The E. coli Nissle 1917 (EcN; O6:K5:H1) strain was purchased as the probiotic drug Mutaflor (Ardeypharm, Herdecke, Germany). The E. coli O6:K13:H1 (EcO) was originally isolated from the urine of a patient with cystitis and was kindly provided by Prof. L. A. Hanson (Department of Immunology, Institute of Medical Microbiology, University of Gothenburg, Sweden). For in vitro adhesivity and invasivity analysis, the adherent-invasive E. coli LF82 strain (a gift from Arlette Darfeuille-Michaud, France) was used as the positive control (Boudeau et al., 2003).

All E. coli strains were cultivated in Luria-Bertani (LB) broth (Himedia Laboratories, USA) overnight at 37°C under aerobic conditions. For in vitro adhesivity and invasivity assays on the Caco-2 cell line, a concentration of 2 x 10⁷ CFU/ml in PBS was used.

Formalin-inactivated bacteria were used for in vitro stimulation of human embryonic kidney cells and bone marrow-derived dendritic cells (BM-DCs), as previously described (Srutkova et al., 2015). Briefly, bacterial cells were plated on LB agar (Himedia Laboratories, USA) to determine the concentration of the bacterial suspension before inactivation. Bacteria were then centrifuged, washed with phosphate-buffer saline (PBS), and inactivated with 1% formalin in phosphate-buffer for 3 h at room temperature. After inactivation, cells were centrifuged, washed three times with PBS, adjusted to 1 x 10⁹ CFU/ml in PBS, and stored at 4°C.

For the in vivo study, EcN or EcO strains were cultivated overnight in LB broth at 37°C under aerobic conditions, centrifuged (4000 RPM for 10 min at 4°C), and diluted in PBS according to optical density to an approximate concentration of 0.5–1.0 x 10⁹ CFU/ml. The final E. coli concentration administered to the mice was determined by bacteria plating on LB agar as 1.9 x 10⁸ CFU/0.2 ml for EcN and 1.1 x 10⁸ CFU/0.2 ml for EcO, respectively.

EcN and EcO strains grown in 5 ml of LB broth at 37°C overnight were processed as follows: The bacteria were fixed in 3% glutaraldehyde with cacodylate buffer for 15 min at room temperature, followed by overnight incubation at 4°C. The extensively washed cells were sedimented overnight onto poly-L-lysine-treated circular coverslips at 4°C. The coverslips with the bacteria were then post-fixed in 1% OsO₄ for 1 h at room temperature.

The washed coverslips were dehydrated through a graded alcohol series (25%, 50%, 75%, 90%, 96%, 100%, and 100%; each step for 20 min), followed by absolute acetone, and then dried from liquid CO₂ in a K850 Critical Point Dryer (Quorum Technologies Ltd., Ringmer, UK). The dried samples were sputter-coated with a 3-nm layer of platinum in a Q150T Turbo-Pumped Sputter Coater (Quorum Technologies Ltd., Ringmer, UK). The final specimens were examined using CBS and TLD detectors in an FEI Nova NanoSEM scanning electron microscope (FEI, Brno, Czechia) at 5 kV (Benada and Pokorny, 1990; Klimentova et al., 2019).

The EcN and EcO strains were grown in 5 ml of LB broth at 37°C until the exponential phase for 3 h. Then, the bacteria were negatively stained on glow-discharge-activated formvar/carbon grids (Benada and Pokorny, 1990) with 1% ammonium molybdate supplemented with 0.1% trehalose (Harris and Scheffler, 2002). The final samples were viewed in a Philips CM100 electron microscope (FEI, formerly Philips EO, The Netherlands) at 80 kV. Digital images were recorded using a Veleta slow-scan camera (EMSis, Muenster, Germany) at a magnification of 25,000 × or 46,000 × and processed with AnalySis 5.2 software (Olympus Soft Imaging Solutions GmbH, Muenster, Germany).

The human epithelial colorectal adenocarcinoma cell line (Caco-2) was kindly provided by Zuzana Jiraskova Zakostelska (Institute of Microbiology of the Czech Academy of Sciences). Cells were cultured in Dulbecco's Modified Eagles high glucose, supplemented with 10% fetal calf serum (FCS), 0.01 M HEPES, 1% penicillin/streptomycin, 1% non-essential amino acids, and 2 mM L-glutamine (all Sigma-Aldrich, CO., St. Louis, USA). The cells were routinely cultured in a 75 cm² tissue culture flask (Techno Plastic, Trasadingen, Switzerland) at 37°C in a humidified atmosphere with 5% CO₂. The medium was changed every second or third day, and cells were re-seeded with trypsin (Sigma-Aldrich, CO., St. Louis, USA) when the monolayer reached 80% confluence. The cell line between the 25th and 30th passages was used for analysis.

Caco-2 cells at a concentration of 2 x 10⁴ per 100 μl of media were added to the apical chamber, and 500 μl of complete media was added to the basolateral chamber of the transwell insert in a 24-well plate (Corning, Sigma-Aldrich, CO., St. Louis, USA). The cells were cultured for 21 days, with medium changes in both chambers every second or third day. Afterward, the medium was removed, and the cells were washed two times with PBS to remove any remaining antibiotics.

Live bacteria at a concentration of 2 x 10⁶ CFU in 100 μl of antibiotic-free medium were added to the apical part, and 500 μl of antibiotic-free medium was added to the basolateral chamber. After 5 h of incubation at 37°C in a humidified atmosphere with 5% CO₂, both chambers were washed three times with 2 ml of PBS.

To determine the adhesive properties of the bacteria, 100 μl of 0.1% Triton-X100 (Sigma-Aldrich) was added to the apical chamber, and the cells with adhered bacteria were harvested by pipetting up and down. The cells were then plated onto Petri dishes with LB agar to determine the number of bacteria adhered to the cells.

To detect invasive properties, transwell inserts with invaded bacteria were washed three times with PBS and then incubated with culture media supplemented with 2% penicillin, streptomycin, and gentamycin (all Sigma-Aldrich) to eliminate adhered bacteria. After 1 hour of incubation at 37°C, the transwell inserts were washed three times using PBS, and the cells were harvested using 100 μl of 0.1% Triton-X100. The cells were plated on LB agar. The results from three independent experiments are pooled and expressed as a percentage of adhered or invaded bacteria from the input bacterial concentrations.

Human embryonic kidney (HEK) 293 cells stably transfected with plasmid carrying human gene TLR4/MD2/CD14 (a gift from Prof. M. Yazdanbakhsh, Leiden, Netherlands), TLR2/CD14- (InvivoGen, USA), or hNOD2 (InvivoGen, USA) were used to test pattern recognition receptor modulatory analysis (Schwarzer et al., 2013). Briefly, cells (2 × 10⁵ cells/ml) were cultured in Dulbecco's complete medium (Sigma-Aldrich) supplemented with appropriate antibiotics (Blasticidin and Hygromycin B Gold from Invivogen, USA) in a final volume of 200 ul per well in a 96-well plate, according to the manufacturer's recommendation. Once the cells adhered to the bottom of the well, EcN or EcO at concentrations of 1 x 10⁸, 1 x 10⁷, and 1 x 10⁶ CFU/ml were added.

hTLRs and hNOD2 agonists were used as a positive control to activate selected signaling pathways: lipopolysaccharide (LPS-EB, 1 μg/ml, InvivoGen, USA) for the HEK 293/hTLR4 pathway, synthetic triacylated lipopeptide PAM3CSK4 (PAM3, 1 μg/ml, InvivoGen, USA) for the HEK 293/hTLR2CD14- pathway, and muramyl dipeptide (MDP, 1 μg/ml, InvivoGen, USA) for the HEK 293/hNOD2 pathway. Cultivation medium was used as a negative control.

Culture supernatants were harvested after 20 h of stimulation with formalin-inactivated EcN and EcO, and the cytokine level of human interleukin 8 (IL-8) was measured by ELISA according to the manufacturer's recommendation (Invitrogen, Thermo Fisher Scientific, USA). Each experiment was repeated at least four times.

BM-DC were isolated from naïve BALB/c mice (8 weeks old, n = 2) and prepared as previously described (Srutkova et al., 2015). Briefly, bone marrow precursors isolated from femurs and tibias were seeded at 2 × 10⁷ cells/ml in RPMI 1640 culture medium (Sigma-Aldrich) containing 10% FBS, penicillin/streptomycin, 0.01 M HEPES, and 20 ng/ml mouse GM-CSF (eBioscience, USA) and incubated for 8 days at 37°C in an atmosphere with 5% CO₂. Fresh medium was added on days 3 and 6. Cells were then harvested, diluted to the final concentration of 2 × 10⁶ cells/ml, and stimulated with the formalin-inactivated EcN and EcO strains individually or in a mixture (1:1) at a concentration of 2 x 10⁷ CFU/ml for 18 h.

For stimulation using subsequently both bacteria, BM-DCs were cultivated with the first bacterial strain (2 x 10⁷ CFU/ml) for 5 h at 37°C in 5% CO₂, washed with culture medium, and then the second bacterial strain was added and cultured for 13 h. Ultrapure LPS (1 μg/ml; InvivoGen, USA) was used as a positive control; unstimulated cells (medium) served as a negative control. The levels of IL-10 and IL-12p70 were determined using Ready-Set-Go! Kit (eBioscience, USA) according to the manufacturer's instructions.

For cell surface marker analysis, BM-DCs were labeled for 30 min at 4°C with anti-mouse FITC-conjugated CD11c, APC-conjugated MHC II, and PE-conjugated CD40, CD80, or CD86 monoclonal antibodies (eBioscience, USA). The data were acquired on a BD FACS Calibur flow cytometer (BD Biosciences, USA) and analyzed with FlowJo software 7.6.2 (TreeStar, USA).

Germ-free (GF), E. coli Nissle 1917, or E. coli O6:K13:H1 monocolonized and experimental groups of BALB/c mice were kept under sterile conditions in Trexler-type plastic isolators in a room with a 12 h light-dark cycle at 22°C. They were fed a 50 kG irradiated sterile pellet diet (Altromin, Lage, Germany) and provided with sterile drinking water ad libitum.

Sterility and colonization status were controlled every 2 weeks by confirming the absence of unwanted bacteria, molds, and yeasts through aerobic and anaerobic cultivation of mouse feces and swabs from the isolators in VL (Viande-Levure), Sabouraud-dextrose, and meat-peptone broth, followed by plating on blood, Sabouraud, and VL agar plates.

Specific pathogen-free (SPF) mice were kept in individually ventilated cages (IVC) (Tecniplast S.P.A., Italy), with a 12 h light-dark cycle at 22°C, fed a 25 kG irradiated sterile pellet diet (Altromin, Lage, Germany), and provided with drinking water ad libitum. SPF mice were regularly checked for the absence of potential pathogens according to an internationally established standard (FELASA). Animal experiments were approved by the committee for the protection and use of experimental animals at the Institute of Microbiology of the Czech Academy of Science (approval ID: 18/2019).

Germ-free (GF) mice were colonized by intragastric gavage with either E. coli Nissle 1917 (Mutaflor) or E. coli O6:K13:H1, and mated after a 2-week colonization period. Their offspring were used for in vivo studies.

Subsequently, 2-month-old female offspring were divided into experimental groups according to bacterial colonization and DSS-colitis induction:

After 15 days of the second colonization, mice received autoclaved 2.5% w/v dextran sulfate sodium (DSS, M.W. 36–50 kDa; MP Biomedicals, Illkirch, France) in drinking water ad libitum for 7 consecutive days to induce acute colitis (see Figure 1, experimental design). Mice colonized by EcN+EcO or EcO + EcN without DSS-induced colitis served as healthy controls. The experiment was repeated three times. To ensure the efficiency of the 2.5% DSS solution, 8–10 weeks-old BALB/c mice reared in SPF conditions were used as a positive control for acute intestinal inflammation, where signs of inflammation such as rectal bleeding and weight loss were apparent (data not shown). At the end of the experiment, occult bleeding was evaluated in stool samples using the hemocult test (Hemocare, Care Diagnostic GmbH, Austria) and scored as follows:

After sacrifice colon were aseptically removed, its length measured, and tissue segments fixed in Carnoy's fluid for 30 min, transferred into 96% ethanol, and embedded in paraffin.

Paraffin-embedded sections (5 μm) were cut and stained with hematoxylin and eosin (H&E) or Alcian Blue/Nuclear Fast Red (all from Vector Labs, Burlingame, USA) for mucin production. The samples were viewed under an Olympus BX 40 microscope equipped with an Olympus Camedia DP 70 digital camera, and the images were analyzed using Olympus DP-Soft. The degree of surface epithelium damage, crypt distortion, and mucin production in individual colon segments was evaluated, and the histopathological score was assessed according to Hudcovic et al. (2001).

Colon-descendent segments were embedded in O.C.T. Tissue-Tek (Sakura Finetek, Alphen aan den Rinj, Netherlands) and frozen in liquid nitrogen. Acetone-fixed cryosections (5 μm thick) were used for immunohistochemistry. Expression of ZO-1 was detected using rabbit anti-mouse polyclonal antibody (Zymed Laboratories, Carlsbad, CA, USA) and secondary antibody Cy3 goat anti-rabbit IgG (Biomeda, Foster City, USA). The samples were viewed under an Olympus BX 40 microscope equipped with an Olympus Camedia DP70 digital camera, and the images were analyzed using Olympus DP-Soft (Jourova et al., 2022).

Fecal samples were collected 1 day before (day-1) intragastric administration of the second E. coli strain, 12 h after the administration of the second strain (day 0), and on subsequent days throughout the colonization of mice as depicted in the experimental design (Figure 1). Samples were stored at −40°C for qPCR analysis. A part of the fecal sample was serially diluted with sterile PBS and plated on modified McConkey agar supplemented with 2% raffinose, prepared according to Splichalova et al. (2014) (see Supplementary materials and methods and Supplementary Figure S2). Total DNA from the feces of colonized mice or pure E. coli Nissle 1917/E. coli O6:K13:H1 culture was isolated using the ZR Fecal DNA Kit according to the manufacturer‘s instructions (Zymo Research, USA). Agarose gel electrophoresis and UV spectrophotometry confirmed the purity, integrity, and concentration of nucleic acids.

To generate a standard curve for qPCR analysis, bacterial DNA was extracted from 150 μl of 10x diluted fecal samples in PBS (w/v) of monocolonized mice (EcN or EcO). Subsequently, the bacterial concentration (CFU/g of feces) corresponding to 1 μl of isolated DNA was determined by plating on LB agar (Himedia Laboratories, USA) and cultivating aerobically overnight at 37°C. A five-point standard curve (10-fold dilutions corresponding to 10⁶ to 10² CFU per 1 μl of DNA) was used for absolute bacterial quantification of E. coli Nissle 1917 or O6:K13:H1 in qPCR reactions. The C_T values at the different dilution points were averaged, and the total number of bacterial cells was interpolated from the averaged standard curve as described elsewhere (Ott et al., 2004; Costa et al., 2022).

Amplification and detection were carried out in 96-well optical plates using the Bio-Rad Real-time PCR Detection System iQ5 (Bio-Rad laboratories, USA) in a final volume of 20 μl per reaction containing 1 μl of a 10x diluted sample of DNA, 2x MM Syto9 (Top-Bio, Czechia), a 0.4 μM concentration of each primer, and PCR water. For the detection of E. coli Nissle 1917, primers Muta-9/Muta-10 (forward: 5'gcg agg ta acct cga aca tg 3'; reverse: 5'cgg cgt atc gat aat tca cg 3') according to Blum-Oehler et al. (2003) were used. The quantitative PCR was performed with an initial hold of 95°C for 3 min, followed by 35 cycles of denaturation at 95°C for 45 s, annealing at 60°C for 45 s, and elongation at 72°C for 45 s, and with a final elongation step at 72°C for 10 min (Blum-Oehler et al., 2003).

For the amplification of E. coli O6:K13:H1, primers detecting the hlyA gene were used (forward: 5'tca gga act ggt ttg aaa aa 3'; reverse: 5'att acc tgc agc tga aat ga 3') (Sheshko et al., 2006). The quantitative PCR was performed with an initial hold of 95°C for 5 min, followed by 35 cycles of denaturation at 95°C for 45 s, annealing at 56°C for 45 s, and elongation at 72°C for 45 s, with a final elongation step at 72°C for 10 min. After each qPCR run, melting curve analysis was performed to verify the presence of the desired amplicon.

Fragments of the distal colon were washed with PBS to remove fecal contents, weighed, cut to approximately 100 mg each, and cultured in RPMI 1640 supplemented with 100 U/ml penicillin, 100 μg/mL streptomycin, 1% HEPES (all Sigma-Aldrich), and 10% heat-inactivated fetal calf serum at 37°C, a 5% CO₂ for 48 h. Supernatants were collected and stored at −40°C until further processing (Hudcovic et al., 2012). The levels of cytokines IL-1β, IL-6, and TNF-α were determined using a Mouse Cytokine/Chemokine Multiplex Immunoassay (Millipore-Sigma, MO, USA) according to the manufacturer's instructions.

Data are expressed as mean ± standard error of the mean (SEM). Data were analyzed using the Student's t-test or one-way ANOVA with Tukey's post-hoc test. All statistical analyses were conducted using GraphPad Prism 8.0 Software (San Diego, CA, USA).

Scanning (Figures 2A, B) and transmission (Figures 2C, D, Supplementary Figures S1A, B) electron microscopy were performed on E. coli Nissle 1917 (EcN) and O6:K13:H1 (EcO) cultures to investigate the surface structures of bacterial cells. An overnight culture with elongated bacterial cells in a stationary growing phase was used for scanning electron microscopy analysis. For TEM analysis, a fresh culture of E. coli cells in the exponential phase was used to ensure the fine structures of both flagella and pili. Both strains exhibited regular rod-shaped structures with a mean cell length of approximately 1–2 μm and a mean thickness of approximately 0.8–1.2 μm (Supplementary Figure S1C).

Scanning electron microscopy of EcN (Figure 2A) and EcO (Figure 2B) displayed a formed filamentous network with flagella connecting bacterial cells. The transmission electron microscopy analysis (Figures 2C, D, Supplementary Figures S1A, B) confirmed the presence of typical filamentous structures on the surface of strains, including flagella and pili. The long and thicker flagella had an average diameter of 17 ± 2 nm (Supplementary Figure S1D), and the thinner and shorter pili had diameters smaller than 10 nm. The presence of pilli and flagella structures was found on both EcN and EcO, ensuring their potential for adhesion to the epithelial cells in the host or cell cultures.

The ability of EcN and EcO to adhere to and invade epithelial cells was tested using a polarized Caco-2 transwell culture, which mimics the intestinal environment. The adherent/invasive strain Escherichia coli LF82, originally isolated from the ileum of a patient with Crohn's disease and capable of disrupting the apical junctional complexes in polarized epithelia, was used as the positive control (Boudeau et al., 2003; Wine et al., 2009). As expected, E. coli LF82 showed high adhesive and invasive ability in the polarized Caco-2 cell line.

Both analyzed EcN and EcO strains displayed weaker efficiency in adhering to (Figure 2E) and invading epithelial cells compared to the positive control (Figure 2F). Nonetheless, EcO was significantly more adhesive and showed a trend toward higher invasiveness than EcN. These results suggest that EcO, as a uropathogenic strain, has better mechanisms for attaching to the epithelial cells of the host.

HEK 293 cells stably transfected with human genes for pattern recognition receptors TLR4, TLR2/CD14- or NOD2 were used to investigate the innate signaling pathways activated by both E. coli strains. Our results showed that both strains are strongly recognized by TLR4 and TLR2/CD14- receptors in a similar dose-dependent manner (Figures 3A, B). There were no differences in signaling through the TLR4 or TLR2/CD14- pathways between the studied strains at the tested bacteria concentrations, suggesting that surface structures such as LPS or lipoproteins are sensed similarly. However, EcO was recognized more intensively by the NOD2 receptor than EcN, suggesting differences in the sensing of peptidoglycan structures (Figure 3C).

BM-DCs derived from naïve BALB/c mice were used as an in vitro model to investigate the immunostimulatory potential of both E. coli strains on immune antigen-presenting cells. The expression of co-stimulatory markers CD40, CD80, and CD86 was assessed to determine the activation status of BM-DCs after stimulation with each bacterial strain. These surface markers were highly induced by both strains, comparable to the positive LPS control (Figure 4A).

BM-DC cultures were stimulated by E. coli strains separately (EcN or EcO), in a mixture of both strains (EcN+EcO), or sequentially pre-incubated with the first strain (EcN or EcO) and after washing, stimulated with the second strain (EcO or EcN). The levels of IL-10 and IL-12p70 were measured in the BM-DCs supernatants (Figure 4B). The data show that stimulation of BM-DCs with the EcN strain alone induces a higher anti-inflammatory cytokine ratio of IL-10/IL-12p70 compared to the EcO strain. This anti-inflammatory effect was maintained even when BM-DCs were pre-incubated with EcN and then stimulated with EcO. However, this effect was not apparent in BM-DC co-cultures with the EcN+EcO mixture or when EcN was added as the second stimulus after EcO (Figure 4B). These results indicate that EcN induces an anti-inflammatory response in immune cells that persists even after subsequent exposure to EcO. Conversely, EcN added as the second strain cannot alter the immune response pre-induced by the pathogenic EcO strain.

To investigate the effect of priority colonization on the host, we used an in vivo mouse model of acute ulcerative colitis induced by administering 2.5% DSS in drinking water. Mice neonatally colonized by EcN received one dose of EcO (1.1 × 10⁸ CFU) intragastrically 15 days prior to colitis induction (EcN+EcO group). To test the anti-inflammatory potential of EcN in neonatally EcO monocolonized mice, one dose of EcN (1.91 × 10⁸ CFU) was administered intragastrically as a second strain 15 days before the induction of DSS-colitis (EcO+EcN group).

The colonization efficiency of both E. coli strains was analyzed in fecal samples by qPCR or bacteria cultivation on modified McConkey agar supplemented with 2% raffinose on indicated days (see experimental design Figure 1, Supplementary Figure S2). We found that the concentration of EcN in neonatally colonized mice reached up to 10¹⁰ CFU/g of feces (Figure 5A, Supplementary Figure S2C) in comparison to EcO neonatal colonization, which reached 10⁸ CFU/g of feces (Figure 5B, Supplementary Figure S2D). The concentrations of EcO bacteria remained stable in both experiment setups, regardless of whether the EcO strain was the first (neonatally colonized) or the second strain in EcN-colonized mice.

Conversely, EcN administered as the second strain did not reach as high a concentration as in the neonatal setup, maintaining a level of 10⁸ CFU/g of feces (Figure 5B, Supplementary Figure S2D).

Further, we investigated the priority colonization effect by colonizing germ-free mice at 6 weeks of age with either individual EcN or EcO strains or with a 1:1 mixture of EcN and EcO (Supplementary Figure S3A). Mice mono-colonized with individual EcN and EcO strains exhibited stable high colonization levels, reaching 10⁹-10¹⁰ CFUs per gram of feces from the second day of the experiment. We detected a high concentration of the EcO strain in monocolonized mice, but in mice colonized by a mixture of EcN+O, the concentration of EcO was 10 times lower. Furthermore, the concentration of EcN strain in feces reached nearly the same level in mice mono-colonized with the individual strain, bi-colonized with an EcN+O mixture (Supplementary Figure S3B), or EcN-neonatally colonized mice with EcO as the second strain (Figure 5A, Supplementary Figure S2C). Despite the stable coexistence of both E. coli strains in the mouse intestine, the EcN strain was able to outcompete the abundance of EcO by a competitive interaction, showing a 100-fold reduction in EcO abundance in neonatally pre-colonized EcN mice or a tenfold reduction in mice colonized simultaneously with an EcN+O mixture.

Disease progression in experimental mice was characterized by rectal bleeding, the shortening of the colon length, and histological evaluation after DSS treatment (Figures 5C–E, Supplementary Figures S3C–E). The histopathological scoring was based on the infiltration of inflammatory cells into the lamina propria and submucosa, submucosa thickening, and the loss of the entire crypt with retained surface epithelium (Hudcovic et al., 2001). Rectal bleeding and the shortening of colon length were most pronounced in mice monocolonized with EcO strain alone (regardless of the colonization of newborn or adult mice) and treated with 2.5% DSS (EcO+DSS groups) compared to healthy control (EcN+EcO or EcO+EcN without DSS) (Figures 5C, D, Supplementary Figures S3C–E).

Histopathological findings included infiltration of inflammatory cells (IIC) into the lamina propria (LP), the thickening of submucosa (SM), the appearance of oedematic loci (OL) in the mucosa (M) structure, and the disappearance of the goblet cells (GC) from the lamina propria in the EcO+DSS group of mice (Figures 5E–G, Supplementary Figures S3F–H). As previously reported for neonatally colonized mice (Hudcovic et al., 2007), adult mice colonized with the EcN strain and treated with DSS exhibited negligible signs of inflammation and histological characteristics comparable to healthy controls, suggesting its anti-inflammatory potential (Supplementary Figures S3C–H).

Similarly, mice that were neonatally colonized with EcN and then administered the pathobiont EcO as a second strain showed lower occurrence of rectal bleeding after DSS treatment and the preservation of colon length similar to healthy controls (Figures 5C, D). Histopathological analysis revealed only negligible changes in colonic mucosa, characterized by small infiltration of cells into lamina propria, preserved structure of villi and crypts with an undisturbed epithelial layer, and mucus production ability (Figures 5E–G). Conversely, the mice treated with EcN as the second colonizing bacteria (EcO+EcN+DSS) showed bleeding scores and colon shortening similar to EcO-monocolonized mice after DSS treatment. Histological analysis of the colon mucosa confirmed macroscopic signs of inflammation, indicating that one-dose treatment with EcN is ineffective in preventing DSS-induced intestinal inflammation in EcO-colonized mice. Similar results were observed in adult GF mice colonized by a mixture of EcN and EcO (Supplementary Figures S3C–H), showing moderate signs of inflammation, rectal bleeding, colon shortening, and mild histopathological changes in the colonic mucosa.

Altered intestinal barrier function through decreased expression of tight junction proteins, such as ZO-1, is a critical factor in the development of intestinal inflammation in mouse models of colitis or human IBD. In this study, we investigated whether neonatal EcN colonization or one-dose EcN treatment influenced the alteration of ZO-1 induced by DSS treatment in EcO-colonized mice. Immunohistochemistry staining (IHC) revealed reduced ZO-1 expression after DSS administration in the mice colonized with only EcO, compared to healthy controls. In contrast, preserved ZO-1 expression was evident in both neonatally EcN-colonized and one-dose EcN-treated mice (Figure 5H).

These results demonstrate the priority effect of EcN on the host, showing that EcN neonatal colonization prior to EcO colonization protects the host from the induction of the inflammatory response driven by the pathogenic EcO strain.

Changes in the cytokine microenvironment in the gut were analyzed in the supernatants of colon tissue fragment cultures from the mice neonatally colonized by EcN or EcO, co-colonized by EcO or EcN, and treated with 2.5% DSS to induce intestinal inflammation. Similar to macroscopic and histopathological evaluation, the mice monocolonized by EcO and treated by DSS showed a significant upregulation of proinflammatory cytokines IL-1β (Figure 6A), IL-6 (Figure 6B), and TNF-α (Figure 6C) compared to healthy controls without DSS treatment. Notably, we found that neonatal colonization by EcN efficiently protected against the upregulation of these proinflammatory cytokines in DSS-treated mice, even when the potentially pathogenic EcO was administered 2 weeks before colitis induction by DSS. Insignificant downregulation of these cytokines was also observed in mice where EcN was administered as the second colonizing stimulus in one dose only. These results highlight the potential of E. coli Nissle 1917 to protect against inflammation, even in the presence of the potentially pathogenic E. coli O6:K13:H1 strain.