|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 MUSCLEBLIND IN DEVELOPMENT AND DISEASE By MICHAEL GUSTAVE POULOS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 Michael Gustave Poulos
3 To my family
4 ACKNOWLEDGMENTS I would like to thank my wife, Ashley Poulos, for her unconditional love and support during my graduate school journey. I am also indebted to my parents, David and Eileen Poulos, and my brother, Nathan Poulos, for providing me with constant support and encouragement throughout my life. I am grateful to the mentors in my undergraduate and graduate studies that helped to foster my scientific develop ment. At Grand Valley State University, Dr. Patrick Thorpe and Dr. Roderick Morgan introduced me to molecular biology and allowed me to explore my scientific curiosity in their laboratories, while teaching me the basic fundamentals of research in an encouraging and fun setting. I would especially like to thank my graduate mentor, Dr. Maury Swanson, who taught me how to critically evaluate and approach science, and has provided me with the necessary tools to be successful at the next level. I would also like to thank my committee members, Drs. Brian Harfe, William Hauswirth, and Jake Streit for providing guidance during my graduate career. Additionally, Molecular Genetics and Microbiology faculty Drs. Henry Baker, Rich Condit, Jim Resnick, and Al Lewin have always found time to talk science or to offer advice, and have greatly enriched my graduate school experience. Fellow Swanson lab graduate students, Rahul Kanadia, Yuan Yuan, Jason ORourke, Jihae Shin, and Ranjan Batr a have always made laboratory an intellectually stimulating and fun experience. I would also like to thank Joyce Conners for her continuous support, help with the administrative aspects of my graduate career, and generally making my life as easy as possible.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF FIGURES .......................................................................................................... 8 LIST OF ABBREVIATIONS ........................................................................................... 11 ABSTRACT ................................................................................................................... 13 CHAPTE R 1 OVE RVIEW OF MYOTONIC DYSTROPHY ........................................................... 16 Myotonic Dystrophy Type 1 .................................................................................... 16 Congenital Myotonic Dystrophy .............................................................................. 18 Haploinsufficiency Model of Disease ...................................................................... 19 Myotonic Dystrophy Type 2 .................................................................................... 21 RNA Dominance and Sequestration Models of Disease ......................................... 21 Multiple Genetic Contributions and the Complex Etiology of DM ............................ 25 2 MUSCLEBLIND PROTEINS REGULATE ALTERNATIVE SPLICING .................... 31 Introduction ............................................................................................................. 31 Introduction to Alternative Splicing ................................................................... 31 Alternative Splicing Misregulation in DM .......................................................... 34 Evolutionaril y Conserved Muscleblind is Important for Muscle Development ... 38 Results .................................................................................................................... 39 Muscleblind Protein Family: Evolutionary Conservation of Structure and Function ........................................................................................................ 40 MBNL1 Proteins Directly Regulate Alternative Splicing of Gene Transcripts Misregulated in Myotonic Dystrophy ............................................................. 41 MBNL1 Interacts Directly with Cis Elements in cTNT pre mRNA ..................... 43 (CUG)n and (CAG)n Repeats Relocalize MBNL1, but only (CUG)n Alter cTNT Splicing ................................................................................................ 44 The Stability of MBNL1:RNA complexes varies between (CUG)n, (CCUG)n and (CAG)n Repeats ..................................................................................... 45 Discussion .............................................................................................................. 46 Muscleblind Proteins Directly Regulate Alternative Splicing ............................. 47 Microsatellite Repeat Expansions Display Variable MBNL1 Stability and Toxicity .......................................................................................................... 51
6 3 LOSS OF MBNL3 4xC3H ISOFORMS ARE NOT SUFF ICIENT TO MODEL CONGENITAL MYOTONIC DYSTROPHY ............................................................. 70 Introduction ............................................................................................................. 70 Overview of Congenital Myotonic Dystrophy .................................................... 70 Potential M odels of CDM .................................................................................. 72 MBNL Family in DM ......................................................................................... 74 Overview of MBNL3 ......................................................................................... 76 Results .................................................................................................................... 77 Mbnl3 mRNA Expression and Alternative Splicing are Spatially and Temporally Regulated ................................................................................... 77 Polyclonal Antisera Raised Against the C terminus of Mbnl3 ........................... 79 Mbnl3 Isoforms Localize to both the Nucleus and Cytoplasm .......................... 80 Mbnl3 Proteins are Primarily Expressed during Embryogenesis ...................... 82 Loss of Mbnl3 Inhibits Myogenic Differentiation in a C2C12 Model .................. 83 Targeting Mbnl3 to Generate a Conditional Allele in Embryonic Stem Cells .... 84 Mbnl3E2/Y Mic e Fail to Express Mbnl3 4XC3H Isoforms .................................. 86 Mbnl3E2/Y Mice do not Recapitulate Cardinal Phenotypes of Congenital Myotonic Dystrophy ....................................................................................... 87 Loss of Mbnl3 4XC3H Isoforms does not Inhibit Skeletal Muscle Regeneration in an Injury Model ................................................................... 88 Discussion .............................................................................................................. 89 Mbnl3 is Expressed in Developing Tissues that are Affected in CDM .............. 90 Mbnl3 is Required for C2C12 Differentiation .................................................... 92 Loss of Mbnl3 4XC3H Isoforms are not Sufficient to Phenocopy CDM or Skeletal Muscle Wasting ............................................................................... 93 4 CONCLUDING REMARKS AND FUTURE DIRECTIONS .................................... 124 5 MATERIALS AND METHODS .............................................................................. 127 MBNL3 .................................................................................................................. 127 RNA Analysis ................................................................................................. 127 Protein Lysate and Fractionation .................................................................... 129 Immunobloting ................................................................................................ 129 Generation of an Mbnl3 Polyclonal Antibody .................................................. 130 Expression Vectors ........................................................................................ 131 Immunoprecipitation and Mass spectrometry ................................................. 132 siRNA and P lasmid Transfections .................................................................. 134 C2C12 differentiation ...................................................................................... 135 Skeletal Muscle Regeneration ........................................................................ 135 Mbnl3E2/Y Mouse Generation ......................................................................... 136 Genomic DNA Isolation, Southern Blotting, Probe Generation and Genotyping .................................................................................................. 139 Skeletal Preparations ..................................................................................... 141 Wild Type and Mbnl3E2/Y Growth Curve ........................................................ 142 Sectioning, Immunoflourescence, and H&E Staining ..................................... 142
7 MBNL1 .................................................................................................................. 144 Transfections and RNA Analysis .................................................................... 144 In Vitro Transcription and UV Crosslinking ..................................................... 145 Plasmids ......................................................................................................... 146 Transfection of siRNA .................................................................................... 146 Immunoblot Analysis ...................................................................................... 147 Flourescent In Situ Hybridization and Immunocytochemistry ......................... 148 Competition Assay ......................................................................................... 149 LIST OF REFERENCES ............................................................................................. 153 BIOGRAPHICAL SKETCH .......................................................................................... 168
8 LIST OF FIGURES Figure page 1 1 Non coding microsatellite repeat expansions in myotonic dystrophy .................. 27 1 2 Genomic DNA replication errors promote microsatellite repeat expansions and contractions ................................................................................................. 28 1 3 Comparison of myotonic dystrophy symptoms ................................................... 29 1 4 MBNL1 sequestration model for DM1 ................................................................. 30 2 1 Patterns of alternative splicing ............................................................................ 54 2 2 Schematic of constitutive and alternative premRNA splicing mechanism ......... 55 2 3 Model of MBNL1 sequestration promoting alternative splicing defects in DM1 .. 56 2 4 Alte rnative splicing produces four distinct Mbl isoforms in Drosophila ................ 57 2 5 Mbl protein isoforms display different subcellular localizations ........................... 58 2 6 Mbl colocalizes with (CUG)300 RNA in nuclear foci ............................................. 59 2 7 Drosophila Mbl regulates alternative splicing of actinin ................................... 60 2 8 Schematic of cTNT and INSR alternative splicing minigene reporters ............... 61 2 9 MBNL1, MBNL2, and MBNL3 regulate alternative splicing of cTNT and INSR .. 62 2 10 Endogenous MBNL1 regulates alternative splicing of cTNT and INSR minigenes ........................................................................................................... 63 2 11 MBNL1 binds upstream of alternatively spliced cTNT exon 5 ............................ 64 2 12 MBNL1 binding site mutations inhibit MBNL1, MBNL2, and MBNL3 responsiveness ................................................................................................... 65 2 13 Alignment of the human and chicken cTNT MBNL1 binding sites reveals a conserved motif .................................................................................................. 66 2 14 MBNL1 colocalizes with (CUG)960 and (CAG)960 RNA in nuclear foci ................. 67 2 15 CUG960, but not CAG960, expression alters cTNT alternative splicing ................. 68 2 16 MBNL1 displays increased stability with (CUG)54 RNA ...................................... 69
9 3 1 Phylogram shows evolutionary proximity of human C3H zinc finger motif containing proteins ............................................................................................. 96 3 2 The MBNL family is composed of three closely related paralogs ....................... 97 3 3 Schematic of potential genetic contributions to DM1 .......................................... 98 3 4 Mbnl3 expression is restricted spatially and temporally ...................................... 99 3 5 The Mbnl3 gene produces multiple isoforms through alternative splicing ......... 100 3 6 Mbnl3 is predicted to encode several different isoforms ................................... 101 3 7 Alternative splicing of Mbnl3 produces multiple isoforms throughout development ..................................................................................................... 102 3 8 Mbnl3 antisera recognizes two distinct bands in E15 placenta whole cell lysate ................................................................................................................ 103 3 9 Mbnl3 antisera does not cross react with other Mbnl proteins ...................... 104 3 10 Mbnl3 distnguishes between Mbnl family members by immunocytochemistry ....................................................................................... 105 3 11 Two distinct Mbnl3 isoforms are expressed in C2C12 and Huh7 cells ............. 106 3 12 MBNL3 localizes to nuclear and cytoplasmic foci ............................................. 107 3 13 MBNL3 isoforms are found in different subcellular compartments .................... 108 3 14 Mbnl3 expression is restricted te mporally and spatially during embryogenesis and postnatally ................................................................................................. 109 3 15 Murine embryonic myogenesis ......................................................................... 110 3 16 Notexin promotes murine skeletal muscle necrosis followed by regeneration in vivo ............................................................................................................... 111 3 17 Mbnl3 is expressed during adult skeletal muscle regeneration ........................ 112 3 18 Adult skeletal muscle regeneration ................................................................... 113 3 19 Loss of Mbnl3 inhibits myogenic differentiation in a C2C12 in vitro model ....... 1 14 3 20 Generation of a conditional Mbnl3 allele ( Mbnl3cond/Y) in ESCs ........................ 115 3 21 Cre mediated recombination of Mbnl3cond/Y allele removes exon 2 containing transcripts from ESCs ....................................................................................... 116
10 3 22 Mbnl3E2/Y mice lack full length Mbnl3 isoforms ................................................ 117 3 23 P1 Mbnl3E2/Y mice do not display defects associated with CDM ..................... 118 3 24 Mbnl3E2/Y neonates do not present with delayed myofibers at P1 .................. 119 3 25 Mbnl3E2/Y mice gain weight normally up to 10 weeks of age ........................... 120 3 26 Mbnl3E2/Y mice do not express full length Mbnl3 isoforms during adult skeletal muscle regeneration ............................................................................ 121 3 27 Normal skeletal muscle regeneration in Mbnl3E2/Y mice ................................. 122 3 28 Mbnl3E2/Y normal skeletal muscle regeneration .............................................. 123 4 1 Primers used for RT PCR, Mbnl3E2/+ genotyping, probe generation, and subcloning ........................................................................................................ 152
11 LIST OF ABBREVIATION S DM myotonic dystrophy CDM congenital myotonic dystrophy HD Huntington Disease SMA spinal muscular atrophy SCA spinal cerebellar ataxia UTR untranslated region Kb kilobase HP1 heterochromatin protein 1 HSALR human skeletal actin long repeats PKR protein kinase R dsRNA double stranded RNA ssRNA single stranded RNA miRNA microRNA MBNL m uscleblind like AAV adenoassociated virus CNS central nervous system NPC nuclear pore complex Dscam Down syndrome cell adhesion molecule snRNP small nuclear ribonucleoprotein particle SF1 splicing factor 1 U2AF65 U2 auxiliary factor 65 SR serine/arginine rich protein hnRNP heterogenous nuclear ribonucleoprotein RRM RNA recognition motif
12 ESE/ESS exonic splicing enhancer/silencer ISE/ISS intronic splicing enhancer/silencer SMN survival of motor neuron protein IR insulin receptor cTNT/TNNT2 cardiac troponin T TNNT3 fast skeletal muscle troponin T CLCN1 muscle specific chloride channel SCN4A voltage gated sodium channel subunit A ClaLC clathrin light chain Mbl muscleblind PTB polypyrimidine tract binding protein E15 embryonic day 15 P1 postnatal day 1 CUGBP1 CUG binding protein 1 LDHA lactate dehydrogenase A Mhc myosin heavy chain GluI glutamine synthetase TA tibialis anterior H&E hematoxylin and eosin ESC embryonic stem cell MEF mouse embryonic fibroblast NeoR neomycin resistance gene TK Herpes Simplex Virus thymidine kinase gene
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MUSCLEBLIND IN DEVELOPMENT AND DISEASE By Michael Poulos May 2010 Chair: Name Maurice Swanson Major: Medical Sciences Genetics Myotonic dystrophy (DM) may be the most variable human disorder described, affecting every age group from newborns to the adult onset form which can range from adolescent manifestations to lateonset disease. Frequent disease characteristics include skeletal muscle weakness, myotonia, cataracts, and distinct changes in alternative splicing patterns. Congenital patients present with additional symptoms at birth, consisting of immature muscle, pulmonary insufficiencies, and central nervous system (CNS) involvement. DM arises from the expansion of two similar noncoding microsatellites in the DMPK and CNBP genes which have been proposed to cause disease through a common mechanism, a toxic RNA gainof function which can either inhibit or activate specific proteins. One of these candidates, the muscleblindlike (MBNL) family of proteins encoded by three genes, MBNL1 MBNL2 and MBNL3 are sequestered into discrete nuclear foci by RNA repeat expansions, preventing interactions with endogenous RNA targets and compromising their activity. However, it is unclear how the inhibition of MBNL function leads to disease and the extent of MBNL involvement in the diverse prese n tation of symptoms in DM. The goal of this study was to investigate the normal function of MBNL proteins which are compromised in DM and
14 their role in congenital disease. Our working hypothesis is that MBNL genes show distinct tem p oral and spatial expres sion patterns that inf luence age of onset and disease associated pathological changes. For MBNL1, we first demonstrate that this protein is an alternative splicing factor that interacts with premRNAs misspliced in DM and that this interaction is necessary for splicing responsiveness. Additionally, we show that MBNL1 interacts with DM1 pathogenic and nonpathogenic repeat RNAs, but inherent differences in these interactions contribute to the ability to promote disease associated changes in alternative spli cing. This and additional evidence supports the hypothesis that loss of MBNL1 function in DM causes defects in the alternative splicing of specific genes during postnatal development which leads to distinct pathological features in adult onset disease, including myotonia and insulin insensitivity. Since MBNL3 shows a restricted expression pattern during development and regeneration, we employed an in vivo approach to test the hypothesis that Mbnl3 is necessary for normal embryonic development and loss of t his protein results in congenital myotonic dystrophy (CDM). Mbnl3 expression patterns, particularly in tissues affected in CDM including skeletal muscle and lung, as well as the cellular localization of Mbnl3 protein isoforms suggest a function that is di stinct from other Mbnl family members. Indeed, siRNA mediated knock down analysis of Mbnl3 in the C2C12 cell culture model of myogenesis demonstrates that Mbnl3 is essential for normal myogenic differentiation. To test the hypothesis that Mbnl3 was required for normal myogenesis in vivo we generated Mbnl3 isoform knocko ut mice that fail to express the major Mbnl3 isoform. While these Mbnl3E2/Y knockout mice do not recapitulate the
15 CDM phenotype, upregulation of other Mbnl3 isoforms occurs in this line suggesting the possibility of functional complementation. A possibl e role for Mbnl3 expression in adult muscle regeneration following injury was also discovered. We conclude that Mbnl3 null lines must be created to determine if expression of this Mbnl gene is essential for normal embryonic muscle development. Overall, t hese results demonstrate that MBNL1 is an alternative splicing factor that regulates gene expression during postnatal life while MBNL3 expression is essential for normal myogenic differentiation in vitro and possibly in vivo
16 CHAPTER 1 OVERVIEW OF MYOTONIC DYSTROPHY Myotonic Dystrophy Type 1 Myotonic dystrophy type 1 (DM1) is a dominantly inherited, multisystemic, neuromuscular disorder. It is the most common adult onset form of muscular dystrophy, affecting 1 in 8000 individuals worldwide (Harper, 2001) Common DM1 symptoms include s ubcapsular cataracts, progressive distal skeletal muscle weakness and wasting, insulin resistance, cardiac arrhythmia, cognitive impairment, myotonia (a failure to relax skeletal muscle following a voluntary contration) (Avaria and Patterson, 1994; Groh et al., 2002; Lacomis et al., 1994; Milner Brown and Miller, 1990; Stuart et al., 1983) and a molecular defect in the alternative splicing of specific premRNAs (Ranum and Cooper, 2006) An interesting hallmark of this disease is the remarkable variability seen in the age of onset and penetrance of these cli nical phenotypes DM1 symptoms have been reported from adolescence to adulthood with early onset cataracts, muscle weakness, and myotonia being the most frequent disease manifestation (Harper, 2001) DM1 also displays genetic anticipation, a phenomenon in which each successive generation afflicted with disease presents with an earlier age of onset accompanied by more severe symptoms. Additionally, at its earliest time of onset, infants are affected with a distinct and severe form of disease (see c ongenital myotonic dystrophy). The dynamic nature of the characteristics of this disease have historically presented problems for defining the disorder in terms of symptoms. However, the variability demonstrated in DM1 became more clear in 1992, when the molecular defect was uncovered (Brook et al., 1992; Buxton et al., 1992; Fu et al., 1992; Harley et al., 1992; Mahadevan et al., 1992)
17 DM1 is caused by a (CTG)n microsatellite repeat expansion in the 3 untranslated region (UTR) of the DMPK gene. In the normal population, CTG repeats can range from 5 to 37. In disease, repeats expand to (CTG)501000 for adult onset patients and (CTG)10004000 for congenital patients (Anvret et al., 1993; Botta et al., 2008; Hedberg et al., 1999) (Fig 1 1A). Individuals with the initial expansion, called the protomutation, generally have smaller repeats and present with more mild symptoms and a later age of onset. Further transmission of the expanded allele results in a longer repeat region and more severe disease in the following generations (Harper, 2001) Interestingly, genetic anticipation, is also seen in other dominantly inherited trinucleotide repeat disorders which are subject to expansions, including Huntington disease (HD) and several spinal cerebellar ataxias ( SCA ) (Koshy and Zoghbi, 1997; La Spada et al., 1994) (CTG)>50 repeats also display somatic instability that result in heterogenous expansions in tissues, likely contributing to the high degree of phenotypic variability seen in DM1. Although (CTG)n size can differ within the sam e individual due to somatic mosaicism, disease severity correlates positively with the length of the CTG expansion itself (Ashizawa et al., 1993; Martorell et al., 1997; Wong et al ., 1995) While mechanism of expansion of CTG/CAG repeats are not well understood, multiple models have been proposed to explain this process and have focused on errors during DNA replication and repair During leading strand replication, CTG/CAG repeats can adopt stable DNA secondary structures that disassociate the 3 end of the synthesizing strand from the parental strand and allow it to reanneal out of register within the repeats of the parental strand (Pearson et al., 2005) (Fig. 1 2). Alternatively, Okazaki fragments located entirely within the expanded repeat itself during lagging
18 strand synthesis, which are not anchored to nonrepetitive sequence, can slip and produce unpaired repeats (Richards and Sutherland, 1994) Repair and further rounds of replication promote expansion. It is, however, unclear what causes the protomutation from (CTG)5 37 to (CTG)>37, making the repeats susceptible to further expansion. Transgenic mice expressing DMPK (CTG)55 demonstrate intergenerational expansions in repeat size, while DMPK (CTG)20 remain static (Seznec et al., 2000) This result substantiates the observation in humans that CTG repeats become unstable between (CTG)3750 and are subject to further expansion in the germline, but it fails to explain the underlying cause of the protomutation. Congenital Myotonic Dystrophy Congenital myotonic dystrophy (CDM) is caused by (CTG)10004000 repeat expansions (Tsilfidis et al., 1992) This is the same CTG mutation in DM1, however, the larger expansion affects newborns with a distinct set of clinical features. Individuals born with CDM present with immature skeletal muscle, poor suckling, talipes (club foot), mental retardation, pulmonary involvement, a nd hyotonia (a lack of baseline muscle tone resulting in movement impairment) (Harper, 1975; Harper, 2001) The main cause of mortality in CDM patients is due to respiratory deficiencies, believed to result from hypotonia of the diaphragm and intercostal muscles, as well as lung involvement (Campbell et al., 2004; Rutherford et al., 1989) CDM children that overcome the early onset symptoms generally present with delays in developmental milestones until 34 years of age (Echenne et al., 2008) Remarkably, delays in skeletal muscle development and pulmonary involvement is resolved in early childhood. However, most CDM patients develop adult onset D M1 symptoms by adolescence.
19 H aploinsufficiency Model of Disease Dominantly inherited microsatellite repeat expansion diseases, i ncluding HD and SCA 1, have CAG repeats located within their coding regions. These repeats code for a polyglutamine expansion, resulting in a toxic gainof function that affects multiple cellular processes (Imarisio et al., 2008; Kang and Hong, 2009) However, DM1associated (CTG)n expansions in the DMPK 3UTR are also inherited in a dominant manner, but do not affected the pre dicted protein product. This raises an intriguing question; how does a noncoding mutation cause a dominantly inherited disease? One potential explanation for this observation is the haploinsufficiency model, which proposes that the (CTG)n repeat expansi on causes downregulation of DMPK and the flanking genes, DMWD and SIX5 DMWD DMPK and SIX5 are located within a 20 kilobase (kb) region of chromosome 19q13. Previous studies have reported evidence of decreased mRNA expression from these genes in DM1 and CDM tissues when compared to unaffected individuals (Eriksson et al., 2001; Fu et al., 1993; Hamshere et al., 1997; Inu kai et al., 2000; Klesert et al., 1997; Novelli et al., 1993; Sabouri et al., 1993) (CTG)n repeats are flanked by CTCF insulator elements, which bind CTCF proteins and establish boundaries preventing the unwanted spread of heterochromatin at the DMPK locus (Phillips and Corces, 2009) In DM1, CTCF binding sites are methylated, preventing CTCF from binding (Filippova et al., 2001) In the absence of CTCF, antisense transcription from the downstream gene SIX5 extends thorugh the CTG repeats and promotes the spread of heterchormatin formation by recruiting heterochromatin protein 1 (HP1 a protein involved in localization of silenced genes to the nuclear lamin (Cho et al., 2005) Additionally, loss of DNaseI hypersensitivity
20 sites at the DMPK locus supports the idea that heterochromatin formation could lead to loss of expression at the DMPK locus (Otten and Tapscott, 1995) To test the idea that haploinsufficieny of the DMPK locus, caused by the expansion of (CTG)n repeats, promoted DM1 and CDM, Dmpk and Six5 knockout mice were generated. Dmpk/ mice display a mild late onset skeletal muscle myopathy, characterized by a reduction in force generation, increased fibrosis, and a d ecrease in organization of the Z line (Reddy et al., 1996) Dmpk/ mice also demonstrate cardiac conduction defects, including first, second and thi rd atrioventricular blocks (Berul et al., 1999) However, only a first degree atrioventricular block and no skeletal muscle defects are observed in Dmpk+/ mice Six 5+ / mice develop oc ular cataracts, while Six 5/ males are st erile and display hypogonadism (Kle sert et al., 2000; Sarkar et al., 2000; Sarkar et al., 2004) However, the age related nuclear cataracts in S ix 5+ / mice are different than the subcapsular cataracts seen in DM1. Six5+/ mice also display mild cardiac conduction defects (Berul et al., 2000; Wakimoto et al., 2002) N either of these models present with cardinal DM1 phenotypes, including skeletal muscle wasting, myotonia, subcapsular cataracts or changes in alternative splicing (Personius et al., 2005) In addition, skeletal muscle phenotypes and hypogonadism are only observed in D mpk/ and Six 5/ mice, not in D mpk+/ and Six 5+ / mice as the haploinsufficieny model would predict. While loss of DM1 associated gene expression may contribute to the disease, it is clear that haploinsufficiency of these genes is not solely responsible for disease onset. To date, no one has generated a Dmwd knockout mouse to assay for its contr ibution to the haploinsufficieny model.
21 Myotonic Dystrophy Type 2 The symptoms of DM1 have been well documented in the literature. However, a small percentage of families with myotonic dystrophy symptoms fail to develop genetic anticipation or a cong enita l form of the disease. Additionally, these individuals display proximal skeletal muscle weakness (as opposed to distal weakness in DM1) and normal (CTG)5 37 repeats in the DMPK 3UTR (Ricker et al., 1995; Thornton et al., 1994) In 2001, a (CCTG)n mutation linked to these families was identified in the first intron o f the CNBP gene (formerly known as ZNF9 ), so the disease was renamed myotonic dystrophy type 2 (DM2) (Liquori et al., 2001) In the normal population, unaffected individuals posses (CCTG)>27, while the DM2 pathogenic expansions contain (CCTG)7511,000 (Fig. 1 1B) These patients comprise approximately 5% of tota l myotonic dystrophy cases Interestingly, DM1 and DM2 share a significant overlap in adult onset disease manifestation (Fig. 13) despite being caused by two independent mutations in seemingly unrelated genes. Cnbp+/ mice have a reduced expression of Cnbp mRNA and show skeletal muscle degeneration and enlarged hearts (Chen et al., 2007) However, unlike DM1, there is no evidence that DM2 expansions alter expression from the locus harboring the (CCTG)n expansion (Botta et al., 2006) Therefore, haploinsufficiency of CNBP is unlikely to be a cause of DM2. Taken together, these observations predict a common primary mechanism of pathogenesis in DM1 and DM2, that is disconnected from the specific dis ease associated loci. RNA Dominance and Sequestration Models of Disease How do two different noncoding C(CTG)n repeat expansions in unrelated genes result in a dominantly inherited disease with remarkably similar phenotypes? The first evidence of a comm on mechanism emerged from the observation that both (CUG)>50
22 and (CCUG)>75 RNAs localize to discrete nuclear foci in DM1/DM2 skeletal muscle sections and cell culture but not in normal controls (Davis et al., 1997; Margolis et al., 2006) R elocalization into nuclear foci inhibits DMPK mRNA from nuclear export and translation (Davis et al., 1997) Additionally, a transgenic mouse expressing (C TG)250 in the 3UTR of a human skeletal actin transgene ( HSALR) developed nuclear RNA foci similar to those seen in DM1 and DM2 and phenocopied characteristic DM skeletal muscle symptoms, including myotonia, centralized nuclei (indicating damaged and regenerating muscle fibers ), and alternative splicing defects (Mankodi et al., 2000) HSASR control mice expressing (CTG)5 were unaffected. Interestingly, the severity of the DM phenotype in mice correlated with expression of the transgene. Highly expressing HSALR mice were more affected than moderately expressing mice, while transgenes inserted into heterochromatic regions that did not express detectable levels of (CUG)250 RNA did not develop DM symptoms. The presence of three unrelated and noncoding C(CUG)n repeat expansions that aberrantly localize in nuclear foci and promote expression dependant DM phenotypes suggest a toxic gainof function at the RNA level. One model proposes that toxic C(CUG)n RNAs activate cellular proteins which, in turn, can contribute to the onset of disease through unwanted downstream events. First, protein kinase R (PKR) is a doublestranded RNA (dsRNA) binding protein that functions in innate immunity by interacting with viral dsRNAs in infected cells (Sadler and Williams, 2007) Activation of PKR through binding of dsRNA promot es cell death by inhibiting cellular translation, via phosphorylation of eIF 2, and activation of other cellular stress response mec hanisms, including RNase L. PKR binds toxic (CUG)n RNA
23 in a lengthdependent manner with increasing affinity in vitr o and activates PKR in cell culture (Tian et al., 2000) However, PKR is not activated in DM1 patient tissues and Pkr/ -; HSALR mice develop DM phenotypes, demonstrating that Pkr is not necessary for onset of disease in the presence of expanded (CUG)250 (Mankodi et al., 2003) Second, expanded (CUG)n repeats can also act as substrates for DICER, a dsRNA binding protein with endonuclease activity that cleaves premicroRNAs into 22 nucleotide fragments in the microRNA (miRNA) pathway (Krol et al., 2007) Processing of (CUG)n repeats by DICER can promote the downregulation of CAG containing transcripts through the siRNA pathway in vitro Nevertheless CUG siRNAs have not been detected in DM1 tissues or in a mouse model expressing (CUG)250 repeats in skeletal muscle (Osborne et al., 2009) No additional evidence has substantiated CUG siRNA involve ment in DM1 pathogenesis. Finally, CUGBP1, a singlestranded RNA (ssRNA) binding protein involved in developmentally regulated alternative splicing and RNA turnover, is upregulated ~ 2fold in DM1 and DM2 skeletal muscle and heart when compared to the nor mal tissues (Timchenko et al., 2001) CUGBP1 upregulation is caused by increased stability of the protein through PKC mediated hyperphosphorylation (Kuyumcu Martinez et al., 2007) Mice expressing a tamoxifeninducible transgene with (CUG)960 repeats in the DMPK 3UTR produce increased levels of CUGBP1, promote skeletal muscle wasting and heart conduction abnormalities, and die within 3 weeks of induction (Orengo et al., 2008) While these mice exhibit skeletal muscle and heart phenotypes, they do not recapitulate an adult onset progressive myopathy seen in DM patients. It is unclear how toxic (CUG)n repeats activate PKC and pr omote CUGBP1 stability. One possibility is that C(CUG)n repeats play a direct role in
24 activating PKC through an unknown mechanism. Alternatively, PKC activation could be a byproduct of acute damage to the heart and skeletal muscle by expression of a toxi c molecule in vivo The sequestration model proposes that toxic C(CUG)n RNAs inhibit cellular proteins by presenting novel binding sites which can inhibit their normal function through through dominant negative interactions. Interestingly, (CUG)n and (C CUG)n repeat expansions are predicted to form similar thermodynamically stable RNA hairpin structures which have been visualized by electron microscopy (Michalowski et al., 1999) As the repeats expand, (CUG)n and (CCUG)n dsRNA would potentially become more toxic by sequestering an increasing amount of cellular protein and leading to more severe disease. To support this hypothesis, an in vitro binding assay using increasing lengths of (CUG)n RNA demonstrated a lengthdependant interaction with a dsRNA binding protein, muscleblindlike 1 (MBNL1) (Miller et al., 2000) Human MBNL1 is homologous to the muscleblind gene that is responsible for terminal muscle and photoreceptor development in Drosophila (Pascual et al., 2006) Moreover, MBNL1 protein colocalizes with pathogenic (CUG)n and (CCUG)n in discrete nuclear f oci in DM1 and DM2 tissues. If inhibition of MBNL1 function by a dominant negative interaction with toxic (CUG)n and (CCUG)n repeat RNAs is responsible for disease (Fig. 14), then loss of Mbnl1 in a mouse model should faithfully recapitulate DM phenotypes. Remarkably, Mbnl1E3/E3 mice, which do not express Mbnl1 protein, phenocopy subcapsular cataracts, electrical myotonia, characteristic skeletal muscle pathology, and alternative splicing de fects seen in DM (Kanadia et al., 2003a) Therefore, loss of Mbnl1 alone in a mouse model, is sufficient to cause multiple DM adult onset
25 phenotypes. In an in vivo complementation assay, adenoassocia ted virus (AAV) mediated overexpression of myc tagged Mbnl1 in the skeletal muscle of the HSALR mouse model, expressing (CUG)250, rescued the myotonia and alternative splicing defects (Kanadia et al., 2006) This result provides further evidence that MBNL1 loss of function is the primary mechanism of adult onset disease While this data supports the idea that loss of MBNL1 function via sequestration by toxic C(CUG)n repeats causes multiple cardinal disease phenotypes in DM1 and DM2, there are many questions that have yet to be addressed. Multiple Genetic Contributions and the Complex Etiology of DM My otonic dystrophy may be the most variable human genetic disorder in terms of i ts clinical presentation. V ariably penetrant symptoms can occur from embryogenesis to late adult ons et with a wide variety of tissues affected. Germline expansion and somatic m osaicism of toxic C(CUG)n repeats likely contribute to the unique nature of this disease. Current data provides considerable evidence that MBNL1 function is compromised in DM and that this loss of function is the primary cause of many of the most characteristic symptoms of the ad ult onset disease. However, this and other models fail to recapitulate key features of disease, including CDM, progressive skeletal muscle wasting, and central nervous system (CNS) phenotypes. C(CUG)n expansions potentially affect many genes by altering local chromatin structure at the DNA level and acting in a dominant negative manner at the RNA level, activating and inhibiting multiple pathways. Like the MBNL1 loss of function model, disruption of a single gene by repeat expansions may be sufficient t o recapitulated unmodeled DM phenotypes including CDM and progressive skeletal muscle wasting Alternatively, combinatorial genetic models may be necessary to fully reconstr uct the onset of disease. The potential
26 contribution of multiple molecular pathw ays disrupted in DM makes determining the specific genetic contributions to disease a daunting task.
27 Figure 11. Noncoding microsatellite repeat expansions in myotonic dystrophy. (A) The DMPK gene with a (CTG)n expansion in the 3 UTR. DMPK exons (boxes) and introns (horizontal line) showing the position of (CTG)n repeats in the 3 UTR and expansion size correlating with disease (coding exons = black box; UTR = open box). (B) The CNBP gene with a (CCTG)n expansion. CNBP exons (boxes) and introns (horizontal line) showing the position of (CCTG)n repeats in the first intron and expansion size correlating with disease (coding exons = black box; UTR = open box).
28 Figure 12. Genomic DNA replication errors promote microsatellite repeat expansions and contractions. (CTG)n DNA expansions (red) form stable hairpin structures on the replicating strand by intramolecular base pairing and promotes slippage on both the leading and lagging strand. Okazaki fragments located entirely within the CTG repeats on the lagging strand can shift out of register, leaving an unpaired 5 end. Repair and replication promote further expansion.
29 Figure 13. Comparison of myotonic dystrophy symptoms. DM1 (red) and DM2 (green) share a unique and overlapping (dark yellow) clinical presentation. CDM (purple) displays distinct neonatal characteristics.
30 Figure 14. MBNL1 sequestration model for DM1. DMPK expression from the DM1 locus (including tightly linked genes DMWD and SIX5 ). Normal DMPK transcripts with (CTG)>37 are processed and exported through the nuclear pore complex (NPC) and translated in the cytoplasm. MBNL1 steady state levels remain diffusely nuclear. DM1 DMPK transcripts with expanded repeats, (CTG)exp, are processed, blocked from transport, and sequestered into r ibonucleoprotein (RNP) foci containing MBNL1. MBNL1 relocalization inhibits its normal function.
31 CHAPTER 2 MUSCLEBLIND PROTEINS REGULATE ALTERNATIVE SPLICING Introduction Introduction to Alternative S plicing Alternative splicing is a remarkable process that allows an organism to extend its protein diversity from a limited gene pool by selectively including or excluding information in mature mRNA, which enables finetuning of protein function by expressing iso forms that are adapted for speci fic physiological requirements (Jin et al., 2008; Licatalosi and Darnell, 2010) In one instance, the Down syndrome cell adhesion molecule ( D scam ) gene can potentially produce 38,016 protein isoforms (Schmucker et al., 2000) Unique Dscam isoforms expressed in each neuron promote repulsion of developing neurites from the same neuron by recognizing like isoforms. It has been estimated that a minimum of 4,752 isoforms are necessary to promote proper development of the Drosophila brain (Fuerst et al., 2009; Hattori et al., 2009; Matthews et al., 2007) Recent experimental approaches using deep sequencing of the entire human transcriptome reveals that nearly every gene is subjected to at least one alternative splicing event, highlighting the remarkable diversity that exi sts within the proteome itself (Pan et al., 2008; Pan et al., 2009; Wang et al., 2008) There are multiple patterns of alternative splicing that are capable of producing distinct mRNAs, including alternative promoters and untranslated regions ( UTRs ) mutually exclusive and cassette exons, retained introns, and cryptic 5 and 3 splice sites (Fig. 21). Each of these alternative splicing decisions is subject to multiple levels of regulation that are important for correctly identifying short exonic sequences within large premRNAs that ultimately govern inclusion or exclusion in the mature transcript
32 The bulk of pre mRNA splicing is catalyzed by the major (or U2 dependant) spliceosome, an RNA:protein complex composed of five small nuclear ribonucleoprotein particles (snRNPs) U1, U2, U4/U6/U5, and a variety of auxiliary protei ns (Black, 2003; Chen and Manley, 2009; W ahl et al., 2009) Although the spliceosome requir es ~145 dynamically interacting proteins (Rappsilber et al., 2002; Zhou et al., 2002) important for functions including the coupling of splicing to transcription and the complex rearrangements of the core component s, we will focus on the basal splicing machinery that is paramount for identifying prospective exons and completing the splicing reaction. Before splicing of pre mRNA can occur, the prospective exons/introns must first be defined within the primary transc ript. Constitutively spliced exons contain three consensus sequences that promote efficient recognition and recr uitment of the spliceosome Initially, U1 binds the canonical 5 splice site (CAG/guragu; uppercase = exon, lowercase = intron, r = any purine) through RNA RNA base pairing. Next, splicing factor 1 (SF1) binds to the branch point (YNCURAY; Y = any pyrimidine, R = any purine, N = any nucleotide), followed by U2 auxiliary factor 65 and 35 (U2AF65 and U2AF35) recruitment to the consensus polypyrim idine tract and the 3 splice site (u3y4unyac/G; lowercase = intron, uppercase = exon, y = any pyrimidine, n = any nucleotide), forming the E complex (Fig. 2 2). U2 interaction with the branch point through RNA RNA base pairing displaces SF1, forming the A complex, a premRNA in which exons and introns have been sufficiently defined to procede with assembly of the basal splicing machinery. Further, introduction of the U4/U6/U5 snRNP, replacing U2AF65 and U2AF35, establishes the B complex, which contains the necessary components to complete the s plicing reaction. Extensive ATP dependant RNA RNA
33 and proteinprotein rearrangement of the B complex leads to the formation of the C complex. In the catalytic C complex, the 5 splice site undergoes a nucleophilic attack from the juxtaposed 2 OH of the branch point adenosine, forming a 5 2 phosphodiester bond within the intron (forming the intron lariat). The 3 splice site subsequently undergoes a second nucleophilic attack from the newly formed 2OH of the 5 splice site, creating a 3 5 phosphodiester bond between the two spliced exons. Following the completion of splicing, the spliceosomal components are recycled and the 5 2 intron lariat is debranched and degraded. Alternatively spliced exons are included in mature mRNAs using the same mec hanism as described for constitutive exons, however, differences in the three defining cis elements used to identify these alternative exons are generally divergent from the canonical sequence. These nonconsensus sequences inefficiently recruit the splic eosome and thus require additional protein co factors to aid in the identification and definition of alternative exons. This subtle difference between constitutive and alternative exons allows for a layer of regulation that can be evolutionarily adjusted to promote tissue and developmental specific patterns of splicing by expressing proteins that assist in identifying or preventing recognition by the spliceosome. Two major families, the serine/arginine (SR) and heterogeneous ribonucleoproteins (hnRNP) pro teins, are intimately involved in these alternative splicing decisions. The SR family of proteins contain multiple copies of an N terminal RNA recognition motif (RRM) which generally recognize ssRNA motifs and C terminal arginine/serine residues (RS) that often promot e proteinprotein interactions (Long and Caceres, 2009; Shepard and Her tel, 2009) SR proteins recognize exonic and intronic splicing enhancer (ESE and
34 ISE) elements and assist in the defi nition and recruitment of the basal splicing machinery to promote inclusion of alternativel y spliced exons (Fig. 22) On the other hand, hnRNP proteins have been demonstrated to bind exonic and intronic splicing silencers (ESS and ISS) and inhibit interactions between alternatively spliced exons and the splicing machinery to promote exclusion of exons (Fig. 22) (He and Smith, 2009) Spatial and temporally restricted expression of these proteins can define a splicing environment, allowing cells to regulate the alternative splicing of genes involved similar pathways and functions (Chen and Manley, 2009) In contrast, misregulation of these factors can affect multiple downst ream splicing decisions in trans, resulting in a widespread misregulation of alternative splicing and disease (Jensen et al., 2009; Ward and Cooper, 2010) Alternative Splicing M isregulation in DM The major molecular defect associated with adult onset DM is the misregulation of alternative splicing for a specific subset of pre mRNAs (Orengo and Cooper, 2007) The absence of a global misregulation in alternative splicing suggests that the defective component is independent of the basal splicing machinery, unl ike the neuromuscular disorder Spinal Muscular Atrophy (SMA), which is caused by mutations in the survival of motor neuron (SMN) protein (Ogino and Wilson, 2004) SMN plays an integral part in snRNP biogenesis and mutations produce a global affect in splicing, manifesting primarily in motor neurons (Chari et al., 2009) Previous studies have indentified common misspliced genes in DM1 and DM2, including the insulin receptor ( I NS R ), cardiac troponin T ( TNNT2, formerly known as cTNT ), fast skeletal muscle troponin T ( TNNT3 ), and the muscle specific chloride channel ( CLCN1 ) (Charlet et al., 2002; Mankodi et al., 2002; Philips et al., 1998; Savkur et al., 2001; Savkur et al., 2004)
35 Interestingly, the predominant pattern of misregulation in DM is the retention of a fetal splicing pattern in adult tissues, indicating that DM symptoms may result from fetal isoform expression in adult tissues that fail to meet the necessary physiological requirements Myotonia, a cardinal characteristic of DM, results from the inability to relax skeletal muscle after a voluntary contraction. To initiate a muscle contraction, acetylcholine is released from the motor neuron, binding nicotinic acetylcholine receptors on the muscle side of the neuromuscular junction, and activating Na+ channels ( SCN4A ). Na+ influx into the muscle triggers an action potential, activating Ca++ channels and ini tiating contraction (Barchi, 1995) After the initial contraction, CLCN1, a skeletal muscle specific voltagegated chloride channel, is activated, allowing an influx if Clinto the muscle (Accardi and Pusch, 2000) The influx of Clreturns the membrane to its resting potential, preventing further contractions. Mutations in the muscle specific Na+ and Clchannels have been shown to cause myotonia in both human disease and animal models (Hudson et al., 1995; Jurkat Rott et al., 2002; Planells Cases and Jentsch, 2009) RT PCR analysis of DM1, DM2, HSALR, and Mbnl1E3/E3 mice adult skeletal muscle revealed missplicing of CLCN1 including the retention of intron 2 and intron 6, as well as inclusion of cryptic exons 7a and 8a (Charlet et al., 2002; Kanadia et al., 2003a; Mankodi et al., 2002) Inappropriate inclusion of these exons and introns insert premature terminat ion codons ( PTCs ) in the mature mRNA and promotes truncated isoforms, turnover of the transcript through nonsense mediated decay ( NMD ) and loss of CLCN1 protein in the muscle sarcolemma in DM, HSALR, and Mbnl1E3/E3 mouse models (Berg et al., 2004) While misregulation of the chloride channel alone is
3 6 sufficient to promote myotonia, previous reports have also implicated reduced activity of the Na+ channel SCN4A in DM1 myocytes and Dmpk/ mice (Benders et al., 1993; Mounsey et al., 2000; Reddy et al., 2002) However, there have been no documented changes in the expression of SCN4A in DM patients (Kimura et al., 2000) indicating that changes in Na+ conductance may be an indirect affect of DM myopathy. To test the hypothesi s that misregulation of the Clchannel alone is sufficient to cause myotonia in DM1, antisense morpholinos designed against the CLCN1 cryptic exon 7a 3 splice site (to inhibit exon 7a inclusion in the mature mRNA) were injected into the tibialis anterior (TA) muscle of the HSALR mouse model and assayed for recovery of myotonia (Wheeler et al., 2007) Injected HSALR mice displayed normal Clcn1 splicing patterns, full length protein correctly localized the sarcolemma, and a reversal of myotonia. This data provides evidence that missplicing of CLCN1 is the primary cause of myotonia in DM patients. Additionally, it lends support to the idea that missplicing events in DM directly promote disease phenotypes associated with disease. Another gene a ffected in DM, the I NS R is a tyrosine kinase receptor composed of two subunits and two subunits that localize to the plasma membrane and regulate glucose metabolism (Kahn and White, 1988) Interaction with its primary ligand, insulin, through the extracellular subunit causes autophosphorylation of the intracellular subu nits, promoting uptake of glucose, glycogen synthesis, and glycolysis Alternative splicing of I NS R exon 11 in the subunit results in two protein isoforms, IR A (exon 11 exclusion) and IR B (exon 11 inclusion) (Moller et al., 1989; Seino and Bell, 1989) which regulate the receptors sensitivity to insulin. The IR B isoform displays increased sensitivity to insulin and a higher signaling capacity (Kellerer et al., 1992; McClain,
37 1991; Vogt et al., 1991) Tissues involve d in maintaining glucose homeostasis primarily express IR B, including skeletal muscle, liver, and adipose tissue. In skeletal muscle of DM1 and DM2 patients, I NS R is misspliced, aberrantly excluding exon 11 which results in expression of the l ess sensiti ve isoform IR A DM patients also display insulin insensitivity in skeletal muscle, suggesting a direct link between I NS R missplicing and disease (Moxley et al., 1978; Moxley et al., 1984; Vialettes et al., 1986) Genes that regulate the protein machinery involved in generating muscle contractions are also misspliced in DM. Cardiac and skeletal muscle contractions are achieved through the regulated interaction of myosin with actin. During skeletal musc le contractions, myosin binds actin in an open state, releasing ADP and inorganic phosphate which promotes a conformational change in myosin to the closed state, resulting in force generation through the shortening of the sarcomere (Geeves and Holmes, 1999) Subsequent ATP hydrolysis returns myosin to the open position for ensu ing co ntractions During muscle relaxation, tropomyosin proteins negatively regulate skeletal and cardiac muscle contractions by occupying myosin binding sites on actin (Gunning et al., 2008) An increase in cellular Ca++ inhibits tropomyosin/actin binding through the trimeric troponin complex, allowing myosin/actin interactions and contraction (Ohtsuki and Morimoto, 2008) The troponin complex consists of three genes, troponin C, troponin I, and troponin T, encoded by TNNC2 TNNI2 and TNNT3 in fast skeletal muscle and TNNC1 TNNI3 and TNNT2 ( cTNT ) in cardiac muscl e Following the initiation of muscle contraction, troponin C binds increased cellular Ca++ promoting conformational changes in troponin T, which allosterically inhibits tropomyosin/actin binding and allow s myosin/actin interaction. cTNT and TNNT3 both
38 contain fetal exons (Townsend et al., 1994; Yuan et al., 2007) that are excluded postnatally during normal development, but are included in adult heart and skeletal muscle in DM. Although the role of cTNT and TNNT3 missplicing in DM is not as clear as CLCN1 and I NS R in disease, mutations in cTN T are associated with cardiac hypertrophy and sudden death (Moolman et al., 1997) Additionally, TNNT3 may be involved in overall skeletal muscle weakness in DM due to inefficient contractions. However, unlike CLCN1 in which exon 7a inclusion display s a direct involvement in myotonia, the missplicing of cTNT and TNNT3 most likely contribute to myopathy in conjunction with other misplicing events. In DM, sequestration and inhibition of an RNA binding protein, MBNL1, by toxic C(CUG)n repeat expansions result in characteristic missplicing of alternative exons that promote disease. MBNL1 was originally identified based on its ability to interact with double stranded (CUG)exp RNA in vitro and loss of Mbnl1 protein in a mouse model recapitulated alte rnative splicing defects in DM (Kanadia et al., 2003a; Miller et al., 2000) One possibility to explain these observations is that MBNL1 is an RNA binding protein involved in promoting normal adult splicing patterns by directly interacting with pre mRNA substrates and regulating the inclusion/ex clusion of alternatively spliced exons (Fig. 23). In DM, MBNL1 is sequestered away from its targets, resulting in the loss of adult specific regulation of alternatively spliced fetal exons and the expression of aberrant isoforms. Evolutionarily Conserved Muscleblind is Important for Muscle Development MBNL1 belongs to a family of proteins, muscleblind, conserved from Caenorhabditis elegans to humans that share a common RNA binding motif, containing three cysteine, and one histidine, residues (C3H) which are important for the
39 coordination of zinc ions and RNA protein interactions (Pascual et al., 2006) Muscleblind ( m bl ) was originally described in a screen designed to identify genes involved in photoreceptor differentiation and development in Drosophila Sev svp2 transgenic flies expressing seven up a hormone receptor transcription factor responsible for photoreceptor R3/4 and R1/6 differentiation, under the sevenless promoter which directs expression of photoreceptor R7, develop a rougheye phenotype due to the inability to correctly pattern photoreceptor subtype (Begemann et al., 1995; Hiromi et al., 1993) Sev svp2 flies were crossed to a collection of flies with UAS containing P elements and screened for gainof function modifiers of the rough eye phenotype. Mb l was identified as a dominant suppressor of the rougheye phenotype, suggesting a required role for terminal photoreceptor differentiation (Begemann et al., 1997) Additionally, m bl null mutants display larval lethality and skeletal muscle phenotypes, including disorganized Z bands and reduced extracellular matrix at muscle attachment sites (Artero et al., 1998) Conservation of muscleblind homologs may suggest that they share a conserved molecular function important for the proper development muscle. If this observation is true, muscleblind should regulate alternative splicing in during both Drosophila and mammalian development. Results The Drosophila m bl gene and its mammalian homolog, MBNL1 play a pivotal role in the normal development of skeletal muscle (Pascual et al., 2006) Inhibition of MBNL1 function via dominant negative interactions with toxic RNA contributes to the adult onset diseases DM1 and DM2, including myotonia and a defect in alternative splicing of a sp ecific subset of pre mRNAs Mbnl1E3/E3 mice, designed to recapitulate
40 loss of MBNL1 function in DM1 and DM2, faithfully phenocopy myotonia and alternative splicing defects (Kanadia et al., 2003a) However, it is not clear if MBNL1 is directly involved in alternative splicing defects. One possibility is that MBNL1 is an alternative splicing factor that directly regulates normal alternative splicing decisions during development by intera cting with those premRNAs misspliced in disease. In this scenario, loss of MBNL1 by sequestration would directly affect missplicing by titrating MBNL1 away from its normal premRNA substrates. Alternatively, loss of MBNL1 could promote downstream events that, in turn, result in disease phenotypes and missplicing. Therefore, we sought to test the hypothesis that MBNL1 directly regulates alternative splicing of premRNAs affected in DM1 and DM2. Muscleblind Protein Family: Evolutionary Conservation of St ructure and Function First, we explored the idea that muscleblind function is evolutionarily conserved in Drosophila and can regulate musclespecific alternative splicing. Mbl can produce four distinct isoforms (A D) varying at their C termini (Fig. 2 4 ), but the functional distinction between these isoforms is unclear. If Mbl isoforms are in fact alternative splicing factors, a reasonable assumption is that they are localized to the nucleus. However, while MBNL1 has been shown to be predominantly nucle ar, a lack of antibodies specific for Mbl isoforms has prevented similar analysis for the Drosophila homolog. Therefore, GFP tagged MblA, MblB, MblC, and MblD isoforms were expressed in COSM6 cells and assayed for cellular localization. MblB and MblC loc alized predominantly to the nucleus while MblA displayed a more cytoplasmic pattern (Fig 2 5 ). MblD was diffuse throughout the cell, possibly due to degradation (Fig 2 5 and Fig. 27C ). Interestingly, MblA and MblC also localized to cytoplasmic foci when exogenously expressed in
41 COSM6 cells, which were later identified as stress granules (Fig 2 5 and data not shown). Mbnl1 and Mbnl3 also localize to stress granules in COSM6 cells while Mbnl2 and MblB were not present in the cytoplasmic foci (data not shown). To further determine if m bl functional interactions are conserved, GFP tagged Mbl isoforms were coexpressed with (CUG)300 repeats in COSM6 cells and assayed for interactions with toxic repeat RNA. MblA, MblB, and MblC colocalized in discrete nuclear foci with repeat RNA (Fig. 2 6 ), similar to MBNL1, MBNL2, and MBNL3 (Fardaei et al., 2002) while MblD remained diffuse throughout the cell. These observations demonstrate that the interactions and relocalization with toxic RNA repeats is conserved within the muslceblind family. Dros ophila actinin is a gene that participates in the organization of the sarcomere and undergoes a developmental alternative splicing switch (Fyrberg et al., 1990; Roulier et al., 1992) To test the hypothesis that m b l regulates alternative splicing during development, we coexpressed an actinin mini gene (Fig. 27 A) with different GFP tagged Mbl isoforms in COSM6 cells and assayed for alternative splicing. MblB and MblC promoted exon 7 exclusion and adult muscle spli cing patterns, while MblA was less efficien t. MblD had no effect (Fig. 27 B). GFP Mbl expression was m onitored by immunoblot (Fig. 27 C). This result provides evidence that the developmental regulation of alternative splicing in muscle by the muscleblin d family and interaction with toxic (CUG)n repeats i s conserved from Drosophila to humans. MBNL1 Proteins Directly Regulate Alternative Splicing of G ene Transcripts Misregulated in Myotonic Dystrophy Many developmentally regulated genes are misspli ced in adult onset DM (Orengo and Cooper, 2007) Loss of Mbnl1 protein in the M bnl1E3/E3 mouse, model
42 reproduces the cardinal DM phenotypes, including missplicing events (Kanadia et al., 2003a) It is, however, unclear if MB NL1 is directly or indirectly involved in the alternative splicing of these genes. We sought to test the idea that MBNL1 directly regulates I NS R and cTNT two pre mRNAs misspliced in DM. I NS R and cTNT minigenes containing the alternatively splice d exons affected in DM (Fig. 2 8 ) where cotransfected into COSM6 cells with GFP MBNL or siRNA specific for MBNL1 and assayed for changes in alternative splicing by RT PCR. Exogenous expression of GFP tagged MBNL1, MBNL2, and MBNL3 promoted exclusion of cTNT fetal exon 5 and inclusion of I NS R exon 11, two premRNAs misspliced in DM, in a minigene assay in cell culture (Fig 2 9 A,B). Alternative splicing of a neuronal specific exon in ClaLC a gene not affected in DM, was not influenced by MBNL expression (Fi g. 2 9 C). GFP MBNL transgene expression was monitored by immunoblot analysis. GFP alone did not alter splicing. Interestingly, all three MBNL family members promote similar splicing patterns in vitro This observation likely reflects the high degree of sequence identity shared by the MBNL family (Pascual et al., 2006) If exogenous expression of MBNL1 promotes a splicing change in cTNT and I N S R then depletion of endogenous MBNL1 should recapitulate the DM splicing pattern. Two MBNL1 specific siRNAs, directed against the coding sequence were used to knock down MBNL1 expression >80 90% in HeL a cells (Fig. 2 10A,B and data not shown). Loss of endogenous MBNL1, via siRNA mediated knockdown, promoted the retention of cTNT fetal exon 5 and inclusion of I NS R exon 11 (Fig. 210C), reproducing the missplicing pattern in DM. ClaLC was unaffected. siRNA directed against GFP was used as a knock down co ntrol and had no
43 affect on splicing. These data verifies that MBNL1 specifically regulates alternative of exons misspliced in DM. MBNL1 Interacts Directly with CisElements in cTNT pre mRNA To determine if MBNL1 interacts directly with cTNT pre mRNA to regulate its splicing, an in vitro binding assay was performed using the alternatively spliced cTNT fetal exon 5 minigene which contains the upstream and downstream introns. Briefly, 32P labeled full length and truncated RNAs were incubated w ith recombinant GST MBNL1, crosslinked with UV light, and resolved RNA:protein complexes by SDS PAGE. RNA:protein interactions were visualized by label transfer. GST MBNL1 was bound to a 41 nucleotide region directly upstream of the alternatively spliced exon (Fig. 211). Scanning mutagenesis further refined the MBNL1 binding site to two 8 nucleotide regions 18 to 26 and 36 to 44 upstream of exon 5 (Fig. 211). Two dinucleotide substitutions that eliminated MBNL1 binding (data not shown), but minimi zed disruption of the intron/basal splicing machinery interactions, were used in subsequent experiments to assay for splicing responsiveness in vitro (Fig. 2 11). We next sought to determine if loss of the MBNL1 interacting cis element in cTNT exon 5 inhi bited MBNL1 responsiveness in a splicing assay. Coexpression of GFP tagged MBNL1, MBNL2, and MBNL3 and the wild type cTNT minigene promoted exon 5 exclusion in HeL a cells (Fig. 2 12). However, coexpression with the mutant cTNT minigene demonstrated a los s of responsiveness f or MBNL overexpression (Fig. 2 12). GFP MBNL transgene expression was monitored by immunoblot analysis and GFP alone did not alter splicing. Alignment of human cTNT and chicken cTNT MBNL1 binding sites (chicken cTNT experiments not s hown) revealed a common YGCU(U/G)Y RNA motif for MBNL1 bi nding (Fig. 213). This data suggests that MBNL1 directly binds
44 to cTNT and IR pre mRNAs and loss of this protein due to sequestration affects the alternative splicing of these genes. (CUG)n and (CAG)n Repeats Relocalize MBNL1 but only (CUG)n Alter cTNT Splicing In the RNA dominance model for DM, loss of MBNL1 function by toxic C(CUG)n RNA sequestration into nucl ear foci results in disease. However, there are multiple trinucleotide repeat expansion diseases, the majority of which are coding region (CAG)n expansions, which fail to present DM specific manifestations (Shao and Diamond, 2007) This observation led us to investigate the idea that MBNL1 sequestration is specifically dependent on C(CUG)n repeat RNAs. To address this question, a DMPK minigene construct with (CTG)960, (CAG)960, or 0 repeats (F ig. 2 14 A) was transfected into COSM6 cells and assayed for repeat RNA and endogenous MBNL1 colocalization in nuclear foci. As expected, DMPK minigene expression with 0 repeats did not induce RNA foci and MBNL1 was localized diffusely t hroughout the nucleus (Fig. 2 14B). Suprisingly, MBNL1 colocalized with both (CUG)960 and (CAG)960 RNAs in nuclear foci (Fig. 2 14B). If (CAG)960 repeats are capable of sequestering MBNL1, do they also alter pre mRNA splicing of genes affected in DM? Increasing amounts of (CUG)960, (CAG)960, and control (0 repeats) plasmids were coexpressed with the cTNT minigene to assay for misregulation of alternative splicing. While (CUG)960 promoted cTNT exon 5 inclusion at only 0.1 g of transfected plasmid, increasing amounts of (C AG)960 had only a minimal affect (Fig. 215). There were no statistical differences in RNA expression or in the number of foci produced from either repeat plasmid in cell culture (data not shown). Interestingly, MBNL1 relocalization into nuclear foci was not mutually exclusive with the misregulation of cTNT alternative splicing. This suggests that while
45 MBNL1 interacts with both (CUG)960 and (CAG)960 RNA, the nature of these interactions may be distinct. The Stability of MBNL1 :RNA complexes varies between (CUG)n, (CCUG)n and (CAG)n Repeats One possibility to explain the discrepancy between MBNL1 sequestration with (CUG)960 and (CAG)960 repeats and missregulation of alternative splicing is that both RNAs are capable of relocalizing nuclear MBNL1, but only (CUG)n repeats trap MBNL1 in foci resulting in their functional depletion. To test this hypothesis, 32P labeled (CUG)54, (CCUG)46, (CAG)54, and T5.45 (an endogenous MBNL1 binding site from TNNT3 intron 8, (Yuan et al., 2007) ) RNAs were incubated with COSM6 whole cell lysate expressing myc tagged MBNL1 in the presence of ATP, incubated for 30 minutes, UV crosslinked, RNase treated, and immunoprecipitated using an antibody speci fic for myc. Labeled RNAs were either challenged with 2000X fold excess of unlabeled RNA concurrently or 15 minutes following the initial labeled RNA:myc MBNL1 incubation period. An unchallenged control was included to assay for the baseline level of RNA :protein crosslinking (Fig. 2 16A). Immunoprecipitated complexes were resolved by SDS PAGE and autoradiographs of RNA:protein complexes were quantified using a phosphoimager. Myc MBNL1 protein retained all of the (CUG)54 RNA when challenged with 2000X fold excess unlabeled competitor, while (CAG)54, (CCUG)46, and T5.45 were less resistant to competition (Fig. 216B,C). MBNL1:(CUG)54 interactions, once formed, effectively trapped MBNL1 protein and did not allow disassociation of the RNA:protein complex in 15 minutes. Interestingly, (CCUG)46, a tetranucleotide repeat expansion that causes DM2, does not trap MBNL1 more effectively than nonpathogenic RNAs (CAG)54 and T5.45. While all repeats tested are capable of relocalizing MBNL1 by interacting
46 with the protein, only (CUG)n repeats trap MBNL1 in vitro Static MBNL1:(CUG)54 interactions suggest that MBNL1 may form different complexes on (CUG)n repeats when compared to the more dynamic interactions with (CCUG)46, (CAG)54, and T5.45. This result predicts t hat (CUG)n repeat expansions are the most toxic RNAs in the sequestrati on model of myotonic dystrophy. Discussion Myotonic dystrophy is a neuromuscular disease caused by two microsatellite repeat expansions in unrelated genes that present with simil ar clin ical manifestations Related C(CUG)n toxic RNA expansions in DM1 and DM2 have been shown to compromise the function of MBNL1, an evolutionarily conserved protein family important for the terminal skeletal muscle development (Artero et al., 1998) Disruption of Mbnl1 in a mouse model mimicking a loss of function in disease, results in myopathy that phenocopies the primary symptoms of adult onset DM and presents with characteristic molecular chang es in alternative splicing (Kanadia et al., 2003a) Previous studies have demonstrated that the transcripts of numerous genes are misspliced in DM, including inappropriate exon inclusion in cTNT and exclusion in I NS R (Philips et al., 1998; Savkur et al., 2001; Savkur et al ., 2004) These misplicing events are particularly interesting because of their correlation to cardiomyopathy and endocrine abnormalities in DM (Belfiore et al., 2009; Fiset and Giles, 2008) However, it is unclear how loss of MBNL1 function contributes to alternative splicing defects in disease. In this study, we sought to test the hypothesis that muscleblind directly interacts with premRNAs to regulate alternative spl icing decisions and (CUG)n RNA expansions inhibit this function.
47 Muscleblind Proteins Directly Regulate Alternative Splicing Drosophila mbl was originally identified in a genetic screen designed to identify modifiers of a rough eye phenotype. Interestingly, mutant flies that fail to express mbl develop severe embryonic muscle defects, including incorrect organization of the sarcomeric Z line (Artero et al., 1998) The Z line component, actinin undergoes normal tissue and developmental alternative splicing transitions, including larval muscle, adult muscle, and nonmuscle isoforms (Roulier et al., 1992) If the muscleblind family regulates developmental alternative splicing decisions, then it is reasonable to think that mbl promotes adult splicing patterns in Drosophila. Ectopic expression of Mbl isoforms A D, which share the 63 N terminal amino acids and 2XC3H RNA binding motifs (Fig. 2 1), in COSM6 cells with an actinin minigene reporter (Fig. 24A) revealed that MblB and MblC preferentially promoted the adult muscle isoform (i.e. larval muscle exon 7 exclusion). Not surprisingly, Mbl isoforms B and C, which localized predominately to the nucleus, demonstrated more efficient exclusion of actinin larval exon 7 than Mbl A, which generally localized to the cytoplasm (Fig. 22). MblD, which contains only the N terminal 84 amino acids, was diffusely distributed throughout the cell (Fig. 22 and 23) and was not stable when assayed for expression by immunoblot analysis (Fig. 24C). This observation may be a byproduct of expressing this protein in a mammalian cell. Interestingly, MBNL1 also shifts splicing of the actinin minigene in vitro, suggesting an overlap in function. Subsequent studies have also shown that mbl regulates adult specific alternative splicing of Drosophila troponin T, a homolog of TNNT2 and TNNT3 genes kn own to be misspliced in DM Additionally, Mbl isoforms A C demonstrated an i nteraction with coexpressed (CUG)300 RNA repeats in vitro, demonstrating an overlap
48 in RNA binding (Fig. 2 3) substrates with MBNL1. Moreover, transgenic flies expressing MBNL1 also block the embryonic lethality associated with mbl mutant flies (Monferrer and Artero, 2006) This data provides evidence that mbl and MBNL1 share a conserved function in promoting adult specific splicing patterns. However, these results do not make the distinction between a mechanism for Mbl promoting exclusion of alternative exons by directly interacting with the premRNA or via an Mbl mediated downstream event. Recent studies using alternative splicing microarray analysis of HSALR and Mbnl1E3/E3 mice reveal a >80% overlap in missplicing events between the two DM models, indicating that loss of MBNL1 function in disease is the prim ary cause of splicing defects. To test the idea that MBNL1 is responsible for regulating these splicing decisions by directly binding the premRNAs affected in DM and promoting adult splicing patterns, we employed in vitro splicing and crosslinking assays GFP tagged MBNL1, MBNL2, and MBNL3 promote adult skeletal muscle splicing patterns in cTNT (exon 5 inclusion) and I NS R (exon 11 exclusion) minigenes, while knockdown of endogenous MBNL1 in cell culture phenocopies aberrant DM splicing. Interestingly, all three MBNL family members promoted similar splicing patterns in vitro, indicating a potential overlap in function, which was also observed in the Mbl alternative splicing analysis. However, Mbnl1E3/E3 mice and endogenous MBNL1 knockdowns reproduce DM splicing abnormalities, suggesting that other family members do not compensate this function in vitro or in vivo The most likely explanation is that the MBNL family shares >90% amino acid identity while Mbl shares 100% identity in their respective C3H m otifs, allowing dominant interactions with MBNL1 binding sites when the proteins are
49 exogenously overexpressed. More importantly, this data demonstrates that MBNL1 can regulate cTNT and I NS R alternatively spliced exons in vitro, recapitulating disease spl icing patterns and allowing for further analysis of the cis elements responsible for mediating these decisions. Using UV crosslinking and mutagenesis of the cTNT alternatively spliced exon 5 and adjacent introns, we identified two MBNL1 binding sites imme diately upstream of exon 5. Loss of either binding site through mutagenesis inhibited binding, suggesting that both are necessary for MBNL1 interaction with the pre mRNA. In addition, two dinucleotide substitutions in MBNL1 binding sites repressed MBNL1 responsiveness in a cTNT minigene alternative splicing assay, demonstrating that inhibition of binding was sufficient to alleviate MBNL1 alternative splicing control of exon 5. Alignment of the human cTNT and conserved chicken cTNT MBNL1 binding sites rev ealed a common MBNL1 binding motif, YGCU(U/G)Y, that is responsive to MBNL1 in vitro Other groups have verified additional MBNL1 binding sites using multiple approaches, including bioinformatics, systematic evolution of ligands by exponential enrichment (SELEX), and X ray crystallography, which when compared with each other unmask a common YGCY sequence motif for MBN L1 binding (Du et al., 2010; Goers et al., 2010; Teplova and Patel, 2008) Interestingly, when Mbnl1 binding sites were mapped to mispliced exons in both HSALR and Mbnl1E3/E3 mice, YGCY sequence motifs were enriched upstrea m of exons included in disease (normally excluded in adult tissues) and downstream of exons excluded in disease (normally included in adult tissues). This pattern is similar to alternative splicing factors Nova and Fox2, in which interaction with sequence motifs upstream of the alternatively spliced exon promotes exclusion and interaction with sequences downstream promote inclusion
50 (Licatalosi et al., 2008; Yeo et al., 2009) These observations suggest that these factors affect splicing decisions of alternative exons using a com mon mechanism. Although the MBNL1 sequence motif was not mapped to I NS R pre mRNA, this observation implies that the interaction would lie downstream of exon 11 and promote inclusion. How MBNL1 promotes exon inclusion is unclear. One possibility is that MBNL1 binds downstream of exon 11, altering the RNA structure of the premRNA, and promoting favorable U1 snRNP binding at the exon 11 5 splice site. Alternatively, MBNL1 could compete with other negative regulators near the 5 splice site or could poten tially assist in actively recruiting U1. These regulatory mechanisms are not mutually exclusive. The role of MBNL1 in exon inclusion may be a little more straight forward. MBNL1 interactions with cTNT exon 5 through the 3 slice site would potentially i nhibit the binding of the basal splicing machinery, more specifically U2AF65/35 or SF1/U2 snRNA, and prevent the assembly of the spliceosome, masking the alternative exon from inclusion. In support of this, Berglund and coauthors demonstrated the MBNL1 bi nds a dsRNA structure directly upstream of cTNT exon 5 and prevents the ssRNA binding protein U2AF65 from interacting with the polypyrimidine tract, therefore inhibiting its ability to recruit U2 (Warf and Berglund, 2007) This mechanism is also likely responsible for promoting fetal exon exclusion of Tnnt3 in which MBNL1 binds a se condary structure directly upstream of the alternatively spliced exon (Yuan et al., 2007) This data provides evidence that MBNL1 is an alternative splicing factor responsible for modulating exon inclusion/exclusion by directly binding to sequences in the premRNA. Loss of MBNL1 through sequestration would therefore remove this layer of regulation and promote missplicing of premRNA targets affected in DM.
51 Microsatellite Repeat Expansions Display Variable MBNL1 Stability and Toxicity Significant evidence supports a MBNL1 loss of func tion model underling DM pathogenesis, in which toxic RNAs that contain similar sequence motifs (YGCY) to normal MBNL1 substrates titrate the protein into insoluble complexes, inhibiting its activity. Colocalization of (CUG)exp RNA and MBNL1 protein in di screte nuclear foci has been observed in DM1 patient cells and tissues, HSALR mice, as well as cell culture in which repeat expansions are ectopically expressed (Kanadia et al., 2006; Mankodi et al., 2003; Mankodi et al., 2001) However, in an unexpected result, exogenously expressed (CAG)960 repeats also relocalized endogenous MBNL1 in nuclear foci in COSM6 cells. Although (CAG)960 RNA is predicted to form a thermodynamically stable dsRNA hairpin, it would not be predicted to interact with MBNL1 based on the deviation of its primary sequence from the consensus YGCY binding site in which purines have replaced pyrimidines. Despite the ability of (CAG)960 to recruit MBNL1 into foci like (CUG)960 repeats, the (CAG)960 RNA expansion failed to appreciably alter splicing of exon 5 in cTNT minigene cotransfection experiments This result is particularly intriguing considering that (CUG)54 and (CAG)54 shared similar affinities (MBNL1:CUG54 Kd=5.3 0.6 nM and MBNL1:CAG54 Kd=11.2 1.5 nM) in a filter binding experiments (Yuan et al., 2007) although this assay does not take into account variable RNA:protein binding modes. Interestingly, MBNL1 proteins form an oligomeric r ing structure, with a prominent central hole that is large enough to accommodate dsRNA (Yuan et al., 2007) These structures are visible in an electron micrograph and show that MB NL1 stacks multiple rings on a (CUG)136 dsRNA. However, these experiments have not been conducted with (CAG)n repeats. One explanation for this observation is that while MBNL1 interacts with both repeats, the nature of the RNA:protein interactions are
52 i nherently different. Upon closer inspection, MBNL1:(CAG)960 foci appeared larger and less compact than MBNL1:(CUG)960 foci when visualized by RNA fluorescent in situ hybridization (FISH). This difference in appearance may be reflective of divergent molecular interactions. Endogenous MBNL1 targets and (CUG)n repeats share similar characteristics, namely dsRNA substrates that contain paired GC dinucleotides with unpaired pyrimidine bulges. While it is likely that MBNL1 recognizes basepaired GC dinucleotides in (CAG)960, the adenosine mismatches may prevent MBNL1 from stably interacting via its ring structure, with the repeats like it does with (CUG)960. In other words, MBNL1:CA G interactions may consist of monomer binding, while MBNL1:CUG repeats are locked in position through oligomeric binding and the subsequent stacking of MBNL1 rings. If this model is correct, then multiple mismatches in the DM2 (CCUG)n repeat could also potentially disrupt stable MBNL1 binding. To investigate these interactions, we performed competition assays in which MBNL1:RNA structures were allowed to form under splicing conditions and then challenged with 2000X fold excess competitor to assay for RN P stability. (CUG)54 demonstrated highly stable MBNL1 interactions, while (CCUG)46,(CAG)54, and T5.45 (an endogenous MBNL1 binding site responsible for fetal exon exclusion in TNNT3 ) displayed less stable binding. This result provides additional evidence that MBNL1 interactions with microsatellite repeats are not the same. However, in this assay, MBNL1:(CCUG)46 is even less stable than T5.45 and (CAG)54. Unlike T5.45 and (CAG)54, (CCUG)n repeats can cause DM symptoms and missplicing events. How do less stable (CCUG)n repeats promote disease while (CAG)n repeats do not? The answer may be a combination of variables that are necessary for
53 (CCUG)n to promote the onset of disease. The largest CCTG expansion in CNBP has been reported to be ~11,000 repeats, making the DM2 expansion ~3 times larger than the DM1 expansion (Ranum and Day, 2002) Although the CCUG repeat displays less stable MBNL1 interactions than CUG, the increase in repeat length may be sufficient to sequester enough MBNL1 to promotes onset of the disease. Additionally, CNBP expression levels have been estimated to be higher than that of DMPK providing more toxic molecules to relocalize MBNL1 (Mankodi et al., 2003) Despite larger repeats and higher expression levels, the later age of onset and generally more mild symptoms characteristic of DM2 may be reflective of reduced CCUG repeat toxicity. In conclusion, the dysregulation of alternative splicing has become a molecular hallmark of DM, affecting multiple cellular pathways and contributing to the clinical manifestations of disease. Our results provide evidence that MBNL1, a protein whose function is impaired by a gainof function at the RNA level in DM, i s an alternative splicing factor that promotes adult specific splicing by directly interacting with premRNA. Additionally, MBNL1 displays variable stabilities when interacting with RNA repeat expansions. Underappreciated differences between (CUG)exp and (CCUG)exp RNAs and their interacting factors could potentially contribute to the immense variability of symptoms seen in DM patients.
54 Figure 21. Patterns of alternative splicing. Alternatively spliced (red) and mutually exclusive (green) exons spliced into constitutive exons (open boxes) produce distinct mature mRNAs via (A) alternative promoters and (B) 3 exons, (C) mutually exclusive and (D) cassette exons, (E) intron retention, and (F) cryptic 5 and (G) 3 splice sites.
55 Figure 22. Sch ematic of constitutive and alternative premRNA splicing mechanism. Constitutively spliced exons (open boxes) contain strong consensus 5 (CAGGU) and 3 (YAGG) splice sites that interact with snRNPs (yellow) and accessory proteins (blue) of the basal spli cing machinery to define exons for splicing in the premRNA. U1 snRNP (5 splice site), SF1 (branch point adenosine), U2AF65 (polypyrimidine tract), and U2AF35 (3 splice site) initially bind the premRNA (E complex), followed by U2 substitution for SF1 ( A complex), introduction of the tri snRNP (U4/U6/U5), and ATP dependant rearrangements of the splic eosome (B comlex). Further ATP dependant sp liceosomal rearrangements, including disassociation of U4 and U1, are required to form the catalytically active C complex to promote transe sterification reactions, join the exons and remove the intron lariat. Alternative spliced exons (grey boxes) often contain weak consensus 5 (GU) and 3 (AG) splice sites and smaller, less defined polypyrimidine tracts. Therefo re, additional proteins are required to promote or inhibit the assembly of the sliceosome. SR proteins (green) interact with exonic splicing enhancers to define an alternatively spliced exon and assist in the recruitment of the sliceosome. Conversely, hnRNP proteins (e.g. hnRNP A1 red) interact with exonic splicing silencers and polypyrimidine tract binding protein (PTB orange) binds the polypyrimidine tract to block assembly of the spliceosome and prevent inclusion of the exon in the mature mRNA.
56 Figure 23. Model of MBNL1 sequestration promoting alternative splicing defects in DM1. In unaffected individuals, DMPK (CUG)5 37 mRNA does not interact with MBNL1, allowing MBNL1 to directly bind premRNA substrates (exons = open boxes, introns = horizontal lines) and promote fetal exon (FE) exclusion (green). In DM1 affected individuals, DMPK (CUG)>50 sequester MBNL1, inhibiting MBNL1 interaction and promoting exon inclusion (red).
57 Figure 24 Alternative splicing produces four distinct Mbl isoforms in Drosophila. All isoforms share the N terminal 63 amino acids (blue) which contain 2XC3H RNA binding motifs (cysteine and hi stidine positions are indicated by vertical white lines). MblA, MblB, MblC share an additional 116 amino acids (red). All isoforms have unique C termini. MblB contains arginine (N), alanine (A), and phenylalanine (F) rich regions. MblC has a putative (*) sumoylation site. Reproduced from Muscleblind isoforms are functionally distinct and regulate actinin splicing ; Vicente M Monferrer L Poulos MG Houseley J Monckton DG ODell KM Swanson MS and Artero RD; Copyright 2007, with permission from Jo hn Wiley & Sons
58 Figure 25 Mbl protein isoforms display different subcellular localizations. GFP tagged Mbl isoforms were exogenously expressed in COSM6 cells. MblA appears cytoplasmic and nuclear while MblB and MblC are predominantly nuclear. Mb l D is diffuse throughout the cell. DAPI stain indicates nuclear location. Reproduced from Muscleblind isoforms are functionally distinct and regulate actinin splicing ; Vicente M Monferrer L Poulos MG Houseley J Monckton DG ODell KM Swanson MS and Artero RD; Copyright 2007, with permission from John Wiley & Sons
59 Figure 26 Mbl colocalizes with (CUG)300 RNA in nuclear foci. GFP tagged Mbl isoforms were cotransfected with (CUG)300 in COSM6 cells. MblA, MblB, and MblC (green) colocalize with (CUG)300 (red labeled with Cy3CAG10 oligonucleotide probe) in discrete nuclear foci. DAPI indicates nuclear location. Reproduced from Muscleblind isoforms are functionally distinct and regulate actinin splicing ; Vicente M Monferrer L Poulos M G Houseley J Monckton DG ODell KM Swanson MS and Artero RD; Copyright 2007, with permission from John Wiley & Sons
60 Figure 27 Drosophila Mbl regulates alternative splicing of actinin An actinin minigene was coexpressed with EGFP tagged Mbl isoforms and assayed for alternative splicing by RT PCR with primers positioned in constitutive exons 5 and 9. (A) The actinin minigene contains exons 59 (exons = open boxes; introns = horizontal lines). actinin is alternatively spliced to produce adult muscle (blue), larval muscle (green), and non muscle (red) isoforms. (B) Mbl isoforms A, B, and C, but not D, promote an adult muscle splicing pattern. EGFP is used as a control. Because actinin adult muscle and nonmuscle spliced mRNAs are the same size, a unique Sac I site is used to digest and resolve the resulting RT PCR products. (C) An immunoblot using an EGFP specific antibody was used to control for EGFP Mbl expression (expected sizes: MblA = 4 9 kDa, MblB = 58 kDa, MblC = 53 kDa, MblD = 35 kDa). Ponse a u S staining was used to control for loading (data not shown).
61 Figure 28 Schematic of cTNT and I NS R alternative splicing minigene reporters. (A) cTNT minigene. A 730 nucleotide fragment including alternatively spliced human cTNT exon 5 (30 nucleotides) and adjacent upstream/downstream introns (red) cloned in between constitutively spliced exons 2 and 4 from the Gallus gallus TNNI2 gene (black) Primers used for RT PCR analysis are located in exons 2 and 4. (B) I NS R minigene. A fragment from the human I NS R locus containing alternatively spliced exon 11 and constitutive exons 10 and 12. Primers used for RT PCR analysis are located in exons 10 and 12. Fetal/DM and adult splicing patterns are indicated by dashed lines. Minigenes are not drawn to scale.
62 Figure 29 MBNL1, MBNL2, and MBNL3 regulate alternative splicing of cTNT and I NS R cTNT and I NS R minigenes were coexpressed with GFP tagged MBNL proteins in primary chicken myoblasts and assayed for inclusion of alternatively spliced exons by RT PCR with primers positioned in constitutive exons. Bands were quantified by phosphoimager and exon inclus ion was calculated as: [(exon inclusion)/(exon inclusion + exon exclusion) X100]. Transgene expression was monitored by immunoblot with an antibody specific for GFP. MBNL1, MBNL2, and MBNL3 promotes (A) fetal exon 5 exclusion of cTNT and (B) exon 11 inc lusion of I NS R minigenes while (C) a neuronal specific exon is unresponsive in the ClaLC minigene. Reproduced from Muscleblind proteins regulated alternative splicing. Ho TH Charlet B Poulos MG Singh G, Swanson MS and Cooper TA ; EMBO Vol. 23, No 15, 31033112. Copyright 200 4 with permission from The Nature Publishing Group.
63 Figure 210. Endogenous MBNL1 regulates alternative splicing of cTNT and I NS R minigenes. cTNT and I NS R minigenes were cotransfected with siRNA directed against the MBNL1 coding sequence (THH31 and THH2) and assayed for alternative splicing by RT PCR (described in Figure 16). (A) siRNA knocks down endogenous MBNL1 41/42 kDa isoform express ion. Immunoblot analysis of HeLa cells treated with THH31 and THH2 using antibodi es recognizing MBNL1 and GAPDH (loading control) or (B) immunocytochemistry using antibodies recognizing MBNL1. A nonspecific GFP siRNA has no affect on MBNL1 expression. (C) Loss of endogenous MBNL1 recapitulates a DM splicing pattern in cTNT and I NS R minigenes. Reproduced from Muscleblind proteins regulated alternative splicing. Ho TH, Charlet B Poulos MG Singh G, Swanson MS and Cooper TA ; EMBO Vol. 23, No 15, 31033112. Copyright 2004 with permission from The Nature Publishing Group.
64 Figure 2 11. MBNL1 binds upstream of alternatively spliced cTNT exon 5. cTNT RNA from the alternatively spliced exon 5, with upstream and downstream introns, were uniformly body labeled with 32P and incubated with recombinant GST MBNL1, UV crosslinked and resolved on an nondenaturing acrylamide gel to assay for binding. Truncation and scanning mutagenesis of the RNA was performed to identify the MBNL1 binding site. (+) and ( ) indicate GST MB NL1 and RNA binding. The putative branch point adenosine (black circle) and RNA mutations (lowercase) are denoted. MBNL1 (two vertical lines) and antagonist CUGBP1 (grey box) binding sites are indicated on the cTNT RNA. Two dinucleotide substitutions (RN A M), which eliminate GST MBNL1 binding, will be used for in vitro splicing assays. Reproduced from Muscleblind proteins regulated alternative splicing. Ho TH, Charlet B Poulos MG Singh G, Swanson MS and Cooper TA ; EMBO Vol. 23, No 15, 31033112. Co pyright 2004 with permission from The Nature Publishing Group.
65 Figure 212. MBNL1 binding site mutations inhibit MBNL1, MBNL2, and MBNL3 responsiveness. Wildtype or mutant (RNA M in Figure 18) cTNT minigenes were cotransfected with GFP tagged MBNL 1, MBNL2, or MBNL3 and assayed for alternative splicing by RT PCR (described in Figure 16). Transgene expression was monitored by immunoblot using an antibody specific for GFP. Reproduced from Muscleblind proteins regulated alternative splicing Ho TH, Charlet B Poulos MG Singh G, Swanson MS and Cooper TA ; EMBO Vol. 23, No 15, 31033112. Copyright 200 4 with permission from The Nature Publishing Group.
66 Figure 213. Alignment of the human and chicken cTNT MBNL1 binding sites reveals a conserved motif. Reproduced from Muscleblind proteins regulated alternative splicing Ho TH, Charlet B Poulos MG Singh G, Swanson MS and Cooper TA ; EMBO Vol. 23, No 15, 31033112. Copyright 200 4 with permission from The Nature Publishing Group.
67 Figure 2 14. MBNL1 colocalizes with (CUG)960 and (CAG)960 RNA in nuclear foci. (A) DMPK minigene containing a CMV promoter and DMPK exons 1115, with 960 interrupted CUG or CAG repeats in the 3UTR. (B) DMPK (CUG)960 or (CAG)960 exogenously expressed in COSM6 cells. Immunocytochemistry with an antibody specific for MBNL1 (green) demonstrates colocalization with both CUG960 and CAG960 RNA (red labeled with Cy5 CAG10 or Cy5 CUG10, respectively). DAPI indicates nuclear location. Scale bar = 10 m. Reproduced from Colocalization of muscleblind with RNA foci is seperable from mis regulation of alternative splicing in myotonic dystrophy Ho TH Savku RS Poulos MG Mancini MA, Swanson MS and Cooper TA ; Journal of Cell Science 118, 29232933. Copyright 2005, with permission from The Company of Biologists Ltd.
68 Figure 215. CUG960, but not CAG960, expression alters cTNT alternative splicing. A cTNT minigene, containing an alternatively spliced fetal exon 5 which is misregulated in DM, is co transfected with either DMPK (CUG)960 or DMPK (CAG)960 and assayed for exon 5 inclusion using RT PCR and primers positioned in constitutive exons. Three independent transfections were done and exon inclusion was calculated as: (exon inclusion)/(exon incl usion + exon exclusion). Increasing amounts of (CUG)960 and (CUG)960 do not significantly alter cTNT splicing. Reproduced from Colocalization of muscleblind with RNA foci is seperable from mis regulation of alternative splicing in myotonic dystrophy Ho TH Savku RS Poulos MG Mancini MA, Swanson MS and Cooper TA ; Journal of Cell Science 118, 29232933. Copyright 2005, with permission from The Company of Biologists Ltd.
69 Figure 216. MBNL1 displays increased stability with (CUG)54 RNA. (A) Schematic of the competition assay. MBNL1myc whole cell lysate (expressed in COSM6 cells) is incubated with 32P labeled RNA (red) alone for 30 minutes (A), concurrently with 2000X fold excess cold RNA (blue) for 30 minutes (B), or alone for 15 minutes and then challenged with 2000X fold excess col d RNA for 15 minutes (C). RNA:protein complexes are UV crosslinked, RNase digested, immunoprecipitated with an antibody specific for myc, and resolved by SDS PAGE. (B) Autoradiographs of RNAs used in the competition assay: (CUG)54; T5.45 (an endogenous MBNL1 binding site on the TNNT3 mRNA responsible for alternative splicing); (CCUG)46; (CAG)54. (C) A representative bar graph indicating the percentage of 32P labeled RNA retained when challenged. Bands were quantified by phosphoimager and retention was calculated as (C/A X100) from the autoradiograph.
70 CHAPTER 3 LOSS OF MBNL3 4XC3H ISOFORMS ARE NOT SUF FICIENT TO MODEL CONGENITAL MYOTONIC DYSTROPHY Introduction Overview of Congenital Myotonic Dystrophy Congenital myotonic dystrophy is perhaps the most severe form of DM1, although the disease likely manifests itself during embryonic development (Harper, 2001) Like the adult onset disease, CDM affects a wide variety of tissues and displays variable penetrance at birth. The most prominent features of CDM include skeletal muscle, smooth muscle, pulmonary, and CNS involvement (Harper, 2001) Unlike adult onset DM1, however, relatively little is known about the root cause of these symptoms at the molecular level. For example, pulmonary insufficiencies are responsible for the majority of infantile mortality in CDM, requiring prolonged ventilation at birth for survival (Campbell e t al., 2004; Rutherford et al., 1989) While it is acknowledged that there is reduced efficiency of the neonatal lung, it is unclear whether respiratory distress is due to the inability of the underdeveloped diaphragm and intercostal muscles t o promote breathing, inherent pulmonary defects, or a combination of both (Harper, 2001) Neonates with CDM also present with significant muscle involvement. Hypotonia is caused by a lack of baseline contractions in skeletal muscle that normall y p romote rigidity and posture (Bergen, 1985; Bodensteiner, 2008) CDM babies are characteristically floppy shortly after birth, lacking significant movements (Harper, 1975; Harper, 2001) Additionally, poor suckling may also be due to poor ly developed to ngue muscle. Histological analysis of CDM skeletal muscle reveals a delay in myogenic development, including the abnormal presence of immature myotubes, small bundles of fibers, thin fibers, and prominent centralized nuclei, while motor neurons
71 appear nor mal (Farkas Bargeton et al., 1988; Sahgal et al., 1983a; Sahgal et al., 1983b; Tanabe and Nonaka, 1987; Tominaga et al., 2010) Difficulties in swallowing postnatally also indicate that the smooth muscle of the esophagus is affected, which may lead directly to another common CDM phenotype, hydramnios, in which an increase of amniotic fluid during pregnancy is caused by a lack of fetal swallowing (Schild et al., 1998; Wieacker et al., 1988) Remarkably, these symptoms, including pulmonary involvements and delays in muscle development, are generally resolved by early childhood and do not affect the prognosis of mobility or respiration later in life (Harper, 2001) CDM patients also display mental retardation and low IQs in adulthood, with one estimated mean of 66.1 +/ 16.2 alt hough there is little evidence as to the cause (Harper, 1974; Harper, 1975) Postmortem analysis of infants afflicted with CDM and mental retardation has yielded no evidence of gross cerebral morphological changes. However, magnetic resonance imaging (MRI) and computed tomography (CT) scans have indicated changes in ventricular size and an increased frequency in intraventricular haemorraging in neonates, although it is not evident if these changes are directly involved in disease (Regev et al., 1987) Finally, talipes, commonly known as club foot, is seen in approximately half of all CDM patients (Ray et al., 1984; Siegel et al., 1984) In CDM, talipes is thought to be caused by in utero muscle weakness (hypotonia) and decreased fetal movements. Surgical correction is commonly used to treat this disability. Ultimately, CDM patients, despite overcoming ea rly muscle and pulmonary deficits, develop early onset adult DM symptoms, often manifesti ng itself in early adolescence While CDM appears clinically distinct from its adult onset counterparts in DM1 and DM2, the causative mutation is the DM1 (CTG)exp that has
72 intergenerationally increased in size to greater than 1000 repeats (Tsilfidis et al., 1992) There is no evidence of a DM2 repeat expansion that contributes to the onset of a CDM phenotype, indicating a link between DM1 and dise ase. This observation begs the question, what causes CDM? Potential Models of CDM Two closely related (CTG)n rich repeat expansions cause very similar adult onset diseases, DM1 and DM2. MBNL1 loss of function through interactions with either (CUG)exp or (CCUG)exp have proven to be the common link between these adult onset disorders. How then does DM1 specifically promote additional CDM phenotypes during embryogenesis and neonatal life? The presence of a second form of DM that does not cause a congenital phenotype provides an opportunity to compare and contrast potential molecular mechanisms involved in congenital disease. Instead of focusing on the similarities of this disease, the answer likely lies in the differences between the two causative mutations. The DM1 and DM2 mutations are located in noncoding portions of two seemingly unrelated genes, DMPK and CNBP respectively. The (CTG)exp promotes adult onset DM1 between 50 and 1000 repeats, while expansions between 10004000 cause CDM. In DM2, (CCT G)exp greater than 75 and less than 11,000 cause the adult onset disorder. One possibility to explain the observation that only (CTG)exp promote CDM is that embryonic expression patterns between the two genes do not overlap during development. In this sc enario, DMPK mediated RNA toxicity could promote the onset of CDM by being preferentially expressed in affected tissues, while CNBP expression is either absent or below a threshold of expression needed to promote disease. However, Dmpk and Cnbp have demonstrated no significant differences in embryonic expression
73 patterns (Swanson lab unpublished data). Using whole mount in situ hybridization, Cnbp is expressed ubiquitously during murine embryonic development while Dmpk displays a more restricted expression pattern. Additionally, Cnbp staining appears more intense than Dmpk suggesting that Cnbp is expressed at a higher level. In an alternative scenario, a more restricted embryonic expression pattern for DMPK may promote the onset of CDM in affected tissu es, while the ubiquitously expressed CNBP results in either an early termination of embryogenesis (i.e. before DMPK is expressed) or widespread expression compromises too many tissues which results in an inability to develop and subsequent lethality. Anot her possibility is that DM1 locus specific affects contribute to neonatal disease. Closely linked DMPK genes, in which expression is inhibited by (CTG)exp, may contribute to the onset of congenital disease through a combinatorial model of haploinsufficieny of DM1linked genes and MBNL1 loss of function model (see General Introduction). Genetic crosses to create Dmpk+/ -; Six 5+/ -; Mbnl1E3/E3 mice (each individual knockout mouse did not develop a CDM phenotype) failed to recapitulate any CDM phenotypes (Swanson lab unpublished data). However, we cannot rule out the contributions of DMWD and any dominant negative effects of the DM1 antisense transcript. The most likely explanation may be differences in the RNA repeat itself. We have previously demonstrated that (CUG)exp and (CCUG)exp display variable stabilities with MBNL1 in vi tro potentially altering its toxicity. Therefore, (CUG)>1000 RNAs may be more inherently toxic. Another distinct difference between (CUG)exp and (CCUG)exp is the asymmetry in repeat expansion length. Somatic expansions of both (CTG)n and (CCTG)n result in a heterogeneous population of repeats in different tissues and time points t hroughout life that can contribute to multiple
74 aspects of adult onset disease, including ageof onset, severity of disease, and penetrance of symptoms (Ashizawa et al., 1993; Martorell et al., 1997; Martorell et al., 2000; Wong et al., 1995) On the other hand, both repeats reach a maximum expansion (CTG)~4000 and (CCTG)~11,000, suggesting that further expansions are incompatible with life. One possibility is that (CUG)n and (CCUG)n could both promote similar CDM phenotypes embryonically, however, unstable (CCTG)n expansions reach a critical length and promote embryonic lethality. This possibility would predict that longer (CCTG)n expansions are more unstable during transmission or development than expanded (CTG)n repeats. While these models are not mutually exclusive, they provide insight into the complexities involved in determining and modeling specific genetic co ntributions to the congenital onset of disease. MBNL Family in DM Considering that (CUG)exp contributes to adult onset DM1 by sequestering a dsRNA binding protein, MBNL1, and inhibiting its function, the same toxic RNA could compromise other factors via dominant negative interactions. Interestingly, MBNL1 belongs to a family composed of three closely related genes, including MBNL2 and MBNL3 (Fig. 3 1). All three genes code for proteins that contain 4XC3H motifs, which mediate RNA interactions (Teplova and Pate l, 2008) MBNL shares >60% conserved amino acid identity in the full length proteins and on average ~90% in their C3H motifs (Fig 3 2). We have previously demonstrated that all three proteins can interact to suppress inclusion of cTNT exon 5 in cell culture (see Chapter 2 Fig. 2 9 ), suggesting that their substrate specificity overlaps. Additionally, GFP tagged MBNL1, MBNL2, and MBNL3 colocalize with (CUG)exp RNAs into discrete nuclear foci when ectopically expressed in cells (Fardaei et al., 2002) Moreover, MBNL1 and MBNL2 have been
75 shown to colocalize with (CUG)exp in the skeletal muscles of DM1 patients, further implicatin g the family in disease pathogenesis (Holt et al., 2009) While Mbnl1E3/E3 knockout mice recapitulate cardinal DM phenotypes, genetically altered Mbnl2 mice have provided variable results. Two different mouse models with decreased levels of Mbnl2 mRNA have been generated using genetrap technology in which a NeoR/ LacZ cassette traps expression using a novel 3 splice site to truncate the full length endogenous transcript. The first mouse contains a genetrap in Mbnl2 intron 2 ( Mbnl2GT2/GT2) which produces a truncated protein containing the first 58 amino acids of Mbnl2 (1XC3H motif) fused with the reporter cassette and displays a significant decrease in overall Mbnl2 mRNA (Ha o et al., 2008) The Mbnl2GT2/GT2 mouse displays multiple phenotypes, including skeletal muscle centralized nuclei, myotonic discharges, minor changes in CLCN1 alternative splicing, and focal loss of CLCN1 localization to the sarcolemma. Interestingly, the second mouse, which contains a genetrap in Mbnl2 intron 4 ( Mbnl2GT 4 /GT 4) and produces a protein containing the first 180 amino acids of Mbnl2 (2XC3H motifs) fused with the reporter cassette, demonstrates normal adult skeletal muscle by histology and no changes in the alternative splicing of CLCN1 or myotonia, despite a >90% loss of Mbnl2 mRNA (Lin et al., 2006) One possibility to account for the observed differences in the two models is that while Mbnl2GT 4 /GT 4 mice have a greater reduction in endogenous Mbnl2 mRNA than Mbnl2GT2/GT2 mice, the pair of Mbnl2 N terminal C3H motifs may be sufficient to prevent the skeletal muscle pathology in vivo Generation of a Mbnl2 knockout mouse would help to clarify these result s. Therefore, while the exact role of MBNL2 in DM is unclear, Mbnl2Gt/Gt mice do not develop CDM or other unmodeled DM phenotypes, including progressive skeletal
76 muscle wasting. There are currently no in vivo genetic models for Mbnl3 involvement in disease. Given the integral involvement of MBNL1 in adult onset DM, compromised function of MBNL2 and MBNL3 are promising candidates for contribution to the o nset and progression of disease However, given the genetic contribution of the many genes that are potentially affected by the (CTG)exp, a comprehensive model may be necessary to fully dissect the constellation of DM phenotypes (Fig. 33). Overview of MBNL3 MBNL3 was first identified in a bioinformatic search for proteins that shared a high degree of identity to MBNL1 and later in a suppression subtractive hybridization (SSH) experiment designed to screen for genes preferentially expressed during myogenic proliferation (Fardaei et al., 2002; Squillace et al., 2002) MBNL3 is an interesting candidate gene in CDM for a variety of reasons. First, Mbnl3 is primarily expressed embryonically. Northern blot analysis of poly A selected RNA from adult murine tissues reveals very l ow Mbnl3 expression in all adult tissues tested, despite a tenfold higher exposure time than Mbnl1 and Mbnl2 which displayed near ubiquitous expression (Kanadia et al., 2003b) Mbnl3 expression was detectable by RT PCR in adult murine lung, spleen, and testis, however, this analysis is not quantitative and likely is reflective of either low level transcription throughout the tissues tested or a subpopulation of cells in these tissues that express Mbnl3 preferentially (Lee et al., 2007) Whole embryo total RNA northern blotting from various murine embryonic stages indicate a high level of Mbnl3 expression, comparable to Mbnl1 and Mbnl2 (Kanadia et al., 2003b) However, the RNA for this analysis was collected from total embryos and does not offer any insight spatially during development. Additionally, Mbnl3 is known to be expressed highly in the placenta, which was included with the embryos, making interpretation of
77 actual Mbnl3 levels a ttributable to embryonic expression difficult. Whole mount RNA in situs demonstrate that Mbnl3 mRNA is present in the limb bud, first brachial arc h, and the neural tube at E9.5 (Kanadia et al., 2003b) This data indicates that Mbnl3 expression, unlike that of Mbnl1 and Mbnl2, is limited to embryonic devel opment. This data is based solely on mRNA expression. Interestingly, Mbnl3 mRNA contains a unusually large 3 UTR (~ 8 kb), making it subject to post transcriptional regulation. In other words, Mbnl3 protein levels may not accurately reflect the corresp onding mRNA levels in vivo Nonetheless, an overlap in Mbnl3 expression and tissues affected in CDM is consistent with an MBNL3 loss of function model for embryonic and neonatal disease. Second, Mbnl3 expression has been shown to affect myogenic differentiation in vitro Mbnl3 mRNA is expressed in proliferating C2C12 cells and is downr egulated after differentiation (Lee et al., 2007; Squillace et al., 2002) Overexpression of Mbnl3 in stably transfected C2C12 cells inhibits differentiation, while 50% knockdown of Mbnl3 using an antisense morpholino promotes the opposite effect, an increase in differentiation. This data suggests a critical role for Mbnl3 during normal myogenesis. Therefore, inhibition of this function due to sequestration may play an important role in the misregulation of embryonic skeletal muscle development seen in CDM. Results Mbnl3 mRNA Expression and Alternative Splicing are Spatially and Temporally Regulated Mbnl3 has been previously reported to be expressed during embryonic d evelopment with the highest levels in the placenta (Fardaei et al., 2002) However, this analysis focused on whole embryos and nonquantitative methods. Therefore, we chose to detail Mbnl3 expression throughout development both spatially and temporally
78 using N orthern blotting (Fig. 34 ). Mbnl3 mRNA is expressed at its highest levels at embryonic day 15 (E15) in placenta, forelimb and tongue with levels reduced in forelimb and tongue by E18 and undetectable in adult tissues. Mbnl3 is also present in E18 and postnatal day 1 lung (P1) and absent in adult. Expression is also noted in P1 and adult spleen. Mbnl3 family member, Mbnl1, is expressed in more tissues during development. This analysis confirms and extends previous reports of Mbnl3 expression to specific time points and tissues during development. It is worth noting that Mbnl3 mRNAs are expressed during embryogenesis in those tissues affected in the congenital form of myotonic dystrophy. Although Mbnl3 N orthern blotting reveals expression, it fails to detail the mRNAs that compose the population. Mbnl1 and Mbnl2 are extensively alternatively splice d, producing multiple isoforms (Pascual et al., 2006) ( Swanson Lab, unpublished data). In order to assay for the alternative splicing patterns of Mbnl3, we amplified isoforms using RT PCR from P1 and adult tissues with primers positioned in constitutively spliced noncoding exons 1 and 8 (5 and 3 untranslated regions, respectively). Subsequent amplicons were subcloned, sequenced, and annotated ( Fig. 3 5 ). Analysis revealed previously unreported exons 7a, 7b, and cryptic 3 splice sites in exons 6 and 7 (exon 7 is referred to as exon 7c from here in). Exons 7a, 7b, and 7c are alternatively spliced producing unique Mbnl3 C termini. Exon 2 is also alternatively spli ced producing a hypothetical N terminal truncation which could produce a protein with either four C3H or two C3H RNA binding motifs (Fig. 3 6 ). Although Mbnl3 undergoes alternative splicing events which produce multiple isoforms, we sought to identify the relative use of these alternative exons in the mature mRNA. To examine Mbnl3 alternative splicing, RT PCR
79 with a forward primer positioned in constitutively spliced exon 3 and reverse primers specific to alternatively spliced exons 7a, 7b, and 7c were us ed (Fig. 3 7 A). Embryonic and P1 brain, as well as placenta, included exons 7a and 7b to the greatest extent while exon 7c isoforms were the most abundant in all tissues examined (Fig. 37 B). Mbnl3 exon 2 assumes a default level of inclusion of approximately 30% (data not shown). Taken together with the sequencing data, Mbnl3 exon 7c containing mRNAs are the predominant isoforms in all tissues while inclusion of exons 7a and 7b vary between tissues. This is particularly interesting considering Mbnl1 uti lizes its Cterminus for self interactions, which suggests that Mbnl3 may interact with multiple proteins in different tissues. Polyclonal Antisera Raised Against the C terminus of Mbnl3 Mbnl3 contains an unusuall y large 3 UTR in the mature mRNA (~ 7800 nucleotides) which contains many predicted cis elements that have been shown to post transcriptionally regulate mRNA fate (data not shown). Of course, Mbnl3 mRNA levels may not accurately reflect the level of protein produced in the tissues assayed. Unf ortunately, there is no antibody available that is cross reactive with murine Mbnl3 which is useful for tissue analysis. The Mbnl family of proteins also share a high degree of amino acid identity, which limits the unique regions available that may be tar geted for antibody production. Therefore, we prepared polyclonal antibodies against the most immunogenic and unique region of the protein, the C terminus (Fig. 37 A). The majority of Mbnl3 isoforms should be recognized by this antibody (Table 31). A 15 amino acid peptide (NVPYVPTTTGNQLKY) was synthesized and conjugated to KLH, four rabbits were immunized, and bleeds were taken to assay for Mbnl3 reactivity. Placenta whole cell lysate (E15) was immunoblotted with either Mbnl3 antisera or preimmune sera from
80 immunized rabbits (Fig. 38 A). Antisera from rabbits A and C recognized two distinct bands at approximately 37 kDa and 27 kDa that were not seen in the preimmune sera. Further evaluation revealed Mbnl3 anti sera A and C recognized these bands with a similar affinity when E15 placenta lysate concentration was varied (Fig. 38 B). Bleeds A and C were chosen for affinity purification and subsequent studies. Mbnl3 polyclonal antibody C ( Mbnl3) will be used for the remainder of this study. Due to th e high degree of identity shared between the Mbnl proteins, it is important to exclude the possibility that Mbnl3 cross reacts with other family members. An Mbnl3 immunoblot shows that antisera reacts with exogenous myc tagged Mbnl3 in COSM6 cells, bu t not Mbnl1 or Mbnl2 (Fig. 39 A). Mbnl3 was also efficiently immunoprecipitated (Fig. 39 B) as well as visualized by immunocytochemistry (Fig. 3 10) in exogenously expressing COSM6 cells. To ensure that Mbnl3 is specifically recognizing Mbnl3 in placenta, protein was immunoprecipitated from placenta whole cell lysate using Mbnl3 and subjected to MALDI TOF mass spectrometry, which correctly identified Mbnl3 peptides (data not shown). Therefore, we have developed a new polyclonal antibody that recogniz es Mbnl3 and which performs well for a wide range of applications. Mbnl3 Isoforms Localize to both the Nucleus and Cytoplasm Mbnl3 has been previously reported to be expressed in the proliferating myoblast cell line C2C12 and is down regulated in response to myogenic differentiation (Lee et al., 2007; Squillace et al., 2002) T his expression pattern suggests an important role for Mbnl3 during myoblast proliferation which must be down regulated in response to exte rnal stimuli to promote proper myotube maturation. Using our Mbnl3 antibody, we
81 sought to confirm Mbnl3 expression in proliferating C2C12 cells. Interestingly, we identified two Mbnl3 proteins at 37 kDa and 27 kDa corresponding to the predicted Mbnl3 i soforms encoded by mRNAs (Table 3 1 and data not shown) that either included or excluded exon 2. Expression of both isoforms are inhibited by Mbnl3 siRNA, but not by nonspecific siRNA (Fig. 3 11). This is the first evidence of an N terminal truncation p roducing a lower isoform in the Mbnl family of proteins. Moreover, a human hepat ocarcinoma cell line, Huh7, exclusively expresses the MBNL3 lower isoform at 29 kDa (Fig. 3 11). siRNA directed against MBNL3 efficiently knocks down MBNL3 expression while S MN siRNA do es not affect expression. The h uman MBNL3 isoform is 2 kDa larger due to the inclusion of a humanspecific exon 8 that is not conserved in mice (data not shown). The appearance of distinct Mbnl3 isoforms suggests that Mbnl3 possesses multiple functions. Examination of MBNL3 localization using immunocytochemistry reveals discrete cytoplasmic foci in Huh7 cells, while C2C12 displays both nuclear and cytoplasmic foci (Fig. 3 12). This result is in direct contrast with Mbnl1 and Mbnl2, which appear predominately nuclear (Fardaei et al., 2002) ( data not shown). Cytoplasmic foci in Huh7 and C2C12 cells do not localize with know n structures, including P bodies and other RNA binding proteins (Fig. 312 and data not shown). In support of the previous observation, lysates from nuclear and cytoplasmic fractionations of Huh7 and C2C12 cells were immunoblotted to assay for cellular localization. Interestingly, the MBNL3 isoform excluding exon 2 localized exclusively to the cytoplasm in both Huh7 and C2C12 cells (Fig. 3 13 A,C). The Mbnl3 isoform including exon 2 localized to both the nucleus and cytoplasm in C2C12 cells. CUGBP1 and l actate dehydrogenase A (LDHA)
82 were used to assay for nuclear and cytoplasmic fractionation, respectively, while MBNL1 served as a family control. We next sought to determine if Mbnl3 localization in vitro accurately reflected it localization in vivo Con sidering Mbnl3 is expressed in a proliferating myoblast cell line, we chose to assay for Mbnl3 localization during embryonic myogenesis. Localization of Mbnl3 isoforms from E15 forelimbs was remarkably similar to C2C12 (Fig. 313B). Unlike Mbnl1, which i s predominately a nuclear RNA binding protein involved in alternative splicing, Mbnl3 encodes several isoforms whose variable cellular distribution implies a nonoverlapping function with other family members. Mbnl3 Proteins are Primarily Expressed during Embryogenesis To be a viable candidate for sequestration according to the RNA dominance model, Mbnl3 protein expression must overlap with DMPK expression. To address this question, brain, skeletal musc le/limb, tongue, and lung tissues were isolated from multiple time points during mouse development and immunoblotted for Mbnl3 expression. Mbnl3 protein isoforms were readily detectable in E15E18 placenta. Forelimb and tongue displayed more moderate levels at E15 and were undetectable by E18, while lung s howed expression at E18 which persisted until P1 (Fig. 314). Mbnl3 expression was also present in forelimb buds as early as E11.5 (data not shown). The single band recognized by Mbnl3 in adult brain was immunoprecipitated, subjected to MALDI TOF mass spectrometry and determined to be glutamine synthetase (Fig. 314 and data not shown). Since Mbnl3 protein is not detectable in the adult brain, this crossreaction did not affect our analysis. Interestingly, Mbnl3 expression in the forelimb and tongue c losely overlaps with embryonic myogenesis (Fig. 315) and is
83 down regulated after completion of secondary myogenesis, indicating a potential role for Mbnl3 during skeletal muscle development. If Mbnl3 is involved in the embryonic development of skeletal muscle, Mbnl3 might also be important for adult myogenesis (i.e. skeletal muscle regeneration). To test this hypothesis, we induced regeneration in vivo by injecting mous e TA muscles with Note xin, a venom peptide isolated from the Australian viper Notechis scutatus. Notexin is a 119 amino acid peptide with phospholipase A2 activity which, when injected into skeletal muscle, has potent myotoxic affects and promotes muscle necrosis, satellite ce ll activation and muscle regeneration (Dix on and Harris, 1996; Harris and MacDonell, 1981) Notexin was injected into the TA muscles of C57BL6/J mice (1012 weeks of age), where they were isolated at two day intervals to assay for regeneration (Fig. 316A,B). Hematoxylin and Eosin (H&E) staining of 10 m transverse cryosections was used to track skeletal muscle regeneration (Fig. 3 16C). Surprisingly, Mbnl3 demonstrated a sharp expression peak at D ay 3 post injection simultaneously with a gene involved in myoblast fusion, myogenin (Fig. 317). Mbnl3 expression at this time point during regeneration is consistent with proliferating myoblasts (Fig. 3 18), but does not rule out the contribution from other cell types However, this pattern of expression is similar to embryonic forelimb/tongue and C2C12 expression, suggesting that Mbnl3 may play an essential role during myogenesis. Loss of Mbnl3 Inhibits Myogenic Differentiation in a C2C12 Model Congenital myotonic dystrophy patients present with immature myotubes per inatally (Farkas Bargeton et al., 1988; Iannaccone et al., 1986; Sarnat and Silbert, 1976) If Mbnl3 plays an important role during myogenesis then conceivably loss of
84 Mbnl3 would potentially inhibit this process. Using the C2C12 myogenesis model, in which immortalized proliferating myoblasts can be induced to differentiate upon the wi thdrawal of growth factors we knocked down Mbnl3 expression with siRNA and assayed for a delay in myogenesis (F ig. 3 19). In a control differentiation, Mbnl3 expression is down regulated during the time course of myogenic differentiation at the same period when myotubes and m yosin heavy chain (Mhc a terminal differentiation marker) appear. However, when Mbnl3, expression is inhibited using siRNA there is a delay in Mhc expression and fewer myotubes at corresponding time points during differentiation (data not shown) Therefore, inhibition of Mbnl3 is suffici ent to delay myogenesis in vitro ; reaffirming that MBNL3 is a viable candidate to contribute to a delayed myogenesis phenotype seen in CDM infants. Targeting Mbnl3 to Generate a Conditional Allele in Embryonic Stem Cells If loss of MBNL3 function by sequestration contributes to the onset of CDM, then removing Mbnl3 in vivo would model the onset of disease. To eliminate those Mbnl3 isoforms that interact with toxic (CUG)n RNAs (Fardaei et al., 2002) we focused our attention on Mbnl3 exon 2. This exon encodes the first translational initiation codon which is responsible for producing the full length four C3H motif isoform (Fig. 3 6 ), which has been previously implicated in disease The Mbnl3 exon 2 was replaced with a Mbnl3 exon 2 flanked by loxP sites in C57BL6 embryonic stem cells (ESC) using stand ard targeting techniques (Fig. 320A). LoxP sites were used to eliminate intervening DNA sequence in vivo by expressing Cre a site specific DNA recombinase. This targeting strategy allows us to remove Mbnl3 exon 2 in tissues and time points of interest by crossing Mbnl3cond/+ mice to mice expressing Cre under a genespecific promoter. Probes located outside the arms of homology were used to identify correctly
85 targeted ESC clones. Out of 120 clones screened, 13 were positive (10.8%) for recombination at the 3 end and 5 were positive (4.2%) for recombin ation at the 5 end (Fig. 3 20B). Less efficient 5 targeting was due to recombination of the conditional exon (instead of the 5 arm of homology), excluding the 5 loxP site from the allele. Considering Mbnl3 is an X linked gene targeted in male ESCs, only one all ele is detectable by s outhern analysis. To verify successful targeting of Mbnl3, Cre was ectopically expressed (Fig. 3 21A) in targeted ESCs and assayed for loss of exon 2. The reduction in size of the Mbnl3 KpnI fragment by southern blot analy sis coincides with removal of exon 2 from the locus (Fig 321B). RT PCR analysis with a forward primer in Mbnl3 exon 1 and reverse primer in exon 8 demonstrate the loss of exon 2 containing transcripts in Mbnl3E2/Y ESCs, while Mbnl3cond/Y ESCs maintain exon 2 splicing (exon 2 inclusion = 1235 bp; exon 2 exclusion = 988 bp). Primary neomycinresistant ( NeoR) mouse embryonic fibroblasts (MEFs) serve as a positive control for Mbnl3 splicing. One concern about leaving the NeoR cassette in the conditional Mb nl3 locus is the possibility that it is included in the mature transcript by uti lizing cryptic 3 splice sites (Nagy et al., 1998; Ren et al., 2002) This scenario could result in hypomorphic Mbnl3 expression instead of a conditional allele designed to express at wild type levels. However, RT PCR did not detect any cryptic splici ng products into the NeoR cassette (Fig. 3 21C). Mbnl3cond/Y ESCs were injected in C57BL/6J Tyrc 2J blastocysts to obtain chimeric mice. Male chimeras were bred to female C57BL6/J mice to attain germline transmission; the resulting F1 mice were C57BL/6J congenic Mbnl3cond/+ females (data not shown).
86 Mbnl3 E2/Y Mice Fail to Express Mbnl3 4XC3H Isofor ms To test the hypothesis that loss of Mbnl3 full length isoforms contribute to CDM phenotypes, Mbnl3cond/+ female mice were crossed to male B6.C Tg( CMV cre)1Cgn/J mice to remove Mbnl3 exon 2 in the germline (Schwenk et al., 1995; Utomo et al., 1999) The resulting Mbnl3E2/+ females were crossed to wild type males to obtain Mbnl3E2/Y males. Mbnl3E2/Y males were used for analysis for the remainder of this study. Because Mbnl3 is primarily expressed during embryogenesis, it is possible that the loss of Mbnl3 exon 2 isoforms during this stage of development result in an embryonic lethal phenotype. However, the numbers of wild type and Mb nl3E2/Y males (wild type, n=36 and Mbnl3E2/Y, n=29) genotyped suggests that Mbnl3E2/Y has not deviated from the expected Mendelian ratio of 1:1 (wild type male: Mbnl3E2/Y male). Therefore, loss of Mbnl3 exon 2 isoforms do not result in an embryonic lethality. To confirm that Mbnl3E2/Y male mice no longer express full length Mbnl3 37 kDa isoforms, E15 forelimb and P1 lung were examined. RT PCR comparing Mbnl3 transcripts in wild type and Mbnl3E2/Y male s indicate a loss of exon 2 transcripts, but an increase in exon 2 excluded transcripts (Fig. 3 22A). This is likely reflective of the steady state levels of total Mbnl3 mRNA in the absence of Mbnl3 exon 2 transcripts. Moreover, Mbnl3E2/Y males show an increase in the expression of the Mbnl3 27 kDa isofor m by immunoblot analysis (Fig. 322B). P1 lung also shows and increase in the Mbnl3 27 kDa isoform. However, this increase does not correlate with the total protein in the wild type P1 lung. This sugg ests that the expression of the Mbnl3 27 kDa isoform is regulated post transcriptionally and independently of the Mbnl3 37 kDa isoform.
87 Therefore, Mbnl3E2/Y mice do not express Mbnl3 37 kDa isoforms and upregulate 27 kDa isoforms during development. Mbnl 3 E2/Y Mice do not Recapitulate Cardinal Phenotypes of Congenital Myotonic Dystrophy The clinical manifestations of CDM at birth include poor suckling, movement deficits (hypotonia), immature skeletal muscle, skeletal deformities in the extremities (talip es), pulmonary insufficiencies, and a failure to reach developmental milestones (Harper, 2001) However, Mbnl3E2/Y mice appear visibly normal at birth (Fig. 3 23A), including indistinguishable neonatal movements and milk spots by P2, indicating normal feeding. Talipes may result from delayed muscle development during embryogenesis To determine if Mbnl3E2/Y mice displayed talipes at birth, skeletal preps were performed on P1 pups; wild type and Mbnl3E2/Y hind limbs and forelimbs were compared, but no differences in the skeletal structures of the limb were observed (Fig. 3 23B). CDM patients also present with immature skeletal muscle at birth, highlighted by smaller myofibers and centralized nuclei, indicating a delay in embryonic myogenesis (Farkas Bar geton et al., 1988) To assay for immature muscle fibers, H&E staining of 7 m transverse paraffin sections from Mbnl3+/Y and Mbnl3E2/Y mice forelimbs were compared. No significant differences in either centralized nuclei or myofiber size were noted (Fig. 324). Normal myofiber appearance was confirmed by wheat germ agglutinin, a fluorescently conjugated lectin that outlines muscle fibers by interacting with glycosylated protei ns in the extracellular matrix (Shaper et al., 1973) and immunohistochemistry using a terminal skeletal muscle marker, muscle specific myosin heavy chain (Fig. 324). No differences were observed postnatally, as Mbnl3E2/Y mice to continue gain weight normally past sexual maturity (Fig. 325). The major cause of
88 mortality in infants with CDM is respiratory distress, most likely from a combination of immature diaphragm and intercostals muscles as well as poorly understood pulmonary insufficiencies (Campbell et al., 2004; Rutherford et al., 1989) Howe ver, no Mbnl3E2/Y mice died (n=65) postnatally, indicating adequate pulmonary function. Taken together, this data suggests that loss of Mbnl3 37 kDa isoforms (i.e. isoforms including exon 2) alone are not sufficient to reproduce the onset of CDM in a mouse model. Loss of Mbnl3 4XC3H Isoforms does not Inhibit Skeletal Muscle Regeneration in an Injury Model Individuals with congenital myotonic dystrophy that survive past childhood go on to develop the adult onset symptoms of myotonic dystrophy. One of t he key characteristics of the adult onset neuromuscular disorder is the progressive wasting of skeletal muscle. Although mouse models have successfully recapitulated the majority of highly penetrate disease characteristics, they have failed to address the cause of muscle wasting in adult onset DM. If Mbnl3 expression in activated satellite cells is required for proper maintenance or repair of skeletal muscle following injury, one interesting possibility is that loss of Mbnl3 due to sequestration by toxic C(CUG)n repeats could inhibit repair/regeneration of muscle. To test this hypothesis, adult skeletal muscle was induced to regenerate by injecting the TA muscle of wild type and Mbnl3E2/Y mice (10 12 weeks of age) with Notexin; TA muscle was pulled at two day intervals to assay for inhibition of regeneration (Fig. 3 26A,B). RT PCR and immunoblot analysis confirmed the loss of Mbnl3 37 kDa isoform expres sion during regeneration (Fig. 3 26C,D). Expression levels of myogenin, a transcription factor invol ved in the commitment and fus ion of proliferating myoblasts ( Buckingham et al., 2003) was not significantly altered during regeneration in Mbnl3E2/Y mice compared to wild type
89 controls. Mbnl1 RNA levels at Day 3 post injection in Mbnl3E2/Y mice were slightly lower, but protein levels r emained unaffected (Fi g. 3 26C,D). Interestingly, Mbnl3 27 kDa isoform was not upregulated at Day 3 post Notexin injection in Mbnl3E2/Y mice. H&E staining of 10 m transverse cryosections to assay for regeneration did not show a reduction in the regenerative capacity of Mbnl 3E2/Y mice (Fig. 3 27A). Wheat germ agglutinin, a lectin that outlines myofibers by interacting with glycosylated proteins of the extracellular matrix and plasma membrane does not show any difference between wild type and Mbnl3E2/Y mice during regeneration (Fig. 3 27B). Desmin, a marker of terminal skeletal muscle differentiation, is unaffected during regeneration (Fig. 3 28). This data suggests that loss of Mbnl3 37 kDa isoforms (i.e. isoforms including exon 2) alone do not inhibit the regeneration of adult skeletal muscle and are not sufficie nt to cause a wasting phenotype. Discussion Mbnl is a family of highly conserved genes that are implicated in developmental maturation of tissues through the regulation of RNA metabolism (Pascual et al., 2006) Most notably, MBNL1 promotes a change in alternative splicing patterns of specific genes during the fetal to adult transition in terminally di fferentiating tissues, which results in protein isoforms that optimally support that t issues physiological needs (Lin et al., 2006) When MBNL1 function is inhibited by toxic C(CUG)n RNA expression in DM, impairment of this splicing transition results in the misregulation of alternative splic ing and disease. Remarkably, the Mbnl1E3/E3 mouse, which mimics loss of MBNL1 protein in DM by removing the exon responsible for initiating translation of full length Mbnl1 protein, recapitulates the majority of highly penetrant DM phenotypes (Kanadia et
90 al., 2003a) This observation is even more extraordinary considering that toxic C(CUG)n repeats likely contribute to disease pathogenesis by com promising not only MBNL1, but also other family members, altering downstream phosphorylation of PKC target CUGBP1, as well as altering expression of genes linked to the DMPK locus (Fu et al., 1993; Inukai et al., 2000; Kuyum cu Martinez et al., 2007; Novelli et al., 1993; Sabouri et al., 1993) Alt hough the Mbnl1E3/E3 mouse has provided important insights into adult onset DM and the molecular pathways affected, many questions remain. For example, what causes deficits in the central nervous system and progressive wasting of skeletal muscle in adul ts, and what molecular events lead to CDM? Modeling these elusive phenotypes will prove critical to determining the underlying mechanism of disease which should allow the development of effective therapies. Mbnl3 is Expressed in Developing Tissues that ar e Affected in CDM Previous studies have reinforced the idea that other MBNL family members are likely comprised in CDM/DM, including evidence that MBNL1, MBNL2, and MBNL3 interact with expanded (CUG)n repeats both in vitro and in vivo (Fardaei et al., 2002; Holt et al., 2009) While Mbnl1 and Mbnl2 expression overlaps in many tissues throughout development, in situ analysis, tiss ue RT PCR, and whole embyo RNA N orthern blot analysis of Mbnl3 has led to the prevailing view point that Mbnl3 expression i s limited to embryogenesis However, this interpretation relies on the assumption that Mbnl3 is not post transcriptionally regulated and protein levels are reflective of mRNA levels. Our data, using a new Mbnl3 antibody, provides evidence for embryonic Mbnl3 expression in those tissues affected in CDM (tongue, lung, forelimb) as well as regener ating adult skeletal muscle. Mbnl3 protein is not detectable
91 in differentiated adult tissues. While we were unable to use the Mbnl3 antibody for immunohistochemistry and localization of Mbnl3 to specific cell types in vivo it appears that Mbnl3 is mos t likely expressed in myoblasts and not differentiated myofibers. Mbnl3 is expressed in proliferating C2C 12 cells and downregulated upon differentiation. Mbnl3 mRNA is also expressed in limb buds by E9.5 (Kanadia et al., 2003b) and protein is present in developing forelimb from E11.5E16.5 until the completion of secondary myogenesis (Fig. 314 and 315 and data not shown) and during adult skeletal muscle regeneration until cell cycle arrest and terminal differentiation (Fig. 317 and 318). An interesting possibility is that Mbnl3 is important for provi ding alternative splicing, or another mRNA regulating event, that is necessary for the embryonic developmental program, similar to Mbnl1 is in adult tissues. Mbnl3 alternative splicing produces multiple isoforms varying at both the N and C termini. This is not uncommon in the Mbnl family, as Mbnl1 and Mbnl2 both undergo alternative splicing of their C termini which can modulate localization and self interaction ( Swanson Lab unpublished data). Like Mbnl1 and Mbnl2, Mbnl3 may utilize alternative splicing a t its C terminus to modulate proteinprotein interactions and function. However, the Mbnl3 N terminal truncated isoform which contains 2XC3H motifs (Fig. 3 2 and Fig. 36 ) is unique among family members. Mbnl1 also produces mRNAs that lack exon 3, which contains the Mbnl1 translation initiation codon, but no truncated proteins have been detected in tissues (Swanson lab unpublished data). N terminal truncated isoforms with 2XC3H motifs may represent an Mbnl3 protein that interacts with a different set of RNAs and plays a different role than Mbnl3 full length 4XC3H protein. Temporal and spatial expression, as well as localization differences between Mbnl3
92 4XC3H and 2XC3H isoforms further suggest a divergent role for these proteins during development. Mbnl3 is R equired for C2C12 D ifferentiation While it has been established that Mbnl3 is expressed in proliferating C2C12 cells (Fig. 3 11), previous studies designed to determine the role of Mbnl3 during myogenesis have provided counterintuitive results. Neonates with CMD present with immature myotubes, which suggests a delay in myogenesis (Farkas Bargeton et al., 1988) Previous studies where a myc tagged 4XC3H Mbnl3 transgene was overexpressed in stably transfected C2C12 demonstrated a delay in differentiatio n as assayed by Mhc expression (Squillace et al., 2002) Moreover, a morpholino antisense oligonucleotide (designed against the 5 UTR/translation initiation codon) that inhibits translation of Mbnl3 also promotes diff erentiation in C2C12 cells If loss of Mbnl3 function by (CUG)n sequestration contributes to CDM, then the expected result of inhibiting translation in C2C12 would be to delay differentiation. However, there are critical controls absent from these experiments. First, the Mbnl family shares a high degree of amino acid identity and have a previously demonstrated redundant function in an in vitro splicing assay (Fig. 2 9) Therefore, overexpression of Mbnl3 would potentially have downstream affects on not only its own targets, but on Mbnl1 and Mbnl2 as well. The lack of control C2C12 cell lines expressing either exogenous Mbnl1 or Mbnl2 prevent proper interpretation of this data. Second, the polyclonal antisera used for these studies was generated using the C terminal 83 amino acids of Mbnl3. The immunizing peptide shares a greater than 60% identify with Mbnl1 and Mbnl2 which greatly increases the chances that this polyclonal antisera reacts against multiple members of the Mbnl family. The Mbnl3 protein also runs at a higher molecular weight by SDS PAGE and
93 immunoblotting than anticipated, so there may be a cross reaction with Mbnl1 or Mbnl2. No further antibody validation studies were performed. Third, the Mbnl3 antisense morpholino experiment to knock down Mbnl3 4XC3H expres sion was inefficient and was not verified in any other experiments. In addition, the Mbnl3 morpholino antisense oligonucleotide is directed against only the full length Mbnl3 4XC3H isoform and does not target the Mbnl3 2XC3H isoform. To clarify these results, we inhibited Mbnl3 expression using siRNAs directed against the Mbnl3 coding sequence/3 UTR, which target all isoforms, and assayed for differentiation of C2C12 cells. Knockdown of Mbnl3 expression inhibited myogenic differentiation of C2C12 cells, which indicates that loss of Mbnl3 alone was adequate to delay myogenesis in vitr o This result is consistent with a delay in myogenesis seen in CDM patients. While it is not clear if the delay seen in C2C12 differentiation is sufficient to cause a delay in embryonic myogenesis in vivo it is important to note that loss of Mbnl3 expression in C2C12 does not promote premature differentiation as reported. While previous studies have established interactions of the Mbnl3 4XC3H isoform with (CUG)n repeat R NA, it is unclear whether the Mbnl3 2XC3H isoform is sequestered. Therefore, the Mbnl3 2XC3H isoform is a potential candidate in the RNA dominance model of CDM/DM. Interestingly, Drosophila Mbl proteins containing only 2XC3H motifs localize with (CUG)n repeat RNAs in vitro and in vivo suggesting the Mbnl3 N terminal truncated isoform may be important for disease pathogenesis. Loss of Mbnl3 4XC3H Isoforms are not Sufficient to Phenocopy CDM or Skeletal Muscle Wasting The MBNL family was originally proposed to be involved in the pathogenesis of myotonic dystrophy by sequestration and loss of protein function based on its length
94 dependent interaction with expanded (CUG)n RNA in vitro and in vivo (Miller et al., 2000) Mbnl3 demonstrates a spatially and temporally restricted expression pat tern in tissues affected in CDM and DM. Knockdown of Mbnl3 expression via siRNA in a C2C12 model of myogenic differentiation delayed myogenesis in vitro Therefore, we tested the hypothesis that CDM and adult onset muscle wasting is caused by sequestrat ion of Mbnl3 4XC3H isoforms, which have been previously shown to interact with (CUG)n repeat RNA. Although all Mbnl3 exon 2 containing isoforms where eliminated, Mbnl3E2/Y mice did not phenocopy the onset of CDM or demonstrate an impaired capacity for sk eletal muscle regeneration. Mbnl3E2/Y mice live to ~6 months of age and appear phenotypically normal at the time of this study. While this data does not rule out a lack of a molecular phenotype in Mbnl3E2/Y mice, it does provide evidence against the idea that loss of function of Mbnl3 4XC3H isoforms alone by toxic RNA sequestration result in CDM or skeletal muscle wasting. What other factors could be potentially involved in this disease? One possibility is that other proteins partially compensate for loss of Mbnl3 4XC3H isoforms in vivo including Mbnl1, Mbnl2, and upregulated Mbnl3 2XC3H isoforms. Alternatively, multiple Mbnl pathways may need to be comprised to promote the onset of CDM or skeletal muscle wasting. As (CUG)n repeats expand to >1000, they may titrate out all Mbnl proteins, and thus compromise both independent and compensatory functions. If so, Mbnl1 Mbnl2, and Mbnl3 conditional and knockout mice must be generated and crossed to test the hy pothesis that sequestration of the Mbnl family contributes to CDM and skeletal muscle wasting. One argument against this hypothesis is that MBNL family members can interact with both (CUG)n and (CCUG)n repeat RNAs
95 in vitro but only DMPK (CUG)>1000 promot e the onset of CDM (Fardaei et al., 2002) In fact, DMPK (CUG)n and CNBP (CCUG)n repeats can expand to > 4,000 and 11,000, respectively (Ranum and Day, 2002) Wh ile DMPK is expressed at its highest levels in the CNS and skeletal muscle, CNBP is expressed ubiquitously and at higher levels than DMPK ( Mankodi et al., 2003) This provides evidence for the idea that either (CUG)n repeat RNA is inherently more toxic than (CCUG)n or repeat expansions in the DMPK locus, alter neighboring gene expression and promote a locus specific contribution to diseas e However, a combinatorial approach using genetic crosses which eliminate Dmpk Six5 and Mbnl1 exon 3 containing isoforms failed to phenocopy CDM in a mouse model (Swanson lab unpublished data). We cannot, however, rule out that another closely linke d Dmpk gene, Dmwd is involved in disease. We have also previously demonstrated that MBNL1:(CUG)n RNA form more stable complexes than MBNL1:(CCUG)n RNA in a competition assay. This implies that there is an inherent difference in the interactions between MBNL proteins and (CUG)n/(CCUG)n RNAs. This observation may explain differences in disease manifestation between DM1 and DM2. In conclusion, CDM and DM are complex diseases that affect nearly every organ of the body. Our results provide evidence agains t a singular role for Mbnl3 4XC3H isoforms in sequestration and onset of CDM and adult onset skeletal muscle wasting. This study highlights the importance of developing a combinatorial genetic model of CDM to elucidate the molecular etiology of this disease. It will be intriguing to unfold the complex genetic contributions and pathways that underlie this disorder.
96 Figure 31 Phylogram shows evolutionary proximity of human C3H zinc finger motif containing proteins. Protein sequences were derived from EMBL EBI database and C3H pairs were analyzed by multiple alignment using MUSCLE (multiple sequence comparison by log expectation). Phylogenetic tree was created using PhyML ( phy logeny analysis using m aximum l ikelihood method ) Bootstrap confidence levels (red) measure robustness of support for a given clade. (0.95 indicates a 95% reproducibility in analysis). MUSCLE and PhyML located at www.phylogeny.fr.
97 Figure 32. The MBNL family is composed of three closely related paralogs. MBNL proteins have conserved C3H RNA binding domains (blue) clustered in two pairs. Percentages below the C3H domains indicate conserved amino acid identity with MBNL1 C3H. Percentages to the right of MBNL2 and MBNL3 indicate amino acid identity of the total protein w hen compared with MBNL1. The hyper variable linker region (green, yellow, red) is the most divergent sequence among the proteins.
98 Figure 33. Schematic of potential genetic contributions to DM1. (A) DMPK (CTG)exp inhibits transcription of closely linked genes DMWD and SIX5 (CTG)exp can act as a strong nucleosomal positioning element, leading to the formation of heterochromatin at the DM1 locus. Antisense transcription through the repeat promotes further chromat in condensation by recruiting heterochromatin associated factors. (B) DMPK (CUG)exp can inhibit or activate downstream protein function. The MBNL family of proteins are sequestered by (CUG)exp, impeding normal cellular functions. Additionally, increased CUGBP1 stability (through hyperphosphorylation) and activation of the RISC complex containing short CUG siRNAs may play roles in disease.
99 Figure 34. Mbnl3 expression is restricted spatially and temporally. RNA blot analysis at different murine developmental time points (embryonic day 15/18 = E15/E18; postnatal day 1 = P1). Coding sequence probes are used to detect Mbnl3, Mbnl1 (family control), and Ppia (loading control). Mbnl3 expression in E15 placenta is used as an exposure control between blots. Expected sizes of mRNAs; Mbnl3 (9 kb, 1.5 kb, 1.3 kb), Mbnl1 (6.5 kb, 5.3 kb), and Ppia (0.7 kb).
100 Figure 35. T he Mbnl3 gene produces multiple isoforms through alternative splicing. Mbnl3 isoforms were amplified from cDNA with primers flanki ng the coding sequence, subcloned, and sequenced. Tissues used, band excised for cloning, and annotation are included.
101 Figure 3 6 Mbnl3 is predicted to encode several different isoforms. Mbnl3 contains two pairs of RNA binding motifs (C3H) encoded by exons 2/3 and 5, as well as a linker region which is highly variable between family members. Exclusion of exon 2 in mature transcripts is predicted to encode a Mbnl3 N terminal trucated isoform with one pair of C3H binding motifs which utilizes an alte rnative initiation codon in exon 3.
102 Figure 37 Alternative splicing of Mbnl3 produces multiple isoforms throughout development. (A) The Mbnl3 locus. The Mbnl3 gene consists of 10 exons (boxes) and 9 introns (horizontal line), constitutive (black) and alternatively spliced exons (red), untranslated regions (open boxes) and coding regions (closed boxes). Alternative splicing of Mbnl3 exons 7a, 7b, and 7c produces a variable C terminus. (B) Exon 7c containing transcripts are predominant in all tiss ues examined. RT PCR of Mbnl3 at different murine developmental time points with a forward primer in constitutively spliced exon 3 and reverse primers in alternatively spliced exons 7a, 7b, and 7c.
103 Figure 38 Mbnl3 antisera recognizes two distinct bands in E15 placenta whole cell lysate. (A) Immunoblot analysis using sera from four rabbits (A D) immunized against Mbnl3 was used at a 1:500 dilution to screen for immunoreactivity in E15 placenta whole cell lysate. Preimmune sera is used as a control. Antisera from rabbits A and C were selected for further analysis. (B) Titration of the input E15 placenta lysate is used to compare affinity between the reactive anti sera A and C at a dilution of 1:1000.
104 Figure 39 Mbnl3 antisera does not cross react with other Mbnl proteins (A) Mbnl3 recognizes Mbnl3, but not Mbnl1 or Mbnl2, by immunoblotting. Myc tagged Mbnl proteins were exogenously expressed in COSM6 cells and immunoblotted with Mbnl3 (experimental), m yc (expression control) or Gapdh (loading control). (B) Mbnl3 antibody immunoprecipitates (IP) Mbnl3, but not Mbnl1 or Mbnl2. Myc tagged Mbnl proteins were exogenously expressed in COSM6 cells, immunoprecipitated with either Mbnl3 (experimental IP) or myc (control IP) antibodies, and immunoblotted with myc. As controls, 10% input was immunoblotted with myc (expression control) or Gapdh (loading control).
105 Figure 310. Mbnl3 distnguishes between Mbnl family members by immunocytochemis try Mbnl3 specifically recognizes Mbnl3 by immunocytochemistry, but not Mbnl1 or Mbnl2. Myc tagged Mbnl proteins were exogenously expressed in COSM6 cells and visualized with Mbnl3 (green) or myc ( red). DAPI (blue) stains DNA and indicates nucle ar location. Scale bar = 5 m.
106 Figure 3 11. Two distinct Mbnl3 isoforms are expressed in C2C12 and Huh7 cells. Immunoblot analysis of C2C12 and Huh7 cells with Mbnl3 showing that siRNA against Mbnl3 (C2C12) or human MBNL3 (Huh7) specifically kno cksdown expression of MBNL3, while non specific siRNA has no effect. Mbnl1 and SMN are used to assay for efficiency of MBNL3 siRNA and mediated knockdowns. GAPDH is used as a loading control.
107 Figure 3 12. MBNL3 localizes to nuclear and cytoplasmic foci. Human MBNL3 (green) is localized to cytoplasmic foci in Huh7 cells and these foci do not colocalize with known cytoplasmic structures such as P bodies, stained with GW182 (red). Mbnl3 (green) is distr ibuted in both the nucleus and cytoplasm in C2C12 cells, but does not colocalize with P bodies (red). DAPI (blue) stains DNA and indicates nuclear location. Scale bar = 5 m.
108 Figure 3 13. MBNL3 isoforms are found in different subcellular compartment s. Immunoblot analysis of nuclear/cytoplasmic fractionations for MBNL3 (experimental), MBNL1 (family control), CUGBP1 (nuclear fractionation control), and LDHA (cytoplasmic fractionation control). The MBNL3 27 kDa/29 kDa isoform (exon 2 exclusion) is fou nd exclusively in the cytoplasm in Huh7 (A), C2C12 (C), and murine E15 forelimb (B). MBNL3 37 kDa/39 kDa isoforms (exon 2 inclusion) are located in both the nucleus and cytoplasm in C2C12 (C) and E15 forelimb (B).
109 Figure 3 14. Mbnl3 expression is res tricted temporally and spatially during embryogenesis and postnatally. Murine tissues were taken at varying developmental time points and immunoblotted for Mbnl3, Mbnl1, and Gapdh. The band in adult brain is a cross reaction with glutamine s ynthetase (Gl uI).
110 Figure 315. Murine embryonic myogenesis. (A) Time line of myogenic developmental windows during gestation (E8 = embryonic day 8). (B) Schematic of embryonic myogenesis. Myoblast precursors originate in the dermomyotome and travel to the devel oping limb bud, proliferate, and fuse to generate myotubes. Single cells represent myoblast precursors/myoblast and elongated tubes represent myotubes. Expression of critical regulatory genes during these time points are indicated by the dashed line. DM = dermomyotome; NT = neural tube; NC = notochord.
111 Figure 3 16. Notexin promotes murine skeletal muscle necrosis followed by regeneration in vivo Timeline of Notexin injection and harvesting of the tibialis anterior (TA) muscle. (A) TA muscle was injected with Notexin and harvested at days 1, 3, 5, 7 post injection (red arrows). A preinjection time point was used as control. (B) TA muscles w ere harvested at indicated time points, divided for immunoblot (IB), cryosection, and RT PCR analysis. (C) H&E staining of 10 m transverse TA cryosections from control and post injection skeletal muscle. Scale bar = 50 m.
112 Figure 3 17. Mbnl3 is ex pressed during adult skeletal muscle regeneration. (A) Immunoblot analysis of time points during Notexin mediated regeneration with antibodies specific for Mbnl3, myogenin (Myog), Mbnl1, and Gapdh. (B) RT PCR of regenerating skeletal muscle with primer s positioned in constitutive exons. Ppia is used as a loading control. Sizes of amplicons are indicated.
113 Figure 318. Adult skeletal muscle regeneration. Quiescent satellite cells (located adjacent to the sarcolemma) express specific markers including M Cadherin. When the sarcolemma is injured or compromised (dashed line), satellite cells are activated. Activated satellite cells proliferate and contribute to both skeletal muscle regeneration as well as satellite cell renewal. Shortly after injury Hepatocyte growth factor (HGF) and Fibroblast growth factor ( FGF) activate quiescent satellite cells Fgf and Insulinlike growth factor (IGF) promote proliferation of myoblasts expressing MyoD/Pax7. Forty eight hours after activation, myoblasts contributing to regeneration downregulate Pax7 and commit to terminal differentiation, while myoblasts repopulating the satellite cell pool downregulate MyoD and enter quiescence. Abbreviations include: Act = activation, Pro = proliferation, Com = commitment, C CA = cell cycle arrest, Comp = completion, Hyp = hypertrophy.
114 Figure 3 19. Loss of Mbnl3 inhibits myogenic differentiation in a C2C12 in vitro mode l. C2C12 cells were either mock treated (control) or transfected with Mbnl3 siRNA (directed against the Mbnl3 coding sequence and 3 UTR). 24 hours post siRNA, C2C12 cells were induced to differentiate by switching to differentiation media (DM). Differentiation status was monitored by immunoblot using terminal marker myosin heavy chain (Mhc). Antibodies against Mbnl3 were used to monitor knockdown, while Mbnl1 and Gapdh served as family and loading controls, respectively.
115 Figure 3 20. Generation of a conditional Mbnl3 allele ( Mbnl3cond/Y) in ESCs (A) Mbnl3 exon 2 targeting strategy. Mbnl3 consists of 10 exons (boxes) and 9 introns (horizontal line); untranslated regions (open boxes) and coding sequence (closed boxes) as well as exon numbers are indicated. Arms of homology from a Kpn I (K) fragment (15.5 kb) were used to target Mbnl3 exon 2 with a conditional exon 2 flanked by loxP sites. A Neomycin resistance cassette ( Neo ) and HSV thymidine kinase cassette ( TK ) were used for positive and negative selection, respectively. Upon successful recombination, Mbnl3 exon 2 (f lanked by loxP sites) and two novel Kpn I restriction sites are introduced in the Mbnl3 locus. (B) Genomic Kpn I digestion followed by Sourthern blotting with probes 1 and 2 (probe positions inticated by dark grey rectangle) reveal successful recombination events (correct 5 recombination = 8.7 kb, incorrect 5 recombination = 9.6 kb; 3 recombination = 7.5 kb; wild types allele = 15.5 kb). Clone identification is indicated above each blot.
116 Figure 3 21. Cre mediated recombination of Mbnl3cond/Y allele removes exon 2 containing transcripts from ESCs (A) Mbnl3cond/Y allele; Kpn I restriction sites (K), probes 1 and 2 for Southern blotting (dark grey rectangles), Mbnl3 exon 2 (open box), neomycin resistance cassette ( Neo ). Mbnl3 exon 2 and Neo are removed during Cre recombination. (B) Southern blotting verifies removal of Mbnl3 exon 2 from the genome ( Mbnl3 conditional allele = 7.5 kb, Cre recombined allele = 6.5 kb). (C) Loss of Mbnl3 exon 2 transcripts from ESCs. RT PCR from mouse embryonic fibroblast (MEF) feeder cells, Mbnl3cond/Y ESCs, and Mbnl3E2/Y ESCs using primers positioned within Neo forward in Mbnl3 exon1reverse in Neo or forward in Mbnl3 exon 1reverse in Mbnl3 exon 8. Expected sizes are ~1000 bp for internal Neo ; 647 bp f or Neo forward Neo reverse; 1235 bp (+ exon 2)/ 988 bp ( exon 2) for Mbnl3 forward exon 1Mbnl3 reverse exon 8.
117 Figure 3 22. Mbnl3E2/Y mice lack full length Mbnl3 isoforms. (A) Exon 2 is absent from Mbnl3E2/Y mRNA. RT PCR of E15 forelimb and P1 lung with forward and reverse primers positioned in Mbnl3 exon 1 and exon 8; Mbnl3 exon 2 and exon 8; Ppia exon 3 and exon 4/5 (RT PCR control). Expected sizes are 1235 bp (+ exon 2)/ 988 bp ( exon 2) for Mbnl3 E1 E2 ; 1151 bp for Mbnl3 E2 E8; 152 bp for Ppia (B) Mbnl3 37 kDa isoforms are not present in Mbnl3E2/Y mice, while Mbnl3 27 kDa is upregulated. Immunoblot analysis of E15 forelimb and P1 lung with antibodies against Mbnl3, Mbnl1, and Gapdh.
118 Figure 3 23. P1 Mbnl3E2/Y mice do not display defects associated with CDM (A) Mbnl3E2/Y P1 pups are normal size at birth when compared to wild type and fail to demonstrate hypotonia, a cardinal phenotype in CDM. (B) Skeletal preps of hindlimbs from P1 wild ty pe and Mbnl3E2/Y mice (bone is stained with Alizarin Red; cartilage is stained with Alcian Blue). Bone deformaties are not present in Mbnl3E2/Y mice.
119 Figure 3 24. Mbnl3E2/Y neonates do not present with delayed m yofibers at P1. Transverse sections from extensor carpi radialis brevis muscle. Mbnl3E2/Y P1 skeletal muscle develops normally. H&E staining (muscle morphology), Wheat Germ Agglutinin (WGA, extracellular matrix), Myosin Heavy Chain (MHC, developmental marker). Scal e bar = 50 m.
120 Figure 3 25. Mbnl3E2/Y mice gain weight normally up to 10 weeks of age. Weights from wild type (n=26) and Mbnl3E2/Y mice (n=17) were determined once per week from 3 weeks to 10 weeks of age (weaning to adults).
121 Figure 3 26. Mbnl3E2/Y mice do not express full length Mbnl3 isoforms during adult skeletal muscle regeneration. (A) Tibialis anterior (TA) muscle was injected with Notexin and harvested at Days 1, 3, 5, 7 post injection (red arrows). A pre injection time point was used as co ntrol. (B) TA muscles were harvested at indicated time points, divided for immunoblot (IB), cryosection, and RT PCR analysis. (C) RT PCR of regenerating TA skeletal muscle with primers positioned in Mbnl3 exon 1exon 8 and exon 2exon 8 demonstrate loss of Mbnl3 exon 2 containing transcripts ( Mbnl3 E1 E8: exon 2 inclusion = 1235 bp; exon 2 exclusion = 988 bp. Mbnl3 E2E8 = 1151 bp). Primers are positioned in constitutive exons of Myogenin (regeneration marker), Mbnl1 (family control), and Ppia (loading control). (D) Immunoblot of regenerating TA muscle with antibodies against Mbnl3, Myogenin, Mbnl1, and Gapdh (loading control).
122 Figure 3 27. Normal skeletal muscle r e generation in Mbnl3E2/Y mice. (A) H&E staining of transverse TA cryosections ( 10 m ) from control and post Notexin injected skeletal muscle. (B) Wheat germ agglutinin (red) highlights the extracellular matrix during regeneration. DAPI (blue) indicates nuclear position. Scale bar = 50 m.
123 Figure 3 28. Mbnl3E2/Y normal skeletal muscle regeneration. Immunohistochemistry of transverse TA cryosections (10 m) stained with an antibody against desmin, an intermediate filament which is expressed in both myoblasts and differentiated skeletal muscle. Scale bar = 50 m.
124 CHAPTER 4 CONCLUDING REMARKS A ND FUTURE DIRECTIONS In this study, we explored the role of MBNL, a family of dsRNA binding proteins that are compromised by microsatellite expansions in the neuromuscular disorder, myotonic dystrophy. Loss of Mbnl1, in a mouse model, is sufficient to reproduce the majority of characteristic phenotypes of adult onset DM, including muscle pathology, cataracts, and a defect in alternative splicing (Kanadia et al., 2003a) We first examined the possibility that MBNL1 regulates the alternative splicing of exons affected in disease by directly interacting with specific cis elements in premRNAs and modulating the i nclusion or exclusion of that exon. Our results demonstrated that MBNL1 interacts with a motif directly upstream of cTNT exon 5, a fetal exon retained in DM, that is required to promote exclusion. Additionally, MBNL1 interactions with repeat RNAs show variable stabilities and toxicities, which may play a role in determining the severity of disease. Further understanding of these MBNL1 interactions with its RNA targets will be important for unraveling the complex alterations in alternative splicing and phenotypes associated with disease, as well as the contribution of other factors that potentially interact with repeat RNAs to promote DM. We also investigated the involvement of Mbnl3 in the onset of CDM and progressive muscle wasting. Our results show that Mbnl3 protein isoforms are restricted to tissues that are affected in CDM during embryogenesis and regenerating skeletal muscle. Importantly, knockdown of Mbnl3 expression in C2C12 cells demonstrated a delay in differentiation, consistent with the phenotype of immature skeletal muscle in CDM. This is in contrast to a previous report that a reduction in Mbnl3 expression promotes terminal differentiation. However, loss of Mbnl3 4XC3H
125 isoforms does not inhibit embryonic muscle development or regeneration of the TA muscle in adult mice. One key difference between these experiments involves the newly identified 2XC3H Mbnl3 isoform. The C2C12 delay in differentiation was dependent on the knockdown of both 4XC3H and 2XC3H isoforms by siRNA, while the Mbnl3E2/Y mouse still expresses 2XC3H isoforms. One possibility to explain this disease is that the 2XC3H isoforms can compensate for loss of the 4XC3H isoforms. This observation highlights the potential importance of the Mbnl3 2XC3H isoforms in development and predicts that loss of all Mbnl3 isoforms by sequestration are necessary to cause the CDM phenotype. As for the normal function of Mbnl3 in development, it will be interesting to establish the interacting RNAs and pathways that are governed by this gene. Unlike DM, very little is known concerning the molecular deficits in CDM, whether they be alternative splicing or other regulatory roles in RNA metabolism. While, the Mbnl3E2/Y mouse does not present with overt CDM or skeletal muscle wasting phenotypes, this does not mean that MBNL3 is not involved in molecular changes that contribute to disease. The unique localization pattern of Mbnl3 isoforms may provide insights into the normal function of these proteins and their dysfunction in disease. The abil ity to identify molecular phenotypes in DM/CDM mouse models are critical to identifying these changes in patients and developing necessary intervention or therapies. DM and CDM represent an interesting class of disease in which a mutation displays a gaino f function at multiple levels, potentially altering the expression and function of multiple genes in both cis and trans in a dominant manner (Fig. 33). This polygenic involvement likely contributes to the constellation of symptoms and
126 unpredictability that is characteristic of this neuromuscular disorder. However, the involvement of many genes allows us to separately model specific contributions of each factor involved to determine their specific contribution t o disease in vivo. While transgenic mice ex pressing toxic (CUG)exp RNAs or knockouts modeling loss of function of have provided valuable insights into disease and the pathways disrupted, a more combinatorial genetic approach may be necessary to unravel additional molecular mechanisms involved in th e onset of disease.
127 CHAPTER 5 MATERIALS AND METHOD S MBNL3 RNA Analysis RNA was isolated from cell culture, staged C57BL6/J murine embryos, postnatal day 1 pups, and adult mouse tissues with Tri Reagent (Sigma) according to the manufacturers protocol. For Northern blot analysis, approximately 15 g of total RNA was denatured in Glyoxal/DMSO buffer (8% deionized glyoxal, 60% DMSO, 12 mM PIPES, 36 mM Bis Tris, 1.2 mM EDTA, pH ~6.5) at 55 C for 1 hr immediately transferred to ice and resolved on a 1.2% agarose gel in 1X BPTE buffer (10 mM PIPES, 30 mM Bis Tris, 1 mM EDTA, pH ~6.5) for 2 hr at 100 V. Following electrophoresis, the gel was rinsed in d2H2O, treated with 50 mM NaOH for 20 min and neutralized in 20X SSC (pH 7) for 40 min RNAs were transferred to HybondN+ nylon membranes (GE Healthcar e) in 10X SSC using a neutral transfer by capillary action. Blots were crosslinked with a UV Stratalinker 1800 (Stratagene) with 120 mJoules and prehybridized for 2 hr in ExpressHyb Hybridization Solution (Clontech) at 68 C. Northern probes were generat ed by [ -32P] dCTP body labeling 50 ng of DNA template ( Mbnl1, Mbnl3 Ppia ) with Ready to Go DNA Labeling Beads dCTP (GE Healthcare) and purified with I llustra ProbeQuant G 50 Micro Columns (GE Healthcare) according to the manufactures protocol. DNA probe in 2 g/ L (1 mL total volume) Sheared Salmon Sperm DNA (Invitrogen) are denatured at 100 C for 10 min and added to blot/hybridization solution and hybridized overnight at 68 C. After hybridization, blots were washed once in 1X SSC, 0.1% SDS at room temperature on a shaker for 10 min followed by three washes in 0.5X SSC, 0.1% SDS at 65 C in a heated/shaking water
128 bath for 10 min each. Blots were exposed to Biomax Film and X Omatic Intensifying Screens (Kodak) at 80 C. For cDNA preparation, approximately 25 g of total RNA + 2.5 g oligo d(T)1218 primers (Invitrogen) were denatured at 65 C for 5 min /ice for 1 min and added to a final volume of 50 L in 0.2 mM dNTPs, 1X 1st Strand Buffer (Invitrogen) + 8 mM DTT + 100 U RNasin (Promega) + 500 U Superscript III reverse transcriptase (Invitrogen). Reverse transcription reaction conditions were 25 C for 5 min 50 C for 60 min 70 C for 15 min in a thermocycler. After reverse transcription, 2 U RNase H (Invitrogen) was added to the RT reaction and incubated at 37 C for 20 min RT PCR analysis was carried out using 2 L template cDNA (i.e. 1 g RNA input) in a 50 L reaction containing 1X High Fidelity buffer (5 Prime), 0.4 mM dNTPs, 30 pmol forward primer, 30 pmol reverse primer, and 2.5 U T riplemaster Taq (5 Prime). PCR reaction conditions were 96 C for 2 min (denaturation step), followed by 96 C for 30 sec 58 C for 30 sec 72 C for 1 min (amplification step see below for cycle numbers per primer set), followed by 72 C for 5 min ( final elongation step). Cycle numbers for amplification step (per primer pair) were as follows: 30X for MSS3655MSS3759, MSS3655MSS3760, MSS3655MSS3763, MSS3586MSS3587, MSS3225MSS3247, MSS3648MSS3247, MSS35803581, MSS3582MSS3583, MSS3584MSS3585; 32X for MSS36483652, MSS29163652, MSS4173MSS4177, MSS27252726 (see Figure 41 for primer identification, position, and sequence). PCR products were resolved on a 1% 1.5% agarose gel, stained with ethidium bromide, and pictures taken on the ImageQuant 400 (GE Healthcare).
129 Protein Lysate and Fractionation Whole cell lysate was prepared from cell culture, staged C57BL6/J murine embryos, postnatal day 1 pups, and adult mouse tissues by disruption with a sterile pestle in lysis buffer (20 mM HEPES KOH, pH 8 .0, 100 mM KCl, 0.1% Igepal, 0.5 M phenylmethylsulfonyl fluoride, 5 g/mL pepstatin A, 1 g/mL chymostatin, 1 M e aminocaproic acid, 1 M p aminobenzamidine, 1 g/mL leupeptin, 1 g/mL aprotinin) on ice. Protein samples were sonicated three times at 40% amplitude and spun at 16.1 RCF for 20 min at 4 C. Supernatant was collected, glycerol was added to a final concentration of 20%, and samples were aliquoted and stored at 80 C. Protein concentrations were determined using the DC Protein Assay (BioRad). Cellular fractionations were prepared using NE PER (Thermo Scientific Pierce) nuclear and cytoplasmic protein fractionation kit according to the manufacturers protocol. Immun obloting 50 g total protein or 35 L immunoprecipitation eluate was resolved on a 12.5% SDS acrylamide gel and transferred to 0.22 m nitrocellulose (GE Water & Process Technologies) by electroblotting (using a Trans Blot SD semi dry transfer cell; BioR ad). Blots were blocked for 45 min in 5% nonfat dry milk in PBS (pH 7.4) at room temperature and immunoblotted in 5% nonfat dry milk in PBS (pH 7.4) with 0.05% Igepal for 1 hr rotating at room temperature with the following primary antibodys: Mbnl3 pur ified antisera C (1:2500 dilution), Mbnl1 (A2764) antisera (1:1000 dilution), CUGBP1 (3B1), c myc (9E10), Myosin Heavy Chain (Sigma MY32 1:1000 dilution), Myogenin (Santa Cruz F5D 1:1000 dilution), Lactate Dehydrogenase A (Cell Signaling 1:1000), Gl utamine Synthetase (Abcam 1:1000), SMN (Santa Cruz 2B1 1:1000
130 dilution), or Gapdh (6C5 Abcam 1:25,000 dilution). Blots were washed three times in PBS (pH 7.4) with 0.05% Igepal for 10 min and incubated for 1 hr rotating at room temperature in blott ing milk with rabbit IgG horseradish peroxidase or mouse IgG horseradish peroxidase (1:5000; GE Healthcare) secondary antibody. The membranes were washed as described above, followed by a PBS (pH 7.4) wash, developed in Amersham ECL detection reagent s (GE Healthcare) or Supersignal West Femto reagents (Thermo Scientific Pierce) and exposed to Biomax Light Film (Kodak). To develop immunoprecipitation/immunoblots, Genscript onestep complete IP Western kit (Genscript) was used according to the manufacturers protocol and exposed to Biomax Light Film (as described above). Generation of an Mbnl3 Polyclonal Antibody Mbnl3 polyclonal anti sera was generated and purified by Genscript (Piscataway, NJ). In short, Mbnl3 C terminal peptide (NVPYVPTTTGNQLKY) w as synthesized and KLH conjugated to the N terminus using an additional N terminal cysteine (final immunizing peptide: KLH CNVPYVPTTTGNQLKY). Four rabbits were injected with Mbnl3 immunizing peptide followed by two additional injections to boost antibody production/affininty. Rabbits were bleed prior to the first immunization for a preimmune sera control for screening. Test bleeds were screened by immunoblot using 50 g E14 placenta whole cell lysate and anti sera at a dilution of 1:500 (immunoblot and whole cell lysate protocols were performed as described in the immunoblot section of materials and methods). Anti seras A and C, with high specificity and low background, were further screened (at a dilution of 1:1000) by immunoblot using decreasing amount of input E14 placenta whole cell lysate (i.e. 50 g, 25 g, 12.5 g, 6.25 g, 3.13 g, and
131 1.56 g) to compare relative affinities. Anti sera A and C were chosen for further affinity purification using the original immunizing peptide. Purified peptides were concentrated using centricon 10 concentrators (Micon) according to the manufacturers protocol. Glycerol was added to a final concentration of 20% to Mbnl3 A and Mbnl3 C (final antibody concentrations: Mbnl3 A = 7.8 mg/mL, Mbnl3 C = 4.8 mg/ mL) polyclonal antibodies and stored in aliquots at 80 C. Expression Vectors Mbnl1, Mbnl2, Mbnl3 coding sequences were amplified using primers MSS3580MSS3581, MSS3582MSS3583, and MSS3584MSS3585 from MEF cDNA in a 50 L reaction containing 2 L cDNA 1X High Fidelity buffer (5 Prime), 0.4 mM dNTPs, 30 pmol forward primer, 30 pmol reverse primer, and 2.5 U Triplemaster Taq (5 Prime). PCR reaction conditions were 96 C for 2 min (denaturation step), 96 C for 30 sec 58 C for 30 sec 72 C for 1 mi n 30 sec (30X cycles amplification), followed by 72 C for 5 min (final elongation step) (see Figure 41 for primer identification, position, and sequence). PCR products were phenol:chloroform extracted, precipitated, and digested in a 40 L reaction containing 1X NEB buffer 3, BSA (100 g/mL), BamHI (40 units NEB), and XbaI (40 units NEB) for 2 hr at 37 C. Following digestion, amplicons were resolved on a 1.0% agarose gel, gel extracted (Qiagen) following the manufacturers protocols, and resuspended in 50 L water. 5 g pcDNA 3.1/myc His MCS A (Invitrogen) was digested in a 50 L reaction containing 1X NEB buffer 3, BSA (100 g/mL), BamHI (50 units NEB), and XbaI (50 units NEB) for 2 hr at 37 C, resolved on a 1.0 % agarose gel, and gel extracted (Qiagen) following the manufacturers protocols, and resuspended in 50 L water. pcDNA 3.1/myc His and Mbnl inserts were
132 ligated in a 20 L reaction containing 1X T4 DNA ligase buffer (NEB), T4 DNA ligase (5X105 units NEB), 12 ng DNA backbone (pcDNA), and 36 ng DNA insert (Mbnl) for 1 hr followed by transformation into Top10 cells (Invitrogen) and plating according to the manufacturers protocol. Individual colonies were picked, grown in 150 m Ls LBamp overnight at 37 C shaking, and maxiprepped (Qiagen) according to the manufacturers protocol. Expression vectors were sequenced (University of Florida ICBR sequencing core) for confirmation. Immunoprecipitation and Mass spectrometry Protein lysat es for immunoprecipitation were prepared from cells or E15 placenta in IPP 150 with 0.1% Igepal (50 mM Tris Cl, pH7.4, 150 mM NaCl, 0.1% Igepal, 1X PicD, PicW) by disruption with a sterile pestle in a microfuge tube on ice, followed by sonicating three tim es at 40% amplitude, and spinning down the lysate at 16.1 RCF for 20 min (4C). Supernatant was transferred to a new microcentrifuge tube on ice and protein concentration was determined using DC protein assay (BioRad). Lysate was left on ice until the pre paration of beads was completed (see below). For bead preparation, ProteinA Dynabeads were vortexed for 30 sec and 100 Ls were added to a microcentrifuge tube. The microcentrifuge tube was placed on the Dynal magnet (Invitrogen) for 1 min to remove the ProteinA Dynabeads from solution, followed by aspiration of the supernatant. The microcentrifuge tube was removed from the magnet and ProteinA Dynabeads were washed with 500 L 0.1 M sodium phosphate buffer (pH 8) three times by repeating the above steps. Next, the beads were resuspended in 100 L 0.1 M sodium phosphate buffer (pH 8.0) followed by the addition of 20 g purified Mbnl3 C polyclonal antibody and incubation for 30 min at RT
133 with rotational mixing to promote IgG capture. Antibody bound beads were washed two times with 0.2 M triethanolamine (pH 8.2) as described above, resuspended, and incubated in 1 mL of 20 mM dimethyl pimelimidate/0.2 M triethanolamine (pH 8.2) (DMP Sigma) for 30 min at RT to covalently crosslink Mbnl3 C antibody to proteinA. The crosslinking reaction was quenched by replacing DMP/triethanolamine with 1 mL 50 mM Tris (pH 7.5) and incubating for 15 min at RT with rotational mixing. Beads were further washed with IPP 150 + 0.1% Igepal three times as described above. For immunoprecipitation, ProteinA: Mbnl3 C IgG beads were resuspended in the 5 mgs of previously prepared protein lysate (~ 400 L total volume) and incubated at 4C for 2 hr with rotational mixing. Beads were collected by placing the microfuge tube o n the magnet for 1 min followed by a flash spin (5 sec 16.1 RCF) to collect them at the bottom of the tube. The beads containing the immunoprecipitate were resuspended in 35 L Laemmli buffer, incubated at 99 C for 5 min on an Eppendorf Thermomixer ( 1100 RPM), followed by removal of the beads from solution using the magnet as previously described. The supernatant was collected for downstream applications, immunoblotting and mass spectrometry. To identify the proteins immunoprecipitated using the Mb nl3 C polyclonal antibody, proteins were resolved on a 12.5% SDS acrylamide gel and stained with the Novex colloidal coomassie blue kit (Invitrogen) according to the manufacturers protocol. Bands of interest were excised from the gel using separate steril e scalpels (BardParker) and stored in separate microcentrifuge tubes at 4C. Samples were submitted to the UF ICBR Proteomic Core facility for liquid chromatography mass spectrometry (LCMS) analysis to identify proteins.
134 siRNA and Plasmid Transfections C2 C12 and Huh7 cells were grown on 10 cm plates in growth media containing DMEM (Invitro gen), 20% FBS (Invitrogen), 1% penicillin/s treptomycin (Invitrogen) and DMEM (Invitrogen), 10% FBS (Invitrogen), and 1% penicillin/s treptomycin (Invitrogen), respectively in a humidified 37 C, 5% CO2 incubator. For siRNA treatment, 1.5 x 105 cells were seeded per well of a 6well plate (Sarstedt) in 2 mLs antibiotic free growth media. Cells were treated with Mbnl3 or MBNL3 siRNA (ONTARGETplus SMARTpool, Dharamcon) 6hr post seeding as follows. Two separate tubes were prepared for siRNA and transfection reagent (Dharmafect 4 Dharmacon). For transfection of one well of a 6 well plate, 5 L of transfection reagent was mixed with 195 L OptiMEM I (Invitrogen) in tube 1. 1 0 L of 20 M siRNA was diluted to 1 M by addition of 90 L of cell culture grade PBS (pH 7.4 Invitrogen) and 100 L of OptiMEM I in tube 2. Tubes 1 and 2 were briefly vortexed (1 sec) and incubated for 5 min at room temeperature (RT). The contents of tube 2 were added to tube 1, vortexed briefly (1 sec), and incubated for 20 min at RT. The entire 400 L of transfection mixture was added dropwise into one well of 6 well plate and swirled to mix. Media was replaced by fresh C2C12 antibiotic free growth media 24 hr post transfection. COSM6 cells were grown in growth media containing DMEM (Invitrogen), 10% FBS (Invitrogen), 1% L glutamine (Invitrogen), and 1% penicillin/streptomycin (Invitrogen) in a humidified 3 7 C, 5% CO2 incubator.. Six hr prior to transfection, 2.0 x 105 cells were seeded per well of a 6well plate in antibiotic free growth media. In a sterile microfuge tube, 6 L Fugene 6 (Roche) was diluted in 180 L OptiMEM I (Invitrogen), vortexed (1 se c) and incubated for 5 min at RT. After incubation, 2 g of
135 plasmid (either myc tagged Mbnl1, Mbnl2, Mbnl3, or empty vector pSP72) was added to the tube containing Fugene 6:OptiMEM I, vortexed (1 sec), and incubated for 15 min at RT. Following incubation, the total contents of each transfection were added to individual wells of the 6well plate, swirled to mix, and returned to the humidified 37 C, 5% CO2 incubator. The following day, cells were washed with PBS (pH 7.4) and antibiotic free growth media w as replaced. Cell lysates were processed 48 hr post transfection for analysis. C2C12 differentiation For C2C12 cell differentiation, proliferating cells were seeded at 1.5 x 105 per well of a 6 well plate in C2C12 growth media. At ~ 9095% confluency, growth media was aspirated, cells were washed twice with 2 mLs of PBS (pH 7.4), and fresh C2C12 differentiation media containing DMEM, 2% horse serum (Invitrogen), and1% Penicillin/ Streptomycin was added to each well. Cells were differentiated for 5 days (media was replaced with fresh C2C12 differentiation media every 24 hr ). For Mbnl3 siRNA experiments, C2C12 cells were differentiated 24 hr post siRNA treatment. Skeletal Muscle Rege neration Lyophilized Notexin (Latoxan Valence, France) was reconstituted in sterile 154 mM NaCl to a final concentration of 100 g/mL. 10 L aliquots (1 g) were stored at 80 C until use. For injections, 1X 10 L aliquot was thawed on ice and sterile 154 mM NaCl was added to a final volume of 1 mL (working concentration is 1 g/mL). For injections, 1012 week old C57BL6 mice were anesthetized with isoflourene (following the University of Florida IACUC approved protocol). Using a 29gauge needle, tibialis anterior (TA) muscles were injected with 100 L (100 ng Notexin) of working
136 concentration Notexin by lateral injection and dragging the needle along the muscle lengthwise (proximal to distal) while injecting. Mice were allowed to recover in their ca ges while monitoring their behavior. TA muscles were taken at days 1, 3, 5, and 7 post injection for analysis. TA muscle was divided in two lengthwise using surgical scissors, half was used for protein preparation and analysis (following whole cell lysat e and immunobloting protocols) and half was snap frozen on a wooden dowel positioned in 10% gum tragacanth (in PBS, pH 7.4 Sigma) in prechilled isopentane ( 40 C) and stored at 80 C in isopentane (Fisher). TA muscle remaining after cryosectioning was used for RT PCR analysis (see RNA Analysis). Mbnl3 E2/Y Mouse Generation The Mbnl3 conditional targeting vector was constructed using standard recombineering bacterial strains and techniques (reagents and protocols can be found at http://recombineering.ncifcrf.gov/ ). In short, an approximately 11 kb fragment (containing 6.8 kb sequence from directly upstream and 4.2 kb sequence from directly downstream of Mbnl3 exon 2) surrounding Mbnl 3 exon 2 was retrieved from a C57BL6 bacteria artificial chromosome (BAC) into a high copy plasmid backbone containing the negativeselection marker Herpes Simplex Virus Thymidine Kinase gene ( HSV TK ) downstream of the Mbnl3 fragment. The Mbnl3 exon 2 was flanked by an upstream loxP site and a downstream NeoR cassestte/loxP site (see Fig. 316). The final targeting construct was sequenced and restriction mapped to confirm its identity. To prepare the targeting construct for ESC targeting, approximately 100 g of the Mbnl3 conditional targeting vector was linearized with NotI (NEB), phenol:chloroform extracted,
137 ethanol precipitated, and resuspended in PBS (pH 7.2) at a final concentration of ~2.2 g/ L. All C57BL/6 ESCs (Millipore, catalog number CMTI 2) in this targeting protocol were thawed, grown, and passaged on mitotically inactivated MEFs using standard ESC techniques in a 37 C, 5% CO2 humidified incubator. ESC growth media contained: 409 mLs DMEM (Invitrogen), 75 mLs Defined FBS (HyClone), 5 mLs 100X nucleoside stock (standard ESC formulation, see Millipore recommendations), 100X Pen/Strep (Invitrogen), 0.76 mLs mercaptoethanol (Invitrogen 55 mM stock concentration), and 50 L LIF (Millipore 107 units/mL). For targeting, 108 ESCs were electr oporated in 0.8 mLs PBS (pH 7.2) containing 100 g Mbnl3 conditional targeting vector using a BIORAD Genepulser (parameters = 0.8 volts, 3 F, 200 Ohms, and 0.2 sec time constant in a 0.4 cm cuvette), allowed to recover for 10 min in the cuvette, and plate d on 15X 10 cm gelatin coated tissue culture plates in ESC growth media. The following day, ESCs were positively selected for neomycin resistance in ESC growth media supplemented with G418 (Invitrogen) at 275 g/mL. Two days post electroporation, ESCs we re positively and negatively selected for successful recombination in ESC growth media supplemented with G418 at 275 g/mL and FIAU (Moravek Biochemicals) at 0.2 M for 72 hr Following selection, individual colonies were picked with a P200 pipet tip, try psinized, and plated in 24well plate for colony expansion. Following expansion, individual wells of the 24well plate were trypsinized and split, with half of the ESCs transferred to 0.5 mL freezing media in cryotubes, containing 60% growth media, 30% FB S, and 10% DMSO (Fisher Scientific), frozen in an isopropanol bath at 80 C overnight, and transferred to N(l) the following day. The other half of ESCs from
138 individual wells of 24well plates were transferred to wells of gelatin coated 12well plates, and grown to confluency for genomic DNA preparation and Southern blot analysis for successful recombinants (see genomic DNA isolation and Southern blotting). Once positive ESC clones were identified, the targeted ESCs were thawed, expanded, and additional s tocks were cryopreserved. Mbnl3cond ESCs chosen for the generation of mice were screened for murine pathogens using the IMPACT I profile at the Research Animal Diagnostic Laboratory (RADIL) at the University of Missouri and karyotyped at the University of Florida Cytogenetics Lab. Mbnl3cond ESCs that tested negative for pathogens and displayed normal chromosome counts were sent to the University of Michigan Transgenic Core for blastocyst injection. In short, 45 B6(Cg) Tyrc 2J/J blastocysts (i.e. albino C57BL/6) were injected with 816 ESCs and transferred in utero to a multiple pseudopregnant C57BL/6 females. The resulting chimeric mice were shipped to the University of Florida through, quarantined in isolators for pathogen testing, and released into s pecific pathogen free (SPF) housing. To obtain germline transmission of the Mbnl3cond allele, Mbnl3cond male chimeric mice were mated with female B6(Cg) Tyrc 2J/J females. Since Mbnl3 is an X linked gene, transmission through the chimeric male germline only appears in F1 females. Therefore, F1 female pups (with a black coat, indicating the gamete originated from BL6 ESCs) from the Mbnl3cond male chimera and female B6(Cg) Tyrc 2J/J cr oss possessed a Mbnl3cond/+ genotype. To eliminate Mbnl3 exon 2 and the 4XC3H Mbnl3 isoforms, Mbnl3cond/+ females were crossed to B6.C Tg( CMV cre )1Cgn/J males (Jackson Labs, catalog number 006054) to obtain F2 Mbnl3E2/+; CMV cre females ( CMV Cre is also X -
139 linked). To obtain knockout mice, F2 Mbnl3E2/+; CMV cre females were crossed to wild type C57BL6/J males, F3 Mbnl3E2/Y males were analyzed for phenotypes. Genomic DNA Isolation, Southern Blotting, Probe Generation and Genotyping Targeted ESC genomic D NA was isolated from individual confluent wells of a 12well plate for Mbnl3cond/Y homologous recombination screening. In short, individual wells were washed in PBS (pH 7.2 Invitrogen) to remove debris and lysed in 0.5 mLs DNA extraction buffer (200 mM NaCl, 100 mM Tris (pH 8.5), 5 mM EDTA, 0.2% SDS, 200 g/mL Proteinase K) overnight in a humidified 37 C, 5% CO2 incubator. The following day, lysates were collected and precipitated by adding an equal volume of 100% ethanol, inverting several times until the genomic DNA was visible, and spinning samples down at 16.1 RCF for 10 min in a table top centrifuge (room temperature). DNA pellets were washed in 1 mL 70% ethanol and centrifuged as before. The supernatant was removed and the genomic DNA pellet was allowed to air dry for one min followed by resuspension in 100 L TE buffer (pellets were not vortexed to avoid shearing) by pipeting with a wide bore tip. ESC Mbnl3 conditional targeting 5 (MSS3256 MSS3272) and 3 (MSS3258MSS3259) southern probes were generated by PCR in a 50 L reaction containing 100 ng 129 genomic DNA, 1X High Fidelity buffer (5 Prime), 0.4 mM dNTPs, 30 pmol forward primer, 30 pmol reverse primer, and 2.5 U Triplemaster Taq (5 Prime). PCR reaction conditions were 96 C for 2 min (denaturation step), 96 C for 30 sec 68 C for 1 min 15 sec (30X cycles amplification), followed by 68 C for 5 min (final elongation step) (see Figure 41 for primer identification, position, and sequence). PCR products were subcloned using the TOP O cloning kit (Invitrogen), grown in 150 mLs
140 LBamp at 37 C shaking overnight, and maxiprepped (Qiagen) according to the manufacturers protocols. Cloned products were sequenced (University of Florida ICBR sequencing core) for confirmation (see Figure 41 for primer identification, position, and sequence). For Southern blotting, 20 L genomic DNA (~1020 g) was digested in a 30 L reaction containing 150 units KpnI (New England Biolabs NEB), 100 g/mL BSA, and 1X NEB buffer 1 overnight in a 37 C water bath. The following day, the digestion was spiked with 10 L digestion buffer containing 50 units KpnI, 100 g/mL BSA, and 1X NEB buffer 1 for 6 hr in a 37 C water bath. Samples were then resolved on a 1X TAE 0.8% agarose gel overnight at 30V, stained in 0.5 g/mL ethidium bromide for 30 min (post staining, pictures were taken on the ImageQuant 400 (GE Healthcare) UV box with a ruler to accurately determine sizes), denatured for 1 hr in 0.5M NaOH/1.5M NaCl (pH ~12), neutralized for 1 hr in 0.5M Tris/1.5M NaCl (pH 7.4). DNA was transferred to HybondN+ nylon membranes (GE Healthcare) in 10X SSC using a neutral transfer by capillary action. Blots were crosslinked with a UV Stratalinker 1800 (Stratagene) with 120 mJoules and prehybridized for 2 hr in Expr essHyb Hybridization Solution (Clontech) at 68 C. Southern probes were generated by [ -32P] dCTP body labeling 50 ng of DNA template (5 probe 1 and 3 probe 2) with Ready to Go DNA Labeling Beads dCTP (GE Healthcare) and purified with illustra ProbeQu ant G 50 Micro Columns (GE Healthcare) according to the manufactures protocol. Body labeled probes are denatured at 100 C for 10 min and added to fresh blot/hybridization solution and hybridized overnight at 68 C. After hybridization, blots were washed once in 2X SSC, 0.1% SDS at 50 C, once in 2X SSC, 0.1% SDS, twice in 0.5X SSC, 0.1% SDS, and
141 once in 0.05X SSC, 0.1% SDS (for 30 min at each wash). Blots were exposed to Biomax Film and X Omatic Intensifying Screens (Kodak) at 80 C. Mbnl3E2/Y mice were genotyped using 100 ng genomic DNA template in a 50 L reaction containing 1X High Fidelity buffer (5 Prime), 0.4 mM dNTPs, 30 pmol forward primer (MSS3377), 30 pmol forward primer (MSS4153), 30 pmol reverse primer (MSS3366), and 2.5 U Triplemaster Taq (5 Prime). PCR reaction conditions were 96 C for 2 min (denaturation step), 96 C for 30 sec 68 C for 1 min 30 sec (40X cycles amplification), followed by 68 C for 5 min (final elongation step) (see Figure 41 for primer identification, position, and sequence). PCR products were resolved on a 1.5% agarose gel, stained with ethidium bromide, and pictures taken on the ImageQuant 400 (GE Healthcare). Expected sizes of bands: wild type Mbnl3 allele = 350 bp, Mbnl3E2 allele = 597 bp. Skeletal Prepar ations Postnatal day 1 mouse pups were asphyxiated with CO2, skinned and eviscerated. Cartilage was stained by incubating embryos in 0.2 g/L Alcian Blue (in 70% EtOH/30% acetic acid) overnight at room temperature with gentle agitation. The following day embryos underwent serial rehydration in 100%, 95%, 70% and 40% EtOH and water for 30 min each with gentle agitation. Bone was stained by in 0.1% w/v Alizarin Red (in 1% KOH) overnight at room temperature. Stained embryos incubated for 1 hr each in 25% glycerol/75% KOH, 50% glycerol/50% KOH, 75% glycerol/25% KOH, and 80% glycerol to clear the embryo. Stained and cleared embryos were stored in 80% glycerol and photographed using a Leica stereoscope.
142 Wild Type and Mbnl3 E2/Y Growth Curve Wild Type and Mbnl3E2/Y male mice were weighed at one week intervals from 3 weeks (weaning) to 10 weeks to assay for postnatal growth (wild type n = 26, Mbnl3E2/Y n = 17). Results were expressed as mean weight for each group SEM at intervals of one week. Statistic al significance was established using a one tailed Students T test with unequal variance. Sectioning, Immunoflourescence, and H&E Staining Forelimbs from postnatal day 1 pups were fixed overnight in 4% formaldehyde in PBS (pH 7.4) at 4 C with agitation. The following day, forelimbs were washed three times (10 min each) in PBS (pH 7.4) and serially dehydrated in 70% ethanol (3X 20 min ), 95% ethanol (3X 20 min ), 100% ethanol (3X 20 min ), CitriSolv (3X 20 min Fisher Scientific). Tissues were then placed in preheated (65 C) paraffin wax (Fisher Scientific) for 3X 1 hr in a 65 C vacuum (1015 mm Hg). Forelimbs were then placed perpendicular (with the foot oriented towards the bottom of the mold) in paraffin molds and allowed to solidify overnight at roo m temperature. For sectioning, 7 m sections were take on a rotary microtome (Leica) and mounted on Superfrost slides (Fisher) and incubated at 65 C for 1 hr in a hybridization oven to remove the paraffin and attach the sections to the slide. Sections w ere then rehydrated by serial incubations in CitriSolv, 100% ethonaol, 95% ethanol, 70% ethanol and d2H2O for 10 min each. At this point, sections were further processed for H&E or immunohistochemical analysis. TA muscle from skeletal muscles from Notexin regeneration experiments were cut (10 m) on a cryostat (Leica) and attached to Superfrost slides (Fisher) and stored at 20 C. For immunohistochemical analysis, sections were thawed at RT for 30 min fixed
143 in 2% formaldehyde in TBS (pH 7.5) at RT for 10 min rinsed in TBS (pH 7.5) and permeablized in 0.1% TritonX 100 in TBS (pH 7.5) for 10 min At this point, sections were further processed for H&E or immunohistochemical analysis. For H&E analysis, sections were stained in Harris Hematoxylin (diluted 1:2 in water Fisher), rinsed in running dH2O, and dehydrated in 95% ethanol for 1 min each. Sections were then stained in EosinY (Thermo Scientific) for 30 sec, followed by 3X 1 min 95% ethanol 3X 2 min 100% ethanol, 3X 5 min CitriSolv dehydrations s teps and mounted in Permount (Fisher). Pictures were taken on a Leica DM 2000 light microscope. For immunoflourescence analysis, sections were processed using the M.O.M (mouse on mouse) kit according to the manufacturers protocol. C2C12 cells and Huh7 c ells were grown on 2well Falcon slides (BD Biosciences), washed in TBS (pH 7.5), fixed in 2% formaldehyde in TBS (pH 7.5) for 10 min at RT, washed 3X 5 min in TBS (pH 7.5), and blocked/permeablized in 3% heat inactivated goat serum (Invitrogen), 0.1% Trit onX 100 in TBS (pH 7.5) for 30 min at RT. Primary antibodies (in blocking/permeablization buffer) were incubated overnight at 4 C in a humidified chamber. The following day, section were washed 3X 5 min in TBS (pH 7.5), incubated in secondary antibody ( in blocking/permeablization buffer) for 2 hr at RT in a humidified chamber, and washed 3X 5 min in TBS (pH 7.5). All sections were counterstained with DAPI mounting media and pictures were taken on a Leica confocal microscope. Antibodies used for staining include, Mbnl3 purified C (1:1000), myc (1:1000 9E10), MHC (1:1000 Sigma), Desmin (1:100 Abcam), GW182 (1:6000 Gift of Edward Chan). Secondary antibodies (used at 1:200) include rabbit IgG Alexaflour 488 (Mbnl3), mouse IgG Alexaflour 488 (Desmin, MHC), mouse IgG Alexaflour 568
144 (myc), human IgG Alexaflour 568 (GW182). For staining of the extracellular matrix, TRITC conjugated WGA (Sigma) was incubated at RT for 1 hr (50 g/mL) in TBS (pH 7.5). MBNL1 Transfections and RNA Analysis For actinin minigene analysis, COSM6 cells were grown to 40% 60% confluency in DMEM (Invitro gen), 10% FBS (Invitrogen), 1% penicillin/s treptomycin (Invitrogen) at 37 C 5% CO2 and w ere co transfected with 2 g of Drosophila actinin minigene and 100 ng of G FP Mbl isoforms in the pMV vector using Lipofectamine 2000 (Invitrogen) and Optimem ( Invitrogen) according to the manufacturers protocol. F our hr after transfection, the media was changed to antibiotic free DMEM media supplemented with 10% FBS (Invitrogen) T o analyze the Drosophila actinin splicing pattern, t otal RNA was extracted from transfected COSM6 cells 48 hr after transfection with Tri Reagent (Sigma) according to the manufacturers protocol. For RT, 5 g of total RNA was t reated with DNAse I (Invitrogen) and RT was performed with Superscript II (Invitrogen), random hexamers and oligo dT12 18 (Invitrogen) following instructions from the manufacturer (Invitrogen) PCR analysis was done with 4 L of the RT reaction as template in a standard PCR reaction. To detect spliced isoforms arising from the Drosophila actinin minigene we used primers MSS 1938 and MSS1956 (see Figure 41 for primer identifica tion, position, and sequence). PCR products were purified by NH4Ac precipitation and resuspended in d2H2O. H alf of the resuspended volume was digested with SacI (NEB) The remaining PCR products, and entire digestions, were resolved on a 2% agarose gel stained with ethidium bromide, and pictures were taken
145 on the ImageQuant 400 (GE Healthcare). Figure 2 4 shows representative blots from three independent experiments For GFP Mbl localization experiments, COSM6 cells were transfected with either 1 g of plasmids expressing GFP tagged Mbl proteins and 1 g carrier DNA (pSP72) or 1 g plasmid expressing DMPK (CUG)300 and 1 g plasmid expressing GFP tagged Mbl pro teins as previously described. For cTNT IR and ClaLC splicing analysis, COSM6 or HEK293 cells were plated at 5X105 cells per well in a six well plate in DMEM (Invitrogen) supplemented with 10% FBS (Invitrogen) and 1X penicillin/ streptomycin (Invitrogen). A t 24 hrs after plating, the cells were transfected with either 1 g of minigene and 2 g of protein expression plasmid or (CUG)960 or (CAG)960 repeat expressing plasmid using Fugene 6 (Roche) according to the manufacturers directions T he total amount of plasmid transfected per well ( pSP72 plasmid was used as a carrier to bring the total DNA to 2 g). P rotein and RNA were harvested 36 48 hr after transfection. Chicken primary muscle cultures were prepared, maintained and transfected as previously described, using 0.5 g minigene reporter and 1 g expression plasmid (Xu et al., 1993) RNA isolation and RT PCR analysis for the cTNT IR and ClaLC minigenes were performed as described previously (Philips et al., 1998; Savkur et al., 2001; Stamm et al., 1999) In Vitr o Transcription and UV Crosslinking Uniformly 32P labeled RNAs were in vitro transcribed using [ -32P]GTP and [ -32P] UTP (Perkin Elmer) from PCR products or cloned regions of the human cTNT introns 4 and 5, as represented in Figure 28 UV crosslinking assays were performed using radiolabeled transcripts standardized for picomoles of G and U. UV crosslinking
146 assays included 1 g of purified GST MBNL1 in the presence of 1 g BSA, 1 g tRNA, 0.3 g heparin, 0.3 mM magnesium acetate, in 2 mM magnesium acet ate, 2 mM ATP, 16 mM HEPES (pH 7.9), 65 mM potassium glutamate, 0.16 mM EDTA, 0.4 mM DTT and 16% glycerol. Binding was for 10 min at 30 C Recombinant GST MBNL1 protein was produced as described (Miller et al., 2000) Plasmids The cTNT IR and ClaLC minigenes were previously described (Kosaki et al., 1998; Philips et al., 1998; Stamm et al., 1999) GFP fusions with MBNL1, 2 and 3 were provided by Dr JD Brook (Fardaei et al., 2002) GFP MBNL1 was found to have a novel MBNL1 isoform lacking exons 7, 9 and 10 and containing a frameshift in exon 12. Plasmids expressing DMPK exons 11 15 containing 960 interrupted CUG or CAG repeat s in exon 15 were cloned using techniques as previously described (Philips et al., 1998) The MBNL mut ant human cTNT minigene was generated by inverse PCR. actinin minigene and GFP Mbl expression vector were constructed as previ ously described (Vicente et al., 2007) Transfe ction of siRNA Two custom siRNA duplexes were designed for RNAi against human MBNL1 using the Dharmacon siDESIGN program (www.dharmacon.com), and were synthesized by Dharmacon. The siRNA sequences (5 to 3) are as follows: THH31 ( AACAGACAGACUUGAGGUAUG ) THH2 ( AACACGGAAUGUAAAUUUGCA ), GFP siRNA duplex (Dharmacon). Prior to transfections, 3X105 HeLa cells were plated in 2 mL of antibiotic free growth media (DMEM supplemented with 10% FBS) per well in a six well plate. At 12 hr after plating, the media was exchanged with 800 L serum free
147 media (DMEM) per well, and siRNA duplex (2.66 g) was transfected using Oligofectamine (Invitrogen) according to the manufacturers protocol. A fter 4 hr 1 mL of 3 X serum containing media (DMEM supplemented with 30% FBS) was added to each well and returned to the incubator. A fter 12 hr the 3 X serum containing media was replaced with antibiotic free growth media and the cells were transfected with 1 g of minigene and 2.66 g of siRNA duplex using Lipofectamine 2000 (Invi trogen ). The media was exchanged with antibiotic free growth media 6 hr later. RNA and protein were harvested 48 hr after transfection of the minigene. Immunoblot Analysis Cells were harvested in protein loading buffer (62.5 mM Tris HCl (pH 6.8), 2% SDS, 1 0% glycerol and 5% 2mercaptoethanol) and the protein concentration was quantitated with the NonInterfering Protein Assay (Genotech). Total protein lysates from HEK293 (20 g) and primary chicken skeletal (30 g) cultures were loaded on a 12.5% acrylami de gel and transferred to ImmobilonP membranes (Millipore). GFP was detected using JL 8 monocl onal antibody (BD Biosciences ) at a dilution of 1:2000. The secondary antibody was a goat anti mouse HRP conjugate (Jackso n Immunoresearch ) at a dilution of 1:50 00. To detect endogenous MBNL1, HeLa (50 g) protein lysates were loaded on a 12.5% acrylamide gel. Blots were probed with the monoclonal 3A4 (16 m g/mL ) at a dilution of 1:500. The secondary antibody was a sheep anti mouse HRP conjugate (Amersham Biosciences) at a dilution of 1:5000. For GAPDH in HeLa cells, 15 g of total protein lysates was run on a 12.5% acrylamide gel, transferred to membranes and detected using the 6G5 monoclonal (Biogenesis) at a dilution of 1:100 000. The secondary antibody was a goat anti mouse HRP conjugate (Jackso n
148 Immunoresearch ) at a dilution of 1:5000. Blots were developed in Amersham ECL detection reagents (GE Healthcare) or Supersignal West Femto reagents (Thermo Scientific Pierce) and exposed to Biomax Light Fil m (Kodak). Flourescent In Situ Hybridization and Immunocytochemistry COSM6 cells (1.2 X 105) were seeded on 2well Falcon culture slides (BD Biosciences) and transfected 24 hr after plating. C ells were washed with 1X Hanks Balanced Salt solution and then w ith CSK buffer (300 mM sucrose, 100 mM NaCl, 3 mM MgCl2, 1.2 mM PMSF, 10 mM PIPES, pH 6.8). The slides were incubated on ice in CSK buffer with 0.5% Triton X 100 and 10 mM vanadyl sulfate for 1 min They were then fixed in 4% paraformaldehyde/PBS (pH 7.4) for 10 min at room temperature and washed with 70% ethanol. Cells were dehydrated with 70% ethanol overnight at 4 C and then rehydrated with 40% formamide/2X SSC for 10 min at room temperature the following day Cy5 labeled oligonucleotide probes (Qiagen) (CAG)10 and (CTG)10 were used to detect CUG and CAG repeats, respectively. The cells were incubated in probe/hybridization buffer (40% formamide, 2 X SSC, 0.2% BSA, 10% dextran sulfate, 2 mM vanadyl sulfate, 1 mg/mL yeast tRNA, 50 ng/mL probe) in a humidified chamber at 37 C for 2 hr After hybridization, the slides were washed three times with 40% formamide/2X SSC for 30 min at 37 C followed by three PBS (pH 7.4) washes for 5 min each They were then pre blocked in 3% BSA/PBS for 15 m in in a humidified chamber at room temperature followed by a wash with PBS (pH 7.4). A fterwards, the slides were incubated with primary anti MBNL1 antibody 3A4 (10 mg/mL; 1:1000 dilution) in 3% BSA/PBS at room temperature for 1 hr in a humidified chamber, washed three times with PBS (pH 7.4) and incubated with the secondary antibody Alexa Fluor488labeled goat anti mouse IgG (2 mg/mL, Molecular Probes ) at a dilution of
149 1:100 in 3% BSA/PBS (pH 7.4) at room temperature for 1 hr Following the incubation, cel ls were washed three times with PBS (pH 7.4) and nuclei were stained with DAPI using Vectashield mounting media (Vec tor Laboratories, Inc ). GFP Mbl proteins were analyzed directly. Cells were analyzed under epifluorescence microscopy using an Axioskop2 m ot plus microscope (Carl Zeiss, Inc). Competition Assay Whole cell lysate was generated using myc tagged MBNL1 (41 kDa isoform) transiently transfected in COSM6 (cells were transfected using 2 g expression vector as previously described). To prepare whole cell lysate for competition assays, 5X 10 cm tissues culture plates containing COSM6 cells ectopically expressing myc MBNL1 were resuspended in 250 mL/plate in cold crosslinking buffer (20 mM HEPES KOH, pH 8.0, 100 mM KCl, 0.1% Igepal, 1X PicD/PicW), so nicated 3X 40% amplitude for 5 sec, spun down at 16.1 RCF for 20 min at 4 C. Following the spin, supernatant was transferred to a new microfuge on ice and glycerol was added to a final concentration of 20%. Whole cell lysate was aliquoted and stored at 80 C until use. To generate uniformly body labeled repeat RNAs, ~1 g linearized plasmid was transcribed with T7 RNA polymerase (Promega) in a 50 L with [ -32P]GTP according to the manufacturers protocol. After in vitro transcription, DNA template was digested by RQ1 RNase free DNase (1 unit/ L Promega) in the presence of ~1 unit/ L RNasin (Promega) and ~0.5 g/ L yeast tRNA (Invitrogen) for 15 min at 37 C. Following digestion, RNA is phenol:chloroform extracted, eth anol precipitated, resuspended in 10 L RNase free water, and resolved on a 5% acrylamideurea gel in 1X TBE (run at 200V for 2 hr in a vertical gel). RNA bands were excised and extracted by disrupting
150 the acrylamide gel slices with a sterile pestle and i ncubation in extraction buffer (0.5 M NH4OAc, 10 mM Mg(OAc)2, 0.1 mM EDTA, 0.1% SDS) at 65 C for 20 min (in a thermo mixer Eppendorf). Extraction was spun down for 1 min at 16.1 RCF at RT, supernatant was passed through a mini spin column (USA Scientif ic) and 0.45 m filter (Ultra Free MC Filter Millipore), ethanol precipitated, and stored at 80 C until use. To generate nonlabeled competitor, ~2.5 g linearized plasmid was transcribed using the MEGAshortscript kit (Ambion) and cleanedup using the MEGAclear kit (Ambion) according to the manufacturers protocol and stored at 80 C until use. Competition assays were done with 0.1 pmol uniformly body labeled RNA in a 35 L reaction composed of 40 mM HEPES KOH (pH 8.0), 2 mM Mg(OAc)2, 20 mM creatine ph osphate 2 mM ATP, 65 mM K glutamate, 0.4 mM DTT, 0.16 mM EDTA, 100 mM KCl, and 15 L myc tagged MBNL1 whole cell lysate at 30 C Incubation times and introduction of 200 pmol cold competitor RNA (2000X fold excess) were done as indicated in Figure 213. Following competition, reactions were UV crosslinked on ice 3X 2.5 min in a UV Stratalinker 1800 (Stratagene). Excess RNA was digested by incubating crosslinked reactions in ~0.5 g/ L RNase A (Sigma) for 20 min at 37 C ProteinA Sepharose (50 L Invitrogen) beads were washed two times by adding 500 L IPP 150 (50 mM Tris, pH 7.4, 150 mM NaCl, 0.1% Igepal),inverting several times in a microfuge tube, flash spinning (5 sec ) in a table top centrifuge, pipeting off the supernat ant, and repeating. ProteinA S epharose was resuspended in 300 L IPP 150 with 1 L myc (9E10 10 mg/mL) and rotated for 1 hr at 4 C Followi ng antibody capture, proteinA Sepharose: myc was washed three times as described above, resuspended in 300 L IPP 150, and incu bated with crosslinked competition reactions
151 for 1 hr rotating at 4 C to immunoprecipitate myc MBNL1:32P RNA. Following immunoprecipitation, complexes were washed five times as described above (four washes with IPP 150 and one wash with IPP 150 without I gepal), resuspended in 35 L 2X Laemmli buffer, resolved by 12.5% SDS PAGE. The gel was fixed for 20 min in destain buffer (10% acetic acid, 20% methanol), dried down using the Hoefer slab gel drier (Amersham), and exposed overnight to Biomax Film and X O matic Intensifying Screens (Kodak) at 80 C. Band intensity was measured using a phosphoimaging screen (Amersham) and the Typhoon 9200 (Amersham).
152 Figure 41. Primers used for RT PCR, Mbnl3E2/+ genotyping, probe generation, and subcloning.
153 LIST OF REFERENCES Accardi, A., and Pusch, M. (2000). Fast and slow gating relaxations in the muscle chloride channel CLC 1. J Gen Physiol 116, 433444. Anvret, M., Ahlberg, G., Grandell, U., Hedberg, B., Johnson, K., and Edstrom L. (1993). Larger expansions of the CTG repeat in muscle compared to lymphocytes from patients with myotonic dystrophy. Hum Mol Genet 2 13971400. Artero, R., Prokop, A., Paricio, N., Begemann, G., Pueyo, I., Mlodzik, M., Perez Alonso, M., and Baylies, M. K. (1998). The muscleblind gene participates in the organization of Z bands and epidermal attachments of Drosophila muscles and is regulated by Dmef2. Dev Biol 195, 131 143. Ashizawa, T., Dubel, J. R., and Harati, Y. (1993). Somatic instability of CTG repeat in myotonic dystrophy. Neurology 43, 26742678. Avaria, M., and Patterson, V. (1994). Myotonic dystrophy: relative sensitivity of symptoms signs and abnormal investigations. Ulster Med J 63, 151154. Barchi, R. L. (1995). Molecular pathology of t he skeletal muscle sodium channel. Annu Rev Physiol 57, 355 385. Begemann, G., Michon, A. M., vd Voorn, L., Wepf, R., and Mlodzik, M. (1995). The Drosophila orphan nuclear receptor sevenup requires the Ras pathway for its function in photoreceptor determination. Development 121 225 235. Begemann, G., Paricio, N., Artero, R., Kiss, I., Perez Alonso, M., and Mlodzik, M. (1997). muscleblind, a gene required for photoreceptor differentiation in Drosophila, encodes novel nuclear Cys3His type zinc finger containing proteins. Development 124 43214331. Belfiore, A., Frasca, F., Pandini, G., Sciacca, L., and Vigneri, R. (2009). Insulin receptor isoforms and insulin receptor/insulinlike growth factor receptor hybrids in physiology and disease. Endocr Rev 30, 586 623. Benders, A. A., Timmermans, J. A., Oosterhof, A., Ter Laak, H. J., van Kuppevelt, T. H., Wevers, R. A., and Veerkamp, J. H. (1993). Deficiency of Na+/K(+) ATPase and sarcoplasmic reticulum Ca(2+) ATPase in skeletal muscle and cultured muscle cells of myotonic dystrophy patients. Biochem J 293 ( Pt 1) 269 274. Berg, J., Jiang, H., Thornton, C. A., and Cannon, S. C. (2004). Truncated ClC 1 mRNA in myotonic dystrophy exerts a dominant negative effect on the Cl current. Neurology 63, 23712375. Bergen, B. J. (1985). Evaluation of the hypotonic or floppy infant. Minn Med 68 341 347.
154 Berul, C. I., Maguire, C. T., Aronovitz, M. J., Greenwood, J., Miller, C., Gehrmann, J., Housman, D., Mendelsohn, M. E., and Reddy, S. (1999). DMPK dosage alterations result in atrioventricular conduction abnormalities in a mouse myotonic dystrophy model. J Clin Invest 103 R1 7. Berul, C. I., Maguire, C. T., Gehrmann, J., and Reddy, S. (2000). Progressive atrioventricular conduction block in a mouse myotonic dystrophy model. J Interv Card Electrophysiol 4 351 358. Black, D. L. (2003). Mechanisms of alternative premessenger RNA splicing. Annu Rev Biochem 72, 291336. Bodensteiner, J. B. (2008). The evaluation of the hypotonic infant. Semin Pediatr Neurol 15, 10 2 0. Botta, A., Caldarola, S., Vallo, L., Bonifazi, E., Fruci, D., Gullotta, F., Massa, R., Novelli, G., and Loreni, F. (2006). Effect of the [CCTG]n repeat expansion on ZNF9 expression in myotonic dystrophy type II (DM2). Biochim Biophys Acta 1762, 329 334. Botta, A., Rinaldi, F., Catalli, C., Vergani, L., Bonifazi, E., Romeo, V., Loro, E., Viola, A., Angelini, C., and Novelli, G. (2008). The CTG repeat expansion size correlates with the splicing defects observed in muscles from myotonic dystrophy type 1 patients. J Med Genet 45 639 646. Brook, J. D., McCurrach, M. E., Harley, H. G., Buckler, A. J., Church, D., Aburatani, H., Hunter, K., Stanton, V. P., Thirion, J. P., Hudson, T., and et al. (1992). Molecular basis of myotonic dystrophy: expansion of a t rinucleotide (CTG) repeat at the 3' end of a transcript encoding a protein kinase family member. Cell 68, 799 808. Buckingham, M., Bajard, L., Chang, T., Daubas, P., Hadchouel, J., Meilhac, S., Montarras, D., Rocancourt, D., and Relaix, F. (2003). The for mation of skeletal muscle: from somite to limb. J Anat 202 59 68. Buxton, J., Shelbourne, P., Davies, J., Jones, C., Van Tongeren, T., Aslanidis, C., de Jong, P., Jansen, G., Anvret, M., Riley, B., and et al. (1992). Detection of an unstable fragment of DNA specific to individuals with myotonic dystrophy. Nature 355 547548. Campbell, C., Sherlock, R., Jacob, P., and Blayney, M. (2004). Congenital myotonic dystrophy: assisted ventilation duration and outcome. Pediatrics 113, 811 816. Chari, A., Paknia, E., and Fischer, U. (2009). The role of RNP biogenesis in spinal muscular atrophy. Curr Opin Cell Biol 21 387393. Charlet, B. N., Savkur, R. S., Singh, G., Philips, A. V., Grice, E. A., and Cooper, T. A. (2002). Loss of the musclespecific chloride channel in type 1 myotonic dystrophy due to misregulated alternative splicing. Mol Cell 10, 45 53.
155 Chen, M., and Manley, J. L. (2009). Mechanisms of alternative splicing regulation: insights from molecular and genomics approaches. Nat Rev Mol Cell Biol 10 741 754. Chen, W., Wang, Y., Abe, Y., Cheney, L., Udd, B., and Li, Y. P. (2007). Haploinsuffciency for Znf9 in Znf9+/ mice is associated with multiorgan abnormalities resembling myotonic dystrophy. J Mol Biol 368 8 17. Cho, D. H., Thienes, C. P., Maho ney, S. E., Analau, E., Filippova, G. N., and Tapscott, S. J. (2005). Antisense transcription and heterochromatin at the DM1 CTG repeats are constrained by CTCF. Mol Cell 20, 483489. Davis, B. M., McCurrach, M. E., Taneja, K. L., Singer, R. H., and Housm an, D. E. (1997). Expansion of a CUG trinucleotide repeat in the 3' untranslated region of myotonic dystrophy protein kinase transcripts results in nuclear retention of transcripts. Proc Natl Acad Sci U S A 94 73887393. Dixon, R. W., and Harris, J. B. ( 1996). Myotoxic activity of the toxic phospholipase, notexin, from the venom of the Australian tiger snake. J Neuropathol Exp Neurol 55 12301237. Du, H., Cline, M. S., Osborne, R. J., Tuttle, D. L., Clark, T. A., Donohue, J. P., Hall, M. P., Shiue, L., Swanson, M. S., Thornton, C. A., and Ares, M., Jr. (2010). Aberrant alternative splicing and extracellular matrix gene expression in mouse models of myotonic dystrophy. Nat Struct Mol Biol 17, 187 193. Echenne, B., Rideau, A., Roubertie, A., Sebire, G., R ivier, F., and Lemieux, B. (2008). Myotonic dystrophy type I in childhood Long term evolution in patients surviving the neonatal period. Eur J Paediatr Neurol 12, 210 223. Eriksson, M., Hedberg, B., Carey, N., and Ansved, T. (2001). Decreased DMPK transcr ipt levels in myotonic dystrophy 1 type IIA muscle fibers. Biochem Biophys Res Commun 286 1177 1182. Fardaei, M., Rogers, M. T., Thorpe, H. M., Larkin, K., Hamshere, M. G., Harper, P. S., and Brook, J. D. (2002). Three proteins, MBNL, MBLL and MBXL, co localize in vivo with nuclear foci of expandedrepeat transcripts in DM1 and DM2 cells. Hum Mol Genet 11, 805814. Farkas Bargeton, E., Barbet, J. P., Dancea, S., Wehrle, R., Checouri, A., and Dulac, O. (1988). Immaturity of muscle fibers in the congenital form of myotonic dystrophy: its consequences and its origin. J Neurol Sci 83 145159. Filippova, G. N., Thienes, C. P., Penn, B. H., Cho, D. H., Hu, Y. J., Moore, J. M., Klesert, T. R., Lobanenkov, V. V., and Tapscott, S. J. (2001). CTCF binding sit es flank CTG/CAG repeats and form a methylationsensitive insulator at the DM1 locus. Nat Genet 28 335 343.
156 Fiset, C., and Giles, W. R. (2008). Cardiac troponin T mutations promote lifethreatening arrhythmias. J Clin Invest. Fu, Y. H., Friedman, D. L., Richards S., Pearlman, J. A., Gibbs, R. A., Pizzuti, A., Ashizawa, T., Perryman, M. B., Scarlato, G., Fenwick, R. G., Jr., and et al. (1993). Decreased expression of myotoninprotein kinase messenger RNA and protein in adult form of myotonic dystrophy. Sc ience 260 235 238. Fu, Y. H., Pizzuti, A., Fenwick, R. G., Jr., King, J., Rajnarayan, S., Dunne, P. W., Dubel, J., Nasser, G. A., Ashizawa, T., de Jong, P., and et al. (1992). An unstable triplet repeat in a gene related to myotonic muscular dystrophy. S cience 255 12561258. Fuerst, P. G., Bruce, F., Tian, M., Wei, W., Elstrott, J., Feller, M. B., Erskine, L., Singer, J. H., and Burgess, R. W. (2009). DSCAM and DSCAML1 function in self avoidance in multiple cell types in the developing mouse retina. Neuron 64, 484 497. Fyrberg, E., Kelly, M., Ball, E., Fyrberg, C., and Reedy, M. C. (1990). Molecular genetics of Drosophila alphaactinin: mutant alleles disrupt Z disc integrity and muscle insertions. J Cell Biol 110 19992011. Geeves, M. A., and Holmes, K. C. (1999). Structural mechanism of muscle contraction. Annu Rev Biochem 68, 687728. Goers, E. S., Purcell, J., Voelker, R. B., Gates, D. P., and Berglund, J. A. (2010). MBNL1 binds GC motifs embedded in pyrimidines to regulate alternative splicing. N ucleic Acids Res. Groh, W. J., Lowe, M. R., and Zipes, D. P. (2002). Severity of cardiac conduction involvement and arrhythmias in myotonic dystrophy type 1 correlates with age and CTG repeat length. J Cardiovasc Electrophysiol 13, 444448. Gunning, P., O'Neill, G., and Hardeman, E. (2008). Tropomyosinbased regulation of the actin cytoskeleton in time and space. Physiol Rev 88, 1 35. Hamshere, M. G., Newman, E. E., Alwazzan, M., Athwal, B. S., and Brook, J. D. (1997). Transcriptional abnormality in myot onic dystrophy affects DMPK but not neighboring genes. Proc Natl Acad Sci U S A 94, 73947399. Hao, M., Akrami, K., Wei, K., De Diego, C., Che, N., Ku, J. H., Tidball, J., Graves, M. C., Shieh, P. B., and Chen, F. (2008). Muscleblindlike 2 (Mbnl2) defic ient mice as a model for myotonic dystrophy. Dev Dyn 237, 403 410. Harley, H. G., Brook, J. D., Rundle, S. A., Crow, S., Reardon, W., Buckler, A. J., Harper, P. S., Housman, D. E., and Shaw, D. J. (1992). Expansion of an unstable DNA region and phenotypic variation in myotonic dystrophy. Nature 355 545 546.
157 Harper, P. S. (1974). Proceedings: Congenital myotonic dystrophy in Britain. Arch Dis Child 49, 820. Harper, P. S. (1975). Congenital myotonic dystrophy in Britain. I. Clinical aspects. Arch Dis Child 50, 505513. Harper, P. S. (2001). Myotonic Dystrophy Third Edition: Harcourt Publishers). Harris, J. B., and MacDonell, C. A. (1981). Phospholipase A2 activity of notexin and its role in muscle damage. Toxicon 19, 419 430. Hattori, D., Chen, Y., Matth ews, B. J., Salwinski, L., Sabatti, C., Grueber, W. B., and Zipursky, S. L. (2009). Robust discrimination between self and nonself neurites requires thousands of Dscam1 isoforms. Nature 461 644648. He, Y., and Smith, R. (2009). Nuclear functions of het erogeneous nuclear ribonucleoproteins A/B. Cell Mol Life Sci 66 12391256. Hedberg, B., Anvret, M., and Ansved, T. (1999). CTG repeat length in distal and proximal leg muscles of symptomatic and non symptomatic patients with myotonic dystrophy: relation to muscle strength and degree of histopathological abnormalities. Eur J Neurol 6 341346. Hiromi, Y., Mlodzik, M., West, S. R., Rubin, G. M., and Goodman, C. S. (1993). Ectopic expression of sevenup causes cell fate changes during ommatidial assembly. D evelopment 118 1123 1135. Holt, I., Jacquemin, V., Fardaei, M., Sewry, C. A., Butler Browne, G. S., Furling, D., Brook, J. D., and Morris, G. E. (2009). Muscleblindlike proteins: similarities and differences in normal and myotonic dystrophy muscle. Am J Pathol 174 216227. Hudson, A. J., Ebers, G. C., and Bulman, D. E. (1995). The skeletal muscle sodium and chloride channel diseases. Brain 118 ( Pt 2) 547563. Iannaccone, S. T., Bove, K. E., Vogler, C., Azzarelli, B., and Muller, J. (1986). Muscle ma turation delay in infantile myotonic dystrophy. Arch Pathol Lab Med 110 405 411. Imarisio, S., Carmichael, J., Korolchuk, V., Chen, C. W., Saiki, S., Rose, C., Krishna, G., Davies, J. E., Ttofi, E., Underwood, B. R., and Rubinsztein, D. C. (2008). Huntington's disease: from pathology and genetics to potential therapies. Biochem J 412 191 209. Inukai, A., Doyu, M., Kato, T., Liang, Y., Kuru, S., Yamamoto, M., Kobayashi, Y., and Sobue, G. (2000). Reduced expression of DMAHP/SIX5 gene in myotonic dys trophy muscle. Muscle Nerve 23, 14211426.
158 Jensen, C. J., Oldfield, B. J., and Rubio, J. P. (2009). Splicing, cis genetic variation and disease. Biochem Soc Trans 37 13111315. Jin, J. P., Zhang, Z., and Bautista, J. A. (2008). Isoform diversity, regula tion, and functional adaptation of troponin and calponin. Crit Rev Eukaryot Gene Expr 18 93 124. Jurkat Rott, K., Lerche, H., and Lehmann Horn, F. (2002). Skeletal muscle channelopathies. J Neurol 249 1493 1502. Kahn, C. R., and White, M. F. (1988). The insulin receptor and the molecular mechanism of insulin action. J Clin Invest 82 11511156. Kanadia, R. N., Johnstone, K. A., Mankodi, A., Lungu, C., Thornton, C. A., Esson, D., Timmers, A. M., Hauswirth, W. W., and Swanson, M. S. (2003a). A musclebl ind knockout model for myotonic dystrophy. Science 302 19781980. Kanadia, R. N., Shin, J., Yuan, Y., Beattie, S. G., Wheeler, T. M., Thornton, C. A., and Swanson, M. S. (2006). Reversal of RNA missplicing and myotonia after muscleblind overexpression in a mouse poly(CUG) model for myotonic dystrophy. Proc Natl Acad Sci U S A 103 11748 11753. Kanadia, R. N., Urbinati, C. R., Crusselle, V. J., Luo, D., Lee, Y. J., Harrison, J. K., Oh, S. P., and Swanson, M. S. (2003b). Developmental expression of mouse m uscleblind genes Mbnl1, Mbnl2 and Mbnl3. Gene Expr Patterns 3 459 462. Kang, S., and Hong, S. (2009). Molecular pathogenesis of spinocerebellar ataxia type 1 disease. Mol Cells 27 621627. Kellerer, M., Lammers, R., Ermel, B., Tippmer, S., Vogt, B., Obe rmaierKusser, B., Ullrich, A., and Haring, H. U. (1992). Distinct alpha subunit structures of human insulin receptor A and B variants determine differences in tyrosine kinase activities. Biochemistry 31, 4588 4596. Kimura, T., Takahashi, M. P., Okuda, Y. Kaido, M., Fujimura, H., Yanagihara, T., and Sakoda, S. (2000). The expression of ion channel mRNAs in skeletal muscles from patients with myotonic muscular dystrophy. Neurosci Lett 295 93 96. Klesert, T. R., Cho, D. H., Clark, J. I., Maylie, J., Adelm an, J., Snider, L., Yuen, E. C., Soriano, P., and Tapscott, S. J. (2000). Mice deficient in Six5 develop cataracts: implications for myotonic dystrophy. Nat Genet 25, 105 109. Klesert, T. R., Otten, A. D., Bird, T. D., and Tapscott, S. J. (1997). Trinucle otide repeat expansion at the myotonic dystrophy locus reduces expression of DMAHP. Nat Genet 16, 402406.
159 Kosaki, A., Nelson, J., and Webster, N. J. (1998). Identification of intron and exon sequences involved in alternative splicing of insulin receptor pre mRNA. J Biol Chem 273, 1033110337. Koshy, B. T., and Zoghbi, H. Y. (1997). The CAG/polyglutamine tract diseases: gene products and molecular pathogenesis. Brain Pathol 7 927 942. Krol, J., Fiszer, A., Mykowska, A., Sobczak, K., de Mezer, M., and Krzyzosiak, W. J. (2007). Ribonuclease dicer cleaves triplet repeat hairpins into shorter repeats that silence specific targets. Mol Cell 25, 575 586. Kuyumcu Martinez, N. M., Wang, G. S., and Cooper, T. A. (2007). Increased steady state levels of CUGBP1 in myotonic dystrophy 1 are due to PKC mediated hyperphosphorylation. Mol Cell 28, 68 78. La Spada, A. R., Paulson, H. L., and Fischbeck, K. H. (1994). Trinucleotide repeat expansion in neurological disease. Ann Neurol 36, 814 822. Lacomis, D., Chad, D. A., and Smith, T. W. (1994). Proximal weakness as the primary manifestation of myotonic dystrophy in older adults. Muscle Nerve 17 687 688. Lee, K. S., Squillace, R. M., and Wang, E. H. (2007). Expression pattern of muscleblindlike proteins differs in d ifferentiating myoblasts. Biochem Biophys Res Commun 361 151155. Licatalosi, D. D., and Darnell, R. B. (2010). RNA processing and its regulation: global insights into biological networks. Nat Rev Genet 11, 75 87. Licatalosi, D. D., Mele, A., Fak, J. J., Ule, J., Kayikci, M., Chi, S. W., Clark, T. A., Schweitzer, A. C., Blume, J. E., Wang, X., et al. (2008). HITS CLIP yields genomewide insights into brain alternative RNA processing. Nature 456 464 469. Lin, X., Miller, J. W., Mankodi, A., Kanadia, R. N., Yuan, Y., Moxley, R. T., Swanson, M. S., and Thornton, C. A. (2006). Failure of MBNL1dependent post natal splicing transitions in myotonic dystrophy. Hum Mol Genet 15 20872097. Liquori, C. L., Ricker, K., Moseley, M. L., Jacobsen, J. F., Kress, W., Naylor, S. L., Day, J. W., and Ranum, L. P. (2001). Myotonic dystrophy type 2 caused by a CCTG expansion in intron 1 of ZNF9. Science 293 864 867. Long, J. C., and Caceres, J. F. (2009). The SR protein family of splicing factors: master regulators of gene expression. Biochem J 417 15 27. Mahadevan, M., Tsilfidis, C., Sabourin, L., Shutler, G., Amemiya, C., Jansen, G., Neville, C., Narang, M., Barcelo, J., O'Hoy, K., and et al. (1992). Myotonic dystrophy
160 mutation: an unstable CTG repeat in the 3' untranslated region of the gene. Science 255, 12531255. Mankodi, A., Logigian, E., Callahan, L., McClain, C., White, R., Henderson, D., Krym, M., and Thornton, C. A. (2000). Myotonic dystrophy in transgenic mice expressing an expanded CUG repeat. Science 289 17691773. Mankodi, A., Takahashi, M. P., Jiang, H., Beck, C. L., Bowers, W. J., Moxley, R. T., Cannon, S. C., and Thornton, C. A. (2002). Expanded CUG repeats trigger aberrant splicing of ClC1 chloride channel premRNA and hyperexcitability of skeletal muscle in myotonic dystrophy. Mol Cell 10, 35 44. Mankodi, A., Teng Umnuay, P., Krym, M., Henderson, D., Swanson, M., and Thornton, C. A. (2003). Ribonuclear inclusions in skeletal muscle in myotonic dystrophy types 1 and 2. Ann Neurol 54 760768. Mankodi, A., Urbinati, C. R., Yuan, Q. P., Moxley, R. T., Sansone, V., Krym, M., Henderson, D., Schalling, M., Swanson, M. S., and Thornton, C. A. (2001). Muscleblind localizes to nuclear foci of aberrant RNA in myotonic dystrophy types 1 and 2. Hum Mol Ge net 10 21652170. Margolis, J. M., Schoser, B. G., Moseley, M. L., Day, J. W., and Ranum, L. P. (2006). DM2 intronic expansions: evidence for CCUG accumulation without flanking sequence or effects on ZNF9 mRNA processing or protein expression. Hum Mol Genet 15 18081815. Martorell, L., Johnson, K., Boucher, C. A., and Baiget, M. (1997). Somatic instability of the myotonic dystrophy (CTG)n repeat during human fetal development. Hum Mol Genet 6 877 880. Martorell, L., Monckton, D. G., Gamez, J., and Bai get, M. (2000). Complex patterns of male germline instability and somatic mosaicism in myotonic dystrophy type 1. Eur J Hum Genet 8 423 430. Matthews, B. J., Kim, M. E., Flanagan, J. J., Hattori, D., Clemens, J. C., Zipursky, S. L., and Grueber, W. B. (2007). Dendrite self avoidance is controlled by Dscam. Cell 129 593604. McClain, D. A. (1991). Different ligand affinities of the two human insulin receptor splice variants are reflected in parallel changes in sensitivity for insulin action. Mol Endocr inol 5 734 739. Michalowski, S., Miller, J. W., Urbinati, C. R., Paliouras, M., Swanson, M. S., and Griffith, J. (1999). Visualization of doublestranded RNAs from the myotonic dystrophy protein kinase gene and interactions with CUG binding protein. Nucl eic Acids Res 27, 35343542.
161 Miller, J. W., Urbinati, C. R., Teng Umnuay, P., Stenberg, M. G., Byrne, B. J., Thornton, C. A., and Swanson, M. S. (2000). Recruitment of human muscleblind proteins to (CUG)(n) expansions associated with myotonic dystrophy. EM BO J 19, 44394448. Milner Brown, H. S., and Miller, R. G. (1990). Myotonic dystrophy: quantification of muscle weakness and myotonia and the effect of amitriptyline and exercise. Arch Phys Med Rehabil 71 983 987. Moller, D. E., Yokota, A., Caro, J. F., and Flier, J. S. (1989). Tissue specific expression of two alternatively spliced insulin receptor mRNAs in man. Mol Endocrinol 3 12631269. Monferrer, L., and Artero, R. (2006). An interspecific functional complementation test in Drosophila for introduc tory genetics laboratory courses. J Hered 97, 67 73. Moolman, J. C., Corfield, V. A., Posen, B., Ngumbela, K., Seidman, C., Brink, P. A., and Watkins, H. (1997). Sudden death due to troponin T mutations. J Am Coll Cardiol 29, 549555. Mounsey, J. P., Mis try, D. J., Ai, C. W., Reddy, S., and Moorman, J. R. (2000). Skeletal muscle sodium channel gating in mice deficient in myotonic dystrophy protein kinase. Hum Mol Genet 9, 23132320. Moxley, R. T., 3rd, Griggs, R. C., Goldblatt, D., VanGelder, V., Herr, B E., and Thiel, R. (1978). Decreased insulin sensitivity of forearm muscle in myotonic dystrophy. J Clin Invest 62, 857867. Moxley, R. T., Corbett, A. J., Minaker, K. L., and Rowe, J. W. (1984). Whole body insulin resistance in myotonic dystrophy. Ann N eurol 15, 157162. Nagy, A., Moens, C., Ivanyi, E., Pawling, J., Gertsenstein, M., Hadjantonakis, A. K., Pirity, M., and Rossant, J. (1998). Dissecting the role of N myc in development using a single targeting vector to generate a series of alleles. Curr Biol 8, 661 664. Novelli, G., Gennarelli, M., Zelano, G., Pizzuti, A., Fattorini, C., Caskey, C. T., and Dallapiccola, B. (1993). Failure in detecting mRNA transcripts from the mutated allele in myotonic dystrophy muscle. Biochem Mol Biol Int 29, 291297. Ogino, S., and Wilson, R. B. (2004). Spinal muscular atrophy: molecular genetics and diagnostics. Expert Rev Mol Diagn 4, 1529. Ohtsuki, I., and Morimoto, S. (2008). Troponin: regulatory function and disorders. Biochem Biophys Res Commun 369, 6273. Orengo, J. P., Chambon, P., Metzger, D., Mosier, D. R., Snipes, G. J., and Cooper, T. A. (2008). Expanded CTG repeats within the DMPK 3' UTR causes severe skeletal
162 muscle wasting in an inducible mouse model for myotonic dystrophy. Proc Natl Acad Sci U S A 105, 26462651. Orengo, J. P., and Cooper, T. A. (2007). Alternative splicing in disease. Adv Exp Med Biol 623, 212223. Osborne, R. J., Lin, X., Welle, S., Sobczak, K., O'Rourke, J. R., Swanson, M. S., and Thornton, C. A. (2009). Transcriptional and pos t transcriptional impact of toxic RNA in myotonic dystrophy. Hum Mol Genet 18, 14711481. Otten, A. D., and Tapscott, S. J. (1995). Triplet repeat expansion in myotonic dystrophy alters the adjacent chromatin structure. Proc Natl Acad Sci U S A 92, 54655 469. Pan, Q., Shai, O., Lee, L. J., Frey, B. J., and Blencowe, B. J. (2008). Deep surveying of alternative splicing complexity in the human transcriptome by highthroughput sequencing. Nat Genet 40, 14131415. Pan, Q., Shai, O., Lee, L. J., Frey, B. J., and Blencowe, B. J. (2009). Addendum: Deep surveying of alternative splicing complexity in the human transcriptome by highthroughput sequencing. Nat Genet 41, 762. Pascual, M., Vicente, M., Monferrer, L., and Artero, R. (2006). The Muscleblind family of proteins: an emerging class of regulators of developmentally programmed alternative splicing. Differentiation 74, 6580. Pearson, C. E., Nichol Edamura, K., and Cleary, J. D. (2005). Repeat instability: mechanisms of dynamic mutations. Nat Rev Genet 6, 72 9 742. Personius, K. E., Nautiyal, J., and Reddy, S. (2005). Myotonia and muscle contractile properties in mice with SIX5 deficiency. Muscle Nerve 31, 503505. Philips, A. V., Timchenko, L. T., and Cooper, T. A. (1998). Disruption of splicing regulated by a CUG binding protein in myotonic dystrophy. Science 280 737 741. Phillips, J. E., and Corces, V. G. (2009). CTCF: master weaver of the genome. Cell 137 11941211. Planells Cases, R., and Jentsch, T. J. (2009). Chloride channelopathies. Biochim Biophys Acta 1792, 173 189. Ranum, L. P., and Cooper, T. A. (2006). RNA Mediated Neuromuscular Disorders. Annu Rev Neurosci. Ranum, L. P., and Day, J. W. (2002). Dominantly inherited, noncoding microsatellite expansion disorders. Curr Opin Genet Dev 12 266 271.
163 Rappsilber, J., Ryder, U., Lamond, A. I., and Mann, M. (2002). Largescale proteomic analysis of the human spliceosome. Genome Res 12, 12311245. Ray, S., Bowen, J. R., and Marks, H. G. (1984). Foot deformity in myotonic dystrophy. Foot Ankle 5 125 130. Reddy, S., Mistry, D. J., Wang, Q. C., Geddis, L. M., Kutchai, H. C., Moorman, J. R., and Mounsey, J. P. (2002). Effects of age and gene dose on skeletal muscle sodium channel gating in mice deficient in myotonic dystrophy protein kinase. Muscle Nerve 25 850857. Reddy, S., Smith, D. B., Rich, M. M., Leferovich, J. M., Reilly, P., Davis, B. M., Tran, K., Rayburn, H., Bronson, R., Cros, D., et al. (1996). Mice lacking the myotonic dystrophy protein kinase develop a late onset progressive myopathy Nat Genet 13 325 335. Regev, R., de Vries, L. S., Heckmatt, J. Z., and Dubowitz, V. (1987). Cerebral ventricular dilation in congenital myotonic dystrophy. J Pediatr 111 372376. Ren, S. Y., Angrand, P. O., and Rijli, F. M. (2002). Targeted insertion results in a rhombomere 2specific Hoxa2 knockdown and ectopic activation of Hoxa1 expression. Dev Dyn 225, 305315. Richards, R. I., and Sutherland, G. R. (1994). Simple repeat DNA is not replicated simply. Nat Genet 6 114 116. Ricker, K., Koch, M. C. LehmannHorn, F., Pongratz, D., Speich, N., Reiners, K., Schneider, C., and Moxley, R. T., 3rd (1995). Proximal myotonic myopathy. Clinical features of a multisystem disorder similar to myotonic dystrophy. Arch Neurol 52, 25 31. Roulier, E. M., Fyrberg, C., and Fyrberg, E. (1992). Perturbations of Drosophila alphaactinin cause muscle paralysis, weakness, and atrophy but do not confer obvious nonmuscle phenotypes. J Cell Biol 116 911 922. Rutherford, M. A., Heckmatt, J. Z., and Dubowitz, V. (1989). Congenital myotonic dystrophy: respiratory function at birth determines survival. Arch Dis Child 64, 191 195. Sabouri, L. A., Mahadevan, M. S., Narang, M., Lee, D. S., Surh, L. C., and Korneluk, R. G. (1993). Effect of the myotonic dystrophy (DM) mutation on mRNA levels of the DM gene. Nat Genet 4 233 238. Sadler, A. J., and Williams, B. R. (2007). Structure and function of the protein kinase R. Curr Top Microbiol Immunol 316 253 292. Sahgal, V., Bernes, S., Sahgal, S., Lischwey, C., and Subramani, V. (1983a). Skeletal muscle in preterm infants with congenital myotonic dystrophy. Morphologic and histochemical study. J Neurol Sci 59 47 55.
164 Sahgal, V., Sahgal, S., Bernes, S., and Subramani, V. (1983b). Ultrastructure of muscle spindle in congenital myotonic dystrophy. A study of preterm infant muscle spindles. Acta Neuropathol 61, 207 213. Sarkar, P. S., Appukuttan, B., Han, J., Ito, Y., Ai, C., Tsai, W., Chai, Y., Stout, J. T., and Reddy, S. (2000). Heterozygous loss of Six5 in mice is sufficient to cause ocular cataracts. Nat Genet 25 110 114. Sarkar, P. S., Paul, S., Han, J., and Reddy, S. (2004). Six5 is required for spermatogenic cell survival and spermiogenesis. Hum Mol Genet 13, 14211431. Sarnat, H. B., and Silbert, S. W. (1976). Maturational arrest of fetal muscle in neonatal myotonic dystrophy. A pathologic study of four cases. Arch Neurol 33 466 474. Savkur, R. S., Philips, A. V., and Cooper, T. A. (2001). Aberrant regulation of insulin receptor alternative splicing is associated with insu lin resistance in myotonic dystrophy. Nat Genet 29, 40 47. Savkur, R. S., Philips, A. V., Cooper, T. A., Dalton, J. C., Moseley, M. L., Ranum, L. P., and Day, J. W. (2004). Insulin receptor splicing alteration in myotonic dystrophy type 2. Am J Hum Genet 74, 13091313. Schild, R. L., Plath, H., Hofstaetter, C., Brenner, R., Mann, E., Mundegar, R. R., Steinbach, P., and Hansmann, M. (1998). Polyhydramnios: an association with congenital myotonic dystrophy. J Obstet Gynaecol 18 484 485. Schmucker, D., Cle mens, J. C., Shu, H., Worby, C. A., Xiao, J., Muda, M., Dixon, J. E., and Zipursky, S. L. (2000). Drosophila Dscam is an axon guidance receptor exhibiting extraordinary molecular diversity. Cell 101, 671 684. Schwenk, F., Baron, U., and Rajewsky, K. (1995). A cre transgenic mouse strain for the ubiquitous deletion of loxP flanked gene segments including deletion in germ cells. Nucleic Acids Res 23, 50805081. Seino, S., and Bell, G. I. (1989). Alternative splicing of human insulin receptor messenger RNA. Biochem Biophys Res Commun 159 312316. Seznec, H., LiaBaldini, A. S., Duros, C., Fouquet, C., Lacroix, C., Hofmann Radvanyi, H., Junien, C., and Gourdon, G. (2000). Transgenic mice carrying large human genomic sequences with expanded CTG repeat mimic c losely the DM CTG repeat intergenerational and somatic instability. Hum Mol Genet 9 11851194. Shao, J., and Diamond, M. I. (2007). Polyglutamine diseases: emerging concepts in pathogenesis and therapy. Hum Mol Genet 16 Spec No. 2, R115 123.
165 Shaper, J. H., Barker, R., and Hill, R. L. (1973). Purification of wheat germ agglutinin by affinity chromatography. Anal Biochem 53 564 570. Shepard, P. J., and Hertel, K. J. (2009). The SR protein family. Genome Biol 10, 242. Siegel, I. M., Eisenberg, B. R., and Glantz, R. H. (1984). Contributory etiologic factor for talipes equinovarus in congenital myotonic dystrophy: comparative biopsy study of intrinsic foot musculature and vastus lateralis in two cases. J Pediatr Orthop 4 327330. Squillace, R. M., Chenaul t, D. M., and Wang, E. H. (2002). Inhibition of muscle differentiation by the novel muscleblindrelated protein CHCR. Dev Biol 250 218 230. Stamm, S., Casper, D., Hanson, V., and Helfman, D. M. (1999). Regulation of the neuronspecific exon of clathrin l ight chain B. Brain Res Mol Brain Res 64 108 118. Stuart, C. A., Armstrong, R. M., Provow, S. A., and Plishker, G. A. (1983). Insulin resistance in patients with myotonic dystrophy. Neurology 33, 679685. Tanabe, Y., and Nonaka, I. (1987). Congenital myotonic dystrophy. Changes in muscle pathology with ageing. J Neurol Sci 77 59 68. Teplova, M., and Patel, D. J. (2008). Structural insights into RNA recognition by the alternativesplicing regulator muscleblindlike MBNL1. Nat Struct Mol Biol 15 13431351. Thornton, C. A., Griggs, R. C., and Moxley, R. T., 3rd (1994). Myotonic dystrophy with no trinucleotide repeat expansion. Ann Neurol 35, 269 272. Tian, B., White, R. J., Xia, T., Welle, S., Turner, D. H., Mathews, M. B., and Thornton, C. A. (2000). Expanded CUG repeat RNAs form hairpins that activate the doublestranded RNA dependent protein kinase PKR. RNA 6 79 87. Timchenko, N. A., Cai, Z. J., Welm, A. L., Reddy, S., Ashizawa, T., and Timchenko, L. T. (2001). RNA CUG repeats sequester CUGBP1 and alter protein levels and activity of CUGBP1. J Biol Chem 276 78207826. Tominaga, K., Hayashi, Y. K., Goto, K., Minami, N., Noguchi, S., Nonaka, I., Miki, T., and Nishino, I. (2010). Congenital myotonic dystrophy can show congenital fiber type dispropor tion pathology. Acta Neuropathol. Townsend, P. J., Farza, H., MacGeoch, C., Spurr, N. K., Wade, R., Gahlmann, R., Yacoub, M. H., and Barton, P. J. (1994). Human cardiac troponin T: identification of fetal isoforms and assignment of the TNNT2 locus to chro mosome 1q. Genomics 21 311316.
166 Tsilfidis, C., MacKenzie, A. E., Mettler, G., Barcelo, J., and Korneluk, R. G. (1992). Correlation between CTG trinucleotide repeat length and frequency of severe congenital myotonic dystrophy. Nat Genet 1 192 195. Utomo A. R., Nikitin, A. Y., and Lee, W. H. (1999). Temporal, spatial, and cell typespecific control of Cre mediated DNA recombination in transgenic mice. Nat Biotechnol 17, 10911096. Vialettes, B., Pouget, J., Viard, R., Moulin, J. P., Serratrice, G., and Vague, P. (1986). Mechanism and significance of insulin resistance in myotonic dystrophy. Horm Metab Res 18, 395 399. Vicente, M., Monferrer, L., Poulos, M. G., Houseley, J., Monckton, D. G., O'Dell K, M., Swanson, M. S., and Artero, R. D. (2007). Muscleblind isoforms are functionally distinct and regulate alphaactinin splicing. Differentiation 75 427440. Vogt, B., Carrascosa, J. M., Ermel, B., Ullrich, A., and Haring, H. U. (1991). The two isotypes of the human insulin receptor (HIR A and HIR B) follow different internalization kinetics. Biochem Biophys Res Commun 177 10131018. Wahl, M. C., Will, C. L., and Luhrmann, R. (2009). The spliceosome: design principles of a dynamic RNP machine. Cell 136 701 718. Wakimoto, H., Maguire, C. T., Sherwoo d, M. C., Vargas, M. M., Sarkar, P. S., Han, J., Reddy, S., and Berul, C. I. (2002). Characterization of cardiac conduction system abnormalities in mice with targeted disruption of Six5 gene. J Interv Card Electrophysiol 7 127 135. Wang, E. T., Sandberg, R., Luo, S., Khrebtukova, I., Zhang, L., Mayr, C., Kingsmore, S. F., Schroth, G. P., and Burge, C. B. (2008). Alternative isoform regulation in human tissue transcriptomes. Nature 456 470 476. Ward, A. J., and Cooper, T. A. (2010). The pathobiology of s plicing. J Pathol 220 152 163. Warf, M. B., and Berglund, J. A. (2007). MBNL binds similar RNA structures in the CUG repeats of myotonic dystrophy and its premRNA substrate cardiac troponin T. RNA 13, 22382251. Wheeler, T. M., Lueck, J. D., Swanson, M S., Dirksen, R. T., and Thornton, C. A. (2007). Correction of ClC 1 splicing eliminates chloride channelopathy and myotonia in mouse models of myotonic dystrophy. J Clin Invest 117 39523957. Wieacker, P., Wilhelm, C., Furste, H., and Schillinger, H. ( 1988). [Polyhydramnios in congenital myotonic dystrophy]. Z Geburtshilfe Perinatol 192 36 37.
167 Wong, L. J., Ashizawa, T., Monckton, D. G., Caskey, C. T., and Richards, C. S. (1995). Somatic heterogeneity of the CTG repeat in myotonic dystrophy is age and size dependent. Am J Hum Genet 56 114 122. Xu, R., Teng, J., and Cooper, T. A. (1993). The cardiac troponin T alternative exon contains a novel purinerich positive splicing element. Mol Cell Biol 13, 36603674. Yeo, G. W., Coufal, N. G., Liang, T. Y., Peng, G. E., Fu, X. D., and Gage, F. H. (2009). An RNA code for the FOX2 splicing regulator revealed by mapping RNA protein interactions in stem cells. Nat Struct Mol Biol 16, 130137. Yuan, Y., Compton, S. A., Sobczak, K., Stenberg, M. G., Thornton, C. A ., Griffith, J. D., and Swanson, M. S. (2007). Muscleblindlike 1 interacts with RNA hairpins in splicing target and pathogenic RNAs. Nucleic Acids Res 35 54745486. Zhou, Z., Licklider, L. J., Gygi, S. P., and Reed, R. (2002). Comprehensive proteomic analysis of the human spliceosome. Nature 419 182185.
168 BIOGRAPHICAL SKETCH Michael Poulos was born in Midland Michigan in 1979. He is the oldest of two sons born to David and Eileen Poulos Michael attended Grand Valley State University from 1997 2002 where he studied biology and chemistry and earned a Bachelor of Science in b iology After graduation, Michael moved to Gainesville, Florida and joined the interdisciplinary program in biomedical sciences at the University of Florida College of Medicine. Michael did his graduate work in Dr. Maury Swansons laboratory of the Molecular Genetics and Microbiology Department and completed his Ph.D. dissertation in May 2010. Michael plans to move to New York in the s pring of 2010 to begin his postdoctoral studies in Dr. Stewart Shumans laboratory at the SloanKettering Institute