Eric T Wang, Ph.D. - US grants

Many trans and cis acting factors that control alternative splicing have been identified. The explosion of next-generation sequencing approaches have identified thousands of regulated splicing events (RNA-seq), and binding sites of many proteins which regulate alternative splicing have been identified via CLIPseq. An important question remaining for the field is: What are the rules governing the behavior of alternative splicing decisions? For example, what are the RNA elements (sequence and structure) that determine if alternatively regulated exons will respond at low concentrations or at high concentrations to a splicing factor, and will the splicing responses exhibit cooperative behavior or not? Addressing these questions is important for providing a framework for understanding how changes in splicing factor concentration can lead to disease. To address these questions, we have created cellular models that allow us to precisely titrate the level of an alternative splicing regulator, the Muscleblind-like 1 (MBNL1) protein. MBNL1 has been shown to regulate thousands of alternative splicing events and is important for the development of skeletal muscle, heart and the central nervous system. This regulation is highlighted by the primary role that MBNL proteins play in the disease myotonic dystrophy (DM), in which MBNL1 and its paralogs (MBNL2 and MBNL3) are sequestered by expanded CUG or CCUG repeat RNAs, resulting in aberrant RNA processing. The mis- splicing of MBNL targets has been shown to be responsible for causing some of the symptoms associated with DM, including the hallmark symptom myotonia. .  

Myotonic dystrophy (dystrophia myotonica, DM) is an autosomal dominant genetic disease, with a diagnosed prevalence of 1:8000 people worldwide, that affects multiple tissues of the body, including skeletal muscle, heart, and related to this proposal, the central nervous system (CNS). DM1 is caused by expanded CTG repeats in the 3' UTR of dystrophia myotonica protein kinase (DMPK). There is substantial evidence in mouse DM1 models and human DM1 postmortem tissue to support an RNA-mediated disease mechanism where toxic intranuclear CUG RNA foci sequester Muscleblind (MBNL) RNA binding proteins that normally play crucial roles to regulate various aspects of post-transcriptional gene regulation. A major gap in our understanding is that we do not know which RNA processing defects underlie specific impairments in DM1 brain function. Recent work together with our new findings suggests that missplicing of RNAs encoding synaptic proteins is responsible for CNS dysfunction in DM1. Our central hypothesis is that CNS phenotypes are directly attributed to loss of MBNL mediated RNA processing and that restoration of MBNL activity and/or splicing can restore brain function. Our goal is to gain a thorough understanding of RNA processing-mediated mechanisms of CNS dysfunction in DM1 and use this to develop and rigorously evaluate novel therapeutic strategies. The overall objectives of this proposal are to use both candidate and genome wide approaches, applied to MBNL KO mice and a new AAV9 based neuronal mouse model, compared to RNAseq analysis of human postmortem brain, to evaluate the role of specific splicing events to drive symptoms, and to comprehensively identify changes in missplicing and RNA processing. Aim 1 will characterize how dysregulation of GABRG2, GRIN1, and SNAP25 splicing events is linked to molecular, cellular, and behavioral phenotypes observed in DM1. Aim 2 will develop a new AAV9 based mouse model to elucidate the set of RNA processing events in neurons that cause DM1 phenotypes, through transcriptional profiling and overlap of human DM brains with DM mouse model brains. Aim 3 will assess the extent to which antisense oligonucleotides or MBNL expression can rescue molecular, cellular, physiologic and behavioral phenotypes in DM1 mouse models. These studies will provide new mechanistic insights into how perturbations to specific RNA processing events can lead to CNS symptoms in myotonic dystrophy, and provide a broader comprehensive view of all transcriptome changes occurring in the DM CNS. The proposed work is significant, as no molecular changes have been linked to any phenotypes in the DM CNS. This provides the framework for future therapeutic efforts aimed at correcting CNS defects.

Transcription of expanded microsatellite repeats is associated with a number of human diseases, including myotonic dystrophy (DM), Fuch's endothelial corneal dystrophy, and C9orf72 ALS/FTD (C9ALS/FTD), among others. Eliminating or reducing production of RNA and proteins arising from these expanded loci holds therapeutic benefit. Here, we will test the hypothesis that a deactivated form of the Cas9 enzyme impedes transcription across expanded microsatellite repeats, in cell and animal models of DM and C9ALS/FTD. We have previously observed a repeat length-, PAM-, and strand-dependent reduction in the abundance of repeat- containing RNAs upon targeting dCas9 directly to repeat sequences. Aberrant Muscleblind-dependent splicing patterns were rescued in DM1 cells, and production of RAN peptides characteristic of C9orf72 ALS/FTD cells was drastically decreased. Pathological CUG-containing RNA foci in DM1 mouse model muscle fibers was reduced by dCas9/gRNA delivered by adeno-associated virus. These observations suggest that transcription of microsatellite repeat-containing RNAs is more sensitive to perturbation than transcription of other RNAs, indicating potentially viable strategies for therapeutic intervention. In this proposal, we will assess the extent to which virally delivered dCas9/gRNA complexes can rescue molecular, cellular, and phenotypic features in to established models of DM1 and C9ALS/FTD. The HSALR model, which exhibits myotonia, centralized nuclei, and altered transcriptomes characteristic of human DM1, will be used to study DM. The C9-500 BAC transgenic model, which exhibits upper and lower motor neuron degeneration, altered gait, paralysis, and premature death, will be used to study C9ALS/FTD. Our proposed experiments will establish whether inhibition of toxic repeat transcription can rescue disease phenotypes, and define a window around which reduction of toxic RNA abundance is therapeutic.

The development and maintenance of neurons and synaptic connections are highly complex processes, in part due to the massive cytoplasmic volume and complex branching morphology of axons and dendrites. As one mechanism, it is well appreciated that RNA localization and local translation are required to precisely regulate protein homeostasis at synapses. Indeed, loss of FMRP in Fragile X Syndrome, or other impairments to RNA localization and local translation at synapses, likely contribute to brain disorders. To better understand RNA localization and local translation in neurons, we must elucidate the RNA cis-elements, RBP trans-factors, and cytoskeletal motors mediating these processes. Although ongoing efforts have demonstrated how RNA binding proteins (RBPs) can regulate local translation at post-synaptic sites, there still exists a major gap in our understanding of how RBPs transport RNAs to regulate synaptic function. Fortunately, recent observations provide clues about fruitful lines of investigation. For example, multiple studies report that distally localized RNAs are enriched for cis-elements targeted by Muscleblind-like (MBNL) proteins. Although these observations suggest that MBNL may be a major player in localizing RNAs to the pre- and post-synapse, we still lack a mechanistic understanding for how MBNL proteins may achieve this task, or what functions depend on MBNL-mediated RNA localization. This line of research has important implications for the neurological disease myotonic dystrophy (dystrophia myotonica, DM), in which MBNLs are depleted by toxic CUG repeats. Therefore, an emerging hypothesis is that RNA localization functions of MBNL are important for proper synapse function, and that mis-localized RNAs might account for some neurological features of DM patients, particularly early in disease. Here, using MBNL depletion and DM-associated models, we propose to identify specific functions for the localization of MBNL targets. Aim 1 will elucidate mechanisms of MBNL-mediated mRNA localization in neurons. We will define the RNA targets that are localized by MBNL in the pre- and post-synapse. We will characterize dynamic properties of motile MBNL RNA granules in live neurons and identify cytoskeletal motors and adaptors associated with these granules. Using genomics, live cell imaging, and biochemical approaches, we will establish mechanisms of how MBNL-interacting RNAs are transported. Aim 2 will define functions conferred by MBNL-dependent RNA localization using models of synapse development and function, and models of myotonic dystrophy. By depleting cytoplasmic MBNL and other proteins required for MBNL-dependent RNA localization, we will assess cellular functions dependent on this process. We will identify specific neuronal functions, such as synaptic vesicle release, that depend on proper localization of mRNAs by MBNL proteins. The impact of this research is to better understand how RNA localization and local translation confers important synaptic functions and how they may go awry in DM. As few RNA binding proteins have been linked to motors, this may evolve into a unifying model for mRNA transport to synapses.