The second century C.E. Greek physician and philosopher Galen advised patients suffering from disorders of the spirit to bathe in and drink hot spring water. Modern day brain scientists have posited that Galen’s prescription delivered more than a placebo effect. Lithium has for decades been recognized as an effective mood stabilizer in bipolar disease, and lithium salts may have been present in the springs Galen knew.

Yet exactly how lithium soothes the mind has been less than clear. Now, a team led by Ben Cheyette, a neuroscientist at the University of California in San Francisco (UCSF), has linked its success to influence over dendritic spines, tiny projections where excitatory neurons form connections, or synapses, with other nerve cells. Lithium treatment restored healthy numbers of dendritic spines in mice engineered to carry a genetic mutation that is more common in people with autism, schizophrenia, and bipolar disorder than in unaffected people, they report today in Molecular Psychiatry. The lithium also reversed symptoms in these mutant mice—lack of interest in social interactions, decreased motivation, and increased anxiety—that mimic those in the human diseases.

“They showed there’s a correlation between the ability of lithium to reverse not only the behavioral abnormalities in the mice, but also the [dendritic] spine abnormalities,” says Scott Soderling, a neuroscientist at Duke University in Durham, North Carolina, who studies how dysfunctions in signaling at brain synapses lead to psychiatric disorders. Soderling adds that the work also sheds light on the roots of these diseases. “It gives further credence to this idea that these spine abnormalities are functionally linked to the behavioral disorders.”

Over the past 2 decades, neuroscientists have built a body of evidence that links not only bipolar disease, but other psychiatric disorders including autism and schizophrenia to abnormal brain development. In particular, they have found abnormalities in the numbers of synapses and in the shape of neurons at the points where they form synapses. Their studies have often implicated abnormal signaling in a brain pathway called Wnt, which is involved both in early brain development and later, more complex, refining of brain connections. The role of Wnt could help explain why lithium is effective: It blocks an enzyme called GSK-3 β, which is an inhibitor on the Wnt pathway. By boosting Wnt signaling, lithium could produce a therapeutic effect in psychiatric diseases in which the Wnt pathway is underpowered.

To find out how that biochemical effect influences the brain, Cheyette and a team of colleagues at UCSF and other institutions worked on several parallel tracks. They compared the genomes of 9554 people with bipolar disease, autism, or schizophrenia with those of 11,361 unaffected individuals, looking for variations in the genetic sequence of one form of DIXDCI, another key gene on the Wnt pathway. The changes were rare in both groups, but they were about 80% more common—0.9% versus 0.5%—in people with the disorders compared with unaffected people, suggesting that the mutations elevate the risk of developing the disorders.

The researchers also created mice with a mutated version of DIXDCI. They subjected the rodents to tests gauging their willingness to interact with other mice, to explore a new environment, and to swim—as opposed to passively float—when placed in water. The mutant mice showed decreased sociability, increased anxiety, and lack of motivation compared with controls—all symptoms with analogs in human psychiatric disease. The scientists imaged the rodents’ living brains and examined their neurons in lab dishes. All the tests showed decreased numbers of dendritic spines. At the same time, biochemical tests showed that Wnt signaling was impaired in the mutant mice.

They then treated the mutant mice with lithium. Although the researchers acknowledge that rodents are an imperfect proxy for human mood disorders, they did observe that the animals’ symptoms markedly improved; studies of their brains also revealed normal numbers of spines. “That’s the key finding,” Cheyette says. “It suggests that lithium could have its well-known therapeutic effect on patients with bipolar disorder by changing the stability of spines in the brain.”