NKG2C⁺CD57⁺ Natural Killer Cells May Play an Important Role in Maintaining Virus-Specific Immune Reconstitution Post-Allogeneic HSCT

Background: Allogeneic hematopoietic stem cells transplantation (HSCT) is the only possible curative treatment for several malignant hematological diseases in adults. The delayed immune reconstitution after HSCT can allow human cytomegalovirus (CMV) to reactivate. This leads to prolonged hospitalization, increased morbidity and even mortality. Natural Killer (NK) cells have been described to undergo a persistent reconfiguration in response to CMV-reactivation. Here we analyzed the presence and expansion of adaptive-NK cells in patients after allogeneic HSCT.

Methods: A multicolor flow cytometry panel for monitoring the adaptive-NK cell (NKG2C⁺CD57⁺) reconstitution was established. Reconstitution of adaptive-NK cells was assessed in peripheral blood samples from 70 CMV-seropositive patients between day 0 and 100 post-HSCT at intervals of 7-10 days. Monitoring of CMV-reactivation was determined by CMV-pp65 antigenemia and reconstitution of CMV-specific T cells (CMV-CTLs) was performed routinely using 7 commercially available, certified CMV-tetramers. For further immunological tests, PBMCs from 7 patients were isolated by density gradient centrifugation. Total NK cells were negatively selected by magnetic bead separation. Additional purification of adaptive-NK cellswas achieved by cell sorting. Selected adaptive-NK cells were expanded by co-culture with irradiated allogeneic PBMCs as feeder cells and the medium was supplemented with PHA, IL-2 and IL-15. Cytotoxicity assays were performed with expanded adaptive-NK cells co-cultured for 5h with CMV-infected human fibroblasts (HFF) or K562 cell line. Degranulation capacity was assessed by CD107a staining and the cytotoxic effect was assessed by lactate dehydrogenase (LDH) assay. Additionally, T cells capacity to migrate towards adaptive-NK cells co-cultured with or without CMV-infected HFF was evaluated using 5.0 µm polycarbonate membranes.

Results:

Our patient cohort consisted of 70 patients after allogeneic HSCT with a median age of 58 years (range: 19-73). Forty-six patients (65.7%) were transplanted for acute leukemia, 58 (82.9%) received reduced intensity conditioning (RIC) and 67 (95.7%) received anti-thymocyte-antibodies globulin (ATG). GvHD-prophylaxis was cyclosporine A (CsA) in combination with mycophenolate motefil (MMF) for 77.1% of the patients and 80% were transplanted from matched donors. Thirty-seven (52.9%) patients reactivated CMV (median age: 61 years, range 23-73; median day of reactivation: day +37 post-HSCT, range: 17-63).

A significant increase in the absolute cell counts of adaptive-NK cells was observed after CMV reactivation, when compared to patients who did not reactivate CMV (p<0.001). Degranulation capacity of in vitro expanded adaptive-NK cells (up to 3680 fold-expansion) was analysed by CD107a staining after co-culture with CMV-infected HFF at three different Effector : Target ratios (E:T) of 1:1; 1:2; 1:5. A statistically significant increase in CD107a expression in adaptive-NK cells was seen for all E:T ratios. In addition, this CD107a expression was not statistically different from the positive control. Furthermore, LDH colorimetric assay was used to supplement the information on cytotoxic capacity of adaptive-NK cells. A significant increase of LDH release was seen in CMV-infected HFF when compared to uninfected HFF (p<0.0001). When comparing adaptive-NK cells with CMV-CTLs, both cell populations show similar kinetics of expansion during CMV reactivation, with an increase in absolute counts. Surprisingly however, adaptive-NK cells continue to expand even after resolution of CMV-reactivation, while CMV-CTLs cell pool starts to retract (p<0.05). We then evaluated the capacity of adaptive-NK cells to recruit T cells. We observed a significantly increased fold difference in the migration of T cells when adaptive-NK cells were co-culture with CMV-infected HFF in comparison to T cell chemoattractant CCL21 (p<0.05).

Conclusion: Taken together, our results indicate that adaptive-NK cells can undergo a dynamic modulation in response to CMV-reactivation. This cell population is not only able to eliminate CMV-infected target cells but can also recruit T cells to the site of infection. It is possible that adaptive-NK cells may substitute for missing CMV-CTLs shortly after HSCT and may also maintain a memory phenotype of CMV-CTLs.