The miR-185/PAK6 axis predicts therapy response and regulates survival of drug-resistant leukemic stem cells in CML

Abstract

Overcoming drug resistance and targeting cancer stem cells remain challenges for curative cancer treatment. To investigate the role of microRNAs (miRNAs) in regulating drug resistance and leukemic stem cell (LSC) fate, we performed global transcriptome profiling in treatment-naive chronic myeloid leukemia (CML) stem/progenitor cells and identified that miR-185 levels anticipate their response to ABL tyrosine kinase inhibitors (TKIs). miR-185 functions as a tumor suppressor: its restored expression impaired survival of drug-resistant cells, sensitized them to TKIs in vitro, and markedly eliminated long-term repopulating LSCs and infiltrating blast cells, conferring a survival advantage in preclinical xenotransplantation models. Integrative analysis with mRNA profiles uncovered PAK6 as a crucial target of miR-185, and pharmacological inhibition of PAK6 perturbed the RAS/MAPK pathway and mitochondrial activity, sensitizing therapy-resistant cells to TKIs. Thus, miR-185 presents as a potential predictive biomarker, and dual targeting of miR-185-mediated PAK6 activity and BCR-ABL1 may provide a valuable strategy for overcoming drug resistance in patients.

Graphical abstract

Figure 1.

Differentially expressed miRNAs are identified in primary CD34⁺ CML cells. (A) DESeq2 analysis of differentially expressed miRNAs in CD34⁺ cells, comparing 3 NBM and 6 CML samples (3 IM responders and 3 IM nonresponders). Plots show the distribution of differentially expressed miRNAs, representing the fold-change between CML relative to NBM (left panel). Red dots represent the differentially expressed miRNAs with adjusted P values < .05. G plots package was used to plot heat maps, accompanied with unsupervised dendrogram analysis, based on the differentially expressed miRNAs with adjusted P values < .05 (right panel). (B) Similar analyses performed for miRNAs found to be differentially expressed between IM responders (R) and IM nonresponders (NR). (C) Differentially expressed miRNAs were determined using a TaqMan qPCR microfluidics device on extracts of CD34⁺ cells obtained from NBM (n = 11) and CML samples (n = 22). Raw Ct values obtained from the 96-well multiplexing microfluidics device were organized using the HTqPCR package and normalized using the quantile method with the limma package. A nonparametric Mann-Whitney U test was performed to compare unpaired samples. Levels of 20 of the miRNAs tested were shown to be different between CD34⁺ NBM and CML cells. Data points represent quantile-normalized Ct values relative to CD34⁺ NBM cells. All comparisons shown are statistically significant (Benjamini-Hochberg-adjusted P value < .05).

Figure 2.

Reduced expression of miR-185 predicts therapy response in CD34⁺ CML cells and miR-185 reexpression restores TKI sensitivity in these cells. (A) TaqMan qRT-PCR was performed to validate differentially expressed miR-185 and miR-340 in CD34⁺ cells from IM responders (n = 11) and IM nonresponders (n = 11). (B). Random forest classifier analysis of miR-185 and miR-340, assessed by the leave-one-out cross-validation (LOO-CV) method, to predict IM responders and IM nonresponders. (C) Validation of miR-185 transcript levels in CD34⁺ cells from 47 responders and 11 nonresponders from a second, separate CML cohort before therapy with the TKI nilotinib (baseline) and 3 months posttreatment. (D) Results of CFC assays (± TKIs) of CD34⁺ CML cells from TKI nonresponders (n = 3) transduced with either a miR-185 expressing vector or a pRRL control vector, with or without 5 µM IM, 150 nM DA, or 5 µM NL treatment. The y-axis shows the frequency of colonies derived from erythroid-burst forming units (BFU-E) and granulocyte/macrophage-colony forming units (CFU-GM; left panel). Results of replating all the cells harvested from the primary CFC assays (left panel) into secondary CFC assays are shown in the right panel. (E) Results of LTC-IC assays performed on the same transduced cells as in panel D, with or without TKIs, as indicated. Data shown are mean ± standard error of the mean (SEM). CFC outputs were measured for cells from 4 individual patients with CML. P values were calculated using a 2-tailed unpaired Student t test.

Figure 3.

Restoration of miR-185 expression reduces the burden of leukemia and sensitizes leukemic blast cells to TKI treatment, with enhanced survival of leukemic mice. (A) Schematic of the experimental design. BV173^YFP/Luc cells transduced with a miR-185-expressing vector, or a pRRL control vector, were injected intravenously into sublethally irradiated nonobese diabetic/severe combined immunodeficiency IL2Rγ-chain–deficient (NSG) mice (2.5 × 10⁶ cells per mouse). Two weeks later, a daily oral gavage treatment of 15 mg/kg DA or vehicle (propylene glycol) was initiated for another 2 weeks. (B) Bioluminescence images of mice taken 2 weeks posttransplantation, before initiation of the oral gavage treatment. (C) Representative bioluminescence images of mice from each group taken 7 weeks posttransplantation. One mouse from each group was then sacrificed to obtain images and weights of spleens and livers. (D) FACS profiles showing human leukemic cell chimerism (GFP⁺ and YFP⁺) in PB, BM, spleen, and liver from mice in each group. (E) Representative hematoxylin and eosin-stained sections of spleens and livers from all treatment groups. (F) Fold-difference in BCR-ABL1 transcripts in hematopoietic tissues from each group. Data shown are mean ± SEM of measurements from 3 technical replicates. P values were calculated using a 2-tailed paired Student t test. (G) Western blots of whole protein extracted from BM cells of mice in each group, after probing with the antibodies indicated. (H) Overall survival of mice from each treatment group (n = 6 mice per group). P values were calculated using log-rank test. ND = not detectable.

Figure 3.

Restoration of miR-185 expression reduces the burden of leukemia and sensitizes leukemic blast cells to TKI treatment, with enhanced survival of leukemic mice. (A) Schematic of the experimental design. BV173^YFP/Luc cells transduced with a miR-185-expressing vector, or a pRRL control vector, were injected intravenously into sublethally irradiated nonobese diabetic/severe combined immunodeficiency IL2Rγ-chain–deficient (NSG) mice (2.5 × 10⁶ cells per mouse). Two weeks later, a daily oral gavage treatment of 15 mg/kg DA or vehicle (propylene glycol) was initiated for another 2 weeks. (B) Bioluminescence images of mice taken 2 weeks posttransplantation, before initiation of the oral gavage treatment. (C) Representative bioluminescence images of mice from each group taken 7 weeks posttransplantation. One mouse from each group was then sacrificed to obtain images and weights of spleens and livers. (D) FACS profiles showing human leukemic cell chimerism (GFP⁺ and YFP⁺) in PB, BM, spleen, and liver from mice in each group. (E) Representative hematoxylin and eosin-stained sections of spleens and livers from all treatment groups. (F) Fold-difference in BCR-ABL1 transcripts in hematopoietic tissues from each group. Data shown are mean ± SEM of measurements from 3 technical replicates. P values were calculated using a 2-tailed paired Student t test. (G) Western blots of whole protein extracted from BM cells of mice in each group, after probing with the antibodies indicated. (H) Overall survival of mice from each treatment group (n = 6 mice per group). P values were calculated using log-rank test. ND = not detectable.

Figure 4.

Restoration of miR-185 expression eliminates TKI nonresponder cells regenerated in transplanted nonobese diabetic-Rag1^−/−IL2Rγc^−/− mice. (A) Detection of human GFP⁺CD45⁺ cells in the BM of mice injected with CD34⁺ CML patient cells transduced with a pRRL control vector (light blue), a miR-185 vector (light red), pRRL plus DA treatment (dark blue), or miR-185 plus DA treatment (dark red) at weeks 4, 12, and 25 posttransplantation. (B) qRT-PCR results for BCR-ABL1 transcript levels relative to GAPDH in total BM cells of all mice analyzed at 25 weeks. (C) Representative FACS plots of human (CD45⁺) and GFP⁺CD45⁺ cells detected in the BM of mice analyzed at 25 weeks posttransplantation (left) and a summary of the levels of transduced GFP⁺CD45⁺ (solid bars) versus non-transduced GFP^-CD45⁺ (white bars, right) human cells in all mice studied. (D) Levels of human GFP⁺CD34⁺, CD34⁺CD38⁻, and myeloid cells detected in the BM of all mice analyzed at week 25 posttransplantation. (E) Representative FACS plots of human GFP⁺CD34⁺ cells in the BM of mice at week 25 posttransplantation. (F) Colony output of FACS-sorted human GFP⁺CD45⁺ cells harvested from the BM of mice 25 weeks posttransplantation of cells from a CML patient sample. Data shown, except where otherwise indicated, are the mean ± SEM of measurements from 2 cohorts of mice injected with transduced cells from 2 independent TKI nonresponder patients. Each dot/triangle represents 1 mouse. P values were calculated using a 2-tailed unpaired Student t test. No. = number.

Figure 5.

RNA-seq analysis identifies increased expression of OXPHOS and ROS gene signatures in TKI nonresponder cells and PAK6 as a target of miR-185. (A) DESeq2 analysis of differentially expressed genes from the same 9 RNA samples used for miRNA profiling and a comparison of genes expressed differentially in CD34⁺ cells from NBM (n = 3) vs CML (n = 6, top panel) samples and IM responder (n = 3) vs IM nonresponder (n = 3, bottom panel) samples. Red dots represent differentially expressed genes with adjusted P values < .05 between these samples. (B) Gene set enrichment analysis plots for OXPHOS and ROS pathways, with nominal enrichment scores. (C) Heat maps of miR-185-predicted target genes identified in CD34⁺ cells from NBM vs CML cells. (D) qRT-PCR analysis of miR-185 expression in IM-resistant K562R cells, CD34⁺ IM responder (n = 3) and IM nonresponder (n = 3) cells transduced with a miR-185-expressing vector or a pRRL control vector and incubated with or without IM (5.0 µM) for 24 hours. (E) Western blot analysis of several key signaling proteins in miR-185-transduced K562R cells, PAK6 siRNA-transfected K562R cells, and control cells cultured with or without IM (5.0 µM) for 24 hours. ACTIN served as loading control. (F-H) qRT-PCR analyses of the transcript levels of PAK6 in CD34⁺ cells from NBM vs IM responders (R) or IM nonresponders (NR). Correlation analysis of PAK6 and miR-185 expression in responder and nonresponder patients (G), and PAK6 transcript levels in CD34 subpopulations from responders vs nonresponders. P values were calculated using a 2-tailed paired (Figure 6F) or unpaired Student t test. Ctrl = control.

Figure 5.

RNA-seq analysis identifies increased expression of OXPHOS and ROS gene signatures in TKI nonresponder cells and PAK6 as a target of miR-185. (A) DESeq2 analysis of differentially expressed genes from the same 9 RNA samples used for miRNA profiling and a comparison of genes expressed differentially in CD34⁺ cells from NBM (n = 3) vs CML (n = 6, top panel) samples and IM responder (n = 3) vs IM nonresponder (n = 3, bottom panel) samples. Red dots represent differentially expressed genes with adjusted P values < .05 between these samples. (B) Gene set enrichment analysis plots for OXPHOS and ROS pathways, with nominal enrichment scores. (C) Heat maps of miR-185-predicted target genes identified in CD34⁺ cells from NBM vs CML cells. (D) qRT-PCR analysis of miR-185 expression in IM-resistant K562R cells, CD34⁺ IM responder (n = 3) and IM nonresponder (n = 3) cells transduced with a miR-185-expressing vector or a pRRL control vector and incubated with or without IM (5.0 µM) for 24 hours. (E) Western blot analysis of several key signaling proteins in miR-185-transduced K562R cells, PAK6 siRNA-transfected K562R cells, and control cells cultured with or without IM (5.0 µM) for 24 hours. ACTIN served as loading control. (F-H) qRT-PCR analyses of the transcript levels of PAK6 in CD34⁺ cells from NBM vs IM responders (R) or IM nonresponders (NR). Correlation analysis of PAK6 and miR-185 expression in responder and nonresponder patients (G), and PAK6 transcript levels in CD34 subpopulations from responders vs nonresponders. P values were calculated using a 2-tailed paired (Figure 6F) or unpaired Student t test. Ctrl = control.

Figure 6.

The PAK inhibitor PF-3758309 sensitizes CD34⁺ nonresponder cells to TKIs and perturbs mitochondrial activity. (A) Representative confocal images showing the morphology of mitochondria stained in CD34⁺ IM nonresponder cells treated with PF-3758309 (PF) and IM, as compared with control and single drug treatment and quantification of mitochondria to nuclear area ratio in these treated cells are shown and each dot represents an individual cell (top, n = 2). The white scale bar represents 20 μm. Intracellular mitochondrial staining of the same cells by MitoTracker Deep Red (MTDR) is also shown (bottom). (B) ROS was assessed using CellROX Deep Red in the same cells as in (A), and representative images and intracellular ROS accumulation are shown. The white scale bar represents 50 μm. (C-D) Viability and apoptosis assays of IM-resistant K562 cells (K562R) and CD34⁺ cells from IM nonresponders (n = 3) for 72 hours in the presence or absence of PF or TKIs (IM, DA, and NL), alone or in combination. (E) CFC assays of CD34⁺ CML cells from TKI nonresponders (n = 3) with or without PF or TKIs, alone or in combination. Data shown are mean ± SEM of measurements from 3 patient samples. P values were calculated using a 2-tailed unpaired Student t test. (F) Model of how restored expression of miR-185 or inhibition of PAK6 activity sensitizes TKI-resistant CML cells to TKIs both in vitro and in vivo, by dual targeting of a novel miR-185-PAK6 axis and BCR-ABL1 activity.