Does N-Acetylcysteine Modulate Post-Translational Modifications of Transthyretin in Hemodialysis Patients?

Abstract

It is assumed that effects of the thiol antioxidant N-acetylcysteine (NAC) are mediated by interaction with protein-associated cysteine residues, however, information on protein level in vivo are missing. Therefore, we analyzed NAC-induced modifications of the protein transthyretin (TTR) in plasma of hemodialysis patients in a randomized, placebo-controlled study. TTR was selected due to its low molecular weight and the free cysteine residue in the polypeptide chain, which is known to be extensively modified by formation of mixed disulfides. The intravenous application of NAC during a hemodialysis session resulted in a substantial increase of native TTR from median 15% (range 8.8%–30%) to median 40% (37–50) and reduction of S-cysteinylated TTR [51% (44–60) vs. 6.6% (2.4–10)]. Additionally the pronounced formation of a TTR-NAC adduct was detected. However, all these modifications seemed to be reversible. Additionally, in vitro incubation of plasma with NAC confirmed the in vivo results and indicated that changes in post-translational modification pattern of TTR were a function of NAC concentration. Based on these observations and the essential metabolic and biochemical role of protein-associated cysteine residues we hypothesize that the interaction of NAC with proteins may explain altered protein functions due to modification of cysteine residues. Antioxid. Redox Signal. 19, 1166–1172.

Introduction

P atients with end-stage renal disease reveal an increased risk for cardiovascular complications, which have been associated with increased oxidiative stress. In this context, N-acetylcysteine (NAC) has been demonstrated to improve antioxidative status and cardiovascular outcome in hemodialysis patients (1). Although the exact underlying mechanisms of NAC function are not completely understood, the effects seem to be based on the antioxidant properties of NAC and protein-associated thiols, particularly cysteine residues, have been suggested as major NAC targets (6). Protein-associated cysteine residues are essential for numerous cellular and metabolic processes including cellular signaling, proliferation, and differentiation (5). Therefore, modification of protein-associated cysteine residues might explain effects of NAC including improvement of endothelial function, modulation of immune system, and anti-cancerogenic properties (6). However, modifications of cysteine residues by NAC have not been analyzed in detail yet. Since the analysis of protein modifications is complex and time-consuming, the selection of a representative protein seems to be appropriate. For this purpose, the plasma protein transthyretin (TTR) might represent an adequate candidate to analyze the effects of NAC on protein-associated cysteine residues directly on protein level. TTR, formerly known as prealbumin, is mainly synthesized in the liver and its main function is the transport of thyroxin and retinol in the blood (4). The visceral protein is composed of four identical subunits with a molecular weight (MW) of ∼14 kDa, which are noncovalently linked. Each TTR subunit reveals a free cysteine residue at position 10 (Cys10), which is not involved in any inter- or intraprotein disulfide linkage (4), but known to be susceptible to post-translational modifications (PTMs) by formation of mixed disulfides (7). To evaluate the effect of NAC on TTR in vivo, plasma from patients with end-stage renal disease receiving NAC intravenously during one hemodialysis session was obtained before and at the end of the dialysis session. Additionally, plasma samples were obtained from the same patients in a session without NAC application. Further, the dose-dependent effects of NAC on PTMs of TTR were evaluated in vitro by incubating human plasma with different NAC concentrations. In all samples TTR was isolated by immunoprecipitation and PTM patterns were analyzed by matrix assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF-MS).

Innovation

This study demonstrated for the first time direct interaction of N-acetylcysteine (NAC) with the cysteine residues of the plasma transthyretin (TTR) in vivo, which were modulated by NAC in favor of the reduced state. Additionally, the formation of TTR-NAC adducts could be detected. All NAC-induced modifications of TTR were reversible and supporting in vitro experiments indicated that the extent of modification is dose-dependent. These findings may explain altered protein functions due to modification of cysteine residues.

Effects of NAC on PTM of TTR During Application In Vivo

A total of six patients with end-stage renal disease (five male, one female) and a median age of 71 (range 46–75) years participated in this prospective, randomized, placebo-controlled study. The clinical characteristics of patients at the start and at the end of hemodialysis sessions, both in the absence and presence of NAC, are given in Table 1. The hemodialysis-induced anthropometrical and biochemical changes were not affected by NAC application.

Table 1.

Anthropometrical and Plasma Parameters of Patients with ESRF at the Start and at the End of Hemodialysis, in the Absence (Placebo Control) and Presence of N-Acetylcysteine, During the Hemodialysis Session in a Randomized, Prospective, Placebo-Controlled Study

	NAC-treatment			Placebo treatment
Characteristic	Predialysis	Postdialysis	p	Predialysis	Postdialysis	p
n (male/female)	6 (5/1)
Age (years)	71 (46–75)
Body weight (kg)	69.7 (50.0–93.6)	69.2 (49.0–91.4)	0.031	71.9 (50.0–92.6)	70.0 (49.5–91.5)	0.031
SBP (mm Hg)	127 (90–161)	137 (77–160)	0.750	140 (90–153)	131 (57–173)	0.438
DBP (mg Hg)	70 (40–75)	65 (49–71)	0.917	64 (46–80)	60 (33–85)	0.281
Creatinine (μM)	588 (406–769)	282 (174–345)	0.031	773 (306–876)	289 (127–522)	0.031
BUN (μM)	28.6 (17.6–36.7)	9.60 (5.90–13.7)	0.031	28.2 (15.3–49.8)	9.40 (5.40–33.0)	0.031
Total protein (g/L)	66.5 (59.0–74.0)	71.0 (67.0–77.0)	0.031	62.5 (58.0–74.0)	72.5 (63.0–80.0)	0.031
TTR (μM)	5.78 (2.81–13.1)	5.58 (2.59–12.3)	0.844	8.29 (4.59–11.0)	6.93 (3.85–8.52)	0.063

Data are presented as median (range).

BUN, blood urea nitrogen; DBP, diastolic blood pressure; ESRF, end-stage renal failure; NAC, N-acetylcysteine; SBP, systolic blood pressure; TTR, transthyretin.

The characterization of TTR by MALDI-TOF-MS analysis revealed the presence of different variants as shown in Figure 1. For the evaluation of changes in PTMs native TTR (index B in Fig. 1) and the mixed disulfides S-sulfonated TTR (index C), S-cysteinylated TTR (index D), S-cysteinylglycinated TTR (index F), and S-glutathionylated TTR (14,064 Da, index G) and the TTR-NAC adduct (index E) were considered as described in detail in the Subjects and Methodology section.

FIG. 1.

Changes in post-translational modification pattern of transthyretin (TTR) in plasma of patients with end-stage renal disease pre- and postdialysis with N- acetylcysteine (NAC) and placebo treatment, respectively. Letters indicate TTR variants with A TTR modified by cleavage of cysteine side chain to form glycine (13,719 Da), B native TTR (13,762 Da), C sulfonated TTR (13,842 Da), D cysteinylated TTR (13,881 Da), E TTR-NAC-adduct (13,923 Da), F cysteinylglycinated TTR (13,938 Da), G glutathionylated TTR (14,065 Da) (7). The symbol *indicates adducts of TTR with sodium (+22 Da).The peak heights are given as intensities determined as arbitrary units (arb.u.).

The relative amounts of TTR variants before and after hemodialysis in the presence and absence of NAC are presented in Figure 2. Before dialysis, no differences in relative intensities of TTR variants between placebo and NAC treatment could be observed and S-cysteinylated TTR represented the major TTR variant in all baseline plasma samples. The predominance of S-cysteinylated TTR is most likely based on the elevated levels of cysteine in plasma of hemodialysis patients, resulting in a pronounced oxidation of protein thiol moieties (9) and might be interpreted as a consequence of increased oxidative stress, which is frequently present in end-stage renal disease (1).

FIG. 2.

Changes in post-translational modifications of TTR in plasma of patients with end-stage renal disease before (grey boxes) and after hemodialysis (white boxes) with NAC and with placebo treatment. Data are presented as box-whisker-plots with whiskers showing interquartile ranges, giving the amount of post-translational modifications of TTR. Open circles indicate outliers (1.5-fold interquartile ranges) and asterisks indicate extreme outliers (3-fold interquartile ranges).

The hemodialysis process per se (placebo treatment) did not induce any changes in PTM pattern of TTR. However, application of NAC during hemodialysis resulted in a significant increase of the relative intensity of native TTR (14.6% vs. 45.1%, p=0.028) and TTR-NAC adduct (11.0% vs. 39.6%, p=0.028) while the relative intensity of S-cysteinylated TTR substantially decreased (51.4% vs. 6.6%, p=0.028). Additionally, S-cysteinylglycinated TTR was no longer detectable after the application of NAC, while S-sulfonated and S-glutathionylated TTR remained unchanged during treatment with NAC. In accordance with these results, reduced cysteine and cysteinylglycine plasma levels during NAC therapy have been reported before and attributed to the displacement of both substances from their protein-binding sites (3). It is assumed that NAC interacts with proteins in a thiol-disulfide exchange reaction, promoting the rapid dissociation of protein-attached aminothiols such as cysteine (2, 3). Accordingly, the amount of proteins with thiol moieties in reduced state is increased as shown in the present study by the increase of native TTR. Since reduced protein-associated thiols represent a major constituent of the extracellular oxidant defense system, this might at least partly explain the improved antioxidant capacity during NAC therapy (1). The presence of TTR-NAC adducts in post-dialysis plasma is presumably based on the interaction of excess NAC with the previously reduced protein-associated thiols of TTR, namely Cys10 residues, as already described for serum albumin (2). The fact that S-sulfonated and S-glutathionylated TTR were not influenced by NAC therapy in the present study might be based on the stronger oxidizing nature of sulfonic acid and glutathione (5), which is apparently not affected by NAC. In this context, it should also be mentioned that the NAC-induced modifications of TTR seem to be reversible (Fig. 2), probably due to the rapid and mainly nonrenal degradation of NAC (6) and the short biological half-life time of TTR (∼2 days) (4). However, the exact underlying mechanism of NAC-TTR interaction remains to be elucidated.

Additionally, the efficacy of TTR modification by NAC seems to vary individually. Although all patients revealed similar tendencies concerning the PTMs of TTR during NAC therapy, the extent of modification differed among the patients, which might be attributed to individual differences of NAC pharmacokinetics (Fig. 2). Based on these observations, TTR might represent a useful marker for evaluation of individual pharmacokinetics of NAC and optimization of NAC therapy and should be addressed in future research. TTR might be superior as pharmacokinetic marker in comparison to other, more abundant proteins such as albumin, due to the short biological half-life time of TTR (4), reducing the bias based on protein turnover, and the noncovalent association of the low-MW TTR subunits, enabling the mass spectrometric evaluation of the entire protein without proteolytical or chemical treatment and thereby reducing the method-caused bias of PTMs characterization.

NAC-Induced Changes in TTR PTM Pattern During In Vitro Incubations

As shown in Figure 3, mass spectrometric analysis of TTR from human plasma after incubation with NAC, revealed the presence of the same TTR variants as observed during in vivo experiments. However, the PTM patterns of TTR substantially differed as a function of the applied NAC concentration and the relative intensities of the considered TTR variants are shown in Figure 4. In detail, the relative intensity of native TTR increased concentration-dependent during incubation with NAC from 36.7% in control (0 μg/ml NAC) to 46.6% after incubation with 500 μg/ml NAC (p=0.003), 57.8% after incubation with 1250 μg/ml NAC (p<0.001), and 86.3% after incubation with 2500 μg/ml NAC (p<0.001). In contrast, the relative intensity of S-cysteinylated TTR steadily decreased from 44.7% in control to 35.4% after incubation with 5 μg/ml NAC (p=0.033), 12.2% after incubation with 50 μg/ml NAC (p<0.001), 4.3% after incubation with 500 μg/ml NAC (p<0.001), 2.6% after incubation with 1250 μg/ml NAC (p<0.001), and 1.2% after incubation with 2500 μg/ml NAC (p<0.001). Additionally, the relative intensities of S-sulfonated and S-glutathionylated TTR decreased, however, the effect was less pronounced than for S-cysteinylated TTR and only significant after incubation with 2500 μg/ml NAC in comparison to control (1.9% vs. 5.2% with p=0.017 and 0.4% vs. 1.5% with p=0.004, respectively). With regard to TTR-NAC adduct a biphasic adduct-formation (inverted U-shape) could be observed with no adducts in control incubations (0 μg/ml NAC) and highest adduct formation capacity after incubation with 50 μg/ml NAC (43.8%, p<0.001 in comparison to control) and 500 μg/ml NAC (45.4%, p<0.001 in comparison to control). At a concentration of 1250 μg/ml NAC the adduct formation capacity was decreased but still higher than in control approaches (35.8%, p<0.001 in comparison to control). The relative intensity of TTR-NAC adduct after incubation with 2500 μg/ml NAC was even more decreased but still higher than in control (11.5%, p<0.001) and comparable to the relative intensity observed in incubations with 5 μg/ml NAC (9.6%, p<0.001 in comparison to control). In any incubation, except control and incubation with 5 μg/ml NAC, S-cysteinylglycinated TTR could not be detected.

FIG. 3.

Changes in post-translational modifications pattern of TTR in human plasma after incubation with different concentrations of NAC for 4 h. Letters indicate TTR variants with A TTR modified by cleavage of cysteine side chain to form glycine (13,719 Da), B native TTR (13,762 Da), C sulfonated TTR (13,842 Da), D cysteinylated TTR (13,881 Da), E TTR-NAC-adduct (13,923 Da), F cysteinylglycinated TTR (13,938 Da), G glutathionylated TTR (14,065 Da) (7). The symbol *indicates adducts of TTR with sodium (+22 Da). The peak heights are given as intensities determined as arb. u.

FIG. 4.

Changes in post-translational modifications of TTR in plasma after incubation with 0, 5, 50, 500, 1250, and 2500 μg/ml NAC for 4 h. Data are presented as box-whisker-plots with whiskers showing interquartile ranges, giving the amount of post-translational modifications of TTR. Boxes with differing superscripts are significantly different from each other with p<0.05 for each analyzed NAC concentration.

These findings are in accordance with previous results obtained by others, reporting a dose-dependent interaction of NAC with plasma proteins (2) and suggesting the displacement of other aminothiols such as cysteine and cysteinylglycine from their protein-binding sites (3). However, the biphasic character of NAC interaction with a plasma protein, namely TTR, is shown in the present study for the first time. The exact underlying mechanisms of this process remain to be elucidated, but the molar excess of NAC might promote the reduction of newly formed TTR-NAC adducts in a downstream reaction (5). Further, it should also be emphasized that the biphasic character of NAC-protein interaction did not seem to influence the dose-dependent effects of NAC concerning other thiol modifications of TTR, which is indicated by the continuous increase of native TTR and the dose-dependent reduction of all other TTR variants with increasing NAC concentrations.

Concluding Remarks and Future Directions

In conclusion, we have shown that NAC directly alters PTMs of TTR in favor of the native protein by interacting with the protein-associated cysteine residue in vitro and in vivo. In this context, we have demonstrated that the extent of NAC-induced modifications of TTR vary individually but nevertheless seem to be reversible. The interaction of NAC with proteins seems to be dose-dependent with a biphasic character of NAC-TTR interaction. Therefore, future studies are warranted considering the essential role of cysteine residues for protein structure and function and the importance of thiol-affecting substances such as NAC in various clinical and experimental setting to highlight their importance for the redox state.

Notes

Subjects and Methodology

Patients

We investigated the effects of NAC during hemodialysis on the PTM pattern of TTR in a randomized, prospective, placebo-controlled study in six patients with end-stage renal disease.

All patients received NAC (5 g in 5% glucose) during one hemodialysis session by continuous intravenous infusion. During another session NAC was replaced by 5% glucose solution as placebo, which was also continuously applied intravenously. The study protocol was in accordance with the Declaration of Helsinki and was approved by the local Ethics Committee. Written informed consent was obtained from all patients before entry into the study. All patients had received hemodialysis for at least 3 months prior to study commencement. Patients were routinely dialyzed for 4 to 5 h three times weekly, using biocompatible membranes with no dialyzer re-use. Blood flow rates were 250 to 300 ml/min, dialysate flow rates were 500 ml/min, dialysate conductivity was 135 mS, and the dialysates used were bicarbonate-based. Kt/V values [the amount of plasma cleared of urea divided by the urea distribution volume) was measured according to the formula Kt/V=−ln (R −0.03)+(4 −3.5×R)×UF/W; using R=post/pre plasma urea nitrogen ratio; UF=ultrafiltrate volume (L) removed and W=post-dialysis weight (kg)]. Blood pressure was obtained by a conventional sphygmomanometric method after a rest of 10 min. Phases I and V of the Korotkoff sounds were considered as representing systolic blood pressure and diastolic blood pressure, respectively. Blood was collected immediately before the start and at the end of the hemodialysis session.

Laboratory procedures

Laboratory values were routinely analyzed. Levels of TTR in plasma were determined by a noncommercial enzyme-linked immunosorbent assay (ELISA) using polyclonal rabbit anti-human TTR antibodies as previously described in detail (8).

In vitro incubation of human plasma with NAC

Human plasma, obtained from healthy volunteers as described above, (10 μl) was mixed with equal amount of TBS (50 mM Tris, 150 mM NaCl, pH adjusted to 7.5 with HCl) containing 0, 10, 100, 1000, 2500, and 5000 μg/ml NAC (Sigma-Aldrich, Munich, Germany), respectively, resulting in final concentrations of 0, 5, 50, 500, 1250, and 2500 μg/ml NAC per approach. The mixtures were incubated at room temperature. After 4 h 10 μl Sephadex G15 (GE Healthcare, Munich, Germany; 3 mg/ml in HPLC-grade water, Carl Roth, Karlsruhe, Germany) and 5 μl polyclonal rabbit anti human TTR antibody (DakoCytomation, Hamburg, Germany) were added for immunoprecipitation and subsequent MALDI-TOF-MS analysis as described in detail below. All incubation mixtures were performed in triplicates on 3 consecutive days, resulting in nine experimental approaches per NAC concentration.

Immunoprecipitation of TTR and MALDI-TOF-MS

Plasma (10 μl) was mixed with 10 μl Sephadex G15 and 5 μl polyclonal rabbit anti human prealbumin antibody. The mixture was incubated overnight at room temperature and then centrifuged at 16,000 g for 20 min. The supernatant was removed and the immunoprecipitated complex of TTR was extensively washed twice with phosphate-buffered saline (Sigma-Aldrich) and once with HEPES (5 mM; Sigma-Aldrich). After a final centrifugation, the pellet was resuspended in 10 μl HPLC-grade water.

MALDI mass spectra of the immunoprecipitated TTR was obtained using an AutoflexSpeed MALDI-TOF mass spectrometer (Bruker-Daltonik, Bremen, Germany). MALDI-TOF MS was performed in linear mode with 19.5 kV for ion source 1, 18 kV for ion source 2, 7.15 kV for lens voltage, and 270 ns pulsed ion extraction using 2′,5′-dihydroxyacetophenone as matrix. The matrix was prepared by dissolving 7.6 mg 2′,5′-dihydroxyacetophenone (Bruker-Daltonik) in 375 μl HPLC-grade ethanol (Carl Roth) and 125 μl diammonium-hydrogencitrate (80 mM; Sigma-Aldrich). For sample preparation trifluoracetic acid (2%; Carl Roth), sample and matrix were mixed in equal amounts and 1 μl of this mixture was applied per spot. The samples were analyzed in triplicates. For ionization, a Smartbeam^™-II laser (355 nm, 1000 Hz) was used and 1500 shots per spot were collected. For spectra calibration Protein Calibration Standard I (Bruker Daltonik) was used as external standard. Spectra were evaluated using the software flexAnalysis version 3.3 (Bruker-Daltonik).

Evaluation of MALDI-TOF-MS spectra

The peaks obtained from MALDI-TOF mass spectra for TTR and its variants were assigned according their MWs as native TTR (13,762 Da), S-sulfonated TTR (13,842 Da), S-cysteinylated TTR (13,881 Da), TTR-NAC adduct (13,923 Da), S-cysteinylglycinated TTR (13,938 Da), S-glutathionylated TTR (14,064 Da), TTR modified by cleavage of cysteinyl side chain to form glycine (13,719 Da), and adducts of TTR with sodium (mass shift of +22 Da) (Fig. 1). The assignment of peaks was based on the work of Kishikawa et al. (7). The assignment of TTR-NAC adduct was based on the mass difference corresponding to the MW of NAC (163 Da).

For semi-quantitative analysis only native TTR and mixed disulfides (sulfonated, cysteinylated, cysteinylglycinated, glutathionylted TTR, and TTR-NAC adduct) were considered. For each mass spectrum the intensities of these six TTR forms were summed and the intensity of each form was then displayed as percentage of the summed intensities as shown here exemplarily for native TTR: relative intensity native TTR=(intensity _native _TTR ×100)/(intensity_native _TTR+intensity_sulfonated _TTR+intensity_{cysteinylated} _TTR+intensity_TTR-NAC+intensity_{cysteinylglycinated} _TTR+intensity_{glutathionylated} _TTR). TTR forms that could not be detected were included in the calculation with an intensity value of zero. Since each sample was analyzed in triplicates the mean percentage for each sample was calculated and used for statistical purposes.

Data analysis and statistics

The results are presented as median and ranges except otherwise stated. Statistical analyses were accomplished by nonparametric procedures using SPSS statistical service pack version 19.0. Mann–Whitney-rank test was used for independent variables and Wilcoxon signed-rank test for related variables. In vitro results concerning the influence of different NAC concentrations were analyzed using Kruskal–Wallis test with Bonferoni correction. Values of p<0.05 (two-tailed) were considered to be statistically significant.

Footnotes

Acknowledgments

We thank Lydia Häußler, Elisabeth Pilz, and Cathleen Friedrich for their technical assistance. This study was supported by Else Kröner-Fresenius Stiftung, Sonnenfeld-Stiftung, European Union Interreg 4A, and the Danish Council for Independent Research.

Abbreviations Used

References

Coombes

, Fassett

. Antioxidant therapy in hemodialysis patients: a systematic review. Kidney Int, 81:233–246. 2011.

Harada

, Naito

, Otagiri

. Kinetic analysis of covalent binding between N-acetyl-L-cysteine and albumin through the formation of mixed disulfides in human and rat serum in vitro. Pharm Res, 19:1648–1654. 2002.

Hultberg

, Andersson

, Masson

, Larson

, Tunek

. Plasma homocysteine and thiol compound fractions after oral administration of N-acetylcysteine. Scand J Clin Lab Invest, 54:417–422. 1994.

Ingenbleek

, Young

. Transthyretin (prealbumin) in health and disease: nutritional implications. Annu Rev Nutr, 14:495–533. 1994.

Jacob

, Battaglia

, Burkholz

, Peng

, Bagrel

, Montenarh

. Control of oxidative posttranslational cysteine modifications: from intricate chemistry to widespread biological and medical applications. Chem Res Toxicol, 25:588–604. 2012.

Kelly

. Clinical applications of N-acetylcysteine. Altern Med Rev, 3:114–127. 1998.

Kishikawa

, Sass

, Sakura

, Nakanishi

, Shimizu

, Yoshioka

. The peak height ratio of S-sulfonated transthyretin and other oxidized isoforms as a marker for molybdenum cofactor deficiency, measured by electrospray ionization mass spectrometry. Biochim Biophys Acta, 1588:135–138. 2002.

Raila

, Henze

, Spranger

, Mohlig

, Pfeiffer

, Schweigert

. Microalbuminuria is a major determinant of elevated plasma retinol-binding protein 4 in type 2 diabetic patients. Kidney Int, 72:505–511. 2007.

Suliman

, Anderstam

, Lindholm

, Bergstrom

. Total, free, and protein-bound sulphur amino acids in uraemic patients. Nephrol Dial Transplant, 12:2332–2338. 1997.