EVER since his heart attack, George's memory has been shot. Has he spoken to the grandchildren lately? He can't remember. Has someone just phoned? He can't recall. The heart attack cut off the supplv of oxvoen to his brain and certain cells were damaged. Permanently, irreparably damaced. Irreparablv? So he was told. Now, however, a few hundred middle-aged rats suggest that one dav it may be otherwise. They also suffered heart attacks and failed memories as a result, but when a team of scientists at the Institute of Psychiatry in London injected them with genetically engineered brain cells from very young mouse embryos, the rats got their memories back. The injected cells migrated to the region of damage, took on the form the brain cells had before they were injured, and started doing their work. "Our cells prefer areas of damage. They avoid the undamaged parts," says Jeffrey Gray, a senior member of the research group. "Whv that is so, I'm not sure," he told corporate and university scientists at a small private seminar in Oxford in January, "but it does look promising."
Promising indeed. If the mouse cells really can detect where a brain needs mending and where it does not, this new technique mav open the door to the prospect not onlv of neuro-repair for humans but of routine neuro-maintenance as well. Why wait until the Alzheimer's is full-blown before injecting some replacement cells? Vvhv ivait till memory is completely ravaged bv old age before seeking treatment? If what the team reports is true, engineered embr-,Io cells should be able to seek out and patch up any old brain damage-whether you knoiv you have it or not. The idea of replacing old, damaged brain cells with spanking new ones is not original. Indeed, the first adult-to-adult transplant in animals was attempted back in 1890. But the operation didn't become truly viable for humans until 1990, when a group led by Anders Bj6rklund at Lund University in Sweden, showed that they could alleviate the worst symptoms of Parkinson's disease by transplanting brain tissue from aborted human fetuses. Since then, hundreds of patients with Parkinson's have received fetal tissue transplants, sometimes with impressive results. That technique has a few limitations, though. Six or seven freshly aborted fetuses are needed for each operation. The fetal cells must be taken at a very specific time during development-just at the point when thev are differentiating into the type of celfs they will replace. Too early, and they won't be the right kind of cells; too late, and you risk damaging the developing axons as you transplant them. This is all fine and well for Parkinson's, since the cells needed for the transplant begin to differentiate at around the sixth or seventh week, falling within the normal range for human abortion. But the hippocampal cells that George needs to restore his memory do not start to specialise until the end of the first trimester, at which point abortion is less common. What's more, when it comes to fetal tissue transplants, the cells must be taken from the very region of the fetal brain that the surgeons aim to repair in the patient-no easy task. Like a handful of similar techniques being developed around the world, the one being pioneered by Gray and his colleagues John Sinden and Helen Hodges has the potential to get around these problems because it uses cells grown in a flask rather than brain tissue from an aborted fetus. And what puts the Institute of Psychiatry team at the head of the pack, is that they have shown with behavioural tests in rats that those laboratory-grown cells actually restore lost brain function. The team started with a very special rodent, the "inunortomouse", developed by a collaborating lab at the Ludwig Institute for Cancer Research in London. Immortomice are genetically engineered so that every cell in their bodies contains a gene that instructs cells to divide, and which is sensitive to temperature and gamma-interferon, a protein that regulates tissue growth. At body temperature, this gene is inert. But when immortomouse cells are grown in a culture flask at 33 'C and gammainterferon is added, the gene is activated, and instructs the cells to keep dividing far longer than normal, ignoring the signals that would usually tell them to die after a few rounds of cell division. Immortomouse cells grown in a flask stop dividing if you raise the temperature back to 37 'C. Gray, Sinden and Hodges knew from their own unpublished experiments that cells taken from the hippocampus of very early immortomouse embryos do something even more extraordinary as -they lose their ability to divide. As long as they are given the necessary growth factors-proteins that direct cell growth-they become specialised, producing everything from nerve cells to glia, the brain's supporting tissue. Unlike the pluripotent stem cells, nerve cells in adult animals are incapable of dividing and spawning new types of cells even in the presence of growth factors.
The researchers wanted to find out whether the stem cells' ability to create different types of tissue might help to repair the kind of brain damage and memory loss that is suffered by about 10 per cent of people who have heart attacks. They used a technique known as a "4vessel occlusion" to trigger a 15-minute heart attack in normal lab rats. The animals survived, but their memories were impaired and they performed exceptionally badly on standard tests, such as remembering the location of a platform submerged in milky water. Postmortem examinations of these heart attack rats showed that cells from the CAl region of the hippocampus were obliterated-the very ceus that are damaged by human heart attacks. Two weeks after the heart attacks, Gray, Sinden and Hodges injected immortomouse stem cells into the rats' brains. When they were tested just six weeks later, the rats did almost as well on the water test as healthy control rats. Subsequent work on marmoset monkeys shows that even four months after damage to the CA1 region, an injection of the engineered mouse stem cells undoes the damage. A transplant from one species to another is possible because the stem cells do not appear to trigger an immune response in a recipient. No one knows for sure why, but it is probably due to the combined effect of the immune system not being very active in the brain, and the stem cans lacking some of the surface proteins that act as a red flag to immune cells.
More remarkable, however, and what holds great promise for medical science, was what an examination of the animals' brains revealed about how the injected cells went about their restoration work. First, even in adult animals that had long since lost the capacity to grow new brain cells, the injected stem cells developed into very specialised CAl cells. Secondly, the cells seemed to know exactly where to go. The researchers had injected the cells close to the site of damage, but not too close in case the injection
As it turned out, not aiming directly for the bull's eye didn't matter at all because the cells compensated, migrating up to 3 millimetres in rats and 8 millimetres in marmosets in pursuit of damage-some of which, Gray hinted at the Oxford seminar, was not the original target of the transplant. Says Gray: "They've gone to the site of damage and taken up their home there." In earlier experiments, where the researchers had transplanted already specialised CAl cells straight from fetuses into rats with damage to their CAl region, the cells simply set up shop where they landed, forming a distinct mass that sat on top of the damaged area. The cell's migration to the area of damage was a "big surprise", but not inexplicable, according to Helen Pitcher, another member of the institute's team. "Stem cells are inherently migratory," she says. After all, the tremendous increase in brain size and complexity during development is largely a function of stem cells dividing and migrating to new positions before they finally become specialised for their permanent tasks. All the available evidence suggests that the developing brain provides a carefully choreographed sequence of chemical markers to show stem cells the way. But how do cells know where to go in the already developed adult brain? "Part of it may be the damaged brain sending out messages saying, 'Come over here'," says Pilcher. For instance, damaged brain cells release various growth factors which could act as a beacon. Once the cells find their target, the local envirorunent, including-presumablylocal growth factors, appears to be au it takes to ensure they become the right sort of specialised brain cells. At least one lab, Evan Snyder's at the Harvard Medical School in Boston, had already shown that brain stem cells from n-dce embryos will become specialised and produce both neurons and glial cells when injected into brains of newbom mice. But in newbom mice, the brain is stffl developing. The Institute of Psychiatry team showed that the stem cells behave the same way in the fully developed brain of middle-aged or elderly animals. 'Perhaps damage [itself] is in some way a condition that resembles the developing brain,' says Sinden. "It remains an interesting finding. It's not something the literature suggested would happen." The Institute of Psychiatry team also checked carefully, but found no signs of cancel in the rats and marmosets. "We've never seen tumour formation in our animals," says Gray. Other tests indicated that the immortomouse gene had switched off once the stem cells reached the animals' brains. Those results are important because the new technique does carry at least the theoretical risk that the genetically-engineered cells would fail to stop proliferating despite the higher temperature of a patient's brain.
"From what I've seen, [their work] is extremely impressive-especially the behavioural data," says Samuel Weiss, a neurobiologist at the University of Calgary, who is following the London team's .progress. But, he cautions, "the key question is whether the same technology can be adapted to humans ... It's a giant leap from mouse to man." But that race is now on. In July last year, backed by E250 000 from Chris Evans, the Welsh biotedi daredevil, Gray, Sinden and Hodges formed a company, ReNeuron, to pursue the idea. This January, Evans's investment company, the Merlin Fund, handed over an additional E5 million. The team has filed for a patent for their "pluripotent neuro-epithelial cells"and, more importantly, the use of a human variant of such cells to treat brain damage caused by all manner of ills that humans are prey to such as "traumatic brain injury, stroke, perinatal ischaen-da including cerebral palsy, Alzheimer's disease ... and Creutzfeldt-jakob disease . . ." Now, the team is identifying a range of human brain cells in 7 to 12-week-old embryos that will grow in culture flasks and have the ability to spawn different cell types. Once they have done that, they plan to insert into the stem cells a temperature-sensitive gene. (The additional sensitivity to gammainterferon that the inimortomouse gene has is probably not necessary, says Sinden.) Finally, they will test the engineered human stem cells on rats and monkeys before trying them in humans. If it proves impossible to grow, engineer, or do effective transplants with human stem cells, the team is even considering a fallback plan. Given that the mouse cells worked in the marmosets, they may work in humans too, says Sinden. Whatever the source of the stem cells for human transplants, if Gray, Sinden and Hodges can make the technique work, no more human fetuses will need to be used for such grafts. Meanwhile, a few other groups also have their eyes on the prize of human stem-cell transplants. 'We believe we have [identified] some human stem cells," says Snyder, who is interested in the potential of the cells for repairing spinal cord injuries, and heritable brain disorders such as Tay-Sachs disease. His team, and two others working on human brain stem cells, one led by Angelo Vescovi at the University of Calgary in Alberta, the other by Ron McKay at the National Institutes of Health near Washington DC, are expected to publish their results in the next few months. And almost everyone in the know agrees that trials using stem cells to repair the human brain are likely to begin within five years. "We are on the verge of seeing this happen," says Weiss. The institute team says that its first target will be acute damage, such as that caused by carbon monoxide poisoning or heart attack, as well as degenerative diseases such as Huntington's disease. But if the cells really are being beckoned by areas of damage, couldn't they sort out the widespread brain injury caused by, say, stroke? Oxygen deprivation at birth? Or even the mental decline that goes with normal ageing? "At this stage," says Gray, "there's nothing we can exclude." Sinden agrees, but cautions that the most severe brain injuries and diseases will get priority and approval first. Already, hundreds of people have asked them for help. 'We have a very, very full postbag," he says.
Further reading: "Immortalized neural progenitor cells for CNS gene transfer and repair" by A. Martinez-Serrano and A. Bjorklund, Trends in Neuroscience, vol 20, p 530 (1997) 'Gene therapy and neurodegeneration" by E. Y. Snyder and J. D. Macklis, Clinical Neuroscience, vol 3, p 310 (1996) 'Recovery of spatial learning by grafts of a conditionally immortalised hippocampal neuroepithelial cell line into the ischaemia-lesioned hippocampus" by J. D. Sinden and others, Neuroscience, vol 81, p 599 (1997)