Study challenges conventional wisdom about cancer cells

May 27 2007

Some scientific discoveries make their biggest impact by disproving "conventional wisdom."

A team of St. Jude investigators challenged conventional wisdom about the eye cancer retinoblastoma by using a mouse model that allowed them to study the tumors as they develop and grow.

Retinoblastoma is a tumor that forms in the developing retina––the layer of nerves at the back of the eye that responds to light. The finding could one day lead to effective new drugs to treat this cancer and fundamentally change the way we understand all brain tumors, according to Michael Dyer, PhD, Developmental Neurobiology. Dyer is senior author of a report on these findings that appeared in the March 15 issue of Cancer Research.

The current thinking states that dividing cells cannot be differentiated (specialized in form and function); instead, only undifferentiated, immature cells divide. In other words, if a cell has matured and taken on a specific identity—like a neuron—that cell doesn't multiply. Only the undifferentiated (immature) progenitor or stem cell versions of that cell do.

But Dyer's team showed that during the early stage of retinoblastoma, the tumor cells are quite differentiated.

The investigators discovered that at the early stage of retinoblastoma the tumor cells look and act like nerves. The cells make the proteins that these nerves ordinarily make and extend tiny projections, called neurites, which form synapses with other nerves. A synapse is the point at which the ending of one nerve communicates with another nerve.

"Most cancer cells are much too busy multiplying to be differentiated," Dyer said. "So seeing synapses in retinoblastomas was amazing. But later these tumor cells dedifferentiate––become undifferentiated––as they invade the surrounding tissue. This raises the possibility that dedifferentiation may contribute to metastasis [spreading of cancer]. Approximately half of the children with retinoblastoma in developing countries still die of metastatic disease, so this is a very important area of research in my laboratory."

The team also disproved another long-held notion about retinoblastoma cells. Until now, researchers thought that structures called rosettes were abnormal photoreceptors—cells that capture incoming light and turn it into electric signals. The St. Jude team showed that rosettes are simply extensive clumps of retinoblastoma cells that stick tightly to each other.

"These tight links are lost as the tumor cells invade the surrounding tissue," Dyer said. "We are now working to determine exactly how this occurs in metastatic retinoblastoma."

The new findings suggest that novel drugs that prevent the maturing nerves in the plexiform layers (two layers of nerves in the retina) from lapsing back into the immature state might prevent retinoblastoma, according to Carlos Rodriguez-Galindo, MD, Oncology.

"The work also demonstrates the great value of the mouse models of retinoblastoma that Dyer's lab developed," said Matthew Wilson, MD, Surgery and Pathology.

"These models allow us to study retinoblastoma at its very earliest stages, which is critical to figuring out how and why this tumor develops and how we might be able to prevent that from happening," Dyer said.

The other St. Jude author of this paper is Jiakun Zhang, MD, PhD, Developmental Neurobiology.