Poster Board II-1003
Natural killer (NK) cells recognize malignant cells through the tumor-associated expression of NKG2D-ligands, including MIC A/B, which are known to be expressed on epithelial tumors, resulting in tumor cytotoxicity (Ayello/Cairo et al, BBMT, 2006). The expression of MIC A/B on these tumors can be induced by in vitro exposure of these cells to HDACi, specifically Romedepsin (RM) (Skov et al, Cancer Res, 2005). Glycogen synthase kinase-3 (GSK-3), a constitutively active serine-threonine kinase with numerous functions including regulation of cellular differentiation, stress and apoptosis, has also been shown to be an important regulatory enzyme in the expression of MIC A/B in response to RM (Doble et al, J cell Sci 2003; Frame et al, Biochem J 2001; Skov et al, Cancer Res, 2005).
We sought to determine of the expression of MIC A/B in response to RM in various leukemia and lymphoma cells (LL), its influence on NK cell mediated cytotoxicity and to investigate the role of the GSK-3 pathway in the regulation of expression of MIC A/B in response to RM.
LL cells (106/ml, RS 4:11 [MLL-ALL], REH [pre-B cell ALL], Jurkat [T-cell ALL], Toledo [DLBCL], Ramos [Burkitt's Lymphoma]) were exposed to RM (10 ng/mL) for 24 hours, followed by FACS staining with PE-conjugated anti-MIC A/B antibody to determine surface expression of MIC A/B. Peripheral blood NK cells (CD3-/56+) were isolated via magnetic separation followed by IL-2 activation (3000IU/ml, 18 hrs). LL cells exposed to RM (generously supplied by Gloucester Pharmaceuticals) were subjected to NK cell mediated cytotoxicity assays (using an europium assay) at effector:target (E:T) ratio of 10:1, as we had previously described (Ayello/Cairo et al, BBMT 2006). LL cells were also pre-treated for 1 hour with 100mM lithium chloride (LiCl), a potent inhibitor of GSK-3 activity (Davies et al, Biochem J, 2000), to determine the role of this regulatory enzyme in the RM mediated expression of MIC A/B in these LL cells. Finally, blocking studies were also performed with anti-NKG2D receptor blockers to determine the specific role of NKG2D signal transduction pathway in NK cell mediated cytotoxicity.
MIC A/B expression significantly increased in LL cells in response to RM ([RS4:11 0.2% vs 19.2%, p< 0.0001], [REH 0.2% vs 46%, p= 0.0003], [Jurkat 1.12% vs 44.7%, p< 0.0001], [Toledo 0.5% vs 15.8%, p=0.0001], [Ramos 0.57% vs 33.6%, p=0.0003]). In addition, the expression of MIC A/B in response to RM was inhibited when LL cells are pre-treated with LiCl (Jurkat [RM vs RM+LiCl] 85% vs 18%, p<0.0001; RS 4:11 [RM vs RM+LiCl] 82% vs 5%, p<0.0001; Ramos [RM vs RM+LiCl] 67% vs 35%, p<0.0001). Cytotoxicity assays revealed significant increases in-vitro cytotoxicty in RS 4:11, Ramos and REH cells at E:T ratio of 10:1. Mean specific release (MSR)(±SEM) was measured for four conditions in the assay: NK cells (A) vs NK cells with 10 ng/ml RM (C) vs IL-2 activated NK cells (B) IL-2 activated NK cells +10 ng/ml of RM (D). MSR for RS 4:11 10%±3.6% vs 108%±47% and 12%±4% vs 145%±45% p<0.05(A vs C, B vs D) for RS 4;11. MSR for REH was 13.2%±14.4% vs 50%±17.6% and 22%±18% vs 132%±29% p<0.01(A vs C, B vs D). MSR for Ramos was 7.6%±7.6% vs 79%±15.6% and 35%±13% vs 90%±22% p<0.05(A vs C, B vs D). NKG2D receptor-blocking experiments resulted in significant decrease in NK cell mediated cytotoxicity in REH (p<0.03) and Ramos cells (p<0.001).
Our data suggests that the surface expression of MICA/B in LL cells is significantly increased by RM leading to enhanced susceptibility for NKG2D-mediated cytotoxicity by NK cells. Furthermore, up-regulation of MICA/B in LL cells secondary to RM exposure is in part regulated by the GSK-3 signal transduction pathway.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.