Genes Or Environment: What Shapes The Sensory Homunculus?

Boston, MA March 30, 2000 -- A single gene expressed in the brain can change a long-standing icon of basic neuroscience that was until now thought to be shaped mostly by neural input from the body's periphery, Harvard Medical School researchers have found.

The sensory homunculus is the familiar textbook caricature of a human being spread out improbably over the surface of the brain. It depicts how much space the brain allocates to our different body parts when it comes to feeling out the world around us. Its wildly uneven proportions are generally thought to arise in response to the activity of sensory neurons feeding information to the brain. In the April Nature Neuroscience, however, researchers led by John Flanagan, HMS associate professor of cell biology, report that a protein previously known to establish maps of the visual world in lower brain centers also plays a prominent role in imprinting a proper body map on the brain's cortex.

The work is intriguing because previous research in other areas has correlated representation in the brain with functional ability¯suggesting, in essence, that more important things get more brain space. "That we can modify this with one genetic change was totally unexpected," says Flanagan.

His group has not yet analyzed how the mice's distorted body map affects them, but he says that generally "it is not too much of a stretch to think that if you change the scale of representation in the cortex, you change behavior."

Most parents marvel at how their children differ in their interests. Where could those differences come from? Genetics partly addresses this question by trying to link genes to specific behaviors, such as a propensity for adventurousness. "But in terms of these maps, there is no previous evidence for a genetic basis to how they could be controlled," says Flanagan.

Long-established species differences in sensory maps already reflect differences in functional importance and, by extension, ability. In mice, for example, the whiskers and snout take up most of the sensory brain space, whereas monkey brains dedicate large areas to the hands and feet. Flanagan speculates that in humans genetic variations in how much space the brain devotes to each body part might help explain the difference between an average person and, say, a gifted musician. The current work does not negate the importance of neural activity by incoming neurons to determining brain maps. It says for the first time, however, that the cortex also has a hand in divvying up brain space, and that this influence is genetic.

Two 1995 papers by German researchers illustrate this joint influence of genetics and the brain's environment. Using magnetic resonance imaging, one study found differences in certain auditory brain areas between musicians with perfect pitch and nonmusicians. These differences, the authors write, develop around the 30th week of pregnancy, and might therefore be genetic. The other study showed that string players had a larger cortical representation of the digits on their left hand but not their right hand than did nonmusicians, and that this difference probably arose when the person began to play.

Pierre Vanderhaeghen, then a postdoc in Flanagan's lab, started this project to learn whether a family of mapping proteins called ephrins did in the cortex what they were known to do in the visual system and elsewhere in the brain. That question was controversial, in part because the developing cortex looks uniform. Indeed, scientists did not even know whether it contained mapmaking molecules, and if so, where. How the sensory homunculus develops is not well understood, and previous work in this area had focused on the incoming neural connections. It argued that densely innervated areas, including the lips and tongue, sent large cohorts of neurons into the brain, and that their intense activity enabled them to claim a large territory in cortical layer 4, which handles incoming sensory information. The cortex itself was thought to have only a small instructive role, if any.

Vanderhaeghen's experiments suggested there must be more to this. First he found that ephrin-A5 was expressed in the mouse's somatosensory cortex in a gradient¯high at the top of the head, low at the sides. Then he found a matching gradient of a receptor for ephrin-A5, EphA4, on incoming neurons from the thalamus, the last relay station of tactile information from the body's surface. Both gradients appear during the developmental stage when these thalamic neurons penetrate the cortex and make connections in layer 4.

Presumably, ephrin-A5 is one of several gradients that overlap to define points in the brain, much as an x-, y- and z-axis define points in a 3-D mathematical graph. The idea is that incoming neurons negotiate this crossfire of labels with the combination of receptors they carry. In the end, each neuron finds its proper spot in a spatial pattern that reflects the outside world.

The molecular mechanism by which this occurs is poorly understood. Yet researchers do know that the ephrins act by repulsion¯that is, neurons studded with many EphA4 receptors avoid areas of dense ephrin-A5 ligand.

Next Vanderhaeghen measured the precise area devoted, in layer 4, to the mouse's whiskers, its most important tactile organ. He found that mice lacking ephrin-A5 had strangely distorted whisker fields, giving some whiskers expanded space while squishing together others.

The overall effect resembles differences in world maps drawn up according to criteria such as oil resources: countries' neighbor-neighbor relationships remain accurate but suddenly Norway looks oddly large next to little Sweden and Finland. Similarly, the mice's body map is basically intact but shifted in scale (see image). That was surprising, because in the visual system ephrins are known to help incoming neurons faithfully maintain these neighbor-neighbor relationships as they make connections in the brain. Without ephrin labels, their spatial order disintegrates, and the neurons grow into the wrong areas altogether. Lastly, lest someone feel misrepresented by the homunculus, Flanagan adds that it is not about mere sensitivity, but ability to resolve tactile information. We do feel pain when pinched in the ribs or lower leg, but would not use these parts to read Braille.