Nuclear Degradation of Wilms Tumor 1-associating Protein and Survivin Splice Variant Switching Underlie IGF-1-mediated Survival*

WTAP (Wilms tumor 1-associating protein) is a recently identified nuclear protein that is essential for mouse embryo development. The Drosophila homolog of WTAP, Fl(2)d, regulates pre-mRNA splicing; however, the role of WTAP in mammalian cells is uncertain. To elucidate a context for WTAP action, we screened growth and survival factors for their effects on WTAP expression in vascular smooth muscle cells (SMCs), a cell type previously found to express WTAP dynamically. This revealed that insulin-like growth factor-1 (IGF-1) uniquely reduced WTAP abundance. This decline in WTAP proved to be necessary for IGF-1 to confer its antiapoptotic properties, which were blocked by transducing the WTAP gene into SMCs. WTAP down-regulation by IGF-1 was mediated by an IGF-1 receptor-phosphatidylinositol 3-kinase-Akt signaling axis that directed WTAP degradation via a nuclear 26 S proteasome. Moreover, by promoting the degradation of WTAP, IGF-1 shifted the pre-mRNA splicing program for the survival factor, survivin, to reduce expression of survivin-2B, which is proapoptotic, and increase expression of survivin, which is antiapoptotic. Knockdown of survivin-2B rescued the ability of IGF-1 to promote survival when WTAP was overexpressed. These data uncover a novel regulatory cascade for human SMC survival based on adjusting the nuclear abundance of WTAP to define the splice variant balance among survivin isoforms.

Smooth muscle cell (SMC) 3 survival is critical to vascular stability (1). This is especially important during atherosclerosis, a condition that depends on replicating SMCs to mechanically stabilize the artery wall. An abundance of proapoptotic stimuli within an atherosclerotic lesion can deplete the SMC population, resulting in plaque rupture and myocardial infarction (1,2). Pathways that enable SMCs to replicate and perform reparative functions and simultaneously resist apoptotic signals are therefore vital to cardiovascular health.
WTAP (Wilms tumor 1-associating protein) is a recently identified nuclear protein that, as its name implies, can interact with the WT1 suppressor protein (3,4). The precise molecular actions of WTAP are not well understood, but the importance of this protein is highlighted by the early lethality of WTAP-null mouse embryos (5,6). WTAP is widely expressed in adult tissues and might therefore play a constitutive role. However, expression of WTAP can also be dynamic. We found that WTAP expression in the adult artery wall varied, depending on whether the vessel was quiescent or remodeling. In particular, WTAP abundance decreased as vascular SMCs were induced to proliferate and increased as remodeling subsided and SMC accumulation terminated. Furthermore, overexpression of WTAP in cultured SMCs suppressed their accumulation, whereas WTAP knockdown stimulated population growth. These effects were mediated, at least in part, by the capacity for WTAP to activate apoptosis (4).
The basis by which WTAP impacts cell function is uncertain, and several possibilities have been raised. One group has reported that WTAP stabilizes cyclin A2 mRNA in human umbilical vein endothelial cells (5). Another level of action relates to the ability of WTAP to prevent WT1 from binding to its transcriptional target DNA (4). There is also evidence that WTAP may act as a splicing regulator (7). In this regard, it is noteworthy that WTAP is the mammalian homolog of Drosophila Fl(2)d (female-lethal-2-d), a protein that is necessary for regulating alternative splicing of several pre-mRNA species. These include transcripts encoded by Sxl (Sex-lethal) and tra (transformer), genes implicated in sexual determination and sexual behavior in Drosophila (8 -10), and also by Ubx (Ultrabithorax) (11), a gene unrelated to sexual determination.
Several lines of evidence suggest that, analogous to Fl(2)d, WTAP may regulate pre-mRNA splicing in mammalian cells. First, WTAP has been found to partly localize to interchromatin granule clusters (3,12), nuclear subregions in which pre-mRNA splicing factors are assembled, modified, or stored. As well, proteomic studies have identified WTAP as one of ϳ145 proteins in purified, functional human spliceosomes (13). Finally, immunodepletion of WTAP from HeLa cell extracts was found to alter the splicing of exogenous tra pre-mRNA (14). Notwithstanding these findings, a functional relationship between WTAP and alternative splicing of specific endogenous transcripts in mammalian cells has yet to be found.
To define a context for WTAP action in mammalian cells, we have capitalized on the finding that WTAP expression changes in vascular SMCs, depending on their growth properties. By screening growth and survival factors for their effects on WTAP expression in human SMCs, we found that WTAP was down-regulated by IGF-1, a potent survival factor for SMCs. This down-regulation of WTAP occurred via an Akt-driven nuclear degradation cascade and proved to be necessary for IGF-1 to confer its antiapoptotic properties on SMCs. Moreover, we found that clearance of WTAP from the nucleus reconfigured the pre-mRNA splicing program of survivin to favor expression of the antiapoptotic survivin and suppress expression of the proapoptotic variant, survivin-2B. The findings implicate WTAP as a gateway protein for a pre-mRNA splicing program that determines apoptosis versus survival. They also identify a novel basis by which IGF-1 enhances the survival of proliferating cells.
Transcript Analysis-Real time quantitative PCR was performed using TaqMan chemistry (Assays-on-Demand; Applied Biosystems, Forest City, CA). Glyceraldehyde-3-phosphate dehydrogenase transcript abundance was used to normalize RNA input. Relative quantification of gene expression was performed by the comparative CT method, using the arithmetic formula, 2 Ϫ⌬⌬CT , where CT is the threshold cycle at which a sufficient number of amplicons have been generated to be statistically significantly above background signal. Total RNA was isolated from SMCs (RNeasy, Qiagen), and 2 g of total RNA was reverse transcribed to cDNA (Applied Biosystems, Foster City, CA). The resultant cDNA was amplified by PCR (PerkinElmer Life Sciences Gene Amp system 9600). A 50-l mixture containing 1 l of cDNA, 2.5 units of Platinum TaqDNA polymerase (10966-018; Invitrogen), 1ϫ PCR buffer (Invitrogen), 25 M dNTPs, 1.5 mM MgCl 2 , and 25 pmol of both 5Ј-and 3Ј-primers were used. The primer pair used to amplify the entire coding region of survivin was as follows: forward primer, 5Ј-GCA TGG GTG CCC CGA CGT TG-3Ј; reverse primer, 5Ј-GCT CCG GCC AGA GGC CTC AA-3Ј (18). ␤-Actin transcript was amplified using the forward primer 5Ј-GCA-CCACACCTTCTACAATGAGC-3Ј and reverse primer 5Ј-TAGCACAGCCTGGATAGCAACG-3Ј.
Assessment of Apoptosis-SMC apoptosis was quantified by flow cytometry of Annexin V-stained cells. SMCs were suspended in 100 l of Annexin V binding buffer to which 5 l of R-phycoerythrin-Annexin V conjugate (Molecular Probes) and 7-aminoactinomycin D (1 g/ml; Sigma) were added. Cells were incubated with fluorescent labels for 15 min in the dark with the subsequent addition of 400 l of Annexin V binding buffer (10 mM HEPES, 140 mM NaCl, and 2.5 mM CaCl 2 , pH 7.4). Apoptosis and necrosis were analyzed by flow cytometry (FACScalibur) using Cell Quest software (BD Biosciences).
Preparation of Nuclear and Cytoplasmic SMC Fractions-SMCs lysates were separated into nuclear and cytoplasmic fractions (catalogue number 78833; Pierce) according to the manufacturer's instructions and then subjected to Western blot analysis for WTAP. Blots were probed with anti-human lamin A/C rabbit polyclonal antibody (sc-20681; 1:500; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and anti-human ␣-tubulin monoclonal antibody (1:25,000; Sigma) to mark the nuclear and cytoplasmic components, respectively.
Protein Degradation Assays-SMCs at 75% confluence were incubated for 1 h in methionine-free Dulbecco's modified Eagle's medium (Invitrogen) and then pulsed for 1 h with 100 Ci/ml [ 35 S]methionine ( 35 S-Promix; Redivue, AGQ0080; GE Healthcare). Cells were then washed five times with Dulbecco's phosphate-buffered saline and "chased" with Dulbecco's modified Eagle's medium containing 4 mM methionine (Invitrogen) in the presence or absence of IGF-I (50 ng/ml). At designated times, the cells were washed with phosphate-buffered saline and solubilized in radioimmune precipitation assay buffer (1% Igepal CA-630 (Sigma), 0.5% sodium deoxycholate, 0.1% SDS in phosphate-buffered saline) containing 0.1 mM phenylmethylsulfonyl fluoride and 10 g/ml leupeptin. Protein extracts were clarified by centrifugation and precleared with 5 l of protein G PLUS-agarose beads (Santa Cruz Biotechnology) for 30 min. Equal amounts of protein were immunoprecipitated using anti-WTAP antibody (10 g/ml) and 25 l of protein G PLUS-agarose beads overnight at 4°C. Immunoprecipitates were washed extensively with radioimmune precipitation assay buffer and resolved on a 12% polyacrylamide gel. Quantification of radioactive bands in dried gels was performed using a PhosphorImager screen and ImageQuant software (Amersham Biosciences).
Protein stability was also assessed in control and IGF1-stimulated SMCs in the presence of cycloheximide (10 g/ml), added 30 min prior to growth factor. WTAP abundance was tracked at designated times by Western blot analysis.
Overexpression of WTAP in Human SMCs-A retroviral gene delivery system was used to generate human cells stably overexpressing WTAP, using methods described previously (4,17). Briefly, full-length cDNA encoding WTAP was amplified from HITB5 SMC mRNA by RT-PCR and subcloned into pQCXIP-IRES-PURO (Clontech), as described previously (4). Retrovirus containing WTAP cDNA was produced using the Phoenix-Amphotropic retrovirus packaging cell line (kindly provided by Dr. G. Nolan (Stanford University Medical School), distributed by ATCC (Manassas, VA)). Stable transductants were selected with 2.8 g/ml puromycin (BD Biosciences Clontech) for 48 h, and overexpression of WTAP was confirmed by Western blot analysis.
Transcript Knockdown by RNA Interference-Knockdown of WTAP and Survivin-2B expression was accomplished by infecting human SMCs with retrovirus-containing sequences encoding short hairpin RNA (shRNA) fragments (4,23). An oligonucleotide containing a 19-base sense sequence starting at nucleotide 1039 of human WTAP (5Ј-CTCTCTCACACAC-CAATCA-3Ј) and the reverse complement targeting sequence with an intervening hairpin loop sequence was generated. Oligodeoxynucleotides were synthesized, annealed, and inserted between BamHI and EcoRI sites of the retroviral expression vector pSIREN-RetroQ (BD Biosciences Clontech). A control insert (nonsilencing RNA (nsRNA)) contained the human WTAP-specific 19-nucleotide sense sequence but not the antisense sequence. Survivin-2B-specific shRNA consisted of a 19-nucleotide sense (starting at nucleotide 245 from the open reading frame) and reverse complement survivin-targeting sequences (24) with an intervening hairpin loop (5Ј-GAAGCTTG-3Ј) (25). Control nsRNA contained a scrambled sense sequence of survivin-2B 19-based nucleotides and corresponding antisense sequence (24) with the same hairpin loop.
Small interfering RNA (siRNA) duplexes purchased from Ambion were used to knock down expression of Akt (ID s659) and PI3K (ID s10536) with Silencer Negative Control 1 siRNA used as a control. To specifically knock down survivin, an siRNA duplex oligonucleotide targeting the exon 2-exon 3 junction was designed and synthesized using Silencer Select chemical modification (Ambion). The siRNAs (1.5 g) were electroporated (Amaxa Nucleofector) into SMCs (2 ϫ 10 5 cells/ cuvette) using siPORT electroporation buffer (Ambion).
Statistics-Values are reported as means Ϯ S.E. of the mean. Statistical significance was determined by t test or analysis of variance with Bonferroni's post hoc test, and -fold changes were assessed by the Wilcoxon signed rank test. Statistical significance was set at p Ͻ 0.05.

WTAP Expression Is Suppressed by IGF-1-Although
WTAP is a widely expressed protein, we previously found that its abundance varied in vascular SMCs, depending on their growth and survival status (4). To identify molecular controls over WTAP expression and elucidate a context for its modulation, human SMCs were rendered quiescent in serum-free medium and then stimulated with various growth or survival factors relevant to the vasculature. This established that WTAP protein abundance was unaffected by the proliferative factors platelet-derived growth factor-BB, fibroblast growth factor 2, epidermal growth factor, or amphiregulin within 12 h of stimulation. Likewise, WTAP expression was unaffected by the antiproliferative factor, transforming growth factor-␤1. TNF␣ and vascular endothelial growth factor also had no effect (not shown). In contrast, WTAP was substantially down-regulated by IGF-1 and, to a lesser extent, by insulin (Fig. 1A). This reduction in WTAP abundance by IGF-1 was concentration-dependent and observed at concentrations as low as 25 ng/ml (Fig. 1B), well within the physiologic range of this growth factor (26,27).
WTAP expression was also observed in human aortic endothelial cells and in human fibroblasts, and in both cell types, expression was inhibited by IGF-1 (Fig. 1B). Interestingly, the effect of IGF-1 on WTAP expression in SMCs was seen at the protein level but not at the mRNA level, the latter assessed using quantitative real time PCR (Fig. 1C). Thus, WTAP protein expression in SMCs is regulatable by the external milieu, although selectively, with a brisk and potent repression by IGF-1.
WTAP Down-regulation Is Necessary for IGF-1-mediated SMC Survival-Although IGF-1 is a multifunctional growth factor, it is unique among the growth factors screened, because its primary action on SMCs is to suppress apoptosis (28,29). This is noteworthy, given previous findings that WTAP knockdown inhibited apoptosis in proliferating SMCs and overexpression of WTAP promoted apoptosis (4). Therefore, we next determined if the downregulation of WTAP by IGF-1 was necessary for IGF-1 to confer its antiapoptotic actions. WTAP cDNA was introduced into human SMCs by retrovirus, which augmented WTAP protein levels in IGF-1-stimulated SMCs such that, following IGF-1 stimulation, they remained at or above that of unstimulated SMCs ( Fig. 2A). Apoptosis was then quantified by assessing surface expression of phosphatidylserine using fluorescence-labeled Annexin V. As shown in Fig. 2B, stimulation of vector-infected SMCs with IGF-1 suppressed apoptosis by 38% (p Ͻ 0.05). However, IGF-1 failed to inhibit apoptosis in SMC-WTAP cells. Assessment of caspase-3 activity in SMCs, using an exogenous fluorogenic peptide substrate, revealed that IGF-1 inhibited caspase-3 activity in SMCs by 57% (p Ͻ 0.05), but this response was abrogated when WTAP suppression was forcibly prevented (Fig. 2C). These findings identify WTAP as an activator of mitochondria-dependent apoptosis in SMCs and implicate WTAP down-regulation as an essential step in IGF-1-mediated SMC survival.

IGF-1-induced WTAP Suppression Is Mediated by an IGF-1 Receptor/PI3K/Akt Signaling Cascade-
We next determined the signaling events responsible for WTAP down-regulation by IGF-1. To determine if the response was mediated by IGF-1R, we incubated SMCs with either IGF-1R inhibitor II, which inhibits ATP binding to IGF-1R (30), or picropodophyllin, which blocks phosphorylation of Tyr 1136 in the IGF-1R activation loop (31). As shown in Fig. 3A, SMCs pretreated with either of these mechanistically distinct inhibitors were resistant to IGF-1-mediated down-regulation of WTAP expression. IGF-1R transmits signals along two major canonical pathways, the mitogen-activated protein kinase and PI3K signaling pathways (32). Interestingly, incubation of SMCs with the ERK1/2 inhibitor, U0126, inhibited IGF-1-mediated phosphorylation of ERK1/2 but had no effect on IGF-1-mediated suppression of WTAP expression (Fig. 3B). In contrast, inhibition of PI3K activity with LY294002 markedly attenuated WTAP down-regulation by IGF-1 (p Ͻ 0.001). A similar response was observed upon inhibiting PI3K with wortmannin (data not shown). Consistent with these findings, a 31% increase in SMC survival afforded by IGF-1 (p Ͻ 0.01) was completely abrogated by LY294002, whereas the antiapoptotic effect of IGF-1 was still evident in the presence of UO126 (20% increase over base line; p Ͻ 0.01). The importance of PI3K was further substantiated by siRNA-mediated knockdown of this enzyme, which prevented the down-regulation of WTAP by IGF-1 (Fig. 3C).
We also evaluated the participation of the PI3K target, Akt, in IGF-1-mediated inhibition of WTAP expression. As shown in Fig. 3D, SMCs incubated with increasing concentrations of Akt inhibitor IV were progressively more resistant to IGF-1-induced WTAP down-regulation. This finding was consistent with the inhibitory effect we observed with picropodophyllin, which blocks phosphorylation of a tyrosine residue (Tyr 1136 ) of IGF-1R required to stimulate the phosphorylation of Akt and activate the PI3K/Akt apoptotic pathway (33). Furthermore, siRNA-mediated knockdown of either PI3K or Akt abrogated the IGF-1-mediated inhibition of WTAP expression (Fig. 3E). Collectively, these data establish that down-regulation of WTAP is a novel consequence of activating the IGF-1R/PI3K/ Akt survival cascade in SMCs.
IGF-1 Induces WTAP Protein Degradation-The relatively rapid decline in WTAP abundance following incubation with IGF-1 suggested that IGF-1 might be increasing WTAP protein degradation kinetics. This possibility was further suggested by the observation that the IGF1-mediated decline in WTAP was associated with stable WTAP mRNA levels (Fig. 1B). We therefore evaluated the effect of IGF-1 on WTAP stability by pulsechase analysis. SMCs were pulsed for 1 h with [ 35 S]methionine, and the amount of labeled WTAP was determined following WTAP immunoprecipitation. In vehicle-treated SMCs, the pool of recently synthesized WTAP was relatively stable and declined by only 23% 8 h after delivering the pulse. In contrast, the amount of radiolabeled WTAP in IGF-1-stimulated SMCs had declined by 86% at this time (p Ͻ 0.01; Fig. 4A).
We also assessed WTAP stability by tracking its abundance after incubating cells with the protein synthesis inhibitor, cycloheximide. This confirmed that WTAP is a stable protein under control, unstimulated circumstances (Fig. 4C). However in the presence of IGF-1, there was an exponential decline in WTAP content, falling to 96.9 Ϯ 1.0% of its resting level after 8 h (p Ͻ 0.001; Fig. 4C). Collectively, these data establish that stimulating SMCs with IGF-1 substantially increases the rate of WTAP protein degradation.
IGF-1-induced WTAP Degradation Is via the 26 S Nuclear Proteasome-Having established that tracking WTAP content following incubation with cycloheximide mirrors the decay kinetics revealed by pulse-chase analysis, we used the former approach to delineate the mode of WTAP degradation by IGF-1. To determine if WTAP was degraded through proteasomal activity, SMCs were incubated in the presence or absence the proteasome inhibitors, MG132 or lactacystin, together with cycloheximide, all added 30 min prior to stimulating cells with vehicle or IGF-1. This revealed that the degradation of WTAP by IGF-1 was inhibited when 26 S proteasome activity was blocked. Under these circumstances, the WTAP degradation rate was similar to that of WTAP in SMCs not subjected to IGF-1 (Fig. 4C).
We next determined if WTAP was degraded in cytoplasmic or nuclear compartments. To do this, cycloheximide-exposed SMCs were incubated in the presence or absence of the nuclear export inhibitor, leptomycin B, and then stimulated with IGF-1. As shown in Fig. 5A, WTAP could not be detected in cytoplasmic fractions either at base line or following IGF-1 stimulation. In contrast, WTAP was abundant in the nuclear fraction and  disappeared from this fraction following IGF-1 stimulation. Inhibition of nuclear export did not prevent the loss of nuclear WTAP following IGF1 stimulation. Leptomycin B did, however, increase the abundance of nuclear p53, a nuclear protein known to be exported to the cytoplasm (Fig.  5B) (34). We also followed WTAP abundance in the cytoplasmic and nuclear fractions of SMCs incubated with cycloheximide and the proteasome inhibitor, MG132. As shown in Fig. 5C, MG132 fully stabilized nuclear WTAP content after IGF-1 stimulation with, again, no evidence for WTAP expression in the cytoplasm. Finally, we observed that in the presence of MG132, WTAP immunoprecipitates from the nucleus of IGF-1-stimulated SMCs but not from unstimulated SMCs, were associated with ubiquitin conjugates (Fig.  5D). Collectively, these data establish that WTAP is exclusively a nuclear protein in SMCs that is ubiquitinated and degraded in a nuclear 26 S proteasome, upon IGF-1 stimulation. Lysates were immunoprecipitated with an antibody against WTAP, separated by SDS-PAGE, and signal-detected by phosphorimaging. Relative WTAP signal intensity is denoted, and data are representative of three experiments. B, Western blots indicating serial abundance of WTAP in SMCs subjected to protein synthesis inhibition with cycloheximide (10 g/ml), added 30 min before stimulating cells with IGF-1 (50 ng/ml) or vehicle. WTAP pool size decay kinetics were fit with monoexponential curves, averaged from three experiments, and statistical comparison between the decay rates was made after natural logarithm transformation (inset). Open triangle, vehicle-stimulated SMCs; closed triangle, IGF-1-stimulated SMCs (p Ͻ 0.0001). C, the IGF-1-mediated decline in WTAP pool size was similarly evaluated in

WTAP Modulates the Balance of Survivin Splice Variants in
SMCs-In view of the importance of WTAP clearance for IGF-1-mediated SMC survival, and having delineated steps in the clearance pathway, we next sought to determine how WTAP clearance enhanced SMC survival. We were particularly interested in determining if WTAP impacted the splicing profile of apoptosis-associated transcripts, recognizing that WTAP has been isolated from the human spliceosome and that the Drosophila homolog of human WTAP (Fl(2)d) affects alternative splicing in that organism. We thus sought out candidate mediators that are known to 1) regulate apoptosis in vascular SMCs, 2) have their expression impacted by IGF-1, and 3) have their expression regulated, at least in part, by alternative splicing mechanisms. One candidate molecule meeting each of these criteria was survivin (also known as BIRC5), a member of the inhibitor of apoptosis protein family. Survivin is abundant in tumors but has also been found to regulate the survival of vascular SMCs (35,36). Using RT-PCR and a primer pair encompassing the entire coding region of survivin (18), we found that human SMCs express the index survivin transcript as well as survivin-2B, an alternatively spliced variant that retains 69 bases from intron 2 as a cryptic exon (Fig. 6A, lanes 1 and 3) (18). Both survivin and survivin-2B were also evident at the protein level (Fig. 6B, lanes 1 and 3).
To determine if WTAP impacted expression of survivin or its variants, we increased WTAP gene dosage in human SMCs by infecting them with retrovirus containing full-length WTAP cDNA, increasing WTAP abundance by 2.4 Ϯ 0.4-fold (Fig. 6B). No additional survivin splice variants were observed following WTAP overexpression; however, there was a striking change in the balance of the existing splice forms. In particular, there was with a substantial increase in survivin-2B mRNA, which has been reported to be a proapoptotic isoform (24), and a reciprocal decrease in abundance of survivin, which is antiapoptotic (Fig. 6A, lane  2). Quantitative assessment revealed that ratio of survivin-2B to survivin was 1.6 Ϯ 0.2 in vector-infected SMCs but 7.7 Ϯ 0.5 in WTAP-overexpressing SMC (p Ͻ 0.01). A strong bias toward survivin-2B following IGF-1 stimulation was also observed at the protein level (Fig. 6B, lane 2).
We next evaluated the effect of reducing WTAP gene dosage on the relative abundance of survivin splice variants. For this, SMCs were infected with retrovirus containing cDNA encoding shRNA specific for WTAP, decreasing the abundance of WTAP to 46 Ϯ 6% of control SMCs (Fig. 6B). As shown (Fig. 6A, lane 4), SMCs expressing WTAP shRNA displayed increased survivin transcript expression, and survivin-2B mRNA was no longer detectable. This increase in survivin and loss of survivin-2B expression also manifest at the protein level (Fig. 6B, lane 4).
To determine if the observed shift in survivin splice forms was part of a generalized splicing shift induced by WTAP, we examined the transcript profiles of Apaf1, Bcl-x, and Mcl, three additional apoptosis regulators, each with isoforms arising from alternative pre-mRNA splicing. Long and short transcripts for each gene were found to be expressed by human SMCs. Interestingly, however, neither overexpression nor knockdown of WTAP changed the relative abundance of the respective splice forms (Fig. 6C).
IGF-1 Shifts the Balance of Survivin Splice Variants via WTAP Clearance-The dependence of survivin isoform expression on WTAP levels raised the possibility that IGF-1 might also stimulate a switch in survivin isoform expression. As shown in Fig. 6D, stimulating SMCs with IGF-1 for 12 h increased survivin transcript expression and simultaneously abolished survivin-2B mRNA expression (lanes 2 and 4). These changes were concordant with those observed with WTAP knockdown. Furthermore, this effect of IGF-1 on survivin isoform abundance was abrogated in SMCs with forced expression of WTAP (lane 6). Abrogation of survivin-2B expression was observed following IGF-1 stimulation, with WTAP knockdown, and with the combination of WTAP knockdown and IGF-1 stimulation (lane 10). These findings thus establish that IGF-1 orchestrates a shift in the survivin pre-mRNA splicing program to favor elaboration of antiapoptotic survivin over proapoptotic survivin-2B. Because this effect was mimicked by WTAP knockdown and abrogated by WTAP overexpression, an essential role for WTAP clearance in mediating this splicing response to IGF-1 can be concluded. Serum-starved SMCs were incubated with cycloheximide (10 g/ml) and the nuclear export inhibitor, leptomycin B (LMB; 10 nM) or vehicle, each added 30 min prior to stimulation with IGF-1 (50 ng/ml). Cytoplasmic and nuclear extracts were prepared at the designated times, and blots were probed for WTAP, lamin A/C, and ␣-tubulin. B, Western blots verifying the inhibitory effect of 10 nM LMB on nuclear export, indicated by the accumulation of nuclear p53. C, Western blots depicting the effect of inhibiting 26 S proteasome activity on WTAP decay in nuclear and cytoplasmic fractions. SMCs were studied as in A with 50 M MG132 added 30 min before IGF-1. D, Western blots showing ubiquitination of WTAP (Ub-WTAP) following stimulation of SMCs for 4 h with IGF-1 (50 ng/ml) in the presence of MG132 (50 M). Nuclear lysates were incubated overnight at 4°C with anti-WTAP antibody and protein A/G PLUS-agarose beads. The precipitated material was resolved on an SDS-polyacrylamide gel, transferred to polyvinylidene difluoride membrane, and immunoblotted using a monoclonal anti-ubiquitin antibody. IP, immunoprecipitation; IB, immunoblot.

Knockdown of Survivin-2B, but Not Survivin, Rescues the Ability of IGF-1 to Suppress Apoptosis in WTAP-overexpressing
SMCs-To determine if the WTAP-dependent modulation of survivin splice variants by IGF-1 was functionally linked to apo-ptosis control, we asked whether knockdown of survivin-2B could rescue the ability of IGF-1 to inhibit apoptosis when WTAP suppression was forcibly prevented. SMCs were thus double-infected with retroviruses containing WTAP cDNA or vector control and also containing cDNA encoding survivin-2B-specific shRNA or scrambled, nonsilencing control RNA. As shown (Fig. 6E), survivin-2B protein was undetectable in SMCs expressing basal levels of WTAP and survivin-2B shRNA. In SMCs overexpressing WTAP, survivin-2B protein content was again modestly increased but not detectable after survivin-2B knockdown.
IGF-1 failed to suppress apoptosis in WTAP-overexpressing SMCs expressing control, nsRNA survivin-2B. When survivin-2B expression was suppressed, however, the ability of IGF-1 to inhibit apoptosis was restored. In survivin-2B knockdown SMCs not stimulated with IGF-1, apoptosis fell by 21.2% (Fig. 6E, bar 1 versus bar 3, p Ͻ 0.01), verifying that this survivin isoform is proapoptotic in human SMCs. Upon stimulation with IGF-1, the proportion of apoptotic SMCs declined further, by 41.4% of that of IGF1-1 stimulated WTAP-SMCs not subjected to survivin-2B knockdown (bar 2 versus bar 4) and by 29.2% of that of survivin-2B knockdown SMCs not stimulated with IGF-1 (bar 3 versus bar 4). The latter increment in survival is consistent with the increased expression of survivin by IGF-1, superimposed on suppressed survivin-2B through knockdown.
This ability of survivin-2B knockdown to restore the survival action of IGF-1 in the setting of WTAP overexpression was mirrored by changes in caspase-3 activity. As shown in Fig. 6E, caspase activity fell in survivin-2B knockdown SMCs, establishing that survivin-2B activates caspase-3 in human SMCs. Moreover, the ability of IGF-1 to inhibit caspase-3 activity was restored in WTAP-SMCs when survivin-2B expression was inhibited, in concert with the further increase in survivin. Thus, the inability of IGF-1 to inhibit caspase-3 activity and suppress apoptosis when WTAP is overexpressed can be attributed to its inability to alter the absolute and relative abundance of survivin-2B and survivin.
We next evaluated the effect of selective knockdown of survivin, using siRNA that targeted the exon 2-exon 3 junction present in the survivin transcript but disrupted in the survivin-2B transcript. Selective knockdown of survivin (Western blot; Fig. 6F) in WTAP-overexpressing SMCs led to increased apoptosis (Fig. 6F, graph, lane 3 versus lane 1), consistent with the unopposed proapoptotic action of survivin-2B. Upon stimulation with IGF-1, apoptosis did not decline and in fact increased somewhat. Similar trends were observed for caspase activation (Fig. 6F).
Collectively, these findings indicate that the survival actions of IGF-1 are mediated by a splice isoform expression program that favors survivin over survivin-2B, a program that depends on WTAP clearance (Fig. 7).

DISCUSSION
We have discovered a novel cascade for regulating apoptosis in proliferating vascular SMCs. Central to this cascade is the abundance of WTAP in the nucleus of human SMCs, which we found governs a switch between SMC survival and apoptosis. When WTAP is abundant, the pre-mRNA splicing program for survivin is biased to favor expression of the proapoptotic variant, survivin-2B. When the abundance of WTAP in SMCs is low, the common survivin splice form is increased, survivin-2B is undetectable, and the antiapoptotic actions of survivin proceed. We further established that the potent survival factor, IGF-1, regulates SMC survival via this WTAP-dependent pathway. Specifically, IGF-1 drives a cascade to clear WTAP from the nucleus of human SMCs by an Akt-dependant and 26 S  proteasome-mediated degradation process. The data thus identify a novel paradigm for controlling cell survival, based on tuning the abundance of WTAP to alter splice variant output.
The actions of WTAP in mammalian cells are only beginning to be understood, and they may be diverse. The current findings that IGF-1 suppresses WTAP expression and that this suppression is essential for IGF-1-mediated SMC survival strongly support a physiologically relevant role for WTAP in regulating apoptosis. The importance of lowering WTAP expression for the antiapoptotic actions of IGF-1 can also be inferred by the rapidity of the response. IGF-1 had no effect on WTAP mRNA abundance but instead briskly depleted the pool of existing WTAP. The latter finding was confirmed using two independent approaches, inhibition of protein synthesis with cycloheximide and tracking recently synthesized WTAP using pulsechase analysis. It is also noteworthy that WTAP was ubiquitinated in the nucleus, and its degradation was dependent on 26 S proteasomal activity that was localized to the nucleus. These findings were established by cell fractionation studies and by the unaltered WTAP decay kinetics when nuclear export was inhibited by leptomycin B. The steps by which WTAP is ubiquitinated are unknown, but Akt-activated ubiquitin ligases have been identified and might therefore be candidates (37). It is also noteworthy that several components of the mRNA splicing machinery, including the splicing factor SC35, the U1-70k small nuclear ribonucleoprotein, and the Sm-B/BЈ protein, have also been found to be degraded by a nuclear proteasome (38). It is reasonable to speculate therefore that nuclear proteolysis of WTAP is part of a broader program by which splicing-associated proteins are cleared, enabling the cell to acutely reconfigure its splicing machinery to meet ongoing demands.
The positioning of WTAP within an IGF-1 signaling cascade is novel. At the same time, the molecular components that transduce the IGF-1 stimulus to suppress WTAP expression (IGF-1R, PI3K, and Akt) have emerged as fundamental players in regulating SMC survival by IGF-1. The IGF-1R has recently been found to be rate-limiting for survival of SMCs in response to oxidative stress (39) and is down-regulated in the apoptosisprone SMCs of atherosclerotic lesions (28,29). An essential role for PI3K/Akt activation in IGF-1-mediated SMC survival has also recently been established (39 -41). In the current study, the need for Akt signaling on WTAP down-regulation was confirmed using both a direct Akt inhibitor, siRNA-mediated Akt knockdown, and an IGF-1R-interfering reagent that prevents the activation of a residue required for Akt phosphorylation and activation (33). Whether WTAP is a direct target of Akt is unknown, but it is an interesting possibility, given that Akt has been found to stimulate the proteasomal degradation of a number of its substrates, including activators of apoptosis, such as p53 (42), FOXO1, and FOXO3a (43). Regardless, the Akt-mediated clearance of WTAP identified in the current study is consistent with an emerging paradigm whereby Akt activation promotes cell survival by driving the degradation of negative regulators of the process.
Our data also establish survivin transcript switching as a novel effector arm for IGF-1-mediated survival. Survivin is a member of the inhibitor of apoptosis family that impacts sur-vival and the cell cycle (44). Although classically studied in tumors, survivin has more recently been found to be a survival factor for proliferating vascular SMCs (35,36). Of the known splice variants of survivin, survivin-2B is noteworthy because, in contrast to other isoforms, it acts as a proapoptotic molecule, and its forced expression in HeLa cells has been found to activate, rather than suppress, caspase-9 and caspase-3 (24). Our findings establish that IGF-1 "resolves" this inherent antagonism among survivin splice variants by terminating the expression of survivin-2B and enhancing survivin expression. Several lines of evidence indicate that this biasing of survivin splice variants by IGF-1 is mediated by WTAP clearance. First, WTAP knockdown mimicked the effect of IGF-1, with respect to both survivin splice variant switching and decreased apoptosis. Second, forced expression of WTAP increased survivin-2B expression and decreased survivin expression and, in so doing, rendered IGF-1 incapable of pushing the splicing program away from survivin-2B. Third, selectively knocking down survivin expression in WTAP-overexpressing SMCs not only maintained the ability of high WTAP content to prevent IGF-1-mediated survival but exacerbated the response. Finally, knocking down survivin-2B in WTAP-overexpressing SMCs decreased caspase-3 activity and restored the ability of IGF-1 to suppress apoptosis, despite high WTAP.
The current data are thus the first to link WTAP to pre-mRNA splicing of an endogenous target. Previous reports have implicated fl(2)d in regulating alternative splicing of Sxl, tra, and Ubx pre-mRNA (8 -11); interestingly, each of these targets is alternatively spliced by a different mechanism. As well, Penn et al. (45) have found that Fl(2)d participates in an early stage of spliceosome assembly. Based on these findings and the current data, we propose that WTAP resides in a molecular milieu that is proximal to the pre-mRNA-spliceosome interface. Accordingly, WTAP may not directly execute splice site decisions. At the same time, our data indicate that WTAP is not simply a housekeeping constituent of the spliceosome but plays a determining role in splicing, presumably through interactions with other splicing factors. Importantly, this role is regulatable by the environment. Because there are orthologs of both WTAP and survivin in Drosophila (fl(2)d and deterin, respectively) (46), further genetic and biochemical studies in this species could help define the steps by which WTAP modulates transcript processing.
Whether WTAP plays a dominant proapoptotic role under all circumstances is uncertain. Indeed, the expression of WTAP in cells of most adult tissues might argue against this (3), although this would need to be examined in the context of the net balance of pro-and antiapoptotic signals. In addition, although WTAP knockdown enhances the survival of proliferating SMCs, we previously found that WTAP was up-regulated in SMCs that had differentiated in culture to a quiescent, contractile state with a low apoptosis rate (4). We speculate therefore that the proapoptotic action of WTAP may be context-dependent and particularly relevant to cells that are proliferating. If, however, the proliferating cells are required for tissue repair and already in an environment with abundant death signals, then it may be critical to abrogate the proapoptotic role of WTAP. Indeed, we previously observed that WTAP expression was strikingly suppressed in SMCs following injury to the rat carotid artery, during which time SMCs are actively repairing the disrupted artery wall (4). Clearance of WTAP may thus be an important step in ensuring that cell replication is coupled to cell survival in inflamed or otherwise vulnerable environments.
Roles for WTAP other than apoptosis are also likely. The phenotype of WTAP-null mouse embryos, for example, has indicated a role for WTAP in endoderm and mesoderm differentiation (6). Although the mechanism for this effect of WTAP is unknown, the current findings might be relevant to this question. This is because, in addition to inhibiting apoptosis, survivin has been implicated in regulating microtubule dynamics (47). Microtubule dynamics, in turn, are vital for developmental processes, including epithelial-mesenchymal transformation (48).
In summary, we have uncovered a novel regulatory cascade for human SMC survival based on tuning the nuclear abundance of WTAP and shifting the splice variant output of survivin. Suppression of WTAP expression by environmental cues, such as IGF-1, may underlie an important means of protecting cells that are actively participating in tissue repair and regeneration.