Systems-Wide Approaches in Induced Pluripotent Stem Cell Models

Abstract

Human induced pluripotent stem cells (iPSCs) provide a renewable supply of patient-specific and tissue-specific cells for cellular and molecular studies of disease mechanisms. Combined with advances in various omics technologies, iPSC models can be used to profile the expression of genes, transcripts, proteins, and metabolites in relevant tissues. In the past 2 years, large panels of iPSC lines have been derived from hundreds of genetically heterogeneous individuals, further enabling genome-wide mapping to identify coexpression networks and elucidate gene regulatory networks. Here, we review recent developments in omics profiling of various molecular phenotypes and the emergence of human iPSCs as a systems biology model of human diseases.

Keywords: disease modeling; induced pluripotent stem cells; proteomics; systems biology; transcriptomics.

Figure 1.. Omics approaches in iPSC models of human diseases.

1. Populations of healthy and/or diseased individuals donate skin or blood cells. 2. iPSCs are derived from donors to capture their genetic backgrounds. 3. The individual-specific iPSCs are coaxed into differentiated cells resembling primary tissues including cardiomyocytes, neurons, and hepatocytes. 4. The resulting iPSC-derived cells are used to profile (4a) live cell functional phenotypes and (4b) molecular expression. 5. Large-scale profiling data are analyzed to discern molecular mechanisms responsible for cellular phenotypes and disease traits.

Figure 2.. Identify disease networks using QTL and co-expression analysis.

1. Gene variants and expression profiles are acquired in large iPSC panels. 2a. QTL studies use of genetic variants as a causality anchor to map out relationships between genes and traits (gene regulatory networks). 2b. Co-expression network modeling take advantage of co-variation in expression profiles of functionally related genes across individuals to generate hypotheses on the underlying regulations of genetic program.

Figure 3.. Genes conceptualized as linear pathways vs. complex networks.

In the linear pathway worldview, disease genes are identified via conspicuous differential expression of one or few molecular markers along known pathways (differential profile). In the complex network view, differences in subnetwork memberships can be analyzed to detect changes in networks or subnetworks between two states and to identify disease modules.