Drug-Induced, NKG2D-Dependent Sensitization of ALL Cell Lines to NK Recognition.

Abstract 2024

Poster Board II-1

The combination of tumor-sensitizing drugs with NK cell infusion is beginning to emerge as a novel anti-tumor therapy. A growing body of in vitro studies show that drugs such as proteosome inhibitors, histone deacetylase inhibitors and thiazolidinediones are able to sensitize tumor cells but not their healthy counterparts to NK-mediated lysis. Drug induced NK-sensitization has shown promise in acute myeloid leukemias but no studies have yet proven this principle in acute lymphoblastic leukemia (ALL); a tumor phenotype reported to be relatively NK-resistant. The mechanisms underlying sensitization have not been fully identified but up regulation of ligands for TRAIL and the NK activating receptor NKG2D: MICA MICB and the UL16-binding proteins, may have a role. We set out to explore ALL susceptibility to NK cytotoxicity and whether this could be modulated by drug treatment. In contrast to published data, untreated ALL cell lines were positive for surface expression of MICB and ULBP2. Median fluorescence intensity ratios (mean ± SD; n = 6) for MICB detection on the cell lines 697, NALM-6, BV173 and SEM were: 3.2 ± 0.9; 3.8 ± 1.3; 4.0 ± 0.5; 2.5 ± 0.9, respectively and for ULBP-2: 2.3 ± 0.4; 55 ± 4.9; 2.9 ± 0.2; 1.8 ± 0.4, respectively. NALM-6 was also positive for ULBP1 (3.3 ± 0.6) while all were negative for MICA and ULBP3. Susceptibility of untreated ALL lines to NK mediated killing was assessed by chromium release assay using an IL-2 stimulated primary NK cell line. At effector to target ratio 40:1, specific release was 2.3% with cell line 697, 12% with NALM-6, 36% with BV173 and 63% with SEM. These results correlated with CD107a exposure in a degranulation assay using IL-2 stimulated peripheral blood lymphocytes: specific degranulation (% CD107a+ target with effector minus %CD107a+ effector alone) was 0.68% (697), 7.1% (NALM-6), 10% (BV173) and 17% (SEM). There was no correlation between baseline expression of NKG2DL and susceptibility to NK killing. Bortezomib, sodium valproate and troglitazone were added to cell cultures at sub-IC50 doses for 48 hours and compared with equimolar vehicle controls. Surface NKG2DL expression was measured by flow cytometry. On NALM-6 troglitazone treatment increased ULBP1 MFI by 2.0 ± 0.33 fold compared with vehicle control and increased percentage of ULBP1 positive cells by 39.6% (paired t-test: p=0.063). Sodium valproate increased MICA expression by 2.91 ± 1.18 fold and percentage of MICA positive cells by 12.3% (p=0.0382). On BV173, sodium valproate treatment increased ULBP2 MFI by 1.55 ± 0.07 fold and percentage of ULBP2 positive cells by 8.6% (p=0.04). There were no significant ligand changes after drug treatment on cell line 697. No NKG2DL changes were seen after Bortezomib treatment on any cell line. The functional significance of NKG2DL changes was assessed by CD107a degranulation assay. NALM-6 treated for 48 hours with drugs yielded the following fold increases in specific degranulation of NK cells compared to NALM-6 vehicle controls: 5.02 ± 5.98 for Bortezomib (mean ± SD), 2.4 ± 0.67 for Troglitazone and 1.44 ± 0.13 for Valproate. Levels of NK degranulation with 697 were very low (<5%) and drug treatment had no effect. Finally, we demonstrated that sensitization of NALM-6 was at least partly dependent on NKG2DL recognition, since blocking antibody to NKG2D reduced CD107a exposure by all three drugs. Compared with controls, blocking reduced CD107a expression by 59 ± 12% for Bortezomib, 47 ± 1.1% for Troglitazone and 48 ±11% for valproate-treated cells. This result was unexpected for Bortezomib as no changes in surface NKG2DL expression were detected after drug treatment. However, we are investigating the possibility that Bortezomib may down-regulate HLA class I expression, thus reducing inhibitory signaling upon NALM-6/NK interaction and unmasking an activation pathway that signals through NKG2D. In conclusion, we found basal levels of expression of NKG2DL on ALL cell lines. There was no correlation between NKG2DL expression and susceptibility to NK lysis, although this was to be expected given that a wealth of other activating and inhibitory receptors that contribute to NK activation. Bortezimib, valproate and troglitazone induced NKG2DL expression and sensitization to NK recognition in a cell line-specific manner. These drugs may therefore be useful to augment conventional chemotherapy or immunotherapeutic approaches to ALL.