Lab Projects Overview
- Mechanism of myotonic dystrophy pathogenesis
- Mechanism of cell-specific alternative splicing
- Mechanisms of coordinated splicing regulation
Myotonic Dystrophy (DM)
DM is the second most common form of muscular dystrophy affecting 1 in 8500 people worldwide. It is dominantly inherited and affects multiple organ systems including skeletal muscle, heart, and the central nervous system. Two microsatellite expansions cause DM: a CTG expansion in the 3' untranslated region of the DMPK gene on chromosome 19 causes Type 1 (DM1) and a CCTG expansion in intron 1 of the ZNF9 gene on chromosome 3 causes Type 2 (DM2). Both expansions cause disease by expression of RNA containing extended tracts of CUG or CCUG from the expanded alleles. These RNAs accumulate in discrete nuclear foci (Figure 1) and have the toxic effects described below.
Disruption of developmentally regulated alternative splicing in DM
The repeat-containing RNA disrupts the functions of at least two families of RNA binding proteins, CUGBP1 and ETR-3 like factors (CELF) and muscleblind like (MBNL). Different members of these two protein families regulate different aspects of gene expression including alternative splicing, translation, RNA editing, and mRNA localization. Particularly relevant to DM, CUGBP1 and MBNL1 regulate a network of alternative splicing transitions during heart and skeletal muscle development (see “ Mechanisms of coordinated splicing regulation ”, below). In DM, disruption of CUGBP1 and MBNL1 functions alters regulated splicing of their pre-mRNA targets resulting in inappropriate expression of fetal protein splice variants in adult tissues ultimately causing features of the disease. For example, we've shown that mis-splicing of the muscle specific chloride channel (ClC-1) and insulin receptor pre-mRNAs cause the myotonia and the unusual form of insulin resistance, respectively (Savkur et al. 2001, Charlet-B. et al. 2002). Twenty-seven splicing events have been shown to be disrupted in DM1 heart, skeletal muscle, or brain; all are normally developmentally regulated and all exhibit expression of the fetal isoform in adult tissues. It is likely that other major features of the disease such as skeletal muscle wasting, cardiac arrhythmias, and CNS dysfunction results from aberrant splicing these events or others that remain to be identified.
Disruption of CUGBP1 and MBNL1 function
MBNL proteins bind to and are sequestered on the repeat RNA resulting in a loss of MBNL function (for review see, Ranum and Cooper, 2006 ). CUG repeat RNA induces phosphorylation of CUGBP1 via a mechanism requiring protein kinase C (PKC). Phosphorylation stabilizes CUGBP1 resulting in higher protein steady state levels and a gain of CUGBP1 function (Kuyumcu-Martinez et al. 2007 and Figure 2).
Inducible mouse models for DM1
Using a Cre-LoxP approach combined with animals expressing tamoxifen-inducible Cre, we have reproduced phenotypic and molecular features of DM1 in heart and skeletal muscle (Wang et al. 2007 , Orengo et al. 2008). The potential roles of such splicing events in muscle wasting is under investigation. Our current goals are to: (i) determine the mechanism(s) by which the expanded repeat RNA induces PKC and CUGBP1 phosphorylation; (ii) use computational, splicing sensitive microarrays (see “ Mechanisms of coordinated splicing regulation ”, below), and RT-PCR screens to identify additional pre-mRNA targets of CELF and MBNL proteins whose mis-regulated splicing contributes to severe manifestations of disease such as skeletal muscle dystrophy, cardiac arrhythmias, and central nervous system symptoms; (iii) determine the pathogenic form(s) of the expanded repeat RNA [are the RNA foci pathogenic, protective, or neutral?]; (iv) use the transgenic mouse models that inducibly express CELF proteins or CUG repeat RNA to investigate the mechanisms of pathogenesis and to test treatment regimes; (v) use cell lines that inducibly express CUG-repeat RNA to determine the cellular and molecular effects of CUG-repeat RNA expression including on signaling pathways required for induction and maintenance of skeletal muscle differentiation.
We are investigating the mechanisms that regulate alternative splicing on several levels. One is the molecular details of how regulatory factors that directly bind the pre-mRNA communicate to the basal splicing machinery and promote inclusion or skipping of a variably spliced region. Another area, described in the " Mechanisms of coordinated splicing regulation " section below, is to determine how the activities of splicing regulators are modulated during development (using mouse models) and during cellular differentiation (using cell culture models), to investigate the regulatory networks that coordinate subsets of alternative splicing events during development.
Alternative splicing decisions are regulated by combinatorial control involving multiple factors that bind to the pre-mRNA. Evidence also indicates that, as in regulation of cell-specific transcription, recruitment (or repression) of the basal machinery requires assembly of a multicomponent complex (Figure 3). The final outcome of a regulated splicing event results from dominance of activation or repression activities.
To investigate the molecular details of regulation, we work on two families of splicing regulators: CELF and MBNL. Each of the proteins examined have been shown to have positive or negative effects on splicing of different alternative exons. Interestingly, CELF and MBNL proteins often regulate splicing of the same alternative exons and in all cases act antagonistically.
Our primary goal is to understand how splicing regulators activate alternative splicing. The specific questions being addressed are, how does binding of a splicing activator recruit or stabilize binding of the basal splicing machinery; what are the required components of the activation complex; what components bind directly to the splicing activator; and what protein-protein interactions link the bound splicing activator to the basal splicing machinery?
CELF interacting proteins have identified by yeast two-hybrid analysis, co-immunoprecipitation, and Tandem Affinity Purification (TAP). Recombinant CELF proteins switch splicing from exon skipping to exon inclusion in a cell-free splicing assay. This provides an excellent experimental system to determine the mechanism for activation. This analysis has identified a direct interaction between the CELF protein, CUGBP2 and a basal splicing component, U2 snRNP. Additional CELF-binding proteins with relevance to splicing as well as transcription were identified and will be investigated.
We have identified CELF protein domains required for activation and repression of different pre-mRNA targets (Singh et al. 2004 , Han et al. 2005). An ongoing analysis of how the same CELF protein activates different pre-mRNA targets and how different CELF paralogues activate the same pre-mRNA has revealed an unexpected diversity of activation mechanisms. This analysis plus comparisons between activation mechanisms of CELF and MBNL proteins will reveal a broad view of the mechanism of splicing activation.
An array of alternative splicing transitions occurs during differentiation and development. Different sets of splicing factors (such as CELF and MBNL proteins) regulate specific subsets of alternative splicing transitions (Figure 4) and this regulation is coordinated by interacting networks similar to those described for transcriptional regulation.
Our current goals are to address the following questions: What is the level of coordinated regulation during development? What regulatory factors are the dominant determinants of specific developmental splicing changes? How are the activities of the splicing modulators modified during development (protein abundance, nuclear:cytoplasmic distribution, post-translational modifications)? What are the signaling pathways responsible for modulating the activities of the dominant regulators? What are the functional consequences of the protein isoform transitions that occur during development?
We are investigating this "next level" of alternative splicing regulation during development of heart and skeletal muscle tissues in mice and during skeletal muscle differentiation in cell culture. For example, using splicing sensitive DNA microarrays and RT-PCR, we identified 67 events that display a robust change between embryonic day 14 and adult heart in the mouse. Most of these splicing transitions are temporally coordinated in one of three patterns before and/or after birth; a period of extensive tissue remodeling. Strikingly, more than 60% of the developmental splicing events examined are conserved in mouse and chicken, strongly supporting their functional significance. Computational analyses for conserved and enriched motifs associated with the variably spliced regions identified binding sites for known splicing regulators, including CELF and MBNL proteins, as well as several highly significant novel motifs. Using transgenic mice overexpressing the CELF protein, CUGBP1, and MBNL1 knock out mice, we established that 24 of 44 developmental splicing transitions tested are regulated by one or both of these proteins. Specifically, reproducing the embryonic expression pattern for CELF and MBNL proteins in adult heart restored the embryonic splicing patterns of all 24 events. The results reveal a previously unrecognized level of alternative splicing regulation during a critical stage of heart development (Kalsotra et al., submitted).
A 10 and 18 fold decrease in CUGBP1 and CUGBP2 proteins during mouse heart development, respectively, correlates with CELF-dependent splicing transitions. Interestingly, CUGBP1 and CUGBP2 mRNA levels remain unchanged during development indicative of regulation of translation efficiency and/or protein stability. CUGBP1 protein stability is regulated by phosphorylation (see, “ Mechanism of myotonic dystrophy pathogenesis ”, above) and CUGBP1 phosphorylation is regulated during heart and skeletal muscle development.
An analysis of splicing transitions that occur during skeletal muscle differentiation in culture using splicing sensitive microarrays and RT-PCR has also revealed a large number of coordinated splicing transitions. Computational analysis has identified enriched binding sites for known splicing regulators. Changes in steady state levels and/or nuclear-cytoplasmic abundance will be analyzed for a panel of these as well as a larger panel of splicing regulators. Most of the splicing transitions occur surprisingly early during muscle differentiation. Nuclear proteins that change in abundance or mobility (indicative of post-translational modifications) during this narrow temporal window of splicing transitions will be identified by Difference Gel Electrophoresis (DiGE) analysis and mass spectrometry. The function of these splicing regulators and ultimately the role of the networks of splicing transitions in muscle differentiation will be determined by shRNA knock downs and overexpression of splicing regulators using inducible lentiviral vectors.