Blood Thiol Redox State in Chronic Kidney Disease

Thiols (sulfhydryl groups) are effective antioxidants that can preserve the correct structure of proteins, and can protect cells and tissues from damage induced by oxidative stress. Abnormal levels of thiols have been measured in the blood of patients with moderate-to-severe chronic kidney disease (CKD) compared to healthy subjects, as well as in end-stage renal disease (ESRD) patients on haemodialysis or peritoneal dialysis. The levels of protein thiols (a measure of the endogenous antioxidant capacity inversely related to protein oxidation) and S-thiolated proteins (mixed disulphides of protein thiols and low molecular mass thiols), and the protein thiolation index (the molar ratio of the S-thiolated proteins to free protein thiols in plasma) have been investigated in the plasma or red blood cells of CKD and ESRD patients as possible biomarkers of oxidative stress. This type of minimally invasive analysis provides valuable information on the redox status of the less-easily accessible tissues and organs, and of the whole organism. This review provides an overview of reversible modifications in protein thiols in the setting of CKD and renal replacement therapy. The evidence suggests that protein thiols, S-thiolated proteins, and the protein thiolation index are promising biomarkers of reversible oxidative stress that could be included in the routine monitoring of CKD and ESRD patients.


Introduction
Chronic kidney disease (CKD) affects approximately 10% of the adult population worldwide [1]. In 2016, it was the ninth most common cause of death in high-income countries, and epidemiological projections-considering the demographic trends and lifestyles of this population-suggest that the mortality rates due to this pathology will more than double globally by 2040 [2]. Therefore, the development of interventional strategies to support the prevention and management of CKD is of primary importance for the health system. According to the Kidney Disease Improving Global Outcomes [3], CKD is classified into five stages based on the glomerular filtration rate (GFR), which represents the best index of renal function, and three stages based on albuminuria levels.
Recently, in subjects without CKD (n = 4745) who participated in the Prevention of End-stage REnal and Vascular Disease (PREVEND) study, elevated serum free thiol concentrations-adjusted to the total serum protein-were shown to be associated with a lower risk of the new-onset of CKD in this population, constituting a possible marker for the identification of the individuals who are most at risk of developing the disease. Therefore, the assessment of the serum free sulfhydryl status would be useful for preventive purposes, in order to mitigate the future burden of CKD in the general population [61].
HcySH can bind reversibly to P-SH, forming S-homocysteinylated proteins (HcySSP). Moreover, HcySH can also be converted to its cyclic thioester, forming the Hcythiolactone, which reacts with the epsilon-amino group of Lys residues in protein, thus generating N-homocysteinylated protein [62]. It has been proposed that the N-homocysteinylation of low-density lipoproteins (LDL) can contribute to an acceleration of the atherogenic process in subjects with hyperhomocysteinaemia [63]. Most of the GSH in human blood is found inside RBCs (approximately 3 mM per RBC), as its plasma concentration is in the µ M range, and leukocytes-which contain 2-4 mM GSH-are much less numerous than RBCs. In healthy erythrocytes, 98% of glutathione is present in its active reduced The plasma levels of total cysteinylglycine (CysGlySH + its oxidised forms) and total homocysteine (HcySH + its oxidised forms) are higher than those of GSH, the main intracellular antioxidant and comprised, respectively, between 18-36 µM and 5-15 µM [52]. Hyperhomocysteinemia is an important risk factor for vascular disease and atherosclerosis [6,[53][54][55][56][57], as well as for autoimmune diseases [58] and neurological disorders [59]. Higher plasma concentrations of total homocysteine are caused by both genetic and nongenetic factors: among the latter, we can mention renal failure [60].
Recently, in subjects without CKD (n = 4745) who participated in the Prevention of End-stage REnal and Vascular Disease (PREVEND) study, elevated serum free thiol concentrations-adjusted to the total serum protein-were shown to be associated with a lower risk of the new-onset of CKD in this population, constituting a possible marker for the identification of the individuals who are most at risk of developing the disease. Therefore, the assessment of the serum free sulfhydryl status would be useful for preventive purposes, in order to mitigate the future burden of CKD in the general population [61].
HcySH can bind reversibly to P-SH, forming S-homocysteinylated proteins (HcySSP). Moreover, HcySH can also be converted to its cyclic thioester, forming the Hcy-thiolactone, which reacts with the epsilon-amino group of Lys residues in protein, thus generating N-homocysteinylated protein [62]. It has been proposed that the N-homocysteinylation of low-density lipoproteins (LDL) can contribute to an acceleration of the atherogenic process in subjects with hyperhomocysteinaemia [63]. Most of the GSH in human blood is found inside RBCs (approximately 3 mM per RBC), as its plasma concentration is in the µM range, and leukocytes-which contain 2-4 mM GSH-are much less numerous than RBCs. In healthy erythrocytes, 98% of glutathione is present in its active reduced form, GSH [64], and protects RBC proteins from irreversible oxidation by maintaining -SH groups of haemoglobin in the reduced state (Hb-SH) [65].
The GSH/GSSG redox potential is lower in the erythrocytes of haemodialyzed patients than in those of healthy controls [66]. In addition, serum from haemodialyzed patients has decreased GSH levels [67] and a diminished activity of the antioxidant enzymes superoxide dismutase and GSH peroxidase [68]. Under conditions of moderate or severe oxidative stress, erythrocytic proteins (e.g., haemoglobin or membrane skeletal proteins) may be prevalently S-glutathionylated in a reversible manner [36,69].
In the plasma of healthy adults, the concentration of P-SH ranges from 400 to 600 µM [41,70], and is mainly due to the Cys34 residue of albumin, which-due to its high plasma concentration (about 600 µM)-accounts for 80% of plasma -SH groups [71]. The plasma albumin pool consists of a mixture of mercaptalbumin, bearing the free thiol and non-mercaptalbumin, bearing the modified thiol. Considering that, in healthy adults, about 70% of total albumin is mercaptoalbumin, its -SH group reaches concentrations higher than 400 µM. Thus, the albumin Cys34 thiol is an important scavenger of ROS in the vascular compartment, and therefore a quantitatively important redox buffer of human blood [41,[71][72][73][74][75], acting as an effective extracellular antioxidant because of its multiple ligand binding capabilities [72]. These properties are closely related to the structure and redox state of the molecule [76]. The non-mercaptoalbumin pool consists mainly of mixed disulphides of albumin with LMM-SH [41]. In the presence of moderate or severe oxidative stress, the reduced albumin (Alb-SH) concentration decreases, and that of S-cysteinylated albumin increases, whereas a low level, or absence, of oxidative stress induces the opposite changes, i.e., a higher level of Alb-SH and a lower level of S-cysteinylated albumin.
form, GSH [64], and protects RBC proteins from irreversible oxidation by maintaining -SH groups of haemoglobin in the reduced state (Hb-SH) [65].
The GSH/GSSG redox potential is lower in the erythrocytes of haemodialyzed pa tients than in those of healthy controls [66]. In addition, serum from haemodialyzed pa tients has decreased GSH levels [67] and a diminished activity of the antioxidant enzymes superoxide dismutase and GSH peroxidase [68]. Under conditions of moderate or severe oxidative stress, erythrocytic proteins (e.g., haemoglobin or membrane skeletal proteins may be prevalently S-glutathionylated in a reversible manner [36,69]. In the plasma of healthy adults, the concentration of P-SH ranges from 400 to 600 μM [41,70], and is mainly due to the Cys34 residue of albumin, which-due to its high plasma concentration (about 600 μM)-accounts for 80% of plasma -SH groups [71]. The plasma albumin pool consists of a mixture of mercaptalbumin, bearing the free thiol and non mercaptalbumin, bearing the modified thiol. Considering that, in healthy adults, abou 70% of total albumin is mercaptoalbumin, its -SH group reaches concentrations higher than 400 μM. Thus, the albumin Cys34 thiol is an important scavenger of ROS in the vas cular compartment, and therefore a quantitatively important redox buffer of human blood [41,[71][72][73][74][75], acting as an effective extracellular antioxidant because of its multiple ligand binding capabilities [72]. These properties are closely related to the structure and redox state of the molecule [76]. The non-mercaptoalbumin pool consists mainly of mixed disul phides of albumin with LMM-SH [41]. In the presence of moderate or severe oxidative stress, the reduced albumin (Alb-SH) concentration decreases, and that of S-cysteinylated albumin increases, whereas a low level, or absence, of oxidative stress induces the oppo site changes, i.e., a higher level of Alb-SH and a lower level of S-cysteinylated albumin.

P-SH and S-Thiolated Proteins in the Moderate-to-Severe Stages of CKD
Several studies have examined the plasma P-SH levels in patients with moderate-to severe stages of CKD compared with healthy subjects, by spectrophotometry (Table 1). A pioneering study by Himmelfarb and colleagues showed that the plasma P-SH levels were significantly lower in the CKD patient group, likely because of oxidative stress [77].

P-SH and S-Thiolated Proteins in the Moderate-to-Severe Stages of CKD
Several studies have examined the plasma P-SH levels in patients with moderate-tosevere stages of CKD compared with healthy subjects, by spectrophotometry (Table 1). A pioneering study by Himmelfarb and colleagues showed that the plasma P-SH levels were significantly lower in the CKD patient group, likely because of oxidative stress [77]. Table 1. P-SH and S-thiolated proteins in the plasma or serum of patients with moderate-to-severe stages of CKD. All of the data are expressed as means ± SD, unless otherwise indicated. (a) Median, with interquartile range in parentheses. The data of S-thiolated proteins are in grey cells. *** p < 0.001, ** p < 0.01, * p < 0.05.

Concentration
Albumin  [86] Successively, lower plasma P-SH levels in CKD patients were correlated to lower plasma albumin levels (hypoalbuminaemia) [78], suggesting a causal relationship between hypoalbuminaemia and lower plasma P-SH levels. A positive association between P-SH and the estimated glomerular filtration rate (eGFR), and a negative correlation between P-SH and creatinine was also demonstrated in adult CKD patients in a recent study, which also detected a significant reduction in P-SH in stage 5 CKD patients [87]. However, the albumin levels were not evaluated, and the control group subjects were significantly younger than CKD patients [87].
The levels of total plasma thiols (P-SH + LMM-SH) were measured in a group of patients with CKD stages 1-5, most of whom were stage 3 (46%) and stage 4 (24%): the levels of plasma thiols, adjusted to the plasma albumin concentrations, decreased dependently on the CKD stage [83]. Moreover, a positive correlation between the eGFR and thiol concentration was observed [83]. A decrease in Alb-SH accompanied by a decrease in renal function was confirmed in non-diabetic CKD patients [88]. Decreased thiol and albumin levels have also been reported in a more recent study conducted on a small group of children with stage 3-5 CKD [89]. This study included and grouped data from both haemodialyzed and non-haemodialyzed patients, making it difficult to understand the real burden of the results differentiated for the two groups. However, a positive correlation between the albumin concentration and levels of thiols, (mixed and symmetrical) disulphides, and total thiols was found in the plasma of CKD children. When the thiol, disulphide, and total thiol levels were normalized to the albumin concentration, only the levels of total thiols and thiols were significantly lower in CKD children than in age-matched control subjects. In the CKD group, the levels of thiols, total thiols, and disulphides correlated positively with the albumin concentration, and negatively with the urea concentration, whereas the thiol and total thiol levels were positively correlated with the eGFR and negatively with the creatinine level [89]. Two other recent studies on adult CKD patients not on HD determined that total plasma thiol, and the ("native") thiol levels were significantly lower in CKD patients than in healthy controls [90,91]. One study also found a significant reduction in the level of disulphides but no significant difference in the disulphide/thiol, disulphide/total thiol, and thiol/total thiol ratios between the CKD and healthy control groups [91]. The other study found a significant variation in disulphide/thiol and disulphide/total thiol ratios only [90]. This discrepancy could be explained by the fact that the control group in the latter study had a significantly higher BMI than the CKD group, causing an increase in oxidative stress in the control group with increased levels of disulphides [92].
Terawaki and colleagues first adopted the redox state of albumin Cys34 measured by an HPLC method as a marker of oxidative stress in pre-dialysis patients with CKD. Their findings indicated that both S-cysteinylated albumin and irreversibly oxidized albumin (Alb-SO 2 H and Alb-SO 3 H) increased as renal function, measured as creatinine clearance, decreased [84]. The authors confirmed their findings in a subsequent study, where they also showed that albumin-SH levels progressively decreased with decreasing kidney function [85]. Conversely, the levels of S-thiolated albumin increased progressively as kidney function decreased, suggesting that the redox state of albumin progressively changes with the worsening of kidney function. Moreover, they found a significant positive correlation between S-thiolated albumin and blood urea nitrogen, and between S-thiolated albumin and serum creatinine. Another recent study analysed the serum fraction of oxidized albumin normalized to total albumin (i.e., oxidized and native albumin) in pre-dialysis CKD patients [93]. The correlation between the oxidized serum fraction of albumin, f(HNA), and numerous clinical parameters related to CKD was examined. The study revealed a positive correlation between f(HNA), kidney function and urea nitrogen in the blood, and also with the degree of anaemia of the CKD patients, independent of age; f(HNA) was also significantly and independently associated with the plasma levels of haemoglobin and ferritin. A negative correlation was demonstrated between the levels of xanthine oxidase-an isoform of xanthine oxidoreductase which is thought to increase the cardiovascular burden among CKD patients via oxidative radical production-and the plasma fraction of Alb-SH and reversibly oxidized albumin [94]. These findings suggest that the analysis of the redox state of albumin Cys34 thiol can be a potentially useful method for the assessment of oxidative stress and the degree of kidney dysfunction in CKD patients, although some analytical and instrumental limitations must be overcome.
Elevated levels of plasma HcySH are commonly present in patients with CKD. In a recent study involving 62 patients in different stages of CKD, a negative correlation was found between the plasma HcySH and eGFR in patients with CKD stages III to V, although a significant increase in HcySH levels was present only in subjects with stage V compared to patients with stage II CKD. The plasma Cys/Hcy ratio turned out to be a more sensitive indicator because it was already significantly decreased at stage III of CKD [95]. Zinellu and colleagues reported that HcySH and CysSH bound to LDL were significantly increased in CKD patients, and the plasma levels of S-homocysteinylated LDL and creatinine were positively correlated with each other [86]. Based on these results, the authors suggested that the increase in S-homocysteinylated LDL levels might account, at least in part, for the excess of cardiovascular risk in CKD patients: S-homocysteinylated LDL can therefore be considered a key risk marker of cardiovascular disease (CVD) in CKD patients [86]. Another study detected a significant increase in S-cysteinylated and S-homocysteinylated plasma proteins in CKD patients compared to healthy subjects [96]. More recently, Zinellu and colleagues evaluated the plasma levels of P-SH and LMM-SH in a group of hypertensive CKD patients at stages 3-4, more than 60% of whom were hyperhomocysteinemic (HcySH > 15 µmol/L), reporting that the thickness of the carotid intima-media was inversely related to plasma levels of P-SH [82].

P-SH and S-Thiolated Proteins in HD Patients Compared to Healthy Subjects
Several studies have examined the plasma levels of P-SH and S-thiolated proteins in ESRD patients on maintenance HD compared to healthy subjects; others detected the concentration of S-thiolated proteins in the RBCs of HD patients (Table 2). Table 2. P-SH and S-thiolated proteins in the plasma or serum of ESRD patients on HD. All of the data are expressed as means ± SD, unless otherwise indicated. Data regarding S-thiolated proteins are in the grey cells. (b) Median with interquartile range in parentheses; (c) median with the 25-75th percentile in parentheses. HbSSCy, cysteinylated haemoglobin; HbSSG, S-glutathionylated haemoglobin; PTI, protein thiolation index.
Several studies examined in HD patients the levels of plasma S-thiolated proteins [102][103][104][105] and S-thiolated haemoglobin in RBCs [66] ( Table 2). The plasma levels of Alb-SH and S-cysteinylated albumin were determined in HD patients subjected to intravenous iron administration by the HPLC method [84]. HD patients had increased levels of S-cysteinylated albumin and irreversibly oxidized albumin (Alb-SO2H and Alb-SO3H), and decreased levels of Alb-SH [102]. On the other hand, the plasma levels of S-homocysteinylated proteins were significantly higher in HD patients than in healthy subjects [104]. This finding could have functional consequences; indeed, the site of interaction with diazepam in homocysteinylated albumin is altered due to the presence of the sulphur amino acid, and the interaction between the drug and the protein is reduced, partly explaining why the albumin-binding capacity is reduced in uraemia [104]. Other investigations determined that ESRD patients show increased levels of HcySH, which have been related to carotid artery atherosclerosis and increased rates of cardiovascular events [109,110]. Other studies have shown that S-glutathionylated haemoglobin (HbSSG) was 46% higher in HD patients than in healthy subjects, and that S-cysteinylated Hb (HbSSCy) was >3-fold higher in HD patients than in controls [66]. In addition, both HbSSCy and HbSSG were positively correlated with blood urea nitrogen, creatinine, and normalized protein catabolic rate, which are parameters related to both uraemia and nutrition [66].

P-SH and S-Thiolated Proteins in HD Patients Before and After an HD Session
Several studies have measured the levels of P-SH and S-thiolated proteins in HD patients before (pre-HD) and after an HD session (post-HD) ( Table 3). Some authors reported that the levels of total plasma thiols (P-SH + LMM-SH) significantly decreased post-HD, returning to the range measured in healthy subjects [111][112][113]. The plasma P-SH levels in ESRD patients on HD were significantly lower than those in healthy subjects. However, the HD session was associated with an increase in the plasma P-SH levels, such that after HD the concentration of P-SH did not differ from that of healthy subjects. Similar results were also obtained when plasma P-SH were expressed as nanomoles per milligram of protein, calculated using total plasma protein concentrations [112]. This result was also confirmed when considering possible dialysis-induced changes in the plasma protein concentration [113], which can dramatically affect the measured values of total plasma thiols Himmelfarb and colleagues reported a significant decrease in plasma P-SH in HD patients [77], in line with what has also been observed in other studies ( Table 2). For example, the plasma P-SH levels (measured in µmol/L) decreased significantly in hypoalbuminemic HD patients, both compared to healthy subjects and to normoalbuminemic HD patients, as well as between the latter and healthy subjects [97]. The levels of inflammation and oxidative stress biomarkers increased in both groups of HD patients compared to healthy subjects, and further selective increases in levels of inflammation and oxidative stress biomarkers occurred in hypoalbuminemic patients, suggesting that hypoalbuminemia, acute-phase inflammation, and oxidative stress could act synergistically to increase cardiovascular morbidity and mortality risk in HD patients [97]. Significantly lower plasma P-SH levels in hypoalbuminemic HD patients could be due, at least in part, to hypoalbuminemia.
Plasma P-SH has also been investigated in HD patients not receiving iron or recombinant human erythropoietin [99]. Iron deficiency is common in HD patients [106]: intravenous iron administration is one of the cornerstones of anaemia's treatment, but it could further aggravate oxidative stress [107]. The plasma P-SH levels were found to be lower in HD patients not receiving intravenous iron or recombinant human erythropoietin vs. healthy controls [99]; however, this study did not specify whether the HD patients were normoalbuminemic or hypoalbuminemic.
Koca and colleagues divided HD patients into four groups according to their HD duration (dialysis vintage) [100], reporting that the total plasma thiols were significantly lower in all of the groups of HD patients than in healthy subjects, but no significant difference was found between the groups of HD patients with different dialysis vintages [100].
The protein thiolation index (PTI, a new biomarker of oxidative stress that is defined as the molar ratio of the sum of all LMM-SH bound to S-thiolated plasma proteins to free P-SH [46,108]) is higher in HD patients than in controls, whereas the concentration of P-SH and that of total plasma thiols are lower in HD patients than in healthy subjects [90,101,105] ( Figure 3). P-SH and PTI provide different but complementary information. A low P-SH concentration indicates an absolute loss of protein reducing power but does not provide information on concomitant changes in disulphide forms, while a high PTI indicates a relative excess of oxidized vs. reduced proteins in patients; however, being a ratio, it offers no indication about the absolute amounts of disulphide and reduced forms. Several studies examined in HD patients the levels of plasma S-thiolated proteins [102][103][104][105] and S-thiolated haemoglobin in RBCs [66] ( Table 2). The plasma levels of Alb-SH and S-cysteinylated albumin were determined in HD patients subjected to intravenous iron administration by the HPLC method [84]. HD patients had increased levels of S-cysteinylated albumin and irreversibly oxidized albumin (Alb-SO 2 H and Alb-SO 3 H), and decreased levels of Alb-SH [102]. On the other hand, the plasma levels of S-homocysteinylated proteins were significantly higher in HD patients than in healthy subjects [104]. This finding could have functional consequences; indeed, the site of interaction with diazepam in homocysteinylated albumin is altered due to the presence of the sulphur amino acid, and the interaction between the drug and the protein is reduced, partly explaining why the albumin-binding capacity is reduced in uraemia [104]. Other investigations determined that ESRD patients show increased levels of HcySH, which have been related to carotid artery atherosclerosis and increased rates of cardiovascular events [109,110]. Other studies have shown that S-glutathionylated haemoglobin (HbSSG) was 46% higher in HD patients than in healthy subjects, and that S-cysteinylated Hb (HbSSCy) was >3-fold higher in HD patients than in controls [66]. In addition, both HbSSCy and HbSSG were positively correlated with blood urea nitrogen, creatinine, and normalized protein catabolic rate, which are parameters related to both uraemia and nutrition [66].

P-SH and S-Thiolated Proteins in HD Patients before and after an HD Session
Several studies have measured the levels of P-SH and S-thiolated proteins in HD patients before (pre-HD) and after an HD session (post-HD) ( Table 3). Some authors reported that the levels of total plasma thiols (P-SH + LMM-SH) significantly decreased post-HD, returning to the range measured in healthy subjects [111][112][113]. The plasma P-SH levels in ESRD patients on HD were significantly lower than those in healthy subjects. However, the HD session was associated with an increase in the plasma P-SH levels, such that after HD the concentration of P-SH did not differ from that of healthy subjects. Similar results were also obtained when plasma P-SH were expressed as nanomoles per milligram of protein, calculated using total plasma protein concentrations [112]. This result was also confirmed when considering possible dialysis-induced changes in the plasma protein concentration [113], which can dramatically affect the measured values of total plasma thiols [105]. Moreover, the total plasma thiol levels decreased significantly even after haemodiafiltration [113]. Table 3. P-SH and S-thiolated proteins in the plasma of HD patients before and after HD. All of the data are expressed as means ± SD, unless otherwise indicated. The data regarding S-thiolated proteins are in grey cells. (b) Median with interquartile range in parentheses. Pre-HD, before HD; post-HD, after HD; PTI, protein thiolation index; M, male; F, female.

Concentration
Albumin  A study reported that the concentration of S-cysteinylated proteins in pre-HD ESRD patients was higher than that in controls, and did not change after HD [115]. This result contrasts with that of other studies showing a decrease in S-thiolated protein levels post-HD. The concentration of Alb-SH and S-cysteinylated albumin were measured in ESRD non-diabetic patients pre-HD and post-HD, compared to healthy subjects, by direct infusion electrospray mass spectrometry [116]. The total amount of albumin isolated from ESRD patients pre-HD was characterized by a significant decrease in the Alb-SH fraction, and by a concomitant increase in the S-cysteinylated albumin fraction; HD significantly decreased the S-cysteinylated albumin and restored Alb-SH in all ESRD patients [116]. A study of 20 patients under dialytic treatment measured 59 ± 3% and 42 ± 2% of S-thiolated albumin before and after dialysis, while 39 ± 3% S-thiolated albumin was found in healthy subjects [117].
Both the albumin thiol content and the conformational status of albumin were evaluated in CKD patients before and after HD, and were compared with those measured in healthy subjects [81]. The results were consistent with those already observed previously, with a reduction in the albumin thiol level, which was further significantly reduced after HD. The susceptibility to oxidation of plasma albumin isolated from controls and CKD patients before and after HD was assessed by paramagnetic resonance electron spectroscopy as a marker of the conformational modification of the molecule. The results show that albumin in CKD patients is highly modified in vivo, and is not vulnerable to oxidation in the same way as normal albumin, suggesting a possible conformational variation in HD subjects.
The total plasma thiols, plasma S-thiolated proteins, and PTI were measured in ESRD patients, before and after a regular HD session, and were compared to age-matched healthy subjects [101]. Colombo and colleagues found a significant decrease in the level of total plasma thiols and, conversely, a significant increase in the level of S-thiolated proteins in ESRD patients. In most patients, the level of post-HD total plasma thiols did not differ from those of healthy subjects, whereas the plasma level of S-thiolated proteins was lower in HD patients than in age-matched healthy controls (Figure 4), suggesting that a single HD session restores the thiol redox status of plasma and the antioxidant capacity of plasma thiols. In addition, HD patients had a significantly higher PTI than the age-matched healthy subjects, and HD was associated with a decrease in PTI to normal, or lower than normal, levels ( Figure 4).  The effects of HD on the total plasma thiol levels before (grey circles) and immediately after (black circles) HD were examined in patients with ESRD who received maintenance dialysis (n = 20). The horizontal solid line represents the mean of the total plasma thiol level in age-matched healthy subjects. The horizontal dashed lines represent the SD of the total plasma thiol level in age-matched healthy subjects. (B) The effects of HD on the S-thiolated plasma protein levels before (grey circles) and immediately after (black circles) HD were examined in patients with ESRD on maintenance dialysis (n = 20). The horizontal solid line represents the mean of the S-thiolated plasma protein levels in agematched healthy subjects. The horizontal dashed lines represent the SD of the plasma S-thiolated plasma protein level in age-matched healthy subjects. Reprinted with permission from Colombo et al. [101].

P-SH and S-Thiolated Proteins in PD
Only a few studies examined the levels of P-SH and S-thiolated proteins in PD. Himmelfarb and colleagues reported that the levels of total plasma thiols in PD patients were significantly different from those measured in healthy subjects (215 ± 19 μM vs. 279 ± 12, p = 0.011, n = 9 and 10, respectively) but not from those measured in HD patients [77]. Similar results were obtained in a subsequent study [118].
One study reported that continuous ambulatory PD restores the normal level of total GSH and total Cys concentration, whereas it decreases the level of S-thiolated proteins (except that of homocysteinylated proteins) in plasma [96].

Conclusions and Perspectives
All of the articles examined in this review provide evidence of a generalized thiol/oxidative stress in patients with CKD, which is reflected in the alterations of the levels of P-SH and S-thiolated proteins compared to healthy control subjects. Analyses of the levels of P-SH and S-thiolated proteins suggest that oxidative stress is present even in the early stages of CKD; it is augmented in parallel with the progression of the disease to ESRD, and is further exacerbated in HD patients. Oxidative stress is involved in the pathogenesis The effects of HD on the total plasma thiol levels before (grey circles) and immediately after (black circles) HD were examined in patients with ESRD who received maintenance dialysis (n = 20). The horizontal solid line represents the mean of the total plasma thiol level in age-matched healthy subjects. The horizontal dashed lines represent the SD of the total plasma thiol level in age-matched healthy subjects. (B) The effects of HD on the S-thiolated plasma protein levels before (grey circles) and immediately after (black circles) HD were examined in patients with ESRD on maintenance dialysis (n = 20). The horizontal solid line represents the mean of the S-thiolated plasma protein levels in age-matched healthy subjects. The horizontal dashed lines represent the SD of the plasma S-thiolated plasma protein level in age-matched healthy subjects. Reprinted with permission from Colombo et al. [101].

P-SH and S-Thiolated Proteins in PD
Only a few studies examined the levels of P-SH and S-thiolated proteins in PD. Himmelfarb and colleagues reported that the levels of total plasma thiols in PD patients were significantly different from those measured in healthy subjects (215 ± 19 µM vs. 279 ± 12, p = 0.011, n = 9 and 10, respectively) but not from those measured in HD patients [77]. Similar results were obtained in a subsequent study [118].
One study reported that continuous ambulatory PD restores the normal level of total GSH and total Cys concentration, whereas it decreases the level of S-thiolated proteins (except that of homocysteinylated proteins) in plasma [96].

Conclusions and Perspectives
All of the articles examined in this review provide evidence of a generalized thiol/oxidative stress in patients with CKD, which is reflected in the alterations of the levels of P-SH and S-thiolated proteins compared to healthy control subjects. Analyses of the levels of P-SH and S-thiolated proteins suggest that oxidative stress is present even in the early stages of CKD; it is augmented in parallel with the progression of the disease to ESRD, and is further exacerbated in HD patients. Oxidative stress is involved in the pathogenesis of CKD progression and complications, particularly cardiovascular diseases. Thus, it represents a non-traditional risk factor for all causes of mortality in patients with CKD and ESRD. For this reason, oxidative stress must become an important diagnostic and prognostic factor, and a target for CKD prevention and treatment [119]. Unfortunately, oxidative stress assessment is still not standardized, and this fact represents an obstacle for its use in everyday clinical practice.
Some studies have shown that the levels of total plasma thiols, measured with Ellman's DTNB method, diminish in moderate-severe CKD (Table 1) in ESRD patients on HD (Tables 2 and 3) or on PD. However, the spectrophotometric analysis of the total plasma thiols with Ellman's method mainly measures P-SH, as they represent the most abundant thiol pool in human plasma, being in the 400-600 µM range, whereas LMM-SH altogether constitute only 12-20 µM [41,50,70]. Therefore, the contribution of LMM-SH to the amount of plasma thiols measured by Ellman's method with DTNB can be considered negligible. Hence, these findings suggest a decrease in the plasma P-SH levels in patients with CKD, and in ESRD patients on HD or PD (Tables 1-3).
Unfortunately, the values of the plasma P-SH levels in CKD (Table 1) and ESRD patients on renal replacement therapy (Tables 2 and 3), as well as in healthy control subjects, are often shown in different units, e.g., µmol/L, µg/L, and µmol/g protein, making it very difficult to compare the results of various studies conducted in different laboratories. Furthermore, in some studies, the CKD and ESRD patients (Tables 1-3) were hypoalbuminemic (with serum albumin levels < 3.8 g/dL), making it particularly difficult to compare P-SH levels, even if they are shown in the same units, between CKD patients and healthy normoalbuminemic subjects. In order to compare the levels of plasma P-SH between CKD patients or ESRD patients on renal replacement therapy and healthy subjects, it is very important to normalize the P-SH concentrations against the total plasma protein or plasma albumin concentrations, e.g., to measure the concentration of the plasma P-SH as µmol -SH/g plasma proteins [44,83,101,114,118]. In this way, the P-SH values would become easily comparable between different studies and, within each individual study, between CKD patients and healthy controls.
Unfortunately, there is no gold standard method for the measurement of the albumin concentration in plasma. Albumin is generally measured using a colorimetric method based on bromocresol-green (BCG) or bromocresol-purple (BCP). Other approaches, e.g., immunonephelometry (INP) or immunoturbidimetry, are used less frequently, as they are more time-consuming and expensive. In addition, there are significant discrepancies when measuring the albumin concentration by different methods (BCG, BCP and INP), especially when determining it in ESRD patients on HD [120]. The albumin concentration measured by BCG is consistently 4-10 g/L higher than that determined by BCP or INP.
Another important point concerning the normalization of the P-SH concentration in relation to the plasma albumin concentration should be mentioned here. In fact, in CKD, albumin can undergo several irreversible post-translational modifications, such as carbonylation [32,121,122] and carbamylation [123][124][125]. Due to the increased urea and isocyanate concentrations in ESRD patients, protein carbamylation (i.e., the covalent binding of isocyanate to the ε-amino group of lysine) increases significantly [126]. For HD patients, the carbamylation of albumin is the main cause of the discrepancy found for values measured by BCP; this leads to an underestimation of 0-6 g/L. Several researchers have proposed that INP, or immunoturbidimetry, would be the preferred method for the measurement albumin in HD patients [120,127,128]. Some have also suggested that the choice of method for the measurement of the plasma albumin concentration (BCP, BCG, INP) could have an important impact on the classification of CKD patients [127].
Based on the results presented in this article, we can conclude that CKD and ESRD patients on HD or PD show lower plasma thiol levels than healthy controls, suggesting that the uremic condition per se plays a key role in oxidative/thiol stress (Table 3). S-cysteinylated Cys34 accounts for most of the oxidized forms of albumin in CKD [84,85,101] ( Table 1) and in ESRD patients on renal replacement therapy [102,103] (Tables 2 and 3). CKD patients are among those with the highest risk for CVD [129,130]. CKD and ESRD patients have a 5-to 10-fold higher risk of developing CVD than age-matched healthy controls [131,132]. The Cys34 thiol of albumin represents by far the largest fraction of all plasma thiols; thus, it can have a protective effect on the vascular endothelium, as was already stated above. In this regard, some researchers have assessed whether changes in the redox state of the Cys34 thiol of albumin influence the incidence of CVD in HD patients [133,134]. These authors reported that a lower concentration of Alb-SH was indicative of an increased CV mortality, demonstrating that the redox state of the Cys34 thiol is closely related to CVD in both HD [133,134] and PD patients [135]. The Alb-SH levels were significantly higher in PD patients with high peritoneal membrane transport properties than in those with low peritoneal membrane transport properties [136].
All of these studies suggest that, together with albumin carbamylation [124,125], the redox state of albumin Cys34 thiol is related to the overall risk factor for CVD among ESRD patients. However, care should be taken when interpreting the outcome of these small-scale studies to define the association of the redox status of albumin Cys34 thiol with CVD, as further large-scale studies are required.
Regarding S-thiolated plasma protein levels in pre-and post-HD ESRD patients (Table 3), following the HD session, the plasma P-SH levels transiently increase because of the de-thiolation of S-thiolated plasma proteins, as there is a decrease in oxidative stress during the HD session. This is very important in the case of albumin, the Cys34-SH of which represents the dominant antioxidant in plasma. However, in the period between two consecutive HD sessions, oxidative plasma conditions induce P-SH to bind LMM-SH, determining the reversible formation of S-thiolated plasma proteins. Therefore, S-thiolated plasma proteins represent a useful marker of thiol-specific reversible oxidative stress in ESRD patients on HD. Therefore, HD has only a limited and transient beneficial effect on the redox state of plasma P-SH, i.e., HD patients are exposed to periodical peaks of severe oxidative stress.
Individual HD sessions have different effects on different types of biomarkers of protein oxidation. For example, the post-HD levels of plasma protein carbonyls are significantly higher than the pre-HD levels, suggesting an increase in oxidative stress during the HD session [32,[137][138][139]. The apparent discrepancy between the low levels of S-thiolated proteins (and, consequently, high plasma P-SH levels) post-HD and the high plasma levels of protein carbonyls post-HD is probably because, while protein S-thiolation is a reversible process, protein carbonylation is, generally, irreversible [29].
In CKD, RBCs are exposed to oxidative stress and uremic toxins released into the blood. This is particularly evident in ESRD patients, where Hb has been found to be S-glutathionylated (Tables 2 and 3). This fact has clinical relevance because the exposure to oxidative stress and uremic toxins may have a role in the development of anaemia in patients with CKD. In fact, it has been widely demonstrated that the RBCs of CKD/ESRD patients have a reduced lifespan, which is caused by a hastened eryptosis, or RBC rupture, compared to healthy subjects. It was observed that, when blood from normal donors was transfused into uremic patients, reductions in the RBC lifespan occurred [140]. Several studies examined RBCs' oxidative stress in renal failure. HbSSG has been suggested to be a clinical marker of oxidative stress in ESRD patients undergoing HD [66,141,142] or PD [142]. Moreover, the structure of Hb in ESRD patients is impaired, and these alterations are enhanced during HD [143]. Regarding RBC thiol stress in HD and PD, LMM-SH (i.e., GSH) in the RBCs of patients on continuous ambulatory PD remain at a physiological level [144], whereas the GSH level in RBCs decreases before HD, and increases after HD [144,145]. However, in HD patients the increase in LMM-SH level is short-lasting and limited to a brief post-HD period, while the increase in the level of LMM-SH in the RBCs of PD patients is continuous and persistent [144]. Thus, both PD and HD therapies increase the antioxidant power of RBC, but PD does so continuously, while the effect of HD is short-lasting and intermittent. These results suggest that, as far as thiol stress is concerned, toxic compounds do not accumulate in PD patients, thanks to the continuous dialytic procedure and the fact that PD patients more often maintain a residual diuresis, lowering the severity of oxidative stress [144].
In conclusion, this review provides evidence indicating a decrease in the plasma P-SH levels and an increase in S-thiolated plasma proteins and S-thiolated RBC proteins in patients with CKD (Tables 1-3). The levels of P-SH and S-thiolated proteins, and the PTI are promising biomarkers of reversible thiol/oxidative stress in moderate-to-severe CKD, and in HD and PD. The successful transposition of these biomarkers into a clinical test requires the availability of a simple, artefact-free assay that should be selective, specific, and sensitive. The measurement of these biomarkers is complicated by the complexity of the methodologies that must be applied in order to avoid procedural artifacts, and that require special care in sample handling and preparation. Nevertheless, several methods are available that minimize artifacts and reproducibly measure the amounts of P-SH, Sthiolated proteins, PTI, and LMM-SH in cells, tissues, and body fluids [146][147][148][149][150][151]. These promising biomarkers, which are serious candidates for the evaluation of oxidative stress in CKD patients, deserve further exploration in large and appropriately designed prospective studies. When this will be done, the levels of blood P-SH, S-thiolated proteins, LMM-SH, and the PTI could be included in the routine monitoring of CKD patients. This will determine important clinical implications, because-at the moment-an assessment of oxidative stress in patients with CKD is not feasible in everyday clinical practice, in spite of its relevance in CKD progression and complications development.
Funding: This research received no external funding.

Conflicts of Interest:
The authors declare no conflict of interest. We further confirm that the manuscript has been read and approved by all of the named authors, and that the order of the authors listed in the manuscript has been approved by all of us.