The functional role of NK cells in leukemia control as well as their role against viruses have prompted numerous studies analyzing NK cell function in allogeneic haematopoietic stem cell transplantation (HSCT) in the recent years. However, most studies refer to unrelated donor, HLA- match or mismatch HSCT and little is known about matched sibling HSCT, which is a more common setting.
Here, we have studied the reconstitution of NK cell in 27 HLA-identical patients. Patients were sampled at early (<2months), intermediate (<4months) and late (>4 months) time points after engraftment of G-mobilized peripheral blood of geno identical siblings. Reduced intensity conditioning (fludarabin 4d+Busulfan 2d) was given to all patients. Post-graft immunosuppression consisted of cyclosporine (CSA). Percentage and numbers of NK cells was monitored together with the relevant markers including CD56, CD16, KIR, NKG2A, CD57 and NKG2C. Functional studies were performed by stimulation with the reference cell line K562 and when available the autologous leukemia. The functional assays evalauted both degranulation (CD107a+b) and cytokines (IFng, TNFa).
Similarly to previous studies, a large percentage of CD56bright NK cells (up to 90%, median 40% of NK cells) was observed during the first months, a percentage that tended to resume to normal after late time points (> 5 months). Remaining CD56dim NK cells mostly expressed NKG2A. CD57 expression was low at the first time points and increased with time to reach healthy controls by late time points.
Overall, these data showed that up to 6 months developing NK cells display a very immature phenotype with high CD56bright frequencies and high NKG2A expression with low frequencies of CD57. KIR expression was monitored by means of KIR2DL1/S1, KIR2DL2/S2/L3 and KIR3DL1 staining. Surprisingly, KIR expression was close to that of healthy controls, but we observed a higher frequencies of NK cells expressing no or only one KIR, which is consistent with an immature phenotype of NK cells. Next, we analyzed the functional capacities of developing NK cells. Degranulation after interaction with a classical, HLA-negative NK target cell, K562, was comparable to that of healthy control NK cells. Similarly, developing NK cells were fully capable to produce MIP1b after interaction with a target cell. By contrast, stimulation by K562 resulted in a lower production of pro-inflammatory cytokines such as IFNg and TNFa whereas chemokine production (MIPb) was conserved. These results are in line with other obtained in unrelated HSCT.
Finally, the normal cytotoxic potential of developing NK cells prompted us to check whether these cells had an anti-tumor potential. For this purpose, we tested the function of the patients NK cells post-transplantation against the aulogous or allogeneic blast (AML, CLL) or cell line (Follicular lymphoma). We analyzed the degranulation capacities of developing NK cells against leukemic cells of the same histology as the original pathology of the patients. A patients treated by HSCT for a follicular lymphoma was tested against RL, a follicular lymphoma cell line. Although, degranulation against K562 was normal (control), these NK cells failed to degranulate against RL, even in the presence of anti HLA blocking antibodies. Similarly, in transplanted patients for B-CLL, NK cell failed to recognize and degranulate against allogeneic primary leukemic B cells. Comparable to healthy controls, only the use of Rituximab allowed the recognition and killing of B-CLL cells. Finally, we tested the reactivity of developing NK cells against autologous AML blast, taken from the patients at diagnosis. Although some degree of reactivity could be observed in some patients, overall degranulation as well as cytokine production was low compared to reactivity against K562. Altogether these data showed that although reconstitution of NK cells after genoidentical allogeneic HSCT was rapid, functionality of NK cells was partially impaired as we observed a dichotomy between degranulation capacities and cytokine production potential. These in vitro data suggest that at the early times post transplantation the immunotherapeutic impact of HSCT does not only rely on the alloreactivity of NK cells from the donors in HLA-matched, sibling. Their role in combination with T cells as well as their anti-viral and GvH function deserve further studies.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.