ and EGTA with other divalent cations (Cr 2+ , Co 2+ , Mn 2+ , Cu 2+ ) and metal chelators (EDTA and sodium citrate) can change the light emission from the Fe 2+ -EGTA-H 2 O 2 system. The Fenton system was 92.6 µmol/L Fe 2+ -185.2 µmol/L EGTA-2.6 mmol/L H 2 O 2 and Fe 2+ was replaced by the same concentrations of the aforementioned cations. In another series of experiments EGTA was replaced by the same concentrations of EDTA or sodium citrate and the total light emission was measured as described. The design of these experiments together with appropriate controls is shown in Table 3.

Determining the Effect of Reactive Oxygen Species Scavengers and Selected Phenolic Acids on Light Emission from Fe 2+ -EGTA-H 2 O 2 System
To determine what ROS are involved in the UPE of 92.6 µmol/L Fe 2+ -185.2 µmol/L EGTA-2.6 mmol/L H 2 O 2 system, 20 µL solution of ROS scavenger was added to the luminometer tube containing FeSO 4 and EGTA in PBS and incubated for 10 min at 37 • C in the dark and then 100 µL of H 2 O 2 solution was injected and the total light emission was measured for 2 min.  Table 4).
The following ROS scavengers were used SOD-an O 2 · − scavenger (final activity of 0.185 U/µL), catalase-an H 2 O 2 scavenger (final activity of 0.185 U/µL), DMSO-a potent • OH scavenger [45] (final concentrations of 0.37 mmol/L to 2.22 mmol/L), mannitol-an • OH scavenger [46] (final concentration of 0.37 mmol/L), NaN 3 -a O 2 ( 1 ∆ g ) scavenger [47] (final concentration of 0.37 mmol/L) and sodium ascorbate (final concentration of 0.46 mmol/L). In another series of experiments the effects of three phenolic acids (ferulic, chlorogenic and caffeic acids, final concentrations in the reaction mixture from 0.09 mmol/L to 0.87 mmol/L) on total light emission from Fe 2+ -EGTA-H 2 O 2 system were studied. The design of these tests and control samples were the same as in the case of ROS scavengers (Table 4). In each series of experiments (repeated at least four times) one ROS scavenger or one phenolic acid was tested. The inhibitory effect of ROS scavengers or phenolic acids on light emission was expressed as percent inhibition (  E-Solution of ROS scavenger in PBS (10 U/µL SOD, 10 U/µL catalase, 20 mmol/L mannitol, 20 mmol/L NaN 3 , 25 mmol/L sodium ascorbate or 20 mmol/L to 120 mmol/L DMSO). Then after gentle mixing the tube was placed into luminometer chain, incubated for 10 min at 37 • C and then 28 mmol/L H 2 O 2 (F) or water (G) was automatically injected with dispenser and total light emission was measured for 2 min.

Statistical Analysis
Results (total light emission or % inhibition of light emission) were expressed as mean (standard deviation) and median and interquartile range (IQR). The comparisons between total light emission from Fe 2+ -EGTA-H 2 O 2 system and light emission from corresponding samples of modified system (e.g., incomplete system, system with addition of ROS scavengers or phenolic acids, system based on other divalent cations or chelating agents, medium alone) were analyzed with independent-samples (unpaired) t-test or Mann-Whitney U test depending on data distribution which was tested with Kolmogorov-Smirnov-Liliefors test. The Brown-Forsythe test 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, the Welch's t-test was used instead of the standard t-test. The comparisons of % inhibition of light emission caused by ROS scavengers and phenolic acids were performed in the same way. A p value < 0.05 was considered significant.

Conclusions
We found that Fenton system composed of Fe 2+ -EGTA-H 2 O 2 emits light within the range from 380 nm to 630 nm. The UPE of Fe 2+ -EGTA-H 2 O 2 depends upon • OH and O 2 − and could be attributed to • OH-induced cleavage of ether bond in the backbone structure of EGTA with consequent formation of 3 (R=C)* and photons emission. Plant phenolic acids with known antioxidant properties (caffeic, chlorogenic and ferulic acids) and ascorbic acid significantly quenched UPE what suggests possible application of this phenomenon as the assay for evaluation of antioxidant activities of various phytochemicals. However, further studies involving optimization of the Fenton system parameters, time of UPE measurement, control tests with other types of antioxidants and validation are necessary before the successful development of this assay.
contributed reagents, materials and analysis tools. Michal Nowak, Wieslaw Tryniszewski and Dariusz Nowak wrote the paper.

Conflicts of Interest:
The authors declare no conflict of interest. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.