Natural killer (NK) cells participate in innate and adaptive immune responses, and upon activation rapidly produce cytokines, chemokines, and growth factors, including IFNγ, TNFα, TGFβ, GM-CSF, MIP1α, MIP1β, IL-10, and others, which can affect the function of other hematopoietic cells. Considering the recent evidences that hematopoietic stem cells (HSCs) respond to cytokine signaling, we hypothesized that NK cell-mediated cytokine production could mediate HSC function.

By the use of co-cultures of purified Ly5.1 murine NK cells and congenic Ly5.2 HSCs, we concluded that NK activity affects HSC frequency in vitro as well as hematopoietic reconstitution in vivo. Sorted NK cells (CD3- NK1.1+) and HSCs (Lin-, Sca1+, ckithi, CD48-, CD150+) were co-cultured in the presence or absence of IL2 over an OP9 stromal cells layer for 14 to 28 days. After 14 days, the addition of NK cells to HSC cultures resulted in an approximate 2-fold reduction of lineage negative cells (Lin-) recovered cells, as compared to control HSC cultures, as determined by flow cytometry analysis. Lin- counts were even lower in HSC+NK long-term cultures when compared to HSC only cultures. Ly5.1 HSCs and/or Ly5.2 NK cells were injected into sublethally irradiated Ly5.1/2 chimeric mice in a ratio of 105 NK to 103 HSCs per mouse. The addition of IL2-stimulated NK to injected HSCs reduced engraftment from 15.7% to 1.82% when the 16 weeks bone marrow (BM) chimerism was analyzed. In agreement, donor CD45.1 cells contribution to the LSK and HSC subpopulations was reduced in the HSC+NK transplanted mice. To test whether NK depletion from BM grafts would affect HSC function, we performed limiting dilution transplantation assays where whole BM from Ly5.2 mice was submitted to immunonagnetic NK1.1 or IgG depletion and injected into lethally irradiated Ly5.1 animals. Donor chimerism after 8 and 16 weeks of transplant showed that depleting NK cells improves the engraftment ability of HSC in a cell dose-dependent manner. When 25 x104 BM cells were injected, chimerism increased from 40 to more than 90% in NK depleted group. Of note, HSC frequency was 1 in 1595 in the control and 1 in 95 in the NK depleted group.

In order to understand the mechanisms by which NK cells could regulate HSCs, we took advantage of a CCAAT/enhancer-binding protein gamma (C/ebpg) knockout (KO) conditional mouse model generated in our laboratory, considering that C/ebpg had been previously shown to regulate NK cytotoxicity. Using similar culture conditions, HSCs and NK cells isolated from control (CT) or Cebpg KO mice were injected into congenic sublethally irradiated recipients. Results showed that Cebpg-deficient NK cells do not harm HSC engraftment as CT NK cells do. For instance, after 8 weeks, the addition of CT non-stimulated and IL-2-stimulated NK cells to normal transplanted HSCs reduced the engraftment from 40% to 20% and 10%, respectively. In contrast, chimerism was not different when HSCs only or HSCs + stimulated KO NK cells were transplanted. Gene expression and cytokine profiles of deficient and normal NK cells revealed the potential players of this HSC-NK regulation. Of these, interferon gamma (IFNg), was lower produced by the C/ebpg deficient NK cells. Therefore, besides controlling NK cytotoxicity, we showed here that C/ebpg also plays a role in the regulation of HSCs by NK-mediated cytokine production.

Next, we investigated whether depletion of NK cells from human BM samples would improve transplantation efficiency. NK cells were removed using CD56 antibody and transplanted into sublethally irradiated NSG mice. Sixteen weeks after transplantation, animals were sacrificed and the percentage of human CD45 cells in blood, BM, and spleen demonstrated that NK depletion from human BM favors engraftment.

Altogether, these findings provide new insights to the knowledge of HSC regulation by NK cells, which are present in BM transplantation (BMT) grafts. Although the alloreactive effect of NK cells against non-identical tumor cells from BMT recipients is well known, its cytokine-mediated effects over identical progenitor cells from the graft were not previously explored. We show that NK-secreted cytokines harm stem cell function, thus suggesting that depletion of NK cells from BM donor cells preparations can improve stem cell engraftment, particularly in the setting of alternative transplants with limiting cell numbers or non-myeloablative conditioning regimens.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution