iPSC Disease Modeling

See allHide authors and affiliations

Science  30 Nov 2012:
Vol. 338, Issue 6111, pp. 1155-1156
DOI: 10.1126/science.1227682

Induced pluripotent stem cell (iPSC) technology has provided previously unanticipated possibilities to model human disease in the culture dish. Reprogramming somatic cells from patients into an embryonic stem cell–like state (1) followed by differentiation into disease-relevant cell types can generate an unlimited source of human tissue carrying the genetic variations that caused or facilitated disease development (2). Yet, despite the excitement over this “disease-in-a-dish” approach, studying genetic disorders in patient-derived cells faces more challenges than studies using genetically well-defined model systems. Here we describe some of these limitations, and also present some solutions for ensuring that iPSC technology lives up to at least some of its promise.

Disease in a dish.

The limitations of iPSC technology and emerging solutions for identifying disease-related phenotypes in vitro.

Individual iPSC lines, independent of disease status, display highly variable biological properties (35). This makes their propensity to differentiate into specific functional cell types unpredictable, thereby limiting their value for studying disease-specific phenotypes. The many reasons for these cell-to-cell differences can be classified into three categories (see the figure): cellular changes resulting from the reprogramming process; culture-induced differences due to the lack of robust differentiation protocols; and differences in genetic background.

Cellular changes caused by reprogramming are especially problematic if the vectors integrate into the host genome. This can result in disruption or dysregulation of nearby genes and often entails residual expression of the reprogramming transgenes (6). Additional complications are caused by incomplete epigenetic reprogramming (7) influenced in part by the stoichiometry of the reprogramming factors and specific culture conditions (8, 9), transcriptional derepression of genes on the inactivated X chromosome (10), and genetic alterations such as point mutations and copy number variations (CNVs) (11, 12). However, more advanced nonintegrating reprogramming technologies such as plasmid or mRNA transfection and protein transduction, improved growth conditions, and more stringent quality control should adequately address these limitations (1, 13).

A major reason for cell-to-cell variability is the lack of robust in vitro differentiation protocols. Most protocols rely on growth factor signals and regulators that play essential roles at specific stages of normal embryonic development. Although fairly efficient in generating some cells of interest, these protocols typically produce a mixture of diverse cell types, which is a crucial limitation when the goal is to create highly controlled conditions. One widely discussed improvement is to introduce reporter or selection genes under the control of lineage- or cell-type–specific promoters (14), thereby allowing the identification, selection, and quantification of specific cell types.

Genetic background variations present a particular impediment to the disease-in-a-dish approach owing to the uncontrolled impact of genetic modifier loci. Epistatic effects from genetic background are observed even in the most prevalent monogenetic disorders causing, for example, variable age of onset and/or disease progression. An individual typically differs from the reference genome at several thousand sites in protein-coding genes, hundreds of which are predicted to alter protein function with many implicated in inherited disorders (1517). Likewise, there is a major influence of genetic variants on gene expression differences across individuals (18). Therefore, the comparison of patient-derived iPSCs to cell lines from healthy donors that share no or only partial genetic background poses considerable risks for interpreting a supposedly disease-related phenotype. This becomes particularly relevant for most late-onset diseases that typically show slow progression of pathophysiological changes and therefore are expected to display only subtle changes in vitro, which may not fully manifest during a short study period.

Recent advances in gene-editing technologies enable the targeted modification of human cells for gene disruptions, genetic repair, or insertion of reporter genes (14). The ability to modify single base pairs, thereby seamlessly correcting or introducing disease-causing mutations in human pluripotent stem cells, allows the creation of genetically controlled experimental model systems in which the disease-causing genetic variation is the sole experimental variable (1921). This could substantially simplify the analysis of the interaction between these genomic variants and disease phenotypes, thereby revealing new insights into the pathophysiology of monogenic and complex diseases.

The combination of iPSC, gene editing, and genome-wide technologies gives us the opportunity to systematically and faithfully model human disease in relevant human cell types. Acknowledging the inherent limitations and subjecting iPSC technology to the same rigor that has become standard in other model systems will make it an indispensable tool for biomedical research.


Stay Connected to Science

Navigate This Article