Discrete Modes of Migration Define Human NK Cell Development from Hematopoietic Stem Cells

Human innate lymphoid cells play a critical role in the control of infection and the shaping of the immune response. Furthermore, natural killer (NK) cells are the first reconstituting lymphocyte population after hematopoietic stem cell transplant and play a key role in shaping engraftment and moderating graft vs. leukemic effects. Relative to their adaptive immune counterparts, the molecular interactions that drive innate immune cell differentiation development are poorly understood. Human NK cells arise from bone marrow CD34⁺ hematopoietic stem cells (HSC) but undergo maturation in distal microenvironments, the best characterized of which is secondary lymphoid tissue. Human NK cells can also be derived in vitro from CD34⁺ HSC through co-culture with developmentally supportive stromal cells. Despite the requirement for direct stromal cell contact in NK cell maturation, the nature of these contacts is not known. We have previously shown that freshly isolated human NK cell developmental intermediates from in vitro culture, secondary lymphoid tissue or peripheral blood undergo spontaneous migration on the EL08.1D2 stromal cell line. We have further defined the NK cell 'developmental synapse', a site of productive and directional signaling in developmental intermediates arrested on stromal cells. Despite these interesting insights, the regulation of, and requirement for, NK cell migration during differentiation has not been defined.

To determine the modes of migration associated with human NK cell differentiation, we performed highly resolved continuous live cell imaging of NK cells developing in vitro from CD34⁺ HSC over 28 days of culture. These analyses were performed in parallel with flow cytometry to monitor the progression of developmental intermediates towards a functionally mature NK cell phenotype. We found that while CD34⁺ HSC had constrained motility, progressive NK cell differentiation was associated with more directed migratory phenotype. NK cell maturation was associated with increased track length and velocity, and decreased frequency of time of cells spent in arrest on stromal cells. We further defined the acquisition of complex, Levy-walk behavior as has previously been described for mature T cells imaged within tissue. The consistency of these behaviors between in vitro derived cells and fresh ex vivo NK cells isolated from healthy human donors confirms that in vitro differentiation effectively recapitulates migratory phenotypes in addition to functional and cell surface phenotypes associated with NK cell maturation.

Together, these data define distinct modes of migration that are acquired by human NK cells. We propose that this reflects a previously unappreciated requirement for cell migration that enables the functional maturation of human NK cells. Better defining the bidirectional signaling between developing NK cells and their microenvironment will enable a complete understanding of the signals that help shape the acquisition of key innate immune cell functions. This understanding is critical for better predicting and therapeutically enhancing lymphocyte reconstitution after transplant.