A novel mouse model based on intersectional genetics enables unambiguous in vivo discrimination between plasmacytoid and other dendritic cells and their comparative characterization

Plasmacytoid dendritic cells (pDC) were identified about 20 years ago, based on their unique ability to rapidly produce copious amounts of all subsets of type I and type III interferon (IFN-I/III) upon virus sensing, while being refractory to infection. Yet, the identity and physiological functions of pDC are still a matter of debate, in a large part due to their lack of specific expression of any single cell surface marker or gene that would allow to track them in tissues and to target them in vivo with high specificity and penetrance. Indeed, recent studies showed that previous methods that were used to identify or deplete pDC also targeted other cell types, including pDC-like cells and transitional DC (tDC) that were proposed to be responsible for all the antigen presentation ability previously attributed to steady state pDC. Hence, improving our understanding of the nature and in vivo choreography of pDC physiological functions requires the development of novel tools to unambiguously identify and track these cells, including in comparison to pDC-like cells and tDC. Here, we report successful generation of a pDC-reporter mouse model, by using an intersectional genetic strategy based on the unique co-expression of Siglech and Pacsin1 in pDC. This pDC-Tomato mouse strain allows specific ex vivo and in situ detection of pDC. Breeding them with Zbtb46GFP mice allowed side-by-side purification and transcriptional profiling by single cell RNA sequencing of bona fide pDC, pDC-like cells and tDC, in comparison to type 1 and 2 conventional DC (cDC1 and cDC2), both at steady state and during a viral infection, revealing diverging activation patterns of pDC-like cells and tDC. Finally, by breeding pDC-Tomato mice with Ifnb1EYFP mice, we determined the choreography of pDC recruitment to the micro-anatomical sites of viral replication in the spleen, with initially similar but later divergent behaviors of the pDC that engaged or not into IFN-I production. Our novel pDC-Tomato mouse model, and newly identified gene modules specific to combinations of DC types and activations states, will constitute valuable resources for a deeper understanding of the functional division of labor between DC types and its molecular regulation at homeostasis and during viral infections.