Large (left, inset) and small (right, inset) centromeres allow for detection of chromosome loss events in ES cells.


Any cell therapy risks introducing mutated and possibly tumorigenic cells. Now results from Peter Stambrook (University of Cincinnati, Cincinnati, OH) and colleagues underline this concern for embryonic stem (ES) cells, which are one possible source for therapeutic cells.

Stambrook constructed a cell line in which he could detect loss or mutation of a heterozygous marker gene. Loss (rather than mutation) of the functional allele was the proximate cause in ∼80% of the events with both mouse embryonic fibroblasts (MEFs) and ES cells. But marker-deficient colonies arose 400-times faster in MEFs than in ES cells—good news for ES cell proponents.

But a closer look revealed complications. Whereas all the loss events in MEFs were caused by mitotic recombination, a full 57% of the ES cell events were apparently caused by loss of the entire chromosome, followed by reduplication of the single remaining chromosome. Such events have the potential to uncover multiple mutant alleles on the single remaining chromosome, and thus pose a greater risk than a more spatially limited recombination event.

Does this mean that ES cells are unfit for use? Stambrook thinks not. “This is not an argument against using ES cells for therapeutic uses,” he says. “It's just a caution that one should be aware about.” Stambrook says it should be possible to screen ES cells for chromosome loss events. Furthermore, the necessary comparisons between cell types have not yet been completed. Loss frequencies for adult stem cells are not yet known, and similar loss events may occur in the MEFs, but be masked by the more frequent mitotic recombination events.

Recombination events between nonidentical sequences may be suppressed in ES cells by the mismatch repair machinery, at least according to evidence from homologous recombination experiments. Before looking for a possible cause for increased chromosome loss in ES cells, researchers will need to confirm that there is indeed a difference in rates between ES and somatic cells. ▪


Cervantes, R.B., et al.
Proc. Natl. Acad. Sci. USA.