Human embryonic stem cells can be grown for months in the laboratory and coaxed into specialized cell types, making them a cornerstone of regenerative medicine research. But every round of cell division carries a small risk of new DNA errors, so researchers must monitor the genomes of these cells carefully before considering them for therapies. A new study published in BMC Genomics used longitudinal whole-genome sequencing to track genetic changes in established stem cell lines over long periods of culture and during differentiation into neurons.
The research team, led by Qian Li, Bo Zhou, Hui Zheng and Tie-Shan Tang, cultured human embryonic stem cell lines in E8 medium, a chemically defined formulation designed to reduce variability associated with complex culture conditions. They sequenced the genomes of these cells across 40 passages, meaning the cells were split and expanded 40 separate times, and also examined them as they differentiated into dopaminergic neurons, the cell type lost in Parkinson’s disease.
The sequencing revealed a low burden of detectable acquired mutations across the culture period. The mutations that did appear were dominated by clock-like signatures, which accumulate steadily with each cell division, and by oxidative stress-related signatures, which point to damage from reactive oxygen species generated during normal metabolism. These patterns are consistent with background replication errors and routine ROS exposure rather than any culture-induced crisis.
Most of the detected variants showed low variant allele frequencies, meaning only a small fraction of cells in each culture carried the mutation. The variants also showed limited recurrence across the sampled passages, and the researchers found no consistent increases in variant allele frequency that would suggest strong positive selection or clonal expansion, in which a single mutated cell multiplies and takes over the population. The authors noted these conclusions depend on the sequencing depth and variant-calling criteria used in the study.
One finding stood out from this generally reassuring picture. A frameshift variant in the BCOR gene, designated p.F914Wfs*2, was detected at passage 40 in the E8-cultured Q19 line with a variant allele frequency of 0.53, indicating the mutation was present in roughly half the cells sampled. However, the researchers emphasized that the longitudinal dynamics and biological significance of this variant remain unresolved, and they could not determine from the available data whether it arose late in culture or was present earlier at undetectable levels.
To test whether the low mutation burden extended beyond a single culture formulation, the team also examined stem cells grown in AIC medium, another chemically defined formulation, across approximately 50 passages. Similar low overall mutation burdens were observed in these cultures, supporting the idea that optimized defined media can support genomic stability during extended propagation in the examined lines.
The study’s conclusions come with important caveats. The findings apply to the specific hESC lines examined, and the authors stressed that larger studies across additional genetic backgrounds are needed before generalizing. They also called for predefined genomic quality-control criteria and continued genomic monitoring as essential steps prior to any clinical application of these cells. The work does not establish that any culture condition guarantees safety, but it provides a detailed baseline for what normal mutation accumulation looks like in well-maintained stem cell cultures.
For the stem cell field, the results offer a measure of reassurance combined with a reminder of vigilance. Chemically defined media such as E8 and AIC appear to keep the detectable mutation load low over dozens of passages, and differentiation into dopaminergic neurons did not reveal a surge of new mutations under the study’s detection limits. Yet the BCOR frameshift finding illustrates why longitudinal sequencing matters: a single variant with a high allele frequency can appear by late passage, and only sustained monitoring across passages can distinguish transient low-level changes from events that warrant deeper investigation before cells are used in translational research or therapy development.
Subject of Research: Longitudinal whole-genome sequencing of human embryonic stem cells during extended culture and neuronal differentiation
Article Title: Longitudinal whole-genome analysis of human embryonic stem cells during extended expansion and differentiation
Article References: Li, Q., Zhou, B., Zheng, H., Li, Z., Wang, L., Wang, S., Wang, L., Feng, L., Li, W., Li, T., Guo, C., Wang, Y., & Tang, T.-S. (2026). Longitudinal whole-genome analysis of human embryonic stem cells during extended expansion and differentiation. BMC Genomics. https://doi.org/10.1186/s12864-026-13507-y
Image Credits: AI Generated
DOI: 10.1186/s12864-026-13507-y
Keywords: human embryonic stem cells, whole-genome sequencing, genomic stability, E8 medium, dopaminergic neurons, mutation signatures, variant allele frequency, BCOR frameshift, clonal expansion, regenerative medicine, chemically defined media, passage number
