2019 Volume 25 Issue 3 Pages 443-448
Accumulation of reactive oxygen species (ROS) in the culture medium may inhibit the growth of microorganisms and prevent the detection of harmful bacteria in foods. In this study, we investigated the relationship between ROS (hydroxyl radical and hydrogen peroxide) and Escherichia coli growth using electron spin resonance, which detects and quantifies multiple oxygen free radicals simultaneously. ROS had little effect on the growth of normal E. coli cells. In contrast, heat-treated E. coli cells were sensitive to ROS, and the number of colony-forming units increased 10-fold with the addition of sodium thiosulfate or sodium pyruvate, both of which scavenge ROS in the culture medium. Both compounds efficiently scavenged hydroxyl radical, while sodium pyruvate eliminated hydrogen peroxide to a greater extent than sodium thiosulfate. Our results imply that hydroxyl radical levels are more critical for the growth and survival of damaged E. coli cells.
In suspected incidents of food poisoning, it is important to clarify the causative microorganism and contamination pathway. However, in approximately 20 % of food poisoning cases, the causative agent and pathway remain unknown i). One reason for this failure is that microorganisms are damaged by environmental stresses, and the detection of damaged bacteria is difficult with current microbiological techniques. Moreover, the existing media used to detect and culture microorganisms, particularly agar-based media, are not necessarily optimal for growth.
Reactive oxygen species (ROS) adversely affect both prokaryotic and eukaryotic cells and potentially prevent the detection of damaged bacteria. Biologically relevant ROS include superoxide radical (O2−), hydrogen peroxide (H2O2), and hydroxyl radical (•OH) (Fig. 1 (A)). These ROS are generated by the autoxidation of cytoplasmic flavoproteins under aerobic growth conditions. Numerous studies have reported various physiological effects of ROS on bacterial cells (McCormick et al., 1998; Pericone et al., 2003; Park et al., 2005). Normally, ROS in cells are rapidly eliminated by ROS scavenging enzymes such as superoxide dismutase. Because of the high efficiency of these enzyme systems, it was previously assumed that ROS are not excreted from cells and do not accumulate in the culture medium; however, recent studies have reported superoxide radical in culture media and the potential to inhibit actinomycete growth (Takahashi et al., 2003; Takahashi, 2004; Nakashima et al., 2012).
A: The generation of reactive oxygen species from O2. B & C: ESR spectra for adducts detected in EM9 agar medium with the spin trapping agent CYPMPO. B: Without E. coli cultivation. C: With E. coli cultivation. The arrows show the signal (g = 2.008) used for the quantitative analysis.
Conventional methods for measuring ROS do not enable the simultaneous detection of multiple ROS products and cannot identify individual molecular species. Accordingly, inadequate data are available for examining the relationship between ROS and microbial growth, and for developing approaches for the control of ROS, despite the need for a culture method that does not inhibit microbial growth.
Electron spin resonance (ESR) is a useful method to detect and identify free radicals. ESR coupled with spin trapping has been applied to detect short-lived free radicals. This method extends the lifetime of short-lived free radicals by trapping the radicals with spin-trapping compounds (spin trap reagents) and transforming them into free-radical adducts (spin adducts) (Kameya, 2013; Kameya, 2017). The use of the ESR spin trapping method enables the simultaneous detection and quantification of multiple oxygen free radicals (such as superoxide and hydroxyl radicals).
In this study, we used ESR to investigate the relationship between ROS and Escherichia coli growth by clarifying the type and amount of oxygen free radicals in the culture medium.
Organism and culture medium The E. coli K-12 strain MG1655 (λ− rph-1) used in this study was obtained from the National BioResource Project (NIG, Japan): E. coli.
Vitamin-free casamino acid, yeast extract, tryptone, tryptic soy broth (TSB), and polypeptone were obtained from Difco Laboratories (Detroit, MI, USA). Other reagents were obtained from Nacalai Tesque, Inc. (Kyoto, Japan). The composition of each medium was as follows: EM9 medium (per liter, 100 mL of EM9 salt [per liter, 88 g of Na2HPO4, 12 g of KH2PO4, and 50 g of sodium chloride], 10 mL of 10 % vitamin-free casamino acid, 10 mL of 20 % glucose, 0.1 mL of 1 M MgSO4•7H2O); lysogeny broth (LB) medium (per liter, 5 g of yeast extract, 10 g of tryptone, 5 g of sodium chloride); TSB medium (per liter, 30 g of TSB); and yeast extract-peptone-dextrose (YPD) medium (per liter, 10 g of yeast extract, 20 g of polypeptone, and 20 g of glucose). Gelling agents (agar, gellan gum, or gelatin) were added at 1.5 % (w/v). The pH values were adjusted to 7.0 before autoclaving. ROS scavengers (sodium thiosulfate (1 %) (Bektasoglu et al., 2006) or sodium pyruvate (0.1 %) (Kładna et al., 2015)) were added after autoclaving. Anaeropack (Mitsubishi Gas Chemical, Tokyo, Japan) was used for anaerobic culture.
Heat treatment Heat treatment was performed according to a previously described method (Katsui et al., 1982). Briefly, a single colony of the E. coli strain was inoculated in 5 mL of LB medium and grown overnight at 37 °C. EM9 medium (5 mL) was inoculated with 50 µL of the overnight culture and incubated at 37 °C for 3 h. Then, 1 mL of the bacterial culture solution was centrifuged at 3 000 rpm for 5 min. The supernatant was removed, and the cells were suspended in 1 mL of phosphate-buffered saline. This washing step was performed twice. The prepared cell suspension was left to stand at 0 °C for 30 min and then heat-treated at 55 °C for 20 min. The colony-forming units/mL were determined by plating serial dilutions.
Sample preparation
Liquid culture A single colony of the E. coli strain was inoculated into 5 mL of LB medium and grown overnight at 37 °C. The liquid medium (5 mL) was inoculated with 50 µL of the overnight culture and incubated at 37 °C for 8 h. Then, 1 mL of the bacterial culture solution was centrifuged at 13 000 rpm for 5 min. Next, 4 mg of the spin trapping agent, 2-(5,5-dimethyl-2-oxo-2λ5-[1,3,2] dioxaphosphinan-2-yl)-2- methyl-3,4-dihydro-2H-pyrrole 1-oxide (CYPMPO) (Shidai Systems, Tokyo, Japan), was added to the supernatant (200 µL). The mixture was measured after 20 min. The control samples without E. coli inoculation were incubated at 37 °C for 8 h and treated similarly.
Plate culture The liquid medium (5 mL) was inoculated with 50 µL of the overnight culture in LB and incubated at 37 °C for 6 h. Then, 100 µL of the diluted bacterial culture (103–107 cfu/mL) was spread on the agar medium and cultured at 37 °C for 18 h. After mechanically scraping the E. coli cells with a spatula, the agar medium was finely crushed, packed in a spin column (Vivaclear centrifugal filter, 0.8 µm; Sartorius, Gottingen, Germany), and centrifuged at 13 000 rpm for 5 min. Then, 4 mg of the spin trapping agent, CYPMPO, was added to the extracted liquid (200 µL). The mixture was measured after 20 min. The control samples without E. coli inoculation were incubated at 37 °C for 18 h and treated similarly.
ESR measurement conditions For ESR, the Bruker EMX-Plus system was used. The measurement conditions were as follows: room temperature (23−25 °C); magnetic field range, 3521 ± 100 G; data points, 1000; microwave power, 1 mW; time constant, 20.48 µs; conversion time, 20.00 ms; sweep time, 20.0 s; scan, 10 times; modulation intensity, 1.00 G.
The measurements were repeated five times under the same conditions, and the signal intensity of the hydroxyl radical-CYPMPO adduct (g = 2.008) was used for quantitative analysis of the hydroxyl radical.
Measurement of the hydrogen peroxide concentration The hydrogen peroxide concentration was measured by colorimetric analysis using the Hydrogen Peroxide Assay Kit (Cosmo Bio Inc., Tokyo, Japan). Samples were prepared as described for ESR measurement.
Statistical processing Statistical analyses were performed using JMP 11 (SAS Institute Inc., Chicago, IL, USA).
Detection and quantification of oxygen free radicals in the culture medium by ESR Based on the spectra obtained by ESR, it is possible to identify the type of oxygen free radical by the shape of the signal and numerical parameters (Oowada et al., 2012). Furthermore, quantification of each radical is also possible based on the intensity of the ESR signals.
Figure 1 (B) and (C) show the ESR spectra for EM9 agar medium without and with E. coli cultivation. A hyperfine coupling constant of the signals was calculated, and the values obtained were consistent with those for hydroxyl radical (AN: 1.38, 1.35 mT; AH: 1.35, 1.23 mT; AP: 4.88, 4.68 mT) and superoxide radical (AN: 1.27, 1.27 mT; AH: 1.10, 1.06 mT; AP: 5.24, 5.08 mT) as reported previously (Oowada et al., 2012). Therefore, the oxygen free radicals in the EM9 medium were identified as hydroxyl and superoxide radicals. There was no change in spectral composition without and with E. coli cultivation; however, the signal intensity reflecting the amount of oxygen free radicals increased significantly with cultivation. Since the signal for superoxide radical was very small (<0.1), we used the signal for the hydroxyl radical (g = 2.008) for further analysis.
Figure 2 shows the results of the quantitative analysis of the hydroxyl radical. Virtually, no hydroxyl radical (<0.01) was observed in the water-agar solution. In contrast, substantial amounts of hydroxyl radical (0.32 ± 0.02) were detected in EM9 agar medium, indicating that the components of the EM9 medium were involved in a non-biological generation of hydroxyl radicals. Upon culturing E. coli, the amount of hydroxyl radical increased significantly to 0.89 ± 0.19, and this could probably be attributed to the excretion of oxygen free radicals by E. coli cells.
Hydroxyl radicals (ESR signal intensity for adducts with CYPMPO) in water agar and EM9 medium without and with cultivation. (**p < 0.01)
Superoxide is the first radical product during the reduction of molecular oxygen (Fig. 1 (A)). This radical is not a more potent oxidizing species than the sequential reduction products, such as hydroxyl radical. Hydroxyl radical has potentially deleterious effects on biological systems via damage to lipids, proteins, and nucleic acids and is generally considered toxic to living organisms (McCormick et al., 1998; Pericone et al., 2003; Park et al., 2005).
Although the primary mechanism underlying their generation in culture media remains unclear, superoxide and hydroxyl radicals were detected. Superoxide radical can arise from non-enzymatic sources in biological systems and is further reduced to produce hydrogen peroxide (Cohen, 1994). When hydrogen peroxide is reduced by the Fenton reaction and by the Haber-Weiss reaction, hydroxyl radical is produced (Fig. 1 (A)) (Cohen, 1994). In addition, hydroxyl radical is also generated by other processes, such as the decomposition of peroxynitrate (Oury and Tatro, 1995).
Differences in oxygen free radicals among culture media Various types of media are used for microbial culture. To determine whether there is a difference in oxygen free radicals among media, several types of media were evaluated with E. coli cultivation.
Figure 3 shows the amount of hydroxyl radical in agar and liquid media for EM9, LB, TSB, and YPD media. In all cases, hydroxyl radical levels in the agar media were >3-fold higher than those in the liquid media. A previous study reported that oxygen radicals are stabilized within crystal structures (Sueishi et al., 2012). Therefore, we measured hydroxyl radical levels in media with different agar concentrations (0.3, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0 %). The amount of hydroxyl radical increased up to agar concentrations of 1.5 % and reached a plateau (data not shown). It is possible that hydroxyl radical released from E. coli cells was entrapped within the crystal structure of agar, their life span was extended, and consequently their level was increased. We also consider an alternative possibility that hydroxyl radical was generated by the reaction between hydrogen peroxide released from E. coli cells and trace metal elements pre-existing in the agar powder (Hara and Isoda, 2012). Since no significant differences were observed among the media examined, EM9 medium was used in subsequent analyses.
Hydroxyl radicals in agar and liquid EM9, LB, TSB, and YPD media with E. coli cultivation. The amount of hydroxyl radicals in the medium without cultivation was used as a control and is shown in white in the bar graph.
Some studies have reported favorable microbial growth when gellan gum was used as an alternative gelling agent (Rule and Alexander, 1986; Nyonyo et al., 2012). The growth promotion effect of gellan gum may be explained by the lower radical levels in the culture media. Therefore, gellan gum and a classical gelling agent, gelatin, were added to EM9 medium, and the amounts of hydroxyl radical in each medium were determined. There were no significant differences in the amount of hydroxyl radical among the three gelling agents, agar (0.67 ± 0.13), gellan gum (0.64 ± 0.09), and gelatin (0.52 ± 0.11). Thus, the primary reason for the observed growth-promotion effect of gellan gum remains unclear. It is possible that certain physical properties of gelling agents are important for microbial growth.
Relationship between external ROS and E. coli growth It is well established that intracellular ROS are generated through the autoxidation of cytoplasmic flavoproteins under aerobic growth conditions. Thus, we examined whether the hydroxyl radical accumulated in the culture media is also generated under aerobic environments. On agar plates, very small amounts of hydroxyl radical were generated under anaerobic growth conditions (Fig. 4), although vigorous cell growth was observed. In addition, the amount of hydroxyl radical was significantly greater in shaking cultures than in standing cultures. From these results, we conclude that hydroxyl radical is generated by aerobic metabolism in E. coli. Given that intracellular ROS are eliminated immediately by catalase and superoxide dismutase within E. coli cells, hydroxyl radical may also be generated in the extracytoplasmic space (periplasm or cell surface).
Hydroxyl radicals in EM9 medium cultured under various conditions. The amount of hydroxyl radicals in the medium without cultivation was used as a control and is shown in white in the bar graph. (**p < 0.01, *p < 0.05)
Next, we investigated whether ROS accumulated in the culture media influence the growth and survival of E. coli. To this end, known ROS scavengers, i.e., sodium thiosulfate and sodium pyruvate, were added to the EM9 medium, and the amounts of ROS (hydroxyl radical and hydrogen peroxide) and the number of viable cells were determined.
Upon the addition of these ROS scavengers, the amounts of hydroxyl radical and hydrogen peroxide were reduced (Fig. 5 (A) and (B), left panel). Both compounds efficiently scavenged hydroxyl radical, while sodium pyruvate eliminated hydrogen peroxide to a greater extent than sodium thiosulfate. Viable cell counts did not increase in the presence of these ROS scavengers (Fig. 5 (C), left panel), indicating that normal (healthy) E. coli cells are not sensitive to extracellular ROS under these experimental conditions.
Effects of the radical scavengers on (A) hydroxyl radical amounts, (B) hydrogen peroxide concentrations, and (C) viable cell counts. Normal or heat-treated E. coli cells were cultured on EM9 agar. Sodium thiosulfate (1 %) or sodium pyruvate (0.1 %) was added after autoclaving. (A) and (B) were measured in agar medium on which the cell viable count was detected approximately 100 cfu.
Damaged (injured) bacterial cells are considered to be sensitive to various stresses, including ROS. Therefore, we also examined the effects of extracellular ROS on the growth and survival of heat-damaged E. coli cells (Fig. 5, right panel). Contrary to what was observed in the normal cells, the heat-treated cells were sensitive to the adverse effects of ROS, and the number of viable cell counts increased >1 log upon the addition of sodium thiosulfate or sodium pyruvate (Fig. 5 (C)). Detectable amounts of hydrogen peroxide were still present in the sodium thiosulfate-treated cultures (Fig 5 (B)). Nevertheless, the viable cell counts were similar between the two treatment groups (Fig. 5 (C), right panel). These results imply that hydroxyl radical levels are more critical for the growth and survival of damaged E. coli cells.
Acknowledgments This study is part of the Project of Food Safety and was supported by a grant from the Ministry of Agriculture, Forestry, and Fisheries (MAFF), Japan.
