Genetic variation is a major source of transcriptional variation in human induced pluripotent stem cells. N. Kumasaka, F. Rouhani, A. Bradley, L. Vallier, D. Gaffney Wellcome Trust Sanger Institute, Cambridge, United Kingdom.

   Induced pluripotent stem cells (iPSCs) are a potentially powerful model system for studying human disease and development. However, iPSC lines have been reported to exhibit substantial transcriptional and epigenetic variability, perhaps driven by stochastic effects during cellular reprogramming or epigenetic memory of the somatic tissue from which they were derived. Here we report the first systematic comparison of the transcriptome of ESCs, iPSCs and their somatic progenitors using high depth sequencing. We performed RNA-seq in a panel of 44 human iPSC lines and somatic tissues obtained from the same four unrelated, healthy subjects. We show that genetic background variation between different individuals is likely to be a major source of transcriptional variation between iPSC lines. In contrast, the transcriptome of iPSCs is largely stable across different tissues of origin and transcriptional memory appears to be uncommon. Our data illustrate that variation between biological replicates of iPSCs derived the same reprogramming event is not dissimilar to that between different passages of adult cells or ESCs. We detect extensive signatures of allele-specific expression at non-imprinted genes in iPSCs and also replicate known expression quantitative trait loci (eQTLs) detected in somatic tissues. Our results suggest that epigenetic and transcriptional variation between iPSCs and ESCs may often reflect underlying genetic differences rather than the effects of cellular reprogramming or residual memory of the tissue of origin. In addition, we clearly show that genetically driven changes in key cellular phenotypes such as transcription are readily detectable using iPSCs as a model system.