UNVEILING HUMAN NATURAL KILLER CELLS ONTOGENESIS IN HOMEOSTATIC CONDITIONS AND DURING THE IMMUNE RECONSTITUTION AFTER HAPLOIDENTICAL HEMATOPOIETIC STEM CELL TRANSPLANTATION

Natural Killer (NK) cells are potent members of the innate immune system, as they are endowed with the capacity of sensing and eliminating tumor-transformed and viral infected cells by integrating a plethora of inhibitory and activating signals coming from a repertoire of germline encoded receptors whose expression is finely tuned throughout NK cell differentiation and licensing. For ontological reasons and similarities in terms of transcription factors and secreted cytokines, NK cells have been grouped in the family of innate lymphoid cells (ILCs), a recently identified heterogeneous cell population now divided into cytotoxic, by means of NK cells, and helper ILCs, further subdivided in subclusters mimicking helper T cells. Despite advances in the classification of ILCs, their ontogenesis remains to date uncertain. Many notions come from murine models, but precise orthologous phenotypical markers often lack, as well some developmental intermediates, likely stimulus microenvironment driven. In fact, the generation of ILCs mostly occurs in the bone marrow niche, in contact with specific soluble and cellular components, but secondary lymphoid tissues and peripheral stimuli are involved in the most advanced phases. Additionally, ILCs are characterized by a huge plasticity, the capacity of transdifferentiating in other subsets NK cells included upon stimulations, further entangling the topic. h-ILC are mainly involved in homeostatic and metabolic processes but act as first line mediators in the tissues where they localize as a prompt source of pro-inflammatory stimuli. Recent evidence also reported ILCs contribution in different solid and hematologic malignancies, with both anti and pro-tumorigenic effects. Often, especially in the case of hematologic tumors an underlying aspect favoring tumor expansion is characterized by alterations in the ontogenetic pathways of both NK cells and h-ILCs, that result in the accumulation of undifferentiated, hypofunctional cells characterized by unconventional phenotypes, that are poorly investigated. Additionally, NK cells concur in granting of a favorable outcome after hematopoietic stem cell transplantation (h-HSCT) being the first lymphocytic population to complete immune reconstitution(IR), thus providing immune-compromised individuals with a certain degree of protection from opportunistic pathogens, still representing the major cause of morbidity and mortality of this potent curative procedure. On the contrary, a little is known regarding h-ILC IR kinetics. Given these premises, in this study we carried out a multi methodologic dissection of human NK cell ontogenesis to evaluate the contribution of rare subsets and h-ILCs. We started by FACS sorting and differentiating in vitro precursors, h-ILCs and NK cell subsets from the BM in presence of an autologous feeder layer of MSC that could help us to recapitulate the interactions occurring within the BM niche. To enlarge our data power and include also the stimuli occurring in the periphery, we performed a scRNA-seq experiment on 6 healthy subjects to characterize rare subsets from a transcriptomic point of view and to perform a pseudotime trajectory analysis that could infer differentiation trajectories based on transcriptional similarities. Last, a consistent part of this study was carried out on HSCT patients, a unique setting reported as a reliable model to study the ontogenesis of immune populations in vivo. This study was carried out by manual gating and applying a novel pipeline capable of processing FACS data as single cell ones. Altogether our data confirm the involvement of h-ILCs in the differentiation pathway from precursors to mature NK cells. These h-ILCs mostly belong to the h-ILC3/p subset, confirming their nature of precursors also for NK cells. additionally, we highlight cCD56bright NK cells as driver of NK cell differentiation both in vitro and in silico. This is true also for the process of IR after h-HSCT, where cCD56bright are accompanied by expanded unCD56dim that concur to the generation of more differentiated NK cell subset few months after h-HSCT. However, in silico data hints that HCs unCD56dim NK cells differ from the unCD56dim identified after transplantation, suggesting the heterogeneity of NK cells with similar CD56 and CD16 expression. This is valid also for the unCD56bright NK cells, seemingly compatible with a more differentiated nature as hinted by our in vitro differentiation but characterized by an early appearance during IR and transcriptional and phenotypic features that make them indistinguishable from cCD56bright or cCD56dim. Another interesting aspect that we observed regards the relationship between NK cell activation, differentiation and licensing. If cells in vitro are uncapable of generating terminally differentiated subsets, circulating NK cells display an activated profile both in silico and ex vivo related with the acquisition of a terminal differentiated, fully licensed phenotype, highlighting the interplay of other stimuli and immune mediators in the homeostatic differentiation of NK cells and during their IR after h-HSCT.