Figure 1.
Priming effects of entinostat and decitabine for ATRA-induced differentiation in PML-RARα-negative AML cell lines.
Effects of entinostat (A) and decitabine (B) on histone acetylation in Kasumi-1 cells. Cells were treated with the indicated doses of entinostat or decitabine in the presence or absence of ATRA (1 µM) as described in Materials and Methods. Whole cell lysates were subjected to immunoblot analysis for pan-histone H3 acetylation and β-actin (loading control). Effects on differentiation in Kasumi-1 (C) and HL-60 (D) cells treated with either decitabine (200 nM) or entinostat (500 nM) alone or in combination with ATRA (100 nM in Kasumi-1, 1 µM in HL-60) added on day 0. Differentiation was quantified by flow cytometry of CD11b expression in Kasumi-1 on day 4 of treatment (no quantification was feasible on day 7 due to low cell numbers) and CD11b and CD11c expression in HL-60 cells on days 4 and 7 of treatment. Bars represent the mean of 2–5 measures and error bars the standard deviation. Statistical significance by ANOVA test: * p<0.05, # not significant. (E) Effects of epigenetic therapy on cell growth. Kasumi-1 and HL-60 cells were treated with ATRA (100 nM in Kasumi-1, 1 µM in HL-60), decitabine (200 nM) and entinostat (500 nM) as described above. Cell proliferation was determined by Trypan blue exclusion; the percentage of cell proliferation is shown. (F) Induction of p21CIP1/WAF1 protein during prolonged cell culture. Kasumi-1 (upper panel) and HL-60 cells (lower panel) were treated with either ATRA, decitabine or entinostat for the times and with the concentrations indicated in (E). Cells were harvested after 4 and 7 days and were subjected to western blot analysis for p21CIP1/WAF1 as described in Materials and Methods. As loading control, β-actin was used.
Table 1.
AML patient characteristics.
Table 2.
Characteristics of AML patients analyzed by pyrosequencing.
Table 3.
A list of the PCR and sequencing primers used for the bisulfite pyrosequencing of the RARβ2 regions 2 and 3.
Figure 2.
Effects of entinostat upon ATRA-induced differentiation and apoptosis in primary AML cells.
(A) Effect of entinostat on histone acetylation in primary AML cells. Bone marrow samples from 2 patients with de novo, normal karyotype AML were treated ex vivo with a single pulse of 1 µM entinostat for 48 hours. Whole cell lysates were subjected to immunoblot analysis for pan-histone H3 acetylation and β-actin (loading control). (B) Effects on differentiation in primary AML blasts. Primary AML samples were treated with entinostat or/and ATRA ex vivo as described in Materials and Methods. Expression of the myeloid differentiation marker CD11c was quantified by flow cytometry after 48 hours of treatment. (C) Representative flow cytometric analysis from primary AML sample #1. Primary AML blasts were treated with 500 nM entinostat or 1 µM ATRA alone or in combination. Forty eight hours after start of the treatment, expression of the myeloid differentiation marker CD11c and the hematopoietic progenitor cell surface marker CD117 (c-Kit) were quantified using flow cytometry. PE, phycoerythrin; APC, allophycocyanin. (D) Effects on apoptosis in primary AML blasts. Representative flow cytometric analysis of FITC Annexin V and 7-AAD staining from primary AML sample #1. Primary AML blasts were treated as above. Cells were stained with FITC Annexin V and 7 AAD viability dye and analyzed by flow cytometry. Annexin V positive/7-AAD negative cells represent early apoptotic cells.
Figure 3.
Epigenetic RARβ2 gene silencing in myeloid leukemia cells is reversed by epigenetically active drugs.
(A) The AML cell lines NB-4, HL-60, Kasumi-1, SKNO-1 and U-937 as well as the CML cell line K562 were treated with 1 µM ATRA for 72 hours, and expression of RARβ2 was determined by RT-PCR. The expected band of 621 bp corresponding to the RARβ2 isoform is shown. Normal CD34+ cells were used as a positive control for RARβ2 mRNA expression. GAPDH served as a control for equal amounts of RNA. (B) Effect of entinostat, alone or in combination with decitabine or ATRA, on RARβ2 expression in Kasumi-1 and HL-60 cells. Upper panel, Kasumi-1 cells were treated as described in Materials and Methods with entinostat and ATRA (1 µM) at the concentrations indicated. RARβ2 and GAPDH (internal control) mRNA levels were assayed using RT-PCR. Lower panel, HL-60 cells were treated with ATRA (1 µM), decitabine (200 nM) and entinostat (500 nM) as described above. Cells were then subjected to RT-PCR as described. (C) Effect of entinostat on chromatin structure in the RARβ2 gene region 2 in Kasumi-1 cells. Cells were treated with 1 µM entinostat for 24 hours, chromatin immunoprecipitation was performed for the indicated chromatin marks, followed by quantitative real-time PCR for precipitated DNA using primers specific for RARβ2 region 2 (RARE element, see Materials and Methods). (D) Cooperative effects of decitabine and entinostat upon RARβ2 re-induction. Kasumi-1 cells were treated with ATRA (1 µM), decitabine (50 nM) and entinostat (50 nM) as shown in Fig. 1A and B. Expression of RARβ2 and GAPDH mRNAs was determined by RT-PCR. (E) Effect of entinostat or ATRA, alone or in combination, on RARβ2 expression in primary AML blasts. Primary AML samples were treated with entinostat and ATRA ex vivo as described in Materials and Methods. RARβ2 and GAPDH (internal control) mRNA levels were assayed using RT-PCR.
Figure 4.
Differential methylation of the RARβ2 promoter/exon1 in human myeloid leukemia cells.
(A) Map of the RARβ2 promoter region with indication of amplicons 1–3 analyzed by bisulfite sequencing (positions of forward (bis F) and reverse (bis R) primers are shown), pyrosequencing and quantitative real-time PCR of ChIP precipitates. RARE: Retinoic Acid Response Element. Vertical ticks indicate single CpGs (region 1∶18 CpGs; region 2∶3 CpGs; region 3∶11 CpGs). Arrow: transcriptional start site. (B) Differential DNA methylation of RARβ2 promoter/exon1 in myeloid cells. Normal CD34+ cells, peripheral blood mononuclear cells and myeloid cell lines NB-4, HL-60, Kasumi-1, SKNO-1, U-937 and K562 were subjected to promoter methylation analyses by bisulfite sequencing of genomic DNA as described. Open circles represent unmethylated, closed circles methylated CpG sites. Percentages indicate mean methylation per region across all sequenced alleles. Region 1 was most heavily methylated in HL-60 cells, region 3 in SKNO-1, HL-60 and Kasumi-1, whereas region 2 disclosed lower overall methylation levels (25 and 21% in Kasumi-1 and HL-60, respectively, between 0 and 12% in the other cell lines). (C) RARβ2 promoter/exon1 DNA methylation in human myeloid leukemia cell lines is partially reversed by DNA hypomethylating treatment. Kasumi-1 and HL-60 cells were treated with decitabine by three 24-hours incubations of decitabine (DAC) at 200 nM. Cells were harvested 48 hours after the last pulse of decitabine and subjected to bisulfite sequencing. The sequenced alleles of untreated HL-60 and Kasumi-1 cells correspond to the experiment depicted in Fig. 4B. (D) RARβ2 5′UTR/exon1 region is rarely methylated in the primary bone marrow blasts of AML patients. Quantitative DNA methylation of the RARβ2 region 2 and 3 (as depicted in Fig. 4A) was analyzed by pyrosequencing in 41 AML patients and is shown as a heatmap. Columns represent single CpGs, each row represents a different sample. White indicates hypomethylation, dark blue indicates hypermethylation. DNA methylation standards (0%, 25%, 50%, 75% and 100% of the in vitro methylated genomic DNA) are depicted at the bottom of the heat map. BM, normal bone marrow.