How cancer cells escape from tumors, spread

New research from a team led by Northeastern's Anand Asthagiri, associate professor of bioengineering and chemical engineering, helps to answer those questions. It provides an astonishing look at the biophysical properties that permit breast cancer cells to "slide" by obsta­les and travel out of their primary tumor toward a blood vessel that will carry them to a new site.

The paper, published in Biophysical Journal, reveals how the abnormal proteinfiber scaffolding of tumors and the agility of the cancer cells themselves come together in a perfect storm to enable the escape. The quantitative method the researchers developed to understand the cells' sliding ability could also lead to a new way to screen for effective cancer drugs and help diag­nose the stage of a cancer early on.

"We are looking at the interaction between cancer cells' migrating and this sliding phenomenon, and how that's influenced by the proteinfiber environments of tumors," says Astha­giri. "In this paper we show that cancer cells migrating on these protein fibers have a unique ability that enhances their invasion capacity: When they bump into other cells -- which the microen­vi­ron­ment is packed with -- they slide around them. Normal cells halt and reverse direction."

An interdisciplinary approach

The researchers' engineering backgrounds shaped their interdisciplinary approach: They set out to explore the mechanics of the sliding ability as well as its molecular components.

To do so they devel­oped a model environment that mimics pro­tein fibers. First they stamped stripes of a protein called fibronectin on glass plates, making sure to represent var­ious widths. "If you treat a fiber as a cylinder, imagine cutting it and opening it up and laying it flat," says Asthagiri. "That's essentially what these long stripes of pro­tein mim­icked." Then they deposited the cells -- alternately hun­dreds of breast cancer cells and hun­dreds of normal cells -- on these fiber­like stripes and used a microscope with timelapse capabil­ties to observe and quantify their behavior.

On fibers that were 6 or 9 microns wide -- the typ­ical size of fibers in tumors -- half the breast cancer cells elongated and slid around the cells they col­lided with. Con­versely, 99 percent of the normal breast cells did an about face.

But why? To under­stand what gave the cancer cells this remarkable agility, Asthagiri and his colleagues, who included Daniel F. Milano, a former grad­uate research assistant at North­eastern, introduced "genetic perturbations" into the mix -- that is, they inserted certain pro­teins into the cancer cells and took the same proteins out of the normal cells. Among them was E-??cadherin, a sticky protein that enables cells to bind to one another.

"Cancer cells often lack E-cadherin," says Asthagiri. "When we introduced it genetically, the cancer cells' ability to slide diminished. And when we took E-cadherin out of normal cells, they acquired some sliding ability once the fibers were wide enough." Together, the varying widths of the fiber paths and the perturbations pro­duced a wealth of quantitative data about how the cells, both cancerous and normal, behaved under different conditions.

"We weren't just showing that cells either slide or don't slide," says Astha­giri. "We were showing that there are dif­ferent levels of sliding ability, and we mea­sured each one."

Multiple applications

Asthagiri's system is relatively easy to con­struct and suited for rapid imaging -- two qualities that make it an excel­lent candidate for screening new cancer drugs. Phar­ma­ceu­tical companies could input the drugs along with the cancer cells and mea­sure how effectively they inhibit sliding.

In the future, the system could also alert cancer patients and clin­i­cians before metas­tasis starts. Studies with patients have shown that the struc­ture of a tumor's protein-??fiber scaf­folding can indi­cate how far the disease has progressed. The researchers found that certain aggressive genetic mutations enabled cells to slide on very narrow fibers, whereas cells with milder muta­tions would slide only when the fibers got much wider. Clin­i­cians could biopsy the tumor and mea­sure the width of the fibers to see if that danger point were approaching. "We can start to say, 'If these fibers are approaching X microns wide, it's urgent that we hit cer­tain path­ways with drugs," says Asthagiri.

Questions, of course, remain. Do other types of cancer cells also have the ability to slide? What addi­tional genes play a role?

Next steps, says Astha­giri, include expanding their fiber­like stripes into three-dimensional models that more closely represent the fibers in actual tumors, and testing cancer and normal cells together. "There are so many types of cells in a tumor environment -- immune cells, blood cells, and so on," he says. "We want to better emulate what's hap­pening in the body rather than in isolated cells interacting on a platform."