Public Release: 

Pulling Strings To Build A Better Brain

Washington University School of Medicine

St. Louis, Jan. 23, 1997 -- A developing infant's brain may fold into a compact shape because its cells behave like rubber bands, says a neuroscientist at Washington University School of Medicine in St. Louis.

David C. Van Essen, Ph.D., Edison Professor and head of the Department of Anatomy and Neurobiology, presents his idea that mechanical tension might shape the brain in today's issue of the journal Nature. The journal publishes hypothesis papers about once a year to spotlight innovative concepts.

"There's a vast literature on how brain cells migrate and establish connections," Van Essen says. "But questions about how the brain attains its distinctive shape have received much less attention from neuroscientists."

The brains of all mammals have a thin outer sheet, the cerebral cortex, wrapped around an inner core of subcortical structures. The human cortex, involved in thinking and other cognitive tasks, is the size of a 16-inch pizza. It fits inside the skull because it is highly convoluted instead of smooth. So why, Van Essen wondered, is the cortex convoluted in some animals such as humans, apes and whales yet perfectly smooth in others such as mice and rats? And what transforms the smooth cortex of a human fetus to a structure that resembles crumpled newspaper by the time a baby is born?

"Previous speculations about cortical folding have tended to focus on differential growth and differential migration of cells -- an outward fold or bulge in the cortex was thought to be a place where there was extra cell growth," Van Essen says. "The underlying idea was that some neurons are programmed to proliferate more exuberantly than others."

Van Essen says the brain may not require a detailed genetic manual to reach its final shape even though cells in its various parts obviously do express different genes. Instead, he proposes that cortical folding relates to the pattern of long-distance connections between different parts of the cerebral cortex. These connections are made by axons -- long "wires" that enable nerve cells to communicate with one another. In many parts of the brain, large numbers of axons or other elongated cell parts run in the same direction, like strings on a guitar. If these neuronal strings are under tension, they would try to contract, just like a rubber band on a slingshot contracts when the tension is released. Indeed, Steve Heidemann and colleagues at the University of Michigan have shown that axons of isolated neurons growing in a dish do generate considerable tension.

As bundles of axons establish connections with other parts of the brain, they could pull the interconnected regions closer together, Van Essen suggests. So the intervening tissue would crumple into a fold, just as strings under tension would crumple a guitar neck that was pliable instead of rigid.

Van Essen, who has studied the brain's visual system for more than 20 years, developed his idea when thinking about the two largest components of visual cortex, areas V1 and V2, which are strongly interconnected. The cortex folds so that V1 and V2 are particularly close to one another in the mature brain, even though they are more widely separated early in development, when the cortex is flat. "Other investigators, particularly Christopher Cherniak at the University of Maryland, have suggested that different parts of the brain are placed so that total wiring length is as short as possible," Van Essen explains.

Given that notion, he realized that the proximity of areas V1 and V2 was a good example and that mechanical tension is an appealingly simple and efficient mechanism for attaining this outcome. "Later, I realized the hypothesis could be readily extended to the brain in its entirety. For example, the accordion-like folding of the cerebellum can be explained by its unique pattern of internal connections," he says.

So why does the cerebral cortex remain flat in some animals? "Species with small brains have a cerebral cortex that's just big enough to wrap around the subcortical structures," Van Essen explains. "In this case, the tendency to fold would be opposed by hydrostatic pressure from the internal structures, just as pressure inside a balloon prevents wrinkling. Folding occurs when the cortex grows sufficiently large that it wraps only loosely around these internal structures."

A clinical condition called hydrocephalus supports this explanation. When cerebrospinal fluid cannot escape from the brain through its normal exit, it enlarges the fluid-filled ventricles within the brain. This hydrostatic pressure forces the brain to expand, and the cortex fails to fold normally.

Another relevant observation comes from Pasko Rakic's group at Yale and colleagues in Japan. Through experiments on transgenic mice, the researchers interfered with the process that prunes away surplus cells in the developing brain. The cerebral cortex of the mice was wrinkled instead of smooth. The retina also was crumpled. "These observations suggest that extra neurons in the cortex and retina alter the balance of forces and allow folding to occur," Van Essen says.

Van Essen's hypothesis also can explain both the consistency and variability in the shapes of different people's brains. "Consistent folding should occur in regions where there are only a few major pathways that dictate the pattern," he says. "But in other regions, there may be a more evenly balanced competition among the hundreds of specific pathways that interconnect nearby cortical areas. So in one person, a slightly stronger pathway or slightly larger cortical area may tilt the balance toward one pattern of folding, whereas another individual with a slightly weaker pathway may progress toward a completely different folding pattern in that part of the cortex."

Van Essen notes that his idea can be tested with computer models as well as with a variety of experimental studies. "This hypothesis is attractive because it can explain the distinctive shapes of many different structures in the central nervous system," he says. "But it needs to be tested as critically as possible in a case-by-case situation."


Van Essen, DC. A tension-based theory of morphogenesis and compact wiring in the central nervous system. Nature 385, 313-318. 1997.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.