Lin-28B transactivation is necessary for Myc-mediated let-7 repression and proliferation

Abstract

Direct control of microRNA (miRNA) expression by oncogenic and tumor suppressor networks results in frequent dysregulation of miRNAs in cancer cells and contributes to tumorigenesis. We previously demonstrated that activation of the c-Myc oncogenic transcription factor (Myc) broadly influences miRNA expression and in particular leads to widespread miRNA down-regulation. miRNA transcripts repressed by Myc include several with potent tumor suppressor activity such as miR-15a/16-1, miR-34a, and let-7 family members. In this study, we have investigated mechanisms downstream of Myc that contribute to miRNA repression. Consistent with transcriptional down-regulation, Myc activity results in the decreased abundance of multiple miRNA primary transcripts. Surprisingly, however, primary transcripts encoding several let-7 miRNAs are not reduced in the high Myc state, suggesting a posttranscriptional mechanism of repression. The Lin-28 and Lin-28B RNA binding proteins were recently demonstrated to negatively regulate let-7 biogenesis. We now show that Myc induces Lin-28B expression in multiple human and mouse tumor models. Chromatin immunoprecipitation and reporter assays reveal direct association of Myc with the Lin-28B promoter resulting in transcriptional transactivation. Moreover, we document that activation of Lin-28B is necessary and sufficient for Myc-mediated let-7 repression. Accordingly, Lin-28B loss-of-function significantly impairs Myc-dependent cellular proliferation. These findings highlight an important role for Lin-28B in Myc-driven cellular phenotypes and uncover an orchestration of transcriptional and posttranscriptional mechanisms in Myc-mediated reprogramming of miRNA expression.

Fig. 1.

Posttranscriptional repression of let-7 expression by Myc correlates with Lin-28B induction. (A and B) qPCR analysis of the abundance of pri-miRNAs (A) and mature let-7g (B) in P493-6 cells with high or low Myc expression. (C) RT-PCR analysis of Lin-28 and Lin-28B expression in P493-6 cells. DNA Plasmids bearing the ORF of Lin-28 or Lin-28B were used as templates for positive controls. (D) qPCR analysis of Lin-28B expression in P493-6 cells. (E) Western blot of Lin-28B abundance in P493-6 cells after removal of tet. Signals were normalized to α-tubulin levels before calculating fold changes. (F) qPCR analysis of Lin-28B expression in murine B cell lymphomas. Two independent tumors from MycON and MycOFF neoplasms were assayed. (G) qPCR analysis of Lin-28B expression in Ras-transformed p53-null colon carcinomas. Four independent tumors from Ras/Myc and Ras-only neoplasms were assayed. For all qPCR analyses, error bars represent standard deviations derived from 3 independent measurements.

Fig. 2.

Lin-28B is a direct Myc target gene. (A) Schematic representation of the genomic region near the transcription start site of Lin-28B. Putative Myc binding sites are indicated (CACGTG or CATGTG); those in red are conserved between human and mouse. The plot depicts evolutionary conservation and was generated from the University of California, Santa Cruz genome browser (human genome March 2006 assembly). qPCR amplicons are represented by numbered lines. (B) qPCR analysis of Myc chromatin immunoprecipitates. Fold enrichment represents signal obtained after Myc immunoprecipitation relative to signal obtained with irrelevant antibody. Error bars represent standard deviations derived from 3 independent measurements. (C) Activity of promoter reporter constructs in P493-6 cells with high or low Myc expression. Values shown represent relative firefly luciferase activity produced from each reporter construct normalized to renilla luciferase activity produced from a cotransfected control vector. Error bars represent standard deviations derived from 3 independent measurements. An internal ribosome entry site (IRES) upstream of luciferase was included in these vectors because several out-of-frame translation initiation codons are present in the cloned genomic fragments. E, potential Myc binding site (E-box).

Fig. 3.

Lin-28B knockdown reverses let-7 repression and slows proliferation. (A and B) P493-6 cells were grown the absence of tet (high Myc state) and transfected with 2 different siRNAs targeting Lin-28B or negative control siRNA. qPCR was used to measure Lin-28B (A) or mature let-7g abundance (B). (C) Proliferation of P493-6 cells after siRNA delivery. Error bars represent standard deviations of measurements of triplicate transfections. (D) Proliferation of P493-6 cells after transfection of siRNA and miRNA inhibitors. Error bars represent standard deviations of measurements of triplicate transfections. P value calculated by 2-tailed t test.

Fig. 4.

Enforced Lin-28B expression fully recapitulates Myc-mediated let-7 repression but has limited effects on other Myc-repressed miRNAs. (A) qPCR analysis of Lin-28B expression in P493-6 cells infected with MSCV-empty or MSCV-Lin-28B. (B) Northern blot analysis of miRNAs in retrovirally-infected cells. We used the conditions described in ref. that prevent cross-hybridization between let-7 family members. tRNA^Lys served as a loading control. (C) Comparison of Myc-repressed and Lin-28B-repressed miRNAs as assayed by microarray. Individual let-7 family members are not distinguishable on the array because of cross-hybridization and are therefore treated as 1 group. (D) Northern blot validation of miRNA microarray results. (E) Motifs complementary to the let-7 seed region are enriched in the 3′ UTRs of Lin-28B-up-regulated transcripts. Expected and observed counts for the corresponding motifs in up-regulated 3′ UTRs are shown. Rank and percentile columns indicate the relative enrichment of a given motif compared with all 16,384 possible heptamer motifs (above dotted line) or all 4,096 possible hexamer motifs (below dotted line). (F) Comparison of Myc-induced and Lin-28B-induced mRNAs as assayed by microarray. Gene sets characteristic of multiple types of stem cells are enriched in the transcripts up-regulated by both proteins. P values were calculated by the Molecular Signatures Database available as part of the Gene Set Enrichment Analysis toolset (www.broad.mit.edu/gsea/msigdb/annotate.jsp) and are not corrected for multiple hypothesis testing. The rank of enrichment of each gene set is shown relative to all 5,452 genes sets tested.