IT MAY seem corny to compare the slithering growth of nerve fibres in the developing human embryo to traffic moving along roads and freeways, but the analogy is apt. Vehicles throng our roads, heading towards their very different destinations through untold acres of streets and buildings, aided by a plethora of signs.
Nerve fibres, too, have some serious navigating to do if a walking, talking, living child is ever to emerge from a fertilised egg. Should a fibre grow up towards the head, or down towards the feet? Should it snake its way towards back or belly? Should it cling to other fibres in a bundle, like a sticky mass of over-done spaghetti, or forge a fresh path to that muscle just south of the spleen? One trillion nerve cells, from motor neuron to sensory neuron, purkinje cell to retinal ganglion cell, must grow right way and link in the right way if the nervous system is to be more than a senseless scramble.
It is a puzzle to challenge the keenest of minds. Consider that the nerve cells of a human embryo have a cool quadrillion connections to make before a basic baby can take shape. And there isn’t much room for sloppiness: linking a fibre (or “axon”) slightly to the right or the left of its intended target could be disastrous. What makes the wiring of the nervous system so precise?
Advertisement
For decades, neuroscientists have been trying to pinpoint the mysterious, microscopic signposts that they assumed must be guiding the tips of growing nerve fibres.
Now, at last, their work is bearing fruit. It’s fast becoming clear that the signposts are actually specialised proteins made by embryos – and these proteins, along with the genes that encode them, are hot news in neuroscience. Every month, it seems, new reports appear in the pages of journals like Cell and Neuron. The science of “neuronal pathfinding”, as it’s known, has finally left the B roads behind and is racing gleefully down the Autobahn.
What’s emerging is a fascinating picture of nerve fibres snaking their way through embryos guided by an army of proteins with names to rival anything in particle physics – semaphorin, netrin, collapsin. There are proteins that drizzle out of tissues to coax certain nerve fibres towards their targets, at the same time sending other nerves fleeing in the opposite direction. Other proteins stick tight to the surfaces of cells as they repel or beckon growing nerve fibres. Not only that, but the same kinds of protein are at work in all animal embryos studied so far, from roundworms and fruit flies to humans.
In short, it’s now clear that a scientist studying the wiring of nerves in a worm or a fly is not simply working on a worm or a fly but on a fundamental biochemistry that is crucial to putting any nervous system together. And this means that from here on out pathfinding research can proceed at breakneck speed, since each animal has its own strengths and weaknesses in molecular studies, and all are vastly easier to work on than humans. This is good news, and not just for the scientists who simply want to solve the puzzle. Zooming in on the molecules that wire up embryos could give a sorely needed boost to the hunt for drugs to repair damaged nerves in adults. Not only that, but many diseases could have their roots in some basic wiring defect. Some of these disorders are already known: the genetic condition called Kallmann’s syndrome is a classic example. Fragrant flowers and Paris perfumes are lost on people with Kallmann’s syndrome because the nerves from their nose failed to hook up to the smell-processing centres in their brain during fetal development. A key “pathfinding” molecule appears to be missing.
Faulty wiring could also be at the root of a rare type of mental retardation called “hydrocephalus as a result of stenosis of the Aqueduct of Sylvius” (mercifully shortened to HSAS). People with HSAS carry a mutation in a gene encoding a protein called L1, which is one of a huge class of proteins that seem to help growing nerves stick to each other. Without the help of the chemical signal normally given by L1, certain nerves seem unable to migrate correctly during development, and the result is an adult brain conspicuously bereft of key structures like the corpus callosum, the bundle of nerve fibres that links the left and right sides of the brain.
Brainworks
There’s even a tentative suggestion that schizophrenics may have congenital misconnections in their brains. And more examples will surely emerge. After all, the making of proper nerve connections, or synapses, is probably central to almost anything the brain cares to do. From the fetus until we die, our synapses are constantly being reworked by life’s experiences – strengthened or weakened, abandoned by one network of neurons, adopted by another. “Any neuroscientist will tend to tell you that their stuff is the most important thing in the world,” says Carla Shatz, a neuroscientist at the University of California at Berkeley. “But this research really is at the core of how the brain works and therefore how things might go wrong.”
For the nervous system to develop in the right way, the correct chemical cues must be in place. Some of the first signs that these chemicals exist came from watching how nerves grow in the lab. Take, for instance, this experiment with the spinal cord of a rat or a chick embryo. First, you slice off a piece of the lowermost portion of the cord – the so-called “floor plate”. Next, you put it in a dish with a blob of tissue from the upper spinal cord. In a matter of hours, that blob will turn into something resembling a sea urchin.
But those tiny, delicate spines are actually nerve fibres growing from certain nerve cells (called commissural neurons) inside the blob. What they’re sensing is something seeping out of the floor plate – something they eagerly grow towards. This makes sense since, in the embryo, commissural neurons must send axons snaking down to the floor plate as the first step in their wiring program. Eventually, when that wiring is complete, they will extend all the way to the brain, and relay messages from the skin about pain, heat, and touch.
Of course, watching nerve fibres being lured this way or that in a lab dish is one thing; identifying the source of the attraction is quite another. The chemical cue pumped out by the floor plate was a mystery until only last year, when it was purified by Marc Tessier-Lavigne’s team at the University of California at San Francisco. This was no mean feat. The researchers had to grind up the brains of no less than 20 000 chick embryos to extract a mere 25 micrograms of protein.
In the end, the team fished out two very similar proteins. What to name them? One firm favourite was “hitherin (as in “come hither”). But turn out to the proteins might repel certain nerve fibres, just as they attracted the commissural ones. So the scientists plumped for a name that implied signalling but didn’t specify a direction: “netrin”, after the Sanskrit word “netr”, meaning “one who guides”.
Signposts
It was a shrewd decision. In the May issue of Cell, Tessier-Lavigne’s team reported that one of the two proteins (netrin-1) does indeed repel nerve fibres as well as attract them. Everything depends on what sort of nerve fibre you are, and where you want to end up. A Los Angeles motorist headed north towards California’s state capital, Sacramento, would do well to follow the “SACRAMENTO” signs. But if his goal is buckets of fun at Disneyland to the south, a “SACRAMENTO” sign should be repellant in the extreme. Similarly, the floor plate may be a great place for a commissural neuron to grow towards, but a trochlear motor neuron has no business going there. Its job is to make its way to the other side of the spinal cord, before shooting off to find the muscles it must link to. And, sure enough, when tested in the lab the trochlear nerve fibre avoids netrin-1 like the plague.
All along the spinal cord, from top of the brain to tip of the tail, the netrins are likely to be enticing some nerve fibres, and rebuffing others. In recent issues of Neuron, for instance, neuroscientists Adrian Pini and Sarah Guthrie in England, as well as Fujio Murakami in Japan, have shown that the floor plate can repel or attract all manner of developing nerves.
But even before this recent blitz of findings, Tessier-Lavigne had a hunch netrins might be strongly “AC/DC” when it comes to pathfinding. That’s because of studies on a remarkably similar protein in the roundworm, Caenorabditis elegans, by two scientists in North America, Ed Hedgecock and Joseph Culotti. That protein, UNC-6, seems to control nerves growing in two opposite directions in the worm – as if it were repelling one group and attracting the other.
Why should nature use one molecule for two different jobs, repulsion and attraction, again and again in the embryo? Economy may be part of the answer. Given the sheer number of nerves in the body, and the sheer number of twists and turns they must make en route to their targets, it’s a wonder that any genome can hold enough genes to supply all its needs. Researchers have wrestled with this problem down the decades.
In the 1960s, for instance, neuroscientist Roger Sperry suggested that every nerve might carry its own unique identity tag. Such tags would ensure that each nerve hooked up to its rightful partners. But as scientists upped their estimates of the number of synapses formed inside a brain, it became clear that this hypothesis couldn’t work. Even Sperry doubted that the genome could code enough “Sperryons,” as some scientists like to call them.
Then came the 1970s, and the pendulum swung sharply in the opposite direction. Now scientists came to believe that developing nerves might begin by wiring up in a promiscuous free-for-all, only fighting it out later on. Those that unleashed electrical signals at the right place and time would remain connected to a particular nerve; those that misfired would be drummed out of the synapse. This way, the brain would only keep the useful synapses.
Electric tuning
Certainly, it’s becoming clear that electrical activity plays a crucial role in the wiring of nerves. Take, for instance, the visual system in mammals. Shatz and others have found that the nerves carrying signals from our eyes need doses of electrical activity to finetune their connections to the vision-processing centres in our brains. Without that finetuning, which takes place early in life, we simply couldn’t see. A baby born with cataracts will never have good vision if an operation is postponed too long. And the visual system is only one example. Any place, says Shatz, where super-precise contacts must be forged, electrical activity could be crucial.
But even in the visual system, nerve connections depend on more than just electrical activity. The fact that fibres sprouting out of the retina know where to grow suggests that chemical signposts must also be operating. And so might it be for much of the nervous system. “In biology, we’ve learned that if you have two opposite extremes of mechanism, it usually winds up that evolution has used both to build up the richness,” reflects neuroscientist Corey Goodman, also at Berkeley.
Goodman’s own hard toil at the bench began when the “activity is everything” theory was in vogue in the 1970s. But as he and others clocked case after case of nerve fibres heading off in the right direction, it became clear that nerves really do know where they’re going. Experimenting with invertebrates, Goodman’s group set out to find the signposts – and in 1992, they found the first of the “semaphorins”, in a grasshopper. Today, whole families of semaphorin genes are turning up everywhere – in fly DNA, in chicks, in mice, in humans. And all the signs indicate that they encode proteins which repel certain nerve fibres as they snake their way through embryos.
In fact, it was the very repulsiveness of semaphorins that led Jonathan Raper, a neuroscientist at the University of Pennsylvania, to discover them in chicks. He would grind up some chick brains and then squirt some of the resulting extract onto certain developing nerves in a petri dish. The tips of those nerves would shrink back. And so, when Raper found his repulsive protein, he called it collapsin, although it’s now clear that it belongs to the wider class of semaphorins.
Earlier this year, Raper’s team showed that when growing nerves bump into tiny beads coated with collapsin, they quickly veer away. In the May issue of Neuron, meanwhile, Goodman, Shatz and Tessier-Lavigne joined forces to describe for the first time the behaviour of a mammalian semaphorin. The protein drives certain sensory nerve fibres away from the “belly” side of the spinal cord, thereby ensuring that they only make connections with nerves in the back of the spinal cord. In embryos, so the thinking goes, different nervous tissues probably produce different semaphorins, each repelling different sets of nerve fibres. And it wouldn’t surprise Goodman if semaphorins turn out to attract some nerves, just as the netrins do.
In search of psycho-path
But semaphorins and netrins are only one part of the picture. Today, research into pathfinding molecules seems set to explode, with other candidates for guidance proteins piling up by the month. There are fifty-odd “cell adhesion molecules” with names like connectin, myelin-associated glycoprotein, laminin, tenascin, janusin. The list gets longer by the day. So, too, does the list of mutant animals with faulty wiring, which will probably lead to even more pathfinding molecules. Goodman’s lab, for instance, has droves of mutant fruit flies with names like “beaten path”, “stranded”, “clueless” and “walkabout”, all sorry victims, it seems, of pathfinding defects in the nervous system. David Van Vactor, a researcher with Goodman who found many of these flies, say he hopes to find at least two more mutants, if only so he can name them “psycho-path” and “path-illogical”.
In the end, researchers believe that a host of cooperating pathfinder molecules will guide each fibre to its proper target, pushing and pulling in all directions, telling a nerve when to turn, when to stall, when to put out branches and when to fashion its final synapses. So far, scientists know next to nothing about the “receptor” molecules that enable the tips of nerve fibres to respond to those chemical signposts.
And then what happens next, inside the cell, once a receptor interlocks with its guidance cue? Somehow the nerve fibre must push its membrane out in the direction it wants to steer, and yank it in at other places. Scientists don’t know how the dense, protein meshes that give the tip of a nerve fibre its structure are altered. Nor do they know if there will, turn out to be key differences between neuronal pathfinding in lowly animals like the worm and in vertebrates like us.
In fact, despite some key breakthroughs, the general feeling among neuroscientists is that their work has just begun, that there is a long road ahead. “There is no single part of any organism, whether it’s a simple or a complex one, where we really understand who the molecular players are, and all the mechanisms going on – where we can predict, based on those molecules and mechanisms what the pattern of connections will be,” says Goodman. Only when that is possible will “the case of the wired-up nervous system” be finally solved.