Knock-down of Kaiso induces proliferation and blocks granulocytic differentiation in blast crisis of chronic myeloid leukemia

K562 and LAMA-84 cell line were maintained in RPMI 1640 medium supplemented with 10% foetal bovine serum (Hyclone), 100 U/ml penicillin (Invitrogen), 100 mg/mL streptomycin (Invitrogen) at 37°C in 5% CO2. K562, established from a CML patient in blast crisis [38], was used as a BCR-ABL-positive cell line. Imatinib-resistant K562 cell line (a gift from A. Mencalha, INCA RJ) was obtained by in vitro passaging of K562 in progressively increasing doses of imatinib. LAMA-84 is a human leucocytic cell line with basophilic characteristic [39].

All samples were obtained from patients admitted to or registered at the Instituto Nacional de Câncer (Rio de Janeiro, Brazil), following the guidelines of the local Ethics Committee and the Helsinki declaration. Diagnoses and follow-up were based on hematologic, cytogenetic and molecular assays.

K562 cell line were exposed to different doses of Imatinib dissolved in Dimethyl sulphoxide (DMSO; Sigma Aldrich). DMSO-treated cells were used as vehicle controls.

The viability of cells was measured using a 4-[3-(4-Iodophenyl)-2-(4-nitrophenyl)-2 H-5-tetrazolio]-1,3-benzene disulphonate (WST-1) assay (Roche). Approximately 2 × 10⁵cells/mL. Cells were plated into 96-well microplates (Corning) for 24 h. After 24 h, 10 μL WST-1 was added to each well, and plates were incubated at 37°C for an additional 2 h. Plates were read on a microplate reader (Bio-Rad, model 550) at 450 nm with a reference wavelength at 630 nm.

All RNA oligonucleotides described in this study were synthesized and purified using highperformance liquid chromatography (HPLC) at Integrated DNA Technologies (Coralville, Iowa), and the duplex sequences are available upon request. RNAi knockdown and transfections were performed following the manufacturer’s protocols of the TriFECTa Dicer-Substrate RNAi kit (Integrated DNA Technologies, Coralville, IA) and the CodeBreaker siRNA Transfection Reagent (Promega,USA). K562 cells (1 × 10⁶ cells per well) were split in 24-well plates to 60% confluency in RPMI media 1 day prior to transfection. The TriFECTa kit contains control sequences for RNAi experiments which include a fluorescent-labeled transfection control duplex and a scrambled universal negative control RNA duplex that is absent in human, mouse, and rat genomes. Fluorescence microscopy and FACS monitored the transfection efficiency according to the manufacturer’s recommendations. Only experiments in which transfection efficiencies were ≥ 90% were evaluated. RNA levels were measured 36 h after transfection, and protein levels were measured 80 h later. All duplexes used were evaluated at 25, 10, 1, and 0.1 nM. All transfections were minimally performed in triplicate, and the data were averaged. Knockdown of Kaiso and P120ctn was performed, and RNA, protein extraction, QRT-PCR, Western blot, and FACS analysis were done as described above.

QRT-PCR Analysis Quantitation of Kaiso, P120ctn, Wnt11, β-catenin, SCF, c-MYB, c-EBPα, Gata-2, PU-1 RNA transcripts was carried out by real time PCR (QRT-PCR). Two micrograms of total RNA from K562 cell line or transfected K562 cell line, were reverse transcribed with Superscript III Reverse transcriptaseVR (Invitrogen). cDNAs were mixed with SYBR Green PCR Master MixVR (Applied Biosystems) and specific primers. Real time PCR was performed in an ABI Prism 7000 thermocycler (Applied Biosystems), with 50 cycles of 15 s at 95°C and 2 m at 68°C. Expression levels were estimated in triplicate with specific and control primers. For each sample, the relative amounts of transcripts of the target gene and the internal control were estimated from a standard curve. Results were expressed in arbitrary units as the ratio of the target gene transcript/internal transcript (data represented by average ± SD of three measurements).

Protein lysates were prepared as previously reported [40]. Protein concentrations were determined by the Bradford method. Approximately 200 μg protein was resolved on 7% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels, blotted onto nitrocellulose membranes (Bio-Rad, CA) and probed with individual antibodies, and visualized by the enhanced chemiluminescence ECL Plus Western Blotting Detection ReagentsVR (GE healthy care, United Kingdom). The following antibodies were used: anti-kaiso (Santa Cruz Biotechnology, Santa Cruz, CA), anti-actin (Santa Cruz Biotechnology, Santa Cruz, CA). The secondary antibodies were horseradish peroxidase (HRP)-conjugated rabbit antimouse IgG (Santa Cruz Biotechnology, Santa Cruz, CA).

K562 cells were incubated in RPMI (Invitrogen, USA), harvested after 16 h, and washed several times in PBS. Normal and imatinib-resistant K562 cells were resuspended at a concentration of 2 × 10⁶/ml in PBS. Normal and imatinib-resistant K562 cells (50,000) were attached to microscope slides by centrifugation for 2 min at 800 rpm at high acceleration in a Cytospin 2 centrifuge (Shandon; Frankfurt, Germany) and dried for 10 min at 37°C in a sterilizer. For immunofluorescence, culture cell were prefixed in formaldehyde vapor by placing the slide into a chamber containing paper towel embedded with formaldehyde for 10 min. Subsequently, the slides were immersed in buffered 4% paraformaldehyde for 15 min. After several washes in phosphate-buffered saline (PBS; pH 7.4), K562 cells were incubated for 72 h at 4°C with primary antibodies diluted in PBS with 0.3% Triton-X 100 (Reagen, Curitiba, PR, Brazil) and 5% normal goat serum (Invitrogen, Carlsbad, CA). Primary antibodies were the following: anti-Kaiso (mouse 1:100; Santa Cruz, CA), anti β-tubulin (mouse 1:500; Sigma), Secondary antibodies were incubated for 2 h at room temperature. Secondary antibodies were the following: goat anti-mouse IgG conjugated with Cy3 (1:800; Jackson ImmunoResearch). Slides were counterstained with DAPI (Molecular Probes). Conventional fluorescence microscopy was performed in an Eclipse TE200 inverted microscope (Nikon, Tokyo, Japan), equipped with a CoolSNAP-Pro cf CCD camera(Media Cybernetics, Silver Spring, MD; monochrome). Images were acquired with the aid of Image-Pro Express software (version 4.5.1.3) and edited with Photoshop CS5.1 (Adobe, San Jose, CA). For FACS analysis, antibodies that recognize cell surface myeloid-specific antigens GPA-phycoerythrin (PE), CD33-fluorescein isothiocyanate (FITC), CD11b-PE, CD15-FITC, CD117-PE, CD71-FITC (Becton Dickinson) were used. Appropriated isotype-matched controls (Becton Dickinson) were used.

Immunohistochemical staining was performed in formalin-fixed, paraffin-embedded bone marrow slides from five CML patients in the chronic phase and six patients in the blastic phase, according to standard procedures. Heat-induced epitopes were retrieved in Tris buffer (pH = 9.9; Dako, Denmark) in a microwave processor. Tissue sections were subsequently incubated with anti-KAISO (1:1000) overnight and with anti-goat immunoglobulin G and peroxidase for 30 minutes at room temperature. Slides were developed using 3,3´-diaminobenzidine/H2O2 (Dako, Denmark) and a hematoxylin counterstain. Slides were analyzed and photographed with a Nikon Eclipse E600 microscope.

Data are expressed as means ± standard deviation (SD). The significance of differences between control and treated groups was evaluated using one-way analysis of variance (ANOVA). Experimental tests were performed at least three times. Differences were considered to be significant when P < 0.05.

The Kaiso protein, like other members of the subfamily POZ-ZF [20, 22, 25, 44‐48], has been implicated in cancer development process when it has been found that Kaiso inhibits activation mediated by β-catenin of the Mmp7 gene (also known as matrilysin), which is well known for metastatic spread [14]. Recently another study suggests that Kaiso can regulate TCF/LEF1-activity, via modulating HDAC1 and β-catenin-complex formation [15]. This shows that Kaiso can directly regulate the signaling pathway of canonical Wnt/β-catenin widely known for its involvement in human tumors. The Kaiso overexpression decreases the ability of TCF/LEF to interact with β-catenin, which implies that Kaiso and TCF/LEF are associated in the nucleus [17].

As expected for a transcriptional factor, the Kaiso protein is often found in the nucleus of several tumor or non-tumor derived mammalian cell lines [28]. Recent studies using immunohistochemistry analysis of normal and tumor tissue revealed that Kaiso protein is predominantly localized in the cytoplasm of the cell or is totally absent, though [49]. These data are consistent with the results found in the K562 cell line in which expression of the Kaiso is predominantly cytoplasmic (Figures 1 and 2). This seems to be unusual because Kaiso has a signal “NLS” highly conserved and required for any protein with nuclear localization. Moreover, Kaiso uses classical nuclear transport mechanisms through interaction with Importin α/β nuclear [50]. One possible explanation is that Kaiso, like other proteins or factors that normally reside in the cytoplasm, require a post-translational modification, to be targeted and translocated to the cell nucleus.

However, 2009 data has shown for the first time that the subcellular localization of Kaiso in the cytoplasm of a cell is directly associated with the poor prognosis of patients with lung cancer (non-small cell), and around 85 to 95% of lung cancers are non-small cell [36]. Such data shows a direct relationship between the clinical profile of patients with pathological expression of Kaiso.

Surprisingly in this paper we describe for the first time a relationship between the cytoplasmic Kaiso to CML-BP. An interesting aspect of our results is the relationship between cytoplasmic Kaiso to the prognosis expected in blast crisis (Figure2). At this stage of the disease, many patients died between three and six months, because they are refractory to most treatments. In CML progression to accelerated phase and blastic phase appears to be due mainly to genomic instability, which predisposes to the development of other molecular abnormalities. The mechanisms of disease progression and cytogenetic evolution to blast crisis remain unknown.

The Wnt11 promoter contains two conserved TCF/LEF binding sites (at −43 for each of the alternative first exons) and one Kaiso binding site (at −775 in the human gene), suggesting that both canonical [42] and non canonical Wnt pathways can down regulate Wnt11 transcription directly. Consistent with this, Kaiso depletion strongly increase Wnt11 expression in Xenopus [29]. On the contrary, in K562 cells, upon Kaiso knock-down we observed a significant decrease in the Wnt11 expression (Figure3A–C). A possible explanation of this controversy is that knock-down of Kaiso, increased β-catenin expression (Figure3A), and this is a likely reason for the maintenance of Wnt11 repression in the absence of Kaiso. As is well known, Wnt11 is actually one of several β-catenin/TCF target genes that contain adjacent putative Kaiso and TCF/LEF binding sites in their promoter, suggesting that Kaiso and TCF/LEF cooperate to repress Wnt11transcription [16].

Our results therefore indicate the cooperation between β-catenin/TCF and Kaiso/p120ctn in negative regulation of Wnt11. A common theme among all these studies is that while Wnt11 expression can be regulated by canonical Wnt signals, this regulation is highly dependent on transcription factors in addition to, or other than, TCF/LEF family members, for example, Kaiso/p120ctn.

The novel anticancer agent, imatinib (Glivec, Gleevec, formerly STI571, CGP57148) has proven to be a highly promising treatment for CML. The drug selectively inhibits the kinase activity of the BCR/ABL fusion protein [4, 5].

Although the majority of CML patients treated with imatinib show significant hematologic and cytogenetic responses, resistance to imatinib is clearly a barrier to successful treatment of CML patients. In some patients, resistance arises due to powerful selective pressure on rare cells that carry amplified copies of the BCR-ABL fusion oncogene or point mutations in the BCR-ABL tyrosine kinase domain that affect binding of the drug to the oncoprotein. However, in a proportion of patients neither mechanism operates, and resistance appears to be a priori, existing prior to exposure to the drug [51]. These mechanisms of imatinib resistance are poorly understood and heterogeneous involving largely BCR-ABL independent mechanisms [51, 52].

Our results show that imatinib-resistant K562 cells has a weak expression of Kaiso in the cytoplasm and with a similar phenotype, but not identical, to Kaiso knock-down cells (Figure1B–D). This result suggests the down regulation of Kaiso as a mechanism of resistance to imatinib. Obviously cannot rule out that weak expression in the imatinib-resistant K562 cell line, is a secondary effect involving other genes that lead to transcriptional and translational repression of Kaiso. So far, no proteomics studies, using high-throughput technologies, identified Kaiso as a gene potentially involved in the acquisition of resistance to imatinib [53, 54].

The result shows a global change affecting the expression of several genes important in hematopoietic differentiation and proliferation, coherently with the genome-wide transcriptional response to Kaiso, characterized during early vertebrate development [55]. Thus, all the changes produced by siRNA indicate a trend towards improvement of cell proliferation and blocks of granulocytic differentiation.

The knock-down of either Kaiso or p120ctn alone or in combination decreased C/EBPα and PU-1 (Figure6A and C) and increased significantly SCF expression (Figure5). The transcription factor CCAAT/enhancer binding protein α (C/EBPα) is a strong inhibitor of cell proliferation [56]. Accordingly we found that in all transfections, C/EBPα levels were reduced by 56-80%, when compared with scrambled knock-down cells.

On the other hand, the transcription factor PU.1 is a hematopoietic lineage- specific ETS (E rythroblast T ransformation S pecific) family member [57] that is absolutely required for normal hematopoiesis [58]. The level of PU.1 expression is critical for specifying cell fate, and, if perturbed, even modest decreases in PU.1 can lead to leukemias and lymphomas [57, 59‐64]. Coherently, our results showed that the PU-1 levels decreased by 57-66% when either Kaiso or p120ctn alone or in combination levels were decreased by siRNA (Figure6).

An important aspect of our analysis is that recent data show a system of autocrine and paracrine activation of c-kit (CD117) by SCF [65]. These mechanisms stimulate the growth of Merkel cell carcinoma in vitro. Analysis of the expression of c-kit on the surface of K562 cells showed a small but significant reduction of the CD117 receptor expression in cells with knock-down of either Kaiso or p120ctn alone or in combination (Figure7). On the other hand, Kaiso/p120ctn double knock-down led to a significant 100 fold increase in SCF expression, important for cell survival and proliferation (Figure5). These results could represent an indirect evidence of autocrine and paracrine stimulation of c-kit in K562 cells and justify the effect on cell proliferation produced by Kaiso/p120ctn double knock-down.

Recent studies demonstrate that Kaiso and N-CoR (interacts directly with Kaiso in the nucleus) have important roles in neural cell differentiation [35, 66, 67]. Also, the POZ-ZF subfamily member BCL6 represses several genes that are necessary for the terminal differentiation of B-lymphocytes [68]. But there is no evidence to support the participation of Kaiso in the hematopoietic differentiation.

Our results showed that knock-down of Kaiso decreased CD15 by 35%, indicating that, reduced expression of Kaiso, can block differentiation of the granulocytic program (Figure7D). We also analyzed the levels of Wnt11, C/EBPα and c-MyB and the results in Figure6 show that the expression of Wnt11 and C/EBPα were also reduced and the expression of c-MyB was increased, which is consistent with the Kaiso contribution to the hematopoietic differentiation.

A major role for Wnt11 in vivo is its ability to promote differentiation, for example, stimulating cardiac differentiation of mouse embryonic carcinoma P19 cells [69], and promoting differentiation of many different types of cells [70‐72]. Moreover, Wnt11 promote the differentiation of QCE6 cells into red blood cells and monocytes at the expense of macrophages [43], suggesting that Wnt11 can modulate hematopoietic stem cell diversification. Thus, the knock-down of Kaiso decreased Wnt11 levels by 78% (Figure3), consistent with the role of Kaiso in the hematopoietic differentiation program.

On the other hand, knock-down of Kaiso reduced C/EBPα that is a critical regulator of hematopoietic stem cell homeostasis and myeloid differentiation [73, 74]. The events leading to the loss of C/EBPα function facilitate leukemogenesis by blocking granulocytic differentiation [75, 76] and coherently the knock-down of Kaiso decreased CD15 used widely as granulocytic marker (Figure7A–D).

Interestingly, in vitro experiments have shown that constitutive overexpression of c-Myb blocks differentiation of myeloid and erythroid cells and the associated growth arrest that occurs with maturation. However, c-myb-antisense-treated HL-60 cells differentiated only into monocytes but not into granulocytes indicating that granulocytic differentiation, unlike monocytic differentiation, requires c-myb-mediated proliferation. Consistent with this, an increase expression of c-MyB resulted in a significant decrease in expression of CD15 in K562 cells transfected with siRNA-Kaiso (Compare Figures 6A and 7A). Finally, the myeloid commitment of hematopoietic progenitors is characterized by the progressive loss of CD34 expression accompanied by the acquisition of CD33 expression at high levels. The knock-down of Kaiso led to a significant decreased by 8% in CD33 expression (Figure7).

These findings provide a comprehensive picture of the changes in proliferation, differentiation, and global gene expression that underlie of the pivotal role of cytoplasmic Kaiso in the blast crisis.