ABSTRACT
The miR-483-3p is upregulated in several tumors, including liver tumors, where it inhibits TP53-dependent apoptosis by targeting the pro-apoptotic gene BBC3/PUMA. The transcriptional regulation of the miR-483-3p could be driven by the ß-catenin/USF1 complex, independently from its host gene IGF2, and we previously demonstrated that in HepG2 hepatoblastoma cells carrying wild-type TP53 the upregulation of the miR-483-3p overcomes the antitumoral effects of the tumor-suppressor miR-145-5p by a mechanism involving cellular glucose availability. Here we demonstrate that in HepG2 cells, the molecular link between glucose concentration and miR-483-3p expression entails the O-linked N-acetylglucosamine (O-GlcNAc) transferase (OGT), which stabilizes the transcriptional complex at the miR-483 promoter. HepG2 cells showed reduced miR-483-3p expression and increased susceptibility to 5-fluorouracil (5-FU)-induced apoptosis in presence of the inhibitor of glycolysis 2-deoxy-d-glucose (2-DG). However, in vivo experiments showed that HepG2 cells with higher miR-483-3p expression were selected during tumor progression regardless of 5-FU treatment. Furthermore, treatment with 2-DG alone did not significantly reduce HepG2 xenograft load in immunodeficient mice. In conclusion, we show that in HepG2 cells glucose uptake increases the expression of the oncogenic miR-483-3p through the OGT pathway. This suggests that depletion of the miR-483-3p may be a valuable therapeutic approach in liver cancer patients, but the use of inhibitors of glycolysis to achieve this purpose could accelerate the selection of resistant neoplastic cell clones.
ABSTRACT
Deregulation of the miR-15a/16-1 cluster has a key role in the pathogenesis of chronic lymphocytic leukemia (CLL), a clinically heterogeneous disease with indolent and aggressive forms. The miR-15a/16-1 locus is located at 13q14, the most frequently deleted region in CLL. Starting from functional investigations of a rare SNP upstream the miR cluster, we identified a novel allele-specific mechanism that exploits a cryptic activator region to recruit the RNA polymerase III for miR-15a/16-1 transcription. This regulation of the miR-15a/16- locus is independent of the DLEU2 host gene, which is often transcribed monoallellically by RPII. We found that normally one allele of miR-15a/16-1 is transcribed by RNAPII, the other one by RNAPIII. In our subset of CLL patients harboring 13q14 deletions, exclusive RNA polymerase III (RPIII)-driven transcription of the miR-15a/16-1 was the consequence of loss of the RPII-regulated allele and correlated with high expression of the poor prognostic marker ZAP70 (P=0.019). Thus, our findings point to a novel biological process, characterized by double allele-specific transcriptional regulation of the miR-15a/16-1 locus by alternative mechanisms. Differential usage of these mechanisms may distinguish at onset aggressive from indolent forms of CLL. This provides a basis for the clinical heterogeneity of the CLL patients carrying 13q14 deletions.