26 OCTOBER
2006
ATLANTA - You can learn a lot about learning by learning how worms learn.
In particular, the tiny roundworm known as C. elegans has taught neuroscientists some basic lessons about how nerve cells form networks that control behavior and how that behavior can be modified by learning.
Just ask neurobiologist Cornelia
Bargmann of Rockefeller University, who studies the little worm's sense of
smell. Her work shows how a spineless creature a mere millimeter long possesses
sniffing
skills that can help explain the power of the human brain.
It's simply a question of teaching old genes new tricks. One of the most important legacies of evolution (and one of the strongest signs of its validity) is the shared genetic repertoire of nearly all the Earth's organisms. Human DNA, the molecule encoding all the body's genetic instructions, comprises something like 30,000 genes -- the blueprints for individual biological molecules. Nearly three-fourths of those genes, Bargmann points out, are also found in invertebrates, with only minor variations from their human forms. Worms like C. elegans, with nearly 20,000 genes, can therefore exhibit genetic systems for performing certain tasks that are similar to the systems found in people.
The
millimeter-long roundworm C. elegans. Photo
by Bill Love,NIH/National
Center for Research Resources
At a recent meeting of the Society for Neuroscience, Bargmann described a series of experiments revealing the intricate abilities of worms to seek food and avoid poison. For a worm, finding food is essential to surviving, as is avoiding toxic substances that merely masquerade as food. So the worm's brain has developed a rather sophisticated strategy for using its sense of smell to guide its movements toward the food and away from the poison.
In its quest for food, a worm moves in a pattern not unlike a walking drunk, changing directions without obvious rhyme or reason. In fact, though, the worm's behavior responds to its environment. As it senses higher levels of good (food) odors, it mostly slithers along in the same direction, making turns infrequently. But if bad odors seem to be increasing, the worm turns more often. The result is an apparently random path that actually gets the worm closer and closer to its dinner.
Such movements are guided by the worm equivalent of muscle cells, which are in turn activated by a network of nerve cells (or neurons). Odor input feeds into this network of neurons, which apparently adjusts the odds of making turns or not. "We think that this network might be generating the probabilities of different behaviors," says Bargmann.
Compared to the human brain, the worm's nerve network is pretty simple. C. elegans possesses a grand total of 302 nerve cells; the human brain has billions and billions. The worm's nerve cells are connected by 7,000 synapses; humans have trillions and trillions. Nevertheless the worm's simplicity exhibits features that underlie the human brain's complexity.
Image from Wikipedia,
photo by Zeynep F. Altun
Even in the worm's simple nervous system, chemical sensing and signaling networks can get pretty intricate. Molecules protruding from a worm's olfactory neuron sense the presence of odor molecules, activating that neuron to send a signal to another neuron in the network. That intermediate neuron them relays a signal to a neuron that tells worm muscle cells whether to move or not. Some neurons send signals that encourage the worm to make its turn; some send signals that inhibit the muscles from activating motion, depending on the whether the odor is absent or present. It's the combination of such stop and go signals that determines the likelihood that a worm will make a turn our not.
It turns out similar kinds of networks operate in higher life forms too, and for different tasks. A very similar system, using some of the same molecular sensors, operates in the vision system of mammals. In vision, the absence or presence of light determines whether a neuron fires a signal, just as some worm neurons fire in the presence of a particular odor signal and some fire in its absence. Thus the worm's odor sensing system is in effect a genetically based module for processing information, adopted by higher organisms for processing information other than odors.
Further experiments by Bargmann and colleagues show that the worm's neural circuits can be modified to respond differently to a given odor. Worms can be taught, for instance to prefer the odor of a particular food (tasty bacteria) to the odor of pathogenic bacteria.
Worms can learn this behavior, though, only if their nerve cells are producing proper levels of the brain chemical serotonin. Serotonin is one of the most important of human brain chemicals -- among many other things, it's a messenger molecule for communication between the gut and the brain. Its relevance for eating and learning no doubt descends from its importance to C. elegans.
In fact, the same sorts of learning systems seem to be used throughout the animal world. Intrinsic nerve cell circuits respond to sensory inputs by changing the odds for different actions. Those odds for different actions can then be modified by experience, and that's what learning is all about, whether in worms, or humans.
"We think that basic learning rules," says Bargmann, "may really apply in all different animals."
Tom Siegfried's latest book, A Beautiful Math, has just been published by the Joseph Henry Press.
E-mail: tsiegfried@nasw.org
