 |
Development and Function of the Nervous System: From Stem Cells to Circuits for Innate Behaviors
Summary: David Anderson is studying the development of the embryonic nervous system, with a focus on neural stem cells. He is also using molecular genetic techniques to map and probe neural circuits that underlie innate behaviors, such as avoidance of aversive stimuli, in both mice and fruit flies.
Stem cells are primitive, undifferentiated cells that have the capacity both to reproduce themselves (self-renew) and to differentiate into specialized cell types, such as neurons or muscle cells. A fundamental problem in neural development is to understand how the stem cells of the nervous system produce all the different types of cells composing the adult brain. These include different subtypes of neurons and of nonneuronal cells called glia. Our approach is to isolate neural stem and progenitor cells, characterize their developmental capacities, and identify some of the molecules and genes that control their differentiation from outside and inside the cell. Our principal accomplishments in recent years have involved (1) isolating neural stem cells from the central nervous system and testing their developmental capacities by direct in vivo transplantation, and (2) identifying "master genes" that control neuronal and glial differentiation and analyzing the functions of these genes.
Testing the Neural Stem Cell Hypothesis In Vivo
Central nervous system stem cells (CNS-SCs) have been defined by their ability to perpetually produce progeny that can generate both neurons and glia, over many generations. Such self-renewing, multipotent CNS-SCs have been defined primarily in petri dishes—in vitro. Do cells with such properties actually exist in the embryonic nervous system—in vivo? In collaboration with Thomas Jessell (HHMI, Columbia University College of Physicians and Surgeons), we have used novel genetic markers to isolate candidate stem cells from mouse embryonic spinal cord and have tested their multipotency and self-renewal by transplanting them directly into host chick embryos. During normal development, embryonic neural progenitor cells first generate neurons and then switch to producing glia. If these cells are true stem cells, they should retain the capacity to generate neurons throughout the subsequent gliogenic phase. Surprisingly, we find that by the time the cells have begun to generate glia, they have lost the capacity to generate neurons. This observation argues that in the embryonic spinal cord, progenitor cells do not retain their multipotency—i.e., the capacity to generate both neurons and glia—throughout development, but rather become gradually restricted in their potential. This suggests that they are not stem cells in the strict sense. Although stem cells undoubtedly exist in a few specific regions of the CNS, such as the hippocampus and olfactory bulb, these cases may be the exception rather than the rule.
Control of the Neuron-Glial Switch
How do neural progenitor cells make the switch from generating neurons to producing glia? By systematically profiling the gene expression changes that occur in a specific population of neural progenitors, as they transition from neurogenesis to gliogenesis, we identified a family of genes, called NFI genes, that are turned on just before the cells begin to produce glia. Through a series of genetic experiments, we established that one member of this family, NFIA, plays a pivotal role in initiating the switch to gliogenesis. Unexpectedly, NFIA not only coaxes progenitor cells into a glial pathway, but also helps to prevent the cells from continuing to generate neurons. An understanding of how NFIA initiates the switch to gliogenesis may suggest ways of artificially reversing the switch. If this could be accomplished, it might be possible one day to use adult glial cells as a source of new neurons for brain repair.
Neural Circuits Involved in Pleasure and Pain
The skin, the body's largest sensory organ, detects both pleasant and painful stimuli. The detection of these stimuli is mediated by sensory neurons, whose fibers innervate the epidermis. Although a great deal has been learned about the neurons that sense painful heat and pressure, much less is known about the neurons that detect pleasant stimuli. We recently discovered a large family of receptors, called Mrgs, that are expressed by specific subsets of sensory neurons. By generating mice in which the neurons that express different Mrgs are selectively labeled with visible tracer molecules, we found that the MrgD gene is expressed by a subset of neurons that exclusively innervate the skin and that likely detect painful thermal and mechanical stimuli. We have applied a similar approach to mapping the neurons that express a different member of the family, MrgB4. These neurons also exclusively innervate the skin, but the anatomical distribution of their nerve endings is remarkably similar to the functional map, in humans, of neurons that detect pleasant, gentle touch. These observations suggest that MrgB4- and MrgD-expressing neurons may sense pleasurable and painful stimuli, respectively, raising the question of how these sensations are distinguished by the brain.
Neural Circuits for Innate Behaviors in Mice and Flies
We are developing molecular genetic tools to trace and functionally manipulate neural circuits in the central nervous system, as well as in the peripheral nervous system (as described in the preceding section). A recent study has focused on the identification of molecular markers for pathways that mediate innate reproductive or defensive behaviors in mice. We discovered that different members of the LIM homeodomain transcription factor family mark the reproductive versus the defensive "branches" of the amygdala-hypothalamic pathway. Using both genetically encoded and classical neuroanatomical axonal tracers, we have found that amygdala neurons activated by reproductive (female urine) or defensive (cat odor) olfactory stimuli converge in a specific region, or nucleus, of the hypothalamus. These two types of projections have opposite "signs" (excitatory versus inhibitory), suggesting that this point of convergence may serve as a "gate" that interrupts reproductive behaviors if threatening stimuli are present. We are developing and testing new methods for inhibiting the firing of these neurons, in order to test their function in reproductive and defensive behaviors.
In parallel with these studies in mice, we have carried out related experiments aimed at mapping neural circuits for innate behaviors in the fruit fly, Drosophila melanogaster. In collaboration with Seymour Benzer (California Institute of Technology), we have shown that traumatized (shaken or shocked) flies emit an odor that repels other flies; we have termed this odor "Drosophila stress odorant" (dSO). Using chemical analysis, we found that one component of dSO is carbon dioxide (CO2), and that flies are strongly repelled by this simple compound. In collaboration with Richard Axel (HHMI, Columbia University), we have mapped the neural circuitry mediating CO2 avoidance. Axel has shown that, in contrast to other odorants, which typically activate multiple classes of olfactory sensory neurons (OSNs), CO2 activates a single class of OSNs. We showed that functional inhibition of these neurons, using genetically encoded inhibitors of nerve transmission, is sufficient to block behavioral avoidance of CO2 in freely moving flies. These studies have assigned a robust behavioral response to a single class of OSNs activated by a known odorant, and provide a starting point for further mapping of the neural circuitry that underlies this innate behavior. More recently, we have identified neurons that respond to a second component of the stress odorant, and we are endeavoring to identify the chemical nature of this component.
Grants from the the National Institutes of Health and the Pritzker Foundation supported portions of this work.
Last updated: January 29, 2007
|
|
|
|
