Modeling Cancer: The Role of Epigenetics

Widely used cancer models do not replicate DNA methylation patterns in tumors

DNA methylation is a key epigenetic modification that influences gene activity without altering the underlying DNA. In a study published in Genome Biology, the Smith and Meissner labs systematically characterized typical DNA methylation patterns in cancers and assessed how well 21 engineered human and mouse cancer models reproduce these features. The researchers found that these models, which are widely used to study tumorigenesis and develop new treatments, fail to recapitulate these key characteristics.

What began as an investigation into the role of epigenetics in cancer soon took an unexpected turn: "In the lab, we have been studying DNA methylation as an epigenetic modification in various contexts," says Sara Hetzel, the study's first author. "One important context is cancer, as DNA methylation changes characteristically in most human tumor types. However, we still don’t know exactly when this happens or what function it serves." Because patient-derived tumors already have altered epigenetic landscapes, they are unsuitable for studying the role of this phenomenon in early tumorigenesis. To pinpoint when and how these changes arise in cancers, the researchers turned to model systems that mimic the transition from healthy to diseased cells. Surprisingly, they initially did not observe any changes in methylation patterns in the models, prompting a deeper investigation.

First, the team analyzed patient data to define the tumor DNA methylation landscape compared to that of healthy human cells. They then used this benchmark to evaluate 19 mouse models and two human cell culture systems through data analysis and bioinformatics. "To our surprise, these model systems rarely adopt a methylation state that faithfully mimics human tumors," says Sara Hetzel. “This is an important finding because it shows that current model systems used to study tumor initiation lack a feature present in nearly all human cancers.”

While in some cases DNA methylation can silence certain tumor suppressor genes and directly contribute to tumorigenesis, the broader role of methylation changes in cancer development and progression remains poorly understood. Nonetheless, tumor methylation data are important for classifying cancer types, identifying biomarkers, and predicting patient responsiveness to certain treatments. The methylation state of model systems, however, is often overlooked. "Our findings demonstrate the need to develop new models that enable us to study the role of epigenetic patterns in cancer development," says Alexander Meissner. "An improved understanding will enable the identification of potential new treatments.”