Hit man: A suspect emerges in the chaos of aggressive brain cancer

New research from UAB oncologist Markus Bredel identifies the splicing enzyme PTBP1 as a key factor
in the spread of glioblastoma multiforme.  

Glioblastoma multiforme is one of the deadliest human cancers. "The tumor can double in size within a few weeks," says Markus Bredel, M.D., Ph.D., a professor in the UAB Department of Radiation Oncology and senior scientist in the neuro-oncology program at the UAB Comprehensive Cancer Center. "Usually, by the time we see a patient, they often have apple-size lesions."

That explosive growth "comes with a substantial amount of genetic chaos," Bredel says. "If you look at the whole genome in a brain tumor, out of the 30,000 genes, you very often have changes in up to 50 percent; they're up or down, lost, amplified, mutated."

A Change for the Worse

Markus Bredel

But in that chaos, patterns emerge with surprising regularity, Bredel says. "When Gene A is up, Gene B is very often down." In two papers published in JAMA in 2009, Bredel's research team argued that "there needs to be a reason why glioblastomas co-select for certain genetic events. The tumor cells must benefit."

In those papers, Bredel's lab identified dozens of gene-gene links that were candidates for additional scrutiny. They focused on one particular pair: The oncogene EGFR, or epidermal growth factor receptor, which is crucial for normal cell growth and wound healing, and the tumor-suppressor ANXA7 or annexin A7. EGFR is of interest in many cancers, because it is often hijacked to fuel the aggressive growth of tumor cells.

"We found that ANXA7 is probably a regulator of EGFR," Bredel says. "So it's to the benefit of the tumor cell to knock down this regulator." But it wasn't clear at the time how this was happening. "ANXA7 resides on a different chromosome from EGFR, so it's a completely independent event, but somehow the tumor cells were disabling it," says Bredel.

Now, in a paper published May 27 in the Journal of Clinical Investigation, Bredel's lab has revealed how ANXA7 normally keeps EGFR in check—and how cancer cells manage to sabotage this system. Those discoveries have also identified a promising new target for treating glioblastoma, a cancer with few therapeutic options.

Taking Out the Trash

Normal cells have ways of dealing with proteins that get too big for their britches. Cellular structures called endosomes degrade the proteins, acting as the "trash cans" of the cell, Bredel explains. "What we found is that ANXA7 promotes the sorting of EGFR into those trash cans."

Here's one way to think about the relationship, Bredel says: "EGFR is kind of the bad guy in the cells. When it's present, it promotes the tumor process. ANXA7 is the police, which under usual conditions constrains the bad guy. But in the absence of the police force, the bad guy can do whatever he wants."

But what is taking out the police force in glioblastoma? Bredel's team started with an observation: A "long form" of the ANXA7 gene exists in normal brain cells, but an altered version appears in glioblastomas.

Genes contain the code that tells the cell's factories how to make their specialized product—usually a protein. The parts of the gene that actually contain instructions are known as exons. Each of the exons in a gene codes for the amino-acid "building blocks" that make up each protein. (In between are non-coding sections: the introns.) In glioblastoma, Bredel found, exon 6 was missing from the ANXA7 gene. "Without exon 6, ANXA7 can't sort EGFR to the cell's trash cans," says Bredel. The "bad guy" has free rein.

Follow the Slices

Clearly, something is snipping exon 6 out of the picture. But that isn't necessarily abnormal. "In the past 15 years, we've realized that gene splicing plays a role in many biological processes, both normal ones and disease processes," Bredel says. Splicing is a way to increase efficiency; it allows the same gene to produce different proteins, depending on which of the underlying amino-acid parts are used.

To turn a gene into a specific isoform of a protein, the cell's copying mechanisms cut out the exons and stitch them together to form an uninterrupted message. The cutting is the job of splicing factors, and Bredel's attention focused on one: PTBP1.

Exon 6 is what is known as a "cassette exon," or "alternative exon," a section of the code that appears in that gene in some body tissues but not others. "A cassette exon might be present in the brain but not in the muscle tissue, for instance," Bredel says. In his team's latest paper, "we figured out that the PTBP1 gene is the splice factor that cuts cassette exon 6 out of ANXA7," he continues. They also established that PTBP1 proteins are overproduced in glioblastoma compared to normal brain tissue.

So PTBP1 turns bad in cancer, guts a key exon from ANXA7, and allows EGFR to replicate like crazy. Well... not quite, says Bredel. It's even more interesting than that. "We initially thought that this splicing was something specific to tumor cells, that it might even be an initiating event that allows the tumors to emerge," he says. But when the researchers looked at a set of normal brain cells called precursor cells or stem cells, they found this same ANXA7 splicing going on. "It wasn't present in mature neurons, but in the immature cells, the stem cells, there it was," says Bredel.

Deadly Inheritance

When you think about it, that makes sense. Neural and glial stem cells power initial brain development and, as is becoming increasingly clear, allow us to learn new things over the course of our lives. They also respond to injury and disease, such as strokes and Parkinson's disease. Having a way to turn off a growth suppressor like ANXA7 is "a useful trait in the stem cell, because the stem cells wants to be able to divide and grow," Bredel says.

The UAB scientists now believe that the ANXA7 splicing in glioblastoma "is something the tumor cells inherited from the stem cells, a potential tumor-initiating ancestor of glioblastoma," Bredel says. "When that stem cell, through accumulation of mutations, develops into a tumor cell, that splicing trait is still there. And then it's exploited further by the accumulation of mutations that enhance EGFR signaling." At this point, Bredel says, "I'm not sure if we can claim this process is involved in the initiation of glioblastoma, but it certainly is involved in the progression of glioblastoma."

PTPB1's role in healthy brain stem cells means eliminating it completely isn't an option. But targeting PTBP1 with the aim of lowering production to normal levels offers exciting treatment possibilities. Bredel's lab is now identifying promising compounds that could act on PTBP1. Restoring tumor suppressors such as ANXA7 directly in cancer hasn't been successful so far, Bredel says. "Having something that is operating in excess that we can target, like PTBP1, is much easier," he notes.

"We haven't been able to make any major, clinically meaningful progress in glioblastoma in the past 20 years," Bredel adds. "We are still a long way off from being able to take this to a clinical trial in patients, but this is an exciting discovery."