Search

Patient-derived induced pluripotent stem cells (iPSCs) provide an opportunity to study human diseases mainly in those cases for which no suitable model systems are available. Here, we have taken advantage of in vitro iPSCs derived from patients affected by amyotrophic lateral sclerosis (ALS) and carrying mutations in the RNA-binding protein FUS to study the cellular behavior of the mutant proteins in the appropriate genetic background. Moreover, the ability to differentiate iPSCs into spinal cord neural cells provides an in vitro model mimicking the physiological conditions. iPSCs were derived from FUSR514S and FUSR521C patient fibroblasts, whereas in the case of the severe FUSP525L mutation, in which fibroblasts were not available, a heterozygous and a homozygous iPSC line were raised by TALEN-directed mutagenesis. We show that aberrant localization and recruitment of FUS into stress granules (SGs) is a prerogative of the FUS mutant proteins and occurs only upon induction of stress in both undifferentiated iPSCs and spinal cord neural cells. Moreover, we show that the incorporation into SGs is proportional to the amount of cytoplasmic FUS, strongly correlating with the cytoplasmic delocalization phenotype of the different mutants. Therefore, the available iPSCs represent a very powerful system for understanding the correlation between FUS mutations, the molecular mechanisms of SG formation and ALS ethiopathogenesis., Summary: Mutated FUS protein is aberrantly delocalized and recruited into stress granules in iPSC-derived motoneurons, which provide a new model system for amyotrophic lateral sclerosis.

Duchenne muscular dystrophy (DMD) is a severe hereditary disorder caused by a loss of dystrophin, an essential musculoskeletal protein. Decades of promising research have yielded only modest gains in survival and quality of life for these patients and there have been no approved gene therapies for DMD to date. There are two significant hurdles to creating effective gene therapies for DMD; it is difficult to deliver a replacement dystrophin gene due to its large size and current strategies to restore the native dystrophin gene likely require life-long administration of a gene-modifying drug. This thesis presents a novel method to address these challenges through restoring dystrophin expression by genetically correcting the native dystrophin gene using engineered nucleases that target one or more exons in a mutational hotspot in exons 45-55 of the dystrophin gene. Importantly, this hotspot mutational region collectively represents approximately 62% of all DMD mutations. In this work, we utilize various engineered nuclease platforms to create genetic modifications that can correct a variety of DMD patient mutations.Initially, we demonstrate that genome editing can efficiently correct the dystrophin reading frame and restore protein expression by introducing micro-frameshifts in exon 51, which is adjacent to a hotspot mutational region in the dystrophin gene. Transcription activator-like effector nucleases (TALENs) were engineered to mediate highly efficient gene editing after introducing a single TALEN pair targeted to exon 51 of the dystrophin gene. This led to restoration of dystrophin protein expression in cells from DMD patients, including skeletal myoblasts and dermal fibroblasts that were reprogrammed to the myogenic lineage by MyoD. We show that our engineered TALENs have minimal cytotoxicity and exome sequencing of cells with targeted modifications of the dystrophin locus showed no TALEN-mediated off-target changes to the protein coding regions of the genome, as predicted by in silico target site analysis. In an alternative approach, we capitalized on the recent advances in genome editing to generate permanent exclusion of exons by using zinc-finger nucleases (ZFNs) to selectively remove sequences important in specific exon recognition. This strategy has the advantage of creating predictable frame restoration and protein expression, although it relies on simultaneous nuclease activity to generate genomic deletions. ZFNs were designed to remove essential splicing sequences in exon 51 of the dystrophin gene and thereby exclude exon 51 from the resulting dystrophin transcript, a method that can potentially restore the dystrophin reading frame in up to 13% of DMD patients. Nucleases were assembled by extended modular assembly and context-dependent assembly methods and screened for activity in human cells. Selected ZFNs had moderate observable cytotoxicity and one ZFN showed off-target activity at two chromosomal loci. Two active ZFN pairs flanking the exon 51 splice acceptor site were transfected into DMD patient cells and a clonal population was isolated with this region deleted from the genome. Deletion of the genomic sequence containing the splice acceptor resulted in the loss of exon 51 from the dystrophin mRNA transcript and restoration of dystrophin expression in vitro. Furthermore, transplantation of corrected cells into the hind limb of immunodeficient mice resulted in efficient human dystrophin expression localized to the sarcolemma. Finally, we exploited the increased versatility, efficiency, and multiplexing capabilities of the CRISPR/Cas9 system to enable a variety of otherwise challenging gene correction strategies for DMD. Single or multiplexed sgRNAs were designed to restore the dystrophin reading frame by targeting the mutational hotspot at exons 45-55 and introducing either intraexonic small insertions and deletions, or large deletions of one or more exons. Significantly, we generated a large deletion of 336 kb across the entire exon 45-55 region that is applicable to correction of approximately 62% of DMD patient mutations. We show that, for selected sgRNAs, CRISPR/Cas9 gene editing displays minimal cytotoxicity and limited aberrant mutagenesis at off-target chromosomal loci. Following treatment with Cas9 nuclease and one or more sgRNAs, dystrophin expression was restored in Duchenne patient muscle cells in vitro. Human dystrophin was detected in vivo following transplantation of genetically corrected patient cells into immunodeficient mice. In summary, the objective of this work was to develop methods to genetically correct the native dystrophin as a potential therapy for DMD. These studies integrate the rapid advances in gene editing technologies to create targeted frameshifts that restore the dystrophin gene around patient mutations in non-essential coding regions. Collectively, this thesis presents several gene editing methods that can correct patient mutations by modification of specific exons or by deletion of one or more exons that results in restoration of the dystrophin reading frame. Importantly, the gene correction methods described here are compatible with leading cell-based therapies and in vivo gene delivery strategies for DMD, providing an avenue towards a cure for this devastating disease.

Background Transcription activator-like effector nucleases (TALENs) have emerged as a tool for enabling targeted gene editing and disruption in difficult systems, such as human pluripotent stem cells (hPSCs). The modular architecture of TAL effectors theoretically enables targeting of any genomic locus and several cloning systems for custom TALEN assembly have recently been established. However, there is a lack of versatile TALEN expression systems applicable to hPSCs. Results Here, we extend an existing TALE assembly system by a dual set of expression vectors for efficient application of TALEN technology in hPSCs. This is characterized by improved TALEN architecture as well as antibiotic resistance and fluorescent reporter cassettes, thus enabling enrichment for transfected cells. Improved functionality of the combined system was demonstrated by targeted disruption of the HPRT1 gene to create isogenic disease models of Lesch-Nyhan-Syndrome. Using female hPSCs, homozygous disruption of HPRT1 occurred at efficiencies of up to 15%. Differentiating isogenic knock-out cells both into central nervous system (CNS) as well as into sensory-like neurons recapitulated previously described phenotypes based on patient-specific induced PSCs and extended these findings to non-CNS neurons, respectively. Conclusion The combined vector system allows for flexible and affordable generation of knock-out hPSCs lines, thus enabling investigation of developmental processes as well as the generation of isogenic disease models without the need for patient material.