High Expression of Decoy Receptor TRAIL-R3 on AML Blasts Is Associated to Shortened Overall Survival. TRAIL-R3 Induced Apoptosis Resistance Can Be Overcome by Specifically Targeting TRAIL-R2 in Vitro.

In vivo, effector T cells and NK cells do express membrane-bound TRAIL (TNF-related apoptosis-inducing ligand). T cells also secrete a soluble form of TRAIL (sTRAIL). Binding of TRAIL to one of the receptors on target cells can induce apoptosis. In human, 4 membrane-bound receptors for TRAIL have been identified: two of them (TRAIL-R1 and TRAIL-R2) contain a functional death domain, while the two other (TRAIL-R3  and TRAIL-R4) lack a functional death domain and function as decoy receptors. Most normal cells express TRAIL-R3 and TRAIL-R4, where many tumor cells express TRAIL-R1 and TRAIL-R2, making activation of the TRAIL mediated apoptotic pathway a possible new model for targeted therapy.
We investigated bone-marrow samples from 119 patients with de novo AML and from 11 healthy donors for TRAIL receptor expression on CD45^dim/SSC^low blasts by flow-cytometry. Expression profiles of patients were correlated to survival data.
Functional data were obtained by incubation of 6 myeloid leukemic cell lines and fresh AML samples with different concentrations of sTRAIL/Apo2L (n=10) and specific antibodies directly targeting R1 and R2 (HGS-ETR1 and HGS-ETR2) (n=18). Apoptosis was assessed by flow–cytometry after staining with 7AAD and annexinV. R3 expression on the cell surface of one cell line (MM6)  was down-modulated with PI-PLC (Phosphatidyl-inositol phospholipase C ).When comparing AML and normal blasts, significantly more pro-apoptotic TRAIL-R1 and TRAIL-R2 receptor expression on myeloid leukemic blasts (p<0.0005) was measured however without association with outcome. About 30% of AML patients had a high anti-apoptotic TRAIL-R3 expression, which was strongly correlated to a shortened overall survival (log rank, p=0.0051). In multivariate analysis, R3 expression remained a significant prognostic factor next to cytogenetics (p=0.03 and p= 0.015 respectively).
In vitro, studies on 6 myeloid leukemic cell lines confirmed sTRAIL sensitivity in cell lines that expressed presumably TRAIL-R1 and TRAIL-R2. Moreover, down-modulation of TRAIL-R3 by treatment with PI-PLC (TRAIL-R3 is the only GPI-linked TRAIL receptor) on cell lines improved sTRAIL sensitivity.
Fresh myeloid leukemic samples were treated with sTRAIL. Mean apoptosis induction was 14% (0- 54%). sTRAIL can bind to both the pro-apoptotic as the decoy receptors. Bypassing of the decoy R3 expression by treatment with the antibody directly targeting TRAIL-R2 (anti-TRAIL-R2) resulted in much higher rates of apoptosis (mean 30%, 0-80%). Apoptosis induction via direct targeting of TRAIL-R2 was, as expected, independent of TRAIL-R3 expression. All leukemic blasts that showed apoptosis with anti-TRAIL-R2 expressed the R2 receptor, but not all blasts that expressed TRAIL-R2 were sensitive for anti-TRAIL-R2, which could be due to inhibitory factors downstream the apoptotic cascade.  In conclusion, AML blasts do express pro-apoptotic TRAIL receptors. However, co-expression with the decoy receptor TRAIL-R3 results in significant shortened overall survival. AML blasts are sensitive to targeting the pro-apoptotic TRAIL-R2 receptor, yielding a new therapeutic option for AML patients.