Studying a deadly type of breast cancer called triple negative, Johns Hopkins Medicine scientists say they have identified key molecular differences between cancer cells that cling to an initial tumor and those that venture off to form distant tumors.
A report on the findings was published Aug. 3 in Science Translational Medicine.
For the current study, the research team looked at molecular differences between initial, or primary, triple-negative breast cancer sites and areas where it spread, or metastatic sites, among three different types of cells: mouse models, human cancers implanted into mice and samples of both primary and metastatic tissues taken from eight patients treated at The Johns Hopkins Hospital.
The researchers used a combination of machine learning, cellular imaging, and biochemical analysis to identify differences in the genetic expression patters of initial and metastatic tumors.
The scientists found that when triple negative breast cancer cells invade other tissues on their way to another part of the body, they gain two cellular properties: better movement and survival.
To do this, breast cancer cells gain a cellular skeleton protein dubbed vimentin, which enhances the migratory ability of so-called mesenchymal cells, a cell type typically found in bones and bone marrow that moves around and makes new cells.
Triple negative breast cancer cells also gain survival advantages by producing a protein called E-cadherin, typically found in epithelial cells that line the ducts and coverings of organs and frequently renew themselves.
When triple negative breast cancer cells gain such survival and migratory qualities, scientists classify their cellular state as so-called hybrid epithelial mesenchymal (EMT) cells.
To look more closely at molecules involved in hybrid EMT states, the scientists sought the help of Elana Fertig, Ph.D., division director and associate director of quantitative sciences and co-director of the Convergence Institute at the Johns Hopkins Kimmel Cancer Center, to track the molecular patterns of individual cells in cell assays that model invasion out of the primary tumor and formation of a colony in a metastatic site.
Fertig’s computational team used machine learning techniques to find patterns among each cell’s expression of RNA, a cousin of DNA involved in protein production. The scientists found that most of the metastatic cells morph into the more mobile, better surviving, hybrid EMT state. Ewald’s team then validated these states in samples from eight patients with triple negative tumors, examining both primary tumors and tissues from metastatic sites of the same patients.
At the molecular level, the most metastatic cells produced five proteins called transcription factors (Grhl2, Foxc2, Zeb1, Zeb2 and Ovol1) that foster the making of proteins involved in either cancer cell invasion or colony formation.
Johns Hopkins Medicine release
