Unravelling the puzzle of malignancy in human pluripotent stem cells

Human pluripotent stem cells (hPSCs), encompassing embryonic and induced pluripotent stem cells, have transformed biomedical research with their unique ability to self-renew indefinitely and differentiate into virtually any human cell type. This makes them powerful tools for disease modeling, regenerative medicine, and tissue engineering. However, a critical barrier to their safe clinical use lies in their malignant potential: under certain conditions, some hPSC lines can form tumors with immature, rapidly dividing components after transplantation. This duality underscores the importance of understanding and detecting malignancy risk.

The dissertation approaches malignancy in hPSCs holistically, addressing both its biological underpinnings and current as well as potential detection methods.

We explore a major driver of malignant transformation: the accumulation of genetic aberrations during long-term culture. Certain chromosomal regions—especially those containing oncogenes—are particularly vulnerable, with recurrent gains on chromosomes 12 and 20 being the most common, observed in roughly 15–20% of hPSC lines worldwide. These genetic changes mirror alterations seen in human cancers, suggesting shared mechanisms of transformation. Evidence from murine pluripotent stem cells, which develop similar aberrations, further supports the existence of conserved pathways across species.

Beyond large chromosomal changes, gene-level mutations also contribute to malignant traits. Mutations in TP53, a key tumor suppressor, exemplify this. We explore this by generating TP53-deficient hPSCs, which initially retain normal pluripotency, but gradually develop genomic instability, enhanced survival under stress, and resistance to chemotherapy—features associated with malignancy. These findings highlight the necessity of routine genomic surveillance during hPSC culture to reduce transformation risk.

Current malignancy assessment relies heavily on the teratoma assay, in which hPSCs are injected into immunodeficient mice to evaluate both pluripotency and tumorigenic behavior. Through a systematic review, we observe that while widely used, the assay suffers from poor standardization, limited reproducibility, and ethical concerns over animal use. Moreover, most studies apply it primarily to confirm pluripotency rather than malignancy. Testing in humanized mice, which possess a human-like immune system, did not significantly improve predictive accuracy, calling into question the utility of modifying in vivo approaches.

To address these limitations, the dissertation also explores in vitro alternatives. Early models based on embryoid bodies distinguished between benign and malignant lines, but their reliance on fetal bovine serum compromised both reproducibility and ethical acceptability. Efforts to transition to serum-free conditions improved pluripotency assessment but failed to reliably capture malignancy traits. More promising results emerged from a pilot co-culture system, where embryoid bodies were grown alongside fibroblasts and endothelial cells, better mimicking the in vivo microenvironment. This approach enhanced differentiation and showed potential to improve malignancy detection without animal-derived components.

Taken together, these findings assemble important pieces of the malignancy puzzle in hPSCs. By clarifying the genomic and molecular drivers of transformation, exposing the shortcomings of current assays, and piloting more ethical and reproducible alternatives, the work brings the field closer to developing a standardized in vitro platform for malignancy detection. Such a system would mark a significant step toward ensuring the safe clinical translation of hPSC-based therapies.