Analysis of the Paternally-Imprinted DLK1–MEG3 and IGF2–H19 Tandem Gene Loci in NT2 Embryonal Carcinoma Cells Identifies DLK1 as a Potential Therapeutic Target

The paternally-imprinted genes insulin-like growth factor 2 (IGF2), H19, delta-like homologue 1 (DLK1), and maternally-expressed gene 3 (MEG3) are expressed from the tandem gene loci IGF2–H19 and DLK1–MEG3, which play crucial roles in initiating embryogenesis and development. The erasure of imprinting (EOI) at differentially methylated regions (DMRs) which regulate the expression of these genes maintains the developmental quiescence of primordial germ cells (PGCs) migrating through the embryo proper during embryogenesis and prevents them from forming teratomas. To address the potential involvement of the IGF2–H19 and DLK1–MEG3 loci in the pathogenesis of embryonal carcinoma (EC), we investigated their genomic imprinting at DMRs in the human PGC-derived EC cell line NTera-2 (NT2). We observed EOI at the IGF2–H19 locus and, somewhat to our surprise, a loss of imprinting (LOI) at the DLK1–MEG3 locus. As a result, NT2 cells express imprinted gene ratios from these loci such that there are i) low levels of the proliferation-promoting IGF2 relative to ii) high levels of the proliferation-inhibiting long noncoding RNA (lncRNA) H19 and iii) high levels of proliferation-promoting DLK1 relative to iv) low levels of the proliferation-inhibiting lncRNA MEG3. Consistent with this pattern of expression, the knockdown of DLK1 mRNA by shRNA resulted in decreased in vitro cell proliferation and in vivo tumor growth as well as decreased in vivo organ seeding by NT2 cells. Furthermore, treatment of NT2 cells with the DNA methyltransferase inhibitor 5-aza-2′-deoxycytidine (5-azaD) inhibited their proliferation. This inhibition was accompanied by changes in expression of both tandem gene sets: a decrease in the expression of DLK1 and upregulation of the proliferation-inhibiting lncRNA MEG3, and at the same time upregulation of IGF2 and downregulation of the lncRNA H19. These results suggest that the DLK1–MEG3 locus, and not the IGF2–H19 locus, drives the tumorigenicity of NT2 cells. Based on these results, we identified DLK1 as a novel treatment target for EC that could be downregulated by 5-azaD.