Tra2β As a Novel Mediator of Vascular Smooth Muscle Diversification
Transformer splicing regulatory proteins determine the sexually dimorphic traits of Drosophila. The role of the vertebrate homologs of Tra-2 in phenotypic specification is undefined. We are using the alternative splicing of the MYPT1 E23 exon as a model for the study of smooth muscle diversification into fast and slow contractile phenotypes. Tra2β mRNA and protein is expressed at up to 10-fold higher levels in fast smooth muscle tissues such as the rat portal vein and small mesenteric artery, in which E23 is spliced, as compared to the slow smooth muscle tissues of the large arteries and veins, in which E23 is skipped. Tra2β is upregulated up to 10-fold concordant with the initiation of E23 splicing as the rat portal vein and avian gizzard implement the fast program of gene expression in the perinatal period. In disease models such as portal hypertension and mesenteric artery high/low flow, the portal vein and first order mesenteric artery dynamically downregulate Tra2β concordant with a shift to E23 skipping and the slow program of gene expression. Tra2β binds to a highly conserved sequence within E23 and transactivates its splicing in vitro and in vivo; this is abolished with mutation or deletion of this sequence. RNA interference–mediated knockdown of Tra2β markedly reduces E23 splicing. We propose that Tra2β has been conserved through evolution and redeployed for the specification of the fast smooth muscle phenotype and may serve as a novel nodal point for the investigation of this process in developmental and disease models.
Smooth muscle exhibits considerable phenotypic diversity that, as is true for striated muscle, may be dichotomized into fast or slow subtypes based on the rates of contraction and relaxation. Vascular smooth muscle tends to be of the slow phenotype, whereas visceral smooth muscle tends to be of the fast phenotype, although muscles of fast, slow, and intermediate phenotypes are present in both vascular and visceral systems. These phenotypes are specified during development, responsive to neural and humoral inputs and mechanical load, and critical to the function of these systems in normal and disease states.1,2
Much diversity in smooth muscle is generated by the alternative splicing of exons. Myosin heavy and light chains, myosin phosphatase, and tropomyosin, caldesmon, and calponin isoforms are generated by alternative splicing of exons and are thought to contribute to the functional differences between fast (also described as phasic) versus slow (also referred to as tonic) smooth muscle.3 We have selected the myosin phosphatase targeting subunit 1 (MYPT1) as a model gene for the study of smooth muscle phenotypic diversity. (1) The cassette-type splicing of a 31-nt 3′ alternative exon (MYPT1 E23) (Figure 1A) is highly tissue-specific. The alternative exon is skipped in the slow smooth muscle of the large arteries and veins and included in the fast smooth muscle of the portal vein (PV) and the small mesenteric arteries (MAs), which have an intermediate phenotype. The splicing of MYPT1 E23 is (2) evolutionarily conserved in mammals and birds and (3) tightly regulated during development. The PV and avian gizzard, prototypical fast tissues, acquire phasic properties in the perinatal period.4,5 At this time, these tissues undergo a complete transition from MYPT1 E23 skipping to E23 inclusion as part of the reprogramming of gene expression to the fast pattern.6,7 In contrast the smooth muscle of the large arteries and veins is of the slow phenotype with E23 skipping throughout development, as are smooth muscle cells (SMCs) in culture,6,8 suggesting that this is a default phenotype. (4) The splicing of MYPT1 E23 modulates in disease models. In models of portal hypertension (PHT) and altered flow, the PV and mesenteric resistance artery switch to E23 skipping9,10 as part of a generalized reversion toward the slow phenotype.
The goal of the present study was to identify the control mechanisms for the tissue-specific splicing of E23 in relation to smooth muscle phenotypic diversity. Depending on the alternative exon and cell type, some have proposed combinatorial control by widely expressed splicing regulatory factors to result in tissue-specific splicing. In a few other instances, a highly tissue-restricted factor has a dominant effect on the splicing of alternative exons in those tissues.11 The most striking example of the latter is sex determination in the fly. This developmental choice is under the control of the Transformer proteins, which through the regulated splicing of Dsx and Fru alternative exons specify the unique physical and behavioral characteristics of female flies.12 The vertebrate homologs of the fly Tra-2, Tra-2α and Tra2β, were subsequently identified and, in the case of Tra2α, shown to be functionally equivalent to the fly Tra2.13–16 Tra2α and -β are atypical members of the SR family of RNA binding proteins and have been suggested to regulate the splicing of a number of vertebrate alternative exons in conjunction with the classic SR proteins.16 However, in contrast to its clear and potent role in phenotypic specification in flies, a role for Tra in phenotypic specification in higher organisms has not been defined.
Materials and Methods
Oligonucleotides used for RT-PCR are listed in the expanded Materials and Methods section in the online data supplement at http://circres.ahajournals.org.
Sequence Analysis of MYPT1 E23 and Flanking Introns
We carried out phylogenetic analysis of the MYPT1 E23 and flanking intronic sequence (rat MYPT1: ENSRNOG00000004925) using the University of California at Santa Cruz genome browser (http://genome.ucsc.edu). The candidate cis elements of E23 splicing were identified either by manually screening the sequence for known binding motifs or by using databases, including Splicing Rainbow (http://www.ebi.ac.uk/asd-srv/wb.cgi?method=8), RESCUE-ESE (http://genes.mit.edu/burgelab/rescue-ese), and ESEfinder (http://rulai.cshl.edu/tools/ESE2). Only those motifs above the threshold values were considered as putative cis elements.
Tissue Samples, Cell Lines, and Cell Culture
Tissue samples were obtained from: (1) normal adult male Sprague–Dawley rats (Charles River Laboratories, Wilmington, Mass) and the offspring of paired matings; (2) rats in which the PV was surgically stenosed to induce PHT9; (3) rats in which alternating second order MAs were ligated to induce high-flow and low-flow states in alternating upstream first order MAs10; (4) white leghorn chickens and embryos before hatching (Squire Valley Farm, Cleveland, Ohio). Animal care and use procedures were approved by the Institutional Animal Care and Use Committees at Case Western Reserve University. The rat samples were derived from previously published studies. SMCs were isolated from chicken gizzard and rat aorta (Ao) and maintained in culture as described.17 The human embryonic kidney (HEK)293 (Quantum Biotechnology Inc, Quebec, Canada) and A7r5 (embryonic rat aortic smooth muscle) cell lines were grown in monolayer in DMEM supplemented with 10% HFBS.
Analysis of Tra2β mRNA Levels and Splice Isoforms
Real-time PCR was performed to measure the total abundance of Tra2β. Values were normalized to SRp20 mRNA, which was invariant between samples. Splice variants of Tra2β in rat tissues were analyzed by standard RT-PCR.
Analysis of Tra2β Abundance by Western Blot
Nuclear extracts (NEs) were prepared from chicken and rat tissues and cultured cells as described.17 NEs were analyzed by Western blot. Antibodies used were as follows: polyclonal antibody to the N terminus of Tra2β (S-18) (1:1000, Santa Cruz Biotechnology, Santa Cruz, Calif); mAb104 against SR proteins; DH7 (gift from Dr Helfman, Cold Spring Harbor Laboratory, New York) against polypyrimidine tract binding protein (PTB); and a GAPDH antibody (Abcam, Cambridge, Mass).
RNA–Protein Binding Assay
The RNA–protein binding assay was carried out as previously described17 using biotin-labeled RNA oligos and NEs collected from rat tissues.
Transfection of Plasmid DNA and RNA Interference Assay
The MYPT1 minigene was constructed by amplifying the 600-nt mouse genomic fragment containing the E23 and flanking introns by PCR and cloned into the second intron of the rabbit β-globin present in pβG plasmid.18 Mutations and deletion in E23 within the MYPT1 minigene construct were generated using the QuickChange II Site-Directed Mutagenesis kit (Stratagene) and confirmed by DNA sequencing. A full-length Tra2β1 cDNA was amplified from mouse liver RNA using the oligonucleotides described previously19 and cloned into the pcDNA 3.1expression vector (Invitrogen). Primary SMCs or HEK293 cells were transfected with pβG plasmid containing a wild-type or mutant MYPT1 minigene and in some experiments a Tra2β expression vector using Lipofectamine 2000 (Invitrogen) as described.17 For in vivo plasmid delivery, 6-day old chickens were anesthetized, and after laparotomy, the gizzards were injected with a wild-type or mutant MYPT1 minigene plasmid with or without Tra2β expression vector as described.20
For RNA interference (RNAi), HEK293 cells were transfected with small interfering (si)RNA synthetic duplex designed for Tra2β (SilencerR siRNA; Ambion, Austin, Tex) as described.17 SilencerR GAPDH siRNA and a nonspecific siRNA (Ambion) were used as controls for the RNAi assay.
RT-PCR and Quantification
The transfected and injected samples were analyzed by RT-PCR for the presence of exon-included and exon-skipped MYPT1 minigene transcripts using the vector specific primer pair as described.17 The 5′ end of the 3′ PCR oligonucleotide was labeled with a Cy3 fluorescent label (Integrated DNA Technologies, Coralville, IA) for the visualization of the RT-PCR product. Data are expressed as mean±SD (n≥3). Data groups were compared by 1-way ANOVA and Student’s t test using Prism version 4.0. P<0.05 was considered statistically significant.
Sequence Analysis of MYPT1 Alternative Exon E23 and Flanking Introns
The sequence of the MYPT1 E23 is highly conserved among birds and mammals (Figure 1A) and has not been identified in the frog (Figure 1B), zebrafish, or worm (not shown) homologs. Characteristics of E23 indicative of a weakly spliced exon include its small size, a 5′ splice site that deviates from the consensus GTRAGT, and the absence of a polypyrimidine tract upstream of the 3′ splice site (Figure 1A and GenBank accession no. AF110176.1). Phylogenetic analysis identified ≈160 nt of intronic sequence upstream and ≈200 nt of intronic sequence downstream of E23 that were also highly conserved (Figure 1B). E23 contains consensus cis elements for the classic SR proteins SRp-40, -55, and SC35 (Figure 1A) that were not affected by the phylogenetic sequence variation at the 2 underlined nucleotides (see Figure 1A). A manual search identified a putative and conserved Tra2β binding site in E23 similar to a proposed Tra2β consensus sequence of GHVVGANR.21
Expression of Tra2β Correlates With Splicing of MYPT1 E23
During the development of the fast smooth muscle phenotype of the rat PV and avian gizzard, Tra2β mRNA is upregulated 4- to 8-fold (Figure 2A), coincident with the complete switch from E23 skipping to E23 inclusion.6,8 The developmental upregulation of Tra2β in the fast smooth muscle is specific because (1) there is no change in the mRNA levels of SRp20, used here for internal normalization; and 2) there is no developmental change in Tra2β mRNA abundance in the slow smooth muscle of the large artery (Ao) (Figure IA in the online data supplement), in which E23 is skipped.7 Tra2β mRNA abundance is 5- to 10-fold higher in the mature fast smooth muscle tissues, the rat PV and chicken gizzard, as compared to the mature slow smooth muscle tissues, the Ao and other large arteries and veins (Figure 2B). SMCs in culture express lower levels of Tra2β and skip E23 regardless of whether they are derived from the slow Ao (Figure 2B, top) or the fast gizzard (Figure 2B, bottom), indicative of a reversion to the slow phenotype. The first order mesenteric artery (MA1) is a small resistance artery with a mixed contractile phenotype.9 It expresses relatively high level of Tra2β mRNA concordant with 80% E23 inclusion (Figure 2B).
In 2 different disease models (1) PV ligature to induce PHT and (2) MA2 ligature to induce high and low flow in the MA1s, we have previously observed shifts from the fast toward the slow programs of gene expression.9,10 In these models, Tra2β mRNA is dynamically downregulated in the PV and MA1, which temporally correlates with changes in E23 splicing (Figure 3A and 3B). Thus, these models demonstrate excellent correlations in the dynamic shifts in Tra2β mRNA abundance and E23 splicing.
Tra2β Isoform Expression Differs Between Fast and Slow Smooth Muscle
The alternative usage of exons 2 to 3 generates 3 major isoforms of Tra2β (Tra2β1, -3, and -4; Figure 4A). Tra2β1 codes for the full-length Tra2β protein, whereas Tra2β3 and β4 are thought to give rise to a truncated protein lacking the first RS domain of Tra2β.15 We assayed the expression of the splice variants of Tra2β. In the PV in the neonatal period, there is a transition from a 1:1 ratio of the transcripts that would code for the β1 and β4 isoforms at day (D)3 and D6 to a 10:1 predominance of the transcript that would code for the β1 isoform at D12 and into adulthood (Figure 4B). The ratio of transcripts that would code for the β1 and β4 isoforms in the Ao is 5:1 (Figure 4B). The transcript that would code for the β3 isoform is abundant only in PV D6. In the PHT PV model, the splicing of the Tra2β alternative exon 2 is reinduced, with a ratio of transcripts for β1to β4 of 2:1 at D3 and 3:1 at D7 and returned to 10:1 by D14 (Figure 4C). Splice variant isoforms of Tra2β are not detected in the chicken.22
Tra2β Protein Abundance During Vascular Smooth Muscle Phenotypic Specification
Tra2β protein is detected as a single band at ≈37 kDa in PV and Ao NEs (Figure 5A). Tra2β protein abundance in this PV neonatal series relative to the adult PV is: D3, 10%; D6, 20%; and D12, 70%. The abundance of Tra2β in the adult Ao is 8% of that in the adult PV, and there is no change in Tra2β protein abundance in the Ao in the neonatal period (supplemental Figure IB). In the PHT model, the abundance of Tra2β protein in the PV is reduced to 10% of control values within 3 days after the PV ligature (Figure 5B) and returned to the control values at D14. These differences in Tra2β abundance are specific as indicated by the invariance of the classic SR proteins (Figure 5A and 5B).
Tra2β Binds to the Thirty-One-Nucleotide MYPT1 E23 Exonic Sequence
The close correlation between MYPT1 E23 splicing and Tra2β expression in these different models suggested that it may be a tissue-specific enhancer of E23 splicing. In the next set of experiments, we tested the ability of Tra2β to bind to and transactivate E23 splicing. Tra2β bound to a 31-nt biotin-labeled RNA oligonucleotide containing the wild-type E23 sequence. This binding is detected with NE from adult PV but not adult Ao (Figure 6A and 6B). On a longer exposure, weak binding of Tra2β is observed in reactions that used PHT PV D3 NE. Tra2β binding is not observed in control reaction that used PV NE and an unrelated RNA sequence (Figure 6B). SRp55 and SRp40 bound to this 31-nt RNA oligonucleotide in NEs from PV, Ao, and PHT PV. A weak singlet at ≈30 kDa is observed only in reactions that used PV NE and likely represents the predicted binding of SC35.
A deletion of 6/8 nt of the putative Tra2β binding site (Figure 6A), leaving the putative SR binding sites intact (ΔTRA), abolished Tra2β binding without affecting SRp55/40/30 binding (Figure 6C). Mutation of 2 key central nucleotides of the putative Tra2β motif, from GA to UU,21 also abolished Tra2β binding without affecting SRp 55/40 binding. When a portion of the Tra2β binding site and overlapping SR p55/40/30 binding site were deleted (ΔTRA-SR), Tra2β and SRp30 binding was not detected and SRp40 binding was significantly reduced, whereas SRp55 binding was only modestly reduced. A mutation of the SR binding site immediately 5′ to the Tra2β binding site (SRmut1) significantly reduced SRp 55/40/30 binding but did not affect Tra2β binding. Mutation of both the 5′ and 3′ SR binding sites, while leaving the Tra2β binding site intact (SRmut2), nearly completely abolished the binding of the SR proteins and Tra2β (a faint band for SRp55 was detected at long exposures). In summary, Tra2β binds to the predicted 8 nt cis element in E23, and this binding is lost when the adjacent SR cis elements are mutated.
Tra2β Transactivates MYPT1 E23 Splicing
To test the functional role of Tra2β in MYPT1 E23 splicing, we inserted a MYPT1 minigene that contains the mouse E23 exon and conserved flanking intronic sequence (≈300 nt each of upstream and downstream sequence) between 2 constitutive β-globin exons in the pβG vector (Figure 7A). Transfection of this construct into cultured chicken gizzard or rat aortic SMCs (RASMCs) resulted in a very low level of E23 splicing (Figure 7B, top), consistent with nearly exclusive skipping of the endogenous E23. In contrast, injection of this construct into the gizzard smooth muscle of 1-week-old chickens resulted in a significant level of splicing of E23 (Figure 7B), consistent with splicing of the endogenous E23 in this fast smooth muscle. In either context, cotransfection of a Tra2β expression plasmid significantly increased minigene E23 splicing (Figure 7B, bottom), with a linear relationship between the amount of Tra2β expression vector and minigene E23 splicing (Figure 7C).
To test the role of the Tra2β and SR binding cis elements in E23 splicing, the mutations and deletions that reduced Tra2β and SR binding (Figure 6) were introduced into the MYPT1 minigene construct. Deletion or mutation of the Tra2β binding site (ΔTRA, TRAmut) completely or nearly completely abolished E23 splicing in the gizzard smooth muscle (Figure 7D). These constructs are unresponsive to Tra2β cotransfection (Figure 7D), demonstrating the necessity of the cis element for Tra2β activation of E23 splicing. A deletion of the 5′ portion of the Tra2β cis element and adjacent SR cis element (ΔTRA-SR) completely abolished E23 splicing, and, again, the cotransfection of the Tra2β expression plasmid had no effect. A 2-nt mutation to the 5′ SR binding site (SRmut1) that severely diminished SR binding but had no effect on Tra2β binding also markedly reduced E23 splicing in the gizzard smooth muscle. This construct is also unresponsive to cotransfection with the Tra2β expression plasmid. In summary, Tra2β transactivates splicing of the MYPT1 E23 dependent on both the Tra2β cis element and the adjacent SR cis element.
Endogenous Tra2β Is Required for MYPT1 E23 Splicing
In screening Tra2β expression in various cell types, we observed that Tra2β is 4- to 6-fold more abundant in the HEK293 cell line as compared to the cultured SMCs. In these cells, and in contrast to the cultured SMCs, there is also increased splicing of E23 in the context of the transfected MYPT1 minigene (Figure 8B). Cotransfection of the Tra2β expression plasmid further increased exon inclusion. To test the function of endogenous Tra2β in the regulation of E23 splicing, we used RNAi to knock down Tra2β. This resulted in a 50% to 75% reduction in Tra2β protein at 48 to 96 hours after transfection (Figure 8A) that was specific, as indicated by the invariance of PTB and the lack of effect of a scrambled (nonspecific) siRNA or siRNA against GAPDH (not shown). Knock down of Tra2β by RNAi reduced E23 splicing to a level similar to that of the cultured RASMCs (Figure 8B). The control siRNAs had no significant effect. These results indicate that endogenous Tra2β is necessary for the splicing of MYPT1 E23 in the context of the minigene construct.
In the present study, we propose Tra2β as a novel mediator of the fast smooth muscle phenotype based on (1) its concordant expression with the splicing of MYPT1 E23, a marker of the fast smooth muscle phenotype, in multiple developmental and disease models. This relationship is conserved between 2 classes of animals, birds and mammals that diverged several hundred million years ago, supporting its functional significance. We also base our proposal on (2) the ability of Tra2β to bind to E23 and transactivate E23 splicing in a sequence-dependent manner. To the best of our knowledge this is the first factor to be implicated in the specification of the fast muscle phenotype, in smooth, or any other muscle lineage.
Previous studies of Tra2β in vertebrates have suggested that it is ubiquitously expressed and may regulate the splicing of a number of exons. A major limitation of these studies is the absence of an analysis of Tra2β in relation to tissue-specific or developmentally regulated splicing of exons. There has also been some suggestion that Tra2β may be enriched in certain tissues, eg, in the nervous system23 and testes.24 This proposed more tissue-specific expression and function of Tra in vertebrates is consistent with its role in the fly. Tra and Tra-2 are required in female Drosophila for the suppression of the development of the male-specific abdominal muscle (also known as the muscle of Lawrence [MOL]),25 whereas the other muscles develop normally, and the development of female-specific behaviors and reproductive organs.26 Interestingly, the MOL is more highly innervated than the other muscles, and it uniquely fails to develop in the absence of innervation,27 perhaps presaging the role of innervation in the development of the fast smooth muscle phenotype.4
A model for the function of Tra2β in tissue-specific splicing of exons in vertebrate smooth muscle specification is suggested by the expression, binding, and minigene splicing data. In cells in which Tra2β expression is low, such as Ao and cultured SMCs, Tra2β binding to E23 was not detected, and mutation of the Tra2β binding site did not affect the low level of E23 inclusion. In tissues in which Tra2β expression is induced up to 10-fold, such as rat PV and avian gizzard, Tra2β binding to E23 was evident, and mutation of this site reduced E23 splicing to the level in the low-Tra2β-expressing cells. In contrast the ubiquitous classic SR proteins bound to E23 in both fast and slow phenotypes but apparently were not sufficient to activate E23 splicing. These results are consistent with a model in which Tra2β and SR proteins synergistically bind and activate alternative exon splicing,28 with Tra2β having a lower binding affinity but more potent activation of splicing of a weak exon.29 In this model, Tra2β plays a dominant role as a tissue-specific splicing factor, but combinatorial control in conjunction with ubiquitously expressed factors is also operative. Although we did not specifically measure the affinity of the cis element for Tra2β binding, the absence of Tra2β binding in NEs in which its abundance is low, eg, the Ao and cultured SMCs, is consistent with a low affinity of Tra2β for this cis element, although we cannot exclude the possibility of an inhibitor of Tra2β binding in these extracts. The Tra2β cis element identified in E23 is based on the consensus sequence of GHVVGANR,21 a modification of the previously described Tra2 binding sequence consisting of GAA repeats.16 The rather degenerate or nonspecific nature of these binding sequences would seem to be at odds with the proposed highly specific role of Tra2β in regulating exon splicing, an issue that may be resolved with further definition of Tra2β target exons and binding sequences.
A limitation of this study is that the tests of MYPT1 E23 splicing were performed in transient transfection assays in vitro and in vivo with an ≈600-nt MYPT1 minigene construct. The level of splicing of the MYPT1 minigene E23 in the gizzard was less than that of the endogenous gene. This is most likely attributable to the out-titration of splicing factors in this transient transfection system, as also observed in our previous experiments with a different construct.20 Consistent with this interpretation, cotransfection of the Tra2β expression plasmid further increased MYPT1 minigene E23 splicing in the gizzard tissue, from 30% to 50%, suggesting that under the conditions of this transient transfection experiments Tra2β levels were limiting. Given that Tra2β does not drive E23 splicing to 100%, as well as the multitude of splicing decisions that are made in phenotypically diverse SMCs, it seems quite likely that additional factors will be involved. A related question is the full repertoire of exon splicing that Tra2β may regulate in vivo. Our initial sequence analysis has identified potential Tra2β binding sites in other smooth muscle alternative exons including Caldesmon, Calponin and m-Vinculin. Identifying the full repertoire of exons under the control of Tra2β in vivo can be addressed by loss-of-function studies. The correlation of MYPT1 E23 splicing and Tra2β expression reported here in the rat and chicken is also observed in the mouse (data not shown), indicating that it will be a useful model to address this question.
In contrast to striated muscle, scarce attention has been paid to the question of the specification of muscle phenotypes in the development of the vascular system and their modulation in disease, despite its likely high significance with respect to vascular function. Tra2β may serve as a novel nodal point to address these questions. A good candidate for upstream control is a 559-nt sequence within the Tra2β intron 1 that is an ultraconserved sequence, with 95% to 100% sequence identity between birds and mammals.30 A high percentage of ultraconserved sequence function in the regulation of tissue-specific gene transcription and developmental phenotypic specification,30,31 and it will be of great interest to test the Tra2β ultraconserved sequence in this regard. Of note, the tissue-restricted expression (splicing) of Drosophila Tra is under the control of Sxl, a pathway that is not evolutionarily conserved, even within Dipterans,32 whereas more distantly related organisms lack homologs of sxl. This is consistent with the theory of “bottom up” evolution of gene regulatory networks,33 in which gene regulatory factors maintain their functional properties, in this case tissue-specific splicing of exons, and are coopted by different upstream inputs to regulate new gene networks as lineages diversify through evolution (also see34). Downstream of Tra, only a few bona fide targets have been identified, including dsx and fruitless in Drosophila, consistent with its specialized role in this organism. These downstream factors have been conserved and expanded through evolution and constitute the DMRT and BTB-ZF transcription factor gene families, respectively, in higher vertebrates. The present study provides a foundation and rationale for investigation of Tra2β gain- and loss-of-function studies in the mouse to further define the program of exon splicing that is regulated by Tra2β in vivo. Such studies may provide a link between exon splicing and transcription in the control of the smooth muscle phenotype in development and disease.
We thank Dr Curthoys (Colorado State University, Ft Collins) for providing pβG plasmid and Michael C. Payne for technical assistance.
Sources of Funding
This work was supported by NIH grant RO1 HL-66171 (to S.A.F.).
Original received April 25, 2008; revision received July 11, 2008; accepted July 17, 2008.
Owens GK. Regulation of differentiation of vascular smooth muscle cells. Physiol Rev. 1995; 75: 487–517.
Ogut O, Brozovich FV. Determinants of the contractile properties in the embryonic chicken gizzard and aorta. Am J Physiol Cell Physiol. 2000; 279: C1722–C1732.
Khatri JJ, Joyce KM, Brozovich FV, Fisher SA. Role of myosin phosphatase isoforms in cGMP-mediated smooth muscle relaxation. J Biol Chem. 2001; 276: 37250–37257.
Dirksen WP, Vladic F, Fisher SA. A myosin phosphatase targeting subunit isoform transition defines a smooth muscle developmental phenotypic switch. Am J Physiol Cell Physiol. 2000; 278: C589–C600.
Payne MC, Zhang HY, Shirasawa Y, Koga Y, Ikebe M, Benoit JN, Fisher SA. Dynamic changes in expression of myosin phosphatase in a model of portal hypertension. Am J Physiol Heart Circ Physiol. 2004; 286: H1801–H1810.
Zhang H, Fisher SA. Conditioning effect of blood flow on resistance artery smooth muscle myosin phosphatase. Circ Res. 2007; 100: 730–737.
Heinrichs V, Ryner LC, Baker BS. Regulation of sex-specific selection of fruitless 5′ splice sites by transformer and transformer-2. Mol Cell Biol. 1998; 18: 450–458.
Dauwalder B, Amaya-Manzanares F, Mattox W. A human homologue of the Drosophila sex determination factor transformer-2 has conserved splicing regulatory functions. Proc Natl Acad Sci U S A. 1996; 93: 9004–9009.
Shukla S, Gatto-Konczak F, Breathnach R, Fisher SA. Competition of PTB with TIA proteins for binding to a U-rich cis element determines tissue-specific splicing of the myosin phosphatase targeting subunit 1. RNA. 2005; 11: 1725–1736.
Patel NA, Eichler DC, Chappell DS, Illingworth PA, Chalfant CE, Yamamoto M, Dean NM, Wyatt JR, Mebert K, Watson JE, Cooper DR. The protein kinase C beta II exon confers mRNA instability in the presence of high glucose concentrations. J Biol Chem. 2003; 278: 1149–1157.
Dirksen WP, Mohamed SA, Fisher SA. Splicing of a myosin phosphatase targeting subunit 1 alternative exon is regulated by intronic cis elements and a novel bipartite exonic enhancer/silencer element. J Biol Chem. 2003; 278: 9722–9732.
Stoilov P, Daoud R, Nayler O, Stamm S. Human tra2-beta1 autoregulates its protein concentration by influencing alternative splicing of its pre-mRNA. Hum Mol Genet. 2004; 13: 509–524.
Venables JP, Bourgeois CF, Dalgliesh C, Kister L, Stevenin J, Elliott DJ. Up-regulation of the ubiquitous alternative splicing factor Tra2beta causes inclusion of a germ cell-specific exon. Hum Mol Genet. 2005; 14: 2289–2303.
Taylor BJ. Differentiation of a male-specific muscle in Drosophila melanogaster does not require the sex-determining genes doublesex or intersex. Genetics. 1992; 132: 179–191.
Currie DA, Bate M. Innervation is essential for the development and differentiation of a sex-specific adult muscle in Drosophila melanogaster. Development. 1995; 121: 2549–2557.
Lynch KW, Maniatis T. Assembly of specific SR protein complexes on distinct regulatory elements of the Drosophila doublesex splicing enhancer. Genes Dev. 1996; 10: 2089–2101.
Sciabica KS, Hertel KJ. The splicing regulators Tra and Tra2 are unusually potent activators of pre-mRNA splicing. Nucleic Acids Res. 2006; 34: 6612–6620.
Bejerano G, Pheasant M, Makunin I, Stephen S, Kent WJ, Mattick JS, Haussler D. Ultraconserved elements in the human genome. Science. 2004; 304: 1321–1325.
Pennacchio LA, Ahituv N, Moses AM, Prabhakar S, Nobrega MA, Shoukry M, Minovitsky S, Dubchak I, Holt A, Lewis KD, Plajzer-Frick I, Akiyama J, De Val S, Afzal V, Black BL, Couronne O, Eisen MB, Visel A, Rubin EM. In vivo enhancer analysis of human conserved non-coding sequences. Nature. 2006; 444: 499–502.
Schutt C, Nothiger R. Structure, function and evolution of sex-determining systems in Dipteran insects. Development. 2000; 127: 667–677.
Olson EN. Gene regulatory networks in the evolution and development of the heart. Science. 2006; 313: 1922–1927.