Involvement of catalase and superoxide dismutase in hydrophobic organic solvent tolerance of Escherichia coli

Escherichia coli strains are generally sensitive to hydrophobic organic solvents such as n-hexane and cyclohexane. Oxidative stress in E. coli by exposure to these hydrophobic organic solvents has been poorly understood. In the present study, we examined organic solvent tolerance and oxygen radical generation in E. coli mutants deficient in reactive oxygen species (ROS)-scavenging enzymes. The organic solvent tolerances in single gene mutants lacking genes encoding superoxide dismutase (sodA, sodB, and sodC), catalase (katE and katG), and alkyl hydroperoxide reductase (ahpCF) were similar to that of parent strain BW25113. We constructed a BW25113-based katE katG double mutant (BW25113∆katE∆katG) and sodA sodB double mutant (BW25113sodA∆sodB). These double-gene mutants were more sensitive to hydrophobic organic solvents than BW25113. In addition, the intracellular ROS levels in E. coli strains increased by the addition of n-hexane or cyclohexane. The ROS levels in BW25113∆katE∆katG and BW25113∆sodA∆sodB induced by exposure to the solvents were higher than that in BW25113. These results suggested that ROS-scavenging enzymes contribute to the maintenance of organic solvent tolerance in E. coli. In addition, the promoter activities of sodA and sodB were significantly increased by exposure to n-hexane.


Introduction
Efficient microbial production of valuable organic compounds including biofuels and fine chemicals from renewable biomass resources is one of the crucial challenges in the establishment of a sustainable society. Some of these valuable chemicals such as advanced biofuels and bulk chemicals, including organic solvents including 1-octanol and styrene, are toxic to various microorganisms such as E. coli and Pseudomonas putida (Akhtar et al. 2015;Lennen et al. 2013;Mukhopadhyay 2015). Accumulation of these toxic products can negatively impact the viability of microbes and impede their efficient mass production of organic compounds (Doukyu and Iida 2020;Doukyu et al. 2003;Mukhopadhyay 2015). Bacterial growth and stress response have been studied in a variety of organic solvents (Heipieper et al. 2007;Inoue and Horikoshi 1989;Ramos et al. 2002). Solvents with higher hydrophobicity tend to increase in toxicity (Kabelitz et al. 2003). At saturated concentrations, the toxicity of the hydrophobic organic solvent is inversely correlated with the logP ow of the solvent (common logarithm of partition coefficient of given solvent in a mixture of n-octanol and water) (Inoue and Horikoshi 1989). In the logP ow range of 2 to 4, increasing the hydrophobicity of the solvent could enhance the level of toxicity (Kabelitz et al. 2003). The effects of organic solvents on membrane structure and fluidity differ depending on the polarity of the solvent (Griepernau et al. 2007).
It has been reported that reactive oxygen species (ROS) are generated in Escherichia coli during exposure to hydrophilic solvents such as ethanol and n-butanol Doukyu and Taguchi AMB Expr (2021) 11:97 (Cao et al. 2017;Chin et al. 2013;Rutherford et al. 2010). A ROS assay using a cell-permeant fluorescent dye showed a significant increase in intracellular ROS levels in n-butanol-exposed E. coli cells (Rutherford et al. 2010). In addition, an E. coli strain expressing metallothioneins, which function in ROS scavenging, exhibited an increased growth rate under n-butanol stress (Chin et al. 2013). Thus, ROS generation has been suggested to be attributable in part to the toxicity of hydrophilic organic solvents. On the other hand, various tolerance mechanisms in E. coli strains have been studied with hydrophobic organic solvents such as n-hexane and cyclohexane (Aono 1998). These findings are useful for improving the production of hydrophobic solvents and fatty acids (Akhtar et al. 2015;Lennen et al. 2013). However, little is known about the ROS generation in microbial cells by these hydrophobic organic solvents.
Various ROS, including superoxide, H 2 O 2 and hydroxyl radical, are generated as by-products in cells grown aerobically. Specific enzymes such as catalase and superoxide dismutase (SOD) decrease the levels of cytotoxic ROS. E. coli harbors two different catalase genes: katG, which encodes hydroperoxidase I (HPI) ) and katE, which encodes HPII ). HPI is induced by H 2 O 2 in an OxyR-dependent manner, while expression of HPII is dependent on a sigma factor, RpoS (Mukhopadhyay and Schellhorn 1994). The activities of both catalases increase as the growing cells enter stationary phase. In addition, the genome of E. coli includes the alkyl hydroperoxide reductase gene (ahpCF), which scavenges H 2 O 2 and organic hydroperoxides (Ferrante et al. 1995;Smillie et al. 1992). E. coli possesses three distinct SOD genes: sodA, which encodes a cytosolic manganese-containing SOD (Touati 1983), sodB, which encodes a cytosolic iron-containing SOD (Sakamoto and Touati 1984) and sodC, which encodes a periplasmic copper and zinc-containing SOD (Imlay and Imlay 1996). SodC is induced in stationary phase and seems to be important to protect the cell from exogenous ROS attacks (Gort Amy et al. 2002).
Both the katE katG double mutant and sodA sodB double mutant are sensitive to H 2 O 2 due to DNA damage (Imlay and Linn 1987;Ruiz-Laguna and Pueyo 1999). In addition, ROS levels significantly increased in sodA sodB double mutant by the addition of p-nonylphenol (Okai et al. 2004). Thus, it was suggested that one of the primary actions of p-nonylphenol in cells is the generation of superoxide.
In this study, we examined the involvement of catalase and SOD genes in organic solvent-tolerance in E. coli. We found that two BW25113-based mutants, the ∆katE∆katG mutant and ∆sodA∆sodB mutant, were highly sensitive to organic solvents. In addition, it was shown that the increases in intracellular ROS levels in these two mutants were larger than that in strain BW25113 when these strains were treated with hydrophobic organic solvents. These results showed that catalase and SOD are implicated in the organic solvent tolerance in E. coli.

Media, culture conditions and materials
E. coli strains were grown aerobically at 37 °C in LBGMg medium consisting of 1% tryptone, 0.5% yeast extract, 1% NaCl, 0.1% glucose, and 10 mM MgSO 4 (Aono et al. 1991). The LBGMg medium was solidified with 1.5% (wt/ vol) agar. Ampicillin (50 μg/ml) or kanamycin (50 μg/ ml) was added to the medium when necessary. Lysogeny broth (LB) agar medium was used for measuring colonyforming units (Neidhardt et al. 1974). M9 medium was used for ROS assay (Neidhardt et al. 1974). Growth of cells in liquid culture was monitored by measuring the optical density at 660 nm (OD 660 ). The tert-butyl hydroperoxide was obtained from FUJIFILM Wako Pure Chemical Industries (Osaka, Japan). The organic solvents used were of the highest quality available (FUJIFILM Wako Pure Chemical Industries). The logP ow values of the hydrophobic solvents used in this study were as follows: cyclooctane (logP ow , 4.1), n-hexane (logP ow , 3.9) and cyclohexane (logP ow , 3.4).

Bacterial strains and plasmids
The E. coli strains and plasmids used in this study are summarized in Tables 1 and 2, respectively. Strain BW25113 and its single-gene knockout mutants were obtained from the National Bio-Resource Project (NIG, Mishima, Japan) (Baba et al. 2006). The plasmid pCP20 was also supplied by NIG. pMC1403 contains a sequence downstream of the 10th codon of lacZ, but it does not contain the lacZ promoter, the Shine-Dalgarno sequence, and the start codon (Casadaban et al. 1980). The plasmid pMW119 was purchased from Nippon Gene (Tokyo).

Construction of BW25113∆katE∆katG and BW25113∆sodA∆sodB
The Km R cassettes in BW25113∆sodB and BW25113∆katE were eliminated with pCP20 (Cherepanov and Wackernagel 1995). Elimination of the Km R cassette was confirmed by PCR analysis using chromosomal DNA. The combination of primers for BW25113∆sodB was sodB-S and sodB-AS, and that for BW25113∆katE was katE-S and katE-AS (Table 3). BW25113∆sodA∆sodB and BW25113∆katE∆katG were constructed from the Km R cassette-eliminated mutants BW25113∆sodB and BW25113∆katE by P1 transduction of kanamycin-resistance with BW25113∆sodA and BW25113∆katG as the donor, respectively. The Km R cassettes in BW25113∆sodA∆sodB and BW25113∆katE∆katG were also eliminated with pCP20. Elimination of the Km R cassette in the sodA and katG region was confirmed by PCR analysis. The combination of primers for the sodA disruption was sodA-S and sodA-AS, and that for the ∆katG disruption was katG-S and katG-AS.

Measurement of organic solvent-tolerance in E. coli
Cultures of E. coli strains in LBGMg medium after 16 h of incubation (OD 660 , 4 to 5) at 30 °C were diluted with

Cloning of the katE, katG, sodA and sodB genes
The regions of katE, katG, sodA and sodB were amplified by PCR using AccuPrime Taq DNA Polymerase (Thermo Fisher Scientific Inc.) with high fidelity and BW25113 chromosomal DNA as the template. The primers used were designed according to the genome sequence of BW25113 deposited in GenBank (accession number CP009273). The combination of primers for katE was katEc-S and katEc-AS, that for katG was katGc-S and katGc-AS, that for sodA was sodAc-S and sodAc-AS, and that for sodB was sodBc-S and sodBc-AS (Table 3). A restriction endonuclease cleavage site was introduced into all primer sequences. The amplified fragments were digested with the relevant restriction enzymes and ligated into the cloning site of pMW119 under the same direction as the lac promoter to construct plasmids pMWkatE, pMWkatG, pMWsodA, and pMWsodB, respectively. The pMWkatE was digested with KpnI and SacI, and then the fragments containing katE were ligated between KpnI and SacI sites of pMWkatG. The resulting plasmid was designated pMWkatEkatG. In addition, the pMWsodA was digested with SphI and SalI, and then the fragments containing sodA were ligated between SphI and SalI sites of pMWsodB. The resulting plasmid was designated pMWsodAsodB.

Enzyme activity assay
E. coli cells grown in LBGMg medium after incubation at 30 °C for 16 h were harvested by centrifugation (4400×g for 10 min at 4 °C) and suspended in 10 mM of Tris-HCl buffer (pH 8.0). The cell suspension was sonicated on ice and centrifuged (10,000×g, 10 min at 4 °C). The supernatant was used for the enzyme activity assay. Catalase activity was determined by following the rate of H 2 O 2 consumption at 240 nm (Claiborne et al. 1979). The enzyme activity was calculated from the molar adsorption coefficient of the H 2 O 2 (ε = 43.6 M −1 cm −1 ). The reaction mixture (1 mL) contained 50 mM potassium phosphate buffer at pH 7.0 and 10 mM H 2 O 2 . The reaction was initiated by adding the enzyme solution (50 μL) to the reaction mixture and the initial velocity of H 2 O 2 disappearance was measured at 30 °C. One unit  of enzyme activity was defined as the amount of enzyme that decomposes 1 μmol of H 2 O 2 per min. SOD activity was measured by following the rate of pyrogallol autooxidation (Marklund and Marklund 1974). The enzyme solution (10 μL) was mixed with 2.48 mL of 50 mM Tris-HCl and 1 mM EDTA at pH 8.2. After preincubation for 5 min at 25 °C, the reaction was started by adding 10 μL of 50 mM pyrogallol solution in 10 mM HCl. The change of absorbance was monitored at 325 nm. One unit of enzyme is defined as the amount of enzyme that inhibits the autoxidation rate of pyrogallol by 50%.

Protein content
The protein concentration was determined by the method of Bradford (Bradford 1976) using bovine serum albumin as the standard.

Sensitivities of E. coli strains to H 2 O 2 and menadione
E. coli cells grown in LBGMg medium (10 mL) at 37 °C to an OD 660 of about 0.6 (approximately 4 to 5 × 10 8 cells/ml) were harvested by centrifugation (4400×g for 10 min at 4 °C), and suspended in PBS buffer (10 mL) consisting of 140 mM NaCl, 8.1 mM Na 2 HPO 4 , 2.7 mM KCl, and 1.5 mM KH 2 PO 4 (pH 7.4). The cell suspension was washed once by centrifugation, resuspended in PBS buffer with H 2 O 2 (1 to 4 mM) or menadione (15 to 60 mM), and further incubated with shaking at 37 °C. After incubation for 1 h, each suspension was plated on LB agar medium. The number of colonies formed on the agar plate was counted after 24 h of incubation at 30 °C.

Detection of reactive oxygen species
The ROS were detected with the 5-(and-6)-carboxy-2' ,7'-dichlorodihydrofluorescein diacetate (carboxy-H 2 DCFDA) (Molecular Probes, Eugene, OR) as reported previously with a slight modification (Rutherford et al. 2010). Carboxy-H 2 DCFDA is a cell-permeable indicator for ROS that does not fluoresce until it is hydrolyzed by esterases and oxidation occurs within cells. A 100 μl culture of overnight-grown E. coli cells was inoculated onto 10 ml of fresh LBGMg medium and incubated at 37 °C with shaking. After incubation for 3 h, 0.5 mL of 7.78 M tert-butyl hydroperoxide (TBHP; used as a positive control for oxidative stress) or 1 mL of an organic solvent (n-hexane or cyclohexane) was added to the culture. The culture was further incubated at 37 °C for 3 h with shaking. Twenty microliters of each culture was added to 1 ml of M9 medium. After incubation at 37 °C for 45 min, 50 μl of 25 mM carboxy-H 2 DCFDA samples was added to the medium. After incubation at 37 °C for 10 min, the OD 660 and the fluorescence excitation/emission at 485/535 nm of each sample were measured by spectrofluorometer (RF-6000; Shimadzu Co., Kyoto, Japan).
Specific fluorescence was calculated as fluorescence/ OD 660 .

Construction of the lacZ reporter fusions
The promoter regions of katE, katG, sodA and sodB were amplified by PCR using AccuPrime Taq DNA Polymerase and BW25113 chromosomal DNA as the template. The combination of primers for the katE promoter region was katEp-S and katEp-AS, that for the katG promoter region was katGp-S and katGp-AS, that for the sodA promoter region was sodAp-S and sodAp-AS, and that for the sodB promoter region was sodBp-S and sodBp-AS (Table 3). The amplified fragments were digested with the relevant restriction enzymes and ligated into the cloning site of pMC1403 to construct plasmids pMCkatEp, pMCkatGp, pMCsodAp, and pMCsodBp, respectively.

Assay for plasmid-borne β-galactosidase activity
E. coli strains were grown in LBGMg containing 50 μg/ ml ampicillin at 37 °C. Cells in the exponential phase of growth were treated with a small volume of chloroform and assayed for β-galactosidase activity as described previously (Miller 1972).

Organic solvent-tolerances of E. coli mutants deficient in ROS-scavenging enzymes
The colony-forming efficiency of the BW25113based ∆katE, ∆katG, ∆ahpF, ∆sodA, ∆sodB, ∆sodC, ∆katE∆katG and ∆sodA∆sodB mutants was investigated using an LBGMg agar plate in the presence of n-hexane (Fig. 1). All strains formed colonies in all spots on the plate without any solvent. The colony-forming A B Fig. 1 Colony-forming efficiency of E. coli BW25113 and its mutants deficient in ROS-scavenging enzymes on LBGMg agar medium. Each strain was grown in the absence of an organic solvent (A) and in the presence of n-hexane (B). Each strain was spotted at a tenfold dilution and incubated at 25 °C for 48 h efficiencies in the single gene mutants were similar to that in the parent strain BW25113 in the presence of n-hexane. In contrast, the double gene mutants, BW25113∆katE∆katG and BW25113∆sodA∆sodB, were highly sensitive to n-hexane, exhibiting 10 2 -or 10 3 -fold lower colony-forming efficiencies than the parent strain in the presence of the solvent. These results indicated that catalase and SOD were involved in the maintenance of organic solvent-tolerance in E. coli. Complementation of organic solvent-tolerances of BW25 113∆katE∆katG and BW25113∆sodA∆sodB by transformation of the catalase-or SOD-coding gene was investigated (Fig. 2). The colony-forming efficiencies in BW25113∆katE∆ka tG(pMWkatE), BW25113∆katE∆katG(pMWkatG), and BW25113∆katE∆katG(pMWkatEkatG) were about 10-, 10 2 -, and 10 3 -fold higher than that in BW25113(pMW119) in the presence of n-hexane, respectively. On the other hand, the colony-forming efficiencies in BW25113∆ sodA∆sodB(pMWsodA) and BW25113∆sodA∆sodB(pMW sodB) were both about tenfold higher than that in BW25 113(pMW119). In addition, the efficiency in BW25113 ∆sodA∆sodB(pMWsodAsodB) was about 10 3 -fold higher than that in BW25113(pMW119). Thus, it was shown that katE, katG, sodA, and sodB genes contribute to the maintenance of n-hexane-tolerance in E. coli.

Activities of ROS-scavenging enzymes and susceptibility to H 2 O 2 and menadione in BW25113∆katE∆katG and BW25113∆sodA∆sodB
The levels of catalase and SOD activities in BW25113∆katE∆katG and BW25113∆sodA∆sodB were compared with the parent strain BW25113 (Table 4). Most of the catalase and SOD activities were eliminated in BW25113∆katE∆katG and BW25113∆sodA∆sodB, respectively. The slight remaining activities of catalase and SOD seemed to be attributable to AhpCF and SodC, respectively. These results showed good agreement with a previous report (Alhama et al. 1998).
Susceptibility of the mutants to ROS was also confirmed by measuring the cell viability after exposure to H 2 O 2 and a redox-cycling agent, menadione (Additional file 1: Fig.  S1). The survival fraction of BW25113∆katE∆katG in 4 mM H 2 O 2 and BW25113∆sodA∆sodB in 60 mM menadione were 40% and 0.3% of those of the parent strain BW25113, respectively.

Growth of the E. coli mutants in liquid medium in the presence of organic solvents
The cell growth of BW25113, BW25113∆katE∆katG and BW25113∆sodA∆sodB in the LBGMg liquid medium in the presence of a hydrophobic solvent including cyclooctane, n-hexane, or a hydrophobic solvent mixture of n-hexane and cyclohexane (9:1 vol/vol) was examined by measuring the turbidity (Fig. 3). In the absence of the solvent, the specific growth rates of BW25113, BW25113∆katE∆katG and BW25113∆sodA∆sodB during the exponential growth phase were 1.6 h −1 , 1.5 h −1 and 1.3 h −1 , respectively. In the presence of cyclooctane, the growth rates of BW25113, BW25113∆katE∆katG and BW25113∆sodA∆sodB were 1.5 h −1 , 1.2 h −1 and 1.0 h −1 , respectively. Thus, the growth of these mutants without any organic solvents and in the presence of cyclooctane was slower than that of the parent strain BW25113. The growth of BW25113∆katE∆katG and BW25113∆sodA∆sodB was highly suppressed compared to that of BW25113 in the presence of n-hexane or the mixture of n-hexane and cyclohexane. These results suggested that each of catalase and SOD is involved in the maintenance of tolerance to hydrophobic organic solvents. BW25113∆sodA∆sodB exhibited an extended lag phase and BW25113∆katE∆katG did not grow during 8 h cultivation in the presence of n-hexane. Thus, BW25113∆katE∆katG was more sensitive to n-hexane than BW25113∆sodA∆sodB. Fig. 2 Colony-forming efficiency of BW25113-based recombinant E. coli strains. Each strain was grown on LBGMg agar medium containing ampicillin (50 μg/ml) and isopropyl-β-d-thiogalactopyra noside (IPTG; 0.5 mM) in the absence of an organic solvent (A) and in the presence of n-hexane (B). Each strain was spotted at a tenfold dilution and incubated at 25 °C for 48 h

Detection of ROS in E. coli cells exposed to hydrophobic organic solvents
We examined the ROS levels in BW25113, BW25113∆katE∆katG and BW25113∆sodA∆sodB in the presence of n-hexane or cyclohexane with carboxy-H 2 DCFDA, a fluorescent indicator for ROS in cells (Fig. 4). No significant difference in ROS levels was observed among these three strains in the absence of TBHP (a known inducer for ROS) or organic solvents. Addition of TBHP enhanced the ROS levels in these three strains. Treatment with n-hexane had little impact on ROS levels in BW25113 but markedly elevated ROS levels in BW25113∆katE∆katG and BW25113∆sodA∆sodB. The ROS levels by exposure to n-hexane in BW25113∆katE∆katG and BW25113∆sodA∆sodB were 1.9-and 3.2-fold higher than that in BW25113, respectively. Cyclohexane-exposure significantly increased ROS levels in all strains. The ROS levels by exposure to cyclohexane in BW25113∆katE∆katG and BW25113∆sodA∆sodB were 2.8-and 2.3-fold higher than that in BW25113, respectively. The ROS levels in the three strains by the addition of cyclohexane were higher than those by addition of n-hexane, respectively.

Discussion
The cell membrane is the main target of organic solvents (Aono et al. 1994). Levels of intracellular ROS in E. coli cells measured by using carboxy-H 2 DCFDA. ROS levels were measured in the strains exposed to TBHP (tert-butyl hydroperoxide), n-hexane or cyclohexane as described in the Methods. The relative specific fluorescence shows the ratio of the specific fluorescence in each strain divided by that in strain BW25113 not exposed to the organic solvent. Abbreviations: BW, BW25113; ∆katE∆katG, BW25113∆katE∆katG; ∆sodA∆sodB, BW25113∆sodA∆sodB. Mean values and standard deviations for three independent experiments are shown wall, secretion of extracytoplasmic proteins, and turgor pressure. Organic solvent molecules intercalate into the lipid bilayer of the membrane. Accumulation of organic solvent in the membrane disturbs a variety of vital cell processes, including those described above. The inner membrane in E. coli contains an electron transport chain that is the major site of the premature electron leakage to oxygen that generates superoxide (O 2 − ). Disturbance of the electron transport chain by organic solvent can increase the level of ROS. ROS have the capacity to damage various biomolecules, including proteins, ribosomes and DNA, and to reduce cellular culturability (Imlay 2003).
In the present study, we investigated the involvement of ROS-scavenging enzymes in hydrophobic organic solvent-tolerance. The colony-forming efficiency of E. coli strains on LBGMg agar showed that the tolerances to n-hexane in single gene-knockout mutants lacking the katE, katG, ahpF, sodA, sodB, or sodC gene were similar to that in the parent strain BW25113 (Fig. 1). AhpF is the peroxiredoxin reductase component of alkyl hydroperoxide reductase (AhpCF) that has been reported to be involved in tolerance to organic solvents such as 1,2,3,4-tetrahydronaphthalene (tetralin), cyclohexane, propylbenzene, and 1,2-dihydronaphthalene (Ferrante et al. 1995). AhpF and AhpC proteins act together (Li Calzi and Poole 1997). AhpF utilizes NADH or NADPH as electron donor to AhpC, which converts alkylhydroperoxides to their respective alcohol forms. AhpC is specifically reduced by AhpF and cannot be reduced by other electron transfer systems such as thioredoxin reductase. However, our results showed that deficiency of ahpF did not influence the tolerance to n-hexane. It has been reported that a sodA sodB double mutant was much more sensitive to paraquat than the wild type, although the absence of only the sodA gene or only the sodB gene had no effect on the sensitivity to paraquat (a superoxide generator) (Carlioz and Touati 1986). Both the sodA sodB double mutant and katE katG double mutant were more sensitive to p-nonylphenol (an endocrine disruptor) than the parent strain (Okai et al. 2004). In particular, the sodA sodB double mutant was highly sensitive to p-nonylphenol. These findings prompted us to construct BW25113∆katE∆katG and BW25113∆sodA∆sodB and then examine their organic solvent-tolerances. BW25113∆katE∆katG and BW25113∆sodA∆sodB lost most of their catalase and SOD activities, respectively (Table 4). In addition, we confirmed that BW25113∆katE∆katG and BW25113∆sodA∆sodB became sensitive to H 2 O 2 and menadione (an O 2 − generator), respectively (Additional file 1: Fig. S1). Both BW25113∆katE∆katG and BW25113∆sodA∆sodB were highly sensitive to n-hexane and a mixture of n-hexane and cyclohexane (Figs. 1 and 3). These results showed that accumulation of either H 2 O 2 or O 2 − in E. coli can exhibit an inhibitory effect on the cell growth. H 2 O 2 and O 2 − are relatively weak cytotoxic radical oxygens compared to other radical oxygens such as hydroxyl radicals (Bruno-Barcena et al. 2010;Fridovich 1986). Therefore, the hydroxyl radical produced via Fenton reaction and Haber-Weiss reaction from H 2 O 2 and O 2 − might be a main cause of the cytotoxicity by addition of hydrophobic organic solvents. Our assays using the fluorescent probe carboxy-H 2 DCFDA showed an increase in ROS after solvent stress (Fig. 4). In addition, the ROS levels in E. coli cells induced by exposure to cyclohexane were higher than those observed by exposure to n-hexane. Thus, the ROS level in E. coli cells is likely to depend on the amount of organic solvents accumulated in the cells, since a larger amount of cyclohexane than n-hexane is accumulated in cells in an organic-aqueous two-liquid-phase system (Tsukagoshi and Aono 2000).

Fig. 5
Effect of n-hexane on the promoter activity of katE, katG, sodA, or sodB. A 100-μl culture of the overnight-grown strain BW25113(pMCkatEp) (A), BW25113(pMCkatGp) (B), BW25113(pMCsodAp) (C), or BW25113(pMCsodBp) (D) was inoculated into 10 ml of fresh LBGMg liquid medium containing 50 μg/ml ampicillin without or with 1 ml of n-hexane. The culture was incubated at 37 °C until reaching an OD 600 of approximately 0.6. Cells were treated with chloroform and assayed for β-galactosidase activity. Values indicate the means and standard deviations of the results from three independent experiments