"Our aim was to make a modelArmenise-Harvard Symposium
Neurons may be the principal actors of the brain but, like many prima donnas, they depend on a retinue of helpers to enhance their function of sending messages. These attendants, called glial cells, perform a variety of tasks--supporting neurons, mopping up excess chemicals from the extracellular space, guiding immature neurons as they migrate through the brain, blanketing their long processes with the communication-enhancing substance called myelin.
This beneficial arrangement is subverted in people suffering from malignant brain tumors. Glia divide uncontrollably, forming masses called glioblastomas that are difficult to treat. Some send out tentacles that cannot be surgically removed. In addition, individual tumor cells may break off and cross the corpus callosum, the band of tissue bridging the two halves of the brain, setting up new centers of disease. Because these migrant cells do not divide during their journey, and may even remain dormant in their new lairs, they elude chemotherapeutic agents designed to track down and kill dividing cancer cells.
"Our aim was to make a model
that allows greater control
of lesions and also dissection
of genetic events that in the
future might allow for better
diagnosis and treatment."
--Harold Varmus
"The enemy here is monstrous," said Xandra Breakefield, professor of neurology at HMS and MGH. She was speaking on June 24 at the Second Annual Symposium of the Giovanni Armenise-Harvard Foundation, in a program titled "The Genetics of Human Brain Cancer." (See p. 7 for coverage of the preceding Armenise program.) After years of frustrated attempts, scientists are beginning to understand how brain tumors form--and how they might be stopped. Speakers at the symposium, attended by over 200 people, pinpointed some of the mutant genes that cause glial cells to relentlessly divide and outlined strategies to arrest the process.
An All-out Attack
Breakefield and her colleagues have been testing a gene-therapy strategy in rats that tackles brain tumors on all fronts--the main mass, migrant tumor cells, and new foci. After removing the tumor surgically, the researchers inject viruses carrying a variety of anticancer genes. Some, hefting toxin genes, kill tumor cells directly. Others carry the gene for an enzyme that converts nonpoisonous substances into toxins and cause the cells to produce toxins that spread to other tumor cells. "We are able to save 50 percent of rats this way," she said.
To target migrant tumor cells, they are taking an equally ingenious approach. Rather than deliver gene-carrying viruses directly, the researchers are packaging them into a line of migratory brain cells, called neural progenitor cells, developed by Evan Snyder, assistant professor of neurology at HMS and Children's Hospital, and his colleagues. In preliminary experiments, the virus-carrying cells migrated and even killed cells at the tumor edge, Breakefield said. Eventually, she hopes to use the virus-carrying cells to "sensitize" migrant tumor cells, that is, to render them susceptible to conventional treatments. She described a plan in which neural progenitors would be used to deliver "activating" genes that prod migrant tumor cells into dividing, thereby making them visible to chemotherapeutic agents.
Finally, to target new foci, she and her colleagues are testing ways of delivering gene-carrying vectors through the new blood supply surrounding tumors. "Unfortunately, you don't get neovascularization until the tumor is a certain size," Breakefield said.
Tumor Formation
A novel answer to the question of how glioblastomas form--and specifically the role played by the well-known tumor-suppressor gene p53--was presented by HMS professor of cell biology Frank McKeon. Researchers have suspected that in addition to monitoring DNA damage, p53 plays a role in maintaining the proper number of chromosomes in the glial cell--perhaps by overseeing their attachment to spindles during cell division--but it was not clear what that role actually was. McKeon and his colleagues have recently conducted an experiment suggesting that rather than monitor the situation directly, p53 may receive distress signals from another protein.
Normally, if a chromosome fails to attach to a spindle, a signal is sent out to halt mitosis. If the problem is not remedied within a period of time, the cell commits suicide. The researchers found that when they blocked the normal action of this protein, Bub1, glial cells would clump rather than die. They did so even if the cell had a normal p53 gene. "So p53 is probably responding to signals from spindle assembly checkpoints," said McKeon.
Other recent findings from his lab suggest p53 is interacting with a whole host of proteins. For example, graduate student Annie Yang has recently discovered a protein, p63, that when truncated prevents the p53 gene from being expressed. "We think we're stumbling and bumbling into a whole network of interactions," said McKeon.
Echoing the theme of multiple interactions, Harold Varmus, director of the National Institutes of Health, described how he and his colleagues transfected two mutant genes--EGFR and cdk4--into mice developed by NIH colleague Eric Holland. Only when both mutants were introduced--and only when they were transfected into mice lacking both copies of the INK4 gene--did they produce tumors. Adding mutant p53 to the mix significantly increased the number of tumors.
Although preliminary, such investigations into the genetic events underlying glioblastomas could point the way toward better diagnosis and treatment, said Varmus. He and his colleagues have designed their mouse model with this aim. The NIH mice develop hydrocephaly when they have brain tumors, making the tumors visible and eliminating the need to sacrifice the animals for diagnosis. Varmus and his colleagues plan to use nuclear magnetic resonance imaging to scan the brains of the hydrocephalic mice and possibly rescue them. "We want to see tumors earlier while the animal is still alive and perhaps select them for therapy," he said.
--Misia Landau
Focus 7/17/98
Armenise-Harvard Researchers Speak International Language
While the Italian soccer team was competing for the World Cup in Paris, a meeting room on Cape Cod was crowded with 130 basic researchers, one third of them from Italy. In the front row sat Count Giovanni Auletta Armenise, flanked by Dean Joseph Martin and former dean Daniel Tosteson. Like everyone else, they watched mesmerized as a culture of epithelial cells, prodded by a dose of extracellular "scatter factors," sprouted branched processes as elaborate as deer antlers. The video was grainy, black and white, and only three minutes long. Yet for this group, it spoke an international language more compelling than sport.
The Second Annual Symposium of the Giovanni Armenise-Harvard Foundation, whose first program was held in Chatham on June 22 and 23, brought together foundation-sponsored investigators for two intense days of lectures, poster presentations, and networking. Sessions centered on neurobiology, cell membrane traffic, gene transcription, cell signaling and cycling, and plant defense and pathogenesis. Although each presentation was tightly focused and highly specialized, the underlying goal was always the same: to find patterns that hold true for different types of cells, tissues, and organisms.
Italian plant biologists Guilia De Lorenzo (left) and Felice Cervone are studying a family of proteins that seem to be involved in plant immune function.
For example, the scatter factors described by Paolo Comoglio, of the Institute for Cancer Research in Turin, are needed for development of many types of tissue. In mammalian embryos, they control branched morphogenesis--in which cells break away, migrate, become polarized, and form tubules--in normal tissues including nerve, muscle, blood vessels, and bone. Additionally, Comoglio's lab has demonstrated that certain point mutations in these specialized factors can transform healthy growth into malignancy, showing that they play a role in pathogenesis, as well.
In a presentation that helped explain the workings of whole organisms rather than tissues, Charles Weitz, assistant professor of neurobiology at HMS, described the molecular mechanisms of endogenous, self-sustaining clocks that have been found in everything from bacteria and fungi to plants, invertebrates, and humans. Only a year ago, researchers at Northwestern University identified the first mammalian circadian gene. Weitz and his HMS colleagues have since found that this gene's protein, CLOCK, forms a heterodimer with a second protein, BMAL1, to turn on a well-known circadian gene called per.
Further experiments indicate that teamwork is the key. Together, CLOCK and BMAL1 have 10 to 15 times the transcription activity of either protein acting alone. Now that it is clear that these two proteins turn on per, Weitz said, the next challenge is to figure out what turns the gene off during normal circadian rhythms.
Speakers at a session on plant defense and pathogenesis took an even broader view, blurring traditional boundaries by demonstrating that the animal and plant kingdoms are more alike, in at least some regards, than previously thought. Frederick Ausubel, professor of genetics at HMS and MGH, described how novel pathogens made in his lab can cause disease in certain model plants, flies, nematodes, and mice. Strains that are especially virulent in plants, Ausubel said, are often even more pathogenic in animals. Felice Cervone and Giulia De Lorenzo, plant biologists from the University of Rome, made the case that plants, like animals, readily identify nonself molecules that may pose a threat to them. The scientists are investigating a family of proteins that appear to have immunologic functions in plants.
At HMS, the Armenise-Harvard science centers are in cell signal transduction, directed by Marc Kirschner; structural biology, directed by Stephen Harrison; neurobiology, directed by Gerald Fischbach; and human cancer viruses, directed by Peter Howley. Programs in plant biology are directed by Frederick Ausubel at MGH.
--Patricia Thomas
Focus 7/17/98