Computational Insights into the Regeneration of Ovothiol and Ergothioneine and Their Selenium Analogues by Glutathione.

Ovothiol and ergothioneine are powerful antioxidants that readily react with oxidants by forming their respective disulfides. In fact, ovothiol is widely considered one of the most powerful natural antioxidants. However, for these antioxidants to be again involved in reacting with oxidants, they must be regenerated via the reduction of the disulfide bonds. In the present work, the regeneration of the antioxidants ovothiol and ergothioneine and their selenium analogues, by the closed-shell nucleophilic attack of glutathione, was investigated using density functional theory. From the calculated thermodynamic data, the attack of glutathione on OSSO and EYYE (where Y = S and/or Se) will readily occur in solution. Moreover, in comparison to the reference reaction GSH + GSSG → GSSG + GSH, all reactions are expected to be faster. Overall, the results presented herein show that the key antioxidant GSH should readily recycle ovothiol, ovoselenol, ergothioneine, and ergoseloneine from OYYO and EYYE (where Y = S and/or Se).

Introduction

Reactive oxygen species (ROSs) have important roles within our bodies. Under normal physiological conditions, ROSs are constantly being produced during metabolic processes.[1] ROSs are also produced pathologically in many ways, for example, from exposure to ionizing radiation and the presence of free-metal ions.[2−4] It is when the concentration of ROSs become too high that negative effects on DNA, proteins, and lipids occur, resulting in oxidative stress that leads to mutations, misfolding of proteins, and possibly cell death.[5−10] Notably, oxidative stress has been shown to be linked to many diseases such as cancer, Alzheimer’s disease, Parkinson’s disease, and type-2 diabetes.[6,11−15]

To combat the harmful effects of ROSs, cells use antioxidants to mitigate, repair, or prevent oxidative damage. Antioxidants may directly interact with ROSs or the molecules that produce ROSs, such as free-metal ions, thus preventing the creation of ROSs.[16−19] One of the major antioxidants in the human body is glutathione (Scheme a), a cysteine-based tripeptide, with a reactive thiol group. To prevent oxidative damage by ROSs, glutathione, undergoes a one-electron oxidation, resulting in a short-lived sulfur radical species that readily joins with another glutathione radical species to form a stable disulfide. The disulfide form of glutathione is then readily acted upon by glutathione reductase via a nucleophilic disulfide exchange reaction to regenerate an active glutathione molecule.[20] In addition to reacting with ROSs, glutathione is known to assist in the regeneration of other antioxidants, such as vitamin E.[21] Although glutathione is one of the most ubiquitous antioxidants, there are many other antioxidants that exist in nature. Two such antioxidants are the mercaptohistidine derivatives, ovothiol and ergothioneine.

Scheme 1

An Illustration of (a) Glutathione, (b) Ovothiol (OvothiolA, n = 0; OvothiolB, n = 1; and OvothiolC, n = 2), and (c) Ergothioneine, in Their Dominant Form at Biological pH

Ovothiol is widely considered one of the most powerful natural antioxidants, and it exists in three naturally occurring forms depending on the degree of methylation of the alpha-amino nitrogen (Scheme b).[22] However, as discussed by de Luna et al.,[23] past experimental work has shown that the deletion of the amino acid portion of ovothiol had little effect on its chemistry and therefore the oxidative power of ovothiol is a result of the mercaptohistidine functional group.[24,25] While ovothiol is not naturally synthesized in humans, it can be taken in through our diet. Given its antioxidant strength, ovothiol and its analogues have been gaining increasing attention as potential therapeutics for use in humans. Indeed, OvothiolA has been reported to induce autophagy in liver carcinoma cell lines, suggesting a potential role in regulating cancer cell growth.[26,27]

First discovered in ergot fungus, ergothioneine (Scheme c), like ovothiol, is not naturally produced in humans but is taken up through our diets. An important source of ergothioneine is mushrooms, where it is proposed to have an anti-oxidative and anti-inflammatory role.[28,29] In humans, the exact role of ergothioneine is unknown, although in animal trials, it has been found in high concentrations in tissues and cells that experience high oxidative stress, such as the liver, erythrocytes, eye cornea, and kidneys.[30−34] It is noted that aerosols containing ergothioneine have been developed to treat chronic inflammatory diseases such as asthma.[25,35,36] At biological pH, ergothioneine predominantly exists in its thione tautomer form, making ergothioneine very stable and resistant to auto-oxidation and a slight resistance to form disulfide.[37] Like ovothiol, past experimental work has shown that the deletion of the amino acid portion of ergothioneine has little effect on its chemistry.[24,25]

Sulfur appears in group 16 of the periodic table, along with the other chalcogens. Selenium lies directly below sulfur on the periodic table, resulting in sulfur and selenium sharing similar chemical and physical properties.[38,39] Like sulfur, selenium has the ability to act as an antioxidant, though a mechanism by which selenium prevents oxidative damage can vary from that of sulfur since selenium’s reductive ability is stronger than that of sulfur.[40−42] However, while the two elements share many similarities, there are some important differences.[43] For instance, thiol/disulfide exchange reactions are accelerated in solution with the replacement of S by Se because of selenolate being both a better nucleophile and a better leaving group than thiolate.[44,45] In fact, it has been stated that almost all chemical reactions involving Se are faster than the analogous reactions involving sulfur.[43] In previous studies,[23,46] the ability of ovothiol, ergothioneine, ovoselenol, and ergoseloneine to prevent the (CuII/CuI) redox cycling in solution was investigated. It is noted that CuI ions catalyze the Fenton-type reaction with H2O2, resulting in the formation of a hydroxide and hydroxyl radical. The results of these past studies found that, in the case of ovothiol and ergothioneine and their selenium analogues, the reduction of CuII to CuI with the concomitant formation of the disulfide or diselenide was thermodynamically favorable.[23,46] It was concluded that ovothiol, ergothioneine, ovoselenol, and ergoseloneine are able to prevent (CuII/CuI) redox cycling and are therefore suitable for the protection of copper-induced oxidative damage.[23,46] However, with the formation of the disulfide and diselenides, it is unclear if the reduced forms of these powerful antioxidants would be regenerated in vivo.

Previous experimental and theoretical work has investigated the dichalcogenide exchange reaction for the nucleophilic attack of a thiolate or selenite on a di-chalcogen bond, in both the gas phase and condensed phase.[45,47−51] However, the past theoretical work used small models, consisting of a methyl or hydrogen as a side chain on the chalcogen.[49] Thus, it is unclear how the presence of the imidazole rings of ovothiol, ergothioneine, and their selenium analogues will have an effect on the kinetics and thermodynamics of the substitution reactions. Herein, the thermodynamics and kinetics of the nucleophilic attack of glutathione to regenerate ovothiol, ergothioneine, and their selenium analogues was investigated using density functional theory (DFT). To the best of our knowledge, this work represents the first investigation of the regeneration of the antioxidants ovothiol and ESH or their selenium analogues by glutathione.

Methods

As discussed above, because past experimental work[24,25] has shown the deletion of the amino acid portion of ovothiol and ergothioneine has little effect on their chemistry, the molecules ovothiol, ergothioneine, ovoselenol, and ergoseloneine have been modeled as 4-thiol-N1-methyl-5-methylimidazole, 4-selenol-N1-methyl-5-methylimidazole, 2-thione-4-methylimidazole, and 2-selone-4-methylimidazole, respectively, to reduce computational costs. For simplicity, they will be referred to as OSH, OSeH, ESH, and ESeH, respectively. In the case of glutathione, to reduce computational costs it has been modeled as methane thiol and will henceforth be referred to as GSH. The disulfide, diselenide, and mixed sulfide selenide molecules will be referred to as OYYO and EYYE, where Y = S or Se. The chemical reactions investigated herein are given in Table .

Table 1

Dichalcogenide Exchange Reactions Investigated in the Present Worka

reaction		Y1	Y2
1	GSH + GSSG → GSSG + GSH	N/A	N/A
2	GSH + OSSO → GSSO + OSH	S	S
3	GSH + OSSeO → GSSO + OSeH	S	Se
4	GSH + OSeSO → GSSeO + OSH	Se	S
5	GSH + OSeSeO → GSSeO + OSeH	Se	Se
6	GSH + ESSE → GSSE + ESH	S	S
7	GSH + ESSeE → GSSE + ESeH	S	Se
8	GSH + ESeSE → GSSeE + ESH	Se	S
9	GSH + ESeSeE → GSSeE + ESeH	Se	Se
10	GSH + GSSO → GSSG + OSH	S	N/A
11	GSH + GSSeO → GSSG + OSeH	Se	N/A
12	GSH + GSSE → GSSG + ESH	S	N/A
13	GSH + GSSeE → GSSG + ESeH	Se	N/A

The Gibbs reaction energies (ΔrG°) and Gibbs activation energies (ΔrG‡), obtained at the SMD-M06-2X/aug-cc-pVTZ//SMD-M06-2X/aug-Cc-pVDZ level of theory, for these reactions are shown in Figures –5. The labels Y1 and Y2 are used in Figures –5.

Figure 1

Gibbs energy surface for Reaction 1 (Table ). Energies obtained at the SMD-M06-2X/aug-cc-pVTZ//SMD-M06-2X/aug-cc-pVDZ level of theory. For the deprotonation in the first step and protonation in the last step the Gibbs energies from Table were used.

Figure 5

Gibbs energy surface (at pH = 0) for Reactions 12 and 13 (Table ) obtained at the SMD-M06-2X/aug-cc-pVTZ//SMD-M06-2X/aug-cc-pVDZ level of theory. For the deprotonation in the first step and protonation in the last step, the Gibbs energies from Table were used.

All calculations were done using the Gaussian 09 and Gaussian 16 software suites.[52,53] Optimized geometries and harmonic vibrational frequencies were obtained at the SMD-M06-2X/aug-cc-pVDZ level of theory, where water was chosen as the solvent.[54−58] The use of SMD continuum method has been shown to be a valid approach to optimize structures in the condensed phase.[59] All calculations were done using the keyword integral = grid = ultrafine. For all disulfide exchange reactions, the transition states (TSs), a relaxed scan approach, was used where the GS–...YEYE, GS–...YOYO, GS–...SMeYE, and GS–...SMeOE distances were initially set to 3.8 Å. As the nucleophilic GS– approached the electrophilic Y, if a maximum in energy was observed in the PES, the structure at this maximum in PE was used as a starting guess for a TS. The standard Berny algorithm was used to optimize the TS. The nature of the TS was visually confirmed via the visualization of a single imaginary frequency using GaussView 5.[60] The optimized geometries of all complexes are provided in the Supporting Information (Table S1). IRC calculations in the forward and reverse direction were done at the SMD-M06-2X/aug-cc-pVDZ level of theory to confirm nature of the TSs. For the IRC calculations, 20 steps were followed in both the forward and reverse directions. Following the 20 steps in the forward and reverse directions, the SMD-M06-2X/aug-cc-pVDZ level of theory was then used to fully optimize the reactant and product complexes. The xyz coordinates of the reactant and product complexes are provided in Table S2. Notably, the IRC calculations confirm that the TSs obtained using the standard Berny algorithm are the TSs for the disulfide exchange reactions discussed herein.

As noted in the Introduction, previous work has investigated the reaction mechanisms that involve S or Se acting as the nucleophile, electrophile, and/or leaving group, where the moiety attached to the chalcogen atom was a hydrogen atom or methyl group.[49] From the work, it was found that, in solution, the substitution reaction can occur via two mechanisms. The first mechanism is a one-step reaction (i.e., SN2-type mechanism), whereas the other mechanism is a two-step process.[49] Regarding the SN2-type mechanism, a single TS exists for the disulfide exchange reaction. The second mechanism, referred to as an addition–elimination mechanism, involves the formation of a hypervalent intermediate (i.e., a sulfur with more than four electron groups surrounding it). However, from the past work, it was concluded that, in solution, the SN2 mechanism is more likely.[49] It is noted that in the present investigation, attempts to find a stable hypervalent sulfur/selenium intermediate failed at the SMD-M06-2X/aug-cc-pVDZ level of theory. Thus, for the disulfide exchange reactions involving OYYO and EYYE, only a single TS, corresponding to the backside attack of the electrophile by the nucleophile, characteristic of a classic SN2 type reaction, was found. Regarding the SN2-type TS, it can proceed via a syn- or anti-conformation of the nucleophile and the leaving group.[49] However, past evidence has shown little difference in the energy of the syn- or anti-TS for the reaction mechanisms regardless of whether sulfur or selenium is acting as the nucleophile, electrophile, and/or leaving group.[49] Given that the past work only looked at systems where the moiety attached to the sulfur/selenium was a methyl group or hydrogen atom, it is expected that the larger and bulkier side group on the electrophile and leaving group would result in the syn-conformers to experience enhanced steric repulsion. Thus, for the reactions discussed herein, the anti-TSs of all disulfides, diselenides, and mixed sulfide–selenides were solely investigated.

Single point energies were obtained at the SMD-M06-2X/aug-cc-pVTZ//SMD-M06-2X/aug-cc-pVDZ level of theory.[58] Gibbs reaction and activation energies were determined by correcting the single point energies by adding the Gibbs energy correction (ΔrG°corr) obtained from the harmonic frequency calculations. Per the work by Galano and Alvarez-Idaboy,[61] the ΔrGs° values were corrected to a standard state of 1 M. The solvent cage effects on the enthalpy and entropy of activation have been included according to the corrections proposed by Okuno that consider the free volume theory.[62,63]

Implicit solvation models do not properly treat hydrogen bonding interactions between solvent and solutes. In aqueous solution, this can result in large errors in the calculated values such as pKas.[64] A number of studies have shown that inclusion of explicit waters to account for hydrogen bonding between solute and solvent improves the calculation of pKas, especially for processes involving ionic species, which are generally prone to more error.[64−74] A previous study demonstrated that for thiols the inclusion of explicit waters near the sulfur was essential for obtaining reasonable pKas when using implicit solvation models to treat effects of the bulk solvent on the solute.[64] Specifically, it was found that when three explicit water molecules were included to hydrogen bond to the acidic sulfur, good agreement between calculated and experimental pKa values was achieved.[64] It was concluded that regarding SMD, the lack of short-range hydrogen bonding interactions is a major factor in the poor performance for calculating thiol pKas.[64] A past study showed that when calculating the pKas of thiols, the wB97XD DFT functional in combination with a basis set includes polarization functions on the hydrogens and diffuse functions on the heavy atoms is required.[64] Thus to calculate all Gibbs deprotonation energies, the wB97XD/aug-cc-pVDZ level of theory was used for geometry optimizations, calculation of Gibbs corrections and electronic energies. The Gibbs reaction energies for the deprotonation of GSH, OSH, ESH, OSeH, and ESeH were calculated, where the standard chemical potential of a proton in a dilute aqueous environment (μ298K0(H+)) was chosen to be −1130.5 kJ mol–1 at a pH of zero.[64]

Results and Discussion

As done in previous work[64] we have included three explicit water molecules that forms hydrogen bonds with both the thiol and the thiolate of methane thiol/thiolate. The water molecules were arranged such that the total number of hydrogen bonds were the same for the thiol and the thiolate. This was done to ensure that the calculated energy difference is a result of the change in the strength of hydrogen bonds and not from the change in the number of hydrogen bonds.[64] Given that the thione tautomer form of ESH is favored over the thiol form, it is the imidazole ring involved in gaining and losing a proton. Thus, in the case of ESH and OSH due to the imidazole ring, an additional water molecule was included to a total of four water molecules.

The commonly accepted pKa for methanethiol is 10.4, whereas for ovothiol, a pKa of 6.7 was used for the SH group.[75] For ESH, a value of 11.2 was used for the deprotonation of the imidazole nitrogen.[76] These values correspond to Gibbs deprotonation energies at the standard pH of 60.0, 38.2, and 78.6 kJ mol–1. As seen in Table the calculated Gibbs deprotonation energy for GSH is in excellent agreement with experiment with an unsigned error of only 0.6 kJ mol–1. For OSH and ESH the calculated values are on an average only 9.0 kJ mol–1 more endergonic and are therefore in reasonable agreement with experiment. It is noted that in the absence of any explicit water molecules, the calculated Gibbs deprotonation energies for GSH, OSH, and ESH are on an average 36.8 kJ mol–1, more endergonic than the experimental Gibbs deprotonation energies. Thus, given the reasonable agreement between calculated and experimental Gibbs energies for deportation of GSH, OSH, and ESH, and given the fact that experimental Gibbs deprotonation energies for the OSeH and ESeH are unavailable we have decided to use the calculated energies in Table for the Gibbs energy surfaces shown in Figures 2345.

Table 2

Gibbs Reaction Energies (ΔrG°) at pH = 0 for the Deprotonation of GSH, OSH, OSeH, ESH, and ESeH Obtained at the SMD-M06-2X/aug-cc-pVTZ//SMD-M06-2X/aug-cc-pVDZ Level of Theorya

reaction	calculated ΔrG°	experimental ΔrG°
GSH → GS– + H+	60.0	59.4b
OSH → OS– + H+	49.3	38.2c
OSeH → OSe– + H+	32.0	N/A
ESH → ES– + H+	70.8	64.0d
ESeH → ESe– + H+	76.4	N/A

All energies are in kJ Mol–1. For the calculation of the Gibbs deprotonation energies μ298K0(H+) was chosen to be −1130.5 kJ Mol–1 (see Computational Methods).[64]

ΔrG° for deprotonation of GSH was determined from the commonly accepted experimental pKa value of 10.4 for methanethiol.

ΔrG° for deprotonation of OSH was calculated from experimental pKa value of 6.7.[75]

ΔrG° for deprotonation of ESH was determined from experimental pKa value 11.2.[76]

Figure 2

Gibbs energy surface for the attack of CH3SH to OY1Y2O to form CH3SY1O + OY2H (Table , Reactions 2–5) at pH = 0. Energies obtained at the SMD-M06-2X/aug-cc-pVTZ//SMD-M06-2X/aug-cc-pVDZ level of theory. For the deprotonation in the first step and protonation in the last step the Gibbs energies from Table were used.

Figure 3

Gibbs energy surface for the attack of CH3SH to EY1Y2E to form CH3SY1E + EY2H (Table , Reactions 6–9) at pH = 0. Energies obtained at the SMD-M06-2X/aug-cc-pVTZ//SMD-M06-2X/aug-cc-pVDZ level of theory. For the deprotonation in the first step and protonation in the last step the Gibbs energies from Table were used.

Figure 4

Gibbs energy surface (at pH = 0) for Reactions 10 and 11 (Table ) obtained at the SMD-M06-2X/aug-cc-pVTZ//SMD-M06-2X/aug-cc-pVDZ level of theory. For the deprotonation in the first step and protonation in the last step, the Gibbs energies from Table were used.

Thermodynamics and Kinetics for the Attack of GSH to OYYO and EYYE

To better understand the effect of the imidazole ring on the thermodynamics and kinetics of the substitution reaction, the Gibbs activation energy (ΔrG‡) for the reference reaction GSH + GSSG → GSSG + GSH was calculated (Figure ). For the reference reaction ΔrG‡ was calculated to be 115.3 kJ mol–1 under standard conditions. Notably, the value of ΔrG‡ as shown in Figure is the largest Gibbs activation energy and is therefore predicted to be the slowest disulfide exchange reaction investigated herein. Given that the reactants and products are the same in Reaction 1, ΔrG° = 0.0 kJ mol–1.

As seen in Figures and 3 the dichalcogenide exchange reactions involving OSSO and ESSE are predicted to occur with a value of ΔrG‡ less than that seen for the reference Reaction 1 (Table ). Regarding the disulfide OSSO, the value of ΔrG‡ for the substitution (i.e., Reaction 2) is approximately 10.5 kJ mol–1 lower in energy than that for Reaction 1. For ESSE (i.e., Reaction 6), ΔrG‡ is 37.5 kJ mol–1 lower in energy than the value for Reaction 1. Thus, it is expected that in solution the rate at which GSH attacks OSSO and ESSE would be greater than that seen for the attack of GSSG.

For OSSeO, the substitution of the S in the Leaving Group (LG) with Se (i.e., Reaction 3) accelerates the substitution further, where ΔrG‡ is 15.8 kJ mol–1 lower in energy than that for Reaction 1. Similarly, for ESSeE, the substitution of the S in the LG with Se (Reaction 7) further accelerates the dichalcogenide exchange reaction, where ΔrG‡ was calculated to be 44.6 kJ mol–1 lower than that for Reaction 1. As noted above for Reactions 4, 5, 8, and 9, a classic SN2-type TS could not be found at the present level of theory. However, considering the reduction in ΔrG‡ for Reactions 2, 3, 6, and 7 relative to Reaction 1 (Table ), the presence of the imidazole ring results in the acceleration of the dichalcogenide exchange reaction. Additionally, comparing Figures and 3, the reactions involving EYYE have Gibbs activation energies that are considerably less than the Gibbs activation energies for the reactions involving OYYO. A partial explanation for the difference in activation energies may be a result of ES– (and subsequently ESH) predominantly existing in its thione tautomer form where cleavage of the dichalcogenide bond results in the negative change of EY– being delocalized into the imidazole ring rather than being localized on the chalcogen atom like that seen for OS–.

For the optimized TSs (Table S1), the average distance between the electrophilic chalcogen atoms and nucleophilic CH3S– was calculated to be 2.84 Å, with a standard deviation of 0.09 Å. The average angle of attack of the nucleophile was calculated to be 172.0°, with a standard deviation of 0.9°. The average dihedral angle between the leaving group and nucleophile (Figure S1) was found to be 127.5°, with a standard deviation of 17.6°. It is noted that because of the hydrogen bond seen in the EYYE system the dihedral angle is approximately 35° smaller than that seen for the OYYO systems. With an average dihedral angle of 127.5° it is apparent that the TSs for dichalcogenide exchange reaction exists in an anti-conformation. Overall, the substitution of the sulfur by a selenium has a marginal effect on the structure of the classic SN2 type TS regardless of choice of chalcogen atom.

From the thermodynamic data provided in Figure , it can be seen that with protonation of OS– and OSe– the attack of GSH on OYYO is exergonic. However, the reaction is most exergonic when Y2 = Se (i.e., Reactions 3 and 5). From Figure , the reactions where GSH attacks EYYE are all considerably exergonic at the present level of theory. Overall, the reactions involving EYYE are generally 66.3 kJ mol–1 more exergonic than the analogous OYYO reactions.

Thermodynamics and Kinetics for the Attack of GSH to GSYO and GSYE

With the formation of GSYO and GSYE, the attack of GSH to form GSSG, OYH, and EYH (Table , Reactions 10–13) was investigated. For the optimized TSs in Figures and 5, the average distance between the electrophile and nucleophile was calculated to be 2.750 Å, with a standard deviation of 0.090 Å. The average angle of attack of the nucleophile was calculated to be 175.0°, with a standard deviation of 0.8°. The average dihedral angle between the leaving group and nucleophile (Figure S1) was found to be 123.8°, with a standard deviation of 18.5°.

The Gibbs energy surfaces for Reactions 10–13 are shown in Figures and 5. As seen in Figures and 5, the attack of GSH to GSYO and GSYE occurs with Gibbs activation energies (ΔrG‡) less than that for Reaction 1. Therefore, at the present level of theory it is expected that the attack of GSH to GSYO and GSYE would occur quicker than that for GSSG. These results are in agreement with the past experimental work that has shown that OSH acts in tandem with glutathione, where OSH neutralizes oxidants forming OSSO with GSH then used to maintain ovothiol in the reduced state.[77] Moreover, previous experimental work has shown that glutathione reductase can reduce ESSE, but only with the aid of glutathione.[78] However, looking at Figures and 5, the second disulfide exchange reaction is generally slower than the respective first disulfide exchange reaction. Specifically, comparing Reactions 10 and 2, Reactions 10 and 3, Reactions 12 and 6, and Reactions 12 and 7, the Gibbs activation energies of the second step are on average 8.2 kJ mol–1 more endergonic. From previous experimental work,[77] it was found that at a pH of 7.2, Gibbs activation energies of 73.0 and 80.9 kJ mol–1 are calculated for GSH + OSSO and GSH + GSSO, respectively. In correcting the values to a pH of 7.2 as in Figures and 4 for reactions GSH + OSSO and GSH + GSSO, the Gibbs activation energies were calculated to be 64.7 and 68.9 kJ mol–1, respectively. Thus, agreement between theory and experiment is good. Moreover, in agreement with the experiment, the second step (i.e., attack of GSH + GSSO) is slower than the first step (i.e., attack of GSH + OSSO).[77]

Regarding the thermodynamic values in Figure , it can be seen that only the dichalcogenide exchange reaction between GSH and GSYO is thermodynamically likely to occur when Y1 = S. In the case where Y1 = Se, the overall change in Gibbs energy is +19.8 kJ mol–1. However, the reactions involving GSYE are considerably exergonic. Looking at Figure , the process of dichalcogenide exchange is most thermodynamically favorable when Y1 = Se and occurs with ΔrG° = −45.2 kJ mol–1. When Y1 = S, the value of ΔrG° is −44.8 kJ mol–1. As noted above, the greater exergonicity for the reaction between GSH and GSYE is likely a result of ESH predominantly existing in its thione tautomer form. Given the exergonicity of the reactions involving EYYE, GSH, therefore, has the thermodynamic driving force to attack EYYE to regenerate EYH.

As noted in the Introduction, previous work has shown that for OSH, ESH, OSeH, and ESeH, the reduction of Cu(II) to Cu(I) with concomitant formation of the disulfide is thermodynamically favorable.[23,46] Importantly, the results presented herein show that the key antioxidant GSH can readily attack OYYO and EYYE (where Y = S and/or Se), resulting in regeneration of the powerful respective antioxidants with Gibbs activation energy less than that for the disulfide exchange reaction between GSH and GSSG. However, from the results presented in Figures –5, the dichalcogenide exchange reaction is exergonic under standard conditions for the reaction of GSH with OSSO or EYYE to form the respective products GSSG + 2OSH and GSSG + 2EYH. Therefore, the ability to regenerate OSH and EYH under standard conditions offers a means to mitigate Cu-induced oxidative damage via the formation of the respective OSSO and EYYE. Moreover, given that the human body naturally produces GSH, the formed OSSO and EYYE is reacted upon by GSH to reform the powerful OSH and EYH antioxidants and GSSG. The importance of it is that, in the human body, GSSG is readily acted upon by glutathione reductase to regenerate the active GSH molecule, whereas OSH and EYH are not naturally produced in humans.

Conclusions

Past work has shown that the reduction of Cu(II) to Cu(I) with concomitant formation of the disulfide (OYYO and EYYE (where Y = S and/or Se)) is thermodynamically feasible for OYH, and EYH.[23,46] In the present work, DFT was used to study the regeneration of the antioxidants OYH and EYH (where Y = S or Se), via the nucleophilic attack of GSH on OYYO and EYYE. Importantly at the SMD-M06-2X/aug-cc-pVTZ//SMD-M06-2X/aug-cc-pVDZ level of theory, the key antioxidant GSH will readily attack OSSO and EYYE (where Y = S and/or Se), resulting in the regeneration of the powerful antioxidants OSH and EYH, respectively in an overall exergonic reaction. Thus, the calculated Gibbs activation energies suggest that the attack of glutathione on OSSO and EYYE (where Y = S and/or Se) will readily occur in solution.