248 Chapter 9 RESULTS Modeling DCM Contractility Defect With hiPSC-CMs To model PLN R14del DCM, we recruited 2 unrelated families carrying the mutation (Figure I in the Data Supplement) to generate hiPSCs from carriers. Isogenic control lines in which the genetic background is identical are important for unequivocal assignment of phenotype to the underlying gene variant. Therefore, we corrected the R14del mutation in both patient hiPSC lines through CRISPR/Cas9–mediated genome editing and similarly introduced the R14del mutation into an hiPSC line derived from an individual with no history of heart disease (healthy donor [HD]; Figure 1A and 1B). We generated 3 pairs of isogenic models that differ only in the PLN R14del mutation (Figure 1C, Figure II, and Tables I-IV in the Data Supplement). Impaired contractility is a pathological hallmark of DCM; therefore, we assessed the contractile function of the hiPSC-CMs. Consistent with previous studies,5 hiPSCCMs carrying the PLN R14del mutation (patient and HD R14del introduced) showed decreased contractility in 3-dimensional engineered heart tissues (EHTs; Figure 1D and 1E) and 2-dimensional monolayer cultures (Figure 1F and 1G). Therefore, the isogenic paired hiPSCs lines recapitulated the contractile deficit, providing a quantifiable model for determining the molecular mechanisms that underlie the development of DCM. scRNA-seq Analysis Reveals UPR Activation Single-cell transcriptomic analysis makes it possible to deconvolute the complex transcriptional responses that occur naturally across populations of cells into clusters of similarly responding cells. We used a high-throughput droplet-based scRNA-seq method (10X Genomics) to examine the transcriptional effects caused by introducing the PLN R14del mutation into an HD hiPSC line. Unbiased t-distributed stochastic neighbor embedding clustering parsed 9244 single-cell transcriptomes from PLN wild-type (WT) hiPSC-CMs (5279 cells) and PLN R14del hiPSC-CMs (3965 cells) into 10 distinct subpopulations (Figure 2A). The heterogeneity of the cardiomyocyte cultures was similar to previous scRNA-seq studies.6 Despite this heterogeneity, t-distributed stochastic neighbor embedding analysis revealed close clustering of the PLN mutant line with its isogenic counterpart, indicating that the transcriptomes and cell type composition are preserved across isogenic populations (Figure 2B). To assign cellular identity, subpopulations were classified on the basis of known marker genes. Most of the cell clusters appeared to be cardiomyocytes based on their specific patterns of gene expression (clusters 1–4 and 6–7 expressed marker genes TNNT2, MYL2, and MYH7). Two clusters, clusters 5 and 8, were identified as fibroblasts (expressing marker genes COL3A1, COL1A1, and FN1), and 1 cluster with very few cells, cluster 9, was identified as smooth muscle cells (with marker genes ACTA2 and TAGLN; Figure 2C and 2D).
RkJQdWJsaXNoZXIy MTk4NDMw