Columbia University Medical Center

Tracing Cancer Back to Its Origins

bladder urothelium

Using different colored “dyes,” researchers are tracing cancers back to their original cells. Bladder image: Cathy Mendelsohn.

Some bladder cancers look like tiny fingers. Others look like flat bed sheets.

The differences are obvious and they have clinical implications. The “fingers” of papillary tumors often grow back after surgery, but flat carcinoma in situ cancers are typically more aggressive and more likely to spread.

“The big question is why,” says Cathy Mendelsohn, PhD, a professor of urology, pathology & cell biology, and genetics & development, who has been studying bladder development and regeneration at Columbia. “Because if we understand how cancers arise, and why some lesions invade and others don’t, we may be able to develop more precise techniques for diagnosis and better therapies for treatment.”

In the last decade, growing evidence suggests that the type of cell in which the cancer originates also shapes the cancer’s susceptibility to treatment.

Most research into tumor diversity delves into the mutations the cancer acquires as it evolves from a single cell to a complex tissue. And indeed, cancers often behave differently depending on the types of mutations they’ve acquired.

But in the last decade, growing evidence suggests that the type of cell in which the cancer originates also shapes the cancer’s traits and susceptibility to treatment.

“If this is true, you can see why identifying the cell of origin might have clinical relevance,” says Michael Shen, PhD, professor of medicine, genetics & development, and urology, who has been searching for the starter cells of prostate cancer.

But identifying each cancer’s “cell of origin” has been tricky. For one thing, says Shen, “looks can be deceiving.” For example, for years it was assumed that “basal-type” breast cancers spring from the breast’s basal cells because most of the cancer’s cells resembled—on a molecular level—normal basal cells. Yet recent work shows that basal-like breast cancers actually originate in a completely different source, luminal cells.

Instead of relying on what cells look like, Mendelsohn and Shen are exploiting newer technology— fate-mapping—that allows them to label specific types of cells with an indelible marker that is passed on to each one of the cells’ progeny.

By marking cells early in mouse development and then waiting for cancer to develop, the researchers simply looked inside the cancer to find the label and identify the original cell type.

In the bladder, this method revealed not one (as originally thought), but two types of cells that give rise to cancer. The bottom-most cells of the bladder’s inner lining—K5 basal cells—give rise to flat carcinoma in situ tumor, considered to be precursors of invasive bladder cancers, while cells in the middle of the lining—called intermediate cells—give rise to the protruding papillary cancers that are considered less invasive than carcinoma in situ.

“These findings help explain why these lesions look and behave differently,” says Mendelsohn “Because we are using mouse models, it will now be possible to pull labeled cells out of these lesions before they become invasive, to see which mutations drive aggressive behavior.”

Organoids made from the cells that give rise to prostate cancer may speed up drug discover. Image: Michael Shen, Nature Cell Biology 16: 951.

Organoids made from the cells that give rise to prostate cancer may speed up drug discovery. Image: Michael Shen, Nature Cell Biology (2014) 16: 951.

Researchers can now also use the cells to make organoids, tiny 3-dimensional cultures of cells that mimic the properties of tumors. “With organoids that behave like tumors, we may be able to speed up drug screening tremendously,” says Shen.

After using fate-mapping to identify the cells that give rise to most prostate cancers, Shen has recently created organoids that respond to chemotherapy in the same way as prostate cancers do in some human patients.

“By reducing prostate cancer to a dish, we can now do experiments that take only weeks, not years, to finish.”