Potential of using an engineered indole lactic acid producing Escherichia coli Nissle 1917 in a murine model of colitis

Abstract

The gut microbiome is a significant factor in the pathophysiology of ulcerative colitis (UC), prompting investigations into the use of probiotic therapies to counter gastrointestinal inflammation. However, while much attention has been given to the therapeutic potential of microbes at the species and strain level, the discovery and application of their metabolic products may offer more precise and controlled solutions in battling disease. In this work, we examined the therapeutic potential of indole lactic acid (ILA) to alleviate inflammation in a murine model of colitis. A previously constructed ILA-producing Escherichia coli Nissle 1917 strain (EcN aldh) and its isogenic non-ILA producing counterpart (EcN) were studied in a murine model of Dextran Sodium Sulfate (DSS) induced colitis. The colitic animals suffered from severe colitic symptoms, with no differentiation between the groups in body weight loss and disease activity index. However, three days after cessation of DSS treatment the EcN aldh-treated mice showed signs of reduced intestinal inflammation, as manifested by lower concentrations of fecal lipocalin-2. Additionally, expression analysis of the inflamed tissue revealed distinct effects of the EcN aldh strain on proteins associated with intestinal health, such as TFF3, occludin and IL-1β expression. These results show no impact of EcN or EcN aldh on acute DSS-induced colitis, but suggest that in particular EcN aldh may assist recovery from intestinal inflammation.

Figure 1

DSS-induced colitis: macroscopic observations and markers of disease. (A) Experimental setup. (B) Body weight loss calculated relatively to the AMT administration day (Day 4). Statistical analysis performed with the mixed effects model (REML), uncorrected Fisher’s LSD. (C) Colon length measurements and (D) spleen weight measurements on termination day. Statistical significance was calculated by Kruskal Wallis, uncorrected Dunn test, and by ordinary one-way ANOVA, Fisher’s LSD test, respectively. (E) Fecal lcn-2 measurements on selective days of the experiment. Statistical analysis was performed with a two-way ANOVA, Tukey correction (adjusted p-values are depicted on the illustration). All data is shown as the mean ± SEM of the biological replicates from each group (n = 10, 10, 9, 8 in the non-colitic, DSS, EcN and EcN aldh groups respectively). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 2

Histopathological analysis: influence of EcN aldh on inflammatory cell infiltrate and intestinal architecture. (A) Histological scores of colon sections. Data was generated by the histopathological analysis of 7 mice from each colitic group and 4 from the non-colitic group. The plots depict the mean ± SEM in each group. Statistical significance was calculated by Kruskal–Wallis, uncorrected Dunn’s test (inflammatory cell filtrate scores) and one-way ANOVA, uncorrected Dunn’s test (intestinal architecture scores). *p < 0.05 (B) Representative images of H&E-stained colon sections. Histological features: a: mucosal immune cell infiltration, a: mucosal and submucosal immune cell infiltration, b: goblet cells loss, c: focal hyperplasia, d: ulcer formation.

Figure 3

Impact of the AMT strains on fecal and serum aLAs. Fold change of (A) ILA, (B) PLA and (C) 4HPLA concentration in the fecal samples from D0 to D10. Values were calculated for each mouse individually. The bars depict the mean ± SEM of each group (n = 7, 9, 6, 4 in the non-colitic, DSS, EcN and EcN aldh groups respectively). Statistical significance was calculated with the Kruskal–Wallis method, uncorrected Dunn’s test. (D) ILA, (E) PLA and (F) 4HPLA concentrations in serum samples on day 9. The graphs illustrate the mean ± SD of the biological replicates (n = 8, 9, 7, 6 in the non-colitic, DSS, EcN and EcN aldh groups respectively). Each symbol represents an individual mouse. Statistical significance was calculated with the Kruskal–Wallis method, uncorrected Dunn’s test. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4

Impact of EcN aldh on molecular markers of intestinal epithelial recovery and inflammation. Relative expression of proteins related to (A–C) intestinal epithelial health (TFF3, OCL, Muc3), and (D–H) immunological markers (IL-1β, iCAM, iNOS, RIPK1, Foxp3, AhR, TRL9, TGF-β) in colonic tissue. The RT-qPCR results were normalized to the expression of NONO. The gene expression is shown relative to the non-colitic animals. Bars represent the mean ± SEM of the biological replicates (n = 9, 7, 7, 7; non-colitic, DSS, EcN and EcN aldh groups respectively). Statistical analysis was performed with one-way ANOVA, Fisher’s LSD test, except for IL-1β, iCAM-1, iNOS and TLR9 data that was Kruskal–Wallis, uncorrected Dunn’s test was used. Asterisks (*) on top of the bars note the p values in comparison to the non-colitic mice. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 5

Compositional changes resulting from the AMT administration. (A) Observed richness of the four experimental groups. Box plots depict the median, minimum, and maximum values from the independent biological replicates on the selected days. (B) The microbiota composition of the experimental groups based on the detected genera on selected days of the experiment. (Relative abundance cutoff 0.1%) *q < 0.05, **q < 0.01, ***q < 0.001, ****q < 0.0001.