Improved tag-switch method reveals that thioredoxin acts as depersulfidase and controls the intracellular levels of protein persulfidation

Abstract

Hydrogen sulfide (H₂S) has emerged as a signalling molecule capable of regulating several important physiological functions such as blood pressure, neurotransmission and inflammation. The mechanisms behind these effects are still largely elusive and oxidative posttranslational modification of cysteine residues (protein persulfidation or S-sulfhydration) has been proposed as the main pathway for H₂S-induced biological and pharmacological effects. As a signalling mechanism, persulfidation has to be controlled. Using an improved tag-switch assay for persulfide detection we show here that protein persulfide levels are controlled by the thioredoxin system. Recombinant thioredoxin showed an almost 10-fold higher reactivity towards cysteine persulfide than towards cystine and readily cleaved protein persulfides as well. This reaction resulted in H₂S release suggesting that thioredoxin could be an important regulator of H₂S levels from persulfide pools. Inhibition of the thioredoxin system caused an increase in intracellular persulfides, highlighting thioredoxin as a major protein depersulfidase that controls H₂S signalling. Finally, using plasma from HIV-1 patients that have higher circulatory levels of thioredoxin, we could prove depersulfidase role in vivo.

Fig. 1. Trx reacts with LMW persulfide NAP-SSH. (A) NAP-SSH undergoes spontaneous re-arrangement in buffer to give persulfide analogs. (B) Kinetics of 50 μM NAP-SSH decay followed by MS. Disappearance of the NAP-SSH parent ion (m/z 240.0347; calculated 240.0359), in the absence (black squares) and presence of 50 μM Trx (red circles), was plotted over time. (C) Time resolved MS spectra of m/z 208.0623 (calculated 208.0638) peak, which corresponds to NAPSH, indicate that Trx cleaves NAP-SSH to form a thiol and release HS^–. See Fig. S1 in ESI. (D) Deconvoluted mass spectrum of [Trx_red + H]⁺ (black) and simulated isotopic distribution for the fully reduced Trx (C₅₂₈H₈₃₈N₁₃₂O₁₅₉S₃; green). See Fig. S2 in ESI. (E) Deconvoluted mass spectrum of Trx mixed with NAP-SSH (red) and simulated isotopic distribution of fully oxidized Trx (C₅₂₈H₈₃₆N₁₃₂O₁₅₉S₃; blue). See Fig. S2 in ESI. (F) Overlay of the starting Trx spectrum and the spectrum obtained after 2 min of incubation with NAP-SSH clearly indicates m/z 2 leftward shift, indicative of the loss of 2H atoms.

Fig. 2. Trx system cleaves cysteine and protein persulfides to form H₂S. (A) H₂S release from 20 μM HSA-SSH upon addition of different concentrations of Trx (E. coli), measured amperometrically by the H₂S sensitive electrode. (B) Plot of the initial rates vs. the concentration of Trx. (C) Combined with TrxR (rat) and NAPDH, Trx (human) efficiently release all sulfur trapped in HSA-SSH as H₂S. Na₂S was injected as internal standard for the comparison. (D) Kinetic analysis of the reaction was performed by measuring the rates of NADPH oxidation varying the concentration of HSA-SSH and keeping the concentrations of other parameters constant: 1 μM Trx (human), 0.01 μM TrxR (rat) and 250 μM NADPH. Experiments were performed in triplicates. (E) Schematic overview of the reaction used for the generation of the CysSS^–/CysSSCys mixture. (F) Kinetics of Trx (E. coli) oxidation (1 μM) with 10 μM CysSSCys (black line) or CysSS^–/CysSSCys mixture (red line), followed by tryptophan fluorescence (λ_ex 280 nm) changes. Inset: Emission spectra before and after the reaction of 1 μM Trx with 50 μM CysSS^–/CysSSCys mixture. See Fig. S5 in ESI. (G) Kinetic analysis of the reaction of CysSSCys (black line) or CysSS^–/CysSSCys mixture (red line) with 1 μM Trx (human), 0.01 μM TrxR (rat) and 250 μM NADPH.

Fig. 3. Improved tag-switch assay labels persulfides in cell lysates and fixed cells. (A) Original tag-switch assay with chemical structures of green fluorescence cyanoacetic acid derivative, CN-BOT and cyanoacetic acid derivative, CN-Cy3. (B) Validation of CN-Cy3 labelling of protein persulfides in tissue extracts from CSE^+/+ and CSE^–/– mice (n = 3, *p < 0.01). Protein ladder in decreasing order: 245, 180, 135, 100, 75, 63, 48, 35, 25, and 20 kDa. (C) Validation of CN-BOT labelling of protein persulfidation in fixed cells. A decrease of intracellular persulfidation in SH-SY5Y neuroblastoma cells could be observed after 1 h incubation with CBS inhibitor AOAA (2 mM). Incubation with 100 μM Na₂S or 2 mM d-cysteine (1 h) leads to the increase of protein persulfidation. Semi-quantification of n = 100 cells was performed using ImageJ software. Data represent mean ± S.E.M., *p < 0.01.

Fig. 4. Persulfidation co-localizes with mitochondria. (A) Co-localization studies of intracellular persulfidation. BAECs were transfected with RFP fused to the leader sequence of E1 alpha pyruvate dehydrogenase and incubated with mitochondria-targeted H₂S donor, AP39. (B) SH-SY5Y neuroblastoma cells were transfected with CellLight® Mitochondria-RFP, BacMam 2.0 to visualize mitochondria and the cells were stained with improved tag-switch assay for persulfidation. Analysis with ImageJ (version 1.45R) revealed that significant co-localization was evident, which was visually represented as white pixels using the co-localization highlighter. Mitochondria-targeted H₂S donor, AP39, lead to a strong increase in mitochondrial persulfidation. Scale bar 10 μm. (C) 100 nM AP39 is much more powerful inducer of persulfidation in the cells than 2 mM d-cysteine. Scale bar 20 μm.

Fig. 5. Reaction of H₂S with sulfenic acids and not with disulfides could be a mechanism for intracellular persulfidation. (A and B) Incubation of SH-SY5Y neuroblastoma cells with 0.5 mM diamide (30 min) does not increase intracellular persulfidation, while the incubation with H₂O₂ (100 μM, 30 min) leads to a higher intracellular persulfide content (A). Treatment with 100 μM Na₂S (1 h) was used as a positive control. At least 5 images were recorded from each experiment performed in triplicate. (B) Semi-quantification of 40 cells was performed using ImageJ software. Data represent mean ± S.E.M, *p < 0.01.

Fig. 6. Trx system is the main regulator of intracellular persulfide levels. (A) Due to the activity of Trx system in the cell lysates obtained under native conditions (HEN buffer, pH 7.4, 1% protease inhibitors, 0 °C), levels of persulfides detected by improved tag-switch assay decrease with the time of protein extraction (gel on the left). This can be completely prevented by the addition of 2 μM auranofin into the lysis buffer (gel on the right). Data are presented as mean ± S.D (n = 3), *p < 0.01. Protein ladder in decreasing order: 200 kDa, 150 kDa, 120 KDa, 100 kDa, 70 kDa, 60 kDa, 50 kDa, 40 kDa, 30 kDa, 25 kDa and 20 kDa. (B) Representative amperometric measurement of H₂S release from BAEC lysate (1 mg mL^–1 protein). Cell lysate was added first (first arrow), followed by the addition of a mixture containing 1 μM Trx, 0.01 μM TrxR and 250 μM NADPH (second arrow). 100 μM ZnCl₂ was added in the end to prove that the response is indeed from H₂S. (C–E) Inhibition of Trx system with 2 μM auranofin for 1 h, leads to an increase of intracellular persulfidation in BAE cells in a similar way like the treatment with 100 μM Na₂S (1 h incubation) as detected by fluorescence microscopy using CN-BOT based tag-switch assay (C) or measured in gel from the cell lysates, using CN-Cy3 based tag-switch assay (D). Scale bar 20 μm. (E) Semi-quantification of in-gel fluorescence intensity. Experiments were performed in triplicates. Data are presented as mean ± S.D., *p < 0.01. (F–H) The effect of 2 μM auranofin or 100 μM Na₂S treatments (1 h, 37 °C) on the intracellular persulfide levels in SH-SY5Y cells, measured by fluorescence microscopy using CN-BOT based tag-switch assay (F) or measured in gel from the cell lysates, using CN-Cy3 based tag-switch assay (G). Scale bar 20 μm. (H) Semi-quantification of in-gel fluorescence intensity. Experiments were performed in triplicates. Data are presented as mean ± S.D., *p < 0.01.

Fig. 7. In vitro and in vivo sulfane sulphur levels are control by thioredoxin system. (A) Quantification of total sulfane sulfur levels in control and BAE cells treated with 2 μM auranofin for 1 h done by isotope dilution MS approach (see Fig. S11 in ESI†). Data are presented as mean values ± S.D. of n = 5, *p < 0.001. (B) Quantification of total sulfane sulfur levels in plasma samples from untreated and ART-treated HIV patients by isotope dilution MS approach. Data are presented as mean ± S.D. *p < 0.001 (see Fig. S12 and Table S1†).