NF-κB and MCL-1 Are Important Determinants for the Effectiveness of Bortezomib In CD34⁺ AML Versus CD34⁻ AML Cells

Abstract 1420

AML patients still have an unfavorable outcome despite the treatment with intensive chemotherapy. New therapeutic options are explored for targeting the limited population of leukemic stem cells that is responsible for the sustained leukemic growth. In particular combination of agents are tested that trigger separate cellular cytotoxic pathways.

In leukemic cell lines it has been shown that the cytokine rhTrail (recombinant human Tumor necrosis factor- related apoptosis inducing ligand) is an effective cytotoxic agent and synergy in effectiveness has been observed with the proteasome inhibitor bortezomib. However, limited information is available about the efficacy of both agents in patient AML cells and more specifically in the CD34⁺ enriched AML stem cell fraction. Firstly, we demonstrated cytotoxic effects of Trail in the leukemic cell line HL60 and resistance in Oci-AML3 cells. We observed synergism in HL60 cells and sensitization towards Trail in Oci-AML3 cells upon bortezomib addition. To investigate the efficacy of both agents in primary AML cells, AML mononuclear cells (n=17) were treated with Trail (100ng/mL) and bortezomib (20nM) and cultured in short- and long-term cultures. Cytotoxic effects of bortezomib were observed in 65% of the patients, whereas Trail was not effective, neither in the absence nor presence of bortezomib despite the expression of the death receptors. To study whether bortezomib also affected the leukemic enriched stem cell fraction, AML mononuclear cells were sorted in CD34⁺ AML and CD34⁻ AML fractions and exposed to bortezomib (n=5). Our results demonstrated that CD34⁻ AML cell fractions were most prone for the effects of bortezomib (survival CD34⁻ versus CD34⁺ = 45% ± 18% versus 75% ± 18%, p = 0.03) which were confirmed with the long-term culture assay. These results could not be ascribed to bortezomib being less effective in non-dividing cells. HL60 cells were used as a model system and exposed to CFSE for making a distinction between dividing and non-dividing cells. CFSE bright and dim cells were equally sensitive to the effects of bortezomib.

Since NF-κB is an important downstream target of bortezomib, EMSA's were performed on CD34⁺ AML and CD34⁻ AML cells. Preliminary data indicates that the highest NF-κB activity is present in the CD34⁻ AML cells. These findings are in line with results of a micro-array study (de Jonge et al, Leukemia 2011) with a separate set of 49 AMLs sorted in CD34⁺ AML versus CD34⁻ AML cells. Cytokines and chemokines with NF-κB binding sites in their promoter region are significantly upregulated in CD34⁻ AML cells versus CD34⁺ AML cells.

The limited effectiveness of bortezomib in CD34⁺ AML cells might be related to an inappropriate downregulation of anti-apoptotic genes in part controlled by NF-κB. It has been shown that MCL-1 downregulation, mediated by bortezomib, has an essential role in the bortezomib-mediated apoptosis. Microarray analysis showed that MCL-1 is significantly elevated in CD34⁺ AML cells compared to normal CD34⁺ cells (p=0.0002). However, western blots demonstrated that these high levels of MCL-1 expression did not decrease in AMLs upon bortezomib exposure. To demonstrate whether downregulation of MCL-1 sensitize AML cells for the effects of bortezomib, Oci-AML3 cells were transduced with MCL-1 shRNA and displayed a 4-fold increased sensitivity to bortezomib (IC₅₀ = 2.7nM versus 10nM). Results in primary AML cells are in progress.

In conclusion, these results demonstrated that CD34⁻ AML cells are more prone for the effects of bortezomib than CD34⁺ AML cells likely due to higher NF-κB activity; the limited effectiveness in CD34⁺ AML cells might be overcome by modulation of the anti-apoptotic protein MCL-1.