Unique response of polysulfide in serum albumin to oxidative stress

Abstract

Supersulfide species have been identified as molecules that have catenated sulfur atoms as persulfides (R-SSH) and polysulfides (R-SS_nH). The physiological activities of intracellular supersulfides have been widely studied, but little is known about the role of supersulfides in extracellular biological fluids. We therefore analyzed the pathological changes and oxidative stress responses of the reduced and oxidized forms of polysulfides in the plasma of patients with diabetic nephropathy. We measured the reduced and oxidized forms of polysulfide with elimination of sulfides using dithiothreitol and ascorbic acid plus alkali conditions, respectively. Oxidation decreased oxidized forms of polysulfide and further polysulfide in human serum albumin, increasing its antioxidative activity. We identified seven cysteine residues that have polysulfides in a reduced form after oxidation in albumin. Further oxidation decreased the levels of the reduced forms of polysulfide and the antioxidative effect of serum albumin. Similar changes were also observed in a mouse model of rhabdomyolysis-induced acute kidney injury; the levels of reduced polysulfide in plasma transiently increased 1 h after glycerol administration, which was accompanied by increased antioxidative activity. In the sera of patients with diabetic nephropathy and chronic renal failure, both forms of polysulfides were decreased compared with those of healthy subjects. Analysis of the stages of renal damage revealed that dithiothreitol-liberated polysulfide increased from stages 1–3 and decreased in stage 5. These results suggest that the antioxidative activity of serum, including that of serum albumin, is regulated by the switch from the oxidized form of polysulfide to its reduced counterparts in response to oxidative stress.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-26702-w.

Introduction

Oxidative stress is critically implicated in the pathogenesis and progression of various diseases, including hepatic, renal, metabolic, and neurological disorders^1–5. Assessments of oxidative stress levels in serum are therefore expected to be applied for the diagnosis and prognosis prediction of oxidative stress-related diseases. One conventional method to measure oxidative stress is to quantitatively analyze the oxidation of thiol (SH) in serum proteins^6,7 and glutathione⁸. These markers reflect oxidative stress as the stage progresses^9,10. However, more sensitive oxidative stress markers that detect the early stage of oxidative stress are thought to be essential for the early diagnosis and prevention of diseases.

The relationship between polysulfides and oxidative stress has been widely researched^11–13. Polysulfide is a thiol that binds an additional sulfur called sulfane sulfur, and the molecules that possess polysulfides are called supersulfides¹⁴. Polysulfide groups have significantly greater antioxidative activity than plain thiol groups¹⁵, so polysulfides are assumed to respond more sensitively to oxidative stress. Protein polysulfide prevents thiols from irreversible oxidation by forming perthiosulfonic acid¹⁶. Sulfur species are therefore likely to be involved in redox homeostasis. Polysulfide compounds are considered reactants of hydrogen sulfide and thiol. Hydrogen sulfide (SH⁻) has been found to reduce oxidized thiols and form persulfide^17,18. Moreover, polysulfides have recently been reported to be homeostatically synthesized in general, and cysteinyl-tRNA synthetase (CARS) has been identified as the enzyme responsible for the biosynthesis of cysteine per- and polysulfide^19,20. Interestingly, cysteine per- and polysulfides synthesized by CARS are transported to cysteinyl-tRNA¹⁹, which suggests that proteins already possess polysulfides during translation. A combination of puromycin labeling and alkylating agents revealed that 63% of the cysteine in GAPDH catenated sulfur during translation¹⁹. A growing body of evidence suggests that the S-persulfidation of proteins significantly impacts their function^21–25. Many studies of S-persulfidation have reported that sulfides are ‘modified’ under inflammation or oxidative stress conditions, despite the majority of cysteine residues in proteins obtaining sulfane sulfur during translation. Moreover, the ‘incorporated’ polysulfides during translation have not been well characterized.

A sulfur pool in serum has been proposed^26,27, with the inference that it is derived from CARS. Human serum albumin (HSA) is one of the identified polysulfide-retaining proteins in the blood^28,29. HSA is the most abundant protein in plasma, and it scavenges reactive oxygen and nitrogen species^30–33. The antioxidative mechanism of HSA involves thiols in Cys34, accounting for 80% of the reduced thiol in the blood circulation³¹. Other cysteine residues in HSA form 17 Cys-Cys bonds and play essential roles in maintaining the unique helix-rich structure of HSA. We have established a novel polysulfide assay called the elimination method of sulfide from polysulfide (EMSP), and we have succeeded in quantifying oxidized forms of polysulfides in blood³⁴. We found more than 10 oxidized forms of polysulfides per molecule of HSA, suggesting that additional sulfur atoms exist as polysulfide bonds in some of the 17 Cys-Cys bonds. Conversely, reduced polysulfides detected by alkylating reagents with or without dithiothreitol (DTT) are considered to constitute only a small fraction of the total^18,35. The polysulfide liberated as hydrogen sulfide by a reducing agent such as tris(2-carboxyethyl)phosphine (TCEP) is called ‘bound sulfur’ and is also smaller than the amount of polysulfide detected by EMSP²⁷. We therefore hypothesized that HSA contains polysulfides with different chemical properties, which are reduced (R-SSH) and oxidized (R-SSS-R′) forms of polysulfides.

In this study, we aimed to understand the reaction of polysulfide in serum or HSA to oxidative stress. An overview of the experimental workflow for HSA and plasma assays is shown in Fig. S1. We measured both the reduced and oxidized forms of polysulfides in serum or HSA via a DTT-liberated assay and an EMSP including ascorbic acid, respectively. Mass spectrometry is used to identify the cysteine residue that responds to oxidation in HSA. HSA is prepared according to a reported method so that its level of polysulfide in an oxidized form decreases without changing cysteine bonds via β-nicotinamide adenine dinucleotide (NADH)³⁶, and here, we tested the effect of oxidation on polysulfides in β-NADH-treated HSA. Furthermore, we analyzed polysulfide changes in acute kidney injury (AKI) model mice, patients with diabetic nephropathy, and patients with chronic kidney disease (CKD).

Results

Effect of oxidation on HSA polysulfides and its antioxidant activities

Measurement of polysulfide content via methylene blue assay

A high concentration of ascorbic acid in the base eliminates reduced polysulfide in its oxidized form to sulfides, and the eliminated sulfide can then be analyzed via a methylene blue assay³⁴. This elimination method of sulfide from polysulfide, the EMSP method, is limited because of the risk of being affected by disulfide bonds that can produce dehydroalanine and sulfides under alkaline conditions³⁷. However, it is thought to be the only method that can be used to quantify the oxidized form of polysulfide in heterogeneous samples. We confirmed that the number of thiols present after the reaction with a thiol-free reductant, tris(2-carboxyethyl)phosphine (TCEP), did not change during EMSP³⁴. Dithiothreitol (DTT) combined with a methylene blue assay was specifically used in this study to detect a reduced form of polysulfide. To determine whether DTT reduces polysulfide in a reduced form, HSA was persulfidated with disodium sulfide as previously reported as a model of polysulfide in a reduced form^18,28. Cys34, which has a free thiol group in HSA, is oxidized with cystine (1). Oxidized HSA is then reduced by H₂S, and HSA or cysteine is persulfidated (2). Persulfidated HSA (HSA-SSH) and HSA-SH were purified by desalting columns. HSA-SSH was treated with DTT, and the elimination of sulfides was measured via a methylene blue assay. The results showed that 3 mM DTT in 50 mM Tris–HCl (pH 8.0) at 37 °C for 1 h eliminated sulfides from persulfides, whereas sulfide was not detected in untreated HSA under the same conditions (Fig. S2). This reductant condition was therefore applied to the following experiments to measure polysulfide in reduced form.

1

2

Effect of ultraviolet (UV) oxidation on polysulfides in HSA

To evaluate the effects of oxidation to HSA on polysulfide levels and antioxidative activity, HSA was oxidized with UV irradiation, as previously described³⁸. The amount of polysulfides in the oxidized HSA was measured via the methylene blue method with an EMSP and DTT. The levels of polysulfide in the oxidized form increased, whereas the levels of the reduced form of polysulfide increased in response to UV irradiation (Fig. 1A). The antioxidative activity of the HSA oxidized by UV was measured via a DPPH radical scavenging assay, which revealed that the antioxidative activity of HSA increased with increasing amounts of exposed UV (Fig. 1B). A positive correlation was observed between the DPPH radical scavenging activity and the levels of the reduced form of polysulfide (Fig. 1C). To further confirm that polysulfide in a reduced form was produced by the UV irradiation of HSA, DN-TG, a fluorescent probe based on dinitrophenol (DN) and TokyoGreen (TG) that reacts with hydropersulfides (R-SSH)³⁹ was treated with the oxidized HSA. The fluorescence intensity was increased by UV irradiation of HSA, suggesting that polysulfide in a reduced form was generated by oxidation via UV (Fig. 1D). Free hydrogen sulfide was not detected by the methylene blue assay without the EMSP/DTT in these experiments. To determine which cysteine residues in HSA possessed a reduced form of polysulfide after UV radiation, cysteine thiol (Cys-SH) and cysteine hydropersulfide (Cys-SSH) were tagged with β-(4-hydroxyphenyl)ethyl iodoacetamide (HPE-IAM), and the adducts were measured via LC‒MS/MS. Both Cys-S-Tag and Cys-SS-Tag were detected on Cys34 after UV irradiation (Fig. 1E, F). Interestingly, other cysteine residues, such as Cys75, Cys101, Cys124, Cys177, Cys265, and Cys487, which usually form disulfide or polysulfide bonds between cysteines, are tagged by HPE-IAM as Cys-S-Tags and Cys-SS-Tags. These data suggest that the oxidized form of polysulfides in HSA can be converted to the reduced form by oxidative stress caused by UV irradiation.

Fig. 1.

Polysulfide contents in HSA and the antioxidative activity changed by oxidation. (A) HSA was irradiated with a UV lamp at 254 nm (0, 4, 8, 12, or 16 × 10³ mJ/cm²) at 25 °C. DTT- and EMSP-liberated polysulfide in each oxidized HSA sample was detected via a modified methylene blue method. n = 3. (B) Antioxidative activity of oxidized HSA was analyzed via a DPPH radical assay. n = 3. (C) Relationship between the reduced form of polysulfide and its radical scavenging activity. (D) HSA was irradiated with a UV lamp at 254 nm (0, 0.5, 2, or 4 × 10³ mJ/cm²) at 25 °C. The hydropersulfide of each oxidized HSA was detected with DN-TG. The data are presented as the means ± S.D.s. *p < 0.05, **p < 0.01 vs. untreated HSA. n = 3. (E) Cys-SH and (F) Cys-SSH of each oxidized HSA sample were detected via MS analysis via HPE-IAM. Each value was obtained from triplicate determinations.

Response of HSA polysulfide to various oxidative stresses

To investigate whether other methods of oxidation cause similar alterations in the forms of polysulfide, HSA was oxidized with chloramine T (CT), as previously reported ⁴⁰. CT is known to generate nitrogen-centered radicals⁴¹. The reaction with CT showed that the polysulfide liberated by DTT increased fourfold 0.1 h after oxidation. The increase in DTT-liberated polysulfide peaked at 1 h and then decreased with time (Fig. 2A). EMSP-liberated polysulfide decreased with CT in a time-dependent manner (Fig. 2B). The antioxidative activity of the oxidized HSA was assessed via the DPPH radical test. There was an increase in antioxidant activity 1 h after the start of oxidation by CT (Fig. 2C), which subsequently decreased. Metal-catalyzed oxidation (MCO) is a method used to oxidize proteins and peptides to generate superoxide (O₂⁻) and hydrogen peroxide (H₂O₂) via dehydroascorbate^42,43. The oxidation of HSA by MCO resulted in the highest level of DTT-liberated polysulfide at 1 h after the start of oxidation, similar to oxidation by CT (Fig. 2D). The levels of EMSP-liberated polysulfide in HSA did not significantly change in response to MCO after 12 h of reaction (Fig. 2E). The antioxidative activity of HSA increased immediately after oxidation by MCO, with a peak at 1 h, and decreased after 12 h (Fig. 2F). The changes in polysulfide in a reduced form and antioxidant activity were similar for both CT and MCO oxidation.

Fig. 2.

Polysulfide contents in HSA and the antioxidative activity changed by oxidation. (A–C) HSA was oxidized by CT at 37 °C for 0.1, 1, 6, 12, 24, and 48 h and then subjected to dialysis after the reaction. (D–F) HSA was oxidized by MCO at 37 °C for 0.1, 1, 6, and 12 h, and dialysis was performed after the reaction. The data are presented as the means ± S.D.s. *p < 0.05, **p < 0.01 vs. untreated HSA. ^#p < 0.05, ^##p < 0.01 vs. 1 h. n = 3.

Effect of β-NADH on HSA polysulfide

To determine whether the reduced form of polysulfide originates from its oxidized form, we tested whether β-NADH can eliminate oxidized polysulfides. The proposed reaction equation between β-NADH and polysulfide reported by Peng et al. is shown in Fig. 3A³⁶. HSA was reacted with β-NADH for 24 h, and the amount of polysulfide was evaluated after purification by dialysis. The results showed that 2 mol of EMSP-liberated polysulfides per mol of HSA were eliminated by β-NADH (Fig. 3B). By measuring the absorbance of β-NADH at approximately 340 nm, we confirmed that no β-NADH remained in the sample after dialysis (Fig. S3). HSA and β-NADH-treated HSA were oxidized with MCO for 6 h, and the amount of polysulfide was evaluated. The amount of DTT-liberated polysulfide acquired from β-NADH-treated HSA by MCO oxidation was 0.15 mol/mol, whereas that from HSA was 0.3 mol/mol HSA (Fig. 3C). Thus, MCO increased DTT-liberated polysulfides at the same cysteine residues that were targeted by β-NADH, indicating that EMSP-liberated polysulfides removed by β-NADH can serve as a source of DTT-liberated polysulfides.

Fig. 3.

β-NADH treatment decreased the oxidative stress caused by polysulfide in HAS. (A) Reaction between β-NADH and polysulfide proposed by Peng³⁶. (B) β-NADH was reacted with HSA in Tris–HCl buffer (pH 8.0) at 37 °C for 24 h, and HSA was purified by dialysis. The amount of EMSP-liberated polysulfide in HSA and β-NADH-treated HSA was measured via a modified methylene blue assay. The data are presented as the means ± S.D.s. **p < 0.01 vs. HSA. n = 3. (C) β-NADH-treated HSA was oxidized by MCO for 6 h, and the levels of DTT-liberated polysulfide were detected. The data are presented as the means ± S.D.s. **p < 0.01 vs. MCO (-). n = 3.

The structure of HSA was confirmed to be affected by β-NADH treatment. The circular dichroism (CD) spectra in the far-infrared region revealed that secondary structures such as the α-helix content changed in β-NADH-treated HSA (Fig. S4A). CD spectra in the near-infrared region revealed a change in the tertiary structure of HSA in response to the β-NADH procedure (Fig. S3B). Trp214 is the only tryptophan residue in HSA and is known to be sensitive to the structural environment. When HSA is denatured and the interior is exposed, the fluorescence of tryptophan is known to decrease, and its peak shifts to a shorter wavelength^44,45. By measuring the fluorescence of tryptophan in HSA, the tryptophan inside HSA was shown to be exposed on the surface by β-NADH treatment (Fig. S5A). The fluorescence probe of a hydrophobic area called bis-ANS fluorescence in the reaction with β-NADH-treated HSA was greater than that with HSA, suggesting that the hydrophobic surface of HSA was exposed by polysulfide extraction (Fig. S5B). HSA has ligand-binding pockets called site I and site II⁴⁶. We evaluated the drug-binding ability of β-NADH-treated HSA via dansylamide and BD140, which are probes that specifically bind to sites I and II of HSA, respectively⁴⁷. These probes emit fluorescence only when bound to HSA. The results showed that in β-NADH-treated HSA, the fluorescence of dansylamide peaked at 485 nm, which was almost the same as that of HSA, whereas the binding activity of BD140, which peaked at 590 nm, was significantly reduced (Fig. S5C). The polysulfide reducible by β-NADH is implied by these results to be required to maintain the structure of HSA, especially in the vicinity of site II.

Change of small molecule supersulfides in oxidation

Effect of UV irradiation on GSSSG

To investigate the abovementioned oxidation reaction of polysulfide in detail, glutathione trisulfide (GSSSG), a model of polysulfide in its oxidized form, was irradiated with UV light. The obtained reaction product was then examined via HPLC. The retention times and standard curves of GSH, GSSG, and GSSSG are shown in Fig. S6. The retention times differed, so we analyzed the levels of GSH, GSSG, and GSSSG in the samples under these HPCL conditions. After UV irradiation, several peaks, including those of GSH, GSSG, and GSSSG, were confirmed via HPLC (Fig. S7A). The levels of GSSSG decreased with increasing UV irradiation time, but GSH and GSSG were generated (Fig. S7B). Other unidentified peaks were increased by UV irradiation (Fig. S7C-H). Free sulfides in the samples were measured via a methylene blue assay, which revealed that the levels of free sulfides were increased by UV irradiation (Fig. S7I). The DPPH radical scavenging activity increased in the GSSSG samples after UV irradiation (Fig. S8A). The antioxidant activity of each fraction is shown in Fig. S8B. The fractions with peaks at 2.6 (not identified) and 4.0 (GSH) min had radical scavenging activities (Fig. S8CD). The thiol concentrations in each fraction were measured, and thiols were detected in the 2′30″ to 4′20″ fractions (Fig. S8E). We hypothesized that some of the products in the peaks contained polysulfides, so polysulfide detection using the SSP4 probe was performed. Polysulfides were shown to exist in samples 2′30″-3′30″ at approximately 7.2 min (Fig. S8F). Considering that hydroxyl radicals are generated in the solution by UV irradiation⁴⁸, the following reaction is inferred.

3

4

5

6

When the polysulfide in an oxidized form is oxidized, these results suggest that it is cleaved to the reduced form and that the compounds have antioxidant ability. The peaks at approximately 2.6 min and 7.2 min could be candidate hydropersulfide compounds. Further investigations are needed, as compounds that could not be identified in this study remain, and it is also possible that UV light directly reacts with trisulfide bonds to generate thiyl radicals.

Supersulfides in desiase conditions

Changes in polysulfide during acute kidney injury

A mouse model of AKI was used to evaluate polysulfide changes during the acute phase of oxidative stress-related diseases. We evaluated rhabdomyolysis-induced acute kidney injury induced by the intramuscular administration of glycerol. In this model, the osmotic pressure produced by glycerol administration causes myolysis and the release of myoglobin into the blood, and the heme iron of myoglobin results in kidney damage with the induction of oxidative stress. Glycerol was administered to both lower limbs of the mice, and blood and urine were collected at 1, 6, 12, and 24 h after administration (Fig. 4A).

Fig. 4.

Changes in polysulfide and antioxidative activity in AKI model mouse plasma. (A) Fifty percent glycerol or saline (as a control) was i.m. injected into each hind limb. Mouse plasma was collected 1, 6, 12, and 24 h after glycerol administration. (B) The levels of DTT-liberated polysulfide in the plasma of glycerol-induced model mice were measured via DTT-MB. (C) Mouse plasma was coincubated with AAPH radical and linoleic acid. Peroxidized lipids were analyzed by measuring the absorption at 234 nm. (D) Relationship between polysulfide in a reduced form and the radical scavenging activity of plasma in glycerol-induced AKI model mice. **p < 0.01 vs. sham. ^#p < 0.05 vs. 1 h. n = 3. (E) The concentrations of protein in the collected urine from the bladders were diluted 40 times and measured via the Bradford method. The levels of (F) blood urea nitrogen (BUN) and (G) serum creatinine (SCre) were measured after treatment with 50% glycerol. (H) Right kidneys were sampled after perfusion with PBS (upper panels). Histological assessment of the kidneys of glycerol-induced rhabdomyolysis patients. The right kidneys of the mice after the administration of glycerol were stained with hematoxylin and eosin (lower panels). The data are presented as the means ± S.D.s. *p < 0.05 and **p < 0.01 vs. sham. n = 3.

Oxidative stress in the blood peaks at 1 h after glycerol administration, and kidney damage is observed at 23 h⁴⁹. Here, we evaluated changes in the concentration of each type of polysulfide in this mouse model. The amount of DTT-liberated polysulfide in the plasma increased, with a peak at 1 h after glycerol administration, and decreased thereafter (Fig. 4B). The antioxidative activity of plasma was evaluated with a 2,2′-azobis(2-amidinopropane) dihydrochloride (AAPH) radical scavenging assay, as the DPPH assay did not work here because of aggregation due to the presence of ethanol. The antioxidative activity of mouse plasma increased 1 h after glycerol administration and decreased in succession (Fig. 4C). There was a correlation (p < 0.0001) between the radical scavenging activity and the concentration of DTT-liberated polysulfide (Fig. 4D). A similar change in polysulfide levels was observed when a polysulfide probe, SSP4, was used (Fig. S9). The fluorescence of SSP4 peaked at 1 h after the induction of rhabdomyolysis. The degree of renal damage in this model is shown in Fig. 4E–H. Blood urea nitrogen (BUN), serum creatinine (SCre), and organ analysis revealed that the myoglobin released from muscle accumulated in the kidneys 1 h after glycerol administration, causing renal damage. Reddish-brown coloration of the kidney due to myoglobin deposition was observed from 1 h after glycerol administration (Fig. 4H, upper). The results of hematoxylin and eosin staining revealed that renal damage gradually intensified beginning at 1 h after administration, which led to morphological abnormalities in the kidneys at 24 h (Fig. 4H, lower).

Changes in polysulfide levels in human plasma caused by kidney diseases

We also evaluated the oxidative stress response of polysulfides in human plasma. We compared the polysulfide content in the plasma of patients with diabetic nephropathy with stage 1–4 renal damage and patients with CKD with stage 5 chronic renal failure receiving dialysis therapy with that of healthy plasma. The levels of EMSP-liberated polysulfide were significantly lower in both groups of patients (Fig. 5A). Similarly, the level of DTT-liberated polysulfide was significantly lower than that in healthy plasma under both conditions (Fig. 5B). Interestingly, the levels of DTT-liberated polysulfides at each stage of renal injury tended to increase in the early stages of renal injury (Fig. 5C). Furthermore, it decreases in the later stages of renal injury. The difference between acute and chronic renal injury makes it difficult to compare the results, but this trend in reduced polysulfide is similar to the results shown in Fig. 4, which used glycerol-induced AKI model mice.

Fig. 5.

Plasma polysulfide levels are decreased in kidney disease patients. EMSP- (A) and DTT- (B) liberated polysulfide levels in plasma from healthy subjects (n = 3), diabetic nephropathy patients (n = 14), and CKD patients (n = 8–11) were measured via the modified methylene blue assay. The data are presented as the means ± S.D.s. *p < 0.05, **p < 0.01 vs. healthy subjects. (C) The levels of DTT-liberated polysulfide in diabetic nephropathy and CKD patient plasma at different stages.

Discussion

In this study, we detected reduced and oxidized forms of polysulfide via dithiothreitol and ascorbic acid, respectively. In HSA, while the amount of EMSP-liberated polysulfide decreased in response to UV irradiation, the amount of DTT-liberated polysulfide increased (Figs. 1, 2). The amount of EMSP-liberated polysulfide decreased in response to β-NADH, whereas there was no change in the amount of DTT-liberated polysulfide (Fig. 3B). However, the number of DTT-liberated polysulfides increased by MCO in β-NADH-treated HSA was significantly lower than that in HSA (Fig. 3C). Therefore, it was suggested that oxidized forms of polysulfide could be switched to a reduced form of polysulfide by oxidative stress. These conversions of polysulfide were caused by oxidation, which produced free radicals such as ·OH, O₂·⁻, and Cl·, but not by H₂O₂ (data not shown). Therefore, radical cleavage, rather than hydrolysis, is also a possibility, but further evaluation is needed to clarify the mechanisms of switching the types of polysulfides.

The physiological significance of the oxidized forms of polysulfides in HSA and their oxidative stress responses are largely unclear. Here, we revealed that they are involved in the antioxidant mechanism of HSA by switching to reduced forms. The increase in the reduced forms of polysulfides in HSA after oxidation was identified via three methods, namely, methylene blue with DTT, a fluorescent probe, and mass spectrometry, and the levels of the oxidized forms of polysulfide decreased (Figs. 1, 2). We suggest the use of polysulfides in oxidized forms that react with oxidative stress to generate stress-triggered antioxidant-responsive (STAR) polysulfides. When HSA is exposed to oxidative stress, STAR polysulfide is changed to a highly sensitive polysulfide that can be measured via the methylene blue method with DTT and can gain antioxidative capacity. This response of STAR polysulfide could constitute a defense mechanism against oxidative stress. This response of STAR polysulfide to oxidative stress was also observed in a mouse model of AKI (Fig. 4). In the case of human serum, the levels of both polysulfides detected by the EMSP and DTT were decreased in patients with CKD and diabetic nephropathy, suggesting that this decrease resulted from the loss of STAR polysulfide activity due to excessive oxidative stress. Under such conditions, polysulfide is unable to exert its antioxidative activity, and the defense system fails. Monitoring the quantity of polysulfide in plasma would therefore lead to the prediction of oxidative stress status and patient prognosis.

Although it may seem contradictory that oxidation increases antioxidative activity, a similar result was reported by Rafael et al.^38,50. These authors reported that oxidative stress increased the antioxidative capacity of HSA by cleaving disulfide bonds to form SH. However, this was not mentioned from the viewpoint of polysulfides, and further investigations are needed to assess whether such cleavage occurs in disulfide bonds or only in the presence of polysulfide bonds. In our study, EMSP-liberated polysulfide extracted by β-NADH made a high contribution to the acquisition of DTT-liberated polysulfide, suggesting that the polysulfide bond is an essential factor in the increase in antioxidative capacity. Medina-Navarro et al. reported that the percentage of oxidized albumin in plasma was correlated with a better outcome of acute ischemic stroke, and they concluded that this phenomenon reflected antioxidative buffering by HSA⁵¹. We suggest that this defense system may be related to the activity of STAR polysulfide. Recently, polysulfides in a reduced form were shown to be present in multiple cysteine residues in HSA without the need for reductants²⁹. This finding would support the possibility of the presence of STAR polysulfide.

The decrease in the amount of EMSP-liberated polysulfide and the increase in the amount of DTT-liberated polysulfide were almost equal when HSA was exposed to UV (Fig. 1A). The levels of DTT-liberated polysulfides continued to increase during irradiation. Conversely, the amount of reduced polysulfide first increased but then decreased after it was oxidized by CT or MCO (Fig. 2A). Considering this difference, we suggest that tyrosine oxidation induced by MCO may lead to polysulfide loss because the tyrosine of HSA is oxidized by MCO⁵², and tyrosine stabilizes polysulfide in a reduced form⁴⁸. Alternatively, STAR polysulfide may react differently, depending on the type of oxidative stress. The intra-assay CV% ranged from 3.61–10.44% for the measurement of EMSP-liberated polysulfide and was 6.89% for DTT-liberated polysulfides, indicating acceptable reproducibility. However, since the calibration curve was susceptible to oxidation, improved strategies for generating more stable calibration curves will be necessary in future studies.

If STAR polysulfide can be measured more specifically in the future, it is expected to become a sensitive method for detecting oxidative stress. In this study, we determined that Cys75, Cys101, Cys124, Cys177, Cys265, Cys392, and Cys487 (all colored in red in Fig. 6) in HSA have STAR polysulfide (Fig. 1E, F). Fu et al. detected persulfidated cysteines in HSA that reacted with H₂S via MS analysis. Cys34, Cys75, Cys101, Cys124, Cys177, Cys200, Cys245, and Cys392 and Cys487 and Cys567 were identified as sulfidated²⁹. The cysteines overlap well, so the cysteine residues that have STAR polysulfides may easily react with H₂S. Notably, in this study, we could identify only a reduced form of polysulfide after UV irradiation. This was not possible for other types of oxidation, perhaps because of denaturation (data not shown). We found that the extraction of polysulfide by β-NADH markedly decreased the ligand affinity at site II (Fig. S4C) and reduced the number of reduced polysulfides that could be acquired by MCO (Fig. 3). These results indicate that the target cysteine residues of β-NADH and oxidation are partly the same, and it is possible that STAR polysulfide exists in the vicinity of site II in HSA. Ibuprofen, one of the representative drugs that binds to site II in HSA, is shown in green in Fig. 6, and it looks close to some of the cysteine identified in Fig. 1E, F. Additional research is needed to confirm whether the STAR polysulfides are common in oxidized methods.

Fig. 6.

HSA structure with coloring cysteine residues identified to have polysulfides. The HSA structure data were obtained from PDB (ID:2BXG)⁶⁰. The cysteines that were identified as having polysulfide after oxidation by UV in Fig. 1 are colored red. Ibuprofen is colored in green.

The oxidative stress response of HSA to STAR polysulfide is self-sacrificial. While oxidized HSA has antioxidant activity, excessively oxidized HSA, such as advanced oxidized protein products (AOPPs), is eliminated from the bloodstream via scavenger receptors^53–55. AOPP occasionally causes inflammation, and it can be said that there are both merits and demerits^56,57. Supplementation with polysulfides is expected to not only exert an antioxidant effect but also help to preserve STAR polysulfide and prevent HSA from becoming an AOPP. The protective effect of polysulfide against kidney damage has been reported^58,59 and might be applied to other diseases related to oxidative stress. Examining the effect of the polysulfide supply on HSA polysulfide would be an interesting focus of future studies.

This study has several limitations. First, it remains unclear whether STAR polysulfides are unique to HSA or are present in other proteins. The observation that UV-irradiated GSSSG produced GSSH suggests possible generality, but this reaction may be specific to UV conditions and requires further confirmation. Second, the number of patient plasma samples was limited, and validation in larger cohorts of renal disease patients will be essential. Future studies should address these limitations by examining STAR polysulfides in a wider range of proteins and by testing the effects of polysulfide supplementation in larger patient populations.

In conclusion, we found that when HSA and plasma were exposed to oxidative stress, polysulfides changed their forms from reduced to oxidized in the antioxidative response. This finding provides an important basis for elucidating the regulation of the redox state by sulfur species and establishing not only diagnostic markers but also novel preventive and therapeutic methods for treating oxidative stress-related diseases.

Materials and methods

Materials

Na₂S and Na₂S₄ were purchased from DOJINDO Chemical Laboratory (Kumamoto, Japan). Human serum albumin and N,N-dimethyl-p-phenylenediamine (DPDA) were obtained from Sigma Aldrich (St. Louis, MI, USA). Spectra pore 7 (MWCO: 3,500) as a sulfur-free dialyzed membrane was purchased from Repligen (Waltham, MA, USA). Ascorbic acid, DTT, chloramine T (CT), zinc acetate dihydrate, 2,2′-azobis(2-amidinopropane) dihydrochloride (AAPH), iron (III) chloride, LabAssay™ Creatinine, and linoleic acid were purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). A BCA protein assay kit and a BUN colorimetric detection kit were obtained from Thermo Fisher Scientific (Waltham, MA, USA). An ultraviolet transfer microplate was obtained from Zell-kontakt GmbH (Hardenberg, Germany). β-(4-hydroxyphenyl)ethyl iodoacetamide (HPE-IAM) was purchased from Molecular Biosciences (Boulder, CO, USA). GSSSG was synthesized and purified as previously reported⁶¹.

Patient plasma collection

The protocol used in this study was approved by the ethics committee of the Faculty of Life Sciences, Kumamoto University (Approval No. 169). This study was performed in accordance with the Ethical Guidelines for Human Genome/Gene Analysis Research of Ministries of Japan. Informed consent was obtained from all the subjects at the corresponding clinic. Eleven subjects with CKD aged 45–76 years (4 women, 7 men) and 14 subjects with diabetic nephropathy (90 > eGFR ≥ 15) were enrolled. For healthy subjects, blood was collected in EDTA-loaded tubes from men and women at 20 s and 30 s and centrifuged at 3000 × g for 10 min. All the plasma used in this study was from Japanese patients.

HSA modification

Oxidation

UV irradiation: UV irradiation of HSA was performed with a UV lamp at a distance of 5 cm from the bottom of the plate. The intensity of the UV lamp at a distance of 254 nm was 614 mW/cm². CT: HSA (300 µM) was reacted with CT (100 mM) in 67 mM sodium phosphate buffer (pH 8.0) for 0–48 h at 37 °C under aerobic conditions and dialyzed against DI water at 4 °C. Metal-Catalyzed Oxidation (MCO): HSA (300 µM) was reacted with ascorbic acid (25 mM) and 100 µM FeCl₃ in 50 mM HEPES (pH 7.4) with 100 mM potassium chloride and 10 mM magnesium chloride under aerobic conditions at 37 °C for 0–12 h. The reactant was dialyzed against deionized water at 4 °C. All the oxidized HSA was used within a day.

Preparation of β-NADH-treated HSA

HSA (150 µM) and β-NADH (100 mM) were mixed in 0.2 M Tris–HCl (pH 8.0) and incubated at 37 °C for 24 h. HSA was purified by dialysis against deionized water at 4 °C after the reaction.

Measurement of the antioxidative activity

The antioxidative activity of plasma and HSA was evaluated by radical scavenging assays. For plasma, 20 μL of tenfold diluted plasma was added to prewarmed PBS (37 °C, 920 μL) and incubated with 16 mM linoleic acid (10 μL, in 50 mM borate buffer, pH 9.0) and 50 mM AAPH or deionized water (50 μL). After 90 min at 37 °C, 200 μL of the reaction mixture was transferred to a UV-transparent microplate, and absorbance at 234 nm was measured⁶².

For HSA, antioxidative activity was assessed by the DPPH assay. Briefly, 250 μM DPPH in ethanol was mixed with an equal volume of 50 mM MES buffer (pH 7.4). Samples adjusted to 40 μM with PBS were incubated with the mixture at 25 °C for 30 min. Absorbance at 517 nm was recorded, and the radical elimination rate was determined relative to PBS-treated controls^63,64.

Polysulfide analysis

Methylene blue-based measurement

Polysulfide levels were measured as previously described³⁴ using two liberation methods (EMSP and DTT). For the EMSP-liberated assay, samples (final HSA 5–15 μM or plasma diluted 20-fold) were incubated with 0.3 M L-ascorbic acid and 1 M KOH at 37 °C for 4 h (final volume 200 μL). After precipitation with 1% zinc acetate and repeated washing, the pellet was resuspended and reacted with 20 mM DPDA (in 7.2 M HCl) and 30 mM FeCl₃ (in 1.2 M HCl) for 30 min at 25 °C. The absorbance at 665 nm was measured, and concentrations were calculated from a Na₂S calibration curve (25–500 μM). The levels of supersulfides per protein were calculated by dividing by the levels of protein concentration (mol/mol) measured using BCA protein assay.

For the DTT-liberated assay, samples (200 μL; plasma diluted 10–20 fold, HSA 20–100 μM) were incubated with 3 mM DTT in 50 mM Tris–HCl (pH 8.0) at 37 °C for 60 min. Precipitation, washing, and subsequent DPDA/FeCl₃ reaction were performed as above, and absorbance was determined at 665 nm using the same calibration curve.

Detection of the reduced form of polysulfide via a probe

HSA (300 µM) in PBS was oxidized with UV irradiation (0, 500, 2000, 4000 mJ/cm²) via a UVP Crosslinker (CL-3000). The reaction mixture (20 µL) was incubated with 1 µM DN-TG (80 µL) in 100 mM HEPES buffer containing 1% Triton X-100 (pH 7.4) at 25 °C for 15 min. The fluorescence intensity at ex. 485 nm and em. The band at 528 nm was measured via a microplate reader.

Mass spectrometry-based detection of the reduced form of polysulfide

Polysulfide detection via mass spectrometry was performed as previously described with slight modifications. HSA was labeled with 4 mM HPE-IAM in 50 mM Tris–HCl (pH 7.5) at 37 °C for 3 h. Four volumes of acetone were added to the reaction mixture and stored at –20 °C for at least 2 h, and the resulting precipitates were collected by centrifugation (13,000 × g, 10 min, room temperature). The precipitated proteins were dissolved in 50 mM Tris–HCl buffer containing 1 M urea (final protein concentration: 0.25 µg/µL), followed by vortexing and sonication. Trypsin Gold, mass spectrometry grade (Promega), was added to the reaction mixture (total protein/trypsin = 50/1, w/w) and incubated at 37 °C overnight (ca. 24 h). The sample pH was adjusted to 3 or less with trifluoroacetic acid to quench the reaction, and the resulting peptides were desalted by homemade StageTip embedded Empore™ SDB-XC disks (GL Science) according to previously published methods⁶⁵. The eluate was dried in vacuo and dissolved in 2% MeCN and 0.1% formic acid in distilled water for LC‒MS/MS analysis.

Nanoliquid chromatography-tandem mass spectrometry (nanoLC-MS/MS)

For proteomic analysis, nanoLC‒MS/MS analyses were performed on an Ultimate 3000 RSLCnano system (Thermo Fisher Scientific) coupled to a Q Exactive hybrid quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific) equipped with a nano ESI source. The nanoLC system was equipped with a trap column (C18 PepMap 100, 0.3 × 5 mm, 5 µm, Thermo Fisher Scientific) and an analytical column (NTCC-360/75-3-125, Nikkyo Technos, Tokyo, Japan). Peptide separation was performed via a 15 min gradient of water containing 0.1% formic acid (mobile phase A) and acetonitrile containing 0.1% formic acid (mobile phase B) at a flow rate of 300 nL/min. The elution conditions were as follows: 0–3 min, 2% B; 3–18 min, 2–40% B; 18–20 min, 40–95% B; 20–30 min, 95% B; 30–32 min, 95–2% B; and 32–45 min, 2% B. Measurements were performed in triplicate. The mass spectrometer was operated in data-dependent acquisition mode. The MS parameters were set as follows: spray voltage, 2.0 kV; capillary temperature, 275 °C; S-lens RF level, 50; scan type, full MS; scan range, m/z 350–1500; resolution, 70,000; polarity, positive; automatic gain control target, 3 × 10⁶; and maximum injection time, 100 ms. The MS/MS parameters were set as follows: resolution, 17,500; automatic gain control target, 1 × 10⁵; maximum injection time, 60 ms; normalized collision energy, 27; dynamic exclusion, 15 s; loop count, 10; isolation window, 1.6 m/z; and charge exclusion, unassigned, 1, 8, > 8.

Data Processing through Proteome Discoverer software

Protein identification and relative quantitation were performed via Proteome Discoverer software ver. 2.4 (Thermo Fisher Scientific). The analytical parameters were set as follows: search engine, sequest HT; protein database, Albumin (Homo Sapiens); enzyme name, trypsin (full); precursor mass tolerance, 10.0 ppm; fragment mass tolerance, 0.02 Da; max missed cleavages, 2; PSM validation, fixed value PSM validator; dynamic modification, oxidation (methionine, + 15.99 Da), 2-((4-hydroxyphenethyl)amino)-2-oxoethyl modification (cysteine, + 177.08; Cys-S-Tag), 2-((4-hydroxyphenethyl)amino)-2-oxoethyl-1-thiol modification (cysteine, + 209.06; Cys-S–S-Tag). The label-free quantification parameters for the detected peptides were set as follows: precursor quantification, precursor abundance on the basis of intensity; normalization mode, total peptide amount. Normalized values (normalized abundance) were used for relative quantification, and the peptide sequences used for relative quantification are summarized in Table 1.

Table 1.

Peptide sequences used for relative quantification.

Peptides sequencea	Cys residue	Cys-S-Tag		Cys-S–S-Tag
		Charge	Theo. MH+ [Da]b	Charge	Theo. MH+ [Da]
LC*TVATLR	C75	2	1053.576	2	1085.5482
QEPERNEC*FLQHK	C101	4	1834.854	4	1866.82616
LVRPEVDVMC*TAFHDNEETFLKK	C124	5	2898.416	5	2930.38853
AAC*LLPK	C177	2	892.4961	2	924.46816
YIC*ENQDSISSK	C265	2	1563.7	2	1595.67162
QNC*ELFEQLGEYK	C392	3	1777.81	2	1809.78223
RPC*FSALEVDETYVPK	C487	2	2030.989	3	2062.96126

a Asterisks in peptide sequences denote tag-labeled cysteine residues. b Theoretical monoisotopic mass ([N + H]⁺) of the corresponding peptide.

Rhabdomyolysis-induced AKI model mice

Six-week-old male ICR mice (28–32 g) were subjected to water deprivation for 12 h, and 50% (v/v) glycerol (10 mg/kg) was intramuscularly administered in half doses in each thigh under isoflurane inhalation as anesthesia. Urine was collected from the bladder (under anesthesia), and blood was collected from the inferior vena cava under isoflurane inhalation anesthesia at 1, 6, 12, and 24 h after administration. Blood was collected from the inferior vena cava via a heparin-loaded tube under the anesthethia with isoflurane, and plasma was obtained after the blood was centrifuged at 2000 × g for 10 min. After blood collection, the mice were euthanized by cervical dislocation. This animal protocol was approved by the animal committee at Tokushima University, and all procedures were conducted in accordance with institutional guidelines and regulations. This study is reported in accordance with the ARRIVE guidelines.

Statistical analysis

All the data are expressed as the mean ± standard deviation (S.D.). The values between groups were compared via the Tukey–Kramer method. Statistical analysis was performed with GraphPad Prism 9 (GraphPad Software, USA). A P value < 0.05 was considered to indicate a statistically significant difference.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AAPH

2,2′-Azobis(2-amidinopropane) dihydrochloride

AKI

Acute kidney injury

AOPP

Advanced oxidized protein products

BUN

Blood urea nitrogen

CARS

Cysteinyl-tRNA synthetase

CD

Circular dichroism

CKD

Chronic kidney disease

CT

Chloramine T

DN

Dinitrophenol

DPDA

N,N-dimethyl-p-phenylenediamine

DPPH

1,1-Diphenyl-2-picrylhydrazyl

DTT

Dithiothreitol

EMSP

Elimination method of sulfide from polysulfide

HSA

Human serum albumin

MCO

Metal catalyzed oxidation

NADH

Nicotinamide adenine dinucleotide

SCre

Serum creatinine

SSP4

Sulfane sulfur probe 4

STAR

Stress-triggered antioxidative responsible

TCEP

Tris(2-carboxyethyl)phosphine

TG

TokyoGreen

UV

Ultraviolet

Author contributions

M.II. participated in all other experiments and analyses. T.M., M.O., M.X., and T.I. assisted with data analysis. Y.I., M.II., M.S., T.A., T.S., and T.I. were involved in the study conception and design. K.U. and Y.M. contributed to polysulfide detection via mass spectrometry. M.II. wrote the manuscript. Y.I., T.A., T.S., M.X., and T.I. critically revised the manuscript. Y.I., T.S. and M.II. provided funds. All the authors reviewed and commented on the manuscript.

Funding

This work was supported, in part, by JSPS KAKENHI Grant Numbers 21H02645 (Y.I.), 21H05259 (Y.I.), 23K17215 (M.II. ), 21H05267 (T.S.), 23K20040 (T.S.), and Grant-in-Aid for JSPS Fellows 18J10935 (M.II.). In part, the work was supported by grants from the Smoking Research Foundation (Y.I.).

Data availability

The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.