Neonatal enteric infection disrupts the microbiota–gut–brain axis through pattern recognition receptors and altered neuroimmune signaling

Significance

Early-life enteric infections are a major health concern that have the potential to impact gut, immune, and brain development. Here we show that such infections reprogram innate immune signaling through pattern recognition receptors on intestinal epithelial cells, disrupting the microbiota–gut–brain axis into adulthood. These findings identify early-life enteric infection as a driver of long-lasting neuroimmune imbalance. Furthermore, we demonstrate that probiotic-derived muropeptides can attenuate inflammation and restore neuroimmune communication. Our work suggests that restoring gut-brain signaling may prevent long-term complications of childhood enteric disease.

Abstract

Early-life enteric infection disrupts the developing microbiota–gut–brain (MGB) axis. Using a neonatal Enteropathogenic Escherichia coli (EPEC) model, we show that intestinal epithelial cell (IEC) nucleotide oligomerization domain (NOD)1 signaling coordinates neuroimmune outcomes throughout early development and into adulthood. Neonatal mice infected at postnatal day (P)7 exhibited ileal inflammation, with increased pro-inflammatory cytokines and monocyte/macrophage infiltration, and loss of myenteric neurons in wild-type mice but not in NOD1 conditional knockout (Nod1^ΔIEC) mice. These deficits persisted into adulthood (P56), with increased intestinal permeability, sustained inflammatory/repair signatures, neuroinflammation, altered neurogenesis, and impaired recognition memory, which were absent in Nod1^ΔIEC mice, establishing a crucial role for IEC NOD1 in long-term MGB remodeling. Last, bioactive peptidoglycan fragments (muropeptides) were isolated from probiotic Lactobacillus strains and tested as possible therapeutics. Excitingly, supplementation of muropeptides attenuated neonatal EPEC-induced mucosal inflammation and enteric neuron loss without altering bacterial burden, potentially via IEC NOD2. Together, these findings identify IEC NOD-dependent signaling linking early-life enteric infection to enduring gut-brain dysfunction and reveal probiotic-derived muropeptides as candidate therapeutics.

Early-life enteric infections are a major global health challenge, contributing significantly to childhood morbidity and mortality, exerting developmental consequences that extend well beyond acute diarrheal illness (1). Enteropathogenic Escherichia coli (EPEC), a noninvasive, attaching-and-effacing bacterial pathogen, remains a leading cause of pediatric diarrheal disease in children under five years of age worldwide (2), causing inflammation and disruptions to mucosal barrier function (3, 4). Importantly, these infections often coincide with critical periods in postnatal development, when the gut microbiome, immune system, and nervous systems are rapidly maturing, and when the host is particularly vulnerable to long-term physiological and neurodevelopmental perturbations (5, 6).

Mounting evidence suggests that enteric infections can disrupt the developing microbiota–gut–brain (MGB) axis, a bidirectional communication network linking the gastrointestinal (GI) tract to the central nervous system (CNS) via neural, immune, and endocrine pathways (7, 8). Perturbations to any of these routes during early life can have persistent effects on intestinal barrier integrity, immune homeostasis, and neurobehavioral trajectories. We previously reported that neonatal EPEC infection disrupts colonization of the gut microbiome and promotes systemic and neuroinflammation (8). While the acute effects of EPEC infection, such as dehydration, malnutrition, and stunted growth, are well characterized (9), the mechanisms linking neonatal enteric infection to persistent gut-brain dysfunction remain poorly defined. Acute GI inflammation and barrier disruption caused by neonatal enteric infection (8, 10–13), may interfere with the maturation of the enteric nervous system (ENS) (14) and neurodevelopment within the CNS, which may reprogram host physiology throughout life.

Innate immune signaling is a key mediator of host–microbe interactions during early life, particularly at mucosal surfaces, where pattern recognition receptors (PRRs) shape these responses (15). Highly expressed in intestinal epithelial cells (IECs), nucleotide-binding oligomerization domain (NOD)1, a member of the Nod-like receptor (NLR) family, plays a critical role by sensing microbial-derived meso-diaminopimelic acid (meso-DAP)–containing motifs in Gram-negative bacterial peptidoglycan (PGN), including EPEC (16, 17). PGN remodeling during bacterial growth and turnover releases small bioactive PGN fragments known as muropeptides. Consistent with this, intestinal NOD1 activity shapes neuroimmune responses (18–21), and stress-associated behavioral regulation (18), suggesting a potential role in MGB axis signaling. Beyond mucosal immunity, PGN sensing pathways have also been implicated in neural development and behavior. Mice lacking PGN recognition protein 2 (PGLYRP2) exhibited sex-specific differences in social behavior and brain-derived neurotrophic factor (Bdnf) expression, a critical regulator of brain function (21). We previously showed that Nod1/Nod2 double knockout (NodDKO) mice exhibit stress-induced behavioral impairments following exposure to acute psychological stress in adulthood (18). This effect was specifically localized to IEC NOD1 expression, identifying the gut epithelium as a potential interface between microbial recognition and neurobehavioral outcomes. Here, we investigated whether IEC NOD1 governs long-term MGB axis disruption following neonatal EPEC infection.

Using IEC-conditional NOD1 knockout mice (Nod1^ΔIEC), we report that neonatal EPEC infection induces acute GI inflammation and loss of myenteric neurons coupled with sustained barrier dysfunction, hippocampal neuroinflammation, impaired neurogenesis, and cognitive deficits in an IEC NOD1-dependent manner. Finally, we identified probiotic-derived muropeptides as a microbial intervention to mitigate infection-induced gut-brain pathology.

Results

Neonatal EPEC Infection Disrupts Intestinal Barrier Structure and Function at P14.

Neonatal wild-type (WT) and Nod1^ΔIEC (Vil.Cre⁺Nod1^f/f) mice were infected with EPEC at postnatal day (P)7 (Fig. 1A) and colonization monitored starting at 7 d postinfection (DPI). Both WT and Nod1^ΔIEC mice were effectively colonized by 7 DPI (P14), with fecal bacterial burden similar in both strains at 7 and 14 DPI. In contrast, EPEC clearance was delayed in Nod1^ΔIEC mice, with mice exhibiting significantly higher bacterial loads compared to WT mice at P28 (Fig. 1B). Despite this delayed clearance in Nod1^ΔIEC mice, no overt histological damage was observed by hematoxylin and eosin (H&E) staining. A reduced villus-crypt axis length was found in WT, but not in Nod1^ΔIEC EPEC-infected mice, vs. sham-infected controls, consistent with impaired epithelial repair and regeneration (Fig. 1C). Confocal microscopy was used to assess IEC proliferation by Ki67 staining, and a reduction in Ki67⁺CDH1⁺DAPI⁺ cells per crypt was found in WT EPEC-infected mice compared to uninfected mice, but not in Nod1^ΔIEC EPEC-infected mice (Fig. 1C).

Fig. 1.

IEC NOD1 is required for efficient clearance of neonatal EPEC infection and drives altered IEC kinetics and ENS deficits. Study design of neonatal EPEC infection in WT and NOD1 knockout (Nod1^ΔIEC) mice (A). Bacterial colonization (CFU/g feces) was assessed at 7-, 14-, and 21-d postinfection (B). The crypt–villus axis measured by H&E and immunofluorescence performed using Ki67 (green), CDH1 (red), and DAPI (blue) to quantify proliferative cells at P14 (C). Epithelial barrier and repair markers (Muc2 and Slc26a3) were measured by qPCR (D). Goblet cells were quantified using PAS staining (E). Neuronal activation and signaling pathways (Bdnf, and S100b) were measured by qPCR at P14 (F). Whole-mount ileal preparations (P14) were stained by ANNA-1 and myenteric neurons counted (G). N = 8 to 20 (colonization), N = 4 to 6 (histology/IF), N = 6 to 8 (qPCR); mean ± SEM; *P< 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 by two-way ANOVA with Bonferroni correction. (Scale bar, 50 μm.)

Expression of genes associated with absorptive or secretory IEC lineages was measured by qPCR. Increased expression of ileal mucin 2 (Muc2; goblet cell marker) and Slc26a3 [down-regulated in adenoma (DRA); Cl⁻/HCO³ exchanger] was observed at P14 in WT EPEC-infected, but not in Nod1^ΔIEC mice (Fig. 1D). Given the increased Muc2 expression, goblet cells were stained using periodic acid–Schiff (PAS), showing increased numbers of PAS⁺ cells in WT, but not Nod1^ΔIEC EPEC-infected mice (Fig. 1E). These results demonstrate that NOD1 signaling reduces IEC proliferation during enteric bacterial infection with EPEC, including an increase in mucus-producing goblet cells.

Neonatal Enteric Bacterial Infection Impairs the ENS.

To assess whether neonatal EPEC infection impacts enteric neurons, expression of neuronal markers was assessed. Bdnf, a key neurotrophic factor in the GI tract (22), was significantly increased in WT, but not in Nod1^ΔIEC EPEC-infected mice (Fig. 1F). S100b, a gene associated with glial activation and neuroimmune signaling, was also increased only in WT mice (Fig. 1F), together implicating potential ENS involvement in enteric neuroimmune signaling. Whole-mount immunostaining and quantification of neurons in the myenteric plexus identified a significant reduction in ANNA1-positive myenteric neuron counts in the ileum following infection in WT EPEC-infected mice vs. WT sham-infected controls, which was absent in Nod1^ΔIEC EPEC-infected mice (Fig. 1G).

EPEC-Induced Intestinal Inflammation Is IEC NOD1-Dependent.

Given the significant impact of neonatal infection on IEC physiology in the distal ileum, early inflammatory responses were assessed. Neonatal EPEC infection elicited a robust inflammatory response in WT EPEC-infected mice, with significantly increased expression of pro/anti-inflammatory and host-protective cytokines, including interleukin-1 beta (Il1β), Il6, Il12, Il10, and Il22, but not transforming growth factor beta 1 (Tgfβ1), compared to WT sham-infected mice, which was absent in Nod1^ΔIEC EPEC-infected mice (Fig. 2A). Expression of additional PRRs was assessed to better understand whether IEC NOD1 signaling was inducing a proinflammatory state in response to bacterial ligands. Increased expression of toll-like receptor 4 (Tlr4) and Nod2 was observed in WT EPEC-infected mice, but not Nod1^ΔIEC mice, indicating amplified innate immune responses in WT mice (Fig. 2A).

Fig. 2.

Early life EPEC causes inflammation and alters mucosal immune cell composition in the distal ileum at P14 via IEC NOD1. Inflammatory tone was assessed for pro-inflammatory cytokines (Il1β, Il6, Il12, Tnfα), regulatory cytokines (Il22, Tgfβ1), and PRRs (Tlr4, Nod2) (A), as well as immune cell recruitment and trafficking (Ccr2, Ccl2, Cxcr1, Cxcl1, Cxcl10, Ccl5) by qPCR (B). Representative gating strategy for flow cytometry to identify live CD45⁺ cells in the ileum and distinguish macrophages (CD64⁺ Ly6C⁻, Ly6G⁻), monocytes (CD64^+/-, Ly6C⁺), neutrophils (CD11b⁺, Ly6G⁺), and T cells (CD3⁺), with quantification presented as % frequency of live CD45⁺cells (C). N = 7 to 8 (qPCR); flow cytometry data are representative of two separate experiments (n = 7 to 10 total mice per group); mean ± SEM; *P < 0.05; **P < 0.01; ***P < 0.001 by two-way ANOVA.

Given the robust inflammatory responses following neonatal EPEC infection, we assessed if chemokine expression and immune cell recruitment occurred in the ileum and whether these are dependent on NOD1 signaling in IEC. Infection in WT mice resulted in significantly increased expression of C-C motif chemokine receptor (Ccr2) and its cognate ligand C-C motif chemokine ligand (Ccl2) and C-X-C motif chemokine receptor (Cxcr1) (P = 0.06) along with its cognate ligand C-X-C motif chemokine ligand (Cxcl1), which were dependent on NOD1 signaling in IEC (Fig. 2B). Infection did not broadly increase the expression of chemokines and receptors, as Ccl5, and Cxcl10, were not impacted in WT EPEC-infected mice (Fig. 2B).

Based on changes in chemokine expression in the ileum, we characterized immune cell recruitment following neonatal EPEC infection using flow cytometry. Neonatal EPEC infection significantly increased the frequency of monocytes (CD64^+/−Ly6C⁺) and macrophages (CD64⁺Ly6C^–) in WT ileum, which was absent in Nod1^ΔIEC EPEC-infected mice (Fig. 2C). No significant differences were observed in T-cells or neutrophils (Fig. 2C). These findings demonstrate that EPEC infection drives the recruitment of specific immune cell populations in the neonatal ileum in a NOD1-dependent manner.

IEC NOD1 Mediates Persistent Disrupted Intestinal Barrier Function and GI Inflammation in Adulthood Following Neonatal Infection.

The long-term consequences of neonatal EPEC infection were determined in adult mice, with ileal physiology assessed in Ussing chamber studies. While ion secretion [short-circuit current (Isc)] was not significantly impacted by neonatal EPEC infection, ileal permeability [conductance (G) and FITC flux] was significantly increased in WT, but not Nod1^ΔIEC mice (Fig. 3A). Despite the lack of effect on ion secretion, increased expression of the chloride-sulfate anion transporter Slc26a3 was observed in adult WT mice previously infected with EPEC (Fig. 3B). Expression of Muc2 and total number of PAS⁺ goblet cells was not different in adult mice neonatally infected with EPEC, WT, or Nod1^ΔIEC mice, suggesting the presence of an intact mucus barrier (SI Appendix, Fig. S1A). Similarly, the villus–crypt axis and the total number of Ki67⁺ IEC were not significantly impacted by neonatal EPEC infection in either WT or Nod1^ΔIEC adult mice (SI Appendix, Fig. S1B), suggesting that EPEC-induced deficits in IEC kinetics were transient, and limited to the period of active infection.

Fig. 3.

Neonatal EPEC infection leads to persistent disrupted intestinal barrier function and inflammation in adult mice, which is absent in Nod1^ΔIEC mice. Ussing chamber studies assessed short-circuit current (Isc), transepithelial conductance (G), and FITC-dextran flux in the ileum (A). Epithelial barrier markers (Slc26a3) were assessed by qPCR (B). Ileal cytokines (Il6, Il12, Ifnγ, Il22, Il1β, and Tnfα), Nod2, S100b, and chemokines (Ccl2, Cxcr1, Cxcl1, and Ccr2) were measured by qPCR (C). Immunofluorescence was performed and Iba1⁺ cells quantified in the ileum (D). N = 8 to 12; mean ± SEM; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 by two-way ANOVA.

As intestinal barrier function is a critical determinant of health, we assessed whether persistent barrier defects in adult mice infected as neonates lead to ongoing inflammation in the ileum. Adult WT, but not Nod1^ΔIEC, mice previously infected as neonates had increased ileal expression of Il6, Il12, interferon-gamma (Ifnγ), and Il22, but not Il1β or tumor necrosis factor-alpha (Tnfα) (Fig. 3C). Persistently increased ileal expression of Nod2 was also found in WT vs. Nod1^ΔIEC EPEC-infected mice, but no changes were found in expression of the glial marker S100b (Fig. 3C). With respect to chemokines, persistently increased Ccl2 and Cxcr1 was found in adult WT, but not Nod1^ΔIEC, mice previously infected as neonates, with no changes in Cxcl1, or Ccr2 expression (Fig. 3C). Evidence of ileal inflammation was further supported by immunostaining, with an increase in the number of Iba1⁺ cells (macrophage) in the mucosa of WT EPEC-infected mice, but not in Nod1^ΔIEC mice (Fig. 3D). Finally, myenteric neurons were quantified in the adult ileum and in contrast to P14 mice, no changes in total neuron numbers were seen in either WT or Nod1^ΔIEC mice neonatally infected with EPEC, suggesting that enteric neurons recover postinfection (SI Appendix, Fig. S1C).

Neonatal Infection Causes Cognitive Deficits, Neuroinflammation, and Altered Hippocampal Neurogenesis in Adulthood.

To assess the role of IEC NOD1 on the long-term neurobehavioral impact of neonatal EPEC infection, we performed behavioral testing in adult mice. Recognition memory, via the novel object recognition (NOR) task, was impaired in neonatally WT EPEC-infected mice, but not Nod1^ΔIEC mice, as determined by a significant reduction in exploration ratio compared to sham-infected mice (Fig. 4A). In contrast, there were no significant differences in anxiety-like behavior as measured by time spent in the light compartment of the light/dark (L/D) box, nor impairments in locomotor activity as assessed by total distance traveled in the open field test (OFT) across any group (Fig. 4A).

Fig. 4.

Early-life enteric infection leads to behavioral deficits coupled with neuroinflammation and decreased neurogenesis in adulthood only in WT mice. Behavior was assessed in adult mice using the NOR task, L/D box, and (OFT) (A). Expression of inflammatory and glial activation markers in the hippocampus of adult mice was performed, including Iba1, Il1β, Il6, Tnfα, Il10, Tlr2, Tlr4, Nos2, and S100b by qPCR (B). Hippocampal neurogenesis was assessed by qPCR (Bdnf) and immunofluorescence [Ki67⁺ and doublecortin (DCX⁺)] in the dentate gyrus (DG) at P56 (C). Representative immunofluorescence images show Ki67 (red), DCX (green), and DAPI (blue) staining. (Scale bar, 100 μm.) Behavior n = 15 to 28, qPCR n = 5 to 8, IF n = 5 to 7; mean ± SEM; *P < 0.05; **P < 0.01; ***P < 0.001 by two-way ANOVA.

Our previous studies found that neuroinflammation developed in adulthood following neonatal EPEC infection, suggesting a delayed CNS response to enteric infection (8). Here, we confirmed the presence of neuroinflammation in adult WT mice neonatally infected with EPEC, with increased hippocampal expression of the microglia marker Iba1, which was absent in Nod1^ΔIEC EPEC-infected mice. Hippocampal expression of Il1β, Il6, Il10, and Tnfα was also significantly upregulated in adult WT mice previously infected with EPEC as neonates in contrast to Nod1^ΔIEC mice, with the exception of Il6, which was elevated in Nod1^ΔIEC mice (Fig. 4B). With respect to PRR expression, Tlr2 remained unchanged whereas Tlr4 expression was significantly increased only in WT EPEC-infected mice compared to sham-infected mice (Fig. 4B). Last, S100b was also increased in WT mice previously infected with EPEC compared to sham-infected mice, suggesting the activation of glial cells (Fig. 4B).

Given the evidence of cognitive deficits and neuroinflammation in adult WT mice neonatally infected with EPEC, hippocampal neurogenesis was assessed. Similar to our prior studies (8), significant increases in hippocampal Bdnf expression were found in EPEC-infected adult WT mice, which were absent in Nod1^ΔIEC EPEC-infected mice (Fig. 4C). Immunofluorescence was used to quantify proliferating (Ki67⁺) and immature neurons [doublecortin (DCX) ⁺] in the hippocampal DG region in adult mice. WT, but not Nod1^ΔIEC, EPEC-infected mice showed a significant increase in Ki67⁺ and DCX⁺ cells in WT EPEC-infected mice relative to sham-infected mice (Fig. 4C).

Early-Life EPEC Infection Results in Sustained Systemic Hypothalamic–Pituitary–Adrenal (HPA) Axis Activation and Altered Serotonergic Signaling in WT mice.

Since serotonin (5-HT) is critical for maintenance of gut-brain communication, we assessed serotonergic signaling in the neonatal ileum and in adult ileum and brain following infection. In neonatal mice, tryptophan hydroxylase (Tph)2, and serotonin reuptake transporter (Sert) expression in the ileum was significantly elevated in WT, but not Nod1^ΔIEC, EPEC-infected mice. In contrast, Tph1 showed no significant changes in neonatal mice, suggesting that 5-HT production by enterochromaffin cells was not impacted by infection, rather it is likely from enteric nerves (Fig. 5A). Increased expression of the 5-HT receptors 5Htr2c and 5Htr4 were also observed only in WT EPEC-infected mice.

Fig. 5.

Neonatal EPEC infection impacted serotonin signaling in both the gut and brain via IEC NOD1. Expression of serotonin (5-HT)-related genes (Tph1, Tph2, Sert, and 5Htr2c) at both P14 (A) and adult (B) ileum and in the adult hippocampus by qPCR (C). N = 5 to 8; mean ± SEM; *P < 0.05; **P < 0.01; ***P < 0.001 by two-way ANOVA.

In adulthood, WT mice neonatally infected with EPEC had increased ileal expression of Tph2, 5Htr2c, and 5Htr4 compared to sham-infected mice (Fig. 5B), suggesting prolonged serotonergic dysregulation in the GI tract, mediated by enteric nerves. In the hippocampus, adult WT EPEC-infected mice also demonstrated significant upregulation of Tph2, Sert, and 5Htr2c, but not 5Htr4, compared to sham-infected controls (Fig. 5C). Impairments in 5-HT signaling were absent in the gut and brain in Nod1^ΔIEC EPEC-infected mice, suggesting that neonatal EPEC infection persistently disrupts 5-HT signaling in both the gut and brain via IEC NOD1.

Given the interaction between 5-HT and HPA axis activation, stress responses were assessed. In the neonatal ileum, EPEC infection markedly increased expression of the stress receptor gene corticotropin releasing factor (Crfr)2 in WT EPEC-infected mice, an effect that persisted into adulthood, but was absent in Nod1^ΔIEC mice (Fig. 6A). In contrast, glucocorticoid receptor (Gr) expression was unchanged in both neonatal and adult ileum. In the adult hippocampus, elevated Crfr2, nitric oxide synthase (Nos)2, and Gr expression was observed only in WT EPEC-infected mice (Fig. 6B). This increased GR expression in the brain may be associated with disrupted HPA axis feedback inhibition. Increased stress-related gene expression was accompanied by increased serum corticosterone levels in WT mice neonatally infected with EPEC, but only in adulthood (Fig. 6C).

Fig. 6.

Early-life EPEC infection results in sustained systemic HPA axis activation. HPA axis signaling markers (Crfr2, Nos2, Gr) were measured in the neonatal (P14) and adult ileum (A), and adult hippocampus (B) by qPCR. Serum corticosterone was assessed in both neonatal (P14) and adult (P56) mice (C). qPCR n = 5 to 6, ELISA n = 8 to 16; mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001 by two-way ANOVA.

Probiotic Muropeptide Supplementation Mitigates EPEC-Induced Neonatal Ileal Pathophysiology and Inflammation.

Muropeptides isolated from certain Lactobacillus strains, particularly M-Tri-Lys (MurNAc-Ala-D-isoGln-Lys), have been shown to ameliorate inflammation caused by chemically induced colitis (23). Muropeptides were therefore isolated from probiotic Lactobacillus strains (Lactobacillus rhamnosus R0011 and L. helveticus R0052) and analyzed by liquid chromatography-mass spectrometry (LC/MS), revealing a complex mixture of monomeric, dimeric, and trimeric muropeptides, including M-Tri-Lys (Fig. 7A). Relative quantification, based on total ion counts, indicated dominance of crosslinked structures, with Tetra-N-Tetra-N trimer being most abundant (32.94% of total ion signal), followed by Tetra-N-Tetra-N dimer (21.56%) and Tri-N-Tetra-N trimer (16.94%). The structural analysis identified the presence of N-acetyl-glucosamine (GlcNAc)–N-acetyl-muramic acid (MurNAc) disaccharide units linked to canonical stem peptides (L-Ala–D-Gln–L-Lys–D-Ala), bridged by peptide crosslinks, characteristic of Gram-positive PGN.

Fig. 7.

Probiotic muropeptide supplementation mitigates EPEC-induced neonatal epithelial cell deficits. Peptidoglycan from probiotic strains (L. helveticus/L. rhamnosus) was isolated and characterized by LC/MS (A). Murine splenocytes were stimulated with bacterial ligands [LPS, M-Tri-Lys, iE-DAP, and Lactobacillus muropeptide (MP)] and Tnfα and Il10 measured by qPCR (B). Probiotic muropeptides (or vehicle) were administered neonatally in vivo (C) and bacterial burden (log10 CFU/g feces) was quantified (D). Ussing chambers were used to assess short-circuit current (Isc) and conductance (G) in the ileum (E). Goblet cells were assessed by expression of Muc2 (qPCR) and PAS staining (F). Ileal histology was used to measure the villus–crypt axis and immunofluorescence for quantifying Ki67⁺ cells [G; Ki67 (green), CDH1 (red), DAPI (blue)]. Splenocytes n = 2 to 3, colonization n = 8 to 9, Ussings n = 5 to 12, IF n = 6 to 10; mean ± SEM; *P < 0.05; **P < 0.01; ***P < 0.001 by one-way ANOVA.

Probiotic muropeptide immunogenicity was assessed using isolated murine splenocytes and measuring cytokine expression (Tnfα and Il10). As expected, Tnfα expression was significantly increased by lipopolysaccharide (LPS) or γ-D-glutamyl-meso-DAP (iE-DAP) stimulation but not following M-Tri-Lys and muropeptide stimulation (Fig. 7B). Interestingly, probiotic muropeptide markedly upregulated Il10 response from splenocytes in contrast to all other groups (Fig. 7B). These data indicate that the probiotic-derived muropeptide fraction is potentially capable of eliciting an anti-inflammatory response.

To assess whether probiotic muropeptides could ameliorate EPEC-induced GI inflammation, mice were orally supplemented daily for 7 d starting at P7 (Fig. 7C). Supplementation with muropeptides (+MP; 10ug/d) did not impact EPEC colonization as determined by fecal bacterial burden at P14, showing no significant differences between EPEC+LB (Luria-Bertani broth; vehicle) and EPEC+MP treated mice (Fig. 7D). Ussing chamber studies revealed that muropeptide supplementation prevented EPEC-induced increased secretory state (Isc) and tight junction permeability (G; Fig. 7E). Ileal Muc2 expression was increased in both EPEC+LB and EPEC+MP treated mice compared to vehicle controls, which was coupled with a significant increase in total PAS-positive cells, findings consistent with mucosal remodeling and goblet cell hyperplasia during infection (Fig. 7F). Last, histological analysis showed that EPEC-induced decreases in the crypt/villus axis were prevented by muropeptide supplementation, which were coupled to restoration in numbers of Ki67⁺ IEC (Fig. 7G). These findings suggest that muropeptides preserve mucus barrier integrity and reinforce epithelial defenses in early life.

The immunomodulatory potential of muropeptides during early-life EPEC infection was evaluated at P14. As expected, EPEC+LB supplementation increased Il1β, Il6, Il10, Il12, Il22, and Tnfα compared to both vehicle control mice, which was prevented in EPEC+MP treated mice (Fig. 8A). No differences were observed between vehicle control and EPEC+MP mice, indicating that the supplementation did not broadly enhance anti-inflammatory responses, but rather rebalanced the inflammatory environment. The increased expression of the glial marker S100b and elevated expression of PRR receptors seen in EPEC+LB mice (Tlr2, Tlr4, and Nod2) was also prevented in EPEC+MP-treated mice (Fig. 8B). Chemokine profiles revealed similar trends, with muropeptide supplementation preventing EPEC-induced expression of Ccr2, Ccl2, Cxcr1, and Cxcl1 compared to EPEC+LB treated mice (Fig. 8C). These findings suggest that muropeptides may limit chemokine-driven recruitment of monocytes and neutrophils to the infected mucosa, preventing inflammation. In contrast, increased expression of Slc26a3 in EPEC-infected mice was not improved by MP treatment, but EPEC-induced elevated expression of Crfr2 was inhibited in the EPEC+MP group (Fig. 8D). The normalization of Crfr2 expression with muropeptide supplementation suggests that these bioactive molecules may modulate GI-derived neuroendocrine signaling. Muropeptide supplementation could potentially mitigate the downstream brain and behavioral alterations we observed in EPEC-infected mice.

Fig. 8.

Probiotic muropeptide supplementation blocked immune responses caused by neonatal EPEC infection. Pro-inflammatory cytokines (Il22, Il6, Il12, Tnfα, Il1β), anti-inflammatory/regulatory cytokines (Il10), the glial cell marker S100b (A), PRRs (Tlr2, Tlr4, Nod2) (B), chemokines (Ccl2, Cxcl1, Ccr2, Cxcr1) (C), the epithelial transporter (Slc26a3), and stress receptor (Crfr2) (D) were quantified in the ileum at P14 by qPCR. Whole-mount ileal preparations were stained by ANNA-1 (green) and myenteric neurons counted (E). Ileal organoids were infected with EPEC ± pretreatment with muropeptide (MP)/NOD2 inhibitor (Nod2i) and qPCR for cytokines performed (F). N = 5 to 8; mean ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001 by one-way ANOVA.

Given the EPEC-induced neuronal loss seen in the myenteric plexus at P14, quantification of enteric neurons was performed. Excitingly, supplementation with muropeptide restored neuron counts in EPEC+MP treated mice to similar numbers seen in Sham+LB treated mice (Fig. 8E), highlighting the ability of muropeptides to protect myenteric neurons following bacterial infection.

Last, to identify a mechanism of action of muropeptides on IEC, ileal organoids were generated. EPEC infection of ileal organoid monolayers caused a robust increase in cytokines (Il1β, Il6, and Tnfα) and chemokine (Cxcl1) production, which could be prevented by 2 h pretreatment with MP, except Tnfα (Fig. 8F). Excitingly, the beneficial impact of MP could be abolished by 1 h pretreatment with a Nod2 inhibitor (Nod2i; Fig. 8F), suggesting that epithelial NOD2 signaling is required for the anti-inflammatory activity of probiotic-derived MP during EPEC infection.

Discussion

Neonatal enteric infections remain a major global health concern, due in part to their potential to impact long-term alterations in GI, immune, and neurodevelopmental outcomes (5, 24). EPEC infection is associated with acute and persistent diarrhea, particularly during the first 2 y of life in children in developing countries, leading to nutritional deficiencies and death (25). In the United States, atypical EPEC isolates, lacking the bundle forming pili, remain a persistent problem, with infection in both children and adults causing significant disease including diarrhea, vomiting, and abdominal pain for multiple days (26). We previously showed that early-life EPEC infection in mice causes acute neonatal GI inflammation leading to disruption of the MGB axis in adulthood (8). Here, we extend these findings by demonstrating that NOD1 signaling in IEC not only regulates local immune responses but also shapes long-term neuroimmune outcomes following neonatal enteric infection, and that supplementation with probiotic-derived muropeptides can mitigate infection-induced MGB axis deficits.

Innate immune responses are essential for clearance of enteric bacterial pathogens (27). EPEC robustly and reproducibly infects neonatal mice, with pathogen clearance typically occurring between 14 and 21 DPI in WT mice (8, 11). In contrast, the delayed EPEC clearance observed in Nod1^ΔIEC mice reflects the importance of epithelial innate immune sensing, with disruption of these pathways compromising early neutrophil responses and recovery. Previous studies showed that NOD1-deficient mice infected with Citrobacter rodentium displayed impaired induction of IL22, leading to delayed pathogen clearance (28). Similarly, NOD1-deficient mice infected with Clostridium difficile also exhibited increased bacterial translocation, reduced CXCL1 production, impaired neutrophil trafficking, and reduced bacterial clearance, ultimately leading to heightened mortality (29). Therefore, this suggests that delayed clearance of EPEC observed in Nod1^ΔIEC mice likely results from impaired chemokine signaling involved in myeloid cell recruitment, as evidenced by reduced chemokine expression and reduced monocyte and macrophage recruitment in the ileum. Together, these findings support a model in which IEC NOD1 functions as a key mediator initiating antimicrobial defense programs, with loss of IEC NOD1 compromising these pathways, resulting in impaired immune cell trafficking and delayed pathogen clearance.

Mucosal damage in the ileum was assessed by histological analysis. As seen in our prior studies (8), significant villus blunting and reduced villus–crypt axis length was observed in neonatally EPEC-infected WT, but not Nod1^ΔIEC mice. These macroscopic changes likely reflect ongoing epithelial stress and impaired regenerative capacity, as suggested by significant reductions in Ki67⁺ IEC in these mice. In addition, goblet cells were impacted in WT, but not Nod1^ΔIEC mice, EPEC-infected mice suggesting that mucosal defenses are mediated by IEC NOD1. Histological changes are consistent with studies in adult EPEC-infected mice pretreated with streptomycin, where infection induced microvillous effacement, neutrophil infiltration, and goblet cell expansion (13). Neonatal EPEC infection also induced robust innate and adaptive mucosal immune responses in the ileum of WT mice with increased expression of pro-inflammatory cytokines and epithelial regulatory genes at P14, which are consistent with our prior study (8). The concurrent upregulation of PRR expression supports a model in which select microbial product can activate innate immune pathways in IEC, amplifying local immune activation. The absence of this immune response in Nod1^ΔIEC mice strongly implicates IEC NOD1 as a necessary upstream regulator of EPEC-induced cytokine and PRR gene expression.

The ENS continues to develop and mature during early life, making it susceptible to the effects of neonatal enteric infection (30). Here, we found significant neuronal loss in the myenteric plexus of the ileum following neonatal EPEC infection (P14). This finding is consistent with enteric pathogens inducing inflammation-driven neurotoxicity in the gut, where cytokines promote oxidative stress, glial activation, and loss of myenteric neurons, particularly in the developing ENS (31). In Salmonella infection, neuronal loss was biased toward excitatory neurons and persisted for months postinfection, resulting in sustained GI dysmotility (31). Neuronal loss, coupled with glial activation, during neonatal infection may reprogram neural circuits and impair immune–neural crosstalk. Together, these data identify the ENS as a critical and vulnerable target of early-life enteric infection and highlight the need for therapeutic strategies to protect or restore enteric neurons during neonatal inflammation.

Following clearance of the pathogen, GI inflammation persisted into adulthood and was coupled with GI pathophysiology (8). Increased ileal permeability in adult WT, but not Nod1^ΔIEC mice, aligned with evidence of ongoing elevated inflammatory gene expression and barrier repair markers in mice neonatally infected with EPEC. Activation of NOD1 may exacerbate infection-induced damage by amplifying NF-κB–mediated inflammatory signaling and stress responses in IEC, leading to persistent epithelial dysfunction. These findings suggest that during infection, IEC NOD1 activation can cause chronic mucosal barrier leakiness and impaired tight junction dynamics, which are known contributors to ongoing mucosal inflammation and systemic immune activation (32). These results suggest a lasting impact of neonatal enteric infection on GI physiology, reinforcing NOD1 as a candidate target for interventions aimed at preserving epithelial function following neonatal infectious insult.

Behavioral studies previously revealed that neonatal EPEC infection causes long-lasting cognitive dysfunction, including recognition memory deficits in adulthood (8). Here, cognitive deficits were confirmed in WT EPEC-infected mice, but not in Nod1^ΔIEC mice, strongly implicating IEC NOD1 in causing the impaired cognitive sequelae following early-life infection. As shown previously (8), evidence of increased microglia activation was coupled with increased cytokines and PRRs in WT mice neonatally infected with EPEC. No evidence of neuroinflammation was present in Nod1^ΔIEC mice, implicating epithelial NOD1 in modulating downstream CNS impairments following neonatal enteric infection. These findings mimic those seen in a viral encephalitis model, where NOD1 deficiency dampened microglial activation and pro-inflammatory cytokine responses, suggesting that NOD1 can directly influence neuroinflammatory tone in the brain following infection (33). Prior studies also demonstrate that microglial NOD1 can amplify neuroinflammation when appropriately stimulated (33, 34), potentially allowing for IEC NOD1-driven signals to converge on central NOD1-dependent pathways. Behavioral deficits and neuroinflammation were also coupled with evidence of increased hippocampal neurogenesis in WT, but not Nod1^ΔIEC mice, neonatally infected with EPEC. Together, these findings support a model in which neonatal enteric infection causes hippocampal neuroinflammation and enhanced neurogenesis, disrupting hippocampal plasticity and detrimentally impacting cognitive function in a NOD1-dependent manner. Collectively, these findings underscore the potential for early microbial insults to exert long-lasting effects on neuroimmune homeostasis and highlight IEC NOD1 as a mechanistic bridge between GI infection and neuroinflammation.

NLRs regulate 5-HT signaling in both the gut and the brain in response to stress (18), likely via both ENS and CNS neurons. Here we found persistent alterations in 5-HT signaling in the ileum, including increased neuronal 5-HT synthesis and 5-HT receptor expression at both P14 and in adult WT EPEC-infected mice, but not Nod1^ΔIEC mice. Similar impairments in 5-HT signaling were found in the adult WT hippocampus, together suggesting that serotonergic dysregulation following neonatal enteric infection occurs both locally in the gut and centrally within the CNS, and may be caused by chronic inflammation (35, 36). In addition, 5-HT is increasingly recognized as a modulator of immune tone, glial reactivity, and neurogenesis (36, 37). Altered HPA-axis activation was identified in WT neonatal and adult mice, which was coupled with increased serum corticosterone levels in adulthood. Importantly, these serotonergic and HPA-axis alterations were absent in Nod1^ΔIEC mice, indicating that epithelial NOD1 is required to propagate signals following neonatal infection to systemic neuroendocrine pathways. Together, these findings strongly support the idea that neonatal EPEC infection reprograms host serotonergic signaling via IEC NOD1, ultimately impairing stress axis regulation and contributing to sustained alterations in neuroendocrine and behavioral outcomes.

Probiotics can beneficially regulate the MGB axis, particularly following infection with an enteric pathogen (38), however, the mechanisms involved remain unclear (39). In contrast to NOD1, which maintains innate immunity and inflammatory programming following infection, activation of NOD2 has been found to be protective against GI inflammation (40). Using LC/MS, we identified the presence of the immunomodulatory M-Tri-Lys peptide, a NOD2 ligand previously reported to ameliorate chemically induced colitis in mice (23). Isolated probiotic Lactobacillus-derived muropeptides induced a robust Il10 response in cultured splenocytes and daily supplementation of neonatal mice prevented EPEC-induced ileal pro-inflammatory cytokine expression, while preserving anti-inflammatory (Il10) expression at P14. This is consistent with prior evidence that muropeptides reduced colitis severity by NOD2-dependent IL-10 production and T regulatory development by dendritic cells (23). Muropeptide supplementation also reduced ileal expression of PRRs and chemokine ligand/receptor expression, further suggesting reduced pro-inflammatory microbial sensing pathways and immune cell recruitment following enteric infection. Subsequent organoid studies confirm the ability of muropeptides to block EPEC-induced cytokine production by IEC and identify Nod2 as a likely primary mechanism for the beneficial impact of muropeptides.

Interestingly, probiotic muropeptide supplementation did not impact goblet cells in EPEC-infected mice. This may serve a protective role following infection (41) and importantly indicates that muropeptide supplementation did not simply block all host responses to enteric infection but rather supported a beneficial response. These data suggest that probiotic muropeptide administration could potentially improve mucus layer functionality–an effect that reinforces the mucosal barrier against luminal pathogens. Last, the ability of muropeptides to protect myenteric neurons from death following enteric infection is highly exciting, proposing a key mechanism by which host–microbe interactions may ensure maintenance of GI physiology, and limit pathology. Together, our data identify probiotic muropeptides as beneficial host-directed immunomodulators of enteric infection-induced deficits. By tuning epithelial PRR signaling, restraining immune cell trafficking, reinforcing barrier integrity, and protecting enteric neurons, muropeptides offer a targeted approach to limit inflammation in the immature gut–one with translational potential for preventing long-term neuroimmune sequelae of early-life enteric infections. The gut microbiome was not assessed in these studies given the use of littermate controls and prior literature suggesting that this study design limits the impact of the microbiome in Nod conditional knockout mice (42, 43), however, future studies to determine the potential impact of muropeptides on the gut microbiome would be warranted.

In conclusion, neonatal EPEC infection causes MGB axis deficits through neuroimmune pathways mediated via IEC NOD1, which can be beneficially modulated by probiotic muropeptides via activation of Nod2. Together, these studies highlight the important role of NLRs in coordinating the host neuroimmune response to infection and subsequently driving gut-brain deficits. Excitingly, this study identifies a potentially therapeutic strategy for pediatric enteric bacterial infections.

Materials and Methods

Detailed methods can be found in SI Appendix.

Mice.

All procedures were approved by the Institutional Animal Care and Use Committee (IACUC #23455) at the University of California, Davis. Nod1^ΔIEC mice and WT littermate controls (Vil.Cre⁻Nod1^f/f) on a C57BL/6 background were bred in house (18).

Bacterial Cultures.

(O127:H6 strain E2348/69) were prepared from glycerol stocks maintained at −80 °C. Cultures were grown in 10 mL of LB broth (37 °C overnight) until logarithmic growth was achieved (OD600 ~ 0.6).

Experimental Design.

EPEC infection.

Mice (male/female) were infected with EPEC (10⁵ CFU; 50 μL) or vehicle (LB) at P7 by orogastric gavage. EPEC colonization and clearance were monitored by CFU counts from P14 until P28.

Muropeptides.

Mice were randomly assigned to one of three groups: 1) Sham+LB, 2) EPEC+LB, or 3) EPEC+MP. Muropeptides were reconstituted in LB broth and administered daily orally by pipette (5 µL; 10 μg). The first dose was delivered 1 h prior to EPEC infection and repeated every 24 h until P14.

Behavioral Testing.

Behavioral tests were performed the same day in order of least to most stressful as previously determined. L/D followed by OFT and finally the NOR Task (8, 18, 44, 45). Data were analyzed using automated software (Ethovision, Noldus).

Statistics.

Results were analyzed by one- or two-way ANOVA with a post hoc Tukey’s or Bonferroni test as appropriate using Prism (GraphPad V10). Normality was assessed using the D’Agostino and Shapiro–Wilk tests for Gaussian distribution and the F test was used to assess whether SD was the same across groups. Data are shown as mean ± SEM and P < 0.05 was determined as statistically significant. Sample size was determined by power analysis [α = 0.05, (1 − β) = 0.8] and N represents individual mice, unless otherwise stated.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

Study data are included in the article and/or SI Appendix.