BCR-ABL1-induced downregulation of WASP in chronic myeloid leukemia involves epigenetic modification and contributes to malignancy

Abstract

Chronic myeloid leukemia (CML) is a myeloproliferative disease caused by the BCR-ABL1 tyrosine kinase (TK). The development of TK inhibitors (TKIs) revolutionized the treatment of CML patients. However, TKIs are not effective to those at advanced phases when amplified BCR-ABL1 levels and increased genomic instability lead to secondary oncogenic modifications. Wiskott-Aldrich syndrome protein (WASP) is an important regulator of signaling transduction in hematopoietic cells and was shown to be an endogenous inhibitor of the c-ABL TK. Here, we show that the expression of WASP decreases with the progression of CML, inversely correlates with the expression of BCR-ABL1 and is particularly low in blast crisis. Enforced expression of BCR-ABL1 negatively regulates the expression of WASP. Decreased expression of WASP is partially due to DNA methylation of the proximal WASP promoter. Importantly, lower levels of WASP in CML advanced phase patients correlate with poorer overall survival (OS) and is associated with TKI response. Interestingly, enforced expression of WASP in BCR-ABL1-positive K562 cells increases the susceptibility to apoptosis induced by TRAIL or chemotherapeutic drugs and negatively modulates BCR-ABL1-induced tumorigenesis in vitro and in vivo. Taken together, our data reveal a novel molecular mechanism that operates in BCR-ABL1-induced tumorigenesis that can be used to develop new strategies to help TKI-resistant, CML patients in blast crisis (BC).

Figure 1

WASP is downregulated in CML patients and BCR–ABL1-positive cell lines. (a) The expression of WASP is suppressed in CML patients comparing with healthy donors (HD). The relative expression of WASP in PBMC was determined by real-time PCR using GAPDH as housekeeping gene (**P<0.01, comparing with HD). Values are plotted as 2^−ΔΔCt. 't' student was used as statistical test. (b) The downregulation of WASP was observed in advanced phases of the disease (AP and BC). (P<0.0001, comparing with HD). Patients in CCyR showed normal WASP levels, according to the healthy donor control group (P<0.001, comparing with BP). Analysis of variance (ANOVA) and Bonferronni posttest were used to statistical analysis. (c) The expression of WASP negatively correlates with BCR–ABL1 expression in CML patients (P<0.0001). Spearman test was used with linear regression. (d) Western blot showing the expression of WASP in cell lines derived from different leukemia types. BCR–ABL1-negative cell lines express high levels of WASP, whereas BCR–ABL1-positive cells express discrete or undetectable WASP. Actin was used as loading control. (e) Stable expression of BCR–ABL1 was induced in HL-60 and Jurkat cell lines by retroviral infection, resulting in downregulation of WASP both at the mRNA and protein levels. GAPDH was used as housekeeping control for qPCR. Western blots show the increased numbers of tyrosine phosphorylated proteins in BCR–ABL-positive HL-60 and Jurkat cell lines, and the complete WASP silencing after BCR–ABL1 expression. Actin protein was used as loading control

Figure 2

WASP downregulation is not rescued by imatinib or calpain inhibitor and involves epigenetic modification. (a and b) Cell lines were treated for 24 h with 10 μM IM and/or the calpain inhibitor N-acetyl-l-leucyl-l-leucyl-l-norleucinal (ALLN), and western blot was performed to reveal the expression of WASP, WIP and the status of BCR–ABL1-TK activity. Actin was used as loading control. (c) Scheme of the distal and proximal promoters of WASP. (d) In silico analysis using UCSC Genome Browser public data (ww.ucsc.edu) show the presence of CpG sites in the promoters of WASP (green boxes). (e) CpG methylation status in the proximal promoter of WASP was evaluated. Black and white circles: methylated and unmethylated CpG sites

Figure 3

5-AZA treatment restored the levels of WASP expression in some but not all CML cell lines. (a) Treatment of CML cell lines with 1 μM 5-AZA alone restored the levels of WASP expression in LAMA-84, KCL22 and BV173, but not in K562 cells. The relative expression of RASSF1A and GST was used as positive controls for 5-AZA treatment. Two way Anova (Analysis of variance) and Bonferroni post test were used to determine statistic significance among the groups. *P < 0.05, **P < 0.01, ***P < 0.001

Figure 4

Lower levels of WASP in CML patients at advanced phases may indicate a worse OS and WASP re-expression negatively modulates BCR–ABL1-induced tumorigenesis. (a) The relative WASP expression values obtained from patients in AP and BP were divided in two groups according to the median: WASP very low for patients exhibiting a strong WASP downregulation (lower than median), and WASP low for patients presenting a mild WASP suppression (higher than median). (b) Patients with higher levels of WASP (grouped in WASP low) presented longer OS (median=61.75 months) comparing with the patients in WASP very low group. Kaplan–Meier was used as statistical test. (c) K562 cell line was transduced with lentivirus vector to induce the re-expression of WASP (eGFP) or WIP (mCherry). (d) K562.WT or re-expressing WASP or WIP were plated and the number of cells was counted in Neubauer chamber according to the time point in the graph. ***P<0.001 comparing K562.WASP with K562.WT. ^##P<0.01 comparing K562.WASP with K562.WIP. (e) K562.WT was co-cultured with K562.WASP in three different initial proportions and the frequency of each population was assessed 7 days later by flow cytometry. (f) K562.WIP (mCherry-positive) and K562.WASP (eGFP-positive) were subcutaneously injected in the right and left flanks of the BALB/c nude mice, respectively, and the development of the solid tumors was followed and imaging of the tumors was assessed at 21^st day. The tumors were collected to determine volume (g) and weight (h). Student's t-test was used for statistical analysis. *P<0.05

Figure 5

WASP sensitizes BCR–ABL1-positive cells to TRAIL- and chemotherapy-induced apoptosis. (a) K562.WT and K562.WASP were cultured in the presence or absence of TRAIL-R2/Fc to prevent TRAIL-mediated apoptosis and PI uptake was assesed 48 h later. (b) qPCR shows the upregulation of DR4, DR5 and TRAIL after re-expression of WASP in K562 cells. (c) K562.WT and K562.WASP were treated with different concentrations of IM and cell death was assessed by the hypodiploid DNA content (Sub G0) (*P<0.05; **P<0.01; ***P<0.001). (d) Cell lines were treated with camptothecin, actinomycin D and ultra violet (UV) radiation and cell death was assessed after 48 h by hypodiploid DNA content. Two way ANOVA was used for statistical analysis. ***P<0.001