ELF4 Is a Target of miR-124 and Promotes Neuroblastoma Proliferation and Undifferentiated State

Abstract

13-Cis-retinoic acid (RA) is typically used in postremission maintenance therapy in patients with neuroblastoma. However, side effects and recurrence are often observed. We investigated the use of miRNAs as a strategy to replace RA as promoters of differentiation. miR-124 was identified as the top candidate in a functional screen. Genomic target analysis indicated that repression of a network of transcription factors (TF) could be mediating most of miR-124's effect in driving differentiation. To advance miR-124 mimic use in therapy and better define its mechanism of action, a high-throughput siRNA morphologic screen focusing on its TF targets was conducted and ELF4 was identified as a leading candidate for miR-124 repression. By altering its expression levels, we showed that ELF4 maintains neuroblastoma in an undifferentiated state and promotes proliferation. Moreover, ELF4 transgenic expression was able to counteract the neurogenic effect of miR-124 in neuroblastoma cells. With RNA sequencing, we established the main role of ELF4 to be regulation of cell-cycle progression, specifically through the DREAM complex. Interestingly, several cell-cycle genes activated by ELF4 are repressed by miR-124, suggesting that they might form a TF-miRNA regulatory loop. Finally, we showed that high ELF4 expression is often observed in neuroblastomas and is associated with poor survival. IMPLICATIONS: miR-124 induces neuroblastoma differentiation partially through the downregulation of TF ELF4, which drives neuroblastoma proliferation and its undifferentiated phenotype.

Figure 1.. miR-124 regulated transcription factors contribute to neuroblastoma proliferation and maintenance of the undifferentiated state.

(A) Transcription factor network regulated by miR-124 (built with STRING). (B-C) BE(2)-C and CHP-212 cells were reverse transfected with siRNAs against each transcription factor. 120 h later: (B) Neurite length was measured utilizing the IncuCyte Neurotrack software module; (C) Quantification of viable cells was assessed by MTS assay. Statistical significance of observed differences was determined by Student’s t-test with a Holm-Šídák correction for multiple comparisons.

Figure 2.. ELF4 is critical for neuroblastoma cell proliferation and reduction of intracellular levels of ELF4 induces differentiation.

(A) Neuroblastoma cell lines were reverse transfected with siRNAs (siELF4 and control) and their proliferation was monitored by live-cell imaging (IncuCyte). (B) Knockdown of ELF4 with siRNA results in a significant decrease in number of viable cells (MTS assay, 120 h post-transfection). (C) Silencing of ELF4 results in differentiation (neurite outgrowth, 120 h post-transfection). (D) ELF4 knockdown significantly diminishes the percentage of Ki67 positive cells (Ki67 quantification, 120 h post-transfection). (E) Loss of ELF4 induces apoptosis (Caspase-3/−7 assay, 48 h post-transfection). Statistical significance of observed differences was determined by Student’s t-test. Proliferation data was adjusted for multiple testing using a Bonferroni correction. * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001.

Figure 3.. Morphological changes of neuroblastoma cells following ELF4 knockdown.

Images of neuroblastoma cells 120 h after reverse transfection with siRNAs.

Figure 4.. ELF4 ectopic expression antagonizes the effect of miR-124 on differentiation of neuroblastoma cells.

(A-B) BE(2)-C cells were first infected with ELF4-expressing lentivirus or control then reverse transfected with miRNA mimics (miR-124 or control). 120 hours later, the impact on differentiation and confluence was measured. (A) Morphology of transfected cells (scale bar = 100 μm). (B) Quantification of neurite outgrowth of treated cells. (C) Confluence of treated cells. (D) Proposed model of ELF4 and miR-124 antagonism. (E) ELF4 and miR-124 expression levels in SK-N-BE(2)-C transfected cells. (F) Expression analysis by qRT-PCR of miR-124 and ELF4-overexpressing shared targets in co-transfected cells were transfected with miRNA mimics; 48 hours later RNA was isolated and qRT-PCR was used to measure expression of a select group of co-targeted genes. Statistical significance of observed changes in neurite outgrowth was determined by a two-way ANOVA with Tukey’s range test for multiple comparisons. A Student’s t-test was used to assess differences in expression. * = p<0.05, **** = p<0.0001.

Figure 5.. Synergy between ALK and ELF4 inhibition on NB cell proliferation.

(A) SH-SY5Y cells were reverse transfected with siRNAs (siALK, siELF4 or control) and their proliferation was monitored with live-cell imaging (IncuCyte). (B) mRNA levels of ALK and ELF4 48 h after transfection. (C) KELLY cells were reverse transfected with siRNAs (siALK, siELF4 or control) and their proliferation was monitored with live-cell imaging. (D) mRNA levels of ALK and ELF4 48 h after transfection. (E) Morphological changes observed. (F) Quantification of the number of viable cells (MTS assay, 120 hours after transfection). (G) Quantification of differentiation (neurite outgrowth, 120 hours after transfection). (H) Proposed model of ALK and ELF4 synergy, based on STRING protein-protein interactions. Statistical significance of observed changes was determined by Student’s t-test with a nominal significance threshold of p = 0.05. * = p<0.05, ** = p<0.01, *** = p<0.001.

Figure 6.. ELF4 regulates a network of cell cycle genes.

(A) Gene ontology analysis of siELF4-downregulated genes with PANTHER reveals that ELF4 targets are preferentially associated with cell cycle pathways. (B) Cell cycle-related genes (GO: Regulation of Mitotic Cell Cycle) regulated by ELF4 form a highly connected network according to STRING. Link thickness is based on the confidence of the interaction. Nodes color-coded based on MCL clustering. (C) siELF4-upregulated genes are highly associated with neuronal processes (PANTHER), supporting the observation that ELF4 loss induces differentiation and that differentiation requires cell cycle exit. (D) Upregulated genes (GO: Neurogenesis) linked to neuronal function form a highly connected network according to STRING. Link thickness is based on the confidence of the interaction. Nodes color-coded based on MCL clustering.

Figure 7.. miR-124 and ELF4 regulate a common set of targets in opposing directions.

(A) miR-124 targets and siELF4-downregulated genes overlap significantly. Statistical significance was determined by hypergeometric distribution test. (B) MSigDb analysis of the 104 overlapping genes reveals highly enriched gene sets centered on cell cycle regulation (e.g. Fisher Dream Complex, GO Cell Cycle). (C) Protein-protein interactions between the co-targeted genes reveals a highly interconnected network. DREAM Complex targets (pink) account for a large majority of the overlapped genes (34/104). Link thickness is based on the confidence of the interaction (STRING).