Expansion of Inhibitory NK Receptor (CD94/NKG2A)-Expressing CD8 T Cells from Various Blood Mononuclear Cells with Cytolytic Activity Against Primary Leukemic Cells.

Leukemic cells and tumor cells can be escaped from allogeneic recognition by usual cytotoxic T cells because of the low expression level of HLA class I molecules. It has recently been shown that inhibitory natural killer cell receptors (NKRs) on not only NK cells but also on T cells negatively regulate NK cell and T cell functions through their binding to MHC class I molecules. The C-type lectin superfamily inhibitory NKR (CD94/NKG2A) heterodimer recognizes an HLA-E that preferably bound to a peptide derived from the signal sequences of most HLA class I. Therefore, CD94 can monitor the global status of HLA class I on the tumor and leukemic cells and induce cytolytic attack without inhibitory signal against HLA class I decreased target cells. In this study, we expanded CD94-expressing T cells from four different sources of blood mononuclear cells (BMCs) and then investigated their cytolytic characteristics against patients’ primary leukemic cells in order to develop a potential strategy of cell therapy for hematological malignancy.

We could get more than 100 fold expansion of CD94-expressing CD8 T cells from normal donor PBMC, apheresed PBMC without G-CSF mobilization from normal donor, apheresed PBMC with G-CSF mobilization from patients after chemotherapy and cord blood after 7 days culture with immobilized anti-CD3 monoclonal antibody (1μg/mL) and IL-15 (5 ng/mL). Cytolytic activities of purified CD94-expressing cells using magnetic cell sorting (MACS) (CD94 > 90%) detected by 4 hours ⁵¹ Cr release assay against HLA class I intermediate primary leukemic cells (AML M0, M2, M4, CML CP, BC, MDS overt) (50 < mean fluorescence intensity (MFI) < 150) were 35.6 ± 12.8 % (n=21). However, CTL activities against HLA class I high primary leukemic cells (ATL, ALL, LBL)(MFI>150) were lower than 10 % (6.5 ± 4.2, n=5). Also, CTL activities against HLA class I very high PHA autoblasts and alloblasts (MFI>200) were lower than 5 % (4.0 ± 3.6, n=11). Although the cytolytic activity of CD94-expressing cells roughly depends on the expression of HLA class I molecules in inverse proportion, adhesion molecules and also activating molecules such as NKG2D on effector cells might be important for the regulation of the killing activity. In fact, anti-NKG2D mAb (50 μg/mL) suppressed the cytolytic activity of CD94-expressing cells against patients’ primary leukemic cells (% reduction of cytolytic activity, 22.5± 5.9, n=13). Furthermore, anti-LFA-1 mAb (20 μg/mL) suppressed the cytolytic activity of CD94-expressing cells much more effectively than did anti-NKG2D mAb(% reduction of cytolytic activity, 74.2±15.5, n=13, p<0.01). Our data indicated that the cytolytic activity of inhibitory NKR-expressing cells depends at least partially on NKG2D-activating NKR and also required adhesion through LFA-1. In this study, we were able to expand CD94-expressing CD8 T cells which have both inhibitory receptors (NKG2A) and stimulatoy receptors (NKG2D) as well as LFA-1 and ICAM-1 from four different sources of BMCs. Therefore, it may be possible to expand CD94-expressing cells from various sources of BMCs with cytolytic activity against both autologous and allogeneic primary leukemic cells for a new strategy of cell therapy.