A ‘tool to understand how tumors work’
Our genes encode the proteins that make our cells and biological functions run. Mutations to a gene — changes to the letters that make up our DNA code — can change the function of the protein it encodes. Sometimes mutations hobble a protein and prevent it from doing its job; sometimes they give it new activities or allow it to function when it shouldn’t. Depending on what the protein usually does, any of these changes could contribute to cancer.
Most studies of the links between mutations and cancer focus on point mutations, single-letter typos in the gene’s DNA sequence. Supported by his new grant, Holland will examine how two other types of mutations that affect how genes get made into proteins could contribute to cancer: gene fusions and changes in gene splicing.
Gene fusions occur when two different genes get stuck together, producing a mashup protein. Splicing is a normal process by which a single gene can encode several similar but distinct proteins; changing which version of a protein is made — and when — can have enormous biological consequences.
A protein produced by a gene fusion can often perform the roles of its component proteins as well as new roles. Holland’s team will focus on fusions that involve the gene for Yap, a protein that, if not properly regulated, can drive the excess cell growth and division needed for tumor growth. YAP gene fusions turn up in many different solid tumor types, including brain cancers.
Using a powerful genetic system that Holland developed to test how mutations promote cancer, his group has developed mouse models of YAP-propelled tumors to study how YAP gene fusions cause tumors to form. They will also use the models to identify and test new treatment strategies. But Holland’s interest in YAP gene fusions extends beyond particular gene pairs or even the specific tumors in which they’re found.
“It turns out that sometimes if you have a gene fusion — even if it's rare — if it’s sufficient to cause cancer, then you learn an awful lot about cancers,” Holland said. “Even if other cancers don’t have that gene fusion, whatever that gene is doing is being achieved by other means. It’s a tool to understand how tumors work.”
Holland has similar hopes for his lab’s studies of splicing.
The gene with various splicing options that Holland and his group study encodes the protein TrkB. Normally, cells choose different splicing options for TrkB depending on their age. Cells in the developing embryo choose one option, but switch to a different splice variant once past the embryonic phase. The embryonic TrkB splice variant is the dominant form of TrkB found in all tumor types, both adult and pediatric.
It’s no coincidence.
“Early development is a highly proliferative, very much cancer-like state,” Holland said.
In fact, tumors often turn on genes or even gene programs that are only supposed to run during our earliest days. These early-development genes can help tumors develop, grow and metastasize, or spread, to other areas of the body. Dr. Siobhan Pattwell, a postdoctoral fellow in Holland’s lab who recently moved to Seattle Children’s Research Institute to run her own research group, showed that the youthful TrkB splice variant is powerful enough to cause many different types of tumors in mice.
“So the question really is, how central is [altered splicing] to all cancer?” Holland said.
It may be that other genetic changes that cause cancer also work by meddling with splicing. If so, the lessons that Holland and Pattwell learn from studying TrkB could help researchers develop new strategies to treat a wide range of tumors.