Crossregulation and Functional Redundancy between the Splicing Regulator PTB and Its Paralogs nPTB and ROD1

1Alternative splicing (AS) allows multiple mRNA isoforms to be produced by individual genes. These commonly encode distinct protein isoforms, thereby enriching the repertoire of proteins available to organisms. AS can also quantitatively control gene expression by producing mRNAs that are degraded by nonsense-mediated decay (NMD), a phenomenon that appears to be widespread among splicing regulators themselves and which we refer to here as AS-NMD.

2Alternative splicing decisions are influenced by enhancers and silencer cis-acting elements that mediate their activity by binding activator and repressor proteins. Activators are commonly members of the SR family of proteins, whereas hnRNP proteins often act as repressors. One of the better-investigated mammalian repressor proteins is polypyrimidine tract binding protein (PTB, also known as PTBP1 and hnRNP-I, reviewed in ). PTB represses many exons, including a number of muscle and neuron-specific exons (e.g., ). PTB binds to splicing silencers with core sequences consisting of UCUU, CUCUCU, or related motifs. In some cases, a single silencer is sufficient to mediate PTB repression, but most PTB-regulated exons have multiple PTB binding sites. Most mammals contain two PTB paralogs that are expressed in a tissue-restricted manner. nPTB/brPTB/PTBP2 is expressed mainly in neurons, whereas ROD1 is expressed in hematopoietic cells. Rodents contain an additional paralog, smPTB. PTB and its paralogs share >70% amino acid sequence identity and a common arrangement of four RRM-type domains. The activity of nPTB has been most extensively studied in the context of the neuron-specific splicing of the N1 exon of SRC where it is less repressive than PTB. Nonetheless, it is likely that nPTB and the other paralogs act as repressors of at least some exons.

3PTB itself is subject to alternative splicing. Inclusion or skipping of exon 9 produces the PTB1 and PTB4 isoforms, which can vary in activity. In contrast, exon 11 skipping produces a frameshifted mRNA that is degraded by NMD. PTB protein itself promotes PTB exon 11 skipping in an autoregulatory feedback loop. The equivalent nPTB exon is also skipped and has a similar arrangement of potential regulatory elements, suggesting that it could also be regulated by PTB and/or nPTB. Indeed nPTB exon 10 lies within one of the genomic ultraconserved elements that have been associated with AS-NMD events.

4A range of global approaches have recently been harnessed with the aim of deciphering “cellular codes” composed of particular complements of splicing regulators and “RNA codes” composed of particular arrangements of regulatory sequence modules that together define cell-specific programs of splicing (reviewed in ). A typical approach involves identification of the set of RNAs bound by a regulatory factor (e.g., ) or the set of splicing events affected by the factor. The latter can be identified by perturbing the cellular levels of the splicing regulator and then analyzing RNA with splice-sensitive microarrays. Such analysis has been carried out for Drosophila SR and hnRNP proteins and the mammalian neuron-specific Nova proteins. As a complementary approach, we decided to use quantitative gel-based proteomics to analyze the consequences of PTB knockdown. Ideally, alterations in the levels of alternatively spliced isoforms would appear as pairs of reciprocally varying spots on 2D gels. Alterations in individual spots would be consistent with AS-NMD events or with PTB's documented roles at other levels of gene expression, including 3′ end processing, regulation of translation, and RNA stability.

5Despite the weight of evidence for the widespread roles of PTB, we observed very little effect of PTB RNAi upon the HeLa cell proteome. The single exception was upregulation of the usually neuronal nPTB, which resulted from a large increase in nPTB exon 10 inclusion. In model systems of PTB-regulated splicing, double knockdown of both PTB and the upregulated nPTB caused greater changes in AS than knockdown of PTB alone. Strikingly, we found that inclusion of cassette exon 2 of ROD1, which is essential for generation of an mRNA with an open reading frame initiating at the first AUG codon, was also upregulated by knockdown of PTB and nPTB together. Moreover, proteomic analysis showed numerous alterations in protein expression upon PTB+nPTB double-knockdown cells and allowed us to identify novel PTB/nPTB-regulated events. Our data indicate that nonproductive AS is used in both autoregulation and crossregulation of PTB, nPTB, and ROD1 and that nPTB is able to replace many of the functions of PTB in HeLa cells.

6Our initial aim was to carry out a proteomic analysis of HeLa cells in which PTB had been knocked down. RNAi was performed using a PTB-specific short-interfering RNA (siRNA) duplex, P1, and a control duplex, C2, resulting in ∼80%–90% PTB knockdown (Figure 1A, top panel, and data not shown). Samples were analyzed by 2D-difference gel electrophoresis (DiGE). The control protein sample was labeled with Cy3 fluorescent dye, and the PTB-RNAi sample with Cy5. The two samples were then pooled and separated on a 2D gel over a pI range of 3–10 (Figure 1B), and two fluorescent images were taken. After false coloring of the merged images, proteins whose levels have changed between the two samples appear red or green, whereas yellow spots indicate proteins whose levels have remained constant. Despite our expectation that knockdown of PTB would cause numerous changes in the HeLa proteome, very few up- or downregulated proteins were observed. The only consistent changes that were observed with reciprocal dye labeling, biological and technical repeats, and with a second PTB siRNA (P2) were a group of spots in the high pI (>9) and 55–60 kDa molecular weight range (Figure 1B and Figure S1 in the Supplemental Data available with this article online). The downregulated spots (green in Figure 1B) corresponded in size and pI to the PTB1 and PTB4 isoforms. We suspected that the adjacent upregulated spots might correspond to one of the PTB paralogs. This might account for the lack of other observed changes if the paralog were able to compensate for the loss of PTB. Comparison of the predicted MWt and pI of nPTB and ROD1 suggested that the upregulated spots were likely to correspond to nPTB. Western blot analysis of control and PTB-RNAi-treated samples confirmed that PTB knockdown was accompanied by an increase in nPTB protein levels (Figure 1A, lane 2, middle panel). Western blot of the 2D gel with nPTB-specific antibodies confirmed that the upregulated spots were nPTB (Figure 1C). The run of spots of the same size but varying pI is likely to correspond to phosphorylation variants of nPTB. Analysis of PTB and nPTB levels after treatment with varying doses of P1 siRNA indicated an inverse relationship between PTB and nPTB levels, with nPTB progressively increasing as PTB decreased (Figure S2). In order to test whether the upregulated nPTB compensates for loss of PTB, we carried out double knockdown of both PTB and nPTB (Figure 1D). 2D DiGE comparing control and PTB+nPTB knockdown cells showed numerous up- and downregulated spots (Figure 1E). By contrast with the limited proteomic consequences of PTB knockdown (compare Figures 1B and 1E), this suggests that the upregulated nPTB buffers against the effects of loss of PTB.

7Skipping of exon 10 of nPTB produces an mRNA that is a predicted NMD substrate. nPTB exon 10 has a similar arrangement of splice site elements to PTB exon 11 with a predicted distant branchpoint 389 nt upstream of exon 10 and multiple surrounding potential PTB binding sites (Figure S3). We therefore analyzed nPTB exon 10 splicing as a candidate cause of the nPTB upregulation in PTB knockdown cells.

8RT-PCR analysis revealed a substantial increase in the percentage of nPTB exon 10 inclusion from 63% to 96% upon PTB knockdown (Figure 2C, lanes 1 and 2), consistent with repression by PTB of nPTB exon 10. Similar results were obtained with a second PTB siRNA (data not shown). Skipping of nPTB exon 10 generates a predicted NMD substrate, so RT-PCR analysis likely underestimates exon 10 skipping. We therefore performed RNAi of UPF1, an essential NMD factor, reducing its levels by 80%–90% (Figure 2B). This resulted in a decrease in the observed levels of nPTB exon 10 inclusion to 16% (Figure 2C, lanes 1 and 3), confirming that nPTB exon 10 skipping leads to NMD. A second UPF1 siRNA showed similar results (data not shown). Comparison of UPF1 knockdown alone with UPF1/PTB double knockdown (Figure 2C, lanes 3 and 4) illustrates the full extent of the effect of PTB on alternative splicing of exon 10 of nPTB—an increase in exon inclusion from 16% to 96%. Quantitative real-time PCR analysis revealed that the absolute levels of nPTB RNA were elevated 4.4-fold upon PTB knockdown, in reasonable agreement with the increases in nPTB protein revealed by 2D DiGE (2.4- ± 0.6-fold, Figure S1) and by western blot (∼3-fold, Figure 1A). Taken together, these data demonstrate that, under the influence of repression by PTB, nPTB in HeLa cells is predominantly spliced to skip exon 10, resulting in NMD of the majority of its RNA.

9We also asked whether splicing of the hematopoietic-restricted PTB paralog ROD1 might be regulated by PTB and nPTB. Unlike PTB exon 11 and nPTB exon 10, ROD1 exon 10 shows no EST evidence of alternative splicing and appears to have a conventionally located branchpoint 21 nt upstream with little sequence conservation further upstream. In contrast, ROD1 exon 2 has a number of features in common with the regulated PTB and nPTB exons. It is identical in length (34 nt), is skipped in many ESTs, and has a conserved region of nearly 200 nt upstream, which encompasses an AG dinucleotide exclusion zone of 168 nt, candidate distant branchpoints up to −160, and multiple potential PTB binding motifs (Figure S3). Indeed, we found exon 2 to be mainly skipped in HeLa cells (Figure 2F). Its inclusion was slightly upregulated upon knockdown of PTB or nPTB alone and was included to the greatest extent upon knockdown of PTB and nPTB together (Figure 2F). Inclusion of ROD1 exon 2 produces an mRNA with a single open reading frame initiating in exon 1, whereas exon 2 skipping leads to introduction of a premature stop codon in exon 3. UPF1 knockdown had no effect upon the ratio of exon 2 inclusion or skipping, suggesting that the exon 2-skipped isoform is not an NMD substrate. This could be because the stop codon is only 20 codons from the initiator AUG. Nevertheless, the exon 2-skipped isoform is unlikely to be efficiently translated as the short open reading frame would repress translational initiation further downstream. Therefore, not only is nPTB expression regulated by PTB-induced AS-NMD in HeLa cells, but ROD1 is regulated by both PTB and nPTB.

10The 2D DiGE data and the analysis of ROD1 exon 2 splicing suggested some functional redundancy between PTB and nPTB. To examine this further, we investigated the effects of PTB and PTB+nPTB knockdown upon three characterized systems of PTB-regulated splicing. Splicing of PTB exon 11 was analyzed by using an NMD-insensitive minigene reporter (pG11; Figure 3B). Knockdown of PTB resulted in an increase in exon 11 inclusion (Figure 3C, compare lanes 2 and 3 to 1). When the increase in nPTB protein levels was prevented by double knockdown of PTB and nPTB, there was an even larger increase in exon 11 inclusion (Figure 3C, lanes 4–7), suggesting that nPTB represses PTB exon 11. As expected, treatment with nPTB siRNAs (N1 and N3) alone had little effect on splicing (Figure 3D).

11We next tested the ability of transfected PTB and nPTB to reverse the effects of PTB+nPTB knockdown (Figures 3E and 3F). P1 siRNA targets the PTB 3′UTR, which is not in the expression constructs, whereas nPTB was expressed from a codon-optimized construct (F. Robinson and C.W.J.S., unpublished data), which was resistant to the nPTB siRNAs. Add back of PTB1, PTB4, nPTB, and smPTB all reversed the effects of double knockdown (Figure 3E, lanes 12–15). This demonstrates that nPTB is equally repressive as PTB for repression of PTB exon 11.

12Splicing of α-actinin was analyzed with the pA minigene, which contains the NM and SM alternative exons flanked by the constitutive EF1a and EF2 exons (Figure 4A). RNAi of PTB had no detectable effects upon pA splicing in many experiments and only minor effects in others (Figure 4B, lanes 2 and 3). Treatment with nPTB siRNAs was also without effect (Figure 4B, lanes 4 and 5). When the increase in nPTB was prevented by double knockdown, there was a substantial switch in splicing, with an increase in double NM and SM inclusion, a decrease in NM inclusion, and an increase in SM inclusion (Figure 4B, lanes 6–9). Overexpressed PTB1, PTB4, nPTB, and smPTB were all able to reverse the effects of knockdown (Figure 4C, lanes 7, 11, 15, and 19).

13At the 5′ end of the αTM gene (HUGO TPM1), exon 2 is specifically included in smooth muscle due to PTB-mediated repression of exon 3. HeLa cells do not transcribe TPM1, and transfected constructs do not skip exon 3. However, the pTΔBP reporter construct with point mutations in the exon 3 branchpoint produces substantial amounts of exon 3-skipped product, which is dependent upon the characterized negative regulatory elements (C. Gooding and C.W.J.S., unpublished data). Whereas knockdown of PTB alone reduced the levels of exon skipping (Figure 5B, lanes 1–3), PTB+nPTB knockdown reduced skipping to less than 2% (lanes 4–7). The effects of PTB knockdown upon endogenous Tpm1 splicing were analyzed in rat PAC1 smooth muscle cells. Knockdown of PTB was again accompanied by an increase in nPTB (Figure 5C, middle panel, lane 5), confirming that the crossregulation of PTB and nPTB is not species or cell specific. Rat nPTB was knocked down with the N1 siRNA (Figure 5C, lane 7). Knockdown of PTB alone resulted in a decrease in TM exon 2 inclusion (Figure 5D, lane 2), whereas N1 siRNA had no effect. Double knockdown of PTB+nPTB once again produced the largest effect with an overall 5-fold decrease in exon 2 inclusion (Figure 5D, lane 4).

14The preceding data show that knockdown of both PTB and nPTB is necessary to maximize the chances of identifying PTB targets. Although 2D DiGE was useful for preliminary characterization of proteomic changes upon PTB+nPTB knockdown (Figure 1E), intergel variation (e.g., Figure S1) can make reliable identification of up- or downregulated spots difficult. We therefore adopted a more sophisticated procedure, biological variance analysis (BVA), which incorporates a Cy2-labeled internal pooled standard comprising equal aliquots of all control and knockdown samples. Six gels were run with the pooled standard and the Cy3- and Cy5-labeled biological replicate samples (Figure S4). To reduce the likelihood of identifying secondary events, protein was harvested 48 hr earlier than in the preceding experiments. We identified 81 regulated spots (p < 0.05), of which 16 were upregulated and 15 downregulated more than 1.2-fold (Figure 6A and Figure S4). Note that, in contrast to the DiGE gels (Figure 1), the gel in Figure 6A is imaged with all three channels, so spots that are confidently identified as up- or downregulated do not appear red or green. A number of these spots were picked for LC-MS/MS sequencing. In some cases, the identified peptides clearly indicated the presence of specific protein isoforms. For example, spot 1782 was upregulated 1.6-fold and contained multiple β-tropomyosin (TPM2) peptides, including some diagnostic of muscle-specific exons 7 and 10. TPM2 had also been identified from the initial DiGE gels (Figure 1E). RT-PCR analysis verified that splicing of the two mutually exclusive exon pairs 6 and 7, and 10 and 11 was altered substantially upon PTB+nPTB knockdown, with an intermediate response to PTB knockdown (Figure 6B). In both cases, splicing was switched toward the muscle-specific exons 7 and 10, consistent with the peptide data. Spot 1490 contained annexin 7 (ANXA7) peptides and was downregulated 1.3-fold. Exon 6 of ANXA7 is a cassette exon used primarily in striated muscle and brain. The position of spot 1490 is consistent with the 50 kDa exon 6-skipped isoform. Peptides diagnostic of exon 6 inclusion or skipping were not identified. Nevertheless, RT-PCR showed a substantial shift toward exon 6 inclusion in double knockdown, but not in single PTB or nPTB knockdown (Figure 6C), again consistent with the proteomic data. Spot 1151 was downregulated 1.3-fold and contained peptides corresponding to two isoforms of pyruvate kinase 3 (PKM2) containing mutually exclusive exon 9 or 10. RT-PCR demonstrated ∼4-fold increased inclusion of PKM2 exon 9 relative to exon 10 in response to either PTB or PTB+nPTB knockdown (Figure 6D). Because the two isoforms comigrate and are present in the same spot, in this case it is not clear how a shift in the ratio of mutually exclusive exons could explain the observed downregulation of spot 1151. α-glucosidase II (GANAB) was identified in a downregulated (∼1.3-fold) spot from the original 2D DiGE gels (Figure 1E) and in spot 485 from the BVA experiment. GANAB cassette exon 6 was found to be partially upregulated by either PTB or nPTB knockdown and to show greater upregulation in the combined knockdown (Figure 6E). Although no isoform-specific peptides were identified, the position of the downregulated protein spot was consistent with the exon-skipped isoform, again indicating agreement between the protein and RNA data. Finally, a charge train of downregulated spots (724, 734, 736–9, −1.2- to −1.5-fold) all contained peptides corresponding to lamin A (LMNA). Extensive alternative splicing at the 3′ end gives rise to the major lamin A/C isoforms. However, an alternative 5′ end exon (chr 1 154,363,150–154,363,330) usually associated with an internal promoter appeared to be a better candidate PTB-regulated exon. Splicing of this exon between the usual exons 1 and 2 would lead to introduction of a premature stop codon, leading to NMD. RT-qPCR indicated that splicing of exon b to both exon 1 and exon 2 (Figure 6F, light bars and dark bars, respectively) increased by 2-fold with PTB+nPTB knockdown and by 4-fold with PTB+nPTB+UPF1 knockdown. Inclusion of this exon could account for the downregulation of the LMNA/C-containing spots. One additional spot (832) containing LMNA/C was upregulated. This could be due to other proteins present in the same spot.

15Therefore, quantitative proteomic analysis has allowed identification of a number of PTB-regulated alternative splicing events, most of which are regulated by both PTB and nPTB, but with varying sensitivities to single or combined knockdown.

16The data presented here shed interesting light upon crossregulation between PTB, nPTB, and ROD1, as well as indicating a high degree of functional redundancy between PTB and nPTB. PTB was already known to autoregulate its own expression by repressing its own exon 11. The feedback loop may serve to prevent overexpression of PTB, to upregulate expression upon PTB export to the cytoplasm, and to reduce stochastic noise in PTB expression. Although splice-sensitive microarray analysis of cells knocked down for UPF1 has been used to argue against the widespread harnessing of regulated alternative splicing to NMD, more recent computational and array analyses have indicated that AS-NMD is widespread among both SR proteins and hnRNP proteins, where it is associated with ultraconserved genomic regions. Similar findings had previously been reported with plant SR protein genes. These reports speculate that the AS-NMD events might generally be involved in autoregulatory feedback loops, as previously demonstrated for PTB and SC35. Our findings demonstrate that such nonproductive splicing events can also be harnessed to crossregulate the expression of families of regulators. In HeLa cells, PTB switches nPTB off by promoting skipping of nPTB exon 10 (Figures 2A–2C), while PTB and nPTB both promote the nonproductive splicing of ROD1 (Figure 2F). Although nPTB exon 10 is associated with an ultraconserved genomic region, PTB exon 11 and ROD1 exon 2 also have extended conserved upstream regions that coincide with extended AG exclusion zones, indicative of branchpoint sequences remote from the exon (Figure S3). The reason for coincidence of ultraconserved regions with AS-NMD events is a mystery. However, for the PTB, nPTB, and ROD1 exons, the extended regions of upstream conservation can be explained in part by the distant branchpoint arrangement and the likely location of regulatory elements.

17nPTB protein is usually expressed with neuronal specificity in the retina and brain, although nPTB transcripts can be detected in cells in which nPTB protein is not present. The regulated skipping of nPTB, PTB, and ROD1 exons provides a mechanism for the cell to respond to excessive levels of any or all of the PTB paralog proteins. The relative sensitivities of the exon-skipping events appear to be tuned to facilitate PTB expression in preference to nPTB or ROD1. Thus the normal levels of PTB protein in HeLa cells result in 80% inclusion of PTB exon 11, 16% inclusion of nPTB exon 10, and 2% inclusion of ROD1 exon 2 (Figure 2). Reduced PTB protein levels may therefore be a prerequisite for nPTB or ROD1 protein expression. In our experiments, this was achieved by PTB knockdown in HeLa and PAC-1 cells, but in vivo, this might result from developmentally programmed reduction in PTB transcription. We also showed that nPTB protein is equally effective as a repressor of PTB exon 11 (Figure 3) and so potentially could modulate PTB levels. Whether this control is actually exerted in vivo is an open question. The crossregulation and partial functional redundancy of PTB and its paralogs will have implications for the analysis of transgenic knockout mice. On the one hand, many of the normal functions of the knocked-out factor may be masked by compensatory upregulation of the paralog(s). However, this may have the advantage of restricting phenotypes to the events that are critically dependent upon the individual paralog that has been knocked out.

18Expression of nPTB appears to be intricately regulated at multiple levels, in addition to the PTB-induced exon 10 skipping. For example, upregulation of the miR133 family of miRNAs during striated muscle differentiation translationally silences nPTB expression. The reduction in PTB+nPTB activity thereby allows various muscle-specific splicing events to occur, such as Tpm2 exon 7 and 10 selection (Figure 6B). Another unusual feature of nPTB expression, shared by ROD1, is that it has an extremely suboptimal codon content, with a high proportion of codons ending in A or U, which are relatively infrequent in the human genome. This limits its ability to be translated in vitro or overexpressed in transfected cells and may also modulate its expression in vivo. Indeed we were only able to overexpress nPTB (Figures 3 and 4) by using a codon-optimized construct with 211 silent mutations (F. Robinson and C.W.J.S., unpublished data).

19The fact that very few changes were observed in 2D DiGE analysis of PTB knockdown samples unless nPTB was also knocked down (Figure 1) suggests a high extent of functional redundancy between PTB and nPTB in HeLa cells. The functional redundancy is not surprising in view of the >70% amino acid identity between PTB and its paralogs. Nevertheless, it contrasts with the example of the SRC N1 exon, where nPTB is much less repressive. PTB and nPTB also have a differential effect upon internal ribosome entry segment (IRES) driven translation. It may be that N1 is representative of a subset of exons that are differentially sensitive to PTB nPTB, whereas many other splicing events are equally repressed by either paralog. Among the AS events that we analyzed (Figures 2–6), the sensitivity to knockdown of PTB alone varied, with substantial effects upon PTB exon 11 (Figure 3) but negligible effects upon actinin splicing (Figure 4). It may be that there is a continuum of AS events with differential sensitivities to repression by PTB or nPTB. Presumably, this differential sensitivity to PTB and nPTB is written into an “RNA code” and might be decipherable upon analysis of a sufficient number of splicing events that are coregulated. A similar situation probably pertains with events regulated by PTB and ROD1 in immune cells.

20The role of PTB in regulating splicing of FGFR2 exon IIIb and PTB exon 11 has been demonstrated previously by RNAi of PTB. However, before uncovering the upregulation of nPTB upon PTB knockdown, we were unable to obtain convincing RNAi evidence for the role of PTB in regulating Tpm1 (Figure 5) or α-actinin (Figure 4) splicing. Likewise the previous evidence for PTB regulation of mammalian TPM2 exon 7 was based upon PTB binding to an intronic silencer rather than functional evidence. Splicing of a chicken β-TM construct in HeLa cells was shown to respond partially to RNAi of PTB. The lack of complete derepression was taken to indicate the involvement of other repressor factors. Our data suggest that the additional repressor is nPTB.

21Our data suggest that quantitative proteomics might provide a useful complementary approach to methods for global analyses of alternative splicing. Here we identified eight splicing events that are targets of PTB in the PTBP2, ROD1, TPM2, ANXA7, LMNA, PKM2, and GANAB genes (Figures 2,6). The TPM2 and PKM2-regulated exons are usually muscle specific, whereas ANXA7 exon 6 is used in muscle and brain. This is consistent with previous indications that PTB commonly represses muscle and brain-specific exons. It will be interesting to see whether a strong enrichment of muscle and neuron-specific exons is maintained as we increase the number of validated PTB-regulated events. Likewise, although the number of validated events currently prevents a robust analysis for enrichment of particular regulatory sequence motifs, it is clear that the newly discovered PTB-regulated events share some common features with the previously characterized examples (Figure S3). These include a high frequency of known PTB binding motifs and potential branchpoint sequences located beyond the conventional ∼40 nt distance from the exon. We aim to compare analysis at the proteome level with the use of splice-sensitive microarrays in order to rigorously determine the extent to which proteomic analysis can contribute to understanding the global roles of splicing regulators. In view of the fact that PTB and other splicing regulators also influence gene expression at other levels, including translation, it seems likely that proteomic analysis will usefully complement splice-sensitive transcriptome analysis. Nonetheless, the data presented here already demonstrate that unanticipated insights can be gained from this approach.

22Minigene reporter constructs pG11, pA, and pTΔBP have been described previously. Constructs for overexpression of PTB1, PTB4, nPTB, and smPTB were generated from pCMV-β-gal by replacing the β-gal ORF with human PTB1, PTB4, nPTB, or rat smPTB. The PTB1 and PTB4 constructs also contained the His6 and XP tags from the pcDNA3.1/HisC vector. Human nPTB was overexpressed by using a codon-optimized construct (F. Robinson and C.W.J.S, unpublished data).

23HeLa and PAC1 smooth muscle cells were cultured under standard conditions in DME medium with glutamax and 10% fetal calf serum. siRNA transfection of HeLa cells was carried out as described with the following siRNAs: 10 pmol P1, N1, N3, and U1 and 110 pmol P2, all made up to total 120 pmol with C2. siRNA transfection of PAC1 cells was the same, except 3 × 104 cells were seeded, siRNA transfection used 2 μl oligofectamine (Invitrogen), and cells were harvested on day 6. mRNA targets for gene specific knockdown were as follows: human PTB P1 and P2 and rat PTB P3, 5′-GCCUCUUUAUUCUCUUCGG-3′; human nPTB N1, 5′-GAGAGGAUCUGACGAACUA-3′; human nPTB N3, 5′-UAAGAAACCUGGAUCCAAA-3′; human UPF1 U1 and human UPF1 U2, 5′- GCUCCUACCUGGUGCAGUAUU-3′ control C2. siRNAs were purchased from Dharmacon.

24Protein was harvested with RIPA buffer supplemented with protease inhibitors. Protein concentration was determined by the Bradford method. Proteins were separated on 15% SDS-PAGE gels, transferred to PVDF membrane, and detected by enhanced chemiluminescence. Primary antibodies used were mouse anti-PTB BB7 monoclonal for HeLa cells, rabbit anti-PTB polyclonal for PAC1 cells, rabbit anti-nPTB, rabbit anti-smPTB, mouse anti-Xpress (Invitrogen), rabbit anti-UPF1, and rabbit anti-ERK (Santa Cruz Biotechnology Inc) with Donkey anti-rabbit and anti-mouse IgG horseradish peroxidase conjugate secondary antibodies (Jackson ImmunoResearch Laboratories Inc). Where necessary, western blot images were scanned with a flat-bed scanner, and band intensities were quantitated with Phoretix 1D Advanced software.

25RNA was harvested with TRI reagent (Sigma), and splicing patterns were analyzed by RT-PCR. RNA was treated with DNase I (Ambion) before reverse transcription using either 200 U Superscript II (Invitrogen) or 10 U AMV RT (Promega). Oligonucleotides for RT were CGRT, 5′-TAGTTGTACTCCAGCTT-3′, for pG11, and SVRT, 5′-GCAAACTCAGCCACAGGT-3′, for pA, oligo dT, or random hexamers for endogenous mRNAs. PCR was carried out on 1/25 of the RT reaction. Standard PCR conditions were 30 cycles 94°C 30 s, annealing 1 min, 72°C 1 min. Other conditions are given in Table 1. In all cases, the reverse primer was 32P end labeled (except for the first round PCR of pA and PKM2), and labeled products were resolved on 5% polyacrylamide gels (12% for Figures 6C and 6D) and quantitated by Molecular Dynamics Storm 840 phosphorimager using Molecular Dynamics ImageQuant software. Quantitative PCR (QPCR) was performed on a Rotor-GeneTM 6000 (Corbett Research) with SYBR green (Applied Biosystems) after reverse transcription of 1 μg total RNA using Superscript II (Invitrogen) and oligo dT. Fold change in mRNA levels was calculated relative to the control sample and normalized to GAPDH levels. QPCR of endogenous nPTB transcripts was performed as a two-step reaction. Except where otherwise indicated, values for exon inclusion are shown as mean ± SD of three biological repeats.

26Protein samples were harvested in ASB14 lysis buffer (8 M urea, 2% ASB14, protease inhibitors 1 mM PMSF, 1 mg/ml aprotinin, and 1 mg/ml leupeptin). CyDye labeling, 2D protein separation, gel imaging, and analysis were performed as described previously. Gels were stained with colloidal Coomassie, and proteins were excised manually. In the BVA experimental set up (Figure 6 and Figure S4), six biological replicates from control or PTB+nPTB siRNA-treated cells were analyzed. CyDye labeling, 2D gels, and analysis were performed as described in. Gels were stained with Sypro Ruby, and proteins were excised with an automated spot picker. Proteins were digested with trypsin before analysis by LC-MS/MS sequencing (Cambridge Centre for Proteomics, Department of Biochemistry, University of Cambridge). MS/MS fragmentation data were used to search the NCBI primary sequence database using Mascot search engine.