Feedback mechanisms limit immune activation to prevent tissue damage. This study maps conserved, cell type-specific regulation of TNFAIP3/A20 and uses conditional mouse models to identify drivers of inflammation when A20 function is impaired. The central question tested whether adaptive immunity or the microbiome are required for systemic inflammation caused by loss of Tnfaip3 in CD11c-expressing cells.
A20, encoded by TNFAIP3, is an NF-κB–inducible regulator that terminates sustained NF-κB signaling and limits cell death via ubiquitin-binding domains. It is induced downstream of receptors such as TNFR, IL-1R, B and T cell receptors, CD40, RANK, TLRs, and NLRs. Full-body A20 knockout mice develop severe, early-onset cachexia and multi-organ inflammation, demonstrating the gene’s central role in immune homeostasis. Conditional deletions have revealed cell type–specific outcomes: myeloid A20 loss causes polyarthritis, DC-specific loss leads to multi-organ inflammation with lymphosplenomegaly and myeloid expansion, and B cell A20 deficiency alters differentiation and predisposes to autoantibodies.
The authors established a data-driven framework describing TNFAIP3 expression and induction across humans and mice using single-cell atlases (Tabula Sapiens, Tabula Muris) and integrated datasets. They evaluated transcription factor regulatory potential and cytokine effects using curated resources. Cytokines were filtered to identify those reproducibly altering TNFAIP3/Tnfaip3, and in vivo cytokine responses were assessed using the Immune Dictionary single-cell resource. These analyses demonstrated strong conservation of cell type–specific expression and inducibility between species, supporting translational relevance of mouse models.
Conditional Tnfaip3 floxed mice (exons 4–5 floxed) were crossed with CD11c-Cre (Itgax-Cre) mice to generate A20CD11c-ko animals. To examine dependence on humoral and cellular adaptive immunity, these mice were further crossed onto Fcgrt-deficient, Ighm-deficient (µMt), and RAG2-deficient backgrounds. Targeting by Itgax-Cre was validated using Rosa26 TdTomato reporter crosses. All experiments used C57Bl/6 background mice maintained under specific-pathogen-free conditions; germ-free animals were generated by embryo transfer into axenic recipients for microbiome studies.
Because autoantibodies are frequently detected in patients with A20 haploinsufficiency (HA20) and in some A20-deficient mouse models, the investigators assessed whether B cells, immunoglobulins, or lymphocytes drive systemic inflammation in A20CD11c-ko mice. They measured serum immunoglobulins and anti-dsDNA or anti-cardiolipin antibodies by ELISA, and used genetic ablation models (µMt, Fcgrt–/–, RAG2–/–) to remove B cells, IgG recycling, or all lymphocytes, respectively. Despite elimination of these adaptive immune components, systemic inflammatory phenotypes persisted in the A20CD11c-ko genetic background, indicating that autoantibodies and adaptive lymphocytes were not required for disease onset in these models.
Given prior links between the microbiome and immune dysregulation, the authors rederived A20CD11c-ko mice as germ-free by embryo transfer and maintained them in axenic isolators. Disease manifestations occurred in the absence of microbial colonization, demonstrating that the microbiome was dispensable for the systemic inflammation observed in this model. These results support a sterile, innate-driven pathogenesis when A20 function is lost in CD11c-expressing cells.
CITE-seq was performed on splenocytes from A20CD11c-ko mice and littermate controls (three mice per group, female, 12 weeks old). Samples were prepared with magnetic depletion of neutrophils, B and T cells for portions of the input and stained with a panel of TotalSeq-A oligo-conjugated antibodies and hashing reagents. Cells were processed with the 10x Genomics Chromium platform; read mapping, barcode processing, and feature counting used Cell Ranger with downstream analyses in Seurat. Quality control steps removed extreme cells and doublets. A weighted nearest neighbor graph combining RNA and ADT modalities was used for clustering and annotation. Differential gene and protein expression and pathway analyses were computed to characterize cellular responses to A20 loss.
Flow cytometry panels and gating strategies were detailed, with enzymatic and mechanical dissociation protocols for spleen, bone marrow, Peyer’s patches, and colonic lamina propria. Cell counting beads and Fc block were used; samples were acquired on BD LSRFortessa or FACSymphony instruments and analyzed in FlowJo. ELISAs quantified serum IgA, IgM, IgG, IL-6, TNF, BAFF, and autoantibodies (anti-dsDNA, anti-cardiolipin) per manufacturer instructions. These assays supported the conclusion that humoral markers can be present but may represent downstream consequences of sterile innate inflammation rather than primary drivers.
Using complementary genetic, gnotobiotic, single-cell, and immunophenotyping approaches, the study demonstrates that loss of Tnfaip3 in CD11c-expressing cells triggers systemic, sterile hyperinflammation that does not require B cells, T cells, immunoglobulins, or commensal microbes. The conservation of TNFAIP3 regulation across species supports translational relevance. The data suggest that autoantibodies observed in TNFAIP3-deficient settings likely arise secondary to primary innate immune dysregulation.
These results favor an autoinflammatory model for many TNFAIP3-associated disorders and imply that therapies targeting innate inflammatory circuits or the specific cell types driving inflammation may be more effective than approaches focused exclusively on adaptive immunity or microbiome modification. The authors note the therapeutic ramifications for treating patients with TNFAIP3 loss-of-function or haploinsufficiency.
All animal procedures were approved by the VIB Center for Inflammation Research animal ethical committee. Mouse strains, sequencing details, antibody panels, and supplementary files are provided in the original article. The GWAS catalog was interrogated to summarize TNFAIP3-associated SNPs. Data availability, funding, acknowledgments, conflict of interest, and ethical statements are reported in the source article.