Investigating the Effect of Hsp90 Inhibitors on Initiating Cells in Childhood Acute Lymphoblastic Leukaemia

Abstract 2952

Enhanced risk stratification protocols and intensified therapies have improved outcomes and reduced the risk of relapse in childhood acute lymphoblastic leukaemia (ALL). Nevertheless, 20% of patients relapse, due to failure to eradicate the disease. Current chemotherapeutic regimens are designed around the properties of the bulk leukaemia cells, which might differ from those of the leukaemia initiating cell populations (LIC). Therefore, if drugs have no effect on LIC, these cells may expand and eventually cause relapse. Since several populations in childhood ALL have been shown to have LIC properties, developing therapies that are effective against all LIC populations may prevent relapse. We have previously shown that the NF-κB inhibitor parthenolide (PTL) could prevent proliferation and engraftment of unsorted cells and all LIC populations in NSG mice, in most cases examined. Heat shock protein (Hsp) 90 inhibitors are promising targets in cancer therapy. Targeting this protein has a combined impact on many oncogenic pathways involved in malignant progression. In order to investigate whether the ablation of LIC that we observed using PTL could be improved, we examined the effects of two Hsp90 inhibitors on childhood ALL cells in this study. 17-DMAG is used in preclinical and clinical phase I and II testing in breast cancers. It targets the binding site of ATP in Hsp90 and has low toxicity and high oral bioavailability. Celastrol is a novel Hsp90 inhibitor, which has recently been shown to eradicate LIC in AML by inhibiting NF-κB survival signals and inducing oxidative stress. Dose-response effects using 0.01–100nM of each Hsp90 inhibitor were assessed in primary T- and B-ALL cases and on normal haemopoietic cells at 24, 48 and 72h. Cells were stained with Annexin V and PI then viability and apoptosis were assessed by flow cytometry. An IC₅₀ was observed in leukaemia samples using 1nM of both 17-DMAG and Celastrol after 72h. Increasing the dose of 17-DMAG to 100nM and reducing the incubation time to 48 hours for each drug further reduced ALL cell survival, without impacting normal cells. At these doses, 17-DMAG reduced the viability of T-ALL cells to 36±30% and B-ALL cells to 32±13%. In T-ALL cases, 43±15% survived treatment with Celastrol and similar results were observed with B-ALL cells (43±16%). Normal haemopoietic cells were relatively unaffected at these doses (17-DMAG: 81±2% and Celastrol: 86±36%). Subsequently, T-ALL cells were sorted for expression of CD34 and CD7 and B-ALL cells were sorted for CD34 and CD19. The subpopulations were treated with 1nM of each inhibitor and the results compared to those observed using untreated controls. However, at these concentrations the drugs had limited effects on the ALL subpopulations; 31–100% and 28–89% T-ALL subpopulations survived treatment with 17-DMAG and Celastrol respectively. Similar results were observed with B-ALL subpopulations, 9–86% survived treatment with 17-DMAG and 62–100% survived following Celastrol treatment. A number of studies have shown that a regulatory loop may exist between Hsp90 and NF-κB in that downregulation of NF-κB could lead to reduction in Hsp90 protein levels. Therefore, we investigated whether using the Hsp90 inhibitors in combination with a NF-κB inhibitor would be more effective. Samples were treated with 100nM 17-DMAG or 1nM Celastrol for 48h and 7.5mM PTL was added for the last 24 hours. In 3 T-ALL cases, PTL reduced the viability to 28±13%, 17-DMAG to 25±12% and Celastrol to 35±15%. When PTL was used in combination, cell survival was further reduced to only 18±9% (PTL + 17-DMAG) and to 19±10% (PTL + Celastrol). In 3 B-ALL cases, PTL treatment reduced viability to 57±9%, similar results were seen with 17-DMAG (59±6%), while Celastrol appeared to be the most effective of the 3 drugs (38±4%, P=0.02). When PTL was combined with the Hsp90 inhibitors the cell killing was increased by 2 fold compared to PTL or 17-DMAG alone (PTL + 17-DMAG: 31±6%, P=0.04 and PTL + Celastrol: 28±3% P=0.01). Results to date indicate a promising role for the use of Hsp90 inhibitors with PTL and data on the functional ability of treated LIC populations will provide further information on the potential of these drug combinations. In conclusion, these Hsp90 inhibitors were as effective as PTL against childhood leukaemia cells and when used in combination with PTL, cell survival was further reduced by up to 20%.