Characterization of p190-Bcr-Abl chronic myeloid leukemia reveals specific signaling pathways and therapeutic targets

Abstract

The oncogenic protein Bcr-Abl has two major isoforms, p190^Bcr-Abl and p210^Bcr-Abl. While p210^Bcr-Abl is the hallmark of chronic myeloid leukemia (CML), p190^Bcr-Abl occurs in the majority of Philadelphia-positive acute lymphoblastic leukemia (Ph + ALL) patients. In CML, p190^Bcr-Abl occurs in a minority of patients associating with distinct hematological features and inferior outcomes, yet the pathogenic role of p190^Bcr-Abl and potential targeting therapies are largely uncharacterized. We employed next generation sequencing, phospho-proteomic profiling, and drug sensitivity testing to characterize p190^Bcr-Abl in CML and hematopoietic progenitor cell line models (Ba/f3 and HPC-LSK). p190^Bcr-Abl CML patients demonstrated poor response to imatinib and frequent mutations in epigenetic modifiers genes. In contrast with p210^Bcr-Abl, p190^Bcr-Abl exhibited specific transcriptional upregulation of interferon, interleukin-1 receptor, and P53 signaling pathways, associated with hyperphosphorylation of relevant signaling molecules including JAK1/STAT1 and PAK1 in addition to Src hyperphosphorylation. Comparable to p190^Bcr-Abl CML patients, p190^Bcr-Abl cell lines demonstrated similar transcriptional and phospho-signaling signatures. With the drug sensitivity screening we identified targeted drugs with specific activity in p190^Bcr-Abl cell lines including IAP-, PAK1-, and Src inhibitors and glucocorticoids. Our results provide novel insights into the mechanisms underlying the distinct features of p190^Bcr-Abl CML and promising therapeutic targets for this high-risk patient group.

Fig. 1. Clinical, genomic and signaling profiles of p190-CML patients.

a Landscape of nonsilent mutations identified by WES from p190-CML patients (n = 4). Variants that has been linked with cancer (COSMIC database) are shown in the figure. Complete list of mutations can be found from Supplementary Table 3. The bottom tracks show clinical (monocytosis (>10% of total leukocytic count), basophilia, additional chromosomal abnormalities (ACAs)) and treatment features of patients (type of treatment, achievement of major molecular response (MMR) at 12 months, sustained MMR). The color of the variant box indicates the type of mutation. b PCA analysis of transcriptional data (protein coding genes) from CML patients. c Volcano plot of protein coding genes between p190-CML (n = 3, right) and p210-CML (n = 3, left). Each gene is represented by a black dot and significant differentially expressed genes (Q < 0.05, Bayesian statistical test) are colored red. d Gene set enrichment analysis (GSEA) output showing upregulation of TNF, IFNα, IL1R, and P53 signaling pathways in p190-CML compared to p210-CML patients. e Western blot analysis of JAK/STAT pathway, Src and PAK1 signaling molecules in CML patients. For each protein a phospho specific antibody was used in addition to an antibody against the protein to assess the overall protein expression.

Fig. 2. Transcriptional regulations in p190 and p210 cell line models.

a Heatmap of top 50 variably expressed genes among Ba/f3 and HPC-LSK cell lines (n = 12). Fading blue colors indicate down-regulation of the gene in the sample and red its up-regulation relative to the mean expression of the genes across all samples. The Explanatory track indicate the isoform type. Clustering was performed for both genes and samples using the Euclidean distance and Ward linkage method. b PCA analysis of transcriptional data (protein coding genes) from Ba/f3 and HPC-LSK cell lines. c Volcano plot of protein coding genes between p190 (n = 6, right) and p210 (n = 6, left) from combined cell lines data. Each gene is represented by a black dot and significantly differentially expressed genes (Q < 0.05, Bayesian statistical test) are colored red. d Venn diagram showing the number of differentially expressed genes between p190 and p210 cell lines in Ba/f3 (violet) and HPC-LSK models(blue). 2316 genes and 650 genes were differentially expressed between p190 and p210 in Ba/f3 and HPC-LSK models respectively. Of these genes, 111 were differentially expressed between the two isoforms in both models. Boxes show depiction of the main molecular pathways significantly deregulated between p190 and p210 from the combined data. The color of the boxes as well as numbers in Venn diagram indicate upregulation in p190 (red) or p210 (Blue). Genes enlisted are significantly differentially expressed between p190 and p210 (Q < 0.05) in the combined model data. The full list of differentially expressed genes and enriched pathways can be found in Supplementary Table 4.

Fig. 3. Phospho-signaling profiles of p190 and p210 cell lines.

Relative abundance of normalized phospho-proteins signals between p190 and p210 in (a) Ba/f3 and (c) HPC-LSK cell lines based on phospho-array results (Log2 scale). Signals have been normalized to the median antibody signal of each slide. The top differentially phosphorylated tyrosine-sites proteins in (b) Ba/f3 and (d) HPC-LSK cell lines between p190 and p210. the cutoff value was set to ±33% in Ba/f3 cells and ±50% in HPC-LSK cells. the color indicates the direction of increased (red) or decreased (blue) phosphorylation in p190/p210 comparison. e Western blot analysis of JAK/STAT pathway, Src and PAK1 signaling molecules in cell line models. For each protein a phospho-specific antibody was used in addition to an antibody against the total protein to assess the overall protein expression. Protein quantifications are presented in Supplementary Fig. 3.

Fig. 4. Isoform specific drug sensitivities in cell line models.

a Scatter plot comparing average drug sensitivity score (DSS) of p190-Ba/f3 and p210-Ba/f3 cells. Drug sensitivity and resistance testing (DSRT) experiments have been performed in triplicates for each setting. Color indicates different drug families (primary targets). b Dose response curves showing drug responses of p190-Ba/f3 (red) and p210-Ba/f3 (blue) cell lines to LCL161 (IAP inhibitor), FRAX486 (PAK inhibitor), dasatinib (TKI with Src inhibitory activity) and dexamethasone (glucocorticoids). Concentrations of drugs are indicated in nM. Drug responses to other drugs from the indicated families are available in Supplementary Fig. 4. c Scatter plot comparing average drug sensitivity score (DSS) of p190-HPC-LSK and p210-HPC-LSK cells. Drug sensitivity and resistance testing (DSRT) experiments have been performed in triplicates for each setting. Color indicates different drug families (primary targets). d Dose response curves showing drug responses of p190-HPC-LSK (red) and p210-HPC-LSK (blue) cell lines to LCL161 (IAP inhibitor), FRAX486 (PAK inhibitor) and dasatinib (TKI with Src inhibitory activity). Concentrations of drugs are indicated in nM.

Fig. 5. Colonogenic potentials and drug combination strategies in p190 and p210 cell lines.

a Number of colonies produced by p190-HPC-LSK and p210-HPC-LSK cell lines. b Density plot showing phenotypic analysis of harvested cells from p190-HPC-LSK and p210-HPC-LSK colonies using flowcytometry. CD19 (lymphoid marker) on vertical axis and CD11b (granulocytic marker) on horizontal axis. c Comparison of the inhibitory effect of imatinib, dasatinib, LCL161, dexamethasone, FRAX486, idasanutlin, and IFNα on colonogenic potentials of p190-HPC-LSK and p210-HPC-LSK. Bar height indicate the average colony counts (normalized to respective control condition) and error bars represent standard deviation. Dose response curves of percent inhibition achieved with imatinib alone and in combination with indicated concentrations of LCL161, FRAX486, idasanutlin, dexamethazone, and IFNα in (d) p190-Ba/f3 and (e) p190-HPC-LSK cell lines. The experiments were conducted in triplicates. f Comparison of the inhibitory effect of imatinib (100 nM) alone and in combination with the indicated concentrations of LCL161, FRAX486, idasanutlin, dexamethazone, and IFNα on the colonogenic potentials of p190-HPC-LSK cells. (*) means p < 0.05, (**) means p < 0.01, (***) means p < 0.001, and (****) means p < 0.0001.