Introduction: Myelodysplastic syndrome (MDS) is a group of clonal myeloid disorders characterized by ineffective hematopoiesis and cytopenias. Treatment strategies are based on the balance between patients’ predicted lifespan and the anticipated risk-benefit of intervention. Current guidelines recommend to classify patients into low-risk and high-risk groups. Hypomethylating agents azacitidine and decitabine are recommended for patients with high-risk MDS, who are not eligible for intensive therapy including stem cell transplantation. Azacitidine showed a survival benefit, however, treatment with this agent has not been standardized. To further optimize the use of hypomethylating agents, the target molecules of the agents should be identified to select patients, who would best benefit from the hypomethylating treatment. In the present study, the cellular drug sensitivity, the cell differentiation, the induction of apoptosis, and DNA methylation status after the treatment with decitabine were extensively investigated using cultured MDS cell line MDS-L that had been previously established.

Methods: MDS-L cells, which had been established in the previous study (Br J Haematol, 87,235-242,1994) were used. This cell line is a blastic subline derived from the bone marrow blasts of a 50-year Japanese male patients with MDS. Trypan dye exclusion assay was used for determining growth inhibition effects. The cellular morphology was observed by staining the cells with May- Grunwald-Giemsa dye. Cell surface markers were detected by using flow cytometry. The transcript of Wilms’ tumor 1 was measured by real time RT-PCR. The induction of apoptosis was determined either by the sub-G1 population in the cell cycle analysis or the caspase cleavage. The enzymatic activity of DNA methyltransferase 1 (DNMT1) was determined by using a DNMT1 activity assay kit (Active Motif, Carlsbad, CA). DNA methylation array was performed using Infinium Human Methylation 450 BeadChip (Illumina, San Diego, CA).

Results: The 50%-growth inhibitory concentration of decitabine was 64 nM in MDS-L cells. The cells were incubated with a non-toxic concentration (5 nM) of decitabine for 6 months (named as MDS-L-DAC cell line). MDS-L-DAC cells proliferated more slowly than untreated MDS-L cells (doubling time:51 h). Morphologically, MDS-L-DAC cells were maturated with a lobulated nuclei and cytoplasmic granulation. Differentiation induction was also indicated by an increased CD15 positivity (9.3→30.5%) and a decreased CD34 positivity (98.8→6.4%) in MDS-L-DAC cells. Wilms’ tumor 1 transcript level was also reduced (1,400,000→420,000 copies/μgRNA) in MDS-L-DAC cells, suggesting a decreased malignant potential by the effect of decitabine. Decitabine induced greater amount apoptosis (sub-G1 :12.3%) was with caspase 3 and caspase 8 cleavage in MDS-L-DAC cells, but not in MDS-L cells (sub-G1: 5.6%). The enzymatic activity of DNMT1 was successfully inhibited by decitabine in MDS-L-DAC cells (11.2→4.5OD/h/mg). The epigenome-wide association study using DNA methylation array compared the methylation status at CpG islands between MDS-L cells and MDS-L-DAC cells. Demethylation occurred in MS4A3 (Membrane-spanning 4-domains subfamily A member 3), PAR-4 (Prostate apoptosis response 4), CD40LG (CD40 ligand), APITD1 (Apoptosis-inducing, TAF 9-like domain 1), LOX (Lysyl Oxidase), and AMPH (Amphipysin) in MDS-L-DAC cells. Western blotting confirmed that the corresponding protein levels were increased.

Discussion and Conclusions: MS4A3 is associated with myeloid differentiation. Caspase-8-mediated PAR-4 cleavage is required for tumor necrosis factor α-induced apoptosis. Low expression of APITD1 interferes with the ability for apoptosis through the p53 pathway. LOX is one of the commonly methylated genes in gastric cancer. Thus, the long-term incubation with decitabine at a low concentration demethylated these gene promoters and regained their protein expression. Decitabine induced the cell differentiation, decreased malignant potential, and increased susceptibility to apoptosis via a death receptor pathway, provably due to the functions of these proteins.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution