Periodontitis is driven not only by bacteria but by how the gingival epithelium responds to them. This study maps that response at spatial and single-cell resolution, showing that a distinct serum amyloid A1 (SAA1)-high inflammatory epithelial subpopulation (SAA1+Epi) sits at the junctional epithelium, where subgingival plaque meets host tissue. The work links lipopolysaccharide (LPS) from Porphyromonas gingivalis (P. gingivalis) to a Toll-like receptor 2 (TLR2)–paired-like homeodomain transcription factor 2 (PITX2) axis that reprograms epithelial cells. It further shows that SAA1+Epi cells instruct complement C3–positive fibroblasts (C3+FB) through transforming growth factor beta (TGF-β) and recruit natural killer (NK) cells via chemokines such as C-X-C motif chemokine ligand 6 (CXCL6), revealing a circuit that sustains inflammation and bone loss.
Epithelial barriers are active organizers of tissue homeostasis, immune surveillance and host defense, yet chronic inflammation can corrupt their repair and communication with connective tissue. In periodontitis, pathogens and virulence factors disrupt junctions such as E-cadherin and occludin, activate epithelial pattern-recognition receptors and trigger signaling that promotes apoptosis, barrier breakdown, osteoclastogenesis and alveolar bone loss. Single-cell and spatial studies have begun to resolve gingival heterogeneity, including keratinocyte states and proinflammatory fibroblasts, but the epithelial subclusters that sense local microbial communities, how they differentiate in situ and how they instruct stromal and immune cells remain poorly defined. Given these challenges, in-depth research is needed on spatially resolved microbe–epithelium–connective tissue crosstalk in periodontitis.
Researchers at the West China Hospital of Stomatology, Sichuan University, report the findings in International Journal of Oral Science , published (DOI: 10.1038/s41368-026-00464-1) online on 29 September 2026. The team combined spatial transcriptomics (ST), single-cell RNA sequencing (scRNA-seq) and metagenomic sequencing of matched gingival tissue and subgingival plaque from healthy controls and patients with periodontitis. By mapping single-cell data onto spatial sections, they localized an inflammatory epithelial zone near the junctional epithelium and identified SAA1 +Epi as a spatially restricted driver of epithelial–connective tissue and epithelial–immune crosstalk.
The study first resolved four epithelial regions in human gingiva. Epithelium-3, at the basal layer of the junctional epithelium, expanded in periodontitis and showed bacterial-response and inflammatory programs. Within this niche, serum amyloid A1 ( SAA1 )-high inflammatory epithelial subpopulation ( SAA1 +Epi) cells expressed SAA1 , SAA2 , MMP13 and CSF3 and were enriched in diseased human and mouse tissue. Metagenomic and host–microbiome analyses pointed to P. gingivalis and its LPS as dominant triggers. In human gingival epithelial cells, P. gingivalis LPS activated TLR2, increased phosphorylation of extracellular signal-regulated kinase 1/2 (ERK1/2) and SMAD family member 2/3 (Smad2/3), and induced PITX2, a transcription factor required for SAA1 +Epi differentiation; PITX2 knockdown reduced and overexpression enhanced this program. Inhibiting TLR2, ERK1/2 or Smad2/3 reduced SAA1 +Epi induction, whereas PITX2 manipulation did not alter upstream phosphorylation, placing PITX2 downstream. Local Pitx2 silencing in mice reduced SAA1 +Epi, C3+FB, NK-cell recruitment, inflammatory infiltration and alveolar bone loss. Mechanistically, SAA1 +Epi-derived TGF-β drove C3+FB differentiation through phosphoinositide 3-kinase (PI3K)/protein kinase B (Akt) signaling, while SAA1 +Epi chemokines including CXCL6 recruited NK cells. Spatial mapping and multiplex immunofluorescence (mIF) showed C3+FB and NK cells accumulating near SAA1 +Epi-rich inflammatory zones, supporting a feedforward epithelial–stromal–immune circuit.
The authors said the work reframes the gingival epithelium as an active organizer of periodontitis rather than a passive barrier. They said the convergence of microbial LPS sensing, a TLR2–ERK1/2–Smad2/3–PITX2 cascade and downstream TGF-β and chemokine signals identifies a druggable epithelial hub. They added that targeting PITX2 reduced both C3+FB induction and NK-cell recruitment in models, suggesting that interrupting this epithelial program could complement conventional antibacterial care. They cautioned that human cohorts were small and that epithelial-specific PITX2 deletion and broader polymicrobial interactions still need study.
The findings suggest that PITX2 and the SAA1 +Epi program could be explored as disease-modifying targets in periodontitis, especially for patients whose inflammation persists despite plaque control. A local RNA-based or small-molecule strategy might aim to quiet the epithelial inflammatory state, reduce C3+FB-driven stromal remodeling and limit NK-cell recruitment, while preserving barrier repair. Because the axis is spatially confined near the junctional epithelium, delivery could potentially be localized to reduce systemic exposure. The work also provides a template for studying microbe–epithelium–connective tissue circuits in other chronic inflammatory diseases. Validation in larger, balanced cohorts and cell-type-specific genetic models will be essential before clinical translation.
###
References
DOI
Original Source URL
https://www.nature.com/articles/s41368-026-00464-1
Funding information
National Natural Science Foundation of China (81991501, 82170949, 32470205); Sichuan Science and Technology Program (2024NSFC1584, 2025ZNSFSC1910); Research and Development Program, West China Hospital of Stomatology, Sichuan University (RD-02-202510).
About International Journal of Oral Science
International Journal of Oral Science (ISSN 1674-2818) was founded in 2009 and aims to publish all aspects of oral science and interdisciplinary fields, including fundamental, applied and clinical research. Covered areas include oral microbiology, oral and maxillofacial oncology, cariology, oral inflammation and infection, dental stem cells and regenerative medicine, craniofacial surgery, dental materials, oral biomechanics, oral, dental and maxillofacial genetic and developmental diseases.
International Journal of Oral Science
Not applicable
Spatially resolved single cell atlas deciphers SAA1 inflammatory epithelial cells
29-Sep-2026
The authors declare that they have no competing interests.