DLK1 is a transmembrane protein of the epidermal growth factor (EGF) family, encoded by a paternally-imprinted gene located within the chromosomal region 14q32. In addition to DLK1, this region contains other paternally expressed genes, including Dio3 and RTl1 and a set of genes expressed from the maternal chromosome, including MEG3, C/D snoRNA and Mirg. In mice, the expression of these paternally- and maternally-imprinted genes is inversely correlated and controlled by the methylation status intergenic differentially methylated region (IG-DMR). Recently it has been reported that DLK1 is expressed at higher than normal levels in myelodysplasia (MDS) and also in acute myeloblastic leukemia (AML). To determine whether loss of imprinting (LOI) could account for DLK1 overexpression and to study the mechanisms that regulate DLK1 imprinting in AML, we analyzed the expression of 3 informative coding SNPs located in exon 5 (rs#1802710, rs#2295660, rs#1058009) in 11 normal bone marrows (NBMs), the 3 cell lines (OCI/AML-5, NB4 and K562) characterized by DLK1 upregulation and 40 AML patients (pts) with DLK1 overexpression. Furthermore, we undertook quantitative methylation analysis, using the MassARRAY system, of 7 CpG rich-reas: 3 located upstream or within MEG3 and correspond to the CpG islands # 30, 45 and 70, 1 corresponds to the putative IG-DMR, and 3 located upstream or within DLK1 and correspond to the CpG islands # 26, 65 and 79. Informative SNPs were found in 6 NBMs, 28 AML pts and the cell line K562. Of these, biallelic DLK1 expression was found in the cell line K562 as well as 22/28 (72%) AML pts. In contrast, all NBMs and 7 AML pts were found to have monoallelic expression. Pts with biallelic DLK1 expression showed higher DLK1/GAPDH than pts with monoallelic expression (median: 0.0057 vs 0.0024). On the other hand, no significant difference in MEG3 levels was found between the two groups (P = 0.49) and no correlation was found between DLK1 and MEG3 levels (r= −0.1212). The quantitative methylation analysis revealed no difference between the pts with monoallelic and biallelic DLK1 expression in the methylation patterns of the CpG islands # 30, 45, 70, 65 and 26 or the IG-DMR. In contrast, significant difference was found in the methylation the CpG island #79, located 18 kb upstream DLK1 (P <0.0001). In fact, there was a strong association between bi-allelic DLK1 expression and the hypermethylation of the latter CpG island, suggesting that this region contains an insulator element that regulates DLK1 imprinting and transcription through a methylation-sensitive mechanisms. Almost all insulator elements use the zinc finger protein CTCF to achieve their silencing activity. Indeed, a bioinformatic search indicated the presence of at least 4 predicted CTCF-binding sites. The CpG dinucleotides located within or in the vicinity of these binding sites were largely hypermethylated in AML with biallelic DLK1 expression, in sharp contrast with NBMs and pts with monoallelic DLK1 expression. Chromatin immunoprecipitation analysis confirmed that the CTCF protein binds to this region in 1 NBM and 2 pts with monoallelic expression, whereas no immunoprecipitation was seen in the K562 and 2 pts with biallelic DLK1 expression. Taken together, our data suggest DLK1 LOI occurs in 72% of AML and the expression of the paternally-imprinted DLK1 and the maternally-imprinted MEG3 genes are not coordinated in AML. Furthermore, an insulator element located 18 kb upstream DLK1 plays important role in controlling this gene imprinting in AML, through interaction with CTCF.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution