Such “cell-in-cell” structures are common in tumors, but the mechanism of invasion was unknown. By labeling human breast cancer cells with different colors, the authors showed that 25% of cells contained other cells within 12 hours of detachment from their substrate. While apoptosis of one cell can drive phagocytosis by another, blockade of either process did not diminish the rate of invasion.
Invasion was suppressed, however, by stopping actin–myosin II contraction in the internalized cell. “The process requires activity of the invading cell,” Brugge says. The group found no evidence that the host initiates the process, and contraction blockade in the host had no effect. Cadherins, which link epithelial cells together, were required and were densest at contacts between the two cells during internalization.
Both cadherins and actin–myosin contraction feature prominently in epithelial compaction, through which multiple layers of attached cells condense into a dense monolayer. The authors suggest that the cell invasion process, which they christened “entosis,” may be epithelial compaction gone awry, with the invader tugging so hard it pulls the other cell right around it. The malaria parasite performs a roughly similar trick when invading its host.
Once inside, most cells were degraded by lysosomes, but about 15% were released apparently unharmed. A few divided inside their hosts, and some hosts apparently turned into invaders themselves, giving rise to a cell within a cell within a cell.
Entosis is not just a novelty of the lab bench: 1–2% of metastatic breast tumor cells contained other cells. Whether entosis promotes tumorigenesis by increasing aneuploidy or inhibits it by killing invasive cells and whether noncancerous cells undergo entosis during development remain to be seen.