Brain Building May Depend On DNA Cutting And Pasting

Researchers at the Howard Hughes Medical Institute at Children's Hospital in Boston and the Center for Blood Research have made a discovery that could help solve one of the central riddles of biology-how the brain, with its dazzling display of cell types, develops from a relatively undistinguished pool of progenitor cells.

For years, immunologists have known that the immune system, with its multitude of T cells and antibodies, is produced not by a finite set of pre-existing genes but through a cutting and pasting of DNA fragments. This recombination process is carried out by a set of proteins that essentially snip genes into bits and then join the broken DNA ends, forming a nearly infinite supply of new genes. It now appears that two of these "paste" proteins are also required during a very specific period in brain development-when progenitor cells are developing into different kinds of neurons. The discovery, by a team led by Fred Alt, a Harvard Medical School professor of pediatrics at the Center for Blood Research and Children's, appears in the December 23 Cell.

"The exciting possibility is that there could be some program-gene rearrangement or something equivalent-that's occurring at a very specific time in brain development," says Alt, the Charles A. Janeway professor of pediatrics at HMS. "It may not be as complicated as putting whole sets of genes together as in the immune system. But one could imagine simpler things like molecular switches-DNA segments that are flipped or deleted to turn genes on and off in an irreversible fashion-which in turn could result in differentiation of progenitor nerve cells."

To arrive at their findings, Alt and his colleagues bred mice lacking genes for the two paste proteins, XRCC4 and ligase IV. In addition to severe immunological deficiencies, both strains of mice displayed holes in regions of their brains, signifying areas of cell death. Overall, the defects were so serious that the mice died before birth.

What was especially striking, says Alt, is that the damage to their brain was inflicted during a very specific stage in neural development. "It's a very narrow window," he says. Another remarkable finding was that the cells died not during division, when chromosomes stretch out and can break or be damaged, but just afterwards.

"The onset of cell death is very tightly linked to the period when dividing progenitors are differentiating into neurons. That makes it potentially very exciting," Alt says. Although he is not certain what is killing the brain cells, he believes it is "very likely" that they are dying because they have DNA breaks that they are not mending.

"The question then is why do they have the DNA breaks during this specific period?" says Alt. "The possibility arises that there's some specific process at that time. That's very exciting to us because we know that process is either happening just before the neural progenitors begin differentiating or it's happening just after. So that means we know where to look for this event."

Alt and his colleagues are developing assays to see if genes expressed during differentiation require breaks for their expression. The first step will be to see if genes expressed during or just after the critical period in normal brains are missing in those of mice without XRCC4 or ligase IV. "The next step would be to see if these genes in normal mice are associated with DNA breaks," he says.

He points out that the DNA breaks occurring during recombination in the immune system do not occur randomly but at specific sites on the chromosome. The cut-and-repair proteins in the brain could work to create flips and deletions with the same precision.

In the immune system, cut-and-paste genes code for T cell receptors and antibodies. What might break-dependent genes in the brain be expressing? "I don't know. It could be a receptor or some other protein that ultimately leads to expression of a surface receptor," Alt says.

So far, total deletions of XRCC4 and ligase IV have not shown up in humans. Presumably, such mutations would be embryonically lethal, Alt says, "but it is possible that more subtle defects in the genes could show up." Defects in other related genes have been found in people with severe immunodeficiency disease. "Maybe now we will evaluate these people for neurological manifestations," he says.

In theory, Alt believes it might be possible to correct these defects with gene therapy. "But that's way down the road," he says.

This research was supported by the Howard Hughes Medical Institute and the National Institutes of Health.