The role of glutathione in periplasmic redox homeostasis and oxidative protein folding in Escherichia coli

The thiol redox balance in the periplasm of E. coli depends on the DsbA/B pair for oxidative power and the DsbC/D system as its complement for isomerization of non-native disulfides. While the standard redox potentials of those systems are known, the in vivo “steady state” redox potential imposed onto protein thiol disulfide pairs in the periplasm remains unknown. Here, we used genetically encoded redox probes (roGFP2 and roGFP-iL), targeted to the periplasm, to directly probe the thiol redox homeostasis in this compartment. These probes contain two cysteine residues that are virtually completely reduced in the cytoplasm, but once exported into the periplasm, can form a disulfide bond, a process that can be monitored by fluorescence spectroscopy. Even in the absence of DsbA, roGFP2, exported to the periplasm, was almost fully oxidized, suggesting the presence of an alternative system for the introduction of disulfide bonds into exported proteins. However, the absence of DsbA shifted the steady state periplasmic thiol-redox potential from −228 mV to a more reducing −243 mV and the capacity to re-oxidize periplasmic roGFP2 after a reductive pulse was significantly decreased. Re-oxidation in a DsbA strain could be fully restored by exogenous oxidized glutathione (GSSG), while reduced GSH accelerated re-oxidation of roGFP2 in the WT. In line, a strain devoid of endogenous glutathione showed a more reducing periplasm, and was significantly worse in oxidatively folding PhoA, a native periplasmic protein and substrate of the oxidative folding machinery. PhoA oxidative folding could be enhanced by the addition of exogenous GSSG in the WT and fully restored in a ΔdsbA mutant. Taken together this suggests the presence of an auxiliary, glutathione-dependent thiol-oxidation system in the bacterial periplasm.


Introduction
Extracellular proteins are stabilized by structural disulfide bonds that are introduced during oxidative protein folding. Thus, the correct folding of most of the proteins translocated into the periplasm and beyond in the model bacterium Escherichia coli depends on the formation of native disulfide bonds. The major pathway for introducing disulfide bonds in the periplasm of E. coli involves the DsbA-DsbB thiol oxidase system [1][2][3][4][5]. DsbA contains a thioredoxin-like domain including an active site CxxC motif with a standard redox potential of around − 122 mV [6,7]. After disulfide introduction, DsbA itself is re-oxidized by the integral membrane protein DsbB, which in turn is connected to the ubiquinone pool of the bacterial plasma membrane [8][9][10][11][12]. As some proteins contain multiple cysteine residues, cells require quality control systems to isomerize non-native disulfide bonds. In E. coli, this system includes the disulfide isomerases DsbC/G and the reductive power from DsbD, which itself is coupled to the cytoplasmic NADPH pool through Thioredoxin A (TrxA) [13][14][15][16][17][18].
In eukaryotes, oxidative folding in the endoplasmic reticulum (ER) is catalyzed in a similar manner by protein disulfide isomerase(s) (PDI), which also contains thioredoxin-like domains. Unlike DsbA, PDI not only acts as thiol oxidase but also functions as a disulfide reductase/ isomerase [19,20]. PDI is re-oxidized for example by the sulfhydryl oxidase Endoplasmic Reticulum Oxidoreductin 1 (ERO1-Lα) [21].
The tripeptide glutathione (GSH) is the most abundant low molecular weight thiol in many domains of life ranging from bacteria to eukaryotes [22][23][24] and usually it is noted for its reductive properties, keeping the cytoplasm reduced. In its reducing role, glutathione is oxidized to glutathione disulfide (GSSG) [25][26][27]. In E. coli, cytosolic GSH levels have been described to be in the millimolar range (up to 10 mM). The GSH:GSSG ratio ranges from 50:1 to 200:1 and is maintained by the glutathione reductase Gor, dependent on the cellular NADPH-pool [22,[27][28][29][30][31].
Exponentially growing E. coli cells extensively secrete GSH into the culture medium. Several ABC-transporters mediate GSH transport across the inner cell membrane and outer membrane porins, allow GSH or GSSG to enter or leave the periplasm with around 10% of all synthesized GSH being secreted. [32][33][34][35][36]. Although glutathione synthesis is restricted to the cytosol, it is exported into the periplasm of E. coli and periplasmic glutathione probably accounts for 10-30% of total cellular GSH [33,34,37]. It has been suggested that the GSH:GSSG ratio in the periplasm reflects the GSH:GSSG ratio of the culture medium, being approximately 16:1 [33,34,[37][38][39][40].
Since redox processes in the cell are highly regulated, the presence of glutathione in the periplasm at such high levels strongly indicates that this molecule also plays a role in the redox homeostasis of the periplasm, similar to its role in the ER [41,42]. In fact, impaired GSH synthesis or reduced GSH import into the periplasm is connected to several phenotypes, similar to those found in cells lacking the DsbA/DsbB-or DsbC-system [33,43]. However, a direct involvement in disulfide bond formation and redox homeostasis for GSH in the periplasm has not yet been established.
Here, we explore the role of GSH in oxidative protein folding and redox homeostasis in the periplasm of E. coli by targeting genetically encoded redox-sensitive probes with an engineered disulfide bond to this compartment and analyzing the periplasmic redox dynamics in the presence or absence of DsbA and/or GSH. We provide evidence that GSH indeed plays a key role in disulfide bond formation and redox homeostasis in the periplasm: We observed that GSSG can complement for the loss of DsbA and, unexpectedly, that GSH accelerates disulfide bond formation in the presence of DsbA. In line, the periplasmic redox balance of GSH-deficient cells was slightly shifted to a more reducing steady state redox potential and the oxidative folding of native DsbA substrates in these cells is impaired.

Strains, plasmids and growth conditions
Bacterial strains and plasmids used in this study are listed in Supplementary Tables S1 and S2. Escherichia coli DH5α [81] served as host for plasmid construction and storage and E. coli BL21 (DE3) [80] was used for DsbA and E. coli MG1655 [78] for roGFP2 recombinant protein production. All E. coli strains used in this study were routinely cultivated at 37 • C in Luria-Bertani (LB) medium, supplemented with antibiotics when required for plasmid maintenance and marker selection (ampicillin 200 μg/mL or kanamycin 100 μg/mL), if not stated differently.
Protein expression in BL21 was induced with 0.4 mM IPTG when the cultures reached an OD 600 of ~0.6-0.8 at 37 • C in LB medium and then incubated for 20 h at 20 • C after induction.
E. coli mutants from the KEIO collection were used for construction of double deletion strains and analysis of the periplasmic redox state [44]. For expression of the redox probes, mutant strains (Supplementary Table 1) harboring roGFP-based sensor protein-coding plasmids (pPT_roGFP2 or pPT_roGFP-iL, Supplementary Table S2) were cultivated in MOPS minimal medium (Technova, Hollister, CA, USA) at 37 • C until an optical density (OD) of ~0.5-0.8 was reached. Sensor protein expression was induced with 0.2 mM IPTG and cells were cultivated for 16 h at 20 • C.

Construction of different roGFP and DsbA expression vectors
For construction of the pPT plasmid enabling periplasmic targeting, the torA signal sequence was PCR amplified using appropriate primers (Supplementary Table S3) from genomic E. coli DNA (strain MC4100) and cloned into the pCC plasmid [45,79] using the introduced NdeI 3' and 5' restriction sites. The NdeI restriction site upstream of torA was then removed by QuickChange mutagenesis (Agilent, Waldbronn, Germany) according to the manufacturer's protocol using primers QC-NdeI-fw and QC-NdeI-rv (Supplementary Table S3) resulting in the pPT plasmid.
In order to target roGFP2 into the periplasm, roGFP2 was PCR amplified using the primer pair listed in Supplementary Table S3 from the pCC_roGFP2 plasmid as template. The introduced NdeI and EcoRI restriction sites were used to clone the gene into the pPT-plasmid resulting in pPT_roGFP2 coding for a TorA_roGFP2 fusion protein.
The roGFP-iL gene was purchased from Addgene (Watertown, MA, USA) in a pQE30 plasmid harboring an ampicillin resistance gene. The roGFP-iL gene was PCR amplified (Primers are listed in Supplementary  Table S3) with simultaneous introduction of XhoI and EcoRI restriction sites. The amplified gene was cloned into the pPT plasmid using the introduced restriction sites resulting in the plasmid pPT_roGFP-iL (pLK9).
For expression of a roGFP2-EcDsbA(ΔSP) fusion protein, a truncated variant of the dsbA (Δnt1-48) gene was PCR amplified lacking the periplasmic signal sequence (ΔSP). The PCR fragment was cut with EcoRI and HindIII. AtPrxA-ΔCP was removed from the p416TEF_roGFP2-AtPrxA-ΔCP [46] plasmid by the same restriction enzymes and EcDsbA (ΔSP) was cloned into the plasmid backbone, resulting in the plasmid pLK16.

Construction of E. coli dsbA, gshA double deletion strain by P1 transduction
For the construction of an E. coli gshA, dsbA double mutant, the kanamycin cassette was removed from the gshA KEIO mutant using the plasmid 709-FLPE as indicated by the supplier (Gene Bridges, Heidelberg, Germany). Removal of the kanamycin cassette was verified by colony PCR using k1, k2 and gshA primers listed in Supplementary  Table S3. The dsbA mutant from the KEIO collection (strain: JW3832) was used as P1 donor. P1 transduction was carried out as previously described [47]. Correct insertion of dsbA deletion in marker free ΔgshA was checked using appropriate primers (Supplementary Table S3).

Periplasmic roGFP2-based measurements in E. coli
For determining the roGFP2 redox state in E. coli WT, ΔdsbA, ΔgshA or ΔdsbAΔgshA, the cells were transformed with the pPT_roGFP2 plasmid encoding the TorA(SP)-roGFP2 fusion construct. Correct periplasmic probe localization was verified by fluorescence microscopy ( Supplementary Fig. 1). After expression of roGFP2 for 16 h at 20 • C, as aforementioned, cells were harvested, washed in HEPES buffer (40 mM, pH 7.4) and adjusted to an OD 600 of 1.0 in HEPES buffer containing either 1 mM Aldrithiol-2 (Sigma-Aldrich, Darmstadt, Germany, CAS-2127-03-9, AT-2) (oxidation control), 10 mM Dithiothreitol (Sigma-Aldrich, Darmstadt, Germany, CAS-3483-12-3, DTT) (reduction control), or buffer. As AT-2 was dissolved in DMSO, the solvent was added to all samples. 100 μL of the E. coli suspensions were added to the wells of a black, clear-bottom 96-well plate (Nunc, Rochester, NY). Fluorescence intensities were recorded using the Synergy H1 multi-detection microplate reader (Biotek, Bad Friedrichshall, Germany) at excitation wavelengths 405 and 488 nm and emission wavelength at 525 nm at 25 • C. The ratios of the fluorescence intensities measured at excitation wavelengths 405 and 488 nm were used to calculate the probe's oxidation state. All values were normalized to fully oxidized (AT-2treated) and fully reduced (DTT-treated) roGFP2 with the following equation [1]: with R ox being the 405 nm/488 nm ratio of oxidized (AT-2-treated) and R red of reduced (DTT-treated) roGFP2 respectively. I 488 ox and I 488 red are the fluorescence intensities of roGFP2 at 488 nm under oxidizing or reducing conditions. R is the measured 405 nm/488 nm ratio of roGFP2 in the respective bacterial strain under the respective condition. For determining the roGFP2 re-oxidation capacity in the periplasm of different E. coli mutants after a reductive pulse, cells expressing roGFP2 were harvested, washed in HEPES buffer and adjusted to an OD 600 of 1.0 in HEPES buffer. Afterwards aliquots of the cell suspension were treated with 10 mM DTT for 10 min at 25 • C and the reductant was washed out by three washing steps with HEPES buffer. The OD 600 was again adjusted to 1.0 in HEPES buffer. After transfer of 100 μL of the respective cell suspensions into a black, clear-bottom 96-well plate (Nunc, Rochester, NY), 5 mM of the substances of interest: buffer, reduced (Sigma-Aldrich, CAS-70-18-8, GSH) or oxidized (Sigma-Aldrich, CAS-27025-41-8, GSSG) glutathione, cysteine, β-mercapthoethanol or DTT were added. Addition of these agents at these concentrations did not change the overall pH of our solutions to values below 7. Fluorescence intensities were recorded as described above with 1 min 41 s intervals for 180 min at 25 • C. Fully oxidized (AT-2-treated), fully reduced (DTTtreated) and untreated cells, that were not pre-reduced were used to calibrate roGFP2. For calculation and normalization of the probe's oxidation degree at each time point, the means of all values recorded at all time points for R red , R ox , I 488 ox and I 488 red were used. Data was processed using Microsoft Excel software and GraphPad Prism.

Determination of the steady state redox potential of roGFP-iL in the periplasm of E. coli
The steady state redox potential of roGFP-iL targeted to the periplasm of E. coli WT, ΔdsbA, ΔgshA or ΔdsbAΔgshA by expression from pPT-roGFP-iL plasmid was determined as described previously for cytosolic roGFP2 [48][49][50] or Grx1-roGFP2-iL located in the endoplasmic reticulum [51]. Correct probe localization was verified by fluorescence microscopy ( Supplementary Fig. 1). Firstly, roGFP-iL was expressed as described above for 16 h at 20 • C, cells were harvested, washed in HEPES buffer (40 mM, pH 7.4) and adjusted to an OD 600 of 1.0 in HEPES buffer. The fluorescence intensity was recorded with continuous stirring every 30 s for at least 5 min at 37 • C in an FP-8500 spectrofluorometer (Jasco, Tokyo, Japan) with a fixed emission wavelength at 510 nm. The scanned excitation wavelength ranged from 350 to 500 nm. Bandwidths of both, excitation and emission were set to 5 nm. The fluorescence excitation ratios (395/465 nm) were used as measurement of probe oxidation [52]. For normalization, 1 mM AT-2 and 10 mM DTT were added to the cells achieving full oxidation or reduction of the probe. For calculation, equation [1] was modified. R ox was 395 nm/465 nm ratio of oxidized and R red of reduced roGFP-iL. I 465 ox and I 465 red represent the fluorescence intensities of roGFP-iL at 465 nm under oxidizing or reducing conditions. R is the measured 395 nm/465 nm ratio of roGFP-iL in the respective bacterial strain under the respective conditions. All values used for calculation of R ox , R red , I 465 ox and I 465 red are the mean of at least four measured time points, respectively.

Fluorescence microscopy
E. coli WT, ΔdsbA, ΔgshA or ΔdsbAΔgshA cells producing either roGFP2 or roGFP-iL from pPT plasmids cultivated as described above were harvested, washed and resuspended to an OD 600 of 1.0 in PBS buffer (1.5 mM KH 2 PO 4 , NaCl 150 mM, 2.7 mM Na 2 HPO 4 -7xH 2 O, pH 7.4). Afterwards, 2 μL of the cell suspension were spotted on a 1.5% Agarose pad covering a microscopy slide. E. coli WT cells producing roGFP2 from the pCC-roGFP2 plasmid served as cytoplasmic control. For fluorescence microscopy and image acquisition an Olympus BX51 microscope equipped with CCD camera (Retiga 3, QImaging), and LED light source (SOLA-365, Lumencor) driven by VisiView 3. × (Visitron systems) software was used. Image acquiring was carried out using a Plan-APO 100 × /1.4 NA oil objective with the following filter set: BP 450-488 nm, FT 495 nm, and BP 512-542 nm. The Image J software [53] was used for image processing.

Intracellular roGFP2-based measurements in yeast cells
YPH499 Δglr1Δgrx1Δgrx2 cells were transformed with a p415TEF-OPT1 plasmid encoding the glutathione transporter, Opt1/Hgt1, and a p416TEF plasmid encoding the indicated roGFP2 fusion construct. Cells transformed with empty plasmids served as controls for the subtraction of fluorescence background. Cells were grown at 30 • C in Hartwell's complete (HC) medium with 2% glucose and lacking leucine and uracil (-Leu-Ura) to ensure retention of both plasmids to an OD 600 of ~3.0. Cells were harvested by centrifugation at 1000×g for 3 min at room temperature and resuspended in 2 mL fresh HC -Ura-Leu medium containing 50 mM DTT to reduce all roGFP2 sensors before the start of the experiment. Cells were then incubated for 4 min at room temperature, centrifuged at 1000×g for 3 min, washed once with 2 mL fresh HC -Ura-Leu medium and finally resuspended in fresh HC -Ura-Leu medium to an OD 600 of 7.5. The cell suspension was subsequently transferred to a flatbottomed 96-well plate, with 200 μL per well. The 96-well plate was centrifuged at 30×g for 5 min so that cells formed an even, loose pellet at the bottom of each well.
To calibrate the probe, fully oxidized and fully reduced samples were generated by the addition of 20 mM N,N,N′,N′-Tetramethylazodicarboxamide (diamide, Sigma Aldrich) or 100 mM Dithiotreitol (DTT, AppliChem), respectively. Glutathione disulfide (GSSG, Sigma Aldrich) was added to experimental samples at final concentrations ranging from 1.5 μM to 100 μM. Probe responses were followed for 500 s in a BMG Labtech CLARIOstar Plus fluorescence plate reader. All experiments were performed at least three times with cells from independent cultures. The degree of roGFP2 oxidation was calculated according to equation [1] with the exception that values were recorded at 400 and 480 nm instead of 405 and 488 nm.

Purification of DsbAΔSP and roGFP2 from E. coli BL21 and enzymatic assay
For DsbA in vitro assays, DsbA was expressed as N-terminally truncated variant lacking the periplasmic signal peptide and N-terminally fused to a Strep-Tag with a TEV-cleavage site in E. coli BL21 cells. The protein was purified with Strep-affinity chromatography (for more details see Supplementary Fig. 2), the tag was removed by TEV-cleavage and the protein was dialyzed into buffer W (100 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, pH 8.0). roGFP2 was expressed from pCC_roGFP2 in E. coli MG1655 and purified as previously described [54]. DsbAΔSP was oxidized with a 10-fold molar excess of AT-2 and roGFP2 was reduced with a 10-fold molar excess of DTT for 10 min at 20 • C and then the oxidant or reductant were removed using Micro Bio-Spin™ P-30 gel columns (BioRad, Feldkirchen, Germany, #7326223) or NAP™-5 columns (Cytiva, Freiburg, Germany, #17085302) according to the supplier's protocol. For kinetic measurements, 0.2 μM roGFP2 were incubated with buffer alone, a 20-fold molar excess of oxidized DsbAΔSP, GSSG or both in a 96-well plate (Nunc, Rochester, NY).
Addition of 2 μM AT-2 or DTT served as calibration marks for fully reduced or oxidized roGFP2 for the calculation of the OxD, according to equation [1]. roGFP2 fluorescence was recorded for 18 h at 25 • C in a Synergy H1 multi-detection microplate reader (Biotek, Bad Friedrichshall, Germany) at the excitation wavelengths 405 and 488 nm and emission wavelength at 525 nm with the following settings: 20 nm bandwidth, gain 75%, intervals: 1 min 41 s, optics: bottom. Data processing was carried out as described above for cellular roGFP2 assays.

PhoA activity assay and verification of PhoA synthesis
Activity of alkaline phosphatase PhoA was measured as described previously [55] with modifications. Briefly, mutant strains were cultivated in 5 mL MOPS minimal medium with appropriate antibiotics at 37 • C without or with supplementation of 5 mM GSH or GSSG for around 18 h. Afterwards, A. dest, 5 mM GSH or GSSG were added again and the cells were incubated for one more hour. After incubation and measurement of the OD 600 , 100 mM iodoacetamide (Sigma-Aldrich, CAS-144-48-9, IAM) was added to 1 mL of the culture and cells were left on ice for 20 min. Cells were harvested (16,000×g, 4 • C, 2 min) and washed two times with 1 mL washing buffer (50 mM NaCl, 10 mM NH 4  with AU being arbitrary units, Abs being the absorption at 420 nm and 550 nm and OD 600 being the culture optical density at 600 nm. For analysis of PhoA accumulation, 1 mL of each aforementioned cultures were harvested and the pellet was resuspended in standard 1x SDS-sample buffer adjusted to 100 μL for an OD 600 of 1.0. PhoA accumulation was analyzed by SDS-PAGE (NuPAGE™ 4 bis 12%, Bis-Tris, Invitrogen, Waltham, MA, USA) and Western blot analyses were carried out using nitrocellulose membranes in the iBLOT2 system (Invitrogen, Waltham, MA, USA), both according to the supplier. For detection on Western blot, a PhoA-specific first antibody (Sigma-Aldrich, #MAB1012 1:5'000) and a goat-anti-mouse IRDye 800CW conjugate (LeiCor, Lincoln, NE, USA, #926-32210) were used (1:5'000) according to the supplier's protocol. Proteins were detected by fluorescence using the Sapphire Azure Multiimager (Azure biosystems, Dublin, CA, USA).

Analysis of E. coli growth
E. coli mutant strains were cultivated in 750 μL MOPS minimal medium in a transparent 48-well plate (Sarstedt, Nümbrecht, Germany) without or with supplementation of 5 mM GSH or GSSG at 37 • C with continuous shaking in an infinite M200 multiplate reader (Tecan instruments, Männedorf, Switzerland). Cells were inoculated to a starting OD 600 of 0.1 and bacterial growth was recorded every 30 min for 18 h.

roGFP2 is completely oxidized in the E. coli periplasm, even in the absence of the major oxidase DsbA
The redox state of cellular compartments is often assessed by targeting genetically encoded redox probes to the compartment of interest. One frequently used redox sensor is roGFP2, a GFP variant, in which two cysteines were introduced at positions 147 and 204 that can form a disulfide bond upon oxidation. This disulfide bond results in a shift in the fluorescence excitation spectrum of the probe, in which the excitation maximum at 488 nm is reversibly decreased with a simultaneous increase of the excitation peak at 405 nm. Excitation at either maxima results in emission at ~510 nm. The ratio of the fluorescence emission intensity upon excitation at 405 nm and 488 nm reflects changes in the dithiol disulfide state of the probe and, in combination with fully oxidized and reduced controls, allows the determination of the degree of the probe oxidation independent of probe concentration and thus the determination of the probe's redox potential [48,50,[56][57][58].
However, biological systems are famously not in equilibrium but in a steady state. Thus the in vivo redox state of a probe protein such as roGFP2 does not reflect a "true" redox potential (which would assume that all redox reactions have finished and thus the system is in equilibrium) but the steady state of thiol oxidation exerted on the probe by cellular redox systems reached at this point of measurement. Therefore, we use the term "steady state redox potential" to refer to the redox potential imposed on our cysteine-containing probe proteins in vivo. In order to assess the periplasmic steady state redox potential of E. coli, we targeted roGFP2 to this compartment. GFP is usually not folding correctly in the periplasm, hence the general secretory pathway (Sec) is unsuitable for targeting roGFP2 to the periplasm [59]. In contrast, E. coli's twin-arginine translocase (or Tat system) transports completely folded substrates over the inner membrane. Tat-translocation requires the presence of an N-terminal signal peptide, characterized by an essential twin-arginine motif [60,61]. To target roGFP2 to the periplasm via the Tat system, we expressed the protein as an N-terminal torA SS fusion in the KEIO wildtype E. coli BW25113 (referred to as WT) [44] (Fig. 1A). We verified translocation of roGFP2 into the periplasm using fluorescence microscopy (Fig. 1B, left panel). Determination of the roGFP2 oxidation state (OxD) revealed complete oxidation of the probe in the periplasm of E. coli WT (Fig. 1C, left panel). As the DsbA/DsbB pair is the major system for oxidative protein folding in the periplasm of E. coli [3], roGFP2 was also targeted to the periplasm of cells lacking DsbA (Fig. 1B, right panel). Surprisingly however, periplasmic roGFP2 was still almost fully oxidized in the absence of DsbA (Fig. 1C, right panel), suggesting the presence of an alternative mechanism for the introduction of disulfide bonds in this compartment. In both cases, due to the almost complete oxidation of the probe, we were not able to determine the steady state redox potential for the roGFP2 dithiol disulfide couple.

The steady state redox potential of the periplasm is significantly more reducing in cells lacking DsbA
In order to determine the periplasmic steady state redox potential, we turned to roGFP variants with a more oxidizing standard redox potential [52]. These probes, termed roGFP-iX were developed by insertion of a single amino acid (denoted by the 'X') adjacent to cysteine 147 resulting in a decreased stability of the disulfide bond and midpoint potentials of − 229 to − 246 mV compared to − 280 mV of roGFP2. These roGFP-iX variants have been successfully used to monitor the oxidation state of the endoplasmic reticulum, a highly oxidized compartment in eukaryotic cells and a major site of oxidative protein folding [51,56,62]. Thus, roGFP-iL with a midpoint potential of − 229 mV was expressed and targeted to the periplasm of the WT and ΔdsbA strain (Fig. 1A, B, right panel). In contrast to roGFP2, roGFP-iL was only around 50% oxidized in the WT periplasm (Fig. 1B, left panel), allowing the calculation of the steady state redox potential using the Nernst equation as described before [50]. In the WT the steady state redox potential of roGFP-iL in the periplasm was − 228 mV (Fig. 1E, left panel). To confirm that the final oxidation ratio of roGFP-iL was reached within the timeframe of our measurement, we determined the in vivo oxidation state of roGFP-iL again 2 and 7 h later but did not find a difference ( Supplementary  Fig. 3). Not surprisingly, the lack of DsbA resulted in a more reduced roGFP-iL probe (Fig. 1E, right panel). The shift in the steady state redox potential was around 14 mV to − 243 mV, supporting the role of DsbA in oxidative protein folding, however, the relatively small size of the shift suggests the presence of an alternative mechanism for the introduction of disulfide bonds.

Periplasmic thiol oxidation is significantly impaired but not abrogated in a ΔdsbA-mutant
Our experiments with roGFP2 and roGFP-iL reflected the steady state of thiol oxidation in the periplasm. However, both stress and physiological situations are often accompanied by increased protein secretion, requiring correct oxidative protein folding and hence, may affect the redox homeostasis of the periplasm. In order to test the capacity of the periplasm to restore its redox balance after reductive challenge, we evaluated the re-oxidation kinetics of periplasmic roGFP2 after a reductive pulse. To explore the role of DsbA in this process, we analyzed the re-oxidation of roGFP2 in cells lacking this oxidoreductase. In these experiments, we treated WT and cells lacking DsbA, both expressing roGFP2 in the periplasm with a pulse of DTT. This was followed by the removal of the reductant and the recording of the oxidation state of roGFP2 over time ( Fig. 2A). In the WT, periplasmic roGFP2 oxidation starts immediately after DTT removal and restores maximum oxidation within 150 min. Cells that lack DsbA also showed sensor re-oxidation, however, roGFP2 was not restored to maximum oxidation in the time frame of our measurement (Fig. 2B).
Additionally, the re-oxidation rate, calculated as the linear slope after DTT removal, was significantly lower in DsbA-deficient cells, however, there was still re-oxidation capacity present (Fig. 2C). These findings suggest that there is an additional factor playing a role in periplasmic redox homeostasis, other than DsbA. To further explore our hypothesis, we used cell lysates derived from E. coli WT and DsbAdeficient cells and explored their capacity to re-oxidize DTT-reduced, purified roGFP2 (Fig. 2D). Similar to the cell re-oxidation assay, we calculated the oxidation rate from the linear slope within the first 2 h of measurement. Although roGFP2 oxidation by cell lysate was slow, when compared to intact cells, lysate from WT was significantly faster in catalyzing roGFP2 re-oxidation than a ΔdsbA lysate. However, in line with our previous findings, indicating another factor capable of catalyzing disulfide bond formation in the periplasm, the ΔdsbA lysate was still significantly faster than a buffer control (Fig. 2E, F). The rather slow oxidation of roGFP2 in vitro may be explained by the fact that DsbA preferably introduces disulfide bonds in unfolded proteins entering the periplasm over folded ones [63].

Oxidized glutathione rescues periplasmic roGFP2 re-oxidation velocity in cells lacking DsbA
Based on our observations, we suspected that glutathione (GSH) and its oxidized dimer GSSG might influence the periplasmic redox balance in E. coli and thus might act as part of an auxiliary system in oxidative folding. Glutathione is one of the most important redox-active molecules inside many cells and it has been shown to be present in the periplasm [33,37].
We thus analyzed the periplasmic capacity for re-oxidizing roGFP2 in the presence of glutathione after a reductive pulse. E. coli WT and DsbAdeficient cells producing periplasmic roGFP2 were thus reduced with DTT, and, after reductant removal, roGFP2 oxidation was recorded in the presence of 5 mM GSH or GSSG (Fig. 3A). Surprisingly, addition of GSH to WT cells resulted in an accelerated roGFP2 re-oxidation in the periplasm, whereas GSSG supplementation had no discernible influence (Fig. 3B). In contrast, in cells that lack DsbA, GSH and GSSG acted in a more expected way; while GSSG supplementation rescued roGFP2 reoxidation, GSH had no influence on the periplasmic redox dynamics in ΔdsbA (Fig. 3C). Calculation of the re-oxidation rate confirmed that while GSH significantly speeds up re-oxidation of periplasmic roGFP2 in the WT, GSSG significantly accelerates the re-oxidation rate of roGFP2 in DsbA-deficient cells, essentially to WT level (Fig. 3D). Based on these findings, we asked ourselves if DsbA is interacting with glutathione, causing the observed opposite responses to GSH and GSSG in WT and DsbA-deficient E. coli cells.

Oxidized glutathione does not directly interact with the oxidase DsbA
To investigate if glutathione interacts with DsbA, we investigated the effect of GSSG on DsbA-dependent roGFP2 oxidation in vitro. Purified DsbA was oxidized and added to reduced roGFP2. Similar to the aforementioned lysate assay, the oxidation rate of roGFP2 by DsbA is rather slow compared to in vivo re-oxidation, although significantly higher compared to buffer alone (Fig. 4A, B). The oxidation rate of roGFP2 in the presence of GSSG alone was slightly slower, but still comparable to roGFP2 oxidation by DsbA. Adding both GSSG and DsbA at the same time did not even double the probe's oxidation rate, indicating an additive effect of GSSG and DsbA and not a GSSG-driven catalytic action of DsbA.
To further confirm the incapability of DsbA to perform GSSGdependent roGFP2 oxidation despite the observed additive effect, we used the cytosol of genetically manipulated yeast cells as "cellular test tubes". This assay is performed in yeast cells, lacking the glutathione reductase (Glr1) and both cytosolic class I dithiol glutaredoxins (Grx1, Grx2), while simultaneously expressing Opt1, a glutathione transporter [46]. This in vivo system should enable us to monitor a potential DsbA-catalyzed oxidation of roGFP2 by GSSG in the absence of any E. coli-specific factors. For this, we engineered a roGFP2-DsbA fusion construct and expressed it in the cytosol of the ΔglrΔgrx1Δgrx2 yeast cells, following the exogenous application of GSSG concentrations between 1.5 and 100 μM. Unfused roGFP2 served as negative and roGFP2 fused to homo sapiens glutaredoxin (roGFP2-HsGrx) as positive control for direct interaction with GSSG. In this assay, GSSG-driven roGFP2 oxidation did not depend on the oxidase DsbA (Fig. 4C, D). However, we cannot exclude that roGFP2 is an inappropriate substrate of DsbA in the fusion construct and therefore roGFP2 may not be oxidized by DsbA in this assay. In contrast, the roGFP2-HsGrx fusion probe strongly reacted to the addition of GSSG (Fig. 4E).
Both the in vitro and the "cellular test tube" approach strongly suggest that DsbA does not interact with glutathione itself, indicating that the role of glutathione in the periplasmic redox homeostasis is independent of the known mechanism for disulfide bond formation in E. coli.

Monothiols accelerate roGFP2 re-oxidation in a DsbA-dependent mechanism
While we excluded the interaction of GSSG with DsbA, we showed that the addition of GSH accelerated the re-oxidation rate of periplasmic roGFP2 after a reductive pulse in a DsbA-dependent manner (Fig. 3B, D). To investigate whether this effect is limited to GSH, we tested the influence of other reduced monothiols and dithiols. Cystein, β-mercaptoethanol, and DTT were added to E. coli WT and ΔdsbA cells and roGFP2 re-oxidation dynamics in the periplasm were measured as described above. The addition of the monothiols cysteine and β-mercaptoethanol to WT significantly accelerated roGFP2 oxidation similar to GSH, as measured by the time after which the probe reached full oxidation. In contrast to monothiols, addition of the dithiol DTT to WT completely inhibited roGFP2 re-oxidation. In the ΔdsbA mutant, the presence of monothiols had no significant impact (Fig. 5C, D). Overall, these findings suggest a DsbA-dependent effect of monothiols, accelerating thiol oxidation in the periplasm, even though we did not observe direct interaction of oxidized glutathione with DsbA itself.

Endogenous glutathione is involved in stabilizing and maintaining the periplasmic redox state
All our experiments thus far were performed with exogenous glutathione. But we also wondered about the role of endogenous glutathione, synthesized in E. coli's cytoplasm. For this, we determined the oxidation state of periplasmic roGFP-iL in cells lacking GshA, the first enzyme of E. coli's glutathione biosynthesis pathway. In glutathione-free media, these cells do not contain GSH [26,64]. Confirming our observation with exogenous GSH, our experiment revealed a slightly, but significantly lower oxidation of roGFP-iL in the periplasm of cells lacking GSH. The steady state redox potential shifted from around − 228 mV (WT) to − 233 mV, a more reducing state, in the absence of GSH (Fig. 6A, B). Nevertheless, the shift in ΔgshA was not as pronounced as in cells lacking DsbA, which had a steady state redox potential of around − 243 mV (Fig. 1E).
We also asked whether the presence of endogenous GSH leads to faster re-oxidation of roGFP2. To address this question, we analyzed the capacity to restore the redox balance after reductive challenge in GSHdepleted cells producing periplasmic roGFP2 as described before. Our data indicates that roGFP2 oxidation rate and end oxidation state in the periplasm of GSH-deficient cells was comparable to WT (Fig. 6C, D) suggesting GSH is not essential for recovery after a reductive challenge.
Next, we asked what happens when both, DsbA and GSH are missing. The redox state of roGFP-iL in a ΔgshAΔdsbA strain showed a steady state periplasmic redox potential (ca. − 244 mV) similar to the steady state redox potential in the dsbA single mutant (ca. − 243 mV) (Figs. 1  and 6B). Intriguingly however, using a ΔgshAΔdsbA strain in a periplasmic roGFP2 re-oxidation assay revealed that exogenous GSSG did not completely restore WT oxidation rate and final oxidation state (Fig. 6E, F) contrary to cells lacking solely DsbA (Fig. 3). Adding GSH to  Fig. 2 and are presented here for context. Significance test was performed using one way ANOVA. ns > 0.05, **p < 0.01, ****p < 0.0001.
a ΔgshAΔdsbA strain even decelerated periplasmic roGFP2 oxidation, something we did not observe in the ΔdsbA single mutant (Fig. 3). These experiments indicate that a fine-tuned balance of reduced and oxidized glutathione plays a role in stabilizing and maintaining the redox environment in the periplasm of E. coli, and cells lacking both DsbA and endogenous glutathione can neither fully utilize the oxidative power of exogenous GSSG nor compensate the reductive power of GSH in their periplasm.

Endogenous glutathione is involved in disulfide bond formation in E. coli's own periplasmic proteins
The observed influence of GSH on growth phenotypes (Supplementary Fig. 4) suggests that our previous observations of disulfide bond formation in heterologously expressed roGFP-based redox sensors also apply to endogenous DsbA substrates. One well-characterized substrate of DsbA is alkaline phosphatase PhoA. PhoA is only active upon formation of intramolecular disulfide bonds essential for correct protein folding. The activity of this enzyme can be measured in a colorimetric assay using the substrate para-nitrophenolphosphate (Fig. 7A) [18,55,65]. We thus tested the activity of alkaline phosphatase PhoA in different E. coli deletion strains grown in MOPS minimal medium. As described above, the lack of GSH, although without effect on re-oxidation of roGFP2 in the periplasm, slightly shifted the periplasmic redox homeostasis to more reducing conditions, as seen in the lowered steady state redox potential of roGFP-iL in those cells (Fig. 6). In accordance with this, PhoA activity was slightly, but significantly lowered in GSH-deficient cells compared to WT. Also, in line with our previous results (Fig. 1), cells lacking DsbA showed only poor PhoA activity, and the same was true for cells lacking both DsbA and GSH (Fig. 7B).
To test whether the observed reduction in PhoA activity is dependent on glutathione levels in the periplasm, we also measured PhoA activity in a cydD mutant. This mutant can still synthesize glutathione, but is not able to produce the inner membrane ABC-transporter CydDC that transports glutathione from the cytosol into the periplasm and hence suffers from reduced periplasmic glutathione levels [33,66,67]. Again, we observed a slight, but significant reduction of PhoA activity in this strain, although less pronounced than in cells completely lacking glutathione biosynthesis (Fig. 7B). Conversely, blocking glutathione transport from the periplasm into the cytosol by deleting GsiC, the inner membrane component of the GsiA-D ABC Transporter [35,36][ did not result in significant changes in PhoA activity compared to the WT.
Next, we supplemented the growth medium with exogenous GSSG to test if, in accordance with the re-oxidation assays (Fig. 3) and the rescue of the growth phenotype ( Supplementary Fig. 4), it can assist in oxidative folding of PhoA in the absence of DsbA. Indeed, GSSG addition restored PhoA activity back to WT level (Fig. 7C). Furthermore, addition of exogenous GSSG also increased PhoA activity in the WT, contrary to what we observed with roGFP2, suggesting that the role of periplasmic glutathione in oxidative folding differs, depending on the protein in question (Fig. 3, Supplementary Fig. 4).
Proteins that are not correctly folded are usually unstable in the periplasm [68], and thus no PhoA protein could be detected on a western blot in DsbA-deficient cells (Fig. 7D). External GSSG addition prevented PhoA from degradation in a ΔdsbA mutant.
Taken together, our data indicates a significant role for glutathione, not only in oxidative protein folding of periplasmic proteins, but also by balancing and maintaining the periplasmic redox homeostasis.

Fig. 4. Oxidation of roGFP2 by oxidized glutathione (GSSG) is not catalyzed by DsbA. (A)
Oxidation of purified reduced roGFP2 by AT-2-oxidized DsbA (red), GSSG (orange) or both (purple). Purified roGFP2 was reduced with DTT, the reductant was washed out and DsbA and/or GSSG were added in a 20-fold molar access. PBS was used to track air oxidation of roGFP2. DTT served as reduction and AT-2 as oxidation controls for calculation of OxD. (B) The initial re-oxidation velocity was calculated as linear regression for the first 1.25 h after the start of measurement (dashed lines A). Values (circles) are the mean of three technical replicates recorded in a minimum of six independent repeats and error bars depict standard deviation. Significance test was performed using one-way ANOVA ns p > 0.05 *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Discussion
The DsbA/B system is the major thiol oxidation system in the periplasm of E. coli and together with the DsbC/D thiol disulfide isomerase system forms the oxidative folding machinery [3,11]. The standard redox potentials of those systems are known [7,69], but the in vivo steady state redox potential seen by protein thiol disulfide pairs in the periplasm is unknown. Here, we used the genetically encoded proteins roGFP2 and roGFP-iL targeted to the periplasm to directly probe the thiol redox homeostasis in this compartment. When expressed in the cytosol of E. coli, the probe roGFP2, with a midpoint redox potential of − 280 mV, is almost completely reduced [48]. However, we found that it is virtually fully oxidized, when expressed in the periplasm, suggesting that its engineered cysteine residues are completely oxidized by the oxidative folding machinery.
To our surprise, roGFP2 was also virtually fully oxidized in cells lacking the major thiol oxidase in the periplasm, DsbA. We thus also analyzed the re-oxidation capacity after a reductive pulse, as the oxidation state in an unperturbed cell solely reflects the steady state. And in the absence of DsbA we did indeed find a significantly diminished re-oxidation velocity underlining the importance of DsbA for oxidative folding.
Using roGFP-iL, we were able to determine the redox potential imposed onto a thiol pair in the periplasm: − 229 mV. This is consistent with the observed, virtually complete oxidation of roGFP2 in the periplasm, which, based on the calculated value at this steady state redox potential should be 98.2% oxidized (Fig. 1C). In the absence of DsbA, the steady state redox potential shifted to a more reducing − 243 mV, again consistent with our findings with roGFP2, which should be 94.7% oxidized under these conditions (Fig. 1C). In a previous study, Messens et al. used an Ag/AgCl electrode to measure the redox potential in E. coli WT periplasmic extracts [70]. Typically, an electrode will measure the redox potential of all species it can chemically interact with and in this case the redox potential was determined to be − 165 mV. Unexpectedly, in their setting, removal of DsbA shifted the redox potential to an even more oxidizing redox potential. Our finding that the steady state thiol disulfide redox potential is significantly below the overall redox potential observed by Messens et al. with an electrode could explain their seemingly paradoxical finding, since in a ΔdsbA strain a component that is reducing in comparison to the overall redox potential measured by the electrode is removed from the overall redox pool. In our case, removal of DsbA did indeed result in an overall more reducing steady state redox potential.
In concordance with our results in the periplasm, roGFP2 targeted to the ER is fully oxidized, as well. Similarly, a roGFP2-iL-Grx1 fusion redox probe targeted to the ER of Arabidopsis thaliana cells lacking Ero1/ 2, the functional homolog of DsbB in eukaryotes was more reduced, and re-oxidation after a reductive pulse was inhibited by the lack of Ero1/2 [51].
After determining the steady state redox potential (a thermodynamic parameter), we focused our attention on the kinetics of disulfide bond formation in the periplasm. We thus monitored the re-oxidation of roGFP2 in the periplasm. While it would be ideal to use the same probe that was used to determine the redox potential in our kinetic experiments as well, the dynamic range of roGFP-iL, especially in a ΔdsbA strain, would be too small to determine a reliable initial velocity The initial re-oxidation velocity was calculated as linear regression for the first 25 min after start of the measurement (dashed lines A). Values (circles) are the mean of three technical replicates recorded in three independent repeats and error bars depict the standard deviation of those means. Significance test was performed using one-way ANOVA ns p > 0.05, **p < 0.01. (Fig. 1D). While roGFP2 re-oxidation was diminished in ΔdsbA, it was not absent, suggesting to us the presence of an alternative pathway for disulfide bond formation. As glutathione is one of the major redox buffers in cells, we assumed a possible role for the small molecule in periplasmic redox homeostasis. Glutathione's cytosolic functions have been studied extensively and, until a few years back, GSH was thought to be absent from the periplasm [33]. However, relatively high glutathione levels in the periplasm of E. coli were discovered recently, and since, there is an ongoing debate on its function in this compartment [33,37,42]. Interestingly, cells lacking the transporter for GSH from the cytosol to the periplasm, CydDC, exhibit several phenotypes, including DTT sensitivity and swarming defects, also found in cells deficient in oxidative protein folding. This is an additional indication for a role of GSH in redox homeostasis and oxidative folding in the periplasm [33,34,43,71]. Accordingly, we found that re-oxidation of periplasmic roGFP2 in ΔdsbA was restored by the addition of exogenous GSSG, while it was not influenced in WT. Somewhat surprisingly, adding exogenous GSH to the WT accelerated roGFP2 re-oxidation, but not in a ΔdsbA strain. We also observed this seemingly paradoxical acceleration of thiol re-oxidation in WT by other reducing monothiols like cysteine or β-mercaptoethanol. However, we did not observe direct interactions of DsbA with glutathione in vitro and in our yeast cell experiments. This observation that reductants, particular in catalytic quantities, can accelerate oxidation reactions is not uncommon. A classic example is the facilitation of the hydroxyl radical-producing Fenton reaction by ascorbate [72].
We next assessed the role of endogenous glutathione in the redox balance of the periplasm. And in line with our observations with exogenous GSH, the oxidation state of roGFP-iL was slightly shifted to a more reducing state in ΔgshA, indicating a role for endogenous glutathione in the periplasmic redox homeostasis as well. It should be noted, however, that roGFP2 re-oxidation in a GSH-deficient mutant was comparable to WT, suggesting that the rate limiting step is DsbA-dependent oxidation and not the resolution of a potential glutathione adduct. Taken together, the presence of GSH and GSSG in the periplasm, driven by the  Fig. 2 and are presented again for context. Significance tests in A and B were performed using ANOVA-test. ****p < 0.0001. (C) Reoxidation of roGFP2 in the periplasm of E. coli WT and ΔgshA. The assay was performed as described in Fig. 2. One representative example out of at least four individual repeats is shown. Error bars represent standard deviation of technical triplicates. (D) The initial re-oxidation velocity was calculated from linear regression for the first 25 min after start of the measurement (dashed lines in B). Values shown as circles are the mean of three technical repeats recorded in at least four independent assays. (E) Re-oxidation of periplasmic roGFP2 in E. coli ΔgshAΔdsbA. One representative example out of at least seven individual repeats is shown. Error bars represent standard deviation of technical triplicates. (F) The initial re-oxidation velocity was calculated from linear regression in the first 25 min after start of the measurement (dashed lines). Values (circles) are the mean of three technical replicates recorded in a minimum of seven independent repeats. Error bars represent the standard deviation. Significance test was performed using one way ANOVA. **p < 0.01, ****p < 0.0001. Values for the WT were already shown in Fig. 3 and are presented again for context. availability of exogenous and endogenous glutathione, seems to be important for the fine-tuning of the periplasmic redox potential.
Expression of a non-native redox sensor might not reflect the natural redox state in the periplasm or might even influence it by diverting oxidative power available for the formation of native disulfide bonds. We thus analyzed the activity of PhoA, a native E. coli protein, which depends on oxidation by DsbA for its activity, in different mutants. This approach revealed that reduced periplasmic glutathione levels indeed resulted in significantly lower PhoA activity, in line with the more reduced roGFP-iL redox state in ΔgshA. While PhoA's oxidative folding is clearly influenced by the presence or absence of GSH, not all DsbA substrates seem to be influenced by glutathione. RNaseI folding and isomerization by DsbA/ DsbC e.g., was not influenced by the loss of GSH [70].
In order to understand the role of glutathione in the periplasm, it is helpful to have a look at the role of GSH in the eukaryotic ER, for which, in contrast to bacteria, more studies are available. In the ER, the glutathione concentration is around 15 mM, higher than in whole cell lysates with around 7 mM [41]. Similar to the periplasm, the GSH:GSSG ratio in the ER is lower (3:1 to 1:1) compared to the cytosol (100:1), resulting in a more oxidizing environment [73]. It is discussed whether GSH is the reductive power for disulfide isomerization by PDI [74] and high GSSG levels could act as oxidant reservoir; however, it is still unclear if GSSG itself is able to oxidize PDI [20,75]. In the periplasm of Gram-negative bacteria, DsbB recycles DsbA, but in contrast to Ero1, it uses the respiratory chain as electron sink. As aforementioned, we observed accelerated roGFP2 re-oxidation in the WT, but not in ΔdsbA, when adding monothiols to the cells and the opposite effect for GSSG, compensating the lack of DsbA, raising the question whether DsbB is somehow regulated by glutathione. However, analyzing a dsbB mutant strain regarding its oxidation state and re-oxidation capacity in presence or absence of GSH or GSSG revealed that GSSG was still able to complement for the loss of DsbB, indicating a DsbB-independent mechanism ( Supplementary Fig. 5).
Overall, we showed that oxidized glutathione can compensate for the loss of DsbA by an unknown mechanism. One possibility is that oxidized glutathione directly oxidizes roGFP2 and other reduced proteins, however previous studies [54] and the current in vitro and yeast data show that direct roGFP2 oxidation by glutathione is very inefficient. Another possibility is the presence of a yet unidentified redox factor in the periplasm that can catalyze oxidative protein folding in the absence of DsbA. One possibility is DsbC, and it has been suggested that reduced glutathione can react with the isomerase, especially by providing reductive power when DsbD is missing [33,34]. However, Messens et al. [70] could also show that DsbC alone was not able to substitute for DsbA in folding of RNaseI. In the ER up to 20 different oxidoreductases, for example the peroxiredoxin Prx4 or the glutathione peroxidase-like enzymes Gpx7 and Gpx8 are found besides PDI and for at least some of them it has been proposed that they possibly oxidize PDI [36,76,77]. Our data suggests that E. coli has a similar backup system for DsbA, presumably a glutaredoxin-like protein, which is coupled to the periplasm's glutathione pool, providing either oxidizing or reducing power. Additionally, different redox couples and redox active small molecules, such as cysteine/cystine could influence the redox homeostasis in the periplasm. PhoA activity depends on oxidation by DsbA and activity of PhoA can be monitored using p-nitrophenylphosphate (pNPP) resulting in formation of the yellow product p-nitrophenol. (B) PhoA activity was assayed in cells cultivated in MOPS medium. Release of p-nitrophenol from the PhoA substrate pNPP was measured and relative PhoA activity in % of WT activity was calculated. (C) GSSG restores PhoA activity to WT level in cells lacking DsbA. Cells were cultivated in the presence of 5 mM GSSG for 18 h at 37 • C. Then fresh GSSG (5 mM) was added and the cells were incubated for 1 h. PhoA activity was assayed as described above. (D) PhoA accumulation was analyzed by SDS-PAGE and western blot in cells cultivated as described in (C). For western blot a PhoA specific first and a fluorophore-coupled second antibody were used and fluorescence signal was scanned. Values in B and C are the mean of at least three independent replicates of duplicate assays with error bars representing the standard deviation. Significance test was performed using one way ANOVA. ns p > 0.05 *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) L.R. Knoke et al.
Taken together, our data underlines the importance of glutathione as a player in redox homeostasis not only in the cytosol, but also in oxidative cellular compartments and it shows that its role in oxidative protein folding did already evolve in bacteria.

Author contributions
LRK and LIL designed the study. LRK planned and performed most of the experiments. The yeast cell assays were designed and conducted by JZ and BM. JFS performed initial experiments, established the reoxidation assay and constructed the pPT and pPT_roGFP2 plasmids. NL and BC purified proteins and BC assisted with DsbA in vitro assays. LRK and LIL wrote the manuscript. All authors consulted on the manuscript and approved the final version.

Funding
LIL acknowledges funding from the German Research Foundation (DFG) through grant LE2905-2 and additional funding through the InnovationsFoRUM Host-Microbe-Interaction IF-018N-22-TP8. We acknowledge support by the Open Access Publication Funds of the Ruhr-Universität Bochum.

Data availability statement
The data supporting the findings of this study is presented within the article and its supplementary materials. Strains and plasmids constructed for this study are available upon request.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.