br Introduction Human induced pluripotent stem

Introduction Human induced pluripotent stem Ezetimibe manufacturer (hiPSCs) carry the whole genetic context of a patient and can be a virtually unlimited source of differentiated cell types of interest. Increasingly, hiPSCs are being applied for disease modeling and drug screening for a number of neurodegenerative disorders, including Alzheimer\'s disease, Parkinson\'s disease, amyotrophic lateral sclerosis, and Huntington disease (HD) (Sterneckert et al., 2014). Indeed, a number of valuable disease phenotypes have been uncovered using differentiated neuronal subtypes of disease relevance (Ichida and Kiskinis, 2015). However, one of the challenges of hiPSC-based disease modeling is the variability in differentiation potential due to variations in their genetic background (Kajiwara et al., 2012). This may result in inappropriate interpretation of disease phenotypes in vitro. Furthermore, the influence of genetic background on disease phenotype can be significant even for monogenic, dominant, and highly penetrant diseases such as HD. Indeed, a recent genome-wide association study identified a number of loci with putative disease-modifying variants that appear to influence the age of neurological onset in HD (GeM-HD Consortium, 2015). Therefore, the use of control hiPSC lines that are genetically identical is crucial to increase confidence in candidate disease phenotypes and mechanisms, and to minimize the chance of missing modifiable effects that are of relevance to the pathogenesis of disease. With recent advances in genome-editing technologies, such as the zinc finger nucleases, transcription activator-like effector nuclease, and the CRISPR-Cas9 system, the establishment of isogenic control hiPSCs has become more feasible. Here, we describe the seamless correction of an HD hiPSC line. HD, the most common genetic cause of dementia, results from an expansion of a polymorphic CAG repeat tract in exon 1 of HTT (Group, 1993). Using a CRISPR-Cas9 and a piggyBac transposon-based selection system (Yusa, 2013), we corrected HD hiPSCs and established isogenic control hiPSCs with seamless excision of the selection cassette. Evaluation of the corrected lines demonstrates that a number of phenotypic abnormalities and gene expression changes in HD hiPSC-derived neural cells are rescued in isogenic controls. Our study highlights the utility of isogenic controls in distinguishing HD-specific molecular phenotypes from those related to the genetic background.
Results
Discussion Using a CRISPR/Cas9 nickase- and piggyBac transposon-based HR approach, we demonstrate that the expanded trinucleotide repeat in HTT can be efficiently corrected in hiPSCs. We show that corrected isogenic hiPSC lines retain pluripotency and normal karyotypes, and can be differentiated into excitable and synaptically active neurons. We further show that a number of phenotypic abnormalities in HD hiPSC-derived neural cells, including impaired neural rosette formation, increased susceptibility to growth factor withdrawal, and deficits in mitochondrial respiration, are rescued in isogenic controls. Importantly, using genome-wide expression analysis, we reveal that a number of apparent differences in gene expression seen when comparing HD with non-isogenic control lines are not seen when comparing HD with isogenic corrected lines, indicating that these differences are likely related to genetic background and are not HD-specific effects. We also show that while gene correction may reduce the phenotypic variability related to genetic background, variability due to clonal differences remains, highlighting the importance of assessing multiple clones, even when using isogenic controls. Genetic correction of HD hiPSCs by HR was reported previously using CRISPR/Cas9-assisted methods (An et al., 2014). However, removal of a selection cassette from the genomic locus of the HTT gene after targeting was not reported. As the presence of a selection cassette may affect HTT expression and regulation, its removal is important to ensure optimal modeling of disease effects.