Response of Escherichia coli to hydrogen nanobubbles: an in vitro evaluation using synchrotron infrared spectroscopy

Abstract

Hydrogen (H2)-rich water, an apparent source of molecular H2, is an emerging functional drink with many purported benefits for human health (Yang et al., 2020; Ostojic, 2021). The preventive and therapeutic effects of H2 on various pathological processes have been intensively investigated in numerous clinical trials; it is commonly believed that the beneficial effects are mainly attributed to its selective antioxidant and anti- inflammatory properties (Lee et al., 2015; Ohta, 2015; LeBaron et al., 2019; Qiu et al., 2020). In recent years, a handful of rodent studies revealed that exogen‑ ous H2 can affect the gut microbiota (Sha et al., 2018; Valdes et al., 2018). For example, H2 was reported to induce a higher abundance of butyrate-producing bacteria in a rat model of Parkinson’s disease (Bordoni et al., 2019). Recent first-in-human trials have explored the effects of the long-term consumption of H2-rich water on antioxidant activity and the gut flora (Sha et al., 2018; Suzuki et al., 2018). Although these promising results suggest that the intestinal microbiota may be another plausible target for molecular H2, more studies are highly warranted to explain the mechan‑ ism(s) of H2 action on bacterial growth and functions. Fourier transform infrared (FTIR) spectroscopy can provide the chemical composition information of samples with the advantages of high sensitivity and good reproducibility (Wang et al., 2017). With the technological advance of synchrotron light source, the higher brightness and smaller source spot size have allowed the FTIR spectra to have higher intensity and better signal-to-noise ratio. As the spectral bands contain molecular fingerprint information, FTIR spectroscopy exhibits sufficient specificity to establish the molecular differences of cells exposed to various types of stimu‐ lations or treatments. The changes of absorption bands, such as those in position or intensity, reflect the com‐ ponent differences. Thanks to this advantage, FTIR spectroscopy has been widely used in the field of microbiology, especially for studying the changes of microbial constituents under stress conditions (Alvarez- Ordóñez et al., 2011). For example, synchrotron infrared spectroscopy has been employed to study the response of Escherichia coli to heat, cold, and ethanol, as well as the interaction between bacteria and metal ions (Hu et al., 2016). In this study, the effect of H2 nanobubbles (NBs) on the viability of E. coli was first tested by measuring the bacterial growth curve. Compared with double dis‐ tilled water (dd-water) and nitrogen (N2) NBs, the growth curve of H2 NB treatment group did not show any significant changes (Fig. 1a). The colony number of live bacteria during the exponential growth phase was further counted on the culture plates. After culture for 8 h, the counts were 5.5×10 9 and 5.9×10 9 CFU/mL (CFU: colony forming units) for the H2 NB treatment group and dd-water group, respectively (Fig. 1b), indi‐ cating no noticeable effects of H2 on the vitality and proliferation of E. coli in vitro. Synchrotron FTIR spectroscopy was further con‐ ducted to explore whether H2 NB treatment induces the biochemical component changes of bacterial cells. A total of 20 spectra obtained from the bacterial samples exposed to dd-water or H2 NBs were collected and analyzed. As shown in Fig. 1c, the differences between the average spectra of the two groups of cells were negligible. To differentiate the component alter‑ ations, the second derivative spectra were processed and principal component analysis (PCA) was carried out. The PCA results and loading plots in the whole regions of 3000–2800 cm −1 and 1800–900 cm −1 are shown in Figs. 1d and 1e. The results showed a variance between the spectra of H2 NB and dd-water treatment groups of E. coli within the 95% confidence interval (CI) (Fig. 1d), which indicated that the chemical components of E. coli changed significantly with the existence of H2 NBs. The loading plots of PC1 (61.72%) and PC2 (25.20%) showed that the bands 2920, 2850, 1626, 1109, 1097, and 1024 cm −1 had the largest contribution (Fig. 1e). Notably, the cellular effect of H2 NBs was also distinct with N2 NBs, and the latter could not induce remarkable spectral changes (data not shown), suggesting a specific effect and molecular mechanism of H2 NBs on bacterial cells. In order to investigate which chemical compon‑ ent exhibited significant change, the infrared band shifts of second-derivative spectra were analyzed in detail. The significantly shifted spectral bands were marked with arrows in Fig. 2a and also listed in Table 1. The results showed that after the H2 NB treatment, some infrared bands of the bacteria in the Amides I and II regions (1694.14, 1656.00, and 1544.85 cm −1) shifted to higher wavenumbers; several spectral bands corres‑ ponding to C–H vibrations of fatty acid components (2921.23, 2851.95, 1467.84, and 1452.17 cm −1) shifted to lower wavenumbers; and the P=O asymmetric stretching vibration band of nucleic acid at 1243.18 cm −1 also shifted, as well as the C–O–C vibrations of carbohydrates at 1123.29 and 1057.68 cm −1. The band shifts of the two spectra indicated that chemical com‐ ponent differences in bacterial cells had taken place following exposure to H2 NBs, and these differences could be detected by synchrotron FTIR spectroscopy. To intuitively reflect the bacterial component variations, PCA was further performed on the spectral protein region (1760–1480 cm −1; Fig. 2b), fatty acid region (3000–2800 cm −1 and 1480–1340 cm −1; Fig. 2c), and nucleic acid and polysaccharide region (1350–1000 cm −1; Fig. 2d). From the PCA diagrams, we could establish that both the protein and fatty acid compositions varied dramatically, while the nucleic acid and polysaccharide compositions showed no significant variations. These results indicated that the effects of H2 NBs on bacterial cells mainly acted on the protein and fatty acid components. Since H2 gas is considered to act as a selective antioxidant, we investigated whether its cellular effect would be comparable with that of natural antioxidant tannins. To this end, bacterial cells were treated with different concentrations of tannins and measured by infrared spectroscopy. As shown in Fig. 3a, tannins at 0.5 g/L had no noticeable effect on bacterial growth (indicated in purple), similarly to the effect of H2. These cells treated with tannin and H2 NB-tannin mix‐ ture were used to obtain FTIR spectra. The averaged FTIR absorption spectra of these four groups of bacteria are shown in Fig. 3b, which showed that the band positions and strength of tannin and H2 NB-tannin treatment groups were significantly different from the untreated and H2 NB treatment groups of bacteria. The corresponding second-derivative spectra are shown in Fig. 3c, where the significantly shifted bands of the tannin or H2 NB-tannin treatment group were labeled with arrows. PCA on full spectral regions (3000–2800 cm −1 and 1800–900 cm −1) could not well discriminate the tannin treatment and the H2 NB-tannin co-treatment groups (Fig. 3d). The loading plots of PC1 and PC2 are presented in Fig. 3e. However, PCA on the protein region showed significant differences between the tannin treatment group (purple) and the co-treatment group (blue) (Fig. 3f), while the fatty acid and nucleic acid and polysaccharide components were not clearly distinguishable (Figs. S1–S3). These results suggested that H2 might have a distinct mech‑ anism of action beyond acting as an antioxidant. In conclusion, we have expanded the application of FTIR to evaluate the cellular effects of H2 NBs on the model bacterium E. coli. The obtained results revealed that molecular H2 mainly induces the cellular changes in protein and fatty acid compositions in vitro, indicating that the possible direct target of H2 lies on the membrane. Given the fact that different effects of H2 NBs on bacterial cells were observed compared with natural antioxidant tannins and the results of our previous report on the binding of H2 gas to the hydrophobic sites of proteins (Cheng et al., 2020), we believe that the cellular mechanism of H2 molecule and/or NB is beyond being an antioxidant, i.e., H2 might have other effects besides the widely described antioxidant effect. Thus, our findings provide a novel insight into the possible mechanism of action of molecular H2, as well as other gases, at the cellular level. To our knowledge, this is the first report on the effects of H2 on bacterial cells at the chemical molecular level in vitro, and in turn justifies further scientific endeavors in this direction.

Study details

Article type

Original Research

Dose comparison

No explicit dose/concentration comparison was performed; only presence/absence of hydrogen nanobubbles and tannins were compared.

Dose-dependent effect

No

Outcome

Neutral

Findings

Hydrogen nanobubbles did not affect the growth or proliferation of Escherichia coli in vitro, but induced significant changes in the protein and fatty acid composition of the bacterial cells, as detected by FTIR spectroscopy. These effects were distinct from those of nitrogen nanobubbles and natural antioxidant tannins, suggesting a unique mechanism of action for hydrogen beyond its antioxidant properties.