The Industrial Organism Corynebacterium glutamicum Requires Mycothiol as Antioxidant to Resist Against Oxidative Stress in Bioreactor Cultivations

In aerobic environments, bacteria are exposed to reactive oxygen species (ROS). To avoid an excess of ROS, microorganisms are equipped with powerful enzymatic and non-enzymatic antioxidants. Corynebacterium glutamicum, a widely used industrial platform organism, uses mycothiol (MSH) as major low molecular weight (LMW) thiol and non-enzymatic antioxidant. In aerobic bioreactor cultivations, C. glutamicum becomes exposed to oxygen concentrations surpassing the air saturation, which are supposed to constitute a challenge for the intracellular MSH redox balance. In this study, the role of MSH was investigated at different oxygen levels (pO2) in bioreactor cultivations in C. glutamicum. Despite the presence of other highly efficient antioxidant systems, such as catalase, the MSH deficient ΔmshC mutant was impaired in growth in bioreactor experiments performed at pO2 values of 30%. At a pO2 level of 20%, this growth defect was abolished, indicating a high susceptibility of the MSH-deficient mutant towards elevated oxygen concentrations. Bioreactor experiments with C. glutamicum expressing the Mrx1-roGFP2 redox biosensor revealed a strong oxidative shift in the MSH redox potential (EMSH) at pO2 values above 20%. This indicates that the LMW thiol MSH is an essential antioxidant to maintain the robustness and industrial performance of C. glutamicum during aerobic fermentation processes.


Introduction
The Gram-positive soil bacterium Corynebacterium glutamicum is widely used as anindustrial workhorse primarily for the production of L-glutamate and L-lysine [1], and has been genetically engineered as a broad platform for production of several important industrial products [2,3]. Currently, C. glutamiucm is mostly used for aerobic production processes, but its facultative anaerobic metabolism allows to design efficient two-stage processes for the production of reduced chemicals including an aerobic growth phase and an anaerobic production phase [4,5].

Strains, Media, and Culture Conditions
The strains used in this study were C. glutamicum ATCC13032 (WT) [37], the MSH-deficient C. glutamicum ∆mshC deletion mutant [26], and the Mrx1-roGFP2 redox biosensor expressing strains C. glutamicum WT_Mrx1-roGFP2 and C. glutamicum ∆mshC_Mrx1-roGFP2 [29]. C. glutamicum strains were pre-cultured in 2xTY medium at 30 • C in 500 mL shake flasks. Prior to inoculation of the main-culture, cells of an overnight culture were washed twice with 100 mM potassium phosphate buffer (pH 7.0). C. glutamicum main cultures were grown in CGXII minimal medium [1] supplemented with 10 g L −1 or 20 g L −1 glucose as a carbon source for growth experiments in shake flasks and bioreactors, respectively.
Growth experiments in bioreactors were performed aerobically at 30 • C as 1 L cultures in 1.5 L jars in a BIOSTAT ® B fermentation system (Sartorius, Goettingen, Lower Saxony, Germany), as described previously [38]. The pH was maintained at 7.0 by online measurement using a standard pH electrode (Mettler Toledo, Giessen, Hessen, Germany) and addition of 4 M KOH and 4 M H 2 SO 4 . Partial oxygen pressure (pO 2 ) was measured online using a polarimetric oxygen electrode (Mettler Toledo), and was adjusted to pO 2 values provided in the text in a cascade by stirring at 200 to 800 rpm as well as by mixing N 2 and air for the inlet gas. For anaerobic condition (pO 2 = 0%), 100% N 2 was used until a pO 2 of 0% was reached. When required, AF204 antifoam agent (Sigma, MI, USA.) was added manually. The data were collected with the software MFCS (Sartorius BBI Systems, Goettingen, Lower Saxony, Germany). Growth in shake flasks and bioreactors was followed by measuring the optical density (OD 600 nm ).

Fluorescence Measurements of Mrx1-roGFP2 Biosensor Oxidation In Vitro and In Vivo
For testing suitable settings for fluorescence measurements of Mrx1-roGFP2 redox biosensor oxidation, C. glutamicum strains harboring genomic integrated Mrx1-roGFP2 [29] were pre-cultured in 2xTY medium until the stationary phase. For preparation of crude cell extracts, cells were harvested by centrifugation (4000 rpm, 10 min., 4 • C), washed twice in potassium phosphate buffer (100 mM, pH 7.0), and finally resuspended in 1 mL of the respective buffer solution. Disruption of the cells was conducted using a Ribolyzer (Precellys TM Control Device, Bertin Technologies, Montigny-le-Bretonneux, Department Yvelines, France) at 6000 rpm, four times for 30 s each. Cell debris were removed by centrifugation (12,000 rpm, 20 min; 4 • C) and 180 µL of the supernatant was transferred to black, flat-bottomed 96-well microplates (Thermo Fisher Scientific, Dreieich, Hessen, Germany) for further fluorescence analysis using a fluorescence spectrophotometer (SpectraMax iD3, Molecular Devices LLC, San Jose, CA, USA). After the addition of 20 µL oxidants (50 mM diamide), reductants (100 mM Dithiothreitol (DTT), and 100 mM potassium phosphate buffer for fully oxidized, fully reduced, and non-treated control samples, respectively, cells were incubated for 15 min at 30 • C as described previously [29]. Subsequently, excitation scans were conducted (360 nm-470 nm) by setting an emission wavelength of 510 nm. For in vivo fluorescence measurements, C. glutamicum strains expressing the Mrx1-roGFP2 biosensor were harvested by centrifugation (4000 rpm, 4 min) and washed in 100 mM potassium phosphate buffer (pH 7.0). Finally, an optical density of 40 was adjusted and 180 µL of the cell suspension was transferred to black, flat-bottomed 96-well microplates for fluorescence analysis. To determine the maximum and minimum oxidation ratios, 20 µL of different concentrated cumene hydroperoxide (CHP) and DTT solutions were added for oxidation and reduction of the biosensor probe, respectively, until the respective ratio reached its minimum and maximum value. For samples, 20 µL potassium phosphate buffer was added instead. Mrx1-roGFP2 fluorescence intensity was recorded at an emission intensity of 510 nm upon excitation at 380 nm and 470 nm. The corresponding biosensor oxidation degree (OxD) was calculated by normalizing to fully reduced as well as oxidized controls, as described previously [29,39,40], with the following Equation (1): Here, I380 sample and I470 sample represent the measured fluorescence intensities received for an excitation at 380 nm and 470 nm, respectively. Fully reduced and oxidized controls are given by I380 red , I470 red and I380 ox , I470 ox , respectively. Calculated OxD values; the standard midpoint redox potential of roGFP2 (E o roGFP2 = −280 mV) [41]; Faraday constant (F: 96,485 C mol −1 ) of electric charge per mole of electrons, where 2 refers to the number of electrons transferred during the redox reaction; the respective temperature in Kelvin (T: 303.15 K); and the universal gas constant (R: 8.314 J K −1 mol −1 ) were used in order to determine the MSH redox potential (E MSH ) via the Nernst Equation (2):

Statistical Analysis
Analysis of one-way variance (ANOVA) with Tukey's test was used to assess differences of biosensor oxidation degrees derived from C. glutamicum WT and MSH-deficient mutant strains harboring the genetically encoded biosensor Mrx1-roGFP2. Differences were considered statistically significant when p < 0.01.

The MSH-Deficient Mutant is Susceptible to Elevated Oxygen Concentrations
To compare growth of C. glutamicum WT and the ∆mshC mutant, batch cultivations in stirred bioreactors were performed at a pO 2 value of ≥30% (regulated in cascade via the stirring rate) and a constant pH of 7.0. These are common conditions used for production and physiological studies with C. glutamicum [42][43][44][45]. Growth of C. glutamicum WT proceeded with a rate of 0.37 h −1 and the cells reached a final OD 600 of 29 after 24 h of cultivation ( Figure 1a). Growth of the MSH-deficient ∆mshC mutant proceeded slower than C. glutamicum WT within the first hours of the bioreactor cultivation, resulting in a cessation of growth after 4 h (Figure 1b). During the course of cultivation with C. glutamicum WT, a stable pO 2 value of 30% was reached after an expected initial phase with a higher pO 2 ( Figure 1a). For the MSH deficient strain, however, the decreased oxygen demand resulted in pO 2 values always above the minimal pO 2 of 30% ( Figure 1b). This decreased oxygen demand of the ∆mshC mutant can be explained by the impaired growth during the first 4 h of cultivation. The rise of pO 2 after 4 h of cultivation coincidences with the stop of growth of the MSH-deficient mutant ( Figure 1b). Moreover, strong foam formation was observed at this phase of the bioreactor cultivation with the ∆mshC mutant. Taken together, the growth deficit of the ∆mshC mutant revealed its susceptibility towards elevated oxygen concentrations present already during the initial phase of the bioreactor cultivation, when a control strategy was chosen to keep pO 2 values ≥30%.
To avoid high pO 2 values during the initial phase of the cultivations, a different strategy for pO 2 control was tested for the ∆mshC mutant, thanks to its susceptibility for elevated pO 2 levels. By mixing air with nitrogen, pO 2 in the bioreactor was adjusted to different levels at an initially constant stirring rate of 400 rpm. The results showed that, at a pO 2 of max 20%, growth of the ∆mshC mutant proceeded with slightly reduced growth rate of 0.31 h −1 (Figure 1d) when compared with a growth rate of 0.37 h −1 (Figure 1c To test for negative effects of elevated pO 2 on growing cultures in bioreactors, C. glutamicum WT and the ∆mshC mutant were cultivated initially at a pO 2 of max 20% for 3 h until optical densities of 7 and 6, respectively (Figure 1e,f). Subsequently, the pO 2 was increased in a single step to 40%. Whereas growth of C. glutamicum WT continued after the increase of pO 2 to a final OD of 28 in the course of cultivation (Figure 1e), growth of the ∆mshC mutant immediately stopped, resulting in a final OD of 9 ( Figure 1f). These experiments showed that the MSH-deficient ∆mshC mutant is highly sensitive to pO 2 levels above 20%. As MSH protects the cells against oxidative stress, these results indicate that, already at slightly increased pO 2 levels, oxidative stress occurs during bioreactor cultivation. Finally, bioreactor experiments were carried out with an initial pO 2 value of 20% during the first 3 h and a second fermentation phase with a pO 2 value of 40% (e,f). Growth was monitored by measuring the optical density at 600 nm. Fermentations were performed in BIOSTAT ® B bioreactors. Data were collected with the software MFCS. OD, optical density.

Oxidation of the Mrx1-roGFP2 Biosensor Allows Monitoring the Changes in the MSH Redox Potential (E MSH ) in C. glutamicum
The observation that the MSH deficient ∆mshC mutant was impaired in growth with elevated oxygen concentrations prompted us to measure the changes in the MSH redox potential (E MSH ) in C. glutamicum during bioreactor experiments. Thus, we applied the recently developed genetically encoded Mrx1-roGFP2 biosensor, which is stably integrated in the genome of C. glutamicum [29]. Redox sensitive GFP2 (roGFP2) harbors two Cys residues that form a disulfide bond upon oxidation, resulting in ratiometric changes of two excitation maxima in the fluorescence excitation spectrum [46]. Mrx1 further was shown to selectively reduce S-mycothiolated proteins as part of the Mrx1/MSH/Mtr electron pathway [19,47]. Moreover, the Mrx1-roGFP2 fusion was well characterized as a redox biosensor with respect to its selectivity towards MSSM in vitro [29]. Upon reaction with MSSM, the MSH moiety is transferred to Mrx1 and roGFP2, followed by intramolecular disulfide formation in roGFP2 and the concomitant change of its fluorescence excitation maxima [29] (Figure 2a). To define suitable settings for ratiometric fluorescence measurements, crude cell extracts of C. glutamicum WT_Mrx1-roGFP2 with integrated Mrx1-roGFP2 were prepared and treated with 10 mM DTT or 5 mM diamide for fully reduced and oxidized controls, respectively, as previously described [29]. The strongest fluorescence intensity alteration (emission intensity at 510 nm) under our settings was detected when the biosensor was excited at 380 nm and 470 nm ( Figure S1). More specifically, upon oxidation of the probe, the excitation maximum at 380 nm increases with the subsequent decrease of the 470 nm excitation maximum, and vice versa upon reduction of the probe. Although the second fluorescence intensity maximum was described at 488 nm previously [29], this was out of the range of measurements of our available microplate reader. Thus, we used the calculation of the 380/470 nm fluorescence intensity ratios in our biosensor settings, which is an indicator of the MSH redox potential changes in C. glutamicum.
For in vivo determination of Mrx1-roGFP2 biosensor oxidation, C. glutamicum WT_Mrx1-roGFP2 and ∆mshC_Mrx1-roGFP2, both harboring the genome integrated biosensor Mrx1-roGFP2, were cultivated in shake flasks until the stationary phase was reached. Prior to fluorescence measurements, cells were harvested by centrifugation, washed twice with potassium phosphate buffer (100 mM; pH 7.0), and an optical density of 40 was adjusted as previously described [29]. Upon treatment with DTT and CHP, for reduction and oxidation of the biosensor probe, respectively, the 380/470 nm excitation ratio of Mrx1-roGFP2 was determined (Figure 2b). For non-treated samples, an equal volume of the respective buffer was added instead. Non-treated shake flask samples of C. glutamicum WT_Mrx1-roGFP2 and ∆mshC_Mrx1-roGFP2 revealed huge differences in terms of the biosensor oxidation ratio with 1.0 ± 0.02 and 1.52 ± 0.03, respectively (Figure 2b). However, the addition of DTT (reducing agent) or CHP (oxidizing agent) eliminated the biosensor ratio differences, resulting in fully oxidized and fully reduced biosensor ratios of 1.5-1.6 and 0.7-0.8, respectively (Figure 2b). As expected, growth of the mutant strain in shake flasks with minimal medium proceeded similar with a growth rate of 0.26 ± 0.02 h −1 when compared with the parental strain C.glutamicum WT_Mrx1-roGFP2 (0.27 ± 0.03 h −1 ) (Figure 2c). Biosensor measurements at the end of the exponential growth phase further revealed that the initial biosensor oxidation degrees (OxD; Equation (1)) were maintained highly oxidized (0.91 ± 0.01; 0.86 ± 0.05) in the MSH-deficient mutant and more reduced (0.6 ± 0.04; 0.49 ± 0.04) in the WT strain (Figure 2d).
This confirms a mycothiol redox potential (E MSH ) (Equation (2)) in C. glutamicum WT_Mrx1-roGFP2 and ∆mshC_Mrx1-roGFP2 of −280 ± 2 mV and −255 ± 7 mV at the end of the exponential growth phase in shake flasks, respectively (Table 1). This observation is in accordance with the previous study, showing that C. glutamicum WT_Mrx1-roGFP2 maintains a highly reducing intracellular environment during the course of cultivation in shake flasks (−280-300 mV) [29]. In contrast, the MSH-deficient mutant showed a more oxidized intracellular environment [29]. It is likely that elevated ROS levels in the MSH mutant caused an oxidation of Mrx1-roGFP2. However, in contrast to bioreactor experiments with oxygen concentrations surpassing the air saturation (pO 2 = 30%), growth of the MSH-deficient mutant was not impaired in shake flasks (Figure 2c), as seen for bioreactor cultivations with pO 2 of 20% (Figure 1d), when compared with the WT strain. This indicates that ROS production under these conditions did not overwhelm ROS detoxification by MSH-independent enzymatic antioxidant systems in the ∆mshC mutant. The addition of the thiol-reactive oxidant NaOCl to shake flask cultures of the ∆mshC mutant was shown to be detrimental in terms of growth patterns [27], as observed during bioreactor experiments with oxygen concentrations surpassing the air saturation. This indicates the sensitivity of the ∆mshC mutant towards increased ROS production in bioreactors, supporting the role of MSH to overcome oxidative stress during fermentation. In contrast, MSH is not essential in aerobic shake flask cultures with lower oxygen tension, which is in agreement with the observed more reduced biosensor signals in the C. glutamicum_Mrx1-roGFP2 strain (Figure 2d) [29].

Mycothiol-Dependent Protection is Important when C. glutamicum is Exposed to Elevated Oxygen Concentrations
To investigate the oxidative response of the Mrx1-roGFP2 biosensor in C. glutamicum WT_Mrx1-roGFP2 at elevated oxygen concentrations, we performed bioreactor experiments with C. glutamicum WT_Mrx1-roGFP2 at a pO 2 of ≥30%. The first fluorescence measurement revealed an almost fully oxidized biosensor.
To ensure that the biosensor response resulted from increased oxygen concentrations, the bioreactor cultivation of C. glutamicum WT_Mrx1-roGFP2 was performed with a stepwise pO 2 gradient (Figure 3a). At the initial pO 2 of 0% (set by providing 100% N 2 as sole gas), the biosensor oxidation ratio was very low, indicating the presence of a reducing environment in the bioreactor at a pO 2 of 0% ( Figure 3a). Subsequently, the pO 2 value was increased in 5% steps in the bioreactor and, at each of the pO 2 steps, the signal ratio of the biosensor was determined 60 min after setting the pO 2 . As depicted in Figure 3a, the ratio of the biosensor increased when setting higher oxygen concentrations (pO 2 ), indicating an oxidative stress response. At a pO 2 of 30%, the biosensor oxidation degree (OxD) was determined as 0.86 ± 0.04 (Figure 3b), representing a highly oxidized environment in C. glutamicum WT under these conditions. Moreover, further increase of the pO 2 value did not lead to enhanced OxD values, which are not significantly different to those determined for the MSH-deficient mutant strain (Figure 3b). In comparison, at lower pO 2 values (pO 2 = 5%; pO 2 = 20%; pO 2 = 25%), OxD values determined for the WT_Mrx1-roGFP2 strain were significantly lower than the fully oxidized biosensor probes for the ∆mshC_Mrx1-roGFP2 mutant strain (Figure 3b).
Notably, under aerobic conditions, the strongest oxidative shift occurred when surpassing the air saturation of 20%, resulting in an OxD shift of 0.33 from 0.53 ± 0.06 (pO 2 = 20%) (which is in the range of OxD values determined for shake flask samples) towards highly oxidized values of 0.86 ± 0.04 (pO 2 = 30%) (Figure 3b). This conforms an oxidative shift of E MSH from −280 ± 6 mV (pO 2 = 20%) to −256 ± 4 mV (pO 2 = 30%) for the C. glutamicum WT_Mrx1-roGFP2 strain, whereas the redox potential of the mutant strain was highly oxidized at every tested pO 2 level (Table 2). This oxidative shift is in agreement with the growth defect of the MSH-deficient mutant strain when surpassing the air saturation both when setting a constant pO 2 = 30% (Figure 1d), but also with a stepwise increase of the pO 2 value for the mutant strain harboring the redox biosensor Mrx1-roGFP2 ( Figure S2).
Taken together, the strong oxidative shift of C. glutamicum WT_Mrx1-roGFP2 in bioreactor cultivations indicates the requirement of the non-enzymatic antioxidant and LMW thiol MSH to overcome oxidative stress when the oxygen concentration surpasses the air saturation.

Discussion
Utilization of respiratory chains for aerobic metabolism comes along with the generation of ROS [6,9]. To eliminate these toxic byproducts, aerobic organisms developed antioxidant defense mechanisms including enzymatic and non-enzymatic protection systems [12][13][14][15][16][17]. The abundant LMW thiol MSH functions to maintain the reduced state of the cytoplasm and represents the main non-enzymatic antioxidant in high-GC Gram-positive bacteria, such as the industrial platform organism C. glutamicum [19,27,48]. Apart from MSH, C. glutamicum encodes highly efficient enzymatic detoxification systems, such as the superoxide dismutase (SOD) [49] and methionine sulfoxide reductases (Msr) [50]; catalase (KatA) and the peroxiredoxins mycothiol peroxidase (Mpx) [15,51]; and thiol-peroxidase (Tpx) [27]. The metalloenzyme superoxide dismutase (SOD) (EC 1.15.1.1) catalyzes the dismutation of superoxide anions to H 2 O 2 and oxygen. An E. coli sodA sodB double mutant was impaired in growth during batch cultivation when the dissolved oxygen concentration was shifted from 30% to 300% air saturation, indicating its importance in ROS detoxification during bioreactor experiments [52]. H 2 O 2 is subsequently converted to H 2 O and O 2 via the H 2 O 2 scavenging systems KatA, Mpx, and Tpx in C. glutamicum [15,27]. This avoids a further conversion to the highly toxic hydroxyl radical.
In contrast to KatA, the peroxiredoxins Mpx and Tpx employ peroxidatic Cys residues for H 2 O 2 detoxification, leading to S-mycothiolation and intramolecular disulfides of Mpx and Tpx. Regeneration or their catalytic activities requires coupling to either the Trx/TrxR or Mrx1/MSH/Mtr pathway, whereas in biochemical studies, the latter was shown to be kinetically favored under oxidative stress [14,15,47,53]. Moreover, the Trx/TrxR pathway is known to be employed by many biochemical systems for regeneration, and thus regarded as rate limited [14]. Thus, the Mrx1/MSH/Mtr pathway functions in de-mycothiolation under stress conditions, when Trx/TrxR is busy with reduction of cellular disulfides. In C. glutamicum, the katA mutant was shown to be highly sensitive towards H 2 O 2 treatment, resulting in a strong oxidation of the Mrx1-roGFP2 biosensor [29]. In contrast, no sensitivity towards H 2 O 2 exposure was observed in the mpx and tpx mutants, which showed similar biosensor responses under H 2 O 2 stress as the wild type [29]. C. glutamicum WT was shown to be resistant towards 100 mM H 2 O 2 and the Mrx1-roGFP2 biosensor did not respond to 10 mM H 2 O 2 [29]. Of note, 1-5 mM H 2 O 2 resulted in a maximal roGFP2 biosensor oxidation in E. coli [12]. This indicates that KatA plays a crucial role for H 2 O 2 detoxification, whereas the contribution of Mpx and Tpx is neglectable.
Notably, KatA of C. glutamicum possesses a remarkably high activity and is even commercially available (Merck, CAS Number 9001-05-2). Despite the extraordinary enzymatic detoxification power of KatA, elevated oxygen concentrations during batch fermentations resulted in cell death of the MSH-deficient C. glutamicum mutant. ROS production rates are proportional to the collision frequency of oxygen and redox enzymes [9]. Consequently, the rate of ROS production inside the cell directly depends on the oxygen concentration in the extracellular environment [7,9]. This indicates that ROS production in bioreactor cultivations overwhelmed the remaining antioxidant systems and that MSH as an additional antioxidant is required to provide protection against oxidative stress at elevated oxygen concentrations. Consistently, a strong oxidative response of the Mrx1-roGFP2 biosensor was observed when C. glutamicum WT was exposed to oxygen concentrations that were shown to be harmful for the MSH-deficient mutant strain in bioreactors. This confirms the requirement of MSH as supporting antioxidant and, consequently, an oxidative redox shift of the 2MSH/MSSM redox couple occurring in C. glutamicum WT strains under these conditions. Thus, the increased MSH synthesis or the overexpression of Mtr for MSH recycling could be beneficial to improve the intracellular production process, providing a promising strategy for the development of highly robust industrial production strains.
MSH has multiple antioxidant functions by scavenging free radicals either directly or as a cofactor for antioxidant enzymes [19]. When the oxygen concentration surpasses the air saturation, MSH becomes a crucial player to overcome oxidative stress, despite the presence of other highly efficient enzymatic antioxidant systems working independently of MSH.