Abstract 4615

Introduction.

Alterations in chromatin organization are a common mechanism of leukemogenesis. Drugs affecting epigenetics have been proposed as therapy of hematopoietic neoplasms. We have previously shown selective activity of histone deacetylase (HDAC) inhibitors on AML1/ETO positive cells (Barbetti, et al- Oncogene, 2007). The AML1/ETO multiprotein complex induces and stabilizes the local DNA methylation pattern by recruiting HDACs and DNA methyl-transferase-1 (DNMT1), resulting in the stable silencing of AML1-controlled genes, like IL3. Our hypothesis is that DNMT inhibitors (DNMTi) 5-Azacytidine (AZA) and Decitabine (DAC) exert a targeted action on AML1/ETO positive leukemic cells.

Methods.

We used U937-A/E-9/14/18 cell line, carrying a ponasterone-A inducible construct of AML1/ETO (kindly provided by M. Luebbert). We analyzed the effects of AZA and DAC at different doses (0.01, 0.1, 1 and 10μM), evaluating cell viability and apoptosis (by Annexin V assay, and by WB caspase 9, 8, 3 -cleavage).

Cells were then treated with AZA 0.1–1μM or DAC 0.1–0.01μM for 24h. We determined by chromatin immunoprecitation (ChIP) assay the association of the IL3 promoter with H4acetylated, H3K4me3, H3K9me2 and H3K27me3 and the quantitative expression of IL3 by Q-RT PCR, calculated by a comparative threshold cycle method, using the formula 2̂(−ΔΔCt). Other promoters controlled by AML1/ETO, like those of GM-CSF, MPO and CEBPalpha were also evaluated. Methylation of IL3 promoter was evaluated by Methylation Specific PCR (MSP).

Confirmatory experiments were carried out in the human leukemic cell line Kasumi-1 naturally expressing AML1/ETO.

Results.

Although both DNMTi were active on leukemic cells, irrespective of the presence of AML1/ETO expression, the dose at which this effect was evident was differing of one logarithm. AZA 0.1μM significantly increased the percentage of apoptotic cells only in U937 AML1/ETO positive cell line. We observed that untreated apoptotic cells percentage was 4.72% (±0.55 Standard Deviation: SD) while after AZA 0.1μM was 5.70% (± 0.24 SD) (P<0.01). In AML1/ETO negative cells AZA 0.1μM did not induce a significant increase of apoptosis (P>0.05). These data were confirmed by the induction of cleavage of caspase 9, 8 and 3. DAC 0.01μM significantly induced apoptosis cells only in U937 AML1/ETO positive cells (NT= 6.27% ±0.24 SD; DAC 0.01μM = 7.50% ±0.57 SD; P<0.01).

Given these results, we investigated whether this different response was correlated with chromatin rearrangements directly involving the AML1/ETO gene.

We evaluated IL3 expression after AZA and DAC treatments, which induced a significant increase, higher in AML1/ETO-expressing than in non-expressing U937 cells. In fact, in AML1/ETO positive cells the amount of IL3 mRNA 2̂(−ΔΔCt) after AZA and DAC respect to NT was 10.5 ± 0.3 DS and 6.1± 0.5 SD, respectively. In AML1/ETO negative cells, after AZA was 5.70 ±2.7 SD (P>0.05) and after DAC 2.28 ± 0.4 SD (P>0.05), both respect to NT.

By ChIP assay we showed that in untreated leukemic cells, both in U937 AML1/ETO-positive and negative cells, at baseline H3K4me3 was not associated with the IL3 promoter, while H3K9me2 was associated with it. H4-acetylated became associated to the IL3 promoter after AZA and DAC treatments but exclusively in U937 cells expressing AML1/ETO. On the other hand, H3K27me3 was co-immunoprecipitated with the IL3 promoter in U937 AML1/ETO negative and in U937 AML1/ETO positive untreated cells. AZA and DAC treatments led to the dissociation of this modified histone from the IL-3 promoter indicating the modification of chromatin rearrangement (Fig 1).

Conclusions.

Our observations demonstrate that AML1/ETO positive cells are more sensitive to DNMTi than cells not carrying such chimeric gene, in terms of biological effects like cytotoxicity and induction of apoptosis. The dissociation of the IL3 promoter from H3K27me3 and its association with H4acetylated upon DNMTi exposure is observed exclusively in U937 AML1/ETO-positive cells and clearly indicates that the epigenetic mechanism of action of these drugs is exerted selectively on cells harbouring the chimeric protein. This result could support a working hypothesis to develop a treatment for AML1/ETO-positive leukaemias including DNMTi in the therapeutic armamentarium.

Disclosures:

Santini:Janssen: Honoraria; Celgene: Honoraria; Novartis: Honoraria.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution