Amplification of glutathione-mediated oxidative stress by catalase in an aqueous solution at hyperthermal temperatures

Introduction

Reduced form glutathione (GSH) has been recognized as a major endogenous antioxidant in living cells. GSH is a tripeptide, consisted of glutamic acid, cystein, and glycine. GSH has a thiol on the cysteine moiety and can work as a hydrogen donor, i.e., a direct reductant. Another important biological antioxidative role of GSH is as a coenzyme of glutathione peroxidases (GSH-Px), which catalyzes decomposition of hydrogen peroxide (H₂O₂) to harmless water (H₂O) using GSH, forming oxidized glutathione (GSSG). The reaction catalyzed by glutathione reductase recovers GSH from GSSG using nicotinamide adenine dinucleotide phosphate (NADPH) as a coenzyme. Therefore, it can be stated generally that the GSH protects living cells from oxidative injury directly and/or indirectly.

On the other hand, there are several reports about oxidative action of glutathione and other thiol compounds. In late 1950s, formation of H₂O₂ in solution of glutathione with oxygen was reported.⁽¹⁾ Lehninger and Schneider⁽²⁾ reported that GSH and cysteine cause mitochondrial swelling and that redox state between cytochrome c and oxygen is as a determinant. The participation of thiol groups in the oxidative destruction of hemoglobin was examined and be concern.⁽³⁾ Hoffsten et al.⁽⁴⁾ reported that the GSH-induced mitochondrial swelling is closely associated with lipid peroxidation. The glutathione induced mitochondrial swelling and lipid peroxidation model has been further studied over subsequent several years.^(5,6) In addition, it was reported that the autoxidation of GSH and/or other thiols yields superoxide.⁽⁷⁾ Oxidative actions of GSH and other thiols has been reported not so much but occasionally and continuously.^(8–12)

Recently, temperature dependent generation of superoxide (O₂^•−) in an aqueous solution containing GSH was reported.^(13,14) Incubating a reaction mixture containing 4-hydroxyl-2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPOL) and GSH at 44°C, the TEMPOL shows a steep reduction following a short time delay. This characteristic TEMPOL reduction was halted by adding superoxide dismutase (SOD) to the reaction mixture or bubbling N₂ gas through the reaction mixture. This result suggests that generation of O₂^•− from dissolved oxygen is related to the observed TEMPOL reduction. Importantly, the temperature dependent reaction of GSH suggests variable behavior of GSH depending on the experimental conditions.

A number of researchers may have experiences that the GSH added as an antioxidant sometimes shows unanticipated paradoxical actions in their test tubes, whether they have reported or ostracized their unexpected results in their experiments. The temperature dependent oxidative actions of GSH through O₂^•− formation may be one reason of the paradoxical biochemical behaviors of GSH. The O₂^•− formation can inevitably generate H₂O₂. However, a clear chemical model mechanism of aggressive oxidative action of GSH has not been proposed.

In this paper, GSH-induced O₂^•− generation in hyperthermal temperatures was again confirmed by the nitroblue tetrazolium (NBT) mehod. The NBT method is a conventional assay for determining O₂^•−, which can reduce NBT to make a blue formazan having absorption at 560 nm. A relation of H₂O₂ simultaneously generated in the reaction mixture and effect of catalase was also investigated. Chemical mechanisms and possible harmful effect of the GSH-induced oxidative effect were described.

Methods

Results and Discussion

For comparison with the previous paper,⁽¹⁴⁾ which used a basic reaction mixture formula containing 0.1 mM TEMPOL and 1 mM GSH, at first in this paper, the same formula of reaction mixture except containing 1.0 mM NBT was tested. Incubating the reaction mixture containing TEMPOL, GSH, and NBT at 44°C, the EPR signal of TEMPOL in the reaction mixture was reduced with the characteristic curve shape (Fig. 1A) similar to as shown in the previous paper.⁽¹⁴⁾ Simultaneously, the yellow/orange color of reaction mixture was changed to blue, and absorption at 560 nm increased, however, no increase of absorbance at 560 nm was observed without GSH (Fig. 1B). The increasing of the absorbance at 560 nm was suppressed when the 1.6 U/ml SOD was added in the reaction mixture or when the reaction mixture was bubbled by 0.5 L/min N₂ gas (Fig. 1C). Therefore, the GSH-dependent increasing absorbance at 560 nm at 44°C in this experiment is suggesting generation of O₂^•−.

The source of O₂^•− must be dissolved molecular oxygen (O₂), since N₂ bubbling in the reaction mixture could suppress the reaction (Fig. 1C and 2B). The GSH may give a hydrogen radical (^•H), and the dissolved O₂ could be directly reduced by a ^•H as below to give a hydroperoxyl radical (HO₂^•) [Eq. 1 and 2]. The HO₂^• is in equilibrium with the O₂^•− in aqueous solution [Eq. 3]. Intermediating ^•H and GS^• is in progress. The reactions 1, 2 and 4 may be one-step reaction. The point is that GSH can reduce O₂ temperature dependently.

GSH ⟷ ^•H + GS^• [1]

^•H + O₂ → HO₂^• [2]

HO₂^• ⟷ H⁺ + O₂^•− [3]

Consequently GSSG can result from a reaction between a pair of GS^•.

GS^• + GS^• → GSSG [4]

The active products in this reaction are actually both HO₂^•, which is a relatively strong oxidant, and O₂^•−, which is basically a reductant.

Halliwell⁽¹⁶⁾ proposed a different reaction mechanism of GSH-dependent O₂^•− generation, which is the reduction of O₂ by GSSG^•−. In their reaction, GS^• was initially given by a reaction of GSH and other free radical species such as ^•OH, and then GSSG^•− was given by a reaction of GS^• and GSH. In our reaction system, however, it is quite unlikely that the other free radicals participate the formation of GS^• and/or ^•H.

The actual trigger of the TEMPOL reduction should be one-electron-oxidation of TEMPOL by HO₂^• to give an oxoamonium cation form of TEMPOL. The oxoamonium cation form of TEMPOL then could react with GS^• to make a stable non-paramagnetic complex and as a result TEMPOL was two-electron-reduced.⁽¹³⁾ On the other hand, NBT can be directly reduced by O₂^•− to give blue formazan dye. Strong reductants, such as ascorbic acid, can also reduce NBT and give the blue formazan (data not shown).

Adding NBT in the reaction mixture made the decaying TEMPOL a little faster and the delay time shorter (Fig. 1A). The nitroxyl radical form of TEMPOL can be one-electron-oxidized by HO₂^• to be the oxoamonium cation form. In an opposite way, the oxoamonium cation can be one-electron-reduced by O₂^•− to recover the nitroxyl radical. The reduction of NBT by O₂^•−, which can also be recognized as the consumption of O₂^•− by NBT, may suppress reduction of the oxoamonium cation form to a nitroxyl radical form, and apparently accelerate one-electron-oxidation of TEMPOL to the oxoamonium cation form. As a result, quasi-one-electron-reduction of TEMPOL, i.e., decay of the EPR signal of TEMPOL, could be accelerated. In this case NBT and the oxoamonium cation form of TEMPOL competes each other for O₂^•−. NBT may decrease the population of O₂^•−, but not deplete and allow the presence of HO₂^•. SOD, however, may deplete O₂^•−, then the equiribrium between HO₂^• and O₂^•− [Eq. 3] would strongly move to the right. Therefore, SOD does not allow that HO₂^• could exist. In other words, HO₂^• rapidly changed to O₂^•− before reacting with other molecules, and then was deleted by SOD.

The next experiment tested whether the TEMPOL had any effect on the generation of O₂^•−. When the reaction mixture containing simply GSH and NBT was incubated at 44°C, the similar increasing of absorbance at 560 nm shown in Fig. 1A was observed at almost the same timing (Fig. 2A). However, no increasing absorbance at 560 nm was observed without GSH (Fig. 2A). The increasing of the absorbance at 560 nm was suppressed when the 1.6 U/ml SOD was added in the reaction mixture or when the reaction mixture was bubbled by 0.5 L/min N₂ gas (Fig. 2B). Since reaction profiles in Fig. 1 and 2 can be evaluated almost similar shapes in considering the variability, the TEMPOL in the reaction mixture has no modification as to the generation of O₂^•−. This result certified that the TEMPOL was working just as a redox detector in the previous paper.⁽¹⁴⁾

No increase of absorbance at 560 nm was also observed when GSSG was used instead of GSH (Fig. 2B). This result is probably the same as the previous paper,⁽³⁾ which demonstrated that no TEMPOL reduction was observed when TEMPOL was incubated at 70°C with GSSG instead of GSH, even though the detection method was different to this paper. Oxygen can not extra-oxidize GSSG to make O₂^•− even in a higher temperature. In other words, it can say that GSSG is a stable enough end product of the anti-oxidative process of GSH. In consequence of stability of the GSSG, the enzymatic system is required to recover GSH from GSSG.

Adding 1.6 U/ml SOD or 0.5 L/min N₂ gas bubbling could stop the quasi-one-electron-reduction of TEMPOL, i.e., EPR signal decay of TEMPOL, in the previous paper. However, a slight increase of absorbance at 560 nm, reduction of NBT, was still observed in the experiment with 1.6 U/ml SOD or 0.5 L/min N₂ gas bubbling (Fig. 1C and 2B) in this paper. In the hypoxic condition, GSH could be able to reduce NBT directly rather reduce O₂. Since this direct reduction of NBT by GSH in hypoxic condition looks saturated, the reaction is reversible. The direct reduction of NBT by GSH in the hypoxic condition might be not negligibly slow.

Next, the temperature dependence of the NBT-GSH reaction was tested. The GSH dependent NBT-derived blue formazan generation was exaggerated temperature dependently (Fig. 3). The results of Fig. 2 and 3 suggest that GSH has worked on the temperature dependent generation of HO₂^•/O₂^•−.

When concentration of GSH in the reaction mixture containing 0.1 mM TEMPOL was varied from 0.5–8.0 mM, decay of TEMPOL became slower depending on the GSH concentration (Fig. 4A and B). At 37°C, TEMPOL reduction was almost stopped by 8.0 mM GSH. At 44°C, the effect of GSH concentration was mild compared to the experiment at 37°C. The timing of steep decay of TEMPOL reduction was delayed when GSH concentration increased. GSH itself may able to reduce HO₂^• directly, in other words the HO₂^•, a relatively strong oxidant, can oxidize GSH as [Eq. 5].

2GSH + 2HO₂^• → GSSG + 2H₂O₂ [5]

In short, GSH can consume HO₂^•, and suppress the process of one-electron-oxidation of TEMPOL. GSH and TEMPOL competes with each other for HO₂^•.

In the previous paper,⁽¹⁴⁾ the steep decay of TEMPOL following the initial delay time was shift to the right, i.e., delayed, by adding albumin and/or ascorbic acid. Albumin, which can be readily oxidized has been known as an antioxidative biomolecule. Ascorbic acid has been known as relatively strong reductant, which can be easily oxidized in other words. Both molecules, therfore, can be oxidized by HO₂^•. This can be said that the both molecules can consume HO₂^•. Initial consumption of HO₂^• can delay the reaction and push the slope to the right.

In an interesting twist, the formation of the blue formazan, i.e., absorbance at 560 nm, increased when the GSH concentration increased in the reaction mixture containing NBT (Fig. 5). The maximum level of blue formazan formation with 8 mM GSH at 120 min incubation was similar level in either 37°C or 44°C experiment, although the reaction at 44°C was slightly faster than that at 37°C. However, the maximum level of blue formazan formation with 0.5 or 1.0 mM GSH at 120 min incubation was around 3 times more at 37°C compared to 44°C experiment. In addition, the formation of blue formazan with 8 mM GSH was suppressed only around 10% by adding 3.2 U/ml SOD in the reaction mixture. This is more like direct reduction of NBT by GSH.

Reaction of NBT and O₂^•− may be sufficiently faster than consumption of HO₂^• by GSH. NBT can consume O₂^•− and brake an equivalent balance of the HO₂^•/O₂^•− population. Therefore, HO₂^• must be rapidly changed to O₂^•− before being reduced by GSH. The reduction of NBT can be exaggerated by increasing GSH concentration, which can work not only direct reduction of NBT but also work on increasing O₂^•− production.

Additionally to the [Eq. 5], H₂O₂ could be yielded by the reaction of a pair of HO₂^• [Eq. 6], and then the H₂O₂ could consume the O₂^•−.

HO₂^• + HO₂^• → H₂O₂ + O₂ [6]

H₂O₂ + O₂^•− → OH⁻ + ^•OH + O₂ [7]

The reaction of [Eq. 7] can make a hydroxyl radical (^•OH). This reaction may be harmful due to the high reactivity of the product, i.e., ^•OH, if the reaction occurs in an intracellular space. However, the extracellular generation of ^•OH may not be lethally important, since ^•OH could immediately be deleted by extracellular molecules due to its super-high reactivity and could not reach to the intracellular bio-functional molecules. Membrane permeable HO₂^• and/or H₂O₂ can have a chance to give lethal effects on cells.

To confirm the reaction of [Eq. 5, 6, and 7], the effect of CAT and H₂O₂ was tested. Addition of CAT to the reaction mixture enhanced the maximum amount of the blue formazan formation dose dependently (Fig. 5A). Conversely, addition of H₂O₂ to the reaction mixture suppressed the formation of the blue formazan with dose dependent manner (Fig. 5B). These results suggest that the generation of H₂O₂ in the reaction mixture was expected and then the H₂O₂ is consuming the GSH-induced O₂^•−. The trigger of the GSH-dependent TEMPOL reduction is one-electron-oxidation of TEMPOL by HO₂^•. Catalase could slightly accelerate GSH-dependent TEMPOL reduction but not significantly (data not shown), since elimination of H₂O₂, which is a consumer of O₂^•−, can move the population of [Eq. 3] left.

Fig. 7 shows the result of the cell survival test using mouse thymus cells. Incubating thymus cells at 44°C for 2 h in PBS could cause apoptosis, i.e., cell size shrank. Incubating thymus cells at 44°C for 2 h in PBS containing 1 mM GSH caused higher apoptosis levels. Incubating thymus cells at 44°C for 2 h in PBS containing 1 mM GSH and 1,000 U/ml catalase caused apoptosis more and the percentage of shrunk cells increased.

Fig. 8 shows a schematic drawing of reactions observed in this paper. GSH may reduce O₂ to make HO₂^•, which is equilibrated with O₂^•− in an aqueous condition. TEMPOL or NBT was used as a detector of GSH-induced HO₂^• or O₂^•− generation at hyperthermal temperatures. However, GSH itself can affect the estimation of GSH-induced HO₂^• or O₂^•−, by reacting with NBT or HO₂^•. HO₂^• can be a source of H₂O₂, which can consume O₂^•−. The reactions were actually quite complicated despite the limited number of players; due to the GSH wearing more than one hat. After all, GSH is still a difficult compound which interferes with detecting self-created HO₂^•/O₂^•−.

GSH exists in living tissues in mM level. The in vivo oxygen level can cause GSH-induced TEMPOL reduction,⁽¹⁴⁾ i.e., GSH-induced HO₂^• or O₂^•− generation, at hyperthermal temperatures. This reaction may be a key for hyperthermia effects. The rapid proceeding of the reaction shown in this in vitro experiment may not go in the living cells, since HO₂^•/O₂^•− can not accumulate in living cells. However GSH-induced HO₂^•/O₂^•− may constantly occurred in in vivo situations and it can be accelerated temperature dependently.

The result in this paper suggests that coexisting catalase with GSH in a reaction mixture could exaggerate GSH-induced O₂^•− generation at hyperthermal temperatures. Exaggerated extracellular O₂^•− generation could give a harmful effect to living cells.

Conclusion

Temperature dependent generation of O₂^•− in an aqueous sample containing GSH was again confirmed by the NBT method. GSH may reduce O₂ to make HO₂^•, which is equilibrated with O₂^•− in an aqueous condition. H₂O₂ was also generated simultaneously in this reaction mixture from HO₂^•, and then H₂O₂ consumed O₂^•−. Coexisting catalase in the reaction mixture, as a result, could increase O₂^•− generation. The catalase-exaggerated extracellular O₂^•− generation could give a harmful effect to living cells. This GSH-induced oxidative stress can be a part of mechanisms of hyperthermia therapy.