Skin-resident innate lymphoid cells converge on a pathogenic effector state

Abstract

Psoriasis pathology is driven by the type 3 cytokines IL-17 and Il-22, but little is understood about the dynamics that initiate alterations in tissue homeostasis. Here, we use mouse models, single-cell RNA-seq (scRNA-seq), computational inference and cell lineage mapping to show that psoriasis induction reconfigures the functionality of skin-resident ILCs to initiate disease. Tissue-resident ILCs amplified an initial IL-23 trigger and were sufficient, without circulatory ILCs, to drive pathology, indicating that ILC tissue remodeling initiates psoriasis. Skin ILCs expressed type 2 cytokines IL-5 and IL-13 in steady state, but were epigenetically poised to become ILC3-like cells. ScRNA-seq profiles of ILCs from psoriatic and naïve skin of wild type (WT) and Rag1 -/mice form a dense continuum, consistent with this model of fluid ILC states. We inferred biological "topics" underlying these states and their relative importance in each cell with a generative model of latent Dirichlet allocation, showing that ILCs from untreated skin span a spectrum of states, including a naïve/quiescent-like state and one expressing the Cd74 and Il13 but little Il5. Upon disease induction, this spectrum shifts, giving rise to a greater proportion of classical Il5-and Il13expressing "ILC2s" and a new, mixed ILC2/ILC3-like subset, expressing Il13, Il17, and Il22. Using these key topics, we related the cells through transitions, revealing a quiescence-ILC2-ILC3s state trajectory. We demonstrated this plasticity in vivo, combining an IL-5 fate mouse with IL-17A and IL-22 reporters, validating the transition of IL-5-producing ILC2s to IL-22-and IL-17A-producing cells during disease initiation. Thus, steady-state skin ILCs are actively repressed and cued for a plastic, type 2 response, which, upon induction, morphs into a type 3 response that drives psoriasis. This suggests a general model where specific immune activities are primed in healthy tissue, dynamically adapt to provocations, and left unchecked, drive pathological remodeling.

Type 2 innate immunity regulates hair follicle homeostasis to control Demodex pathosymbionts

“…The increase in blood ILC2s in type 2 immune-deficient mice was associated with elevations of serum IL-5 and IL-13, but also in serum IL-22 (Supplementary Fig. 4c), consistent with studies revealing transition of activated skin ILC2s to an inflammatory Type 2/3 phenotype in both mice and humans 16, 17 .…”

“…The ability of Rag-deficient mice to restrain Demodex infestation in association with activation and expansion of skin ILC2s, focally arrayed about the hair follicles, suggests a critical role for innate type 2 immunity in control of these prevalent symbionts. Expansion of skin ILC2s was accompanied by acquisition of an ILC2/3 phenotype marked by co-expression of IL-13 and IL-22, in agreement with natural trajectories of these resident immune cells during perturbation 17, 25 , and mimicking trajectories of skin microbiota-specific adaptive CD8 T cells 26 . As shown here, IL-13 attenuates proliferation of epithelia throughout the pilosebaceous unit, and perhaps better integrates with the effects of IL-22 on epithelial differentiation 27 to stabilize synchronous follicular development during infection that otherwise leads to aberrant follicles that allow overgrowth of the resident mites.…”

AMULET: a novel read count-based method for effective multiplet detection from single nucleus ATAC-seq data

“…3 a, b). To further study the library depth effects on AMULET’s performance, we applied it on snATAC-seq data with different library depth: (1) publicly available mouse CD45 + cells [ 16 ] and innate lymphoid cells (ILCs) data [ 17 ] and (2) down-sampled PBMC1 and PBMC2 samples at 5% increments starting at 50%. To quantify AMULET’s performance, we simulated multiplets in each sample by randomly selecting nuclei pairs and used them to calculate recall.…”

Single-Cell Transcriptional Heterogeneity of Lymphatic Endothelial Cells in Normal and Inflamed Murine Lymph Nodes

“…The imiquimod-driven inflammation model we employed mimics many features of psoriasis at the histological and transcriptional levels [17,18]. Many studies, based on both mouse models and psoriasis patient samples, have illustrated alterations in whole skin, keratinocytes, resident lymphoid cells, macrophages, and other cell types, while changes in the LECs lining the draining LNs have not been studied before [56][57][58]. It is worth noting that many LN LEC-subtype markers remained differentially expressed in their respective locations upon inflammation, indicating that the general distinction between LEC subsets is maintained.…”