Understanding Dendritic Cell Diversity with Direct Cell Reprogramming

Abstract

Dendritic cells (DCs) are a heterogenous family of professional antigen presenting cells that specialize in the uptake, processing and presentation of antigens to T cells to promote downstream immune responses, therefore acting as orchestrators of immunity. Conventional type 1 DCs (cDC1) are cross-presenting cells, cDC type 2 (cDC2) cells govern type 1, 2 and 3 immunity, and plasmacytoid DCs (pDCs) are critical in type-I IFN (IFN-I) secretion for antiviral responses. However, the intricacy of DC diversity has become more apparent with recent studies highlighting the range and redundancy of some DC functions previously thought as restricted. In the context of cancer, the effectiveness of classical immunotherapy approaches, such as immune checkpoint blockade, has been associated with cDC1, but also with the presence of cDC2 and pDCs in the tumor microenvironment. Understanding how DC diversity is generated is crucial to predict and promote immune responses. However, the transcription factors (TFs) driving DC subset identity and functional divergence remain unclear.

By allowing the direct conversion of a somatic cell type into another, direct cell reprogramming offers not only a strategy to generate cells for regenerative medicine or immunotherapy, but also a platform to dissect the TF codes underlying heterogeneous cell lineages. The minimal network composed of PU.1, IRF8, and BATF3 was shown to be sufficient to induce cDC1s from fibroblasts, opening opportunities to understand DC diversity with reprogramming.

Here, I applied direct cell reprogramming to uncover the TF networks underlying the diversity of DCs, while also generating diverse DC subsets for cancer immunotherapy. Using an additive screening strategy, I identified two TF triads that induce cDC2 or pDC identities. PU.1, IRF4, and PRDM1 (PIP) induced cDC2s that acquired a pro-inflammatory cDC2B transcriptional program with an interferon gene signature and the ability to present antigens to T cells. SPIB, IRF8, and IKZF2 (SII) induced pDCs that displayed an immature pDC pro-inflammatory profile which could be primed to IFN-β secretion by the addition of IRF4. We further elucidated TF engagement at the onset of the reprogramming process, highlighting the need for cooperative action between an ETS and an IRF factor as the base program for DC induction, further aided by a third factor to establish subset-specific identities. The injection of all induced DCs into tumors promoted anti-tumor immunity in mice, solidifying their immunotherapeutic potential. In vivo reprogramming of two different tumor models revealed a differential response to cDC1 and cDC2 reprogramming, highlighting the specificity of each DC-reprogramming combination in driving tumor-protective responses. Additionally, I developed a combinatorial barcoded TF-based single-cell screening platform to further identify TF combinations for DC reprogramming. A set of 48 DC-specific barcoded TFs allowed the simultaneous induction of multiple DCs and other immune cell types and the construction of a TF hierarchy map to inform DC reprogramming.

Collectively, these findings contribute to better understanding TF dynamics in DC specification, heterogeneity, and function, paving the way for advancing precision cancer immunotherapies based on DC reprogramming. Additionally, I developed a multiplexed technology that opens opportunities to harness immune cell reprogramming for broader applications, ranging from cancer to autoimmunity and beyond.