Specific protein 1, c-Abl and ERK1/2 form a regulatory loop

c-Abl (Abelson tyrosine kinase, Abl) is a non-receptor tyrosine kinase that plays an essential role in regulating a variety of cellular functions, including cell adhesion, migration, cytokinesis and smooth muscle contraction (Anfinogenova et al., 2007; Chen and Tang, 2014; Cleary et al., 2014; Hu et al., 2005; Wang et al., 2016; Wang, 2014). Furthermore, c-Abl promotes the proliferation of smooth muscle cells (Jia et al., 2012; Wang et al., 2013a) and fibroblasts (Mitra et al., 2008) during stimulation with growth factors. c-Abl has also been implicated in airway smooth muscle layer thickening in asthma (Cleary et al., 2013; Rhee et al., 2011; Tang, 2015). Its mutant form BCR-Abl contributes to the development of leukemias, such as chronic myeloid leukemia (Wang, 2014).

The expression level of c-Abl is associated with various pathological and physiological conditions including cancer, pregnancy and asthma (Tang, 2018; Tang and Gerlach, 2017; Weigel et al., 2013; Yaba et al., 2011). Although the roles of c-Abl in cell functions and disease pathogenesis have been extensively investigated, fundamental knowledge regarding the regulation of its expression is surprisingly limited. Recent studies have suggested that c-Abl expression is regulated in part by the microRNA (miR)-203 in T cell lymphoma (Bueno et al., 2008) and smooth muscle cells (Liao et al., 2015). The mechanisms that regulate c-Abl transcription in human cells are largely unknown.

Specific protein 1 (Sp1) is a founding member of the Sp1 transcription factor family, which participates in the regulation of gene expression. Sp1 was initially identified as a promoter-specific binding factor that is essential for transcription of the SV40 major immediate early (IE) gene (Dynan and Tjian, 1983). For many years, Sp1 was thought to be a basal transcription factor that regulates expression of housekeeping genes, and it has been implicated in tumor cell proliferation, differentiation, apoptosis and angiogenesis (Beishline and Azizkhan-Clifford, 2015). The DNA-binding domains of Sp1 possess three highly homologous zinc fingers, which preferentially bind to the same GC consensus site [5′-GGGCGG-3′] in target promoters and regulate transcription of housekeeping genes (Beishline and Azizkhan-Clifford, 2015; Nagaoka et al., 2001). Moreover, Sp1 phosphorylation at Thr-453 and Thr-739 has been implicated in modulating its transcriptional activity in SL2 Drosophila cells (Milanini-Mongiat et al., 2002). However, how Sp1 is regulated in human cells is still elusive.

The mitogen-activated protein kinase (MAPK) pathway is critical for the regulation of various cellular functions including gene expression (Jia et al., 2012; Wang et al., 2013a; Widmann et al., 1999). Upon stimulation with growth factors within minutes, ERK1 and ERK2 (ERK1/2, also known as MAPK3 and MAPK1, respectively) become phosphorylated and activated, which eventually can regulates gene expression and cell proliferation (Chang and Karin, 2001; Jia et al., 2012; Wang et al., 2013a; Widmann et al., 1999). Interestingly, c-Abl can regulate ERK1/2 activation in smooth muscle cells (Jia et al., 2012; Wang et al., 2013a) and fibroblasts (Mitra et al., 2008) upon activation by growth factors.

In this study, we used the PROMO online tool (http://alggen.lsi.upc.edu/recerca/menu_recerca.html; for prediction of transcription factor-binding sites) to analyze potential transcription factor-binding sites on c-Abl promoter and found that there are five Sp1-binding sites on the essential region of the c-Abl promoter. Moreover, we used loss-of-function and rescue approaches to evaluate the role of Sp1 in c-Abl expression. We discovered that Sp1 regulates c-Abl promoter activation, c-Abl expression and cell proliferation. Furthermore, c-Abl conversely controls ERK1/2 activation and Sp1 expression.

Because c-Abl is one of the major players in regulating smooth muscle cell proliferation during activation with growth factors (Liao et al., 2015; Tang, 2015; Tang and Gerlach, 2017; Wang et al., 2013a), we questioned whether PDGF affects c-Abl expression in cells. Human airway smooth muscle (HASM) cells were treated with 10 ng/ml PDGF for 24 h, and immunoblot analysis was used to assess c-Abl protein expression. The protein level of c-Abl in PDGF-treated cells was higher than in untreated cells (Fig. 1A). These results suggest that treatment with PDGF increases the expression of c-Abl protein in HASM cells. Because ERK1/2 and AKT (also known as AKT1) are also phosphorylated and activated upon growth factor stimulation (Jia et al., 2012; Liao et al., 2015; Wang et al., 2013a), we evaluated the effects of PDGF stimulation on ERK1/2 and AKT phosphorylation. Phosphorylation levels of ERK1/2 and AKT were higher in cells treated with PDGF as compared to untreated cells (Fig. 1B,C). Because PDGF treatment increases cell number, we evaluated whether cell density affects PDGF-dependent c-Abl expression. Cells with different densities were treated with PDGF for 24 h. c-Abl protein expression was higher in PDGF-treated cells with 25%, 50% and 75% densities (Fig. 1D).

We also determined the role of PDGF in cell proliferation. HASM cells were treated with 10 ng/ml PDGF for 24 h. DNA synthesis in the cells was evaluated using the BrdU incorporation assay (Jiang and Tang, 2015). BrdU incorporation was higher in cells treated with PDGF than in control cells (Fig. 1E). In addition, exposure to PDGF increased cell number (Fig. 1F). The results are consistent with previous results that PDGF promotes cell proliferation (Ambhore et al., 2018; Liao et al., 2015; Wang et al., 2013a).

We used the PROMO online tool to analyze potential transcription factor-binding sites and found that there were five Sp1-binding sites on the essential region of c-Abl promoter (NCBI Reference Sequence: NG_050744.1) (Fig. 2A). This raised a possibility that Sp1 may regulate c-Abl transcription and expression. HASM cells were transfected with either control siRNA or Sp1 siRNA for 24 h. They were then cultured in medium containing 0.25% fetal bovine serum (FBS) (to prevent cell death) with or without PDGF for additional 24 h. Immunoblot analysis verified that Sp1 knockdown (KD) occurred in cells treated with Sp1 siRNA for a total of 48 h (Fig. 2B). Moreover, Sp1 KD did not affect the expression of proliferating cell nuclear antigen (PCNA) and vimentin (Fig. 2B), two proteins implicated in cell proliferation (Cheng et al., 2016; Cleary et al., 2013). The results indicate that Sp1 siRNA selectively inhibits Sp1 expression. More importantly, Sp1 KD inhibited the PDGF-induced c-Abl expression at the protein level by ∼40% and mRNA level by ∼50% (Fig. 2C). Because Sp1 is critical for cell proliferation (Beishline and Azizkhan-Clifford, 2015), complete knockdown of Sp1 can impede cell viability. As such, we used the cell model in which Sp1 had a partial, although still substantial, downregulation, for the following experiments. In addition, Sp1 KD reduced c-Abl expression in cells not treated with PDGF (Fig. 2C). This is not surprising because cells were cultured in medium with a low concentration (0.25%) of FBS. These results suggest that Sp1 regulates both basal and PDGF-treated c-Abl expression.

Furthermore, we performed the KD and rescue experiment to test the specificity of Sp1 siRNA. HASM cells were treated with control siRNA, Sp1 siRNA and Sp1 siRNA plus RNAi-resistant Sp1 construct for 2 days. Immunoblot analysis verified Sp1 protein expression in the Sp1 KD and rescue cells (Fig. 2D). More importantly, rescue of Sp1 in the KD cells restored c-Abl protein and mRNA expression (Fig. 2D). Taken together, the findings indicate that Sp-1 promotes c-Abl expression in cells.

HASM cells were treated with either control siRNA or Sp1 siRNA for 24 h followed by PDGF treatment for an additional 24 h. DNA synthesis was reduced in cells treated with Sp1 siRNA, as revealed by a BrdU incorporation assay (Fig. 2E). Moreover, Sp1 KD diminished basal and PDGF-induced cell proliferation (Fig. 2F). Furthermore, Sp1 was located in the nucleus of smooth muscle cells as evidenced by immunfluorescence analysis (Fig. 2G).

Next, we determined whether PDGF stimulation affects c-Abl promoter activity. HASM cells were transfected with c-Abl promoter reporter (Fig. 3A) for 48 h followed by PDGF stimulation for 24 h. Luciferase activity of the c-Abl promoter in the cells was then determined. PDGF exposure increased c-Abl promoter activity (Fig. 3B), indicating a role for PDGF in activating the c-Abl promoter. Because there are five Sp1-binding sites on the essential region of c-Abl promoter (Fig. 2A), we determined the role of Sp1 in c-Abl promoter activation. HASM cells were co-transfected with the c-Abl promoter reporter, and with either control siRNA or Sp1 siRNA for 48 h followed by PDGF activation for 24 h. Basal and PDGF-induced luciferase activity of the reporter in Sp1 KD cells was reduced compared to that in control cells (Fig. 3C). These results indicate that Sp1 has a positive role in regulating c-Abl promoter activity.

To assess the role of the Sp1-binding sites in c-Abl promoter activation, we mutated the five GC-rich sites from GGGCGG to GTTCGG. Cells were transfected with wild-type or mutant c-Abl promoter reporters plus Sp1 construct and cultured for 72 h. The luciferase activity of the promoter with mutated Site 1, 4 or 5 was significantly reduced compared with that of the wild-type promoter (Fig. 3D). In contrast, the luciferase activity of the promoter with mutated Site 2 or 3 was not diminished significantly (Fig. 3D). Furthermore, the luciferase activity of triple mutant (Site 1/4/5) is reduced by 80% (Fig. 3D). The results suggest that Site 1, 4 and 5 are important for c-Abl promoter activation and not all binding sites on the promoter are necessary for Sp1-mediated activation. The results are not surprising because not all binding sites are accessible to transcription factors in certain cell types (Soufi et al., 2015).

To assess whether PDGF affects the interaction of Sp1 with endogenous c-Abl promoter, smooth muscle cells were treated with PDGF for 24 h. Chromatin immunoprecipitation (ChIP) analysis was performed using Sp1 antibody. In cells treated with PDGF, the amount of c-Abl promoter fragment precipitated with Sp1 was higher in cells treated with PDGF than in control cells (Fig. 3E). Quantification analysis showed that the Sp1-precipitated DNA was enhanced in cells treated with PDGF; however, IgG-precipitated DNA was not altered after PDGF treatment (Fig. 3F).

ERK1/2-mediated Sp1 phosphorylation at Thr-453 and Thr-739 has been shown to promote Sp1 transcriptional activity in SL2 Drosophila cells (Milanini-Mongiat et al., 2002). Previous studies have shown that PDGF exposure activates ERK1/2 in various cell types including smooth muscle cells (Jia et al., 2012; Wang et al., 2013a). Our current results showed that PDGF-induced ERK1/2 phosphorylation was increased as early as 5 min after stimulation, and slightly reduced and sustained for 24 h (Fig. 4A,B). Furthermore, we assessed whether PDGF could increase Sp1 phosphorylation at the two residues. HASM cells were treated with PDGF for different times, and Sp1 phosphorylation at the residues in cells was evaluated by immunoblotting. We unexpectedly found that the phosphorylation of Sp1 at Thr-453 and Thr-739 was not increased in cells treated with PDGF (Fig. 4A,C,D).

Next, we sought to explore whether PDGF affects expression of total Sp1 protein. Cells were treated with PDGF for different times, followed by immunoblotting. PDGF slightly increased the expression of total Sp1 protein in cells 4 h after stimulation, but significantly enhanced Sp1 expression 24 h after treatment (Fig. 4A,E). Moreover, PDGF increased the expression of c-Abl protein 24 h after stimulation (Fig. 4A,F). These results suggest that Sp1 transcriptional activity may be regulated by its protein level rather than its phosphorylation in smooth muscle cells.

As described earlier, ERK1/2 phosphorylation plays an important role in the signaling pathways that control gene expression. We questioned whether ERK1/2 may regulate Sp1 and c-Abl in cells. First, we found that PDGF exposure increased ERK1/2 phosphorylation (Fig. 1B), which is supported by previous studies (Jia et al., 2012; Jiang and Tang, 2015; Liao et al., 2015; Tang, 2015). Immunoblot analysis demonstrated that exposure to U0126 (an inhibitor of MEK proteins, which are upstream of ERK1/2) inhibited ERK1/2 phosphorylation in smooth muscle cells (Fig. 5A). However, treatment with U0126 did not reduce the amount of total ERK1/2 (Fig. 5A). Therefore, it is unlikely that U0126 nonspecifically inhibits all cell functions. To determine the functional role of ERK1/2, HASM cells were treated with 10 µM U0126 with or without PDGF for 24 h. Treatment with U0126 inhibited basal expression of Sp1 and c-Abl at the protein and mRNA level (Fig. 5B–D). More importantly, exposure to U0126 attenuated the PDGF-induced upregulation of Sp1 and c-Abl at the protein and mRNA level (Fig. 5B–D). These results indicate that ERK1/2 has a role in regulating Sp1 and c-Abl expression at the transcription level in cells. Additionally, exposure to U0126 attenuated the basal and the PDGF-induced BrdU incorporation, as well as the amount of basal and PDGF-stimulated cell proliferation (Fig. 5E,F).

c-Abl has been shown to regulate ERK1/2 activation in various cell types including smooth muscle cells (Jia et al., 2012; Mitra et al., 2008; Wang et al., 2013a). Because ERK1/2 regulates Sp1 expression, we speculated that c-Abl may conversely regulate Sp1 expression. We have previously generated and characterized stable c-Abl KD HASM cells (Chen and Tang, 2014; Wang et al., 2013a). To test the potential effect of c-Abl on Sp1, stable c-Abl KD HASM cells and control cells were treated with PDGF for 24 h and assessed by immunoblotting. Sp1 protein expression in c-Abl KD cells was lower than in control cells (Fig. 6A). Furthermore, we also confirmed that c-Abl KD inhibited basal and PDGF-induced ERK1/2 phosphorylation (Fig. 6B). However, c-Abl KD did not inhibit the expression of total ERK1/2 in cells (Fig. 6B).

The roles of c-Abl in cell functions and disease pathogenesis have been extensively characterized. The knowledge regarding its transcriptional regulation is astonishingly limited. In this report, we used bioinformatics and loss-of-function approaches to show for the first time that Sp1 regulates c-Abl transcription, expression and the proliferation in smooth muscle cells. Since c-Abl is ubiquitously expressed in human cells, Sp1-mediated c-Abl transcription could be common.

We used site-directed mutagenesis to disrupt the five GC-rich domains and assessed the luciferase activity of individual mutants. We found that Site 1, 4 or 5 affected the activity of c-Abl promoter. In contrast, Site 2 or 3 did not affect promoter activity significantly. These results imply that not all binding sites are critical for promoter activation. This could be due to inaccessibility of certain binding sites on promoters in vivo (Soufi et al., 2015). In addition, the location of DNA-binding motifs could also influence the interaction of transcription factors with promoters (Westholm et al., 2008).

We also discovered that growth factor activation coordinately promotes the occupancy of Sp1 on the endogenous c-Abl promoter and c-Abl expression in smooth muscle cells. As such, growth factor-mediated c-Abl upregulation is modulated through c-Abl transcription in cells. However, expression of miR-203 (a microRNA that downregulates c-Abl) is reduced in T-cell tumors (Bueno et al., 2008) and asthmatic human airway smooth muscle cells (Cleary et al., 2013; Liao et al., 2015). Moreover, growth factor concentration or signaling is upregulated in asthma (Ammit and Panettieri, 2003; Booy et al., 2011). We do not rule out the possibility that miR-203 reduction may be also involved in growth factor-induced c-Abl expression.

Sp1 has been implicated in the regulation of the vascular endothelial growth factor promoter, and phosphorylation of Sp1 at Thr-453 and Thr-739 by ERK1/2 increases Sp1 promoter activity in SL2 Drosophila cells (Milanini-Mongiat et al., 2002). Because PDGF stimulation increases ERK1/2 phosphorylation (Jia et al., 2012; Wang et al., 2013a) (Fig. 1B), we evaluated the effects of PDGF stimulation on Sp1 phosphorylation. We unexpectedly found that PDGF activation did not induce Sp1 phosphorylation at these two residues. Thus, it is less likely that Thr-453 and Thr-739 have a role in modulating Sp1 activity in this cell type. This discrepancy may stem from different cell types and/or different experimental conditions (growth factors versus inducible Raf-1 activation). Intriguingly, we found that Sp1 total protein was upregulated by PDGF in smooth muscle cells. The time course results suggest that ERK1/2 activation precedes the upregulation of Sp1 and c-Abl. Furthermore, ERK1/2 inhibition attenuated Sp1 expression, c-Abl expression and cell proliferation. These results led us to suggest that ERK1/2 activation may upregulate Sp1 expression, which may subsequently promote c-Abl expression and cell proliferation.

The mechanisms by which ERK1/2 modulates Sp1 expression are currently unknown. ERK1/2 has been shown to phosphorylate and regulate E20 transformation-specific family transcription factors (Roberts and Der, 2007). Thus, ERK1/2 could regulate unidentified transcription factors, which in turn control Sp1 expression. Activated ERK1/2 has also been implicated in non-coding RNA regulation (e.g. miRNA) (Qu et al., 2017). ERK1/2 could control unidentified non-coding RNAs that modify Sp1 expression. Future studies are required to test these possibilities.

Our previous studies have demonstrated that c-Abl promotes ERK1/2 phosphorylation by activating the Raf1–MEK1/2 pathway (Jia et al., 2012; Wang et al., 2013a). In this study, we verified that c-Abl positively regulates ERK1/2 activation. Moreover, c-Abl regulated Sp1 expression in cells. Thus, we propose that c-Abl tyrosine kinase conversely regulates ERK1/2 activation and Sp1 expression in cells during growth factor stimulation.

In conclusion, we provide the first evidence that Sp1 regulates c-Abl tyrosine kinase at the transcription level. We also unveiled a novel transcriptional regulation of c-Abl. In response to growth factor activation, Sp1 increases c-Abl promoter activity, which subsequently enhances c-Abl expression. Conversely, c-Abl regulates ERK1/2 activation and Sp1 expression (Fig. 7). Hence, Sp1, c-Abl and ERK1/2 form a regulatory loop.

Human airway smooth muscle (HASM) cells were prepared from human bronchi and adjacent tracheas obtained from the International Institute for Advanced Medicine (Wang et al., 2014a,b, 2013a,b, 2015). Human lungs were non-transplantable, and informed consented was obtained from all subjects for research. This study was approved by the Albany Medical College Committee on Research Involving Human Subjects. All clinical investigation have been conducted according to the principles expressed in the Declaration of Helsinki. Briefly, muscle tissues were incubated for 20 min with dissociation solution [130 mM NaCl, 5 mM KCl, 1.0 mM CaCl₂, 1.0 mM MgCl₂, 10 mM Hepes, 0.25 mM EDTA, 10 mM D-glucose, 10 mM taurine, pH 7, 4.5 mg collagenase (type I), 10 mg papain (type IV), 1 mg/ml BSA and 1 mM dithiothreitol]. All enzymes were purchased from Sigma-Aldrich. The tissues were then washed with Hepes-buffered saline solution (composition in mM: 10 Hepes, 130 NaCl, 5 KCl, 10 glucose, 1 CaCl₂, 1 MgCl₂, 0.25 EDTA, 10 taurine, pH 7). The cell suspension was mixed with Ham's F12 medium supplemented with 10% (v/v) fetal bovine serum (FBS) and antibiotics (100 units/ml penicillin, 100 µg/ml streptomycin). Cells were cultured at 37°C in the presence of 5% CO₂ in the same medium. The medium was changed every 3–4 days until cells reached confluence, and confluent cells were passaged with trypsin/EDTA solution (Li et al., 2006; Wang et al., 2013a). Smooth muscle cells from four non-asthmatic subjects were used for the experiments. In some cases, duplicated experiments were performed for cells from a donor. Cells were recently authenticated by morphological analysis and immunoblotting/immunostaining of α-actin. No contamination was found for these cells.

Cells were lysed in SDS sample buffer composed of 1.5% dithiothreitol, 2% SDS, 80 mM Tris-HCl pH 6.8, 10% glycerol and 0.01% Bromophenol Blue. The lysates were boiled in the buffer for 5 min and separated by SDS-PAGE. Proteins were transferred onto nitrocellulose membranes. The membranes were blocked with bovine serum albumin or milk for 1 h and probed with use of primary antibodies followed by horseradish peroxidase-conjugated secondary antibodies (ThermoFisher Scientific). Proteins were visualized by enhanced chemiluminescence (ThermoFisher Scientific) using the GE Amersham Imager 600 system.

Anti-total Sp1 antibody (1:1000) was purchased from Invitrogen (#PA529165, L/N RJ2284607). Anti-c-Abl antibody (1: 1000) was purchased from Cell Signaling (#2862S, L/N 15). Anti-Sp1 and anti-c-Abl antibodies were validated by using corresponding knockdown cells. Anti-phospho-Sp1 (Thr-453) antibody (1:500) was purchased from ThermoFisher Scientific (#PA5-19658, L/N RH-2250163) and validated by examining the molecular mass of detected bands. Anti-phospho-Sp1 (Thr-739) antibody (1:500) was purchased from Life Span Biosciences Inc. (#LS-C205086, L/N 59177) and validated by examining molecular mass of detected bands. Anti-total vimentin antibody (1:1000) was custom-made as previously described (Li et al., 2006; Tang et al., 2005) and validated by examining the molecular mass of detected bands. Anti-α-tubulin antibody (1:1000) was purchased from Santa Cruz Biotechnology (#SC-32293, L/N G0114). Anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) antibody (1:1000) was acquired from Santa Cruz Biotechnology (#SC-32233, K0315). Anti-phospho-ERK1/2 antibody (1:1000) was purchased from Cell Signaling (#4370S, L/N 38). Anti-total ERK1/2 antibody (1:1000) was purchased from Cell Signaling (#4695S, L/N 14). Antibodies against α-tubulin, GAPDH, phospho-ERK1/2 and total ERK1/2 were validated by the examining molecular mass of detected bands. Finally, vendors have provided data sheets to show that antibodies were validated by positive controls.

The levels of proteins were quantified by scanning densitometry of immunoblots (Fuji Multigauge Software or GE IQTL software). The luminescent signals from all immunoblots were within the linear range.

For Sp1 knockdown, control siRNA (SC-37007/Lot# B1517) and Sp1 siRNA (SC-29487/Lot# J0616) were purchased from Santa Cruz Biotechnology. HASM cells were transfected with siRNA according to the manual of the manufacturer (Santa Cruz Biotechnology). Stable c-Abl knockdown and corresponding control cells were generated using lentivirus encoding control or c-Abl shRNA and cultured as previously described (Wang et al., 2014a, 2013a, 2018). For protein expression, cells were transfected with pN3-Sp1FL (Addgene #24543) or pEGFP (control, Clontech) using the FuGene HD transfection reagent (Promega).

c-Abl promoter reporter (p-EZX-PG04) was purchased from GeneCopoeia (MD). The reporter uses a dual-reporter system, which uses Gaussia Luciferase (GLuc) as the promoter reporter and SeAP (secreted alkaline phosphatase) as the internal control for signal normalization. Cells were transfected with the reporter plus siRNAs or expression constructs using the FuGene HD transfection reagent. They were then cultured in the growth medium for 48–72 h. The luciferase activity was determined by using the secrete-pair dual luminescence assay kit (Gene Copoeia, #LF032).

Cells were treated with human platelet-derived growth factor (PDGF)-BB (Sigma, 10 ng/ml) in F12 medium containing 0.25% FBS. Additional cells were cultured in the medium with 0.25% FBS as a control. The 5′-bromo-2′-deoxyuridine (BrdU) cell proliferation assay kit (Millipore) was used to evaluate DNA synthesis. BrdU is an analog of thymidine, which is able to be incorporated into newly synthesized DNA. Cells in 96-well plates were treated with BrdU for 24 h. They were fixed and reacted with BrdU antibody followed by incubation with secondary antibody conjugated to peroxidase. They were then reacted with peroxidase substrates. The reaction was detected using a Promega GloMax-Multi Microplate reader. The number of viable cells were evaluated with a Trypan Blue exclusion test. Triplicated samples were averaged for each experiment (Jiang and Tang, 2015).

Total RNA from smooth muscle cells was purified by using the GeneJET RNA Purification Kit (Thermo Scientific) according to the manufacturer's instructions. Reverse transcription was performed using the iScript cDNA Synthesis Kit (Bio-Rad). Quantitative real-time PCR was performed by using the SsoAdvanced Universal SYBR Green Supermix kit (Bio-Rad) and a real-time PCR system (Bio-Rad). For the detection of human c-Abl mRNA, the 5′-primer sequence was 5′-AGCTCTACGTCTCCTCCGAG-3′; the 3′-primer sequence was 5′-CAGCTTGTGCTTCATGGTGA-3′. Human β2-microglobulin (B2M) mRNA was used as a control. The 5′-primer sequence of B2M was 5′-TGCTGTCTCCATGTTTGATGTATCT-3′; the 3′-primer sequence of B2M was 5′-TCTCTGCTCCCCACCTCTAAGT-3′. The expression level of c-Abl mRNA was expressed as the ratio of c-Abl mRNA to that of B2M mRNA.

Different c-Abl promoter reporter mutants were generated by using QuikChange II XL site-directed mutagenesis kit (Agilent Technologies) as previously described (Jiang and Tang, 2015). The construct (p-EZX-PG04) was used as a template. For mutant 1, the sequence of forward primer was 5′-CCAGAGCCGGGAGGTTCGGCGGTGTCCCGGG-3′. The sequence of reverse primer was 5′-CCCGGGACACCGCCGAACCTCCCGGCTCTGG-3′. For mutant 2, the sequence of forward primer was 5′-CCGGACGTCGCCGTGTTCGGGGCCGAGGGCGG-3′. The sequence of reverse primer was 5′-CCGCCCTCGGCCCCGAACACGGCGACGTCCGG-3′. For mutant 3, the sequence of forward primer was 5′-GTGGGCGGGGCCGAGTTCGGGGCCTGGCCTCG-3′. The sequence of reverse primer was 5′-CGAGGCCAGGCCCCGAACTCGGCCCCGCCCACG-3′. For mutant 4, the sequence of forward primer was 5′-CCCCTACCGGCGGGGTTCGGCTGGGTCCCTCGG-3′. The sequence of reverse primer was 5′-CCGAGGGACCCAGCCGAACCCCGCCGGTAGGGG-3′. For mutant 5, the sequence of forward primer was 5′-GATGTGACTGCCTGAGTTCGGTGGTGGTGTCAGC-3′. The sequence of reverse primer was 5′-GCTGACACCACCACCGAACTCAGGCAGTCACATC-3′. Plasmids were purified by using the QIAPrep Spin Miniprep kit (Qiagen, Germany). DNA sequencing was performed by Genewiz.

ChIP assay was performed according to the protocol at Aparicio et al. (2004) with modifications. Briefly, adherent cells were incubated in culture medium containing 1% formaldehyde with gentle shaking for 10 min at room temperature and crosslinking was stopped by addition of 2.5 M glycine to a final concentration of 0.125 M glycine. After two washes with cold PBS, cells were harvested in ice cold lysis buffer (10 mM Tris-HCl pH 8.0, 85 mM KCl, 0.5% NP-40, 5 mM EDTA and fresh proteinase inhibitor cocktail) and incubated on ice for 10 min. Nuclei were collected and suspended in cold RIPA buffer (10 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.1% SDS, 0.1% sodium deoxycholate, 1% Triton X-100, 5 mM EDTA and fresh proteinase inhibitor cocktail). The mixtures were sonicated to shear the genomic DNA to an average of 200–300 bp. Cleared extracts were blocked with protein A/G beads (Upstate Biotechnology) and aliquots of the supernatants were used for immunoprecipitation with the anti-Sp1 antibody or IgG. After seven washes by RIPA buffer with gentle rotation for 5 min each time, the proteins were eluted from the beads by 0.5 ml elution buffer (0.1 M NaHCO₃ and 1% SDS). The DNA samples were recovered by phenol extraction and ethanol precipitation after reversal of crosslinking. The purified DNA was then analyzed by PCR within linear amplification range followed by agarose gel electrophoresis. The sequence of forward primer was 5′-GGGAAAGCGGCTCTTGGG-3′. The sequence of reverse primer was 5′-TCAGGCACAGACACCAAAC-3′.

All statistical analysis was performed using Prism software (GraphPad Software, San Diego, CA). Differences between pairs of groups were analyzed by Student–Newman–Keuls test or the Dunn's method. Comparison among multiple groups was performed by one-way or two-way ANOVA followed by a post hoc test (Tukey's multiple comparisons). Values of n refer to the number of experiments used to obtain each value. Power and Sample Size Calculation software (Vanderbilt University, http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize) was used to determine the sample size. P<0.05 was considered to be significant.

The authors thank Ruping Wang for technical assistance.