Metallothionein III (MT3) is a putative tumor suppressor gene that is frequently inactivated in pediatric acute myeloid leukemia by promoter hypermethylation

Background Acute myeloid leukemia (AML) is the second most common form of leukemia in children. Aberrant DNA methylation patterns are a characteristic feature in various tumors, including AML. Metallothionein III (MT3) is a tumor suppresser reported to show promoter hypermethylated in various cancers. However, the expression and molecular function of MT3 in pediatric AML is unclear. Methods Eleven human leukemia cell lines and 41 pediatric AML samples and 20 NBM/ITP (Norma bone marrow/Idiopathic thrombocytopenic purpura) control samples were analyzed. Transcription levels of MT3 were evaluated by semi-quantitative and real-time PCR. MT3 methylation status was determined by methylation specific PCR (MSP) and bisulfite genomic sequencing (BSG). The molecular mechanism of MT3 was investigated by apoptosis assays and PCR array analysis. Results The MT3 promoter was hypermethylated in leukemia cell lines. More CpG’s methylated of MT3 was observed 39.0% pediatric AML samples compared to 10.0% NBM controls. Transcription of MT3 was also significantly decreased in AML samples compared to NBM/ITP controls (P < 0.001); patients with methylated MT3 exhibited lower levels of MT3 expression compared to those with unmethylated MT3 (P = 0.049). After transfection with MT3 lentivirus, proliferation was significantly inhibited in AML cells in a dose-dependent manner (P < 0.05). Annexin V assay showed that apoptosis was significantly upregulated MT3-overexpressing AML cells compared to controls. Real-time PCR array analysis revealed 34 dysregulated genes that may be implicated in MT3 overexpression and apoptosis in AML, including FOXO1. Conclusion MT3 may be a putative tumor suppressor gene in pediatric AML. Epigenetic inactivation of MT3 via promoter hypermethylation was observed in both AML cell lines and pediatric AML samples. Overexpression of MT3 may inhibit proliferation and induce apoptosis in AML cells. FOXO1 was dysregulated in MT3-overexpressing cells, offering an insight into the mechanism of MT3-induced apoptosis. However, further research is required to determine the underlying molecular details.


Introduction
Acute myeloid leukemia (AML) is a type of cancer in the myeloid cell. It is the most common form of acute leukemia in adults and the second most common form of leukemia in children after acute lymphoblastic leukemia (ALL) [1,2]. The incidence of AML increases with age and is characterized by the rapid growth of abnormal white blood cells which accumulate in the bone marrow and interfere with the production of normal blood cells [3].
DNA methylation is an important regulator of gene transcription. It is an epigenetic modification that typically occurs at CpG (cytosine-phosphate-guanine) sites in mamalian cells. It is catalyzed by DNA methyltransferase which results in the conversion of the cytosine residue to 5-methylcytosine, leading to the formation of Me-CpG [4]. Alterations in DNA methylation frequently occur during cell development and are common in tumors. Aberrant DNA methylation patterns have been reported in a variety of cancers including AML [5,6]. Hypermethylation within the promoter regions of tumor suppressor genes leads to gene silencing, and global hypomethylation has also been recognized as a cause of oncogenesis [7,8]. As such, the role of methylation in carcinogenesis has been the focus of considerable research. However, the mechanisms leading to aberrant DNA methylation are poorly understood.
Genome-wide studies have been performed to obtain details on the genome and epigenome in various solid tumors. Examination of clonal evolution patterns have suggested that these may shape epigenetic dysregulation in AML. Investigations on genome-wide methylation in AML have shown variations in methylation patterns between AML and normal healthy tissues [9][10][11]. Although several studies have found that molecular subtypes in AML can exhibit highly distinct DNA methylation profiles [12,13], a small set of common genes have also been shown to exhibit consistent aberrant methylation across several hundred cases of AML [14,15], such as PML-RARα and AML1-ETO [12,13,16]. Together, these findings implicate distinct aberrant DNA methylation patterns in the pathogenetic processes in cancer, rather than evolving homogeneously across cancer types, Deep sequencing of patients with different myeloid malignancies has revealed recurrent mutations in epigenetic regulator proteins, such as DNMT3A and TET2, and several major driver mutations in normal karyotype AML [17,18]. Functional studies of these mutations have revealed specific methylation signatures. These mutations carry prognostic relevance and are likely to be contributors to AML pathogenesis.
Metallothionein (MT) is a family of cysteine-rich, low molecular weight (500-14,000 Da) proteins [19]. Ten functional isoforms of MTs have been identified, which are divided into four classes (MT1-4) based on small differences in protein sequences and characteristics. MTs have been proposed to play important roles in protecting against DNA damage, apoptosis and oxidative stress. MT3 was first identified as a neuronal growth inhibitory factor [20]. In vitro studies have demonstrated that MT3 can inhibit neurite formation and survival in neurons [20]. It has been shown to protect against apoptotic neuronal death in the brains of patients with Alzheimer's disease and in the hippocampus of SAMP8 mice, suggesting that MT3 may inhibit the development of neurodegeneration and may influence neuronal regeneration during the recovery process [21,22]. Its endogenous overexpression in glial and tumor cells stably transfected with MT3 was found to inhibit cell growth. In addition, MT3 plays a role in regulating lysosomal functions; in the absence of MT3, reductions in specific lysosomal enzymes associated with decreased autophagic flux have been reported [23].
In relation to cancer, downregulation of MT3 has been reported as one of 17 changes in gene expression which are associated with metastasis and poor clinical outcome in a range of solid tumors, including gastric cancer [24] and primary esophageal squamous cell carcinoma (SCC) [25]. This was supported by the observation that treatment with 5-aza-2'-deoxycytidine, an inhibitor of DNA methylation reduced the degree of methylation and increase the level of MT3 expression, in esophageal SCC cell lines [25]. Significant downregulation of MT3 has been most frequently reported in tumors that exhibiting MT3 methylation, suggesting that MT3 may act as a tumor suppresser via promoter hypermethylation [26].
However, reports on the methylation status of MT3 in the blood system are rare, and its expression and role in pediatric AML remains unclear. The aim of this study was to analyze the methylation profile and molecular function of MT3 in pediatric AML.

Patients and samples
Bone marrow specimens were obtained at the time of diagnosis during routine clinical assessment of 41 pediatric patients with AML, who presented at the Department of Hematology and Oncology, Children's Hospital of Soochow University between 2000 and 2010. Ethical approval was provided by the Children's Hospital of Soochow University Ethics Committee (No. SUEC2000-021), and informed consent was obtained from the parents or guardians. AML diagnosis was made in accordance with the revised French-American-British (FAB) classification. The main clinical and laboratory features of the patient cohort are summarized in Table 1. Additionally, bone marrow samples from 12 healthy donors and 8 patients with Idiopathic thrombocytopenic purpura (ITP) were analyzed as controls. Bone marrow mononuclear cells (BMNCs) were isolated using Ficoll solution within 2 h after bone marrow samples harvested and immediately subjected for the extraction of total RNA and genomic DNA.

CD34 + cell purification
For CD34 + cell selection, the Miltenyi immunoaffinity device (VarioMACS 130-046-703) was used according to the manufacturer's instructions (Miltenyi Biotech, Auburn, CA). Briefly, the CD34 + cells are magnetically labeled with CD34 MicroBeads. Then, the cell suspension is loaded onto a MACSR Column which is placed in the magnetic field of a MACS Separator. The magnetically labeled CD34 + cells are retained within the column. The unlabeled cells run through; CD34 + cells were adsorbed on the magnetic poles. After removing the column from the magnetic field, the magnetically retained CD34 + cells can be eluted as the positively selected cell fraction.

Sodium bisulfite modification of genomic DNA
High-molecular-weight genomic DNA was extracted from cell lines and biopsies by a conventional phenol/chloroform method. The sodium bisulphite modification procedure was as described [27] with slight modification. In brief, 600 ng of genomic DNA was denatured in 3 M NaOH for 15 min at 37°C, then mixed with 2 volumes of 2% lowmelting-point agarose. Agarose/DNA mixtures were then pipetted into chilled mineral oil to form agarose beads. Aliquots of 200 μl of 5 M bisulphite solution (2.5 M sodium metabisulphite, 100 mM hydroquinone, both Sigma, USA) were added into each tube containing a single bead. The bisulphite reaction was then carried out by incubating the reaction mixture for 4 h at 50°C in the dark. Treatments were stopped by equilibration against 1 ml of TE buffer, followed by desulphonation in 500 μl of 0.2 M NaOH. Finally, the beads were washed with 1 ml of TE buffer and directly used for PCR.

Methylation-specific PCR
The methylation status of the MT-3 promoter region was determined by methylation-specific PCR. Primers distinguishing unmethylated (U) and methylated (M) alleles were designed to amplify the sequence:  methylation-specific PCR reactions, in vitro-methylated genomic DNA treated with sodium bisulphite served as a positive methylation control. PCR products were separated on 4% agarose gels, stained with ethidium bromide and visualized under UV illumination. For cases with borderline results, PCR analyses were repeated.

Leukemia cell cells treated with 5-aza-2'-deoxycytidine
De-methylation was induced with 5-aza-dC (5-Aza, Sigma-Aldrich, St Louis, MO, USA) treatment at a concentration that induced de-methylation of the DNA without killing the cells. Culture media for HL-60 and MV4-11 cells contained 5 μM 5-Aza. DNA and RNA were extracted after 72 hours of 5-Aza treatment for the following analysis.

MT-3 lentiviral expression constructs and lentivirus production
Briefly, an approximately 250 bp fragment containing the human MT-3 gene was directly synthesized, and cloned into the pMD18-T vector. Positive clones were confirmed by sequencing and subcloned into the pLVX-IRES-ZsGreen vector (Clontech). The vector plasmids, pLVX-IRES-ZsGreen1, pLP1, pLP2 and pLP/VSVG were amplified in E.Coli and purified using the Endofree Maxiprep Kit (Qiagen). 270 μg of transfer vector, 176 μg of pLP1, 95 μg of pLP/VSVG and 68 μg of pLP2 was mixed with 0.25 M CaCl2 (Sigma) and added to same volume of 2 × HEPES (Sigma) and mixed while bubbling for 20 min to allow a precipitate to form. This was then added to a 175 cm2 flask of approximately 60% confluent 293 T cells containing 20 mL DMEM supplemented with 10% fetal calf serum, 100 U/mL penicillin, 100 μg/mL streptomycin and 2 mM glutamine and incubated for 48 h at 37°C in 5% CO2. The supernatant was centrifuged at 1,700 g for 10 min to pellet cell debris, and ultracentrifuged at 121,603 g for 2 h. The pellet containing concentrated virus was resuspended in DMEM without supplements and stored at -80°C.

Cell proliferation analysis
Acute myeloid leukemia cells were seeded in 96-well plates at 2 × 10 4 cells per well. 20 ul CCK-8(Dojindo Molecular Technologies, Japan) was added to each well and incubated at 37°C for a further 4 hours. The optical density (OD) values were measured at 450 nm on a scanning multi-well spectrophotometer (BioRad Model 550, USA). Compared with the control group, Cell proliferation was calculated as proliferation values. All experiments were performed in triplicate and repeated twice. The results were analyzed using ANOVA and the Student-Newman-Keuls tests, p < 0.05 were considered significant.

Apoptosis assay
Apoptosis assay was according to the manual operation of BD Annexin V Staining Kit (Cat: 556420, BD Biosciences, Franklin Lakes and NJ USA). Briefly, wash cells twice with cold PBS and then resuspend cells in 1 × Binding Buffer at a concentration of~1 × 10 6 cells/ml. Transfer 100 μl of the solution (~1 × 10 5 cells) to a 5 ml culture tube. Add Annexin V and PI 5 μl/test. Gently mix the cells and incubate for 15 min at RT in the dark. Add 400 μl of 1 × Binding Buffer to each tube. Analyze by flow cytometry as soon as possible (within 1 hour).

Western blot analysis
For western blot analysis, cellular proteins were extracted in 40 mM Tris-HCl (pH 7.4) containing 150 mM NaCl and 1% (v/v) Triton X-100, supplemented with a cocktail of protease inhibitors. Equal amounts of protein were resolved on 12% SDS-PAGE gels, and then transferred to a PVDF membrane (Millipore, Bedford, MA

Results
The MT3 promoter is hypermethylated in AML cells Our long-term research is focused on epigenic modification in pediatric AML and we have found a series of abnormal methylated genes related with AML [29,30]. In this study we conducted CpG island array analysis to explore promoter methylation in pediatric AML. The hypermethylated and hypomethylated genes between AML and NBM were clustered in Figure 1 and Additional file 1. The results implied that the MT3 promoter was hypermethylated in AML. Subsequent analyses identified four CpG islands in the MT3 promoter region (Figure 2A). Therefore, we conducted an MSP assay in 11 leukemia cell lines using a primer for MSP analysis that encompassed the CpG islands in the MT3 promoter. Our results showed that the MT3 promoter was hypermethylated in 7/11 leukemia cell lines, with the highest methylation levels observed in HL-60, MV4-11 SHI-1, U937 and K562 cells; whereas it was unmethylated in 4/11 cell lines, 697, SHI-1, THP-1 and Jurkat ( Figure 2B). To confirm methylation of the MT3 promoter, we treated the leukemia cell lines with 5-Aza. This demethylation reagent is an epigenetic modifier that inhibits DNA methyltransferase activity resulting in hypomethylation and gene activation. PCR analysis showed that methylation status of MT3 was decreased in leukemia cells following 5-Aza treatment compared to control cells treated with DMSO. Our results showed that MT3 expression was significantly upregulated in leukemia cells following 5-Aza treatment compared to control cells treated with DMSO ( Figure 2C

The MT3 promoter was methylated in patients with pediatric AML
In order to examine the methylation status in the MT3 promoter in pediatric AML, we obtained samples from 41 patients with pediatric AML and 20 control patients with NBM/ITP. Aberrant MT3 methylation was observed in  Figure 3A). Six NBM samples and six AML samples were selected for further analysis by BGS ( Figure 3B). Consistent with the MSP results, these confirmed that the CpG islands in the MT3 promoter were methylated in the AML samples (68%-86.7%); whereas they were unmethylated in the NBM samples (49.3%-61.3%). Examination of the clinicopathologic characteristics in patients with pediatric AML revealed that there were no   significant differences in clinical features, such as sex, age, initial hemoglobin level, white blood cell counts, platelet counts and chromosomal abnormalities between those with methylated MT3 and those with unmethylated MT3 (Table 1).
Real-time qPCR was employed to examine the transcript levels of MT3 in the 41 pediatric AML samples and 20 NBM/ITP control samples ( Figure 3C; Table 1). MT3 expression was found to be significantly decreased in the AML samples compared to the control samples    Figure 3D). In summary the hypermethylation status of the MT3 promoter in samples from patients with pediatric AML was consistent with results in human myeloid leukemia cell lines.

Overexpression of MT3 inhibited proliferation and induced apoptosis in leukemia cells
Transfection of the PLVX-MT3 lentivirus into HL-60 and MV4-11 leukemia cells was found to significantly upregulate expression of MT3 ( Figure 4A) and significantly inhibit cell proliferation ( Figure 4B). A CCK-8 assay in HL-60 and MV4-11cells showed that the inhibition rate at 5 days post-transfection was 43.7 ± 30.1% and 51.1 ± 26.8% in MT3-overexpessing cells compared to the mock transfection group (P < 0.05).
To further investigate the apoptotic effect of MT3 in HL-60 and MV4-11 cells, we investigated the expression levels of cleaved PARP, a marker of apoptosis, by Western blotting. The results were consistent with the Annexin V data, confirming that MT3 induced apoptosis in leukemia cells ( Figure 4A).

Dysregulation of apoptosis-related genes in HL-60 cells overexpressing MT3
In order to identify the apoptosis-related genes implicated in MT3 overexpression in HL-60 cells, we analyzed expression and clustering of 370 key genes associated with apoptosis by real-time PCR array analysis, cells harboring empty vector or a vector overexpressing MT3 were subjected to real-time PCR array ( Figure 5A). The genes most significantly downregulated or upregulated are shown in Figures 5B and  5C, respectively. Examination of the array data revealed that 17 genes were significantly upregulated and 17 genes were significantly downregulated in the MT3-overexpressing group compared to the control group (Tables 2 and 3, respectively). These included TNF (tumor necrosis factor) ligands and their receptors, members of the bcl-2 family, BIRC (baculoviral IAP repeat) domain proteins, CARD (caspase recruitment domain) proteins, death domain proteins, TRAF (TNF receptor-associated factor) domain proteins, and caspases. FOXO1 was found to be the most upregulated gene in MT3-overexpressing group.
The up-regulation of FOXO1 and CDKN1A in MT3overexpressing group was vivificated with western-blot analysis ( Figure 6).

Discussion
In this study, we found that the MT3 promoter was hypermethylated in pediatric AML. Our results showed that the MT3 promoter was hypermethylated in seven out of eleven human myeloid leukemia cell lines. Furthermore, in agreement with reports in esophageal SCC, treatment with 5-Aza, a specific inhibitor of DNA methylation, led to a significant increase in MT3 expression in AML cells (39.8 fold in HL-60 and 26.8 fold in MV4-11; P < 0.05 and P < 0.01, respectively). These results demonstrated that the MT3 promoter was consistently significantly methylated in human myeloid leukemia cells, suggesting that the MT3 promoter may also be methylated in pediatric AML. Consistent with the cell line results, we found that the MT3 promoter was methylated in samples obtained from patients with pediatric AML. Aberrant methylation of MT3 was observed 39.0% (16/41) of pediatric AML samples compared to 6.7% (2/30) of NBM control samples. In addition, MT3 was significantly downregulated in the pediatric AML samples compared to the control samples. Further analysis revealed that patients with pediatric AML exhibiting methylated MT3 showed lower levels of MT3 expression compared to those with unmethylated MT3. These results confirmed that hypermethylation of the MT3 promoter occurs with high frequency in both AML cell lines and pediatric AML samples. However, comparisons between the two groups of patients showed no significant differences in MT3 methylation status and patient characteristics, including sex, age, initial hemoglobin level, white blood cell counts, platelet counts, and chromosomal abnormalities. Together, these results implied that various mechanisms may be involved in the downregulation of MT3 in pediatric AML, such as different post-transcriptional modifications, gene deletions, copy number reductions and histone code modifications.  Further research will be required to elucidate the details of the underlying mechanisms. A previous study in prostate cancer PC-3 cells revealed that overexpression of MT3 significantly increased cell proliferation, invasion and tumorigenic activities both in vitro and in vivo [31]. However, there were no reports on the role of MT3 in leukemia cells. In this study, we demonstrated that HL-60 and MV4-11 leukemia cells overexpressing MT3 inhibited cell proliferation in a dosedependent manner. Further analysis by Annexin V assay revealed that there were a greater proportion of apoptotic cells in MT3-overexpressing leukemia cells compared to mock transfected cells. The apoptotic effect of MT3 in leukemia cells was confirmed by Western blot analysis which showed that MT3 overexpression led to enhanced expression of cleaved PARP, a marker of apoptosis. These findings suggested that MT3 may possess promising antitumor activity in AML cells.
Real-time PCR array analysis is an effective technique for quantifying the expression of a focused panel of genes [32,33]. Therefore, in order to explore the underlying mechanisms of MT3 antitumor activity, we carried out a real-time PCR array assay on 370 apoptosis-related genes in order to identify those genes that were dysregulated in AML following MT3 overexpression. The findings showed that 17 of the genes were significantly upregulated and 17 genes were significantly downregulated in MT3-overexpressing cells relative to control cells. The gene that was most significantly upregulated was found to be FOXO1, which belongs to the family of forkhead box transcription factors. These are downstream targets in the serine/threonine protein kinase B (PKB)/Akt pathway which is involved in the regulation of cell proliferation and survival [34][35][36]. It has been shown that withdrawal of growth factors leads to inactivation of the PI3K-Akt pathway, FOXO1 dephosphorylation at its Akt sites, nuclear translocation and activation of FOXO target genes [35][36][37]. FOXO1 plays a central role in initiating apoptosis by inducing expression of death genes, such as FASL. In the nucleus, FOXO1 is a key mediator of tumor suppression downstream of PTEN. Exogenous expression of PTEN induces FOXO1 to relocate to the nucleus, restoring its transcriptional activation. In addition, a constitutively active form of FOXO1 that cannot be phosphorylated by Akt has been found to induce apoptosis in PTEN-null cells, indicating that it has the same effect as reconstitution of PTEN [38]. Our results indicated that FOXO1 is a downstream target of MT3 in pediatric AML. A study in prostate cancer cells suggested that overexpression of FOXO1 resulted in apoptosis via increased expression of TRAIL (TNF-Related Apoptosis-Inducing Ligand) [39]. However, the mechanism and the role of these genes in MT3-induced apoptosis in AML remain to be elucidated.

Conclusions
In this study, we identified epigenetic inactivation of MT3 in both AML cell lines and pediatric AML samples via hypermethylation of the MT3 promoter. Our findings also showed that transcriptional overexpression of MT3 could inhibit proliferation and induce apoptosis in AML cells. We identified 34 dysregulated apoptosisrelated genes in MT3-overexpressing, including FOXO1. These results may provide new insights into the molecular mechanism of MT3-induced apoptosis; however, further research will be required to determine the underlying details. Together, our findings suggest that MT3 may act as a putative tumor suppressor gene in pediatric AML.