To ‘feel comfortable in one’s own skin’ is an idiom referring to one’s confidence in interacting with others. However, when the skin is inflamed, as in atopic dermatitis or psoriasis, patients carry a sub-stantial burden leading to opposite effects. Current therapies target redundant, late-stage inflammatory events but not the disease drivers, leading to heterogeneous and insufficient efficacy. Understanding the proximal mechanisms of inflammation will stimulate the development of better therapies. Among the innate immune sensors for stress and microbes in keratinocytes, mutations in the NLRP1 and NLRP10 inflammasomes are linked to skin disorders. These molecules and the pro- and anti-inflammatory IL-1 family members they regulate are differentially expressed in the different layers of the epidermis. We hypothesize that inflammasome signaling in keratinocytes needs context-dependent and spatio-temporal control to avoid inflammation, which poses unique analytical and conceptual challenges. Therefore, to understand how inflammasome signaling in specific keratinocytes drives skin inflammation, 4D-SkINFLAM will i. optogenetically activate specific inflammasome components with spatio-temporal precision and perform a spatial analysis of transcriptomes and proteomes in neighboring cells. With loss-of-function approaches and pathway activity reporters, we will ii. define the ‘sensome’ and the activity of inflammasomes in different areas of the epidermis. Using mouse models, we will iii. evaluate how spatial inflammasome activity drives skin inflammation. Through iv. AI-driven deep visual proteomics combined with an analysis of inflammasome activity, we will discover spatial in-flammasome activation and its effects in inflammatory skin disorders. A precise understanding of spatio-temporal inflammasome signaling in the skin will be critical for se-lecting therapeutic targets acting as upstream drivers of prevalent diseases with high unmet needs.

The pool of innate immune effector cells is wired to rapidly respond to pathogens, whereas only few specificities within the naïve adaptive repertoire expand clonally, undergo epigenetic remodelling and differentiate into effector and memory cells. However, innate cells can differentiate upon pathogen encounter and remember past experiences as well, thereby challenging this strict dichotomy. In particular, us and others have shown that human memory Natural Killer (NK) cells with global epigenetic remodelling can be generated in response to specific signals during cytomegalovirus (CMV) infection. We have recently identified two major types of open chromatin domains in human memory NK (mNK) cells ex vivo: first, a shared signature featured by all mNK cells across CMV-seropositive donors (“public memory”); second, a diverse set of unique open chromatin regions associated with the drastic expansions of individual and stable NK cell clones (“private memory”). Based on this unexpected finding, we hypothesise that the shared and the unique clonal memory might provide mNK cells with increased fitness and high effector potential, but also enhance the risk of oncogenic mutations. The ultimate goal of this project is to identify the signals and molecular mechanisms driving acquisition, selection and maintenance of human NK cell public and private clonal memory. To this end, we will combine multiomic single cell assays and lineage tracing of human NK cells from healthy donors and patients ex vivo, or under various stimuli in vitro, with genome-wide CRISPR perturbation studies to directly link ex vivo features with functional read outs. Success of this project will not only lead to new insights into the key networks promoting persistence and effector functions of mNK cells, but also reveal promising novel targets for cellular anti-tumour therapies.