Search

Individual elements within a superenhancer can act in a cooperative or temporal manner, but the underlying mechanisms remain obscure. We recently identified anIrf8superenhancer, within which different elements act at distinct stages of type 1 classical dendritic cell (cDC1) development. The +41-kbIrf8enhancer is required for pre-cDC1 specification, while the +32-kbIrf8enhancer acts to support subsequent cDC1 maturation. Here, we found that compound heterozygous Δ32/Δ41 mice, lacking the +32- and +41-kb enhancers on different chromosomes, show normal pre-cDC1 specification but, surprisingly, completely lack mature cDC1 development, suggestingcisdependence of the +32-kb enhancer on the +41-kb enhancer. Transcription of the +32-kbIrf8enhancer-associated long noncoding RNA (lncRNA) Gm39266 is also dependent on the +41-kb enhancer. However, cDC1 development in mice remained intact when Gm39266 transcripts were eliminated by CRISPR/Cas9-mediated deletion of lncRNA promoters and when transcription across the +32-kb enhancer was blocked by premature polyadenylation. We showed that chromatin accessibility and BATF3 binding at the +32-kb enhancer were dependent on a functional +41-kb enhancer located incis. Thus, the +41-kbIrf8enhancer controls the subsequent activation of the +32-kbIrf8enhancer in a manner that is independent of associated lncRNA transcription.

As a cell-based cancer vaccine, dendritic cells (DC), derived from peripheral blood monocytes or bone marrow (BM) treated with GM-CSF (GMDC), were initially thought to induce antitumor immunity by presenting tumor antigens directly to host T cells. Subsequent work revealed that GMDCs do not directly prime tumor-specific T cells, but must transfer their antigens to host DCs. This reduces their advantage over strictly antigen-based strategies proposed as cancer vaccines. Type 1 conventional DCs (cDC1) have been reported to be superior to GMDCs as a cancer vaccine, but whether they act by transferring antigens to host DCs is unknown. To test this, we compared antitumor responses induced by GMDCs and cDC1 in Irf8 +32–/– mice, which lack endogenous cDC1 and cannot reject immunogenic fibrosarcomas. Both GMDCs and cDC1 could cross-present cell-associated antigens to CD8+ T cells in vitro. However, injection of GMDCs into tumors in Irf8 +32–/– mice did not induce antitumor immunity, consistent with their reported dependence on host cDC1. In contrast, injection of cDC1s into tumors in Irf8 +32–/– mice resulted in their migration to tumor-draining lymph nodes, activation of tumor-specific CD8+ T cells, and rejection of the tumors. Tumor rejection did not require the in vitro loading of cDC1 with antigens, indicating that acquisition of antigens in vivo is sufficient to induce antitumor responses. Finally, cDC1 vaccination showed abscopal effects, with rejection of untreated tumors growing concurrently on the opposite flank. These results suggest that cDC1 may be a useful future avenue to explore for antitumor therapy.See related Spotlight by Hubert et al., p. 918

Supplementary Figure from cDC1 Vaccines Drive Tumor Rejection by Direct Presentation Independently of Host cDC1

The divergence of the common dendritic cell progenitor

The events that initiate autoimmune diabetes in nonobese diabetic (NOD) mice remain poorly understood. CD4 + and CD8 + T cells are both required to develop disease, but their relative roles in initiating disease are unclear. To test whether CD4 + T cell infiltration into islets requires damage to β cells induced by autoreactive CD8 + T cells, we inactivated Wdfy4 in nonobese diabetic (NOD) mice (NOD. Wdfy4 −/−- ) using CRISPR/Cas9 targeting to eliminate cross-presentation by type 1 conventional dendritic cells (cDC1s). Similar to C57BL/6 Wdfy4 −/− mice, cDC1 in NOD. Wdfy4 −/− mice are unable to cross-present cell-associated antigens to prime CD8 + T cells, while cDC1 from heterozygous NOD. Wdfy4 +/− mice cross-present normally. Further, NOD. Wdfy4 −/− mice fail to develop diabetes while heterozygous NOD. Wdfy4 +/− mice develop diabetes similarly to wild-type NOD mice. NOD. Wdfy4 −/− mice remain capable of processing and presenting major histocompatibility complex class II (MHC-II)-restricted autoantigens and can activate β cell-specific CD4 + T cells in lymph nodes. However, disease in these mice does not progress beyond peri-islet inflammation. These results indicate that the priming of autoreactive CD8 + T cells in NOD mice requires cross-presentation by cDC1. Further, autoreactive CD8 + T cells appear to be required not only to develop diabetes, but to recruit autoreactive CD4 + T cells into islets of NOD mice, perhaps in response to progressive β cell damage.

CD40 signaling in classical type 1 dendritic cells (cDC1s) is required for CD8 T cell-mediated tumor rejection, but the underlying mechanisms are incompletely understood. Here, we identified CD40-induced genes in cDC1s, including Cd70, Tnfsf9, Ptgs2 and Bcl2l1, and examined their contributions to anti-tumor immunity. cDC1-specific inactivation of CD70 and COX-2, and global CD27 inactivation, only partially impaired tumor rejection or tumor-specific CD8 T cell expansion. Loss of 4-1BB, alone or in Cd27(−/−) mice, did not further impair anti-tumor immunity. However, cDC1-specific CD40 inactivation reduced cDC1 mitochondrial transmembrane potential and increased caspase activation in tumor-draining lymph nodes, reducing migratory cDC1 numbers in vivo. Similar impairments occurred during in vitro antigen presentation by Cd40(−/−) cDC1s to CD8(+) T cells, which were reversed by re-expression of Bcl2l1. Thus, CD40 signaling in cDC1s not only induces costimulatory ligands for CD8(+) T cells but also induces Bcl2l1 that sustains cDC1 survival during priming of anti-tumor responses.

The events that initiate autoimmune diabetes in NOD mice remain poorly understood. CD4 and CD8 T cells are both required but whether either cell initiates disease is unclear. To test whether CD4 T cell infiltration into islet required damage to β cells induced by autoreactive CD8 T cells, we selectively inactivated Wdfy4 in NOD mice (NOD.Wdfy4-/-) using CRISPR/Cas9 targeting. Similar to C57BL/6 Wdfy4-/- mice NOD.Wdfy4-/- mice develop type 1 conventional dendritic cells (cDC1) that are unable to cross-present cell-associated antigens required to activate CD8 T cells. By contrast, cDC1 from heterozygous Wdfy4+/- mice can cross-present normally. Heterozygous NOD.Wdfy4+/- mice develop diabetes similar to NOD mice, but NOD.Wdfy4-/- mice neither develop diabetes nor prime autoreactive CD8 T cells in vivo. By contrast, NOD.Wdfy4-/- mice can process and present MHC-II-restricted autoantigens and can activate β cell specific CD4 T cells in lymph nodes, and yet do not develop CD4 T cell infiltration in islets. These results indicate that the priming of autoreactive CD8 T cells in NOD mice requires cross-presentation by cDC1. Further, autoreactive CD8 T cells are required not only to develop diabetes, but to recruit autoreactive CD4 T cells into islets of NOD mice, perhaps in response to progressive β cell damage.

CD40 signaling in classical type 1 dendritic cells (cDC1s) is required for CD8 T cell-mediated tumor rejection, but the underlying mechanisms are incompletely understood. Here, we identified CD40-induced genes in cDC1s, including Cd70, Tnfsf9, Ptgs2 and Bcl2l1, and examined their contributions to anti-tumor immunity. cDC1-specific inactivation of CD70 and COX-2, and global CD27 inactivation, only partially impaired tumor rejection or tumor-specific CD8 T cell expansion. Loss of 4-1BB, alone or in Cd27

As a cell-based cancer vaccine, dendritic cells (DC), derived from peripheral blood monocytes or bone marrow (BM) treated with GM-CSF (GMDC), were initially thought to induce antitumor immunity by presenting tumor antigens directly to host T cells. Subsequent work revealed that GMDCs do not directly prime tumor-specific T cells, but must transfer their antigens to host DCs. This reduces their advantage over strictly antigen-based strategies proposed as cancer vaccines. Type 1 conventional DCs (cDC1) have been reported to be superior to GMDCs as a cancer vaccine, but whether they act by transferring antigens to host DCs is unknown. To test this, we compared antitumor responses induced by GMDCs and cDC1 in Irf8 +32–/– mice, which lack endogenous cDC1 and cannot reject immunogenic fibrosarcomas. Both GMDCs and cDC1 could cross-present cell-associated antigens to CD8+ T cells in vitro. However, injection of GMDCs into tumors in Irf8 +32–/– mice did not induce antitumor immunity, consistent with their reported dependence on host cDC1. In contrast, injection of cDC1s into tumors in Irf8 +32–/– mice resulted in their migration to tumor-draining lymph nodes, activation of tumor-specific CD8+ T cells, and rejection of the tumors. Tumor rejection did not require the in vitro loading of cDC1 with antigens, indicating that acquisition of antigens in vivo is sufficient to induce antitumor responses. Finally, cDC1 vaccination showed abscopal effects, with rejection of untreated tumors growing concurrently on the opposite flank. These results suggest that cDC1 may be a useful future avenue to explore for antitumor therapy. See related Spotlight by Hubert et al., p. 918

During dendritic cell (DC) development, Myc expression in progenitors is replaced by Mycl in mature DCs, but when and how this transition occurs is unknown. We evaluated DC development using reporters for MYC, MYCL, and cell cycle proteins Geminin and CDT1 in wild-type and various mutant mice. For classical type 1 dendritic cells (cDC1s) and plasmacytoid DCs (pDCs), the transition occurred upon their initial specification from common dendritic cell progenitors (CDPs) or common lymphoid progenitors (CLPs), respectively. This transition required high levels of IRF8 and interaction with PU.1, suggesting the use of EICEs within Mycl enhancers. In pDCs, maximal MYCL induction also required the +41kb Irf8 enhancer that controls pDC IRF8 expression. IRF8 also contributed to repression of MYC. While MYC is expressed only in rapidly dividing DC progenitors, MYCL is most highly expressed in DCs that have exited the cell cycle. Thus, IRF8 levels coordinate the Myc-Mycl transition during DC development.

Efforts to determine the unique functions of the two types of classical dendritic cells, cDC1 and cDC2, are hindered by limited understanding of the divergence of the common dendritic cell progenitor (CDP). Some transcription factors act in commitment of already specified progenitors, such as Batf3 which stabilizes Irf8 autoactivation at the +32 kb Irf8 enhancer, but how other factors control CDP divergence remains unknown. Here, we report the transcriptional mechanism of CDP divergence and describe the first requirements for pre-cDC2 specification. Genetic epistasis analysis suggested that Nfil3 acts upstream of Id2, Batf3, and Zeb2 in cDC1 development but has not revealed its mechanism or targets. Analysis of newly generated NFIL3 reporter mice showed extremely transient NFIL3 expression during cDC1 specification. CUT&RUN and ChIP-seq analysis identified NFIL3 binding in the −165 kb Zeb2 enhancer at sites that also bind CCAAT-enhancer-binding proteins (C/EBPs). Mutation of these NFIL3/C/EBP sites by in vivo CRISPR/Cas9 targeting revealed functional redundancy, with C/EBPs and NFIL3 competing in binding these sites to support or repress Zeb2 expression, respectively. Mutation of these three NFIL3/C/EBP sites eliminated Zeb2 expression in myeloid, but not lymphoid progenitors, producing mice that completely lacked pre-cDC2 specification and mature cDC2 in vivo. These mice failed to make an appropriate TH2 response against H. polygyrus infection, consistent with cDC2 supporting TH2 responses to helminths. Notably, Zeb2 expression is not required to maintain cDC2 identity, but acts only to block cDC1 specification. Thus, the competition between C/EBPs and NFIL3 binding at the −165 kb Zeb2 enhancer determines CDP divergence.

T follicular helper cells (TFH) participate in germinal center (GC) development and are necessary for B cell production of high-affinity, isotype-switched antibodies. In a forward genetic screen, we identified a missense mutation in Prkd2, encoding the serine/threonine kinase protein kinase D2, which caused elevated titers of immunoglobulin E (IgE) in the serum. Subsequent analysis of serum antibodies in mice with a targeted null mutation of Prkd2 demonstrated polyclonal hypergammaglobulinemia of IgE, IgG1, and IgA isotypes, which was exacerbated by the T cell-dependent humoral response to immunization. GC formation and GC B cells were increased in Prkd2-/- spleens. These effects were the result of excessive cell-autonomous TFH development caused by unrestricted Bcl6 nuclear translocation in Prkd2-/- CD4+ T cells. Prkd2 directly binds to Bcl6, and Prkd2-dependent phosphorylation of Bcl6 is necessary to constrain Bcl6 to the cytoplasm, thereby limiting TFH development. In response to immunization, Bcl6 repressed Prkd2 expression in CD4+ T cells, thereby committing them to TFH development. Thus, Prkd2 and Bcl6 form a mutually inhibitory positive feedback loop that controls the stable transition from naive CD4+ T cells to TFH during the adaptive immune response.