New and exciting advances in our capacity to manipulate biology often seem to be pegged back by our realization that we did not understand quite as much as we thought. This currently looks to be the case with human induced pluripotent stem cells (HiPSCs) — cells that have been reprogrammed from a differentiated state to a pluripotent state and can, as far as we can tell, function similarly to human embryonic stem cells (HESCs). However, as discussed in a Research Highlight on page 232 and a Review on page 268, HiPSCs differ from HESCs in worrying ways. It would seem that the reprogramming process is a harsh one that results in genetic changes, and in some regions of the genome epigenetic reprogramming seems to be incomplete or aberrant. Moreover, the pattern of the mutations in protein-coding genes in HiPSCs is similar to that evident in cancer cells. In addition, as Uri Ben-David and Nissim Benvenisty point out in their Review, HiPSCs can form teratocarcinomas, as opposed to the benign teratomas that are formed by HESCs, and might also give rise to somatic tumours. The finding that protein-coding changes and aberrant methylation patterns are retained in HiPSCs during prolonged culture, and that the differences in methylation are not lost during differentiation, clearly indicates that much more study of HiPSCs is needed before we can truly enter a world of regenerative medicine.

However, these findings are not all doom and gloom. If the mutational similarities between HiPSCs and cancer cells pan out, then understanding in detail what happens during the reprogramming process might provide another means of appreciating how tumour cells arise and might identify new markers for the early detection or treatment of pre-neoplastic lesions.