The role of dendritic cells (DCs) in intestinal immune homeostasis remains incompletely defined. Here we show that mice lacking IRF8 transcription-factor-dependent DCs had reduced numbers of T cells in the small intestine (SI), but not large intestine (LI), including an almost complete absence of SI CD8αβ(+) and CD4(+)CD8αα(+) T cells; the latter requiring β8 integrin expression by migratory IRF8 dependent CD103(+)CD11b(-) DCs. SI homing receptor induction was impaired during T cell priming in mesenteric lymph nodes (MLN), which correlated with a reduction in aldehyde dehydrogenase activity by SI-derived MLN DCs, and inefficient T cell localization to the SI. These mice also lacked intestinal T helper 1 (Th1) cells, and failed to support Th1 cell differentiation in MLN and mount Th1 cell responses to Trichuris muris infection. Collectively these results highlight multiple non-redundant roles for IRF8 dependent DCs in the maintenance of intestinal T cell homeostasis.

T-helper 2 (Th2) cell responses defend against parasites. Although dendritic cells (DCs) are vital for the induction of T-cell responses, the DC subpopulations that induce Th2 cells in the intestine are unidentified. Here we show that intestinal Th2 responses against Trichuris muris worms and Schistosoma mansoni eggs do not develop in mice with IRF-4-deficient DCs (IRF-4f/f CD11c-cre). Adoptive transfer of conventional DCs, in particular CD11b-expressing DCs from the intestine, is sufficient to prime S. mansoni-specific Th2 responses. Surprisingly, transferred IRF-4-deficient DCs also effectively prime S. mansoni-specific Th2 responses. Egg antigens do not induce the expression of IRF-4-related genes. Instead, IRF-4f/f CD11c-cre mice have fewer CD11b+ migrating DCs and fewer DCs carrying parasite antigens to the lymph nodes. Furthermore, CD11b+CD103+ DCs induce Th2 responses in the small intestine, whereas CD11b+CD103− DCs perform this role in the colon, revealing a specific functional heterogeneity among intestinal DCs in inducing Th2 responses.

Low- and high-dose infections with the murine large intestinal nematode Trichuris muris are associated with induction of adaptive Th1 and Th2 responses, respectively, in mesenteric lymph nodes (MLN). Classical dendritic cells (cDC) accumulate in the large intestinal mucosa and MLN upon T. muris infection, yet their role in driving adaptive responses to infection remains largely unknown. We performed low- and high-dose T. muris infections of mice deficient in defined cDC subsets to investigate their role in induction of adaptive immune responses. Mice lacking IRF4-dependent cDC failed to clear a high-dose infection and displayed impaired Th2 responses. Conversely, mice lacking IRF8-dependent cDC cleared a low-dose infection and displayed an impaired Th1 response while increased production of Th2 cytokines. Finally, mice lacking both IRF4- and IRF8-dependent cDC were able to generate a Th2 response and clear a low-dose infection. Collectively, these results suggest that IRF4- and IRF8-dependent cDC act antagonistically during T. muris infection, and demonstrate that intestinal Th2 responses can be generated towards T. muris in the absence of IRF4-dependent cDC.

The pleiotropic TNF-α:TNFR axis plays a central role in the immune system. While the cellular expression of TNFR1 is broad, TNFR2 expression is mainly restricted to immune cells and especially high on T regs. The therapeutic potential of targeting TNFR2 for cancer treatment has been previously indicated but the mechanism-of-action (MoA) of these reagents remains unclear, with conflicting data reported by different investigators. To gain further insight, we identified and characterized a wide panel of human and mouse TNFR2-specific antibodies, generated from the n-CoDeR F.I.R.S.T™ phage display platform. Based on their ability to block TNF-α:TNFR2 binding and to agonize or antagonize TNFR2 signaling, we identified parallel human and mouse TNFR2-specific antibodies with activities ranging from agonist to antagonist, and from complete ligand blocking to non-blocking. Two antibody variants with distinctly opposing in vitro activities (complete ligand-blocking antagonist versus non-blocking agonist), were expressed in various IgG formats preferentially engaging activating FcγR (mIgG2a), inhibitory FcγR (mIgG1), or no FcγR (N297A Fc-mutated) and screened for in vivo antitumor activity. Both anti-TNFR2 antibody clones displayed anti-tumor efficacy but showed strikingly different FcγR-dependence for optimal antitumor activity. Further characterization demonstrated potent anti-tumor efficacy across several syngeneic in vivo cancer models (CT26, MC38 and B16), both as single agents, and when combined with anti-PD-1. In vivo mode-of-action studies indicated different initial events are evoked by the two antibodies but that they ultimately converge to elicit a similar immune modulation within the tumor that is associated with anti-tumor efficacy. The antagonist antibody caused early intra-tumoral T reg depletion, while the agonist caused dramatic CD8+ T cell increases. Over time, both antibodies induce an increase in effector T cells at the tumor site, improved CD8/T reg ratios, and tumor regression. In addition, the two antibodies similarly modulated the tumor myeloid content. Based on careful MoA-characterization, two human lead candidate anti-TNFR2 antibodies are being developed for treatment of solid cancer; BI-1808, a ligand-blocking T reg depleting antibody and BI-1910, a TNFR2 agonist. Citation Format: Linda Mårtensson, Kirstie Cleary, Monika Semmrich, Mathilda Kovacek, Petra Holmkvist, Caroline Svensson, Mimoza Demiri, Therese Blidberg, Ulla-Carin Thornberg, Vincentiu Pitic, Osman Dadas, Sean H. Lim, Stephen A. Beers, Mark S. Cragg, Björn Frendéus, Ingrid Teige. Targeting TNFR2 for cancer immunotherapy: Ligand blocking depletors versus receptor agonists [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 936.

BackgroundThe pleiotropic TNF-alpha:TNFR axis plays a central role in the immune system. While the cellular expression of TNFR1 is broad, TNFR2 expression is mainly restricted to immune cells. The therapeutic potential of targeting TNFR2 for cancer treatment has been previously indicated and to gain further insight, we characterized a wide panel antibodies, generated from the n-CoDeR F.I.R.S.T™ target and antibody discovery platform. We identified parallel human and mouse TNFR2 specific, complete ligand (TNF-alpha) blocking antibodies and could show potent anti-tumor activity in several immune-competent models, both as single agent and in combination with anti-PD1 using a BI-1808 murine surrogate. The mechanism-of-action was shown to be FcgR dependent and likely mediated through a combination of intra-tumor T reg depletion, CD8+ T cell expansion and modulation of tumor-associated myeloid cells. These findings were confirmed using BI-1808 in a humanized mouse model.MethodsTo address safety of the human lead-candidate BI-1808 two toxicological studies were performed in cynomolgus monkeys. The first study was a dose-range-finding study and the second a GLP study where three doses (2, 20 and 200 mg/kg) were given weekly for four consecutive weeks followed by a recovery period of eight weeks. In addition, cytokine release was further studied in T cell stimulation assays and in a humanized mouse model. Moreover, the BI-1808 murine surrogate was used to study the relationship between dose, receptor occupancy (RO) and efficacy in immune competent mouse cancer experimental models.ResultsFour weekly administrations of BI-1808 to cynomolgus monkeys were well tolerated at all doses, with no associated clinical signs, and no histopathological changes. Non-adverse and reversible increases in neutrophil counts and decreases in T cells were observed at all dose levels. No drug-related adverse events were observed and consequently the NOAEL for BI-1808 was determined to be 200 mg/kg. Pharmacokinetic studies demonstrated an expected half-life of two weeks at receptor saturation. There were no indications of cytokine release in any of the systems tested. Finally, we could show that to achieve max therapeutic effect, sustained RO was needed for approximately two weeks, covering the time it takes to generate a full adaptive Immune response.ConclusionsThere is a clear association between RO and therapeutic effect and BI-1808 is well tolerated at doses associated with high and sustained RO. Collectively, these studies were used to determine the starting dose in upcoming phase I/II study in solid cancer aiming for first-patient in during December 2020.Ethics ApprovalThe study on cynomolgous monkeys was conducted by Citox/Charles River Laboratories in compliance with animal health regulations, in particular: Council Directive No. 2010/63/EU of 22 September 2010 and French decree No. 2013-118 of 01 February 2013 on the protection of animals used for scientific purposes. Studies in mice were approved by the Swedish Animal Experiment Ethics Board, ethical permit/ethical license numbers 5.2.18-17196/2018 and 5.8.18-03333/2020