"It would be ironic if the shortfall in a same gene brokered the path to brain disease or rejuvenation," says Nottebohm, who is Dorothea Leonhardt Professor and head of the Laboratory of Animal Behavior. "But naturally replaced neurons may hold the key to understanding processes of neurodegeneration."
The results appear in the May 23 Early Edition issue of Proceedings of the National Academy of Sciences.
The potential medical implications of the study came as a surprise to the researchers. They set out to answer fundamental questions about the small number of brain neurons that can be replaced when they die: Are these cells similar to other neurons that are not replaced in adult life? Or, in terms of which genes are "turned on," are they marked as transient?
Nottebohm's laboratory has pioneered the study of the brain pathways controlling how songbirds learn to sing. A discrete region of the brains of songbirds, known as the high vocal center, controls their singing behavior. It was in this brain region that, more than 20 years ago, Nottebohm discovered new neurons being born in adult birds, overturning the conventional wisdom that all neurons in vertebrates are created by birth or soon after.
In the new study, the researchers looked at cells in the high vocal center region of the brains of 19 zebra finches, a kind of songbird native to Australia that has long been used in studies of birdsong. Since replaceable neurons sit side-by-side with neurons that are not replaced, the first step was to label the two kinds of cells in order to tell them apart.
The replaceable neurons send projections to another part of the brain a few millimeters away from the high vocal center to form part of the brain pathway that controls muscle movements related to singing. The neurons that are not replaced send their projections to a different brain region, one that controls the song learning pathway. In order to label the two cell types, the scientists injected dyes of different colors, one green and one red, into the brain regions where the neurons project. The dyes then traveled back to the bodies of the cells in the high vocal center. As a result, replaceable and nonreplaceable neurons could easily be distinguished under the microscope.
Then, looking at very thin slices of brain tissue, the researchers used a technique called laser capture microdissection to collect cells of each type. From these they purified RNA, the quantity of which corresponds to the activity of the particular gene producing it. The microdissection technique uses a low-power laser to melt a small circle of film into the tissue. When the film is lifted, the cell body - including its RNA - comes with it, as if removed by a hole punch.
After collecting, purifying, and amplifying the RNA from 3,500 cells for each bird, the researchers used a microarray created in the lab to find out which of the genes in the cells were turned on. This technique allowed them to identify which genes were active, or expressed, by comparing them with 800 known genes from the zebra finch.
They expected to find a few overactive genes in the replaceable neurons. But instead, one gene turned up in consistently low levels.
The gene was UCHL1.
"We expected UCHL1 would be high in all kinds of neurons," says first author Anthony Lombardino, Ph.D., a postdoctoral fellow in Nottebohm's laboratory. The gene is known to be highly expressed throughout the brain, and in fact, the UCHL1 protein is thought to make up as much as 2 percent of total soluble brain protein. But in the zebra finch brains, nonreplaceable neurons produced, on average, about two and a half times the amount of UCHL1 as replaceable neurons.
In addition, studies from other laboratories have pointed to deficiencies in UCHL1 in degenerative diseases such as Alzheimer's and Parkinson's.
Experiments in mice confirmed the finding that low levels of UCHL1 correlate to the ability of neurons to be replaced. The researchers used similar techniques to assess UCHL1 in hippocampus and olfactory bulb neurons in mouse brains, areas where neurons are known to undergo spontaneous replacement. They found markedly low levels of UCHL1 in the replaceable neurons of both these regions.
"Low levels of UCHL1 appear to be a feature of replaceable neurons wherever they occur," says Lombardino.
The scientists also knew that, in songbirds, most neurons born during adulthood die, but a bird's active singing increases the survival rate of new neurons. They reasoned that, if UCHL1 levels are related to the death or survival of neurons, these levels should increase when birds are singing and when more new neurons thus have a chance of surviving. In another experiment the scientists allowed male finches to sing to a female before analyzing the birds' brain neurons. As predicted, the expression of UCHL1 in replaceable neurons, but not in nonreplaceable ones, increased in birds that were singing.
"These findings suggest that rising levels of UCHL1 may be associated with a reduced risk of neuronal death," says Nottebohm. "This study draws attention to the replaceable neurons of adult brain as a wonderful model system in which to study changes in genomic expression related to learning, aging and neurodegeneration, in addition to those that are related to the choreography of replacement itself."
"In addition, because we study birds that learn their songs, the story relates to a fascinating, learned behavior, so it offers the opportunity to show how integration at different levels of biology can be made to pay," Nottebohm says. "It allows us to think about the brain in ways that no single level of analysis would allow."
Proceedings of the National Academy of Sciences