Intracellular glutathione pools are heterogeneously concentrated

Abstract

Glutathione is present in millimolar concentrations in the cell, but its relative distribution among cellular compartments remains elusive. We have chosen the endoplasmic reticulum (ER) as an example organelle to study compartment-specific glutathione levels. Using a glutaredoxin sensor (sCGrx1pER), which rapidly and specifically equilibrates with the reduced glutathione (GSH)-glutathione disulfide (GSSG) redox couple with known equilibrium constant, we showed that the [GSH]:[GSSG] ratio in the ER of intact HeLa cells is less than 7:1. Taking into consideration the previously determined value for [GSH](2):[GSSG] in the ER of 83 mM, this translates into a total glutathione concentration in the ER ([GStot]=[GSH]+2[GSSG]) of greater than 15 mM. Since the integrated, intracellular [GStot] was measured as ~7 mM, we conclude the existence of a [GStot] gradient across the ER membrane. A possible homeostatic mechanism by which cytosol-derived glutathione is trapped in the ER is discussed. We propose a high [GStot] as a distinguishing feature of the ER environment compared to the extracellular space.

Keywords: DTT, Dithiothreitol; EGSH, Half cell reduction potential of glutathione; ER, Endoplasmic reticulum; Endoplasmic reticulum; GSH, Reduced glutathione; GSSG, Glutathione disulfide; Glutaredoxin; Glutathione; NEM, N-ethylmaleimide; OxD, Percentage of oxidation; PDI, Protein disulfide isomerase; PERK, Double stranded RNA-activated protein kinase (PKR)-like ER kinase; RGS, [GSH]:[GSSG]; Redox Homeostasis; Redox compartmentalization; Redox, Reduction–oxidation; Reduction potential; TMM(PEG)12, Maleimide-activated polyethylene glycol; UPR, Unfolded protein response; XBP1, X-box binding protein 1; [GStot], Total glutathione concentration; sCGrx1p, C30S mutant of yeast glutaredoxin 1.

Graphical abstract

Fig. 1

Targeting of sCGrx1p to the ER. (A) Schematic representation of sCGrx1p_ER. The active site Cys²⁷ is either reduced (–SH) or glutathionylated (–SSG). SP, ER signal peptide; HA, hemagglutinin epitope; KDEL, ER retrieval motif. (B) HyPer_ER (green) was transfected into HeLa cells and the cells stained with αPDI followed by a red-fluorescent goat-anti-mouse antibody. (C) HeLa cells were co-transfected with HyPer_ER (green) and sCGrx1p_ER, which was stained with αHA/goat-anti-mouse (red). Merged images are shown in the bottom panel, and white arrows highlight examples of co-localizing structures. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

sCGrx1p_ER neither causes ER hyperoxidation nor ER stress. (A) HeLa cells were transfected with the indicated constructs for 24 h or, to obtain fully reduced or oxidized control samples, treated with DTT (10 mM for 5 min) or diamide (5 mM for 5 min), respectively. Free cysteines were alkylated in situ with NEM. After cell lysis, proteins were reduced with TCEP and re-alkylated with AMS, and the modified lysates analyzed by SDS-PAGE and anti-ERp57 western blot (WB). The AMS-modified, oxidized fraction of ERp57 (ox) runs slower than the reduced fraction (red). Oxidized fractions, as determined by densitometry, are indicated in percent (% ox). Results of two independent experiments (EXP I and EXP II) are shown. (B) HeLa cells were transfected as in panel (A) or treated with DTT (2 mM for 1 h). The phosphorylation/activation of PERK was analyzed by anti-PERK western blot based on the decreased mobility of the phosphorylated protein (asterisk). In equivalent cell samples, total RNA was isolated and subjected to rtPCR analysis using primers specific for XBP1. Splicing of XBP1 mRNA is evident by the appearance of the spliced (s) and the hybrid (h) forms, as opposed to the unspliced (u) form.

Fig. 3

Pulse-chase analysis of sCGrx1p_ER. (A) HeLa cells transfected or not with sCGrx1p_ER were labeled with ³⁵S-methionine overnight and chased for the indicated time periods. sCGrx1p_ER was isolated by anti-HA immunoprecipitation and analyzed by SDS-PAGE and phosphorimaging. (B) Densitometric analysis of sCGrx1p_ER band intensity (n=3; mean±SEM).

Fig. 4

Redox state analysis of sCGrx1p_ER. HeLa cells were transfected or not with sCGrx1p_ER, pulsed overnight with ³⁵S-methionine and chased for the indicated time periods. NEM-modified cell lysates were subjected to anti-HA immunoprecipitation, and the immunoprecipitates treated with TCEP and TMM(PEG)₁₂ as described in Materials and methods. Samples were separated by SDS-PAGE and visualized on a phosphorimager. Where indicated, cells were treated with DTT or diamide prior to NEM-alkylation as in Fig. 2A.

Fig. 5

Schematic cartoon of a detail of a HeLa cell depicting the trapping of cellular glutathione in the ER. GSH can freely enter the ER (light brown) from the cytosol (cyto, light gray) where some of it is converted to GSSG through the action of the resident machinery for oxidative protein folding. GSSG cannot (or only inefficiently) diffuse back into the cytosol. Thus, as the concentration of GSH can equilibrate between luminal and cytosolic side of the ER membrane, [GS_tot] is significantly higher in the ER than in the cytosol.