Scientists have discovered a molecule that shrinks brain tumors in mice, raising hopes that the compound could one day become a pill to treat glioblastoma, one of the world’s most pernicious cancers.
The molecule, known only as compound A, attacked the animals’ glioblastoma by hitting a target previously thought to be difficult, if not impossible, to successfully drug. The researchers, led by Hui Li, Ph.D., of the University of Virginia, are now working to tweak compound A into a form fit for human trials—and have already crafted 160 new variations, Li told Fierce Biotech.
The results were published in Science Translational Medicine on Jan. 27.
Glioblastoma is an aggressive adult brain cancer with an extremely poor prognosis—most patients succumb within a year to 18 months, and five-year survival rates are only 5%. Most patients are treated with some combination of surgery, radiation and a chemotherapy drug called temozolomide. However, many glioblastomas develop resistance to temozolomide, Li said, and even a successful temozolomide and radiation regimen only extends life by about 10 weeks.
Li’s team first identified that a protein called advillin, or AVIL, was superabundant in glioblastoma in 2020. After revealing the protein’s role in the devastating disease, Li began fielding calls and emails from glioblastoma patients and family members who were desperate for a new treatment, he said.
“They would describe how horrible it is, how it's affecting their life or their loved ones,” Li said. “I got depressed. I had to say ‘no’ to so many people.”
That awful feeling spurred Li to try to turn his discovery into a new medicine, a daunting task for numerous reasons. First, any brain cancer drug has to effectively cross the blood-brain barrier to reach the tumor, and, even once there, the drug shouldn’t interfere with all the important work done by the brain. Any new drug would also hopefully not be easy for the cancer to evolve resistance to, the same way it does with temozolomide.
Li teamed up with Angela Koehler, Ph.D., who runs a cancer drug discovery lab at the Massachusetts Institute of Technology, to screen a library of 50,000 compounds for their ability to bind and inhibit advillin. This was a relatively small library to use for such a big problem, Li admits, because advillin is not a typical drug target. For starters, there’s no published structure of the protein available; Li said he tried to use artificial intelligence program AlphaFold to crack advillin’s structure, but with little success.
Then there’s the matter of advillin’s job. Advillin’s role in cells is to bind to and regulate actin, a filamentous protein that forms the cell’s skeleton. Actin-binding proteins tend to not have pockets or active sites that can serve as landing spots for a drug looking to disrupt them, the researchers noted in the paper.
But Li and colleagues got lucky. Their screen turned up compound A, and further assays and lab experiments in both healthy and cancerous cells revealed that the molecule can target advillin.
The result “challenges an assumption that cytoskeletal regulators are not tractable drug targets,” Mustafa Khasraw, M.D., a neuro-oncologist at the Duke University School of Medicine who was not involved with the study, told Fierce. “This mechanistic validation is notable in a field where few first-in-class, non-kinase oncogene inhibitors have advanced.” An oncogene is any that, like the gene that makes advillin, can cause cancer if mutated or overexpressed.
The researchers then ran compound A through a gauntlet of glioblastoma mouse models and found the molecule met the mark at reducing the size and volume of brain tumors, even when the cancer was resistant to temozolomide, with no signs of toxicity. Better still, glioblastoma stem cells, which drive the cancer’s ability to resist temozolomide, were even more susceptible to the compound, Li said, which means tumors may have trouble evading the advillin inhibitor’s effects.
However, more work is needed to ensure that the cancer can’t evolve resistance to compound A through other means, Khasraw said.
Compound A’s success at killing multiple glioblastoma cell lines and treating multiple mouse models while also able to cross the blood-brain barrier and to be given by mouth is notable, Khasraw said. “It is uncommon for a single experimental agent to meet so many translational benchmarks at the preclinical stage.”
That said, both Khasraw and Li caution that compound A is far from a cure for glioblastoma. A startup Li founded, aptly named Avil Therapeutics, is now working to optimize a version of the molecule as a pill for human trials, which Li hopes will begin in one or two years.
Li estimates getting a drug candidate to the clinic will cost about $3 million to $5 million, he said, but he hasn’t really thought about fundraising much yet—he’s been too busy trying to get the proof-of-concept paper published. He already has some outside enthusiasm in his work, and he expects the paper will generate even more.
“We'll be looking for some fundraising or some partnership,” Li said. “I'm open to different options.”
In his academic lab, Li’s team is also working to identify other cancers characterized by advillin activity. They’ve already found the villainous protein is involved in rhabdomyosarcoma, a soft tissue cancer that mainly affects children, and Li said it seems to play a part in gastric cancer, too.
Advillin may end up becoming a fruitful target across a suite of cancers, because it’s only expressed at very low levels in certain healthy cells—meaning disrupting it shouldn’t cause many side effects. Cells limit how much advillin they make because of the protein’s power over the cell’s actin skeleton, Li said.
“It can bind to actin, it can cap actin, it can nucleate actin,” Li listed. “It can polymerize. It can depolymerize. It can sever, it can bundle, it can branch.”
But with advillin so important for glioblastoma and so sparsely used by healthy cells, the protein’s power may soon become the cancer’s liability.