Role of microRNAs in epigenetic silencing of the CHD5 tumor suppressor gene in neuroblastomas

Abstract

Neuroblastoma (NB), a tumor of the sympathetic nervous system, is the most common extracranial solid tumor of childhood. We and others have identified distinct patterns of genomic change that underlie diverse clinical behaviors, from spontaneous regression to relentless progression. We first identified CHD5 as a tumor suppressor gene that is frequently deleted in NBs. Mutation of the remaining CHD5 allele is rare in these tumors, yet expression is very low or absent, so expression is likely regulated by epigenetic mechanisms. In order to understand the potential role of miRNA regulation of CHD5 protein expression in NBs, we examined all miRNAs that are predicted to target the 3'-UTR using miRanda, TargetScan and other algorithms. We identified 18 miRNAs that were predicted by 2 or more programs: miR-204, -211, -216b, -17, -19ab, -20ab, -93, -106ab, -130ab, -301ab, -454, -519d, -3666. We then performed transient transfections in two NB cell lines, NLF (MYCN amplified) and SY5Y (MYCN non-amplified), with the reporter plasmid and miRNA mimic, as well as appropriate controls. We found seven miRNAs that significantly downregulated CHD5 expression in NB: miR-211, 17, -93, -20b, -106b, -204, and -3666. Interestingly, MYCN upregulates several of the candidates we identified: miR-17, -93, -106b & -20b. This suggests that miRNAs driven by MYCN and other genes represent a potential epigenetic mechanism to regulate CHD5 expression.

Keywords: CHD5; MYCN; miRNA; neuroblastoma; tumor suppressor.

Figure 1. Nucleotide sequence of the 3′ untranslated region (UTR) of CHD5

Three locations that microRNAs are predicted to bind in the CHD5 3′-UTR are shown in blue, and the miRNAs that target each of these regions are shown below. The regions cloned into the targeting vector are highlighted in yellow.

Figure 2. Graphic representation of miRNA regulation of CHD5 3′-RNA reporter construct

Graphical representation of “CHD5 levels” in the NLF cell line after transfection with the MYCN-driven subset of miRNAs. Each sample was transfected with miRNAs from the MYCN driven subset, and either no 3′-UTR inserted (no Insert), wild-type (WT) 3′-UTR insert, or mutated 3′-UTR insert (MUT). Therefore, each value represented in the bar graph reflects the ratio of Renilla to firefly normalized to Allstars siRNA. A. NLF and MYCN-driven miRNAs. B. NLF and non-MYCN-driven miRNAs. C. SY5Y and MYCN-driven miRNAs. D. SY5Y and non-MYCN-driven miRNAs. Each experiment included miR-211 mimic as a positive control, Allstars siRNA (Qiagen) as a negative control, and no miRNA mimic as a transfection control. Each transfection was carried out in triplicate, and each experiment was repeated at least 3 times. Statistical analyses were performed using the Prism two-way ANOVA method followed by a Sidak post-test. Data are expressed as the standard error mean (SEM). Values are the mean of triplicates readings from four independent experiments and p-values were reported (* P<0.05, ** p<0.01 and ns= non-significant).

Figure 3. Summary of miRNA regulation of a CHD5 3′-UTR reporter construct in NLF and SY5Y lines with miRNA mimics predicted to target CHD5

miRNAs in green circles caused downregulation of the CHD5 3′-UTR target construct in at least one of the two lines. miRNAs in red circles did not significantly downregulate the CHD5 3′-UTR. Purple arrows indicate miRNAs shown to be upregulated by MYC/MYCN expression. Asterisk (•): miRNA 106b downregulated CHD5 3′-UTR in NLF only, not in SY5Y.

Figure 4. CHD5 protein expression in NBLS cells following transient transfection with miRNA mimics

A. Western blot analysis of transfected cells with indicated microRNAs. Post transfection, cells were washed twice with PBS and isolated cell extracts as described in methods [41]. Whole cell extracts (100 μg) either transfected with indicated miRNAs or mock transfected were subjected to polyacrylamide gel electrophoresis (4-12% SDS-PAGE), using NuPAGE Bis-Tris gels with MOPS-SDS Running Buffer Allstars siRNA and miRNA-454 were used as negative controls. Proteins were transferred on to nitrocellulose membranes (GE Healthcare Life Sciences) and probed with antibodies using rabbit polyclonal CHD5, actin (Santa Cruz Biotechnology, CA 1:1000), rabbit polyclonal CHD4 (Bethyl 1:2000), and MYCN monoclonal (1:5000; BD Biosciences). Almost complete reduction of CHD5 protein levels were observed for miR-20b, miR-93, miR-17, and miR-211 as indicated, but no change in CHD4, actin or MYCN levels were seen. B. Densitometric analysis of CHD5 protein expression in NBLS cell line. The number of pixels from each band was measured, and a bar graph was created using the Prism to indicate the difference in CHD5 expression upon miRNA transfection. Data are expressed as the standard error mean (SEM). Statistical analysis was performed using the Prism one way ANOVA method followed by Tukey's post-test. Statistical significance relative to the control Allstar siRNA is indicated: *p<0.05; **p<0.01.