50 Million euro investment for LUMC spin-off RougeTx advances therapies for inherited bleeding disorders.
Researchers from the groups of Associate Investigator Richard Davis and Principal Investigator Christine Mummery at reNEW Leiden, together with biotech company Sartorius, investigated how the chemotherapy drug doxorubicin harms the heart using advanced stem cell models and Artificial Intelligence (AI)-based image analysis. Published in Stem Cell Reports, the study shows that interactions between different cardiac cell types can influence toxic responses and highlights the importance of multi-cell type models for studying cardiotoxicity effects.
Chemotherapy drugs such as doxorubicin can be lifesaving, but they can also lead to long-term cardiac side effects in cancer survivors. Much remains unknown about how cumulative exposure to these drugs triggers heart damage, limiting opportunities for earlier detection and prevention.
Together or alone
To learn more about these mechanisms, the team differentiated human pluripotent stem cells into three cardiac cell types: cardiomyocytes, fibroblasts, and endothelial cells. Davis explains why this is important: “Many heart models used in the laboratory contain only cardiomyocytes – the cells responsible for contraction of the heart. However, we know that these models are often too simplistic to provide all the necessary insights into the harmful effects of drugs on the heart.”
The researchers exposed each cell type alone and in combinations to repeated short exposures of doxorubicin to mimic cumulative dosing and followed their responses over a week using live imaging.
AI analysis software
“Together with Sartorius (who produce the Incucyte® Live-Cell Analysis System that was used in this study), we developed machine learning based software able to perform automated analysis of large imaging datasets. Importantly, we could distinguish cardiomyocytes from the other cell types in co-cultures, and quantify early signs of cell injury in various cell types completely label-free,” says first author Dr. Marcella Dias Brescia. “This enables monitoring of changes in our cultures in near real time.”
Different than expected
The analysis revealed an unexpected finding. “Contrary to what we expected, endothelial cells showed the earliest and strongest sensitivity to doxorubicin, and their presence altered how quickly cardiomyocytes became injured in co-culture,” Davis explains. “This is a finding we could only make by including multiple cell types in our model systems.”
A foundation for the future
The results emphasize the importance of multi-cell type heart models in studying cardiac toxicity and lay a foundation for future research.
“By combining multi-cell type human heart models with label-free imaging and AI, we now have a scalable platform to compare the cardiotoxicity of related compounds in the same drug family as doxorubicin. In addition, we can use the platform to explore patient-specific responses to a variety of drugs such as these chemotherapeutics using cardiac cells derived from different individuals,” concludes Mummery.
