Differential expression of RBM4 and PTB/nPTB proteins during muscle cell differentiation. C2C12 cells were cultured under differentiation conditions for 0–3 d. Protein samples were fractionated on separate gels followed by immunoblotting sequentially with antibodies specific for PTB, nPTB, and tubulin (top) or with RBM4, phosphorylated RBM4, and tubulin (bottom). Values given below the bands represent relative protein expression levels of each protein throughout differentiation. The graph shows the change in the ratio of total RBM4 to (n)PTB during cell differentiation.

RBM4 modulates PTB/nPTB protein expression levels in C2C12 cells. C2C12 cells were mock transfected (lanes 1 and 3) or transfected with vector expressing FLAG-RBM4 (lane 2) or RBM4 shRNA (lane 4). Immunoblotting was performed on independent membranes (separated by blank space) with pairs of antibodies in parallel that were specific for the FLAG epitope (FLAG-RBM4) or endogenous proteins as indicated.

RBM4 and PTB differentially modulate alternative splicing of muscle cell–specific exon selection. (A) Diagrams showing the alternatively spliced regions of premRNAs encoding mouse PTB, nPTB, palladin, fTnT, MEF2c (exons α1/2 and β), cTnT, IR, Ryr, and myocardin. M and NM represent muscle-enriched and nonmuscle-enriched mRNA isoforms, respectively. C2C12 cells were cultured in growth (lane 1) or differentiation (lane 2) medium for 3 d or transfected with the mock vector (lane 3) or expression vector of FLAG-tagged PTB, RBM4, or ASF (lanes 4–6). Total RNA was subjected to RT-PCR using gene-specific primers (Table II), and the PCR products were blotted for probing with a ³²P-labeled forward primer. Efficiency of muscle-enriched isoform production is reflected by the ratio of M/NM; the values shown below the lanes are averaged from three independent experiments. (B) C2C12 cells were transfected with the empty vector (lanes 1 and 2) or the PTB shRNA expression vector (lanes 3 and 4) and cultured under growth (lanes 1, 3, and 4) or differentiation (lane 2) conditions for 2 d. RT-PCR was performed to examine terminal exon inclusion of α-TM as described in Lin and Tarn (2005). GAPDH was used as the control. Protein samples were fractionated on separate gels followed by immunoblotting sequentially with anti-PTB and anti-tubulin (middle) or with anti-RBM4 and anti-tubulin in parallel (bottom). (C) C2C12 cells were transfected with the pαTM reporter and then cultured under growth (lane 1) or differentiation (lane 2) conditions for 3 d. In addition to the pαTM reporter, C2C12 cells were also transfected with an empty vector (lane 3) or the FLAG-PTB expression vector alone (lane 4) or together with increasing amounts of the FLAG-RBM4 expression vector (lanes 5–7). α-TM and cTnT transcripts were detected as in B and A, respectively. GAPDH was used as the control. Values represent the relative level of α-TM normalized to GAPDH in for each sample. Immunoblotting was performed using anti-FLAG. In all panels, band sizes for RT-PCR products are shown in parentheses.

Model for RBM4-mediated control of muscle cell–specific alternative splicing. (A) RBM4 activates exon 11 exclusion via binding to exon 11 together with its upstream or downstream intron. The resulting premature termination codon-containing PTB transcripts are targeted for degradation through NMD. PTB activates exon 11 skipping of its own transcripts via binding to exon 11 as well as upstream intronic elements. Purple and blue rectangles depict representative RBM4 and PTB binding sites, respectively. RBM4 and PTB may share a part of CU-rich binding sites. (B) RBM4 activates alternative exon skipping when it binds simultaneously to both exonic and intronic elements, whereas it may promote exon inclusion via binding only to (downstream) intronic elements. (C) In myoblasts, RBM4 and PTB may function similarly in splicing regulation of the transcripts (such as MEf2c) that are expressed at an early stage of differentiation to promote muscle cell–specific isoform expression. However, overexpression of PTB in general suppresses the production of mRNA isoforms that are specific to late stages of differentiation. RBM4 down-regulates PTB via AS-NMD and meanwhile acts oppositely to PTB or antagonizes its suppressive activity in alternative splicing control to achieve maximal yield of muscle cell–specific mRNA isoforms.

AS-NMD of PTB/nPTB transcripts occurs during myogenesis. C2C12 cells were cultured under differentiation conditions for 0–3 d. Before harvest, cells were mock treated or treated with cycloheximide (CHX). Total RNA was isolated for RT-PCR analysis followed by probing with ³²P-labeled primers specific to PTB exon 10 (lanes 1–8) or nPTB exon 9 (lanes 9–16). The identity of the transcripts is indicated. Asterisks represent nonspecific bands. Exon exclusion efficiency (percentage) is represented by the ratio of exon-skipped transcripts (PTB-1⁻¹¹ and PTB-4⁻¹¹ for PTB; nPTB⁻¹⁰ for nPTB) to total RNA species; the averaged values shown below the lanes were obtained from three independent experiments.

RBM4 activates exon skipping of PTB/nPTB premRNAs and is involved in differentiation-induced alternative splicing of PTB/nPTB. (A) C2C12 cells were transiently transfected with the mock vector or vector expressing FLAG-tagged PTB, RBM4, or ASF. (B) C2C12 cells were transfected with the mock vector (lanes 1–4) or RBM4-targeting shRNA expression vector (lanes 5–8), and then cultured in growth (G) or differentiation (D) medium. Before harvest, cells were mock treated or treated with cycloheximide (CHX). RT-PCR analysis and calculation of exon exclusion efficiency were performed in Fig. 1. Asterisks represent nonspecific bands. Immunoblotting was performed with anti-FLAG (A) and with anti-RBM4 and anti-tubulin in parallel on the same membrane (B).

Ultraconserved elements of PTB are responsible for RBM4-mediated regulation of PTB AS-NMD. (A) Diagram showing the mouse PTB genomic fragment encompassing exon 11 and a part of its adjacent introns. Putative intronic RBM4 and PTB binding sites are indicated by red and blue vertical lines, respectively. The CU-rich element in exon 11 is indicated by a thick orange line. The PTB genomic fragments of the pαE11 and its derived reporters (see panel B) are depicted. (B) Diagrams showing the chimeric reporters in which exon 9a of human α-TM was replaced by mouse PTB exon 11 with different lengths of the flanking intron sequences. (C) Each reporter, as indicated, was transiently transfected into C2C12 cells. Cells were cultured under growth (G) or differentiation (D) conditions for 1 d. Cells were treated with cycloheximide before harvest. For C–E, total RNA was isolated from the transfectants and subjected to RT-PCR using a forward primer specific to SV40 and reverse primers specific to α-TM intron 8 (i8) and exon 9b (9b) to yield products e8/11/9b and e8-i8, respectively. The PCR products were probed with ³²P-labeled SV40 primer. Exon 11 exclusion of the reporter transcripts is represented by the ratio of e8/11/9b to e8-i8; the averaged values shown below the lanes were obtained from three independent experiments, and then normalized to that of mock transfectants. GAPDH was used as the control. (D) Each indicated reporter was cotransfected with an expression vector encoding FLAG-tagged PTB, RBM4, or ASF. (E) Diagram showing the mutant pαE11 reporter in which the CU-rich sequence of exon 11 was mutated. The reporter assay for lanes 1–4 and 5–12 was performed as in C and D, respectively.

Overexpression of RBM4 promotes muscle cell differentiation. (A) C2C12 cells were mock transfected or transiently transfected with the expression vector encoding FLAG-tagged RBM4. Immunoblotting was performed with antibodies against myosin heavy chain (MHC), FLAG-epitope, and α-tubulin. (B) C2C12 cells were cultured under growth (G) or differentiation (D) conditions or transiently overexpressed FLAG-tagged RBM4. Indirect immunofluorescence was performed with anti-MHC. Bar, 10 µm.