The study, led by researchers at NYU School of Medicine, is published in the January 2, 2004 issue of Science magazine.
"What we found was a complete surprise," says Gordon Fishell, Ph.D., Associate Professor in the Department of Cell Biology at New York University School of Medicine. "No one had believed that it was possible to push back the birth date of a cortical neuron. There is this central tenet governing the process of brain development, which says that late progenitor cells [forerunners of mature cell types] cannot give rise to cell types produced earlier in development," he explains.
"Consequently, while some populations of stem cells exist in the adult brain, these cells are restricted to producing only a subset of cell types," notes Dr. Fishell. "If one's goal is to produce cells for replacement therapy, some method must be found to turn back the clock and allow adult stem cells to give rise to the wide variety of cells made during normal brain development."
Eseng Lai, Ph.D., of Merck & Co. and one of the study's co-authors, cloned the Foxg1 gene while he was working at Memorial Sloan-Kettering Cancer Center in New York. He also did seminal work in the late 1990s showing that when the gene is eliminated in embryonic mice, the brain's cerebral hemispheres barely develop. Subsequent work demonstrated that the gene played a role in the early phases of cortical development.
The cerebral cortex is massively folded gray matter incorporating billions of neurons. Despite its complexity, the cortex comprises six orderly layers of cortical cells that are laid down during development at a precise time and in a precise sequence. In the study, the researchers asked which cortical cell types embryonic mice lacking Foxg1 can generate. Carina Hanashima, Ph.D., a postdoctoral fellow in Dr. Fishell's laboratory who had previously worked with Dr. Lai, conducted a series of experiments that made the analysis possible.
The progenitor cells for the cortex are born in a zone deep in the brain, and migrate to their assigned layer, depending on the time they are born. So a cortical cell's identity is based on the date of its birth. The first cortical cells to be born populate layer 1, the most superficial layer, which is made up of special Cajal-Retzius (CR) cells. The next cells born migrate to the innermost layer, layer 6. Subsequent layers pile up above layer 6 (between layers 6 and 1), and in descending numerical order from 5 to 2. Each layer has a specific type of neuron associated with it.
The researchers looked at the cortical layers in embryonic mice at a time in their development when layers 1, 6, and 5, would normally have already been formed. In mice lacking the Foxg1 gene, the researchers found that only layer 1, which is made up of CR cells, was present, and these cells were abundant. The absence of other cell types implicated the gene in producing later-born cortical cell types.
One of the ways the scientists were able to identify CR cells was by the expression of a protein called reelin, which plays a vital role in building the developing brain and is only present in CR cells. Mice lacking the protein stumble around so much that they were named "reelers." The cortical layers in these mice are scrambled. In recent years, reelin deficiencies have been linked to such human disorders as schizophrenia and epilepsy.
In subsequent experiments, the researchers asked how the overproduction of CR cells occurs. They used a clever biochemical manipulation that served as a kind of genetic stop watch, allowing them to temporarily turn off the Foxg1 gene in late progenitor cells, after the normal birth date of CR cells had passed. In this way, they observed that cortical cells destined to become layer 5 became CR cells instead. Apparently, the gene orchestrates the program responsible for ensuring that the cortical layers of the cerebral cortex are laid down in a precise sequence. When the gene is inactivated or turned off, the program seems to revert to its earliest stage.
The researchers do not know how late in the game they can play their genetic tricks. If they turn off the Foxg1 gene at a later time in development, such as when cortical layers 2 or 3 are forming, will progenitor neuronal cells still become CR cells? Are there other genes that control the developmental clock? If such genes exist, it may be possible to turn these genes off in adult neural stem cells, and thereby generate a far broader array of tissue than otherwise possible. "I would say that the chances of this happening are very remote," says Dr. Fishell, "but then again, I never thought that the clock could have been turned back in neuronal progenitor cells."
The authors of the study are Gordon Fishell and Carina Hanashima of NYU School of Medicine; Eseng Lai of Merck; Suzanne Li of Memorial Sloan-Kettering Cancer Center; and Lijian Shen of Weill Medical College of Cornell University.