Genome Engineering Human iPS Cells to Model and Treat Disease

− この都医学研セミナーは終了しました。 −

講演要旨

Our research uses genome engineering in iPSC-derived tissues to gain mechanistic insights, and develop treatments via pharmaceutical or genetic correction. We have engineered over 30 different human isogenic iPS cell lines that are now yielding phenotypes in iPS-derived cardiomyocytes, and help to explain the molecular basis of several types of cardiomyopathy. All the mutations were made in the wild-type iPSC line (WTC) that used as a reference by several public resources (www.ConklinLab.org and www.allencell.org). We made an allelic series that include multiple versions of RBM20, BAG3, PRKAG2, PKP2, MYBPC3, PHOX2B, as well as several versions of the CRISPRi system. By studying an allelic series in an isogenic iPSC background, the relative effects of different disease mutations are easier to decipher that studying the same mutations in mixed genetic backgrounds. The iPS-derived cardiomyocytes allow dynamic imaging of electrical activity, contraction, cellular structure and protein quality control. Initial studies have revealed the potential mechanism of chemotherapy toxicity, and unique therapeutic targets.

We are also developing methods to “fix” disease genes with the goal of developing methods for genome surgery is specific tissues. Genome surgery faces major challenges to ensure that only a specific part of the genome is edited, without causing “off-target” DNA damage. Our initial genome surgery efforts are focused on the retina and motor neurons since these tissues have the advantage of a limited number of cells that need to be safely targeted for therapy with RNPs (RNA protein complexes of Cas9 and gRNA). There are hundreds of dominant mutations in genes (e.g. BEST1, MFN2, NEFL2, HSPB1), resulting in incurable disease. Most mutations involve relatively few patients so it is likely that many genome surgeries will require unique Cas9-gRNA combinations. This challenge can be simplified by targeting common ancestral SNPs so that just 10 RNPs may be able to treat >90% BEST1 patients, based on and the analysis of >2,500 phased genomes. We are modeling genome surgery in iPS-derived tissues from patients so that off-target analysis can be done prior to surgery. We anticipate that lessons from each patient will provide lessons that will eventually provide a path to cost-effective therapy. We are hopeful that proof-of-concept therapeutic editing methods could be expanded to other tissues in the future.