Genetic insights may explain retinal growth, eye cancer

Investigators at St. Jude Children’s Research Hospital have discovered the role of several key genes in the development of the retina, and in the process have taken a significant step toward understanding how to prevent or cure the potentially deadly eye cancer retinoblastoma. Retinoblastoma is the third most common cancer in infants after leukemia and neuroblastoma (nerve cancer). Retinoblastoma that has spread outside the eye is among the deadliest childhood cancers, with an average survival rate of less than 10 percent.

A key finding of the new study is that humans are more susceptible to developing retinoblastoma than mice, because mice can compensate for the loss of a gene critical to normal retinal development while humans cannot. The results of the study appear in the open-access journal BMC Biology.

“Our study gives us important new information on the normal development of the retina and suggests new studies that could lead to the design of more effective drugs to treat retinoblastoma,” said Michael Dyer, Ph.D., an associate member of the Department of Developmental Neurobiology at St. Jude and senior author of the paper.

The researchers discovered that during the development of the retina in mice, three genes that belong to the Rb gene family are expressed at different times. Specifically, the p107 gene is active before birth in cells that are going to become the retina. This gene ensures that the retinal cells stop multiplying at the proper time during development of this tissue. The Rb gene is expressed after birth in those cells that are actively multiplying as they also help form the retina.

In addition, the St. Jude team found that when Rb was inactivated during development of the mouse retina, the two p107 gene copies were up-regulated–made more active–therefore compensating for this loss of Rb activity. Importantly, this compensation required the presence of both p107 genes. In turn, when p107 was inactivated, Rb activity was upregulated; but unlike with p107, this compensation required only one copy of Rb.

The St. Jude team proposes that the ability of Rb and p107 to compensate for the absence of each other in mice prevents the developing retinal cells from multiplying uncontrollably and causing retinoblastoma. Also, the expression of both Rb and p130 might prevent this cancer in mice.

However, researchers learned that conditions in humans are not the same as in mice. They found that the primary Rb gene family member active in the developing human retina is RB1, and unlike in the mouse, little p107 is expressed in the developing human retina. In addition, p107 is not up-regulated to compensate for a loss of RB1 activity.

“This could explain why humans are susceptible to retinoblastoma following RB1 gene mutations, while mice require inactivation of both Rb and p107, or both Rb and p130,” said Dyer.

The discovery by the St. Jude team that p107 is not expressed during development of the retina in humans suggests that it might be possible in the future to prevent retinoblastoma by “turning on” that gene, Dyer noted.

“Because the eye is visible to researchers studying retinoblastoma, it’s possible to watch a tumor grow from a single cell,” said Stacy Donovan, Ph.D., a postdoctoral fellow in Dyer’s laboratory. “This could tell us which type of cell in the developing eye causes this cancer.”

“Knowing which cell causes retinoblastoma would give researchers a specific target for a novel retinoblastoma drug,” added Brett Schweers, Ph.D. a postdoctoral fellow in Dyer’s laboratory at St. Jude. “The biochemical pathway driving the multiplication of a cancer cell of origin would differ, depending on whether it was a progenitor cell or one of the more specialized cells. So it would be important to know which type of cell is giving rise to the tumor. That way you could design a drug to knock out the pathway driving the abnormal growth in that particular cell.”

Donovan and Schweers are the first and second authors respectively of the paper and contributed equally to the work.

Dyer’s team previously developed the first reliable mouse models of retinoblastoma that could be used to test new drug therapies for this tumor: http://www.stjude.org/media/0,2561,453_5485_11388,00.html.

Subsequently, the team used these models to demonstrate that a combination of topotecan and carboplatin were superior to the current treatment being used to treat retinoblastoma: http://www.stjude.org/developmental-neurobiology/0,2522,414_2041_19593,00.html.

Other authors of the paper include Rodrigo Martins of St. Jude and Dianna Johnson (University of Tennessee).