The immune system balances protective immunity against pathogens with tolerance to self and selected foreign antigens. Emerging evidence identifies erythropoietin (EPO) as a systemic immunoregulatory signal that links physiological stress to immune adaptation. Beyond its canonical role in erythropoiesis, EPO–EPO receptor (EPOR) signaling reprograms specialized myeloid cells, including macrophages and type 1 conventional dendritic cells (cDC1s), promoting efferocytosis, inflammatory restraint, and maturation-associated cDC1 fate. Through these mechanisms, EPO shapes antigen-specific T-cell responses across transplantation and cancer. We propose that the EPO–EPOR axis functions as a systemic regulator of immune adaptation through myeloid-cell programming.

Nature, Published online: 15 September 2026; doi:10.1038/s41586-026-11137-8 Author Correction: Erythropoietin receptor on cDC1s dictates immune tolerance

Nature, Published online: 15 September 2026; doi:10.1038/s41586-026-11137-8Author Correction: Erythropoietin receptor on cDC1s dictates immune tolerance

In the version of this article initially published, in Fig. 2e, the macrophage values for the day 28 +DT (15–28) group contained a copy-and-paste error; in Fig. 3i, the processed qPCR values for Scube3 in the EporΔXcr1 T–A group were inadvertently duplicated from the EporΔXcr1 UNT group; in Fig. 5k, day 15 values for the EporΔXcr1 + Iso group were inadvertently duplicated from day 13. In the Source Data for Fig. 3, the data corresponding to Fig. 3i were incorrectly labelled “Fig. 3k” instead of “Fig. 3i”; in the Source Data for Extended Data Fig. 2c, the unit “ng/ml” should have read “pg/ml”; in the Source Data for Extended Data Fig. 3, “p4E-BP4” should have read “p4E-BP1” in row 2, corresponding to Extended Data Fig. 3d, and row 11, corresponding to Extended Data Fig. 3e. In the Fig. 5k legend, the sample size for the “EporΔXcr1 + anti-PD-1” group was shown as n = 6 instead of n = 7. In the Methods subsection “qPCR, RNA extraction and cDNA synthesis”, “omental tissue” should have read “cDC1s”. Due to a production error, the Fig. 5i label “Gated on CD45+TCRβ+CD8+ T cells” should have read “Gated on CD45+TCRβ+CD4+ T cells”. The figures,...

The authors identify peripheral dendritic cell priming of T cells as a key driver of brain inflammation and neurodegeneration in a mouse model of tauopathy, providing new insights into immune mechanis...

In a mouse model of tauopathy, conventional type 1 dendritic cells (cDC1s) cross-presented brain-derived antigens in deep cervical lymph nodes, priming CD8+ T cells that accumulated in the brain and p...

cDC1 development depends upon a cDC1-specific super-enhancer +32-kb of the Irf8 gene14. cDC1s are absent in Irf8 +32−/− mice even under pro-inflammatory conditions, without affecting Irf8 function in other leukocytes14,15. We crossed TE4 mice with Irf8 +32−/− mice to generate cDC1-deficient mice on the TE4 background (TE4Δ+32). Four cohorts of mice were generated: E4WT (non-tau, cDC1 sufficient), E4Δ+32 (non-tau, cDC1 deficient), TE4WT (tau expressing, cDC1 sufficient) and TE4Δ+32 (tau expressing, cDC1 deficient; Extended Data Fig. 1a). We assessed cDC1 development in 9.5-month-old TE4 mice, which exhibited robust neuropathology, by sorting common dendritic cell progenitors (CDPs), pre-cDC1 progenitors (pre-DC1s) and pre-cDC2 progenitors (pre-DC2s) from bone marrow and culturing these cells with Flt3L ex vivo for 7 days (Extended Data Fig. 1b,c). CDPs derived from TE4WT mice differentiated into both cDC1s and cDC2s, whereas pre-DC1s and pre-DC2s exclusively generated cDC1s and cDC2s, respectively (Extended Data Fig. 1d). In contrast, TE4Δ+32 CDPs did not generate cDC1s. TE4Δ+32 CDPs still specified into pre-DC1s but failed to mature into cDC1s, whereas cDC2s remained intact (Extended Data Fig. 1d). As a result, cDC1s were absent in tissues including spleen, brain-draining deep cervical lymph nodes (dCLNs), dural meninges and brain parenchymal tissues in TE4Δ+32 mice (Extended Data...

Read the full paper: CLEC-1 promotes early IFN response in cDC1s and favors immune dysregulation during sepsis in mice

Mattiuz et al. assessed the role of dendritic cells in TLS formation and maintenance. Mature cDC1s were found to play essential roles, with TLS formation being dependent on maturation of cDC1s and their cross-presentation to T cells in the TDLN, followed by T cell recruitment to the tumor. Over time, maintenance of TLSs required the presence of cDC1s and cDC1 migration to CCR7 ligand-enriched stromal hubs, MHC-I and MHC-II antigen presentation to T cells, and CD40 signaling.

Zhang et al. investigated the role of the CARM1 in cDC1s and found that while its effect on cDC1s in healthy tissues was minimal, inactivation of Carm1 in tumor cDC1s enhanced their activation and cross-presentation of tumor antigens, resulting in increased CD8+ T cell-mediated antitumor efficacy. Carm1 expression was found to be downregulated by type 1 IFN and TNFα, and upregulated by TGFβ via Smad2/3. CARM1 altered chromatin expression, with its loss leading to enhanced chromatin access at regions associated with NF-κB and AP-1 binding, which supported enhanced activation and antigen presentation. When used in combination with a neoantigen vaccine, inhibition of Carm1 enhanced antitumor responses.

While activation of the UPR sensor IRE1 is typically associated with protein folding, cDC1s show basal IRE1 activity in the absence of ER stress. Bosteels et al. uncover that this is due to their uniq...