Deni S. Galileo

Using the developing chick optic tectum (midbrain) as the model for vertebrate brain development, a variety of methods are used to investigate these processes. A main technology we use for this is one that I helped to develop: retroviral gene transfer. Here, a recombinant retroviral vector carrying a marker gene and another cDNA (or its antisense copy) can be used in vivo to infect brain progenitor cells that line the ventricular cavity. The retroviral vector will incorporate the recombinant DNA into the infected cell's genome and express it. Then, we can analyze how the expressed single protein (or the antisense-attenuated endogenous protein) affects the processes we are interested in. For example, we were the first to show that integrin extracellular matrix receptors were involved in brain development by using retroviral transduction in vivo of antisense sequences against β1 integrins. This caused infected neuroblasts to fail to migrate into superficial brain laminae, and then, to die. Similar experiments followed to show specifically that integrin heterodimers α6β1 and α8β1 were responsible for correct cell migration and survival, respectively. Replication-incompetent retroviral vectors are used when cell-autonomous effects are desired, and now we are using replication-competent vectors and in ovo electroporation to achieve widespread misexpression of several proteins in developing brain.We are investigating the roles of suspected integrin extracellular matrix substrates in migration and survival. We are also investigating the roles during brain development of other adhesion molecules (e.g. L1/NgCAM), contactin, the proto-oncogene c-src, and its viral counterpart oncogene v-src.

A focal point of the lab now is to further develop and utilize the chick embryo as a model for the study and manipulation of basic mechanisms of abnormal brain cell migration- i.e. glioma tumor cell local invasiveness and breast cancer spread to brain. This is a logical extension of our v-src, integrin, and NgCAM/L1 work, and is of interest because the insidious nature of glioma cells lies in their capacity to extensively migrate throughout the brain, particularly along axon tracts and blood vessels. This extensive migration usually precludes successful resection of the tumor by surgery. I believe that some of the same molecules and mechanisms that are used during normal neuronal migration are usurped by glioma and other cancer cell types that spread in the brain. These tumor cells are also more amenable to culture and in vitro assays than are normal brain cells, which allows meticulous dissection of postulated migratory mechanisms under a variety of in vitro conditions. Toward this new focus, we have shown that human and rat glioma tumor cell lines are capable of producing invasive tumors in the developing chick brain and are now beginning such experiments to determine whether primary human glioma tumor cells from surgical samples do this as well. This novel chick model ultimately should be as useful as immunodeficient mice for studying certain mechanisms of invasive brain tumors, and it has several practical advantages like accessibility, ease of manipulation, and cost. We also have shown that human breast cancer cells injected into the extraembryonic blood vessels of the chick embryo extravasate and invade the brain within days. Brain metastases from breast cancer often kill within weeks to months, and this new model undoubtedly will prove useful in studying certain aspects of metastasis, like homing to brain and extravasation. In collaboration with Dr. John Koh (UD Dept. of Chemistry and Biochemistry) we also have begun a project determining the effects on axon outgrowth of patterns of gene expression in cell monolayers that are induced by light.