Experimental compound offers fragile X treatment

Fragile X syndrome (FXS) is the most common inherited form of autism and intellectual disability, characterized by hyperactivity, repetitive behaviors and seizures. Although there are a number of medications that can help individuals to cope with the behavioral and mental health issues associated with the syndrome, as yet there are no proven therapies that directly impact the root cause of the disease. However, this could now be about to change with the RIKEN-led discovery of a compound that not only treats the symptoms in adult mice but also reverses the morphological defects associated with the syndrome¹.

The international research team, led by Susumu Tonegawa of the RIKEN–MIT Center for Neural Circuit Genetics (CNCG) at the Massachusetts Institute of Technology (MIT) in the United States, recently demonstrated the remarkable effectiveness of their promising compound—called FRAX486—in model FXS mice. “The effect was dramatic,” says Tonegawa, a Nobel laureate and also director of the RIKEN Brain Science Institute. “With only one dose, the mouse was cured.”

The development of FRAX486 began six years ago when Tonegawa and his colleagues were studying mice engineered to have low activity of an enzyme known as p21-activated kinase (PAK). Unexpectedly, they noticed that the animals had relatively few dendritic spines in their brains, and the spines that were there tended to be short and stubby. Dendritic spines are small neuronal protrusions that assist with neuron-to-neuron communication in the synapse. The researchers recognized that the absence of these spines in PAK-mutant mice was the exact opposite of what is seen in people with FXS and in mouse models of the disease. In FXS mice, which lack a working copy of the gene that encodes the fragile X mental retardation protein (FMRP), researchers typically see a profusion of long and thin dendritic spines in the brain.

The observation that mice lacking PAK and FMRP display mirror-opposite phenotypes led Tonegawa’s team to test an intriguing hypothesis: perhaps PAK and FMRP antagonize each other to regulate spine morphology. Inhibiting PAK activity might therefore correct some of the defects associated with FXS. To test this idea, the researchers engineered a mouse to express neither PAK nor FMRP. To their surprise and delight, these mice had normal numbers of dendritic spines, unlike those only missing FMRP. Mice lacking both proteins were also less anxious and displayed fewer repetitive behaviors than FXS model animals. “It looked like the mouse was cured,” says Tonegawa. “We never thought it would come out that nicely,” he adds.

In 2007, soon after the team’s first promising results were published, Tonegawa was contacted by Jay Lichter of Avalon Ventures—a US venture capital firm. Lichter was interested in working with Tonegawa to develop a drug that could chemically inhibit PAK. They hoped that such an agent could yield the same therapeutic benefit as the genetic manipulation reported by Tonegawa’s research team but be applicable as a treatment in humans. Later that year, Tonegawa and Lichter founded the biotechnology company Afraxis to further develop the PAK inhibitor concept.

Scientists at Afraxis embarked on a search for a potent inhibitor of PAK by screening a library of 12,000 small molecules known to block kinase enzymes. Their search yielded one promising hit: FRAX486, a compound that selectively inhibits its target and also readily crosses the blood–brain barrier, staying active in the brains of mice for many hours. The researchers found that a daily dose of FRAX486 achieved a steady state of active drug levels in the brain.

Next, Tonegawa’s research group undertook further characterization of the FRAX486 compound. This required new tests to be developed to quantify the effectiveness of the drug. Correspondingly, Bridget Dolan, a member of Tonegawa’s laboratory at the CNCG, created an assay that involves the measurement of sound-induced seizures in FXS mice. She found that most mice injected with higher doses of FRAX486 were immune to seizures 8 hours after treatment, whereas all untreated FXS mice suffered seizures, with around half subsequently dying from respiratory arrest.