Cut off from their usual social group, the three macaque monkeys fell into simian depression. They no longer took pleasure in anything. They lost status and did not seem to care.
Columbia University researchers gave three other exiled monkeys the antidepressant Prozac, and they showed no signs of depression. Later examination showed that in a key area deep in the monkeys' brains - the seahorse-shaped hippocampus - myriad new cells had sprouted.
Then the scientists treated four more monkeys with X-ray radiation that blocked the hippocampus from making new cells. When those monkeys were sent into depressing exile, Prozac couldn't help them. And their brains later showed no signs of new cells in the hippocampus.
That preliminary study, presented earlier this month at the annual Society for Neuroscience conference, adds the latest scientific backing to a hot theory of depression that has been gaining momentum - and drawing debate - for several years.
It goes like this: Depression, which affects at least 19 million Americans a year, can involve problems not only with chemical messengers such as serotonin, but with the very structure of the brain, with the neurons and their connections.
Research suggests that stress and depression can actually shrink parts of the brain and that anything that successfully lifts depression - be it exercise, drugs, or shock therapy - appears to involve a burst of new neurons in key areas.
The catchword for the hot theory is "neurogenesis" - meaning the creation of new neurons - and even enthusiasts hasten to caution that it is still only a theory. But it is compelling enough that some drug companies are already focusing on neurogenesis and various substances that boost it in hopes of generating better antidepressants.
The theory could explain, for example, why drugs like Prozac often take weeks to kick in, even though they change the brain's chemistry within hours: It could take that long for the new neurons to develop. The theory also fits with findings that the longer a person has been depressed, the smaller their hippocampus. And with findings that depression can impair memory and learning, which depend in part on forming new neurons.
Depression and other mood disorders could actually act in the brain something like milder versions of neurodegenerative diseases - Alzheimer's or Parkinson's - with their atrophy and cell loss, said Dr. Ron Duman, a professor of psychiatry and pharmacology at Yale University. But with one key difference, he said: Neurodegenerative diseases are not normally reversible. Mood disorders usually are.
The theory "tends to tie in everything," said Dr. Tarique Perera, the psychiatrist who presented the macaque study. "It's the one theory that can encompass the neurotransmitters, the stress hormones, the structure problems and even certain behavioral aspects of depression."
And recent research raises an exciting new possibility, he said: boosting neurogenesis may not only lift depression but actually prevent it.
The neurogenesis theory leans on surprising findings from the late 1990s: Contrary to longstanding dogma that adult brains are fully developed, scientists discovered that the human hippocampus, a center of learning and memory, continues to make new neurons throughout life.
Neurogenesis pioneer Dr. Fred Gage, of the Salk Institute, cautioned against oversimplifying the link with depression. At this point, it is not clear that the link between neurogenesis and depression is actually one of cause and effect.
Evidence from various avenues of research suggests that increasing neurogenesis helps alleviate depression, and reduced neurogenesis could exacerbate it, he said.
How could that be? How might the birth of new neurons affect mood?
That question tends to bring a lot of hand-waving from scientists, and possible answers remain sketchy. But the known link between neurogenesis and learning has led some to speculate that in a depressed state, a person becomes overfocused on the bad and unable to register the good, and in effect, that inability to shift gears is a failure to learn.
Researchers also note that though the hippocampus is seen as a hub of memory, it also appears to influence anxiety and has connections to centers of emotion in the deeper limbic brain, and could affect their activity.
Some people suffer from depression so intractable that no treatment works. According to the neurogenesis theory, part of the problem may be that because of chemical or structural alterations in their brains, they can no longer produce enough new neurons. But research has yet to address that possibility.
The neurogenesis theory also has troubling holes.
At this month's giant Society for Neuroscience conference in San Diego, one challenge to the theory came from Shawn Kohler, a researcher at the University of Illinois at Urbana-Champaign.
In mice, it takes three or four weeks for new neurons to mature - about as long as it takes for people with depression to feel antidepressants kick in. Many researchers had assumed that the time scale must be the same in humans, but Kohler found that in monkeys, new neurons took a good 24 weeks to mature.
"We would only expect it to be longer in humans," he said.
Some studies, like Perera's, have shown that blocking neurogenesis also blocks the buoying action of antidepressants. "But other papers have come out recently showing that's not always the case," said Duman.
What is clear is that the evidence is very strong that increasing neurogenesis is at least a bellwether or a biomarker" that an antidepressant treatment is starting to work, he said.
It is also clear, researchers say, that new tools that may prove the neurogenesis theory true or false are quickly coming on line.
William Greenough, a prominent University of Illinois brain scientist is beginning to be able to label neurons in animals by when the cells were born, he said, and follow them through time.
And just this month, a team of researchers led by Grigori Enikolopov of Cold Spring Harbor Laboratory in New York announced that using technology related to MRI and a telltale marker for new cells, they had developed a noninvasive method to track neurogenesis in living humans. The first tool with such a capability, it could help researchers study brain cell changes in a variety of diseases from depression to cancer and stroke.
If such tools succeed, future psychiatrists may use signs of neurogenesis in their patients' brains to tell quickly whether a treatment holds promise, instead of waiting weeks as they often must now.
"You could do this in the office to speed people's antidepressant treatment," said Dr. John Denninger, a Massachusetts General Hospital psychiatrist doing research related to neurogenesis. "You'd improve the odds that you'd actually give them something that would make them better.