Decreased SATB1 expression promotes AML cell proliferation through NF-κB activation

Background Special AT-rich sequence-binding protein 1 (SATB1) is a chromatin-remodeling protein that regulates gene expressions in different types of cancer. Up-regulation of SATB1 is linked with progression of tumors. Our previous study showed that SATB1 expression was decreased in T cell leukemia/lymphoma. The contrary roles of SATB1 in solid organ tumors and hematology malignancy may provide hints to study the function of SATB1. Methods To characterize SATB1 mRNA and protein expression in acute myeloid leukemia (AML), we performed qRT-PCR and Western blot on bone marrow mononuclear cells from 52 newly diagnosed AML patients. Stable HL-60 cell lines with knockdown of SATB1 by shRNAs sequences (HL-60 SATB1-shRNA1 and HL-60 SATB1-shRNA2) were established. Cell proliferation, cell cycle and cell invasiveness were analyzed. Murine model was established using HL-60 SATB1-shRNAs treated nude mice and tumorigenicity was compared to study the role of SATB1 in vivo. Global gene expression profiles were analyzed in HL-60 cells with SATB1 knockdown to investigate the mechanisms underlying the regulation of AML cell growth by SATB1. Results We found that SATB1 expression was significantly decreased in patients with AML compared to normal control, and was increased after complete remission of AML. Knockdown of SATB1 enhanced the proliferation of HL-60 cells and accelerated S phase entry in vitro, and promoted the tumor growth in vivo. Global gene expression profiles were analyzed in HL-60 cells with SATB1 knockdown and the differentially expressed genes were involved in NF-κB, MAPK and PI3 K/Akt signaling pathways. Nuclear NF-κB p65 levels were significantly increased in SATB1 depleted HL-60 cells. Conclusions Decreased SATB1 expression promotes AML cell proliferation through NF-κB activation. SATB1 could be a predictor for better response to treatment in AML.


Background
Acute myeloid leukemia (AML) is a clinically and genetically heterogeneous disease that is fatal for over 80% of patients, particularly for those older than 60 years of age, who have less than 5% to 10% 5-year survival [1][2][3]. Unlike the booming development of therapies for lymphoid malignancies, myeloid drug development lagged behind due to less progress in understanding the genetic basis and pathophysiology of AML. For years, recurrent genetic abnormalities or some genetic mutations were found to clearly interact to drive the initiation, progression and relapse of AML. For example, AML with mutations of FLT3-ITD, c-Kit, TP53, etc. often has poor response to chemotherapy, high relapse rate and aggressive progression [2][3][4][5][6][7][8]. Chemotherapy for AML remained largely unchanged for decades, clinical response of targeted therapy is limited and the survival improvements over time were attributable mostly to hematopoietic stem cell transplantation (HSCT) [2,5,9]. New chromosomal aberrations, epigenetic changes and mutations of AML continue to be reported, and risk stratification is now based on genetic, genomic, and molecular characteristics which help determine the effect on prognosis, as well as provide new insights into the mechanisms of AML.
Special AT-rich sequence-binding protein 1(SATB1) is a chromatin-remodeling protein that has been shown to play an important role in tumor progression and metastasis. Aberrant expression of SATB1 was reported in various types of cancers, including laryngeal squamous cell carcinoma, gastric cancer and breast cancer [10,11]. High SATB1 expression is usually correlated with high metastatic potential and poor prognosis of cancer [12][13][14][15][16][17]. Contrary to observations in solid tumors, SATB1 appears to suppress the progression of leukemia and lymphoma. SATB1 deficiency causes severe immunodeficiency and multiple defects in T-lineage development [18][19][20][21][22][23]. We previously reported that SATB1 expressions were decreased in T cell leukemia/lymphoma (T-ALL), Knockdown of SATB1 significantly enhanced invasiveness of Jurkat cell in vitro [24]. It is reported that SATB1 binding to the enhancer of Sfpi1 resulted in reduced PU.1 expression in myeloid progenitor cells, which indicated that SATB1 dysfunction is associated with a subset of human AML patients [25]. However, less is known about the association between SATB1 and AML, whether SATB1 is implicated in the development of AML warrants further exploration.
In this study, we investigated the role of SATB1 in AML. We determined differences in SATB1 expression of initial AML and the control samples, paired initial and complete remission (CR) samples to find whether SATB1 played a role in the development of AML. We found that SATB1 expression was decreased in AML patients and increased after these patients received CR. We further studied the roles of SATB1 on AML in vitro and in vivo by shRNAs mediated knockdown of SATB1. We showed that knockdown of SATB1 enhanced the proliferation of AML cells and accelerated S phase entry in vitro and promote tumorigenicity in vivo. We then analyzed the global gene expression profiles to investigate the mechanisms underlying the regulation of AML cell growth by SATB1 (part of the results were published as abstract in the 59th ASH annual Meeting: Blood 2017 130:1238).

Patients and clinical characteristics
Seventy-three AML patients from Oct 2015 to Jun 2017 and 38 healthy donors as control were enrolled in this study. Characteristics of newly diagnosed patients are described in Table 1. Peripheral blood samples from

Plasmid Construction and Retroviral Infection
After synthesis of the oligoduplexes, the GV248.puro plasmid vector was digested with AgeI and EcoRI restriction endonucleases. Next, the SATB1-specific shRNAs were cloned into the GV248.puro vector. The successful plasmid construction was verified by DNA sequencing. Production of lentivirus was performed according to the instructions; HL-60 cells were subjected with infection of lentivirus expressing SATB1-shRNA1 and SATB1-shRNA2 or GV248 (empty vector). HL-60 SATB1-shRNA and HL-60-CTR cells were cultured with puromycin for 3 days to produce stable SATB1 knockdown cell line. Targeted SATB1 sequence is as follows: SATB1-shRNA1, GGA TTT GGA AGA GAG TGT C; and SATB1-shRNA2, GTC CAC CTT GTC TTC TCT C.

Cell Proliferation assay
HL-60 SATB1-shRNA1 cells, HL-60 SATB1-shRNA2 cells and HL-60 CTR cells were planted at a density of 2000 cells/well in 96-well plates and maintained at 37 °C in humidified 5% CO2 incubator for 24 h. 10 µl of the Cell Counting Kit-8 (CCK8, Dojindo, Japan) solution was added per well and the plate was incubated for 4 h, the absorbance at 450 nm was measured on a microplate reader.

Cell invasion assay
The invasive potential of HL-60 SATB1-shRNA1 cells, HL-60 SATB1-shRNA2 cells was examined using transwell inserts fitted with polycarbonate filters (8-μm pore size, Costar, Cambridge, MA) coated with matrigel (BD Biosciences, Bedford, MA). Matrix solutions within transwells were polymerized at 37 °C for 1 h and dried onto the transwells overnight at room temperature. Cells were seeded in the upper compartment without FBS medium while lower wells contained 10% FBS medium. After 48 h of incubation, the cells in the upper chamber were removed while other cells, which had passed through the filter, were fixed with 3.7% paraformaldehyde, stained with 0.2% crystal violet, counted, and captured at ×100 and ×400 magnifications using the camera on inverted microscope. Contents of the lower compartments were collected, and migrated cells were also counted.

Cell Cycle Analysis
Cell cycle distribution of HL-60 SATB1-shRNA1 cells, HL-60 SATB1-shRNA2 cells and HL-60 CTR cells was determined by flow cytometric analysis. Cells were resuspended into 5 × 10 5 cells/ml and were collected for nuclear staining which was performed according to the manufacturer's instructions using Flow Cytometry Analysis of Cell Cycle Kit (GENMED, Shanghai). Following staining, cells were immediately analyzed by flow cytometry.

Nuclear and Cytoplasmic Proteins Extraction
The nuclear and cytoplasmic proteins were extracted using an Nuclear

Gene expression profiling and the KEGG pathway analysis
Integrity of total RNA was checked using the Agilent 2100 Bio-analyzer (Agilent, Santa Clara, USA). Gene expression profiling was performed by the Dana-Farber Microarray Core Facility using Affymetrix U133 Plus 2.0 chip (Affymetrix, Santa Clara, USA) following the manufacturer's guidelines, and data was analyzed by DAVID software. Gene expression was compared between HL-60 SATB1-shRNA1 cells and HL-60 CTR cells and analyzed by KEGG Pathway. Genes that displayed two-fold upregulation or 0.5-fold down-regulation in expression in comparisons were selected for further analyses. Arrays with poor quality according to the manufacturer's recommendations were excluded from further analysis.

Statistical analysis
The data were reported as mean ± S.D. Differences among three groups were determined by analysis of Mann-Whitney U, whereas differences between two groups were evaluated by the Student's t test. Chi square test was done to study the association of SATB1 expression with multiple factors. P values less than 0.05 were considered statistically significant. Statistical analysis was performed by the SPSS 19.0 (SPSS Inc., Chicago, USA) statistical software programs.

SATB1 expression was decreased in AML patients
To evaluate the association between SATB1 and AML, SATB1 expression was detected in 52 newly diagnosed AML patients and the normal control. Results showed that SATB1 mRNA expression was significantly lower in AML samples (P < 0.0001, Fig. 1a). In 52 newly diagnosed AML patients, 29 patients received CR after one or two cycles of chemotherapy. We then evaluated the correlations between SATB1 and AML status. SATB1 expression was increased in CR patients compared to their initial samples (P = 0.0306, Fig. 1b), indicating that SATB1 could be a predictor for better response to treatment. According to the average expression of SATB1 of 52 AML patients, patients were then categorized into SATB1 high and SATB1 low , of which the CR rate was 76.47% and 56.14%, respectively (P = 0.1891, Table 1). Next, we analyzed the correlation between SATB1 expression and FAB category to assess whether low SATB1 expression is indicated in some specific subtypes of AML with poor clinical outcome. We observed no difference among FAB subtypes of AML regarding SATB1 expression which also has not been related to age, sex, blood cell count and percentage of blast (Fig. 1c).

Knockdown of SATB1 enhanced the proliferation of AML cells
To define the association between SATB1 expression and the development of AML, we evaluated SATB1 expression in HL-60, NB4, HEL, THP-1 cells and samples from healthy donors as control. Results showed that SATB1 expression was decreased in all AML cell lines than in control (P < 0.0001, Fig. 1d). SATB1 exhibited highest expression in HL-60 cells among 4 different cell lines (P < 0.0001, Fig. 1e, f ), we investigated its role in HL-60 by shRNAs mediated knockdown of SATB1. As shown in Fig. 2a, b, C HL-60 SATB1-shRNA1 and SATB1-shRNA2 cells exhibited decreased SATB1 expression compared to the control (HL-60-CTR) both in the mRNA (P = 0.0286, Fig. 2 SATB1 knockdown led to a progressive increase in the proliferation of HL-60 cells and accelerated S phase entry, but did not increase the invasiveness of AML cells. a SATB1 mRNA expression was decreased in HL-60 SATB1-shRNA1 and SATB1-shRNA2 cells (P = 0.0286, P = 0.0017). b, c Decreased SATB1 protein levels were confirmed by western blotting (P < 0.0001, P = 0.0002). d CCK assay was performed and the absorbance at 450 nm was measured. Increased proliferation of HL-60 cells was detected from day 2 to day 5. e Knockdown of SATB1 Cell cycle was analyzed by flow cytometry in SATB1-shRNA1, 2 and HL-60 CTR group. f The percentage of cells in SATB1-shRNA1/2 was reduced in the G0/G1 phase but increased in S phase (P = 0.0045, P < 0.0001). g Cell Invasion was analyzed by transwell assay and the result showed that down regulation of SATB1 did not increase the penetration of AML cells through the matrigel-coated membrane (P = 0.2134) P = 0.0017) and protein levels (P < 0.0001, P = 0.0002), suggesting that the stable SATB1 knockdown cells were successfully established. To define the role of SATB1 in HL-60 cells, we performed CCK8 assay to evaluate the proliferation of HL-60 SATB1-shRNA cells. SATB1 knockdown led to a progressive increase in the proliferation of HL-60 cells both in SATB1-shRNA1 and SATB1-shRNA2 groups since day2 (P < 0.0001, Fig. 2d).

Down regulation of SATB1 accelerated S phase entry but did not increase the invasiveness of AML cells
To study the effect of SATB1 on AML cell cycle, we compared the number of quiescent and actively cycling HL-60 cells in SATB1 shRNA1, SATB1 shRNA2 and CTR group using flow cytometry. The number of quiescent cells in the G0/G1 phase of the cell cycle was significantly reduced in the absence of SATB1 (P = 0.0045). Consistently, the number of cells in S phase was increased (P < 0.0001), which suggested that SATB1 promoted quiescence of AML cells while knockdown of SATB1 enhanced cell activation (Fig. 2e, f ). In order to further evaluate the impact of SATB1 on the invasiveness of AML cells, we cultured HL-60-CTR cells and HL-60 SATB1-shRNA cells in transwell matrigel-coated chambers under the same conditions and compare the invasiveness between both cell lines. We found that shRNA mediated knockdown of SATB1 did not cause significant increase of penetration through the matrigelcoated membrane, which showed similar invasiveness as HL-60-CTR cells (Fig. 2g, P = 0.2134).

Knockdown of SATB1 promoted tumorigenicity in nude mice
To further test if SATB1 knockdown could promote AML progression in vivo, we compared the tumor size and weight in HL-60 SATB1-shRNA1 cells treated mice and the CTR mice. Mice were injected subcutaneously with 1 × 10 7 HL-60 SATB1-shRNA1 cells (N = 6)) or HL-60 cells (CTR, N = 6) and all mice were executive 3 weeks after injection (Fig. 3a). Tumor was formed in 4 SATB1-shRNA1 mice and 4 CTR mice at the average time of 9 ± 1 day and 10 ± 1 day, respectively. Tumor size was compared from day 9 to day 16 and SATB1 shRNA mice showed a larger tumor volume compared to CTR mice (Fig. 3b). At the end of the experiment, tumor weight was measured and showed a significant increase in SATB1 shRNA mice compared to CTR mice (P = 0.0286, Fig. 3c).

Fig. 3
Knockdown of SATB1 promoted tumorigenicity in nude mice. a HL-60 SATB1-shRNA1 and HL-60 CTR cells (1 × 10 7 ) were injected subcutaneously in SATB1-shRNA1 and CTR mice, respectively. Tumor was formed in 4 mice in each group. All mice were executive 3 weeks after injection. b Tumor size was measured everyday and the result showed that tumor volume of SATB1-shRNA1 was larger compared to CTR mice. c Tumor weight showed a significant increase in SATB1 shRNA mice (P = 0.0286)

SATB1 regulated AML cell growth via the activation of NF-κB signaling pathway
To investigate the mechanisms underlying the enhanced proliferation of HL-60 SATB1-shRNA cells, we analyzed the global gene expression profiles of HL-60 SATB1-shRNA cells and HL-60-CTR cells. Genes were identified as differentially expressed if their expression showed at least two-fold difference in each of the comparisons. Using these criteria, we identified 862 downregulated and 377 upregulated genes in HL-60 SATB1-shRNA cells (Fig. 4a). Downregulated genes were involved in the regulation of cell cycle, proliferation, apoptosis and cell death. Upregulated genes were involved in chromatin assembly and disassembly, protein-DNA complexes assembly, cellular cation homeostasis and inflammatory response. KEGG based analysis was used to determine whether particular biological pathways gene expression was more highly represented in HL-60 SATB1-shRNA cells versus HL-60-CTR cells. Differentially expressed genes were involved in NF-κB, MAPK and PI3 K/Akt signaling pathways. We therefore assessed the protein levels of p38 MAPK, phospho-p38 MAPK, Akt, phospho-Akt and total NF-κB p65 in both SATB1-shRNA and CTR cells.
Results showed that total NF-κB p65 levels were significantly increased in HL-60 SATB1-shRNA cells (Fig. 4b, c). Since nuclear and cytoplasm NF-κB p65 play different roles in cell proliferation, we next compared both nuclear and cytoplasm NF-κB p65 levels among groups. Nuclear NF-κB p65 levels were significantly increased in HL-60 SATB1-shRNA cells, while cytoplasm protein levels showed no difference (Fig. 4d). Together, these results indicate that shRNA mediated knockdown of SATB1 regulated AML cell growth via the activation of NF-κB signaling pathway.

Discussions
SATB1 is a global genomic organizer that participates in activation and inactivation of genes and the differentiation of embryonic stem cells, thus is implicated in a variety of cancers and cancer progression [4,5,15,17,22,[29][30][31]. Up-regulation of SATB1 was found in many epithelial tumors and negatively correlated with the prognostic properties and clinical outcome [6,15,32]. Recent studies in cells with SATB1 knockdown have emphasized the functional relevance of SATB1 to carcinogenesis [17,32,33]. The mechanism by which SATB1 regulate cancer development remains unclear and has been extensively studied. It's reported that SATB1 might cooperatively regulate expression of the anti-apoptotic BCL2 and pro-apoptotic NOXA genes to participate in disruption of apoptosis, which is a hallmark of cancer that reinforces tumorigenesis and resistance to cytotoxic cancer therapies [6,34]. However, SATB1 expression was decreased rather than increased in hematology malignancy. We previously reported that SATB1 expressions were up-regulated in T-ALL. Knockdown of SATB1 have enhanced invasive potential in Jurkat cells [24]. These observations raised interesting questions of whether SATB1 play the similar role in AML as in T-ALL or in epithelial tumors and whether it could be one of the potential treatment targets for AML. In this study, our data showed that SATB1 expression was deregulated in 52 initial AML samples, which is in line with our own findings on T-ALL. By comparing SATB1 expression of pair initial AML and CR patients, we found that SATB1 expression was increased in CR patients. Although the CR rate was not significantly increased as expected as SATB1 expression was up-regulated, it's higher in SATB-1 high patients and the difference could be statistical significant if the sample size is larger. We then confirmed decreased SATB1 expression in different AML cell lines and found that SATB1 was more specific in HL-60 cells. These data suggested that down-regulation of SATB1 could be an important factor in leukemogenesis. SATB1 is predominantly expressed in the thymus and highly associated with thymocyte differentiation. It is thought to induce lymphopoiesis and regulate the expression of various genes during lymphocyte especially T cell development, so SATB1 is thought to be correlated to the development of ALL [18-22, 28, 31, 35-40]. SATB1 regulates many genes involved in HSC lineage decisions and development, but evidence that supports the role of SATB1 in AML remained elusive [41,42]. It is reported that SATB1-deficient HSCs were found to be less quiescent and more prone to differentiate to myeloid-erythroid lineages, which indicated that SATB1 is associated with the regulation of myeloid lineages and its dysfunction could be linked to AML [1-6, 25, 29].
One aim of this study was to describe the association between SATB1 and AML. SATB1 was knockdown in the selected HL-60 cell line. We observed that knockdown of SATB1 resulted in a progressive proliferation of HL-60 cells. HL-60 SATB1-shRNA cells were then injected subcutaneously in mice. As expected, progression of leukemia was proved by significant increased tumor weight, suggesting that SATB1 knockdown could increase AML proliferation both in vitro and in vivo. In addition, cell cycle analysis was then performed and we showed that down regulation of SATB1 could promote cell activation by accelerating S phase entry. These results suggest an indispensable role of SATB1 in the development of AML. However, we did not observe a significant increased in the invasiveness of AML cells after knockdown of SATB1. It is difficult to make a conclusion that SATB1 is dispensable for the invasiveness of AML because AML arise as a consequence of a number of genetic aberrations and multiple molecular events; the invasiveness of different subtypes of AML varied and results from only one AML cell line can not tell the whole story. More detailed analysis is necessary for the mechanisms underlying the enhanced proliferation of AML cells by SATB1.
SATB1 regulates multiple genes in lymphocyte differentiation. It is reported that SATB1 interacted with β-catenin which acts as a transcriptional regulator mediating Wnt signaling thus inducing target gene expressions [24,36,43]. SATB1 also associate with epigenetic modifiers such as histone deacetylase which is important for lymphocyte development [20]. However, what target genes and the associated signal pathways are possibly involved in the regulation of AML by SATB1has not been clarified. In this study, we compared the global gene expression profiles of HL-60 SATB1-shRNA cells and HL-60-CTR cells. For those differentially expressed genes which were involved in NF-κB, MAPK and PI3K/Akt signaling pathways, we assessed the protein levels and found that nuclear NF-κB p65 levels were significantly increased in HL-60 SATB1-shRNA cells, while cytoplasm protein levels showed no difference when compared to HL-60-CTR cells. These findings suggested that knockdown of SATB1 might regulate AML cell growth via the activation of NF-κB signaling pathway.

Conclusions
We demonstrated in this study that impairment of SATB1 is associated with AML. Knockdown of SATB1 promoted cell proliferation and tumorigenicity of AML via the activation of NF-κB signaling pathway. Further study about gene regulatory network modulated by SATB1 may lead to the development of a new treatment of AML.