Toll-like receptor 9 deficiency induces osteoclastic bone loss via gut microbiota-associated systemic chronic inflammation

The study investigates whether TLR9 signaling is required for normal bone remodeling in vivo and how its absence affects bone metabolism. Inflammation and bone turnover are tightly linked through osteoimmunology: activated immune cells (especially T and B cells) secrete cytokines (TNFα, IL1β, IL6, IL17) and RANKL that drive osteoclastogenesis, while OPG counterbalances RANKL. Low-grade systemic inflammation contributes to bone loss in aging and endocrine conditions, and emerging evidence implicates the gut microbiota as a modulator of systemic immunity and bone. The role of TLR9 in inflammation is context-dependent and controversial, and its specific contribution to bone remodeling in physiological conditions is unclear. The purpose of this study is to determine the impact of TLR9 deficiency on bone mass, define the inflammatory and hematopoietic changes underlying any bone phenotype, and test whether gut microbiota alterations mediate these effects.

Prior work establishes that inflammatory cytokines from immune cells promote osteoclast differentiation and bone resorption in diseases such as RA, IBD, and HIV, and that low-grade inflammation contributes to bone loss in aging, estrogen deficiency, and hyperparathyroidism. T and B cells regulate bone via RANKL and OPG, with dysregulated RANKL/OPG linked to bone loss. The gut microbiota modulates systemic immunity and influences bone, including estrogen deficiency-induced bone loss. TLRs recognize microbial patterns and shape innate immune responses; TLRs 1–9 are expressed in osteoclast progenitors and can either promote or inhibit osteoclastogenesis in vitro. Reports on TLR9 are conflicting: exogenous CpG activation can suppress RANKL-induced osteoclastogenesis in naïve precursors but enhance it in primed cells, often via TNFα. In vivo roles of TLR9 in bone homeostasis remain unresolved, and TLR9 has shown both pro- and anti-inflammatory roles across models of autoimmunity and inflammation, suggesting context-specific functions.

- Animals and genetics: Global TLR9−/− mice (C57BL/6J background) and WT controls; Rag1−/− mice as lymphocyte-deficient recipients; germ-free (GF) derivation of TLR9−/− and WT via hysterectomy and axenic fostering; cohousing of newly weaned WT and TLR9−/− for microbial transfer; bone marrow chimeras by lethal irradiation of WT recipients followed by i.v. transfer of TLR9−/− or WT BM cells; adoptive transfer of 1×10^7 splenocytes from TLR9−/− or WT into Rag1−/− recipients; 5-fluorouracil (5-FU) single-dose challenge to probe hematopoietic recovery. - Anti-TNFα therapy: Continuous delivery using subcutaneous Alzet minipumps loaded with neutralizing anti-TNFα antibody or isotype control for 6 weeks in TLR9−/− and WT mice. - Bone phenotyping: Micro-CT (femurs, L3 vertebrae) for trabecular (BMD, BV/TV, Tb.N, Tb.Th, Tb.Sp) and femoral cortical indices; histomorphometry (TRAP for osteoclasts, osteocalcin IHC for osteoblasts; dynamic calcein labeling for MAR); serum bone turnover markers CTX and P1NP by ELISA. - In vitro osteoclastogenesis: Primary BM nonadherent mononuclear cells cultured with M-CSF and RANKL for 5 days; TRAP staining to quantify multinucleated OCLs; qPCR of osteoclast genes (CTSK, RANK, NFATc1); Western blots in early OCPs (RANK, c-Fos, NFATc1, CSF1R); proliferation by BrdU; TLR9 antagonist (ODN-2088) titration assays. - Cytokines and immune phenotyping: ELISAs for TNFα, IL1β, IL6, IL17, IFNγ, RANKL, OPG in serum and bone marrow supernatants; spleen size; flow cytometry of spleen, bone marrow, MLN, and Peyer’s patches for T cells (CD4, CD8, activation marker CD69), intracellular TNFα/IFNγ, B cells (CD19/B220), macrophages (CD11b, F4/80), and OCPs (B220−CD11b+CD115+, B220−CD11b+Ly6Chi); culture supernatants from purified CD4+ T cells and CD19+ B cells for RANKL/OPG. - Transwell cocultures: Rag1−/− BMNCs in lower chamber with activated TLR9−/− or WT splenocytes/CD4+ T cells/B cells in upper chamber to assess soluble-factor-driven osteoclastogenesis. - Microbiota and gut inflammation: 16S rRNA sequencing (V4) of fecal microbiota; PCoA and LEfSe; measurement of circulating LPS and IgA; large intestine length; fecal Lipocalin-2 (Lcn-2); mucosal immune cell profiling (MLN, PP); assessment after anti-TNFα, GF, and cohousing interventions. - Hematopoiesis: Single-cell RNA-seq (10x Genomics) of BMNCs from TLR9−/− and WT (two independent experiments), Seurat clustering, UMAP/tSNE, Monocle pseudotime, SCENIC regulon analysis, ligand–receptor inference; KEGG/GO enrichment. Flow cytometry for progenitors (HSPCs; CMP, GMP, cMoP, CLP), CD11b+ myeloid cells, OCPs; ex vivo differentiation of FACS-sorted LSK (Lin−Sca1+Kit+) under myeloid vs lymphoid cytokines; in vivo hematopoietic response post 5-FU; analysis in BM chimeras, GF, and cohoused mice. - Statistics: Two-tailed t-tests, Mann–Whitney for nonparametric data, one-/two-way ANOVA with Tukey’s tests for multiple comparisons; significance at P<0.05.

- Bone phenotype: TLR9−/− mice displayed significantly reduced trabecular bone mass in distal femur and vertebrae by µCT, with no cortical differences. Circulating CTX and P1NP were elevated, indicating high bone turnover. Histomorphometry showed increased osteoclast numbers and resorption indices; osteoblast numbers/surface were unchanged, but MAR modestly increased. - Osteoclastogenesis: BMNCs from TLR9−/− generated more TRAP+ multinucleated OCLs in vitro; osteoclast genes (CTSK, RANK, NFATc1) and OCP proteins (RANK, Fos, NFATc1, CSF1R) were upregulated; OCPs proliferated faster to CSF1. Pharmacologic TLR9 blockade in WT BMNCs did not recapitulate the KO phenotype, arguing against an osteoclast-intrinsic requirement for TLR9 signaling as the primary driver. - Systemic inflammation: TLR9−/− mice had splenomegaly and elevated serum/bone marrow TNFα, IL1β, IFNγ, and RANKL, with decreased OPG and increased RANKL/OPG ratio; IL6/IL17 were not consistently increased. Flow cytometry showed expanded and activated (CD69+) CD4+ and CD8+ T cells, with more TNFα+ and IFNγ+ producers; B cell counts were reduced. Purified TLR9−/− CD4+ T cells secreted more TNFα, IFNγ, and RANKL; TLR9−/− B cells secreted more RANKL and less OPG. - Causality of immune-derived factors: Soluble factors from TLR9−/− splenocytes and specifically CD4+ T cells enhanced osteoclastogenesis from Rag1−/− BMNCs in Transwells. Adoptive transfer of TLR9−/− splenocytes into Rag1−/− recipients induced trabecular bone loss, elevated CTX, splenomegaly, T cell activation, and increased systemic cytokines (TNFα, IFNγ, IL1β, RANKL). - Bone marrow compartment: WT recipients of TLR9−/− bone marrow (KOchim) exhibited reduced trabecular bone mass, higher CTX, increased in vitro osteoclastogenesis from BMNCs, splenomegaly, and elevated serum TNFα, IL6, RANKL with reduced OPG, indicating osteoclast-driven bone loss mediated by hematopoietic cells. - Anti-TNFα therapy: Neutralization of TNFα in TLR9−/− mice restored trabecular bone mass to WT levels, reduced CTX (without altering P1NP), and lowered circulating TNFα and IL6 (with trends for lower RANKL and IFNγ) and reduced splenic CD4+/CD8+ frequencies. - Gut microbiota dysbiosis: 16S profiling showed distinct fecal microbial communities in TLR9−/− vs WT. TLR9−/− had increased Deferribacteraceae, Odoribacteraceae, Rikenellaceae, Staphylococcaceae and decreased Erysipelotrichaceae, Turicibacteraceae. Gram-negative taxa including Mucispirillum schaedleri (Deferribacteres) and Parabacteroides distasonis (Bacteroidetes) were enriched. Circulating LPS and IgA were elevated. TLR9−/− showed shortened large intestines, increased fecal Lcn-2, and increased CD4+ (and PP CD8+) T cells in MLN/PP, consistent with low-grade intestinal inflammation. Anti-TNFα reduced fecal Lcn-2 and serum LPS and normalized colon length. - Microbiota dependence: GF TLR9−/− mice had restored trabecular bone mass comparable to GF WT, normalized CTX (P1NP remained higher), and equivalent in vitro osteoclastogenesis. Systemic cytokines (TNFα, IL1β, IFNγ) and spleen size were markedly reduced to WT-like levels; T cell activation differences were minimized. Cohousing normalized microbiota composition and eliminated differences in bone density, CTX, spleen/colon size, fecal Lcn-2, LPS, and most cytokines and T cell populations. - Myeloid-biased hematopoiesis: scRNA-seq of BMNCs revealed fewer B cells and increased myeloid lineages (monocytes, neutrophils), T cells, and HSPCs in TLR9−/−. Pseudotime showed accelerated lineage commitment and faster monocyte differentiation in KO. TLR9−/− HSPCs upregulated myeloid genes (Lyz2, Mpo, Fos, Elane, Cd9) and downregulated lymphoid genes (Sox4, Satb1, Vpreb3, Myl10); regulon activity of Cebpb, Fos, Jun, Egr1, Hlf increased, while Pax5, Ebf1, Klf2 decreased; Cebpd (inflammation-associated) activity was elevated. HSPC subclusters (HSCs, CMPs, MDPs, cMoPs) were KO-enriched; markers indicated myeloid priming. - Flow cytometry corroborated increased CMP, GMP, cMoP, CD11b+ myeloid cells, and OCPs (B220−CD11b+CD115+; B220−CD11b+Ly6Chi) in TLR9−/− BM/spleen. LSK assays showed skewing toward myeloid over lymphoid differentiation in TLR9−/−; after 5-FU, KO mice recovered with more myeloid and fewer B220+ cells. In KOchim mice, BM myeloid progenitors and OCPs were increased, with fewer B cells, indicating transferability. Anti-TNFα normalized KO myeloid/OCP expansion; GF and cohousing abrogated differences in CD11b+ cells, OCPs, CMP/GMP, and macrophages. - Monocyte/OCP molecular program: scRNA-seq subclustering showed that immature monocyte-like OCP populations had higher Fos and Spi1 (PU.1) in TLR9−/−, with increased RANK+ cells; regulon activities of Cebpb, Fos, Spi1, and Cebpd were elevated; KEGG pointed to lysosome/antigen presentation and osteoclast differentiation pathways. - T cell programs: Bone marrow T cells in TLR9−/− had increased Tcf7/Lef1 activity (memory T cell programs) and reduced Foxp3 (Treg). Ligand–receptor analysis suggested increased CD40–CD40L signaling between T and B cells, potentially enhancing B cell RANKL production; CCR7 upregulation in CD4 central memory T cells may facilitate bone marrow homing.

The data demonstrate that TLR9 maintains bone homeostasis under physiological conditions by constraining gut microbiota-driven systemic inflammation. In TLR9−/− mice, dysbiosis (notably enrichment of gram-negative taxa) elevates LPS and provokes low-grade intestinal and systemic inflammation. This inflammation expands and activates T cells, elevating osteoclastogenic cytokines (TNFα, IL1β, IFNγ, RANKL) and reducing OPG, thereby directly stimulating osteoclast differentiation and activity. Concurrently, chronic inflammatory signals reprogram hematopoiesis, biasing HSPCs toward myelopoiesis and expanding OCP pools, as evidenced by scRNA-seq transcriptional programs (myeloid regulons including Cebpb/Fos/Spi1 and inflammatory Cebpd) and increased myeloid progenitors and OCPs by flow cytometry. The dual impact—cytokine-driven differentiation and increased precursor availability—amplifies osteoclastogenesis and bone resorption. Causality is supported by adoptive transfer and BM chimera experiments, reversal with anti-TNFα, normalization in germ-free and cohoused conditions, and in vitro cocultures implicating CD4+ T cells. These findings integrate microbiota, immunity, hematopoiesis, and skeletal remodeling, highlighting TLR9 signaling as a critical node in osteoimmunology and suggesting that targeting inflammatory circuits or microbiota may mitigate inflammatory bone loss.

This study establishes that TLR9 deficiency causes trabecular bone loss by fostering gut microbiota–driven systemic chronic inflammation that both elevates osteoclastogenic cytokines and induces myeloid-biased hematopoiesis, expanding osteoclast precursor pools. scRNA-seq reveals inflammatory priming and myeloid transcriptional programs in HSPCs and monocyte-lineage cells, while in vivo interventions (anti-TNFα, germ-free rederivation, cohousing) validate the central roles of TNFα and the microbiota. The work identifies TLR9 signaling as a physiological regulator linking microbiota, immune homeostasis, hematopoiesis, and bone remodeling. Future directions include defining specific microbial taxa and metabolites driving inflammation and myelopoiesis, dissecting cell-intrinsic roles of TLR9 across immune compartments, and translating these insights to human inflammatory bone loss and therapeutic modulation of microbiota–immune axes.

- The study relies on mouse models; translational relevance to human bone remodeling and inflammatory bone diseases requires validation. - Global TLR9 knockout affects multiple cell types; while adoptive transfer and chimeras implicate hematopoietic compartments, cell type–specific contributions (e.g., intestinal epithelium, B cells, myeloid subsets) were not fully dissected genetically. - Microbiota analyses used 16S rRNA profiling; causal roles of specific taxa (e.g., Mucispirillum schaedleri, Parabacteroides distasonis) were not tested with gnotobiotic colonization or fecal microbiota transfer. - Anti-TNFα therapy improved bone and inflammation, but contributions of other elevated cytokines (IL1β, IFNγ, RANKL) remain incompletely separated. - Elevated P1NP in GF TLR9−/− mice indicates bone formation changes whose mechanisms were not explored in depth.