Scientists Strike a Cord

Courtesy of SR Eng BABY STAINS: The head of a transgenic murine embryo in which a marker enzyme has been specifically expressed in the sensory neurons of the trigeminal and dorsal root ganglia. The marker allows staining of the projections of these neurons into, among other areas, the hindbrain and spinal cord. (S.R. Eng et al., "Defects in sensory axon growth precede neuronal death in Brn3a-deficient mice," J Neurosci, 21:541-9, 2001.) Somewhere in the 200 million bases of the human genome is the information necessary to establish the basic wiring pattern of the nervous system. But just how that genetic information gets translated into specialized neurons, properly connected circuits, and the brain's vast information-processing capacity remains largely a black box. Scientists have been whittling away at the problem for decades, and some mechanisms that control neural specification are becoming clear. One reason: Scientists are studying the brain through the spinal cord. "We are all studying the spinal cord and the sensory ganglia .... [because] it is simpler," says Eric Turner, assistant professor at University of California, San Diego. "We can define the system and know what its normal connections look like at a much more complete level than we can know what the forebrain and its connections look like." While he ultimately wants to understand how brain wiring works, Turner says that he and many colleagues prefer studying the spinal cord system as it has a well-defined developmental pattern. "Significantly, many of the same proteins are active in both regions of the nervous system, so what is learned in the spinal cord will likely be transferable to the brain." Another explanation for the continuing clarity regarding neural specification is that scientists can now perform more subtle genetic and biochemical manipulations. "Particularly in the mouse system, the tools for looking at these developmental processes are getting ... clever," says Turner. Several years ago, researchers were removing genes in knockout experiments, but often the damage was so extensive that the mice died. Turner, who spoke about his work at the recent Society for Neuroscience (SFN) meeting, told the group that while a lethal phenotype shows the gene is important, it doesn't really reveal its function. HOW DOES IT DO THAT? Researchers are learning which transcription factors play a critical role in determining the identities of various neurons. For example, Sam Pfaff, assistant professor at The Salk Institute for Biological Sciences, told the meeting attendees that if a ventral motor neuron in the spinal cord expresses the transcription factor Islet1, it will become a visceral motor neuron, innervating the autonomic system that controls involuntary movement. On the other hand, if it expresses both Islet1 and Islet2, then it will become a somatic motor neuron, connecting to the skeletal muscle system and acting to control voluntary movement. What is unclear is how these and other nucleus-confined transcription factors give rise to functionally different neurons. The transcription factor never leaves the nucleus, but somehow it directs the cell to project axonal extensions. These neurons produce axons that head in distinct directions and innervate different muscle groups. While they know when and where these transcription factors are expressed, researchers now want to see what genes they activate, a question called the target gene problem. "A lot of effort has gone into just characterizing expression of the transcription factors, so that tells you they are there and gives you the tools to follow the [individual] cells," says Jane Johnson, University of Texas Southwestern. But researchers now want to use that information to find out how the transcription factors drive a neuron to act in one manner or another and build the specific networks that they do. For example, Turner wants to know what genes the transcription factor Brain3 controls downstream. Already his team knows that if they knock out this gene in mice, the neural anatomy remains about the same, because the gene is only expressed in postmitotic (no-longer-dividing) cells. Nonetheless, the animals die as neonates with severe neurological defects.1 However, knowing what killed the animals doesn't tell the researchers how the protein works. Turner's team initiated a microarray experiment to determine which genes change their expression patterns in response to the loss of Brain3. To do this, the group isolated RNA from trigeminal ganglia that were removed from either wild-type or Brain3 knockout murine embryos at 13.5 days after fertilization. (At this point in the development of these sensory neurons, Brain3 should have turned on or off whichever genes it controls.) The researchers then used microarrays to probe the two RNA populations to see which genes were differentially expressed in one genotype versus the other. To confirm that Brain3 regulates any putative hits gleaned from the microarray data, the team performed chromosome immunoprecipitation (ChIP) assays using antibodies against Brain3. If the microarray turns up an altered gene and the gene is really regulated by Brain3, then immunoprecipitating, broken-up chromatin pieces with Brain3 antibodies should bring down that same gene sequence. Thus far, out of 12,000 expressed sequence tags tested on the microarray, the team has identified several intriguing target genes. Although they are still analyzing the data, Turner presented preliminary results at the SFN meeting. Both calretinin and somatostatin appear to be upregulated along with several receptor genes, whereas galanin, advillin and persyn were all downregulated, as were some ion channel genes. These genes, Turner said, influence neural function somehow, indicating that the researchers really are getting at functional Brain3 target genes. Furthermore, it appears that Brain3 acts as an all-or-nothing switch for some genes because the relative increase or decrease in expression comparing the wild-type and mutant mice was between five- and 23-fold; many researchers consider microarray expression-level changes significant if the difference between two conditions is greater than a two-fold change. Courtesy of Samuel Pfaff Lateral image of a mouse embryo (head to left) showing GFP labeling of peripheral nerves as they navigate toward their targets. A QUESTION OF GRADATION Although some researchers are already studying forebrain development, the system's complexities are formidable. There are more cells; they migrate more during the course of development; the connections are more complex; and everything has to be considered in a three-dimensional pattern. "If we know 20% of what it takes to make a spinal cord," says Turner, "then they only know 2% of what it takes to make a forebrain." In other words, perturbing one gene and trying to interpret the results is akin to jumping into a landfill to locate a paper dropped two years ago; the information might be there, but finding it is a Herculean feat. Layered onto all of this genetic complexity is the issue of how much refinement or fine-tuning the system undergoes after the initial pattern is laid down; it's a question of gradation. Learning and memory acquisition require changes in wiring and synaptic connections, and those proceed throughout most of adult life. But even during some systems' initial organizations, extensive rewiring occurs in response to neural firing. Carla Shatz, Harvard Medical School, has shown that the visual system achieves its adult wiring pattern only if retinal neurons fire repeatedly during eye development, prior to the eye opening or even receiving any input signals.2,3 How much plasticity exists in the spinal cord and peripheral nervous system is a big question. In general, researchers think that the spinal cord layout is more completely directed by gene expression than that of the brain, but the question is open. Pfaff, who says that it's hard to know where to draw a line between what systems are hardwired and which require fine-tuning, is looking for the answer in spinal cord motor neurons. Considerable evidence exists, he says, showing that these neurons fire extensively during development prior to any stimulus, just like the retinal neurons do. The hypothesis suggests that, as in the eye, these firing neurons repeatedly strengthen their synaptic connections, and those that don't gradually lose their synaptic connections and die. Although Pfaff says that numerous pharmacological experiments have been done to block the firing, the results are unclear. His strategy is to engineer mice that have extra hyperpolarizing ion channels, which the researchers can switch on and then off again during development. Presumably, when the channels are expressed, they will block the normal depolarization that happens when a neuron fires, blocking the firing. If the random firing in the motor neurons is necessary for normal development, the mice with the blocked firing will have abnormal motor control, even when the extra channels are turned off. "This is still all on paper," Pfaff says, but if the experiment works, it could tell scientists whether such firing is necessary to build the normal wiring circuitry in the spinal cord--just like in the visual system in the brain. Such refinement probably wouldn't surprise researchers. Nobel laureate Eric Kandel, Columbia University, who was awarded for his work on signal transduction in the nervous system (Nobel in Medicine, 2000), says that the level of genetically determined hardwiring and fine-tuning exists on a continuum. To think, he says, that any part of the nervous system is completely at one end of the scale or another is to probably exaggerate the situation's simplicity. Rabiya S. Tuma (rabiya@nasw.org) is a freelance writer in Brooklyn, New York. References 1. S.R. Eng et al., "Defects in sensory axon growth precede neuronal death in Brn3a-deficient mice," J Neurosci, 21:541-9, 2001. 2. A.A. Penn et al., "Competition in retinogeniculate patterning driven by spontaneous activity," Science, 279:2108-12, 1998. 3. D. Stellwagen, C.J. Shatz, "An instructive role for retinal waves in the development of retinogeniculate connectivity," Neuron, 33:357-67, 2002.

Scientists Strike a Cord

Interested in reading more?

Become a Member of

Courtesy of SR Eng
BABY STAINS: The head of a transgenic murine embryo in which a marker enzyme has been specifically expressed in the sensory neurons of the trigeminal and dorsal root ganglia. The marker allows staining of the projections of these neurons into, among other areas, the hindbrain and spinal cord. (S.R. Eng et al., "Defects in sensory axon growth precede neuronal death in Brn3a-deficient mice," J Neurosci, 21:541-9, 2001.)