|
The Giovanni Armenise-Harvard Foundation Session 2: Cellular Differentiation Overview One of biology's greatest wonders is that a fertilized egg gives rise to an embryo made up of myriad cell types that are not only chemically different, but also arranged in a specific, three-dimensional pattern. Cells that inherited exactly the same genetic material from the egg diverge into brain and bone, hair and heart. Such chemical and architectural variety is possible because genes are switched on or off, and are expressed differently in diverse tissues. This session, like the opening one, also concerned development,
Dr. Tullio Pozzan said in his introduction. The first presentation focused
on a novel method for identifying genes that are activated in normal
and malignant growth of epithelial cells. The other three explored various
facets of neuronal tissue development. Presentations
In normal development, scatter factors stimulate epithelial cells to execute a complex program that culminates with polarization and the formation of tubules; in invasive tumor growth, this normal process is subverted and cells proliferate and invade in abnormal ways. One way to understand the differences between normal and malignant growth is to compare which genes are switched on in each. In order to do this, Dr. Medico's team created a "gene trap" - a novel fusion protein that can be used to screen a cell's entire genome for activated genes, whether or not their sequence and function are known. They built a promoterless retroviral vector carrying a reverse-oriented splice acceptor (or ROSA) gene and the sequence for a green fluorescent nitro-reductase (GFNR) fusion protein that serves as a marker. When this gene trap encounters an active promoter, the trap construct will integrate itself downstream and the marker will be expressed. Cells that have taken up the trap can then be identified using FACS analysis. Conversely, the nitro-reductase moiety allows pharmacological selection against constitutive GFNR expression. A mouse liver cell line was stimulated with hepatocyte growth factor (HGF), a scatter factor, and screened with some traps set to select for HGF-induced genes, and others designed to pick out genes suppressed by exposure to HGF. Some 60 different traps were used to categorize genes and pick out the most promising HGF targets. Several responsive clones were isolated, and regulated expression of the trapped gene was confirmed at the RNA level. When Dr. Medico and his colleagues sequenced the regions around trap sites, they found genes that had never before been linked to scatter factor biology. The goal of future studies will be to shed light on transcriptional response in normal and cancerous cells.
The long-range goal of Dr. Thor's work is to understand the molecular genetic mechanisms that control establishment of motor neuron identities. His laboratory uses Drosophila as its primary model, and although the fly has a relatively simple nervous system it still features about 100 distinct types of cells that can be classified as neurons, glia, or interneurons. Recent experiments have focused on LIM homeodomain proteins, a family of transcription factors that are expressed in discrete subsets of developing neurons throughout the animal kingdom. Dr. Thor's experiments indicate that three LIM-HD genes, islet (isl), lim3, and apterous act in a combinatorial code to specify motor neuron subtype identity. By attaching markers to mutant versions the genes, he has found that they control two basic hallmarks of neuronal identity - they guide axons toward target cells and specify which neurotransmitters are turned on. Additional genes are probably required to establish the ultimate, unique identity of neurons, and current research focuses on identifying them. Because LIM-HD programs appear to be highly conserved, Dr. Thor hopes that his findings will ultimately help medical scientists understand vertebrate motor neuron generation and differentiation. With this knowledge in hand, it may someday be possible to replace cells that are lost in spinal cord injuries or neurodegenerative disorders.
A classic family of helix-loop-helix (HLH) transcription factors are the myogenic proteins, which are well known for their role in the differentiation of muscle cells. Dr. Consalez' lab has a long-time interest in a different subclass of transcription factors with the HLH DNA-binding motif, called the Ebfs. This gene family was originally implicated in B-cell maturation and olfactory function. Several years ago, his group identified two family members in the mouse (Ebf2, Ebf3); more recently they found two more in Xenopus laevis (Xebf2, Xebf3). Other investigators have cloned Ebf family members from the nematode C. elegans. Just as myogenic proteins can trigger events that turn epidermal cells into myoblasts, Dr. Gonzalez' team has demonstrated that overexpression of Ebf genes in Xenopus laevis embryos can transform presumptive epidermis into neurons. Ebf expression in frogs begins very early in embryonic development and continues through the tadpole stage, with different genes acting at different times. Xebf2 operates at early stages of neuronal differentiation, upstream of NeuroD, whereas Xebf3 is a target of NeuroD and plays a role in terminal neuronal differentiation. In the mouse, three Ebf genes have been shown to advance neuronal differentiation after primary neurogenesis is underway. A tantalizing feature of Ebf proteins is that they have intrinsically different functions, and act at different stages of development, despite having very similar molecular structures.
For nearly 100 years, scientists have viewed the development of the central nervous system as a life and death matter. Cells are initially produced in huge excess, then whittled away by cell death as development proceeds. Which cells live or die is regulated by extracellular growth factors, such as neurotropins, and Dr. Bonni's laboratory focuses on exactly how these life-or-death decisions are carried out. Using cerebellar granule neurons obtained from rat pups and grown in culture, he has been able to pinpoint the pro-life activities of a polypeptide growth factor called brain-derived neurotropic factor (BDNF). BDNF appears to promote cell survival in several ways. It activates the Ras-MAPK and PI-3 NKT cascades, which team up to modify BAD, one of a class of proteins that are known to act as gatekeepers of the cell-death machinery. In its native form, BAD promotes cell death by binding to and suppressing pro-survival members of the Bcl-2 family. BAD can no longer kill developing cells, however, if signaling proteins in the Ras-MAPK and PI-3 NKT pathways phosphorylate it at two specific sites. In addition to this transcription-independent mechanism, Dr. Bonni found that BDNF also promotes cell survival through transcription-dependent means. The MAPK-Rsk path acts on CREB, a transcription factor known to promote cell survival. He hypothesized that while transcription-independent mechanisms might allow cerebellar granule neurons to survive shortly after they are generated on the outer surface of the developing brain, transcription-dependent mechanisms might act later on, as these neurons differentiate and mature in the brain's interior.
|
The
transcriptional response of epithelial cells to scatter factors
