Identification and Ex Vivo Expansion of a Circulating NK Cell Progenitor Population That Leads to Sustained Production of CD56+ NK Cells

Natural Killer (NK) cells are a component of the innate immune system that can lyse virally infected cells and tumor cells. Although IL-2 activated and ex vivo expanded NK cells are being developed for adoptive cell therapy of cancer, recent data suggests significant heterogeneity exists within these NK cell populations as defined phenotypically, functionally and developmentally. In order to maximize the efficacy of adoptive NK cell therapy, further insights into the ontogeny and behavior of various NK subsets are needed.

NK cells have been historically defined as lymphocytes expressing CD56 but lacking CD3. However, we now report the identification of a population of human peripheral blood mononuclear cells that lack expression of the canonical NK markers CD56 and CD16, but otherwise share phenotypic and functional properties with NK cells, which we term “double negative (DN)” killer cells. Besides CD56 and CD16, these cells lack other standard lymphocyte markers including CD3, CD14, CD19, CD20, and CD33, although the majority of these DN cells express surface markers considered to be NK-specific, including CD8α, CD94, CD224, NKp30, CD127, NKp46, NKG2D, and NKG2A.

We sorted CD3^-CD19^-CD14^-CD20^- human PBMNCs into CD16^-CD56^Bright ,  CD16⁺ CD56^dim, CD16⁺CD56^- and CD16-CD56-(DN) cellular subsets by flow cytometry, then expanded these subsets in vitro by co-culture with irradiated EBV-transformed LCL feeder cells and IL-2 (500IU/mL).  DN cells expanded 40-60 fold after 14 days in culture, with 55% of cells retaining the DN phenotype while 40% of cells acquired expression of CD56 but not CD16 (termed CD56^SP). Generation of CD56^SP cells from DN cells cultured in vitro suggests that circulating DN cells contain NK cell precursors. When DN cells were expanded for 14 days in culture and the residual remaining DN cells were then resorted and re-expanded for an additional 14 days, exponential cellular growth continued.  After this addition 14 days of culture, 10-40% of cells became CD56^SP NK cells, while the remainder remained DN. Repeated cycles of resorting residual DN cells demonstrated that the DN subset could continually regenerate additional DN cells as well as produce CD56^SP NK cells for at least 45 days, suggesting unique proliferative and differentiative properties of this cell population. 

CD56^SP NK cells generated in a 14 day cell culture from DN cells had up-regulation of a variety of cell surface receptors including CD30 (+57%), CD127 (+0.7%), CD161 (+69%), CD57 (+5.7%), CD8α (+40%), CD244 (+99%), NKp30 (+79%), NKp44 (+53%), CD25 (+3.5%), CD69 (+95%), and had higher intracellular IFN-γ (+35%), and perforin (+60%), suggesting both NK differentiation and education occurred in these progeny (figure).   In contrast, expression of inhibitory receptors in the DN population, such as CD94 (88%), CD158e (12%) and NKG2A (24%) showed minimal change in expression levels in their progeny DN cells or CD 56^SP NK cells at culture-day 14, consistent with prior studies showing inhibitory receptors are less environmentally influenced.

CD56^SP cells generated from the DN cells also had higher levels of bone marrow and lymph node homing receptors CXCR4 and CCR7 and possessed superior degranulation capacity against K562 cells compared to pre or post expanded DN or pre-expanded CD56^SP cells (figure). Further, gene expression analysis comparing DN vs. CD56^SP cells at day 0 and DN and CD56^SP cells generated from DN cells at day 14, showed DN cells had a gene expression profile with enhanced expression of genes involved in processes such as cell division and replication, while CD56^SP cells generated from DN cells had a more cytotoxic or “active phenotype”.

Conclusion: CD56^-16^- DN cells share many characteristics with canonical NK cells, have long-term in vitro differentiation and proliferation capacity, and may represent a cell population that is a self-renewing progenitor for highly activated and cytotoxic CD56^SP NK cells. Isolation and expansion of a progenitor population with long-term NK cell repopulating potential in adoptive NK cell immunotherapy trials could offer advantages over trials transferring more differentiated NK cells with limited in vivo proliferative capacity.