Synergistic Dual Inhibition of BCR-ABL1 and the Unfolded Protein Response Causes p38 MAPK-Mediated Cell Death and Sensitizes BCR-ABL1⁺ Acute Lymphoblastic Leukemia to Dexamethasone

Introduction

The IRE1-XBP1 branch of Unfolded Protein Response (UPR) was previously shown to be pivotal for the survival of BCR-ABL1⁺ Acute Lymphoblastic Leukemia (ALL) cells. In this study, we identified a link between the IRE1 and BCR-ABL1 pathways which we harnessed for the pre-clinical application of a novel IRE1 inhibitor, MKC-8866, in combination with Nilotinib. The underlying mechanism was investigated by analysis of the phosphoproteome of BCR-ABL1⁺ ALL cell lines SUP-B15 and TOM-1. Upon dual treatment, the main force driving the cells toward death was ascribed to the activation of p38 MAPK. Additionally, the set of proteins affected under dual therapy suggested a possible regulation of glucorticoid receptor (GR) signaling that could modify the response to dexamethasone.

Methods and Results

SUP-B15 and TOM-1 cell lines were treated with vehicle, MKC-8866 (MKC, 30 µM) and nilotinib (NL, 0.5 µM) as single agents or in combination. Viability assays were performed in order to verify the synergistic effect of the treatment administered, evaluated using the Bliss formula. The combination of NL with MKC showed a striking synergism in both SUP-B15 and TOM-1 (Bliss calculation. Assumed additive effects: SUP-B15 24.2±20.6%, TOM-1 40.9±18.4% versus experimental values: SUP-B15 68.7±12.7%, TOM-1 78.8±4.9%).

To validate the effect on a genetic level, pre-B cell derived from conditional Xbp1^+/fl mice were transduced with BCR-ABL1and with either inducible cre or empty vector. The ALL-like cells were then treated with NL 0.5 µM and/or 4-hydroxytamoxifen (4OHT) 1µM to induce deletion. In agreement with the synergistic effect observed with the human cell lines, NL was shown to be significantly more effective in the presence of heterozygous deletion of Xbp1 (viable cells after 24 hours. NL: 65.2±0.3%, 4OHT + NL: 6.87±1.2%).

To unravel the basis of this synergism, we performed a phosphoproteomic analysis of the two human cell lines treated for 16 hours with vehicle, MKC, NL and dual treatment, respectively. For this analysis, SUP-B15 and TOM-1 were lysed in 8M urea and digested using LysC/trypsin. The obtained peptides were labeled using the dimethyl isotope labeling method. Phosphopeptides were enriched using TiO₂ beads and analyzed by nanoLC-MS/MS (Orbitrap Elite). The raw data were analyzed using MaxQuant and the Andromeda search engine against the human Uniprot database. The resulting information was further interpreted using the Perseus software package as well as the String and PhosphoPath apps in Cytoscape. The analysis of the regulated phosphorylated proteins highlighted a p38 MAPK differential regulation, with MKC alone or in combination causing its activation, and, in contrast, NL causing inhibition. Western Blot analyses investigating the activation of p38 MAPK showed increased phosphorylation of its main downstream target, heat shock protein beta-1 (HSPB1), in the presence of MKC alone or together with NL. Pharmacological inhibition of p38 MAPK activity by adding p38 MAPK inhibitor BIRB-796 (BIRB, 10µM) was able to rescue the cells from the concerted action induced by MKC and NL.

Furthermore, the phosphoproteome pattern observed in the cells upon dual treatment suggested a regulation of crucial GR interactors. Therefore, we tested whether the studied combination would render the cells more responsive to the GR agonist dexamethasone (DEXA, 10 mg/L), cornerstone in ALL therapy. The viability assay revealed a striking enhancement of its cytotoxic effect. However, most of the effect was conveyed by the action of DEXA together with NL, with MKC adding only a small and non-significant accessory effect (percentages of viable cells compared to DMSO values after 24 hours treatment. SUP-B15: DEXA 83.9±4.2%; MKC + DEXA 81.6±5.1%; NL + DEXA 66.7±7.1% and MKC + NL + DEXA 59.8±7.3%. TOM-1: DEXA 78.2±8.4%; MKC + DEXA 74.1±5.1%; NL + DEXA 45.3±10.1% and MKC + NL + DEXA 45.2±8.1%).

Conclusion

By studying the phosphoproteome in BCR-ABL1⁺ ALL, we identified a possible underlying mechanism explaining the success of the combined action of NL and MKC-8866. The p38 MAPK axis was found responsible for the observed cytotoxic effect, elucidating important aspects of BCR-ABL1⁺ ALL biology. Overall, the successful treatment regime with NL and MKC in vitro, and its positive interaction with DEXA, provide a promising basis for further pre-clinical evaluation.