As recently as two decades ago, the scientific world was pretty sure that
you thought with the brain you had, not the brain you built.
Conventional wisdom held that brain cells were born only during fetal or
early postnatal development, but never once a human reached adulthood.
However, new research shows that neural stem cells not only exist in the
adult brain, but divide, differentiate, and find a way to settle right in with
the rest of the neurons in some parts of the brain, said Fred Gage, Ph.D., a
professor in the Laboratory of Genetics at the Salk Institute in La Jolla,
The old dogma gave two reasons why there could be no adult neurogenesis.
For one thing, neurons were so complex that it was thought they could not
divide. In addition, the brain was the seat of memory, the locus of who humans
are, so that if there were structural changes—aside from the familiar
degeneration of old age—how could humans maintain a sense of self?
In fact, both of these propositions have been proved untrue, thanks to the
research of Gage, his colleagues, and others in the field.
They found that while neurons do not divide, new ones do arise from neural
stem cells in two areas of the brain, the subventricular zone and dentate
gyrus of the hippocampus.
"Neurogenesis is not an event, it's a process," Gage told a
packed lecture hall at APA's 2007 annual meeting in San Diego in May."
How does a new neuron mature and integrate into a fully adult brain?
The process is different than in normal development."
That process begins when the neural stem cells divide. Seven days later,
they send out dendritic branches, and by 14 days they are fully bipolar, with
dendritic branches and an axon. After 28 days, they display spines on their
surface and have become electrophysiologically active.
By comparing images of cells born prenatally and postnatally, researchers
have found that there are no functional physiological differences between the
two sets of cells—although the adult-born neurons have a lower threshold
for firing and so are more easily excitable, said Gage.
At 14 days after the new cells are born, they have no connections with
their surroundings, but two days later, they have begun integrating with their
environment. By day 30, every dendrite makes thousands of connections with
neurons from the cortex into the gyrus. Gage suggested that glutamate spilling
out from "leaky" synapses may serve as an attractant to draw the
new cells into association with existing synapses.
GABA also plays an important role. The new neuron reads GABA as excitatory,
rather than inhibitory, as happens with mature neurons.
The new neurons don't just nestle in with their existing brethren, said
Gage. "They don't fill existing spaces, but outcompete existing cells
that are not making it." This competition results in the death of the
losing cell, new or old.
The process of developing and integrating new cells persists throughout
life, but it doesn't happen all the time. Genetics plays the biggest role, but
environmental circumstances exert a complex influence, too, said Gage. Acute
stress can suppress cell proliferation, although it restarts in days. Chronic
stress can suppress neurogenesis for months, but not permanently.
Physical activity also can affect neurogenesis. Mice allowed to exercise
produced two or three times as many new cells and did better on memory tests,
compared with sedentary mice. Although aging usually slows neurogenesis, older
mice given the chance to exercise also showed enhanced memory retention
compared with their peers.
"This shows that the mechanism for new cells still exists in old
brains," said Gage. "New stem cells are not gone; they are just
not activated. We see that environmental stimuli can help activate new cells
and integrate them into the circuitry of the brain."
Gage's lab has used the antiviral drug gancyclovir to dampen neurogenesis
in transgenic mice. Without neurogenesis, animals have difficulty remembering.
Once the drug is withdrawn, however, that ability returns in three weeks.
Implications of neurogenesis studies are widespread and may mark a path for
future research and eventually clinical applications.
For instance, there is an increase in neural proliferation after strokes,
and although it is accompanied by massive death of new cells, the process may
represent an attempt at compensation and a potential area for
Imaging studies have shown that new neurons don't die in people with
epilepsy, but their dendrites do appear to be misoriented, so that perhaps
aberrant integration of new cells into the circuitry is a cause of the
Finally, of special interest to psychiatrists, tests with the SSRI
fluoxetine indicate that blocking neurogenesis may also block the effects of
the drug. Thus, the development and integration of the new cells may be needed
for its action and may explain why fluoxetine takes so long to begin working,
said Gage. ▪