MicroRNAs regulate the expression of the alternative splicing factor nPTB during muscle development

Abstract

Alternative pre-mRNA splicing determines many changes in gene expression during development. Two regulators known to control splicing patterns during neuron and muscle differentiation are the polypyrimidine tract-binding protein (PTB) and its neuronal homolog nPTB. These proteins repress certain exons in early myoblasts, but upon differentiation of mature myotubes PTB/nPTB expression is reduced, leading to increased inclusion of their target exons. We show here that the repression of nPTB expression during myoblast differentiation results from its targeting by the muscle-restricted microRNA miR-133. During differentiation of C2C12 myoblasts, nPTB protein but not mRNA expression is strongly reduced, concurrent with the up-regulation of miR-133 and the induction of splicing for several PTB-repressed exons. Introduction of synthetic miR-133 into undifferentiated C2C12 cells leads to a decrease in endogenous nPTB expression. Both the miR-133 and the coexpressed miR-1/206 microRNAs are extremely conserved across animal species, and PTB proteins are predicted targets for these miRNAs in Drosophila, mice, and humans. There are two potential miR-133-responsive elements (MRE) within the nPTB 3' untranslated region (UTR), and a luciferase reporter carrying this 3' UTR is repressed by miR-133 in an MRE-dependent manner. Transfection of locked nucleic acid (LNA) oligonucleotides designed to block the function of miR-133 and miR-1/206 increases expression of nPTB and decreases the inclusion of PTB dependent exons. These results indicate that miR-133 directly down-regulates a key splicing factor during muscle development and establishes a role for microRNAs in the control of a developmentally dynamic splicing program.

Figure 1.

nPTB protein is nearly eliminated during differentiation of C2C12 myoblasts, while mRNA levels increase. (A, top panel) Immunoblot of total protein lysates from C2C12 cells using α-nPTB IS2 (lanes 1,2) or α-PTB-NT antibodies (lanes 3,4). Cells were cultured in either growth or differentiation medium as indicated above each lane. (Bottom panel) GAPDH loading control. (B, top panel) Immunoblot of myosin heavy chains, a mature muscle marker in the same cells used in A and B. The bottom panel is blotted for GAPDH. (C) The top panel shows RT–PCR of the nPTB (lanes 1,2) and PTB (lanes 3,4) alternatively spliced regions in mRNA from proliferating (−) and differentiating (+) cells (same samples as used in A). (Bottom panel in all lanes) Equivalent amounts of RNA were used as indicated by the RT–PCR of GAPDH. (D) mRNA quantification for PTB and nPTB relative to β-actin mRNA between days 0 and 8 of differentiation treatment as determined by real-time PCR.

Figure 2.

nPTB contains conserved MREs for miR-133 and miR-1/206. (A) The MREs in both mouse and human nPTB are shown as red boxes. Untranslated regions are indicated in blue and coding sequence are in yellow. Each MRE is numbered, the predicted structure of each base-paired MRE/miRNA hybrid is diagrammed, and the predicted free energy of hybridization of the miR with the MRE is indicated. The top strand in each diagram represents the MRE sequence in the target mRNA, 5′–3′, and the bottom strand represents the indicated microRNA. Paired bases are indicated by a black oval, G:U pairs are indicated by two dots. All nucleotides in the miR and almost all in the MREs are identical between the mouse and human; where they differ, the mouse-specific nucleotides are shown in blue. (B) In mouse and human, the miR-133 (green arrows) and miR-1/206 (red arrows) microRNAs are coexpressed from three distinct loci. The chromosomes and the nucleotide distances between the pre-miR sequences are indicated. The arrows indicate the 5′–3′ direction of the mature microRNA. The miR locus on chromosome 18 is on the reverse strand. The blue underline represents a human EST sequence containing the miRNA precursors.

Figure 3.

miR-133 and miR-1/206 expression is restricted to muscle-derived cell lines. Size-fractionated samples of total RNA from tissue culture cells were probed for the presence of microRNAs. Note that the probe for miR-133 cannot distinguish between miR-133a and miR-133b, and the probe for miR-1/206 cannot distinguish between miR-1 and miR-206. miR-16 is a widely expressed microRNA detected in most cell lines. 5S rRNA was stained with ethidium bromide to normalize for loading. (A) Small RNA Northern probed for miR-133 expression in cell lines. The lane marked “C2C12” contains RNA from proliferating cells, whereas “C2C12 Diff” is RNA from cells treated for 4 d in differentiation medium. Similarly, “P19” is RNA from undifferentiated cells and “P19 Diff” is RNA from retinoic acid-treated P19 cells. (B) Small RNA Northern blot of RNA from differentiating C2C12 cells probed for miR-133 and miR-1/206. Differentiation medium was added at time 0 and RNA was harvested every 48 h through day 10. Note that the miR-1/206 and miR-16 panels are probings of the same blot and are normalized to the same 5S RNA panel.

Figure 4.

miR-133b represses expression of nPTB through elements in the 3′ UTR. (A) Proliferating C2C12 myoblasts were transfected with the indicated control siRNA (Luc), wild-type miR-133 (miR-133), or mutant miR-133 (miR-133 mut). At 48 h post-transfection, cells were harvested and Western blots were performed on total cell lysates are shown. Fluorescent secondary antibodies were used to quantify nPTB levels relative to a U1 70k control, which are graphed below. (B) A luciferase reporter plasmid carrying the Renilla luciferase coding sequence attached to the entire human nPTB 3′ UTR with either wild-type (wt) or mutant (mut) MREs is diagrammed. For each MRE/miR pair, the calculated free energy of hybridization (ΔG) in kilocalories per mole is shown on the right. (C) Luciferase reporters were cotransfected with the artificial miR-133 or control RNAs. The ratio of reporter (Renilla luciferase, Rn) to control plasmid (firefly luciferase, Ff) in relative luminescence units was normalized for each reporter to the buffer control and plotted as a percentage of the control value. Error bars represent the standard error for n = 3.

Figure 5.

nPTB and PTB proteins are increased in differentiating myoblasts when miR-133 is functionally blocked. (A) Undifferentiated cells (lane 1) or cells grown in differentiation medium for 72 h (lanes 2–5) were immunoblotted for nPTB (top panel), PTB (middle panel), or GAPDH (bottom panel). Note that the residual nPTB seen after 72 h is eliminated by longer culture in differentiation medium (Fig. 1A). Cells were transfected with the indicated LNA oligonucleotides and/or siRNAs before differentiation treatment. (Lane 3) An LNA-modified oligonucleotide against miR-124 was used as a control (α-miR-124). LNA oligos designed to block miR-133, miR-1, and miR-206 are labeled above the lanes (α-miR-133, α-miR-1, and α-miR-206, respectively). (B) Immunoblot showing nPTB protein levels relative to a β-actin loading control. (C) Quantification of fluorescent secondary antibody signal comparing nPTB expression against U1 70k as a loading control; based on three separate experiments.

Figure 6.

Alternative splicing of PTB-dependent muscle-specific exons is altered by blocking miR-133. (A) Lysates of cells from Figure 5 were processed for total mRNA, and the splicing of the indicated transcripts was assayed by RT–PCR. Lane 1 is undifferentiated cells (UD). Lanes 2–6 are cells differentiated for 72 h after transfection with the indicated LNA oligos and/or siRNAs. Polyacrylamide–urea gels with RT–PCR products incorporating 5′-³²P-end-labeled primers are shown to the left, with quantification of the exon inclusion graphed to the right. The exon included and skipped products are indicated by arrows. All data are graphed as the average of three separate experiments, and the standard error is indicated by error bars. (B) Bars showing the fold increase in each exon in response to double knockdown of PTB and nPTB by RNAi extend to the right. This is measured against cells treated with anti-miR-133 LNA oligo and a control siRNA. Bars showing the fold decrease in exon inclusion as a result of anti-miR-133 LNA oligo treatment extend to the left. Error bars show the variation between three separate transfections.