The Aurora kinase/β-catenin axis contributes to dexamethasone resistance in leukemia

Abstract

Glucocorticoids, such as dexamethasone and prednisolone, are widely used in cancer treatment. Different hematological malignancies respond differently to this treatment which, as could be expected, correlates with treatment outcome. In this study, we have used a glucocorticoid-induced gene signature to develop a deep learning model that can predict dexamethasone sensitivity. By combining gene expression data from cell lines and patients with acute lymphoblastic leukemia, we observed that the model is useful for the classification of patients. Predicted samples have been used to detect deregulated pathways that lead to dexamethasone resistance. Gene set enrichment analysis, peptide substrate-based kinase profiling assay, and western blot analysis identified Aurora kinase, S6K, p38, and β-catenin as key signaling proteins involved in dexamethasone resistance. Deep learning-enabled drug synergy prediction followed by in vitro drug synergy analysis identified kinase inhibitors against Aurora kinase, JAK, S6K, and mTOR that displayed synergy with dexamethasone. Combining pathway enrichment, kinase regulation, and kinase inhibition data, we propose that Aurora kinase or its several direct or indirect downstream kinase effectors such as mTOR, S6K, p38, and JAK may be involved in β-catenin stabilization through phosphorylation-dependent inactivation of GSK-3β. Collectively, our data suggest that activation of the Aurora kinase/β-catenin axis during dexamethasone treatment may contribute to cell survival signaling which is possibly maintained in patients who are resistant to dexamethasone.

Fig. 1. Glucocorticoid-induced regulation in gene expression.

SUP-B15 cells were treated with 1 µM dexamethasone, 2 µM prednisolone, or DMSO for 6 h. Global gene expression was measured by Affymetrix Human Gene 2.0 ST Array. a The number of downregulated genes in dexamethasone- and prednisolone-treated cells. b The number of upregulated genes in dexamethasone and prednisolone-treated cells. c The heatmap displaying clusters of upregulated genes in dexamethasone- and prednisolone-treated cells was generated by the heatmap.2 function of Gplots library in R. d Gene set enrichments in dexamethasone-treated cells were measured by GSEA (Hallmarks and Oncogenic signatures) and visualized by the pyplot.scatter function of Matplotlib. e Gene set enrichments in prednisolone-treated cells were analyzed and visualized as described above.

Fig. 2. Glucocorticoid-induced regulation of kinase activation.

SUP-B15 cells were treated with 1 µM dexamethasone or DMSO for 6 h before lysis. a To measure protein tyrosine kinase activity in dexamethasone-treated cells, lysates were applied in Pamgene peptide substrate-based tyrosine kinase array and analyzed for tyrosine kinase activity using Pamgene software. b To measure protein serine/threonine kinase activity in dexamethasone-treated cells, lysates were applied in Pamgene peptide substrate-based serine/threonine kinase array and analyzed for serine/threonine kinase activity using Pamgene software. c Changes in the expression of Sprouty family genes in dexamethasone and prednisolone-treated SUP-B15 cells compared to DMSO-treated cells. Error bars show standard deviation. d–i SUP-B15 cells were treated with 1 µM dexamethasone for the indicated time period before lysis. Lysates were analyzed by SDS-PAGE and western blotting using specific antibodies as labeled. j SUP-B15 cells were treated with 1 µM dexamethasone and with different concentrations of Tozasertib for 24 h before lysis. Lysates were analyzed by SDS-PAGE and western blotting using specific antibodies as labeled. Blots shown in each panel were from the same experiment and processed similarly.

Fig. 3. Dexamethasone sensitivity prediction model.

a Deregulated gene signatures from dexamethasone- and prednisolone-treated SUP-B15 cells were combined with genes displaying the highest level of variation in CCLE and TARGET (ALL) datasets and 500 genes were selected. The 500 genes from 138 cell lines of hematological malignancies were used to build a deep learning model. The model was tested using three sets of samples. b Confusion matrix for all three groups of samples. c The performance of the model was calculated using three test sample groups.

Fig. 4. Dexamethasone sensitivity prediction in ALL patient samples.

The TARGET dataset for ALL was used to predict dexamethasone sensitivity. a In total 205 patient samples were used to predict dexamethasone sensitivity by the deep learning model. b Event-free survival between dexamethasone-sensitive and -resistant groups using 205 ALL patient samples was determined by GraphPad Prism. c Event-free survival for ALL patient samples collected from bone marrow. d Event-free survival for ALL patient samples collected from peripheral blood.

Fig. 5. Pathway enrichment in dexamethasone-resistant ALL patient samples.

Pathway enrichment in dexamethasone-resistant ALL patient samples was analyzed using GSEA. a Hallmarks and b Oncogenic signatures gene sets were used for pathway enrichment analysis. c Upregulated and downregulated genes in dexamethasone-resistant ALL patient samples were determined by SAM. The bar graph shows selected top-listed genes. All the upregulated and downregulated genes are included in supplementary table T2. d The enrichment of KEGG cytokine and cytokine receptor interaction pathway in dexamethasone-resistant ALL patient samples was determined by GSEA. e, f SUP-B15 and TANOUE cells were lysed. Lysates were analyzed by SDS-PAGE and western blotting using specific antibodies as labeled. Blots shown in each panel were from the same experiment and processed similarly.

Fig. 6. The synergy between dexamethasone and kinase inhibitors.

a Using the same 500-gene signature and drug synergy data from the DrugComb database, a deep learning binary classification model was developed to predict synergy between dexamethasone and other inhibitors. b Model performance was tested using 108 test samples. c Confusion matrix showing the model’s performance using 108 test samples. d In silico synergy prediction between dexamethasone and 1454 kinase inhibitors using the deep learning model. e In vitro synergy measurement between dexamethasone and kinase inhibitors in TANOUE cells using Cell Titer Glo after 48 h of incubation with drug combinations. The figure shows representative kinase inhibitors. f In vitro synergy measurement between dexamethasone and β-catenin inhibitors in TANOUE cells using Cell Titer Glo after 48 h of incubation with drug combinations.