The potential of human induced pluripotent stem cells (hiPSCs) for research, drug development and, eventually, cell-based therapy, is now widely acknowledged, but there are many hurdles that remain to be crossed. One relatively unexamined question about these cells is whether, like human embryonic stem cells, they may have a propensity for acquiring genetic aberrations.

In collaborative work conducted by researchers at the Hebrew University in Jerusalem, the Cedars-Sinai Medical Center in Los Angeles and the University of California, Los Angeles, a group led by Nissim Benvenisty isolated embryonic stem cells from human embryos determined by preimplantation genetic screening to have trisomies of specific chromosomes. The scientists noticed during the course of this work that these trisomic lines had expression levels well above normal for genes on the trisomic chromosomes. This mirrored observations that have been made in other systems, notably in yeast and in cancer cells, and it led the scientists to the idea that a meta-analysis of gene expression profiles of hiPSCs might identify aneuploidies and other aberrations in the genome.

The researchers combined microarray-derived gene expression profiles for pluripotent stem cell lines from 18 independent studies, with 38 human embryonic stem cell lines and 66 hiPSC lines (some of them subclones, in the latter case). They then statistically examined normalized expression levels of about 12,000 genes in each sample, and identified regions on the chromosomes for which expression was unusually high or low. These regions were candidates for harboring a genetic change, a duplication or a deletion.

For a subset of the samples, the researchers could compare the results of the gene expression analysis with either karyotyping or higher-resolution array-based studies of copy-number variation. In all cases for which data were available on comparable samples, the results matched. However, especially for smaller, subchromosomal changes, the number of samples was not large enough to draw conclusions about the sensitivity of the approach. “For trisomies and chromosomal arms, [using gene expression is] very sensitive and specific,” says Yoav Mayshar, first author on the paper reporting the study, “but for small changes, to really know the limits, we'd need a larger dataset.”

One of the strengths of using gene expression profiles to identify genetic aberrations is that the data are readily available. “Ideally we would do the same experiment using [single-nucleotide polymorphism] arrays or some such approach,” says Mayshar, “but this was really the only feasible way to get our hands on data from so many different lines, different labs, different passage numbers, different types of reprogramming.” What is more, once a chromosomal region has been identified as aberrant in this way, one already has some information about functional expression changes that may result.

This is not to say that the approach does not have disadvantages as well. It is not as high-resolution as some of the more direct approaches: the smallest change the researchers detected so far was of about ten megabases. Furthermore, it could be confounded by epigenetic effects on gene expression, and it is likely to be less successful than other approaches on heterogenous samples.

Nevertheless, Mayshar, together with Uri Ben-David and their colleagues, identified abnormalities in about a fifth of the 66 hiPSC lines tested. The data at this stage are probably too limited in scale to suggest that particular reprogramming methods are more or less prone to problems—the analyzed dataset included lines generated using retroviral integration, protein-based reprogramming and episomal vectors—but it remains to be seen with future work on larger numbers of lines whether such patterns emerge.