Impact of the defined Oligo-MM12 microbiota on intestinal colonisation and dissemination of Listeria monocytogenes
Monica Cazzaniga, Kardokh Kaka Bra, Mathias K. M. Herzog, Wolf-Dietrich Hardt, Marcus J. Claesson, Harsh Mathur, Cormac G. M. Gahan
TL;DR
This study compares how a simplified microbiota and a conventional microbiome affect Listeria infection in mice, showing that microbiome complexity influences infection outcomes.
Contribution
The study demonstrates that the Oligo-MM12 microbiota model is useful for studying L. monocytogenes infection and colonization resistance mechanisms.
Findings
Oligo-MM12 mice showed higher L. monocytogenes shedding in feces compared to SPF mice.
SPF mice reduced L. monocytogenes levels over time, while Oligo-MM12 mice did not.
Ex vivo fermentation confirmed in vivo patterns, validating the Oligo-MM12 model.
Abstract
Listeria monocytogenes is a foodborne pathogen of global concern, particularly for immunocompromised individuals at risk of severe disease. In mice, infection outcomes are strongly influenced by host immunity and gut microbiome composition. The Oligo-MM12 defined microbiota mouse model, containing a simplified community of 12 bacterial strains, offers a controlled system to study L. monocytogenes pathogenesis and microbiome interactions. Defined or reduced-complexity microbiota models are increasingly used to investigate colonisation resistance and identify protective taxa. In this study, we compared Oligo-MM12 mice with conventionally raised Specific Pathogen Free (SPF) mice to assess how microbiome complexity shapes infection. This allowed us to explore how microbiome complexity affects resistance to L. monocytogenes. We performed an in vivo infection study to assess host responses…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 5
Figure 6- —https://doi.org/10.13039/501100001711Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung
- —https://doi.org/10.13039/100010686H2020 European Institute of Innovation and Technology
- —https://doi.org/10.13039/501100001602Science Foundation Ireland
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsListeria monocytogenes in Food Safety · Salmonella and Campylobacter epidemiology · Vibrio bacteria research studies
Introduction
Understanding the pathogenesis of enteric pathogens and the protective role of the gut microbiota is central to advancing treatments for gastrointestinal infections. A range of murine models have been used to study these interactions, each varying in their ability to replicate human disease outcomes^1^. For some food-borne pathogens such as Campylobacter spp*., conventionally colonised murine models are completely resistant to infection whereas humanised mice are more amenable to pathogen colonisation^2^. Despite similarities at the phylum level, human and murine microbiomes differ significantly at lower taxonomic levels and in functional gene content^3^, complicating efforts to select specific microbial taxa responsible for colonisation resistance. To circumvent this, mice colonised with a defined microbiota allow a better understanding of colonisation resistance, through the addition or removal of specific taxa. The Oligo-MM (Oligo-Mouse-Microbiota) consists of a defined gut microbiome comprising 12 sequenced and culturable strains of mouse-derived bacteria^4,5^. These include Clostridium innocuum, Bacteroides caecimuris, Limosilactobacillus reuteri, Enterococcus faecalis, Acutalibacter muris, Bifidobacterium animalis, Muribaculum intestinale, Flavonifractor plautii, Clostridioformicola clostridioformis, Akkermansia muciniphila, Turicimonas muris and Blautia coccoides^6^.* This community remains stable over time and is vertically transmissible under gnotobiotic conditions, making it an ideal platform for dissecting microbiota-pathogen interactions. Originally developed for studying colonisation resistance against Salmonella enterica serovar Typhimurium^4^, Oligo-MM^12^ has since been adopted for broader investigations into host-microbiota-pathogen dynamics^7,8^.
However, mice (C57BL/6 J, 129S1) with a complex microbiome display colonisation resistance against human enteropathogens and often exhibit a limited disease phenotype^1^. These mice are referred to as “specific pathogen-free” (SPF), and their microbiome is only defined by the absence of certain potential pathogens and thus varies between animal facilities. However, their complexity makes them well-suited to modulation experiments involving antibiotics, diet, or prebiotic supplementation. Reduced-complexity models such as germ-free (GF), Oligo-MM^12^, or low-complexity microbiota (LCM) mice offer alternative approaches, where individual taxa can be added or removed to understand their specific contributions to pathogen inhibition. Previous studies have shown that GF mice are highly susceptible to L. monocytogenes due to the absence of microbiome-derived colonisation resistance^9^^.^
Such findings underscore the protective role of the microbiota against enteric infections. The gut microbiome itself comprises several dominant bacterial phyla, most notably Firmicutes and Bacteroidota, alongside smaller yet functionally relevant populations of Verrucomicrobia, Actinobacteria, and Proteobacteria^10^. The Firmicutes phylum includes diverse bacterial classes such as Lactobacillales and Clostridiales. Lactobacillales are facultative anaerobes that contribute to mucosal homeostasis, antimicrobial peptide production, and amino acid metabolism, making them valuable as probiotics^11–13^. Clostridiales, mainly obligate anaerobes, dominate the cecum and colon and, along with Lactobacillales, aid gut microbiota recovery after disturbances, enhancing colonisation resistance and protecting against food allergies and inflammatory bowel disease (IBD)^9,13^. They produce short-chain fatty acids (SCFAs) like butyrate, supporting colonocyte health, and secondary bile acids that shape microbial ecology^14^. Bacteroidota, comprising 50–60% of the colonic microbiota, include families such as Bacteroidaceae, Prevotellaceae, Rikenellaceae, and Porphyromonadaceae. These organisms play critical roles in polysaccharide degradation, pathogen resistance, vitamin synthesis, and anaerobe support^15,16^. Bacteroides spp. in particular are central to carbohydrate metabolism, producing SCFAs like acetate and propionate, essential for host energy and metabolic balance^17^. Verrucomicrobia is mainly represented by Akkermansia muciniphila, known for enhancing gut barrier integrity, mucus production, and exerting anti-inflammatory and metabolic benefits^18^. Actinobacteria, especially Bifidobacterium and Eggerthella, are integral to gut barrier function, immune modulation, and resistance against pathogens like enterohaemorrhagic Escherichia coli (EHEC), particularly in the infant gut^19,20^. Lastly, Proteobacteria is a diverse phylum that includes opportunistic pathogens such as Escherichia coli, Salmonella spp., Klebsiella spp., and Enterobacter spp.^21,22^. It is likely that this complexity and the potential for variations in microbiome community structure across mouse facilities^23^ has limited the development of a reproducible paradigm outlining interactions between L. monocytogenes and the host microbiome. There is clearly a need to examine simplified model systems that could provide insights into the bi-directional interaction between L. monocytogenes and the microbiome during infection.
To investigate microbial responses to L. monocytogenes infection, we used both in vivo and ex vivo approaches. While in vivo models capture host-microbiota interactions under physiological conditions, host factors like immune responses can obscure microbial-specific effects. Ex vivo systems such as the micro-Matrix bioreactor simulate the distal colon under controlled anaerobic conditions and allow for host-independent analysis of microbiome-pathogen dynamics^24^. Currently, there are no published studies that specifically utilise the micro-Matrix bioreactor system to investigate L. monocytogenes infections.
In this study, we extended previous work^8^, by investigating how the gut microbiome modulates L. monocytogenes infection in SPF and Oligo-MM^12^ mice during a three-day infection period. These models enabled us to dissect both the early colonisation resistance and host-mediated dissemination of L. monocytogenes. To decouple microbial interactions from host influence, we complemented in vivo experiments with an ex vivo infection model using the micro-Matrix system, recreating microbial communities from faecal slurries of both mouse types. This multi-model strategy provides a comprehensive view of microbiome-mediated resistance and microbial dynamics during L. monocytogenes infection.
Methods
Animal experiment
All animal experiments were conducted in compliance with ethical and legal requirements and were reviewed and approved by the Kantonales Veterinäramt Zürich under project license ZH158/19. All methods and procedures were carried out according to Swiss national and cantonal regulations. The experiments were conducted using the same C57BL/6^Bern^ Oligo-MM^12^ and SPF mice as described in Herzog et al.^8^. All experiments were conducted with adult male and female mice 8–12 weeks of age (typically 18–27 g) at the start of the experiment. Experiments are reported according to ARRIVE guidelines. In order to prevent cross-contamination with different pathogens used in the original study (Herzog et al.^8^), full blinding during experimentation was not feasible. However, downstream processing of samples and data analyses were conducted in a blinded manner. During the experiment, all mice were kept in individually ventilated cages and received sterile mouse chow (3437, KLIBA NAFAG, Kaiseraugst, Switzerland) and water. L. monocytogenes EDG-e (murinized InlA) was cultured overnight in liquid BHI medium (Brain Heart Infusion broth; Thermo Fisher Scientific), washed in PBS (Phosphate Saline Buffer, Thermo Fisher Scientific), and used for inoculation. The inoculum size was verified by plating on selective Listeria agar (Oxoid Limited, Wade Road, Basingstoke, Hampshire, RG24 8PW). The mice were orally gavaged with 100 µL of 10^9^ CFU L. monocytogenes EGD-e InlA^m^ (vehicle: PBS). A fresh faeces pellet was collected from each mouse every day (days 1, 2 and 3) after the infection into a sterile and pre-weighed microcentrifuge tube. At the endpoint, mice were euthanised by CO_2_ asphyxiation using a clear Perspex chamber and were observed during the procedure, followed by cervical dislocation. Organs were immediately removed by dissection. Faeces, liver, spleen, mesenteric lymph node (mLN) and cecum content were transferred to sterile and pre-weighed microcentrifuge tubes for CFU quantification. BHI plates were used to determine CFU in liver, spleen and mLN, whereas Listeria selective Agar was used to measure L. monocytogenes densities in faeces and cecum content.
Faecal slurry preparation
Buffer 1 and Buffer 2 were prepared according to Table 1 below. After autoclaving, 5% L-cysteine hydrochloride (Merck Life Science Limited, Vale Road, Arklow, Co Wicklow, Ireland) was added to each buffer at a 1:100 ratio. For Buffer 2, glucose (Merck Life Science Limited) was added at a 1:1 ratio.Table 1. Composition of Buffer 1 and Buffer 2, indicating the volume (mL) of each compound used.Buffer 1 (ml)Buffer 2 (ml)0.1 M NaH_2_PO_4_25538.250.1 M NaH_2_PO_4_24536.75H_2_O50023Final volume100098
Oligo-MM^12^ faecal samples were collected into pre-weighed 1.5 mL microcentrifuge tubes from 22 different mice at the ETH Zurich mouse facility, snap-frozen in Buffer 2, and shipped on dry ice to University College Cork (UCC, School of Microbiology). SPF faecal samples were collected fresh from the APC mouse facility in University College Cork. For this, 2–3 pellets of faeces were placed into pre-weighed 1.5 mL microcentrifuge tubes containing 400 µL of Buffer 1. The tubes were then weighed, kept on ice, and processed further. Both fresh SPF and thawed Oligo-MM^12^ faecal samples were transferred, along with the corresponding buffers, into sterile bags containing a 70 μm filter (Fisher Scientific). The faeces were mashed with the addition of Buffer 1 to facilitate the mashing process. The resulting mixture, containing the bacteria, was collected into 50 mL Falcon tubes (Sarstedt, Wexford, Ireland). A corner of the bag was cut (approximately 4 mm) to allow for easier passage of the sample. The samples were then centrifuged at 4000 rcf for 10 min at 4 °C. After centrifugation, the SPF pellet was resuspended in 300 µL of Buffer 2 and immediately frozen. The aliquots were stored at − 80 °C. Before use in experiments, the SPF frozen samples were thawed in a 37 °C incubator for 30 min. The Oligo-MM^12^ pellets, on the other hand, were resuspended in Buffer 1 and used for the experimental setting. Faecal samples from different mice were pooled to create three biological replicates. For each mouse model, SPF and Oligo-MM^12^, three biological and two technical replicates were used in the experiment (15 mg faeces per mL Buffer).
Bacterial growth
L. monocytogenes EGD-e InlA^m^ was cultured overnight in liquid BHI medium in a shaking incubator at 190 rpm at 37 °C, washed twice in PBS, and 200 µL were used for mouse oral gavage. L. monocytogenes EGD-e was used to infect both SPF and Oligo-MM^12^ faecal slurry into the micro-Matrix bioreactor. An overnight inoculum of L. monocytogenes EGD-e was prepared in BHI medium and incubated at 37 °C in a shaking incubator at 190 rpm. The following day, the inoculum was washed twice with 10 mL PBS and then resuspended in a final volume of 10 mL of PBS. From this suspension, 1 mL was transferred to a fresh tube containing 9 mL of PBS, creating the starting L. monocytogenes inoculum. 100 µL of the initial inoculum was plated onto a BHI agar plate to determine the initial CFU/mL count, corresponding to 5 × 10⁷ CFU/mL.
16S rRNA amplicon sequencing and analysis
The DNA extraction was performed from pellets after 0 h, 24 h, 48 h, 72 h, 96 h, and 120 h for three biological replicates of SPF and Oligo-MM^12^ in the micro-Matrix system and control and Listeria groups), using the QIAamp PowerFecal Pro DNA Kit (Qiagen GmBH, QIAGEN Strasse 1, 40,724 Hilden, Germany) at University College Cork. The variable V3 and V4 regions of the 16S rRNA genes were amplified using 16S Amplicon PCR Forward Primer = 5’ TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG and 16S Amplicon PCR Reverse Primer = 5’ GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC. All libraries were diluted and pooled at 4 nM final concentration before being sequenced by the NextSeq 2000 (Illumina, Inc. San Diego, USA) platform using the P1 600 cycle (2 × 300 bp) reagent kit (Illumina, Inc. San Diego, USA) at Teagasc Next Generation DNA Sequencing Facility in Moorepark Food Research Centre, Teagasc, Cork, Ireland. Demultiplexed, paired FASTQ files were returned from the sequencing facility. The 16S rRNA gene sequencing analysis elucidated the microbial communities present, focusing initially on rigorous quality control of raw sequence data using FastQC, and the removal of adapters and primers via Cutadapt to ensure sequence integrity. The cleaned sequences were further processed and quality-filtered through the DADA2 pipeline within the QIIME2 framework. The forward and reverse reads were shortened to 280 base pairs and 260 base pairs, respectively, and required a minimum of 20 base pairs of overlap for merging reads, thus enhancing sequence assembly accuracy. Accurate Amplicon Sequence Variants (ASVs) were derived after quality trimming, filtering, error correction, and chimera removal. A custom-trained classifier, developed in QIIME2 based on the Greengenes2 2022.10 reference database for the V3-V4 region, was employed for precise taxonomic classification. Microbial community composition was analyzed using Principal Coordinates Analysis (PCoA) based on Aitchison distance metrics. Differences in beta diversity between groups were assessed using PERMANOVA. To compare microbial abundances, the Wilcoxon rank-sum test (also known as the Mann–Whitney U test) was used. This non-parametric test, which does not assume a normal distribution of the data, was applied in three contexts: to compare treatment groups (Control vs. Listeria) on the same day, to assess changes across time points within the Control group, and to assess changes across time points within the Listeria group. P-values resulting from multiple comparisons were adjusted using the Benjamini–Hochberg procedure. Statistical significance was visualized on boxplots by adding significance stars using the geom_signif () function in R.
Ex vivo micro-Matrix fermentation setting
The micro-Matrix bioreactor platform (Applikon Biotechnology, Delft, Netherlands) was used as an ex vivo model of the mouse distal colon. The ex vivo colon model experiments were conducted in sealed micro-Matrix cassettes, having 24 wells/cassette (1–10 mL). Before the test, Faecal Fermentation Medium (FFM) was prepared and a total of 7 mL were added to each well of the micro-Matrix cassette with 100 μL of faecal slurry under anaerobic conditions. The FFM was prepared according to the Fooks and Gibson’s protocol^25^. Faecal slurries were generated from pooled faecal material collected from multiple mice of the same model (SPF or Oligo-MM ), with three independent biological replicates prepared per condition using distinct pools of mice. For the Listeria-treated group, an additional 100 µL of L. monocytogenes inoculum was introduced into each well, while the control group received no bacterial challenge. Following cassette setup, ex vivo colon model experiments were performed according to a previous described protocol^24^. Samples were collected from each well at 24 (d1), 48 (d2), 72 (d3), 96 (d4), and 120 h (d5). At each time point, 1 mL was withdrawn from the anaerobic chamber for DNA extraction, and 100 µL was collected for CFU/mL enumeration. A schematic overview of the study design is presented in Fig. 1.Fig. 1micro-Matrix experimental setting. Faecal samples were collected from SPF mice (University College Cork, UCC) and Oligo-MM^12^ mice (ETH Zurich). Faecal slurries were prepared, and 100 µL of each was inoculated into individual wells of the micro-Matrix, each containing 7 mL of faecal fermentation medium. L. monocytogenes EGD-e was added to designated wells at a concentration of 10⁶ CFU per well. The micro-Matrix system was run for 120 h, with daily replacement of the Faecal Fermentation Media (FFM). At each 24-h interval, 1 mL of culture was removed for downstream 16S rRNA gene sequencing and colony-forming unit (CFU) enumeration.
Statistical analysis
Before conducting statistical analyses, data normality was assessed using the Shapiro–Wilk test, and homogeneity of variances was evaluated with Levene’s test. These assumption checks guided the selection of appropriate statistical methods. A two-way ANOVA was performed to compare CFU/mL counts between SPF and Oligo-MM^12^ mice in both in vivo organ samples and ex vivo micro-Matrix experiments. Post hoc comparisons were conducted using the Benjamini, Krieger, and Yekutieli two-stage linear step-up procedure to control the false discovery rate. To assess differences across time points within the same mouse model, pairwise t-tests were performed. The resulting p-values were corrected for multiple testing using the Benjamini–Hochberg procedure to control the false discovery rate.
Results
Colony forming units (CFU) of L. monocytogenes in a in vivo mouse model and ex vivo micro-Matrix fermentation setting
To investigate the role of the gut microbiome as a first line of defence against L. monocytogenes, we first assessed bacterial dynamics in an ex vivo micro-Matrix fermentation system. Fermentations were conducted using faecal slurries from Oligo-MM^12^ and SPF mice under controlled anaerobic conditions, and L. monocytogenes density was monitored over five days by plating at days 1–5 post-inoculation (Fig. 2A). No significant differences in bacterial growth (CFU/mL) were observed between the two microbiota systems over this period, highlighting that initial proliferation under ex vivo conditions is similar across microbiome complexities. Building on earlier in vivo study^8^, we further analysed bacterial burden in SPF (n = 6) and Oligo-MM^12^ (n = 10) mice during day 3 post-infection. This time point was selected to limit animal suffering, as Listeria infections extending beyond 4–5 days are typically lethal due to progressing systemic infection. As already published^8^, quantification of L. monocytogenes in primary tissues (cecum, faeces) and secondary organs (liver, spleen, mesenteric lymph nodes) revealed no significant differences in overall bacterial burden between the two groups (Supplementary Figure S1). Notably, faecal shedding dynamics differed: Oligo-MM^12^ mice showed lower Listeria levels on day 1 (p = 0.0001), similar levels on day 2, and higher levels on day 3 (p = 0.0048), suggesting that colonisation resistance diminishes over time and that a more complex SPF microbiota may clear the pathogen more efficiently (Fig. 2B). However, when comparing the Listeria dynamics across both the mouse and micro-Matrix experimental settings we can see similar trends. In both SPF mice and the SPF micro-Matrix systems the numbers of Listeria declined over time. In contrast, in Oligo-MM^12^ mice and the micro-Matrix containing the Oligo-MM^12^ microbiota we found reduced control of Listeria numbers over time. This suggests functional similarities between the in vivo (murine) and ex vivo (micro-Matrix) experimental settings.Fig. 2. Quantification of L. monocytogenes colonisation in vivo and ex vivo. (A) CFU/mL counts of L. monocytogenes recovered from micro-Matrix Faecal Fermentation Media (FFM) inoculated with SPF or Oligo-MM^12^ faecal slurry at days 1, 2, 3, 4, and 5 post-infection (B). Colony-forming unit (CFU) counts of L. monocytogenes were measured in faecal samples from SPF and Oligo-MM^12^ mice at days 1, 2, and 3 post-infection. The experiment included six biological replicates for SPF mice and ten for Oligo-MM^12^ mice. It should be noted that the day three data was published previously in Herzog et al. 2025. Three biological replicates were used for each condition. * for P < 0.05, ** for P < 0.01, *** for P < 0.001, and **** for P < 0.0001. Two-way ANOVA was used as a statistical test.
Microbiome analysis
To investigate how the gut microbiota changes during L. monocytogenes infection and to characterize temporal dynamics within the bacterial community, 16S rRNA gene sequencing was performed on DNA extracted from faecal samples of Oligo-MM^12^ and SPF mice at ETH Zurich, as well as from micro-Matrix samples. In the in vivo mouse model study, the PBS group consisted of mice that received an oral gavage of PBS, serving as the uninfected control, whereas the L. monocytogenes group received an oral gavage of L. monocytogenes. In the ex vivo micro-Matrix fermentation study, the control group was composed of faecal slurries prepared from SPF or Oligo-MM^12^ mice without any added pathogen, while the L. monocytogenes group consisted of corresponding faecal slurries inoculated with L. monocytogenes, enabling the assessment of microbiota-pathogen interactions in the absence of host influence.
Alpha and beta diversity analysis
Alpha diversity was assessed using both the Shannon index and Chao1 richness estimator for the in vivo mouse experiment and the ex vivo micro-Matrix fermentation study. Shannon incorporates species evenness and richness, while Chao1 estimates species richness. In the in vivo study, significant differences in alpha diversity were observed (Figs. 3A and 3B). The SPF group had a significant increase in microbial diversity in Listeria-infected and PBS groups at day 2 and 3 (p < 0.001) compared to day 1, indicating recovery or expansion in both richness and evenness following initial disruption. In the Oligo-MM^12^ group, an increase in richness and evenness was reported on day 2 compared to day 1 and day 3 in both Listeria-infected and PBS groups (Figs. 3A and 3B). Statistical analysis showing p-values between each condition are reported in Supplementary Fig. S2.Fig. 3. Alpha and beta diversity analysis in both mouse models and ex vivo micro-Matrix experimental settings. (A)–(B): Alpha diversity metrics were calculated using the Chao1 richness estimator and the Shannon diversity index for both SPF and Oligo-MM^12^ mouse models during the in vivo Listeria infection study. (C), (D), (G), (H): Beta diversity was assessed using Aitchison distances and visualised through principal component analysis (PCA). Pairwise PERMANOVA (ADONIS) showed significant clustering by treatment (Listeria vs. control) in both SPF and Oligo-MM^12^ models (p = 0.001). In the Oligo-MM^12^ ex vivo model, both treatment and time significantly influenced microbial composition (p = 0.001 and p = 0.049, respectively). (E)–(F): In the ex vivo micro-Matrix fermentation model, alpha diversity was calculated using the Chao1 and Shannon indices.
Beta diversity was calculated using Aitchison distances and visualised via principal component analysis (PCA). In the in vivo study, the pairwise PERMANOVA (ADONIS) analyses revealed that treatment (Listeria vs. PBS) significantly affected microbiome composition in both SPF and Oligo-MM^12^ models (p = 0.001), indicating that Listeria infection altered the overall structure of the microbial communities (Figs. 3C and 3D).
In the ex vivo micro-Matrix fermentation study, in Chao1 analysis, the alpha diversity remained stable across treatments (Listeria vs. control) in both SPF and Oligo-MM^12^ slurries, with the exception of day 5 in the Oligo-MM^12^ group, which showed a marked increase in diversity compared to earlier time points (Fig. 3E). Shannon diversity followed a similar trend, showing a gradual increase over time in both groups, with no significant differences between treatment and control. However, by day 5, the Oligo-MM^12^ group displayed a significantly higher Shannon index than at previous time points (Fig. 3F). This late-stage increase in both richness (Chao1) and evenness (Shannon) may reflect the expansion of low-abundance taxa, leading to a more even microbial distribution within the simplified Oligo-MM^12^ microbiota. This was, however, not observed in the SPF group, which was likely due to the greater stability and competitive buffering provided by its complex microbial community.
Beta diversity analysis showed that in the SPF model, time significantly influenced community composition (p = 0.001). However, in the Oligo-MM^12^ model, both treatment (p = 0.001) and time (p = 0.049) had a significant effect on microbial community structure, highlighting the impact of both infection and temporal dynamics under ex vivo conditions (Fig. 3G and 3H).
The distinct patterns observed in alpha diversity during the first three days of infection between the in vivo and ex vivo studies highlight the impact of L. monocytogenes on the gut microbiome. In the in vivo model, rapid shifts in microbial diversity suggest that even a short, two-day exposure to Listeria was sufficient to significantly alter community structure, likely reflecting the dynamic interplay between pathogen, host immune responses, and microbiome composition. In contrast, the gradual and linear increase in alpha diversity observed in the ex vivo micro-Matrix fermentation study may be attributed to the controlled, nutrient-rich environment of the fermentation system, where microbial growth and succession occur without interference from immune responses or other host factors.
Taxonomic analysis at genus level
Following the alpha and beta diversity analyses, relative abundance profiles were assessed to investigate taxonomic shifts in the gut microbiome at the genus level (Fig. 4).Fig. 4. Relative abundance of key genera during Listeria infection. Relative abundance bar plots showing the distribution of the top 8 genera in faecal samples from SPF and Oligo-MM^12^ mice collected at days 1, 2, and 3 post-infection (A) and (B) and until day 5 from the micro-Matrix setting (C) and (D). Each bar represents the average relative abundance per group at each time point, highlighting dynamic shifts in microbial composition throughout infection.
The eight most abundant genera were visualised for each condition in the ex vivo study, and the fifteen most abundant genera in the in vivo study, to explore compositional shifts in microbial communities. In the in vivo mouse model, within the Oligo-MM^12^ group, no major differences appeared between infection days or between the PBS and Listeria groups, aside from minor changes in the Listeria group, such as an increase in Akkermansia and Enterocloster at day 3 (Fig. 4A). In contrast, the SPF mice exhibited a different pattern over time and between groups, showing a clearer effect of Listeria on reducing bacterial composition and diversity. Even at the earliest post-infection timepoint (day 1), the relative abundance profiles differed between the two groups, with an increase in Lactobacillus, Faecalibaculum, and Dubosiella in the Listeria group. By day 3, the Listeria-infected group showed a marked reduction in diversity compared to the control. Alloprevotella increased in the Listeria group, whereas in the PBS group, we detected an increase in Akkermansia and Lactobacillus (Fig. 4B). In the ex vivo micro-Matrix fermentation study, a more gradual and progressive shift in taxonomic composition was observed over time. At day 0, we noticed the presence of only a few genera in both mouse models, indicating that the fermentation system allowed the expansion of other taxa that were initially grouped under “Others” (grey bar). In the Oligo-MM^12^ model, Blautia and Bacteroides appeared to be the most abundant (Fig. 4C), whereas in SPF mice, Lactobacillus and Ligilactobacillus were dominant (Fig. 4D). During the experimental time period, both microbial communities underwent compositional changes, indicating not only the direct effect of Listeria but also a potential fed-batch effect, likely due to the addition of fresh fermentation media every 24 h post-sample collection. This practice can influence microbial dynamics by introducing fresh nutrients and shifting competitive balances^26^. Notably, by day 5, a clear restructuring of the microbial community was observed, with the emergence of a distinct set of dominant genera compared to earlier time points (Figs. 4C and D). Interestingly, we detected the presence of Escherichia/Shigella at day 1 in SPF microbiota in the micro-Matrix setting, suggesting that E. coli was capable of rapid growth in this fermentation model. The fermentation environment may offer favourable conditions, such as reduced competition or available carbon sources, that facilitate its growth^27^.
Distinct sets of dominant genera were observed across the two mouse models and experimental settings, indicating that both microbiota composition (SPF vs. Oligo-MM^12^) and environmental context (in vivo vs. ex vivo) influence community structure.
To identify bacterial genera significantly impacted by L. monocytogenes infection, statistical comparisons were performed between the control/PBS and the Listeria-infected groups across various time points in both in vivo and ex vivo settings (Fig. 5). In this analysis, PBS samples from days 1 and 3 (in vivo) and Control samples from days 1 and 5 (ex vivo) were clustered together, forming a distinct group. Notably, numerous changes were observed both within and between groups during Listeria infection; however, we focus here on the most relevant alterations in the context of this pathogen.Fig. 5. Comparative analysis of key genera during Listeria infection. Boxplots comparing the relative abundance of selected genera showing significant changes between Listeria-infected and PBS groups (A) and (B) or Control group (C) and (D) within each mouse model. Listeria infection days are indicated as L1, L2, L3 and L5, whereas the Control group is indicated as CTRL. Data are shown for individual timepoints, with statistical significance indicated by * for P < 0.05, ** for P < 0.01, *** for P < 0.001, and **** for P < 0.0001. The Wilcoxon rank-sum test was used to assess microbial abundances.
In the in vivo model using Oligo-MM^12^ mice, Akkermansia levels significantly decreased on day 2 (p = 0.017), followed by a marked increase by day 3 (p = 2.2e-05) compared to the PBS group (Fig. 5A). In contrast, Akkermansia levels remained unchanged in SPF mice throughout the infection period. On day 1, Faecalibaculum (p = 0.04), Lactobacillus (p = 0.014), and Prevotellamassilia (p = 0.013) were significantly reduced in the Listeria-treated SPF group compared to PBS controls. By day 3, Bacteroides and Blautia were also significantly decreased (p = 0.017), whereas Prevotellamassilia showed a significant increase (p = 0.00025) in the Listeria-infected group (Fig. 5B), indicating ongoing microbial restructuring during infection. Notably, when comparing these taxa across both mouse models, no consistent temporal patterns were observed. Listeria abundance in the in vivo model was also assessed using sequencing data and showed temporal trends consistent with faecal CFU/mL plate counts (Fig. 1) (Supplementary Figure S3).
The ex vivo micro-Matrix fermentation model was used to investigate further microbiota alterations induced by Listeria exposure over five days post-infection. In the Oligo-MM^12^ not many taxa exhibited a significant increase or decrease with respect to the control (Fig. 5C). In the SPF-derived ex vivo system, Listeria abundance consistently increased from day 1 through day 5 (p = 0.017) relative to the Control group (Fig. 5D). A comparison between the two faecal slurry fermentation models revealed shared trends, including increased Listeria abundance and a common rise in Lysinibacillus by day 5.
Taxonomic shifts in response to L. monocytogenes infection differ between mouse models and experimental settings
Given the complexity of in vivo model systems and the influence of the host on gut microbiome dynamics, it was important to examine L. monocytogenes- associated changes in both SPF and Oligo-MM^12^ mice. We aimed to identify microbial taxa potentially associated with the enhanced early colonisation resistance observed in faecal shedding (Fig. 2B), which appeared to diminish by day 3 post-infection. To investigate this, we compared the relative abundances of the twelve most abundant genera within each mouse model across the infection time course in the Listeria-infected groups. This approach enabled the exploration of taxonomic shifts potentially linked to microbiota-mediated protection or susceptibility during the progression of infection. Given the extensive number of changes observed, we focused our analysis on the most relevant alterations. In Oligo-MM^12^ mice, Akkermansia significantly increased at day 3 (p = 1.1e-05) (Fig. 6A). In contrast, in SPF mice, Akkermansia levels decreased after the first day of Listeria infection (Fig. 6B). Both Bacteroides and Blautia decreased over time in both mouse models (p = 0.0037). Lactobacillus in SPF mice significantly increased from day 1 to day 2, followed by a marked decline at day 3 (p = 0.0043) (Fig. 6B). Notably, Prevotellamassilia was found to be significantly abundant only in SPF mice and increased consistently over the infection period (p = 0.0022).Fig. 6. Temporal changes in gut microbiota composition during L. monocytogenes infection in SPF and Oligo-MM^12^ mice. A/B: Relative abundances of the twelve most abundant genera were analysed in faecal samples collected on days 1, 2, and 3 post-infection, indicated as L1, L2 and L3. Data are presented separately for SPF and Oligo-MM^12^ mice across the time course. * for P < 0.05, ** for P < 0.01, *** for P < 0.001, and **** for P < 0.0001. The Wilcoxon rank-sum test was used to assess microbial abundances.
A closer examination of genus-level taxonomic profiles in the ex vivo micro-Matrix fermentation study revealed no significant changes over time or between treatment groups (Supplementary Figure S4). This stability suggested that, in the absence of host factors, L. monocytogenes alone was insufficient to drive substantial restructuring of the gut microbiome, highlighting the critical role of host-microbe interactions in modulating community dynamics during infection.
Discussion
This study evaluated the Oligo-MM^12^ defined microbiota mouse model as a model system for the analysis of L. monocytogenes pathogenesis. As Oligo-MM^12^ mice harbour a reduced complexity microbiota from birth, they demonstrate normal development of the gut epithelium and associated immune system, and thereby overcome the limitations of the germ-free mouse model^4,28^, but still exhibiting some immune differences compared to SPF mice^29^. Previous work used Oligo-MM^12^ mice to examine the impact of supplementation with specific species upon L. monocytogenes infection but did not carry out a full comparison with SPF mice over time, or examine the microbiota or other aspects of infection, such as the host response^30^. In our recent publication, we demonstrated that Oligo-MM^12^ mice represent a valuable model for studying enteropathogenic infections^8^. Here, we extend these findings by directly comparing Oligo-MM^12^ and SPF mice over the course of L. monocytogenes infection, with a particular focus on microbiota composition. In parallel, we investigated bacterial community dynamics in the absence of the host using a micro-Matrix fermentation system.
Our previous results showed that L. monocytogenes infection in both mouse models (Oligo-MM^12^ and SPF) followed a similar course in the internal organs, suggesting that systemic growth of L. monocytogenes is not strongly influenced by microbiota complexity^8^. In the current study analysis of L. monocytogenes numbers in the faeces indicated interesting differences in infection dynamics between Oligo-MM^12^ and SPF mice. On day 1 following infection, Oligo-MM^12^ mice appear to control L. monocytogenes proliferation better than SPF mice. We speculate that the Oligo-MM^12^ microbiota may provide an early impediment to infection, potentially due to the high overall abundance of taxa that have been shown previously to impede L. monocytogenes infection in more complex models. For instance, supplementation with Akkermansia muciniphila^31^, Clostridium spp.^9^ or Lactobacillus spp^32–34^ has been shown to impede L. monocytogenes infection in various murine models. In Oligo-MM^12^ mice these are present in elevated absolute numbers relative to SPF mice in which they form only a limited subpopulation of a more complex microbiota.
In SPF mice, faecal levels of L. monocytogenes were controlled over the three-day time course of the infection. In contrast, Oligo-MM^12^ mice were more permissive to Listeria proliferation with an increase in numbers in the faeces over time. L. monocytogenes proliferates in internal organs, including the liver and gallbladder, from which it is regularly released into the GI tract. Faecal densities of L. monocytogenes therefore represent both proliferation in the gut and seeding from growth in internal organs^35^. Since the numbers of L. monocytogenes in internal organs were similar in both models, we conclude that the Oligo-MM^12^ microbiota was permissive to the growth of L. monocytogenes over time. An additional and critical aspect of L. monocytogenes pathogenesis that may be influenced by microbiota complexity is invasion of the intestinal epithelium^36,37^. Microbial communities can modulate epithelial susceptibility to invasion through effects on mucin production and goblet cell function, expression of host surface receptors, and maintenance of tight junction integrity^38^. These parameters were not assessed in the present study, and therefore, potential differences in epithelial invasion between Oligo-MM^12^ and SPF mice cannot be excluded. Future studies incorporating analyses of epithelial barrier function and host–pathogen interactions at the mucosal surface will be important to determine how microbiota composition influences the early stages of L. monocytogenes infection. Our recent work showed that metabolomic analysis of faeces from infected SPF or Oligo-MM^12^ mice provides evidence that there were indeed functional differences between these two bacterial communities^8^. It is likely that the more diverse SPF microbiota provides increased competition for L. monocytogenes or produces hitherto undetected metabolites that impede the growth of the pathogen^9,39^. Overall, Oligo-MM^12^ may provide a model which allows a stepwise increase in microbiome complexity in order to understand the inhibiting conditions for L. monocytogenes growth in the mouse gut, as was utilised for Salmonella enterica serovar Typhimurium in the Oligo-MM^12^ model system^40^.
During the progression of L. monocytogenes infection, we only detected minor changes in the Oligo-MM^12^ microbiota composition. A. muciniphila relative abundance was reduced on day 2 after L. monocytogenes infection and increased on day 3 post-infection in Oligo-MM^12^ mice. In SPF mice, we found a significant reduction in Blautia spp. during infection. Evidence from analysis of the human microbiota suggests that Blautia populations may be negatively impacted by inflammatory responses^41^. Other studies examining L. monocytogenes infection in mice demonstrated that Blautia abundance is reduced in both young-adult and aged mice following L. monocytogenes infection^42^. The functional implications (if any) of this microbiome change for L. monocytogenes pathogenesis are currently unknown. L. monocytogenes EGD-e is known to produce a bacteriocin that actively reduces Prevotella ssp. in the gut^30^. In SPF mice at day 1 post-infection when L. monocytogenes numbers were initially relatively high in the gut, we determined a reduction in Prevotella potentially indicative of this interaction. In contrast, as L. monocytogenes numbers decreased, the abundance of Prevotella increased in the system (on day 3 of infection). It was important to note that Prevotella species were not present in the Oligo-MM^12^ microbiota, and their addition may therefore be necessary in order to mimic the infection dynamics found in normal SPF mice.
To model the luminal phase of L. monocytogenes growth and interaction with the microbiota, we used a mini-bioreactor system (micro-Matrix). To our knowledge, this has not been used previously to model Listeria gastrointestinal growth dynamics in association with the microbiota. Other systems such as Simulator of the Human Intestinal Microbial Ecosystem (SHIME)^43^, EnteroMix and TIM-2 (TNO Intestinal Model 2) have been utilised^44,45^. The micro-Matrix platform is commonly used following seeding with human faecal microbiota and herein we adapted the system for use with murine samples from either control Oligo-MM^12^ or SPF mice. L. monocytogenes densities as determined by direct plating on selective media were gradually reduced over time in the micro-Matrix system. The dynamics broadly mirror those seen in the murine models, with the Oligo-MM^12^ microbiota controlling L. monocytogenes proliferation in the early phase of growth (notably on day 1) and showing a reduced capacity to control L. monocytogenes over longer time periods (notably on day 5). L. monocytogenes competed efficiently with the microbiota from both Oligo-MM^12^, and SPF mice and the bacterial count reduced gradually over time.
The establishment of the microbiota from faecal slurries in the micro-Matrix naturally differed from the murine faecal microbiota. The reduction in abundance in A. muciniphila likely reflects the lack of a mucous compartment in the system^18^. The micro-Matrix system mimics semi-continuous feeding, with fresh nutrients introduced at regular intervals. This dynamic likely created feedback effects that initially favoured fast-growing facultative anaerobes such as Enterococcus and other Bacillota species, but over time selected for taxa better adapted to nutrient limitation or capable of utilising secondary metabolites that accumulated in the system. The establishment of Clostridioides, Clostridium and Blautia species indicates that strict anaerobes were maintained within the system. Using the micro-Matrix system, we detected minimal impact of L. monocytogenes on the composition of the Oligo-MM^12^ or SPF-derived microbiota, which suggests that host factors (such as fluctuations in bile acids and immune-inflammatory responses) are potential drivers of perturbations seen in mice. Moreover, the fermentation system did not fully replicate in vivo conditions, not only due to the absence of the host but also because of differences in microbial growth environments. This limitation is particularly relevant when studying a pathogen such as L. monocytogenes, which rapidly disseminates to secondary organs and interacts with the intestinal environment during the early stages of infection. Nevertheless, studying single bacterial interactions over shorter infection periods in the fermentation system can provide valuable insights into the dynamics between the pathogen and selected commensals, offering hypotheses about processes that may occur in vivo.
Conclusions
This study provides an integrated perspective on the complex interactions between L. monocytogenes and the gut microbiota, expanding upon the work of Herzog et al.^8^. Using SPF and Oligo-MM^12^ mouse systems alongside a micro-Matrix fermentation platform, we reveal how microbiota complexity shapes infection outcomes. Importantly, we noted an early inhibition of L. monocytogenes proliferation (during the first 24 h post-infection) by the Oligo-MM^12^ microbiota that suggests that the taxa present broadly inhibit the pathogen. This enhanced protection is lost as the pathogen replicates, with the more diverse SPF microbiota providing greater protection in the system at later points post-infection. Key microbiota changes, including Prevotella, Akkermansia and Blautia species, may reflect interactions with Listeria. Such changes were not observed in the micro-Matrix system, suggesting that a more complex environment, including mucus and a functional immune system, is necessary to model the intricacies of colonisation resistance against L. monocytogenes. Future studies could manipulate specific taxa within the Oligo-MM^12^ system to identify the species responsible for the functional differences observed between the two murine models in their ability to inhibit L. monocytogenes.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Hernández-González JC, Martínez-Tapia A, Lazcano-Hernández G, García-Pérez BE, Castrejón-Jiménez NS. Bacteriocins from Lactic Acid Bacteria. A Powerful Alternative as Antimicrobials, Probiotics, and Immunomodulators in Veterinary Medicine. Animals. 2021 Apr;11(4):979.10.3390/ani 11040979 PMC 806714433915717 · doi ↗ · pubmed ↗
- 2Fooks, L. J. & Gibson, G. R. Mixed culture fermentation studies on the effects of synbiotics on the human intestinal pathogens Campylobacter jejuni and Escherichia coli. Anaerobe 9 (5), 231–242 (2003).10.1016/S 1075-9964(03)00043-X 16887709 · doi ↗ · pubmed ↗