McGowan Institute for Regenerative Medicine
faculty member G. Bard Ermentrout, PhD, University of Pittsburgh Professor of Mathematics and Adjunct Professor of Neurobiology, worked with University of California, Berkeley, scientists on computer models that retrace how mollusks build dazzling shells. The work illustrates how memory and sensory input influence action.
Dr. Ermentrout, and lead authors, graduate student Alistair Boettiger and Dr. George Oster of Berkeley, modeled the neural network of mollusks and designed a computer program that can generate the complex patterns and shapes of most mollusk shells. The researchers traced the trail of brain activity that begins with a mollusk’s tongue-like organ called a mantle and leads to the cells that produce the shell and pigmentation.
The team supposed that as mollusks build their shells, they retrace their previous work with the mantle and use those “memories” to continue the pattern. At the same time, the new pigment and shell growth are influenced by external factors that result in the varied patterns and shell structures.
Mr. Boettiger and Drs. Ermentrout and Oster simulated the neural network with integral equations that retrace the previous pattern but can be manipulated to accurately predict how a shell will form under specific conditions. The resulting models help illustrate how neural networks — including mammalian cortices — function in response to a combination of sensory information and experience.
“These models demonstrate the combined influence of sensory input and memories on brain activity,” Dr. Ermentrout said. “Brains convert sensory information into action. If a ball is thrown at you, you duck or catch it because you know that the ball could hit you. That knowledge and the sight of the ball coming at you dictate your action. A mollusk collects sensory information from its previous pigmentation and converts it into motor action by producing more pigmentation and continuing the pattern.”
To construct their model, the team studied electron microscope images of mollusk mantles to understand the neurons that connect the mantle’s sensing cells with the cells that secrete calcium carbonate and pigmented proteins. The team found that the excitatory and inhibitory synapses, which promote or diminish cell activity, surrounding the secretory cells and the cells’ firing thresholds act as a neural network that determines how much calcium and pigment the mollusk secretes. Different rates of calcium carbonate secretion determine the shape of the shell, while different amounts of pigment result in a pattern unique to each species.
For instance, shell ridges result from one cell increasing calcium carbonate secretion while depressing secretion from surrounding cells. With striped shells, a pigment-secreting cell inhibits secretion of pigment by neighboring cells but not itself, so that the same pattern is repeated day after day, yielding a stripe. Bands parallel to the growing edge form when pigment secreted one day inhibits secreting cells for a few days, resulting in an on/off pattern.
“Traveling wave” patterns of diamonds, zigzags, arrowheads, and other shapes come about when a pigment inhibits future secretion at that site but excites secretion in surrounding cells, so that pigment moves laterally on successive days like a wave.
Illustration: The researchers' computer model can reproduce a wide variety of shell shapes, colors and patterns, including Conus vicweei. In the photo, the real shell is at left and the computer-generated image is at right. (Alistair Boettiger/UC Berkeley)
University of Berkeley Press Release (w/video) (04/01/09)
Science Daily (04/02/09)
Biology News Net (04/02/09)
University Times (04/16/09)
Bio: Dr. Bard Ermentrout
Abstract (Proceedings of the National Academy of the Sciences of the United States of America. Published online 04/07/09.)