1. Introduction
Inflammatory bowel disease (IBD), encompassing Crohn’s disease (CD) and ulcerative colitis (UC), comprises chronic, relapsing inflammatory disorders of the gastrointestinal tract with complex and incompletely understood etiologies [
1]. The global prevalence of IBD has risen steadily over recent decades, imposing a substantial and growing burden on healthcare systems worldwide [
2], [
3], [
4]. Clinically, IBD is characterized by a fluctuating course of remission and relapse, with disease flares commonly presenting as hematochezia, abdominal pain, diarrhea, and fever, along with a broad spectrum of extraintestinal manifestations (EIMs), including arthritis, uveitis, dermatologic and oral mucosal lesions, hepatobiliary disease, and osteoporosis [
5], [
6].
Although the precise mechanisms driving IBD remain elusive, accumulating evidence implicates intestinal microbial dysbiosis as a central contributor to disease pathogenesis [
7], [
8], [
9]. This dysbiosis is typically marked by reduced microbial diversity, loss of beneficial commensals, and expansion of pro-inflammatory pathobionts. Advances in next-generation sequencing have revealed profound alterations in gut microbial composition in patients with IBD [
10], [
11]. However, whether these microbial shifts represent a primary cause of intestinal inflammation or arise as a consequence of mucosal injury and immune dysregulation remains an area of active investigation [
12]. Beyond the gut microbiome, increasing attention has been directed toward the role of the oral microbiota as a potential upstream regulator of intestinal inflammation. The oral cavity harbors one of the most diverse microbial communities in the human body and serves as a continuous source of microbes that are swallowed and introduced into the gastrointestinal tract. The concept of the oral-gut axis has thus emerged as a critical framework for understanding bidirectional microbial and immunological interactions between these two compartments [
13]. Oral microbial dysbiosis and translocation have been implicated in a range of systemic diseases, including diabetes, cardiovascular disease, and gastric cancer [
14], [
15], [
16]. In IBD, recent studies suggest that oral-derived pathobionts can ectopically colonize the intestine, disrupt mucosal homeostasis, and exacerbate intestinal inflammation [
17], [
18].
The tongue coating represents a distinct, relatively stable, and easily accessible microbial niche within the oral cavity. It consists of a complex matrix of bacteria, desquamated epithelial cells, salivary components, and dietary residues [
19], [
20], and may reflect both local oral conditions and systemic inflammatory states. Clinical observations have reported thicker tongue coatings in patients with IBD, raising the possibility that alterations in tongue coating microbiota are related to intestinal inflammation [
20], [
21]. Despite its potential diagnostic and mechanistic relevance, the composition and functional significance of tongue coating microbiota in IBD remain largely unexplored.
In this study, we sought to systematically characterize the tongue coating microbiota in patients with IBD and to elucidate its contribution to intestinal inflammation. By integrating large-scale human cohort analyses with complementary murine models, we identified a distinct dysbiotic signature in the tongue coating of IBD patients, characterized by enrichment of Streptococcus, Prevotella, and Gemella. This oral microbial profile was recapitulated in a spontaneous enteritis mouse model. Importantly, transplantation of tongue coating microbiota from patients with CD into antibiotic-pretreated mice exacerbated colitis, establishing a causal role for disease-associated oral microbiota in driving gut inflammation. Mechanistic studies further demonstrated that oral-derived Streptococcus strains promote T helper 1 (Th1) cell polarization and aggravate intestinal pathology. Collectively, these findings identify the tongue coating microbiota as a reservoir of pro-inflammatory pathobionts and highlight its potential utility as a non-invasive biomarker for CD, providing new insights into the pathogenic relevance of the oral-gut axis in IBD.
2. Materials and methods
2.1. Patients
Tongue coating and fecal samples were collected from participants recruited at the Department of Digestive Diseases, The First Affiliated Hospital of University of Science and Technology of China (USTC), between December 2020 and December 2022. The study protocol was approved by the Ethics and Investigation Committee of the First Affiliated Hospital of USTC (approval No. 2021KY02), and written informed consent was obtained from all participants prior to enrollment.
Hospitalized patients with a clinical diagnosis of IBD were eligible for inclusion if they had not received antibiotics or traditional Chinese medicine within seven days before sample collection. Diagnoses of UC and CD were established in accordance with the 2010 World Gastroenterology Organization practice guidelines and the European Crohn’s and Colitis Organisation (ECCO)-European Society of Gastrointestinal and Abdominal Radiology (ESGAR) consensus guidelines for the diagnosis and management of IBD [
5], [
22]. These guidelines integrate clinical presentation, endoscopic findings, radiologic assessment, and histopathological evaluation to ensure standardized and evidence-based disease classification.
Demographic characteristics, disease phenotype and duration, clinical symptoms, EIMs, coexisting primary sclerosing cholangitis, and medication histories were extracted from electronic medical records. Patients with a history of surgical resection for intestinal lesions were excluded. Disease activity was assessed using the partial Mayo score for UC and the short Crohn’s disease activity index (sCDAI) for CD. Clinical remission was defined as a partial Mayo score ≤ 2 for UC and an sCDAI ≤ 4 for CD. Healthy controls (HCs) were recruited from individuals undergoing routine health examinations. Participants with a history of chronic metabolic or inflammatory diseases (including diabetes, hypertension, and hyperlipidemia) were excluded. Absence of IBD was confirmed by baseline endoscopic and histopathological evaluations.
2.2. Sample collection
Participants were instructed to refrain from tooth brushing on the night before tongue coating sample collection. On the morning of sampling, prior to oral hygiene, tongue coating was gently scraped from approximately two-thirds of the central and anterior regions of the tongue using sterile instruments. Samples were immediately suspended in an equal volume of phosphate-buffered saline (PBS) containing 20% (w/v) glycerol, snap-frozen in liquid nitrogen, and stored at −80 °C until further analysis. Fecal samples were collected from the same participants on the day of tongue coating sampling and processed using identical cryopreservation procedures. Comprehensive clinical data were recorded for all participants, and all sample collections were conducted on a voluntary basis.
2.3. Animals
All animal experiments were performed using age- and sex-matched C57BL/6 mice (6-10 weeks old) obtained from Gempharmatech Co., Ltd. (China). C57BL/6-TNF
ΔARE mice (TNF: tumor necrosis factor; ARE: AU-rich element) and their wild-type (WT) littermate controls were used as previously described [
23]. Mice were maintained under specific pathogen-free conditions with a 12-h light-dark cycle (lights on from 08:00 to 20:00), controlled temperature and humidity, with
ad libitum access to food and water. All animal procedures were approved by the Ethics Committee of USTC (approval No. 2024-N(A)-114) and conducted in accordance with institutional and national guidelines.
2.4. Tongue coating microbiota transplantation
Tongue coating microbiota transplantation was performed as previously described with minor modifications [
24]. Fresh tongue coating samples from patients with CD or HCs were resuspended in sterile PBS under anaerobic conditions with vigorous agitation and allowed to settle for 2 min. The supernatant containing microbial suspensions was collected and immediately administered to recipient mice. To deplete endogenous commensal microbiota, recipient mice were treated with a broad-spectrum antibiotic cocktail (1 g·L
-1 ampicillin, 1 g·L
-1 neomycin sulfate, 1 g·L
-1 metronidazole, and 0.5 g·L
-1 vancomycin; Sangon Biotech (Shanghai) Co., Ltd., China) in drinking water for seven consecutive days prior to transplantation. Two weeks after microbiota transplantation, experimental colitis was induced by administration of dextran sodium sulfate (DSS) in the drinking water.
2.5. DSS-induced colitis
Experimental colitis was induced by administering 2.5% (w/v) DSS (MP Biomedicals, USA) in the drinking water for two consecutive days, followed by replacement with regular drinking water until the end of the experiment. Mice were monitored daily for changes in body weight and clinical signs of colitis. At sacrifice, colons were excised and measured to assess disease severity.
2.6. 2,4,6-Trinitrobenzene sulfonic acid (TNBS)-induced colitis
TNBS-induced colitis was established as previously described with minor modifications [
25]. Briefly, mice were presensitized by shaving a 1.5 cm × 1.5 cm area of dorsal skin between the scapulae and topically applying 150 μL of TNBS presensitization solution. This solution was prepared by vortexing acetone (Sangon Biotech (Shanghai) Co., Ltd.) and olive oil (MedChemExpress, USA) at a 4:1 (v/v) ratio, followed by mixing four volumes of this solvent with one volume of 5% (w/v) TNBS (Sigma, Germany) to achieve a final TNBS concentration of 1% (w/v). Control mice received an identical vehicle solution lacking TNBS. Seven days after presensitization, mice were weighed, anesthetized, and administered 100 μL of TNBS solution into the colonic lumen using a 3.5-French catheter inserted approximately 4 cm proximal to the anus. The TNBS enema solution was prepared by mixing equal volumes of 5% (w/v) TNBS in water and absolute ethanol. Following instillation, mice were maintained in a vertical, head-down position for 60 s to ensure retention of the solution before recovery.
2.7. Bacterial isolation and culture
Bacterial isolation was performed as previously described [
26]. Tongue coating samples from patients with CD were serially diluted in sterile PBS (10
-1 to 10
-4) and plated onto non-selective Gifu anaerobic medium (GAM) agar (Solarbio, China). Plates were incubated under anaerobic conditions (80% N
2, 10% H
2, 10% CO
2) in an anaerobic chamber (Electrotek, UK) at 37 °C for 2-4 d. Individual colonies were picked and identified by amplification and sequencing of the 16S ribosomal RNA (rRNA) gene using universal primers (27F: 5′-AGRGTTTGATYMTGGCTCAG-3′, 1492R: 5′-GGYTACCTTGTTACGACTT-3′). Resulting sequences were compared against the National Center for Biotechnology Information (NCBI) database for taxonomic assignment. For bacterial administration, individual strains were cultured in GAM broth (Solarbio), harvested during the logarithmic growth phase, and orally gavaged into antibiotic-pretreated mice at approximately 1 × 10
8 colony forming units (CFUs) per strain in 200 μL sterile PBS per mouse. For preparation of heat-killed bacteria,
Streptococcus strains were cultured twice with sterile PBS, and heat-inactivated at 121 °C for 30 min.
2.8. In vitro T cell differentiation
Naïve cluster of differentiation 4 positive (CD4) T cells were isolated from mouse spleens using a CD4 naïve T cell isolation kit (BioLegend, USA) according to the manufacturer’s instructions [
27], [
28]. Cells were seeded at equal densities into 96-well plates pre-coated with anti-CD3 antibody (2 μg·mL
-1; BioLegend) and stimulated in the presence of bacterial culture supernatants or control media (pH adjusted to 7.0) at a final concentration of 2.5% (v/v), or with heat-killed bacteria at a 1:1 cell-to-bacteria ratio. For Th17 cell differentiation, cultures were supplemented with anti-CD28 monoclonal antibody (1 μg·mL
-1; BioLegend), mouse interleukin (IL)-6 (25 ng·mL
-1; BioLegend), mouse IL-23 (20 ng·mL
-1; BioLegend), human transforming growth factor-β1 (TGF-β1; 2 ng·mL
-1; BioLegend), anti-mouse interferon-γ (IFN-γ; 5 μg·mL
-1; BioLegend), and anti-mouse IL-4 (5 μg·mL
-1; BioLegend). For Th1 cell differentiation, cultures received anti-CD28 (1 μg·mL
-1), human IL-2 (50 U·mL
-1; BioLegend), mouse IL-12 (20 ng·mL
-1; BioLegend), and anti-mouse IL-4 (5 μg·mL
-1). Cells were cultured at 37 °C in a humidified incubator with 5% CO
2 for 4 d. Following stimulation, cells were harvested and intracellular IFN-γ and IL-17A expression was analyzed by flow cytometry.
2.9. Isolation of lamina propria lymphocytes (LPLs)
LPLs were isolated from colonic tissues immediately following euthanasia. Excised colons were flushed with ice-cold PBS (KeyGEN BioTECH, China), opened longitudinally, and subjected to two sequential 30-min incubations in PBS supplemented with 2 mmol·L
-1 ethylenediaminetetraacetic acid (EDTA; Sangon Biotech (Shanghai) Co., Ltd.) and 2% fetal bovine serum (FBS; VivaCell, China) at 37 °C with constant shaking at 250 r·min
−1 to remove epithelial cells. Tissues were then extensively washed with PBS to eliminate residual EDTA. Subsequently, colonic tissues were minced into fragments < 5 mm in size and enzymatically digested for 30 min at 37 °C in Roswell Park Memorial Institute (RPMI)-1640 medium (KeyGEN BioTECH) containing 0.5 mg·mL
-1 collagenase type II (Sigma), 0.5 mg·mL
-1 DNase I (Sigma), and 2% FBS, with continuous agitation at 250 r·min
−1. The resulting cell suspension was filtered through a 70-μm nylon mesh, washed twice with PBS containing 2% FBS, and resuspended in complete RPMI-1640 medium supplemented with 10% FBS and 1% penicillin-streptomycin. Cell viability consistently exceeded 95%, as assessed by trypan blue exclusion [
27]. Isolated cells were immediately processed for downstream analyses.
2.10. Flow cytometry analysis
Single-cell suspensions were incubated with fluorochrome-conjugated monoclonal antibodies (BioLegend) diluted in PBS containing 0.5% FBS and 2 mmol·L-1 EDTA for 20 min at 4 °C in the dark for surface marker staining. For intracellular cytokine detection, cells were stimulated for 4 h in RPMI-1640 medium supplemented with 10% FBS and cell stimulation cocktail (1:500 dilution; eBioscience, USA), according to the manufacturer’s instructions. Following stimulation, 1 × 106 cells were fixed and permeabilized using the BD Cytofix/Cytoperm™ kit (BD Biosciences, USA) for 20 min at 4 °C, followed by intracellular staining with cytokine-specific antibodies as per the manufacturer’s protocol. Cells were washed with Perm/Wash™ buffer (BD Biosciences), resuspended in magnetic-activated cell sorting (MACS) buffer (PBS containing 0.5% FBS and 2 mmol·L-1 EDTA), and analyzed on a CytoFLEX S flow cytometer (Beckman Coulter, USA). Data acquisition and analysis were performed using FlowJo™ software v10.8 (BD Life Sciences, USA). Gating strategies were standardized across experiments, and all antibodies were validated for specificity and used at a dilution of 1:200 unless otherwise indicated. The following antibody conjugates were used: CD45.2-APC/Cy7 (clone 104), CD3-FITC (clone 17A2), CD4-Super Bright 600 (clone RM4-5), IL-17A-PE (clone TC11-18H10.1), and IFN-γ-PE/Cy7 (clone XMG1.2). Cell viability was assessed using 4′,6-diamidino-2-phenylindole (DAPI; BioLegend) and MitoTracker™ deep red FM dye (Thermo Fisher Scientific, USA).
2.11. Histological analysis
Distal colon segments from colitis model mice were fixed in 4% paraformaldehyde (PFA) and cryoprotected through a graded sucrose series (10%, 20%, and 30%). For each sample, four transverse sections representing distinct anatomical regions were prepared and stained with hematoxylin and eosin (H&E) according to standard protocols. Histological evaluation was performed in a blinded manner using established scoring criteria encompassing three parameters: ulceration severity (0, absent; 1, punctate; 2, minimal; 3, moderate; and 4, extensive), inflammatory cell infiltration (0, none; 1, minimal; 2, mild; 3, moderate; and 4, severe), and depth of inflammation (0, none; 1, mucosal; 2, mucosal and submucosal; 3, transmural; and 4, full-thickness) [
29]. The cumulative histological score for each animal was calculated by summing the scores of these parameters across three randomly selected sections.
2.12. 16S rRNA gene sequencing and microbiota analysis
All samples were stored at −80 °C until processing. Genomic DNA was extracted from tongue coating, intestinal mucus, and fecal samples using the QIAamp Fast DNA Stool Mini Kit (Cat. No. 51604; QIAGEN, Germany) according to the manufacturer’s protocol. The V4 hypervariable region of the bacterial 16S rRNA gene was amplified using the primer pair 515F/806R, as recommended by the Earth Microbiome Project 2. The barcoded forward primer sequence was 515F: AATGATACGGCGACCACCGAGATCTACACGCTXXXXXXXXXXXXTATGGTAATTGTGTGYCAGCMGCCGCGGTAA, and the reverse primer sequence was 806R: CAAGCAGAAGACGGCATACGAGATAGTCAGCCAGCCGGACTACNVGGGTWTCTAAT. Amplicons were quantified, pooled at equimolar concentrations, purified using the QIAquick PCR Purification Kit (Cat. No. 28104; QIAGEN), and sequenced on an Illumina MiSeq platform using the v3 kit with 250-bp paired-end reads. Sequencing yielded an average depth of 29 292 ± 11 228 (mean ± standard deviation (SD)) reads per sample. Raw sequencing data were processed using a customized bioinformatics pipeline integrating USEARCH (v8.1), VSEARCH (v2.13.0), and QIIME (v1.9.1). Quality filtering, chimera removal, and clustering were performed to generate an operational taxonomic unit (OTU) table. Taxonomic assignment was conducted using the Ribosomal Database Project (RDP) classifier. For β-diversity analyses, unweighted UniFrac distances were calculated and visualized by principal coordinates analysis (PCoA) based on a rarefaction depth of 5000 reads per sample. Statistical differences in microbial community structure among groups were assessed using permutational multivariate analysis of variance (PERMANOVA) implemented in the vegan package (v2.5-7) in R (v4.0.5).
2.13. Accession numbers
The 16S rRNA gene sequencing datasets have been deposited in the NCBI Sequence Read Archive (SRA) under BioProject accession numbers PRJNA1103394, PRJNA1103950, and PRJNA1102901.
3. Results
3.1. Distinct microbial profiles in oral and fecal microbiota of patients with IBD
A total of 596 participants were included in the study, comprising 91 patients with UC (68 active and 23 in remission), 278 patients with CD (145 active and 133 in remission), and 227 HCs (
Table 1).
We analyzed 16S rRNA sequencing data from 414 tongue coating samples (195 HCs, 173 CD and 46 UC patients) and 81 fecal samples (9 HCs, 47 CD and 25 UC patients). PCoA based on unweighted UniFrac distances revealed a significant separation of tongue coating microbial communities between CD patients and HCs (PERMANOVA, CD vs HC:
P = 0.001), whereas no significant difference was observed between UC patients and HCs (UC vs HC:
P = 0.168;
Fig. 1(a)). Consistent with these findings, CD patients exhibited a significant reduction in α-diversity compared with HCs (
Figs. 1(b) and
(c)). At the genus level, both CD and UC patients showed a marked enrichment of
Streptococcus in the tongue coating relative to HCs (
Figs. 1(d) and
(e)).
Prevotella was also significantly increased in CD patients. Further differential abundance analyses identified enrichment of
Gemella and concomitant depletion of commensal genera, including
Fusobacterium and
Neisseria, specifically in the tongue coating of CD patients (
Figs. 1(f) and
(g)). In UC patients, elevated abundances of select
Prevotella species and
Gemella were similarly observed
(Figs. 1(f) and
(h))
. Collectively, these results demonstrate that IBD, particularly CD, is associated with a distinct dysbiotic signature in tongue coating microbiota, characterized by expansion of oral-associated genera such as
Streptococcus,
Prevotella, and
Gemella. The consistent enrichment of these taxa suggests a potential role for tongue-derived oral pathobionts in amplifying mucosal immune responses and contributing to intestinal inflammation.
PCoA of fecal microbiota based on Bray-Curtis dissimilarity revealed clear separation among CD, UC, and HC groups (PERMANOVA: CD vs HC,
P = 0.002; UC vs HC,
P = 0.001;
Fig. 1(i)). Both CD and UC patients exhibited significantly reduced microbial richness and diversity compared with HCs, as reflected by decreased Shannon and Simpson indices (
Figs. 1(j) and
(k)), indicating a global loss of commensal bacterial complexity in IBD. At the genus level,
Prevotella was significantly enriched in fecal samples from CD patients (
Fig. 1(l)). Linear discriminant analysis (LDA) effect size (LEfSe) analysis further identified disease-specific dysbiotic signatures. CD-associated fecal microbiota was characterized by enrichment of potentially pro-inflammatory genera, including
Escherichia-Shigella,
Veillonella, and
Fusobacterium, accompanied by marked depletion of beneficial commensals such as
Bifidobacterium,
Eubacterium hallii, and
Anaerostipes (
Fig. 1(m)). In contrast, UC fecal microbiota showed increased abundances of
Morganella,
Erysipelatoclostridium,
Romboutsia, and
Flavonifractor, along with decreased levels of
Anaerostipes,
Eubacterium hallii,
Bacteroides massiliensis, and
Alistipes (
Fig. 1(m)). Collectively, these findings demonstrate that both CD and UC are associated with a shared reduction in fecal microbial diversity and a consistent depletion of key butyrate-producing taxa, despite each disease exhibiting a distinct compositional dysbiosis. This loss of short-chain fatty acid (SCFA)-producing bacteria may represent a common microbial mechanism contributing to impaired mucosal homeostasis in IBD.
3.2. Animal models recapitulate oral—gut microbiota dysbiosis associated with IBD
To validate the generalizability of the oral-gut microbiota dysbiosis observed in patients with IBD and to explore its potential pathogenic relevance, we examined the TNF
ΔARE spontaneous enteritis mouse model [
23]. Compared with WT littermate controls, TNF
ΔARE mice exhibited marked alterations in microbial community structure in both fecal and tongue coating samples, as evidenced by distinct β-diversity clustering in PCoA analysis (
Figs. 2(a) and
(b)). Importantly, TNF
ΔARE mice recapitulated several key microbial features characteristic of CD. In the intestinal compartment, TNF
ΔARE mice displayed CD-like dysbiosis, including significant enrichment of
Prevotella, particularly
Prevotellaceae UCG-001 and
Prevotellaceae NK3B31 (
Fig. 2(c)). Concordantly, the tongue coating microbiota of TNF
ΔARE mice showed increased abundance of
Streptococcus accompanied by a depletion of
Neisseria (
Fig. 2(d)), similar to the oral microbial shifts observed in CD patients. Notably,
Prevotellaceae NK3B31 was simultaneously enriched in both fecal and tongue coating samples from TNF
ΔARE mice (
Fig. 2(e)), suggesting coordinated oral-gut expansion of this taxon.
Prevotellaceae UCG-001 was significantly increased in fecal samples and exhibited a similar increasing trend in the oral microbiota (
Fig. 2(f))
. Together, these findings support a potential pro-inflammatory role for specific
Prevotella lineages across oral and intestinal niches. Consistent with these observations, LEfSe analysis identified enrichment of multiple
Prevotella species in the gut of TNF
ΔARE mice, along with a wide depletion of key commensal and SCFA-producing genera, including
Bifidobacterium and
Anaerostipes. In the oral compartment, LEfSe further revealed a significant reduction in
Neisseria abundance (
Fig. 2(g)). Collectively, these results demonstrate that the TNF
ΔARE model faithfully reproduces the oral-gut microbial dysbiosis observed in human IBD, supporting its utility for mechanistic studies of microbiota-driven intestinal inflammation.
3.3. Tongue coating microbiota from CD patients exacerbates colitis in mice
To directly assess the pathogenic potential of tongue coating microbiota in IBD, we transplanted tongue coating microbial communities from patients with CD or HCs into antibiotic-treated (Abx) mice, followed by induction of colitis using DSS (
Fig. 3(a)). Mice colonized with CD-derived tongue coating microbiota (Abx-CD) developed significantly more severe colitis than mice receiving HC-derived microbiota (Abx-HC) or antibiotic-treated controls without transplantation (Abx), as evidenced by greater body weight loss, shortened colon length, and more pronounced histopathological damage (
Figs. 3(b)-(f)). In contrast, no significant differences were observed between the Abx-HC and Abx groups, indicating that tongue coating microbiota from healthy individuals did not exacerbate intestinal inflammation. To determine whether these phenotypic differences were associated with alterations in intestinal microbial composition, we performed 16S rRNA gene sequencing of fecal samples following transplantation. PCoA revealed distinct microbial community structures among the Abx-CD, Abx-HC, and Abx groups (
Fig. 3(g)), accompanied by significant differences between the Abx-CD and Abx-HC groups in α-diversity (
Figs. 3(h) and
(i)). Notably, mice receiving CD-derived tongue coating microbiota exhibited enrichment of potentially pro-inflammatory taxa, including
Klebsiella, along with depletion of SCFA-producing commensals such as
Akkermansia,
Lachnoclostridium, and
Anaerostipes, compared with Abx-HC mice (
Figs. 3(j) and
(k);
P < 0.05). These findings suggest that specific oral bacteria present in CD patients possess a selective advantage for colonization and expansion within the inflamed intestinal environment. Consistent with this notion,
Klebsiella species, previously isolated from oral and salivary microbiota, have been shown to induce Th1 cell responses and promote intestinal inflammation [
18]. In contrast, tongue coating microbiota from healthy individuals exerted minimal effects on gut microbial composition and disease severity, likely reflecting the absence of pathogenic oral taxa capable of establishing ectopic colonization in the intestine. Together, these results provide functional evidence that CD-associated tongue coating microbiota can reshape gut microbial communities and directly exacerbate colitis.
3.4 Colonization with Streptococcus vestibularis (S. vestibularis) from CD patients exacerbates colitis
Having demonstrated the pathogenic potential of the tongue coating microbiota as a whole, we next sought to identify specific oral bacterial taxa responsible for driving intestinal inflammation. From the tongue coating samples of CD patients, we successfully isolated two streptococcal species,
Streptococcus salivarius (
S. salivarius) and
S. vestibularis, a genus consistently enriched in both CD patients and spontaneous enteritis (TNF
ΔARE) mice. Although
Prevotella species were also enriched in tongue coating and fecal samples from CD patients, repeated attempts to isolate viable
Prevotella strains were unsuccessful. We first evaluated the immunomodulatory capacity of the isolated
Streptococci in vitro. Under Th1 cell-skewing conditions, both
S. salivarius and
S. vestibularis significantly promoted Th1 cell differentiation, as evidenced by increased IFN-γ production, while exerting minimal effects on Th17 cell polarization (
Figs. 4(a)-(d), Fig. S1 in Appendix A). These findings suggested a preferential role for streptococcal species in driving Th1 cell-biased immune responses. To assess the pathogenic relevance of these observations
in vivo, antibiotic-treated mice were colonized with
S. vestibularis and subsequently subjected to DSS-induced colitis (
Fig. 4(e)). Mice colonized with
S. vestibularis developed markedly more severe colitis than control mice, as indicated by enhanced body weight loss, increased histopathological scores, and a significant accumulation of Th1 cells in the colonic lamina propria (
Figs. 4(f)-(h), Figs. S2(a)-(d) in Appendix A). Notably, the proportion of IFN-γ-producing CD4
+ T cells was selectively increased, whereas Th17 cells were not substantially affected. The pathogenic role of
S. vestibularis was further validated using an independent TNBS-induced colitis model (
Fig. 4(i)). Consistent with the DSS model,
S. vestibularis-colonized mice exhibited greater body weight loss, more pronounced colon shortening, exacerbated histopathological injury, and significantly increased infiltration of colonic Th1 cells compared with controls (
Figs. 4(j)-(l), Figs. S2(e)-(h) in Appendix A). Collectively, these findings establish a direct causal relationship between a specific CD-associated oral bacterium and intestinal immunopathology. They identify
S. vestibularis as a potent Th1 cell-inducing pathobiont capable of ectopic intestinal colonization and exacerbation of colitis, thereby implicating oral
Streptococcus species as key contributors to the oral-gut axis in IBD.
4. Conclusions
This study provides evidence that the tongue coating microbiota serves not only as a potential diagnostic biomarker but also as a functional contributor to IBD, particularly CD. By integrating comprehensive microbial profiling with experimental validation, we demonstrate that oral-gut microbial interactions constitute a previously underappreciated pathway in IBD pathogenesis. Our findings advance mechanistic understanding of the oral-gut axis and highlight opportunities for non-invasive diagnostics and targeted microbial interventions in chronic intestinal inflammation. We identified a distinct dysbiotic signature in the tongue coating of IBD patients, characterized by enrichment of
Streptococcus,
Prevotella, and
Gemella. Corresponding alterations were observed in fecal microbiota, with CD patients showing increased
Fusobacteriota and
Prevotella, along with depletion of butyrate-producing taxa such as
Anaerostipes and
Eubacterium hallii, suggesting a potential ecological linkage between oral and gut microbial communities. Although
Prevotella is generally associated with fiber-rich diets and SCFA production [
30], [
31], [
32], several strains (e.g.,
Prevotella copri,
Prevotella intermedia, and
Prevotella nigrescens) have been implicated in pro-inflammatory responses, immune modulation via Th17 cell pathways, and diseases including periodontitis, rheumatoid arthritis, and IBD [
33], [
34], [
35], [
36], [
37], [
38], [
39], [
40]. This dualistic role highlights the importance of future studies to resolve strain-level functions and their context-specific impacts on host health [
30]. The biological relevance of these oral-gut microbial alterations was further supported in a spontaneous model of chronic intestinal inflammation. TNF
ΔARE mice recapitulated key microbial features observed in CD patients, including enrichment of
Streptococcus in the tongue coating and selective
Prevotella strains in both oral and gut niches, reinforcing the plausibility of oral-derived bacteria in driving intestinal inflammation. Functionally, we established causality through oral microbiota transplantation experiments. Mice receiving tongue coating microbiota from CD patients developed more severe DSS-induced colitis than those receiving HC microbiota. This was accompanied by expansion of pro-inflammatory genera such as
Klebsiella [
41], [
42], [
43], and depletion of beneficial SCFA-producing taxa, including
Akkermansia and
Anaerostipes [
44], [
45], [
46]. These results align with previous evidence that oral pathobionts can modulate intestinal immunity and exacerbate chronic gut inflammation. Mechanistically, we show that
Streptococcus strains isolated from CD patients promote Th1 cell polarization
in vitro and aggravate colitis
in vivo, providing direct support for a pathogenic role of oral
Streptococci in gut inflammation. Previous studies have linked specific
Streptococcus species to intestinal or systemic pathology:
Streptococcus pneumoniae disrupts gut microbial homeostasis and triggers Toll-like receptor (TLR)-mediated systemic inflammation;
Streptococcus mutans can exacerbate dysbiosis following translocation [
47], [
48], [
49];
Streptococcus gallolyticus promotes colorectal cancer and inflammation via IL-1β/Th17 cell signaling [
50]; and commensal oral
Streptococci such as
Streptococcus sanguinis and
Streptococcus oralis impair epithelial barrier integrity [
51]. Accordingly, our identification of pathogenic
S. vestibularis in CD patients highlights that a consortium of oral
Streptococci may cooperatively fuel intestinal inflammation through diverse immunological mechanisms, establishing a mechanistic relationship between oral microbiota and CD pathogenesis.
While our study provides novel insights into the role of tongue coating microbiota in IBD, several limitations should be acknowledged. First, focusing exclusively on tongue coating samples may not fully capture the complexity of oral-gut axis dysbiosis, as the oral cavity comprises multiple distinct ecological niches (e.g., subgingival crevices and saliva) with specialized microbial communities that may differentially influence intestinal inflammation [
52], [
53], [
54]. Future studies employing multi-niche sampling strategies are needed to comprehensively define the oral microbiome’s contribution to IBD pathogenesis. Methodologically, we faced challenges in isolating oral
Prevotella strains, which precluded direct functional validation of their proposed pro-inflammatory role. Additionally, although our mouse models recapitulated key features of human IBD, inherent interspecies differences in host-microbe interactions warrant caution when extrapolating these findings to humans. From a clinical perspective, inconsistent documentation of oral manifestations within our cohort limited our ability to stratify patients based on this potential confounder. It remains possible that the observed oral dysbiosis reflects local oral inflammation, which may in turn influence gut homeostasis. To address these gaps, future work should integrate multi-niche oral sampling with shotgun metagenomics to achieve strain-level resolution, enhanced culturing techniques for functional characterization of specific isolates, concurrent host immune profiling, and systematic oral phenotyping. Such integrative approaches will be essential for establishing causality and elucidating the precise mechanisms through which oral microbes contribute to intestinal inflammation in IBD.
PDF
(4639KB)
