Deciphering oxidative stress responses in human gut microbes and fecal microbiota: a cultivation-based approach

Abstract

Chronic inflammation creates an oxidative environment, altering the gut microbiota. However, the mechanisms underlying oxidative stress-induced community changes remain poorly understood, owing to the complexity of the host environment, high inter-individual variability, and a lack of comparative data on stress tolerance across intestinal taxa. To address this, we developed an in vitro cultivation approach to assess the effects of oxidative stress, induced by 12 concentrations each of hydrogen peroxide (H₂O₂) and oxygen (O₂), on 41 intestinal strains and seven adults' fecal microbiota. Fusicatenibacter saccharivorans and Lachnospira eligens emerged as particularly sensitive taxa in both pure cultures and complex communities. Oxidative stress also reduced butyrate-producing taxa, like Agathobacter and Anaerostipes, along with total butyrate levels. In contrast, facultative anaerobes, like Escherichia-Shigella and Enterococcus, were largely unaffected, and Bacteroides showed high resilience. Notably, the impact of oxidative stress varied among individuals, with numerous genera showing taxon-specific changes depending on the host microbiota composition. These findings underscore the importance of considering individual microbiota backgrounds when assessing oxidative stress effects on microbial communities. Our study provides a tolerance profile of gut microbes to oxidative stress, reveals overlooked taxa involved in community restructuring, and introduces a screening tool to characterize individual microbial and metabolic responses.

Keywords: anaerobic cultivation; gut microbiota; hydrogen peroxide; intestinal microbes; oxygen; stress tolerance.

Figure 1.

Experimental setup for oxidative stress testing on pure intestinal strains and fecal microbiota using batch cultures. Oxidative stress was induced using a range of H₂O₂ concentrations (0–7.3 mM), and O₂ transfer rates modulated by agar content (0%–0.15%), generating distinct stress dynamics (i.e. initial spike vs. gradual increase). To evaluate responses in pure strains and fecal microbiota, a three-step cultivation protocol was applied. Pure isolates or fecal dilutions (1% v/v) were first inoculated into modified YCFA medium (YCFA-3C for pure strains; bYCFA-6C+muc for complex samples). Following 24 h anaerobic pre-cultivation at 37°C, cultures were transferred into oxidative stress conditions for 24 h (Passage 1), then into post-stress conditions for an additional 24 h (Passage 2).

Figure 2.

Effect of H₂O₂ stress on pure strains (n = 41) and fecal microbiota cultures (n = 7). (A) A panel of taxonomically representative gut strains was tested to determine the maximum H₂O₂ concentrations permitting growth under both stress and post-stress conditions, as indicated by fill levels. Strains are ranked by post-stress tolerance. All strains were tested in biological (n = 3) and technical (n = 2) replicates. Growth data across all conditions are shown in Fig. S5A. Data normality was assessed using the Shapiro–Wilk test; comparisons were made using a t-test (P > 0.05) or Wilcoxon test otherwise. (B) Impact of H₂O₂ on growth, metabolic output, and community diversity across the pre-culture and two subsequent passages of fecal microbiota (n = 7). Lines represent mean values; shaded areas indicate standard deviations. For clarity, only metabolic data of stressed cultures is represented, while the complete set of metabolic data with the corresponding statistical results displayed in Fig. S6. Bray–Curtis dissimilarities were calculated between stressed cultures and the corresponding control. (C) Median clr-abundances of all genera significantly altered under stress conditions compared to controls, as analyzed by ALDEx2. Asterisks denote significance compared to controls (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001). All fecal microbiota data were derived from pooled technical replicates (n = 3).

Figure 3.

Effect of O₂ stress on pure strains (n = 41) and fecal microbiota cultures (n = 7). (A) A panel of taxonomically representative gut strains was tested to determine the maximum O₂ transfer level permitting growth under both stress and post-stress conditions, as indicated by fill levels. All strains were tested in biological (n = 3) and technical (n = 2) replicates. Growth data of all conditions are displayed in Fig. S10A. Data normality was assessed using the Shapiro-Wilk test, followed by a t-test for P > 0.05 or a Wilcoxon test otherwise. (B) Impact of O₂ on growth, metabolic output, and community diversity across the pre-culture and two subsequent passages of fecal microbiota (n = 7). Lines represent mean values; shaded areas indicate standard deviations. For clarity, only the metabolic data of stressed cultures is represented, while the complete set of metabolic data and corresponding statistical results are displayed in Fig. S12. Bray–Curtis dissimilarities were calculated between stressed cultures and the corresponding control. (C) Median clr-abundances of all genera that were significantly altered under stress conditions compared to controls, as analyzed by ALDEx2. Asterisks denote significance compared to controls (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001). All fecal microbiota data were derived from pooled technical replicates (n = 3).