Probing the Mysteries of Neuron Growth and Rebirth


by Carl Sherman

December 18, 2009

Beneath the brain’s countless neurons and the synapses that connect them lies a maze of even greater complexity—the intricate interplay of neurochemicals that feed brain cells.

Now researchers are beginning to map out this landscape, from how neurons form to how they wire themselves into networks. As the molecular pathways that create and shape the brain come into sharper relief, they may reveal important insights into brain disease and injury and lead the way toward more effective ways to address such health problems.

The birth of brain cells

During fetal development, brain cells arise out of a closely coordinated pair of processes: Progenitor stem cells proliferate to provide raw material, but at a certain point they stop dividing and start to differentiate into neurons and support cells such as astrocytes.

Researchers at the University of North Carolina may have identified what they call “the master regulator” that flips the switch from proliferation to differentiation: the enzyme glycogen synthase kinase 3 (GSK-3).

“GSK-3 was originally studied in the context of glycogen synthesis [feeding energy to the body], but now people take a much broader view of its functions,” says William Snider, director of the Neuroscience Center at University of North Carolina and senior author of the paper reporting the research, which appeared in the November issue of Nature Neuroscience. “It was found right in the middle of a signaling pathway that regulates neural progenitors.”

When the researchers engineered mice not to produce the enzyme in the developing brain, the resulting animals had much bigger heads than normal mice—heads that were filled with progenitor cells but had few neurons.

Further experiments suggested that at the molecular level, GSK-3 normally keeps proliferation and differentiation in balance by braking proteins involved in the first and releasing those that promote the second. Removing GSK-3 produced a kind of arrested neurogenesis. “The progenitor cells stayed in the division phase,” Snider said.

Because neurogenesis does not end at birth, learning about GSK-3 may help explain the processes that underlie some psychiatric disorders, which could improve their treatment. For example, “we know that lithium, a drug widely used to treat bipolar disorder, inhibits GSK-3,” Snider says. “Bipolar disorder is increasingly diagnosed among children, and it should make us consider whether to be more cautious about giving lithium to them.”

 Other researchers have found a key role for GSK-3 in a neurochemical process that may lead to schizophrenia. The scientists first identified a gene, DISC-1, associated with schizophrenia in certain extended family groups and showed how silencing DISC-1 in rat brains impaired neurogenesis and produced schizophrenia and depression-like symptoms. But when the rats were injected with a GSK-3 inhibitor, neurogenesis—and behavior—returned to normal.

“We think deregulation of GSK-3 may contribute to behavior abnormalities,” says Li-Huei Tsai, director of the Picower Institute for Learning and Memory at the Massachusetts Institute of Technology and lead author of the research, published in Cell in March. “Our study suggests that it may be a good target for psychiatric drug development.”

The nervous system repairs itself

Another mystery surrounding the biochemistry of brain cells has to do with their self-healing capacities—or lack thereof. Many tissues in the body can regenerate after injury, but neurons lose this capacity around the time of birth and why “remains a major question in neuroscience,” says Jeffrey Goldberg, of the University of Miami. Researchers in Goldberg’s lab may have found an explanation in certain proteins that regulate the growth rate of axonal connections between brain cells.

Searching for possible culprits in this process, the researchers identified 111 genes whose expression increased sharply in the period between late gestation and a few days after birth—the time when neurons lose their ability to regenerate. When the team tested the genes in brain cell cultures, it found that one produced KLF-4, which reduced the growth of axons and dendrites by 50 percent.

KLF-4 is a transcription factor—a molecule that regulates the generation of mRNA (the genetic ‘blueprint’ used to build proteins) from DNA. The increase in KLF-4 around the time of birth seemed to stop cellular production of proteins that spur axons and dendrites to grow. When the researchers added an extra KLF-4 gene to retinal ganglion cells in culture, the result was far shorter axons than normal. When the gene for KLF-4 was disabled, axons grew longer.

In addition, when the researchers crushed the optic nerves of mice and examined them two weeks later, signs of regeneration appeared in the axons of mice that had been genetically modified to produce no KLF-4, but not in normal mice. The team reported its research in the Oct. 9 Science. Experiments since then have shown that chemically similar transcription factors also influence axon growth—some increase it, others decrease it the way KLF-4 does.

“The whole family of KLFs—there are 17 of them—appear to regulate axon growth,” Goldberg says. “If we manipulate multiple KLFs simultaneously, we may get bigger effects.” His team is now doing studies to explore this possibility, he says.

Much in the development process remains unknown, however, observes Peter Richardson, a neuroscientist at Barts and The London Centre for Neurosciences.

“What comes between the transcription factors and growth itself is still a black box,” He says. But research along this line could lay the foundation for a new approach to nerve cell regeneration that might ultimately revolutionize treatment not only of nerve injury, but also of neurodegenerative diseases such as Alzheimer’s and Parkinson’s.

How neurons connect

Another key question concerns the molecular process that forges synaptic connections between neurons. Several years ago, researchers found one piece of the puzzle in thrombospondins, proteins secreted by astrocytes that initiate the growth of excitatory synapses—the most common type of synapse in the brain.

Thrombospondin 1 and 2 are abundant during brain development, when most such synapses are formed, but largely absent from the adult brain. They reappear when the nervous system is damaged and new synapses are formed. But that process doesn’t always go smoothly. Abnormal synaptogenesis may be involved in the epilepsy and chronic neuropathic pain that sometimes follow nervous system injury.

Until recently, researchers had been unable to make much further progress on synapse growth because they couldn’t identify the thrombospondin receptor. A group at University of California at Irvine finally tracked it down using molecular biology techniques.

“We chopped the thrombospondin molecule into smaller units, and found one portion that was as active as the entire protein in inducing synapse formation,” says Cagla Eroglu, who led the research team. Then they looked for a molecule on the neuron surface with a section that fit the active part of thrombospondin.

As they reported in the Oct. 16 Cell, the match was the receptor α2δ-1, best known as a neuron’s calcium flow regulator. The researchers confirmed that α2δ-1 was the missing link by showing that reducing  its number through genetic manipulation resulted in fewer excitatory synapses being formed. When the researchers modified rats to express extra α2δ-1, more synapses appeared.

The discovery could explain why the drugs gabapentin and pregabalin work to treat some nervous system problems, including epilepsy and neuropathic pain. α2δ-1 is the receptor for gabapentin and pregablin; this research suggests that the drugs work, at least in part, by blocking the formation of new excitatory synapses.

“It’s a totally new, totally different way of looking at what gabapentin does,” says Allan Basbaum, who studies the neurobiology of pain at University of California, San Francisco, and is a member of the Dana Alliance for Brain Initiatives. “It was always assumed that the drug’s pain-relieving properties were [only] related to its effect on calcium flow.”

Learning more about underlying molecular events could revise our understanding of chronic pain “not just as a symptom, but as an independent disease of the nervous system,” Basbaum says. “Instead of going after the α2δ-1 receptor, maybe we should be going after synaptogenesis itself.”

 The interaction between thrombospondin and its receptor is only an initial event in synapse creation—the first step, perhaps, on a winding pathway. “We want to know what happens afterwards… what other molecules are involved,” Eroglu says. “With α2δ-1 we caught a clue, and if we continue pulling the thread we may find out what’s on the other end.”