Neurons Seek Their Own Solution
Computer models find that various ion channel arrangements can produce the same firing pattern
Each cell in our nervous system is an instrument in a complex symphony of electrophysiologic communication. A neuron’s signaling abilities arise from its array of ion channels—tunnels within the cell’s membrane that act as gatekeepers of electrical charge. But how does a cell determine the types of channels it needs and where in its membrane they should sit? The results of a new computer model suggest that even with markedly different patterns of ion channels, neurons still can come to play the same tune.
The work supports a growing paradigm shift in neurophysiology, says Erik De Schutter, MD, PhD, professor of neurobiology at the University of Antwerp, Belgium, lead author of the study. “We used to think of [ion channels] as LEGO blocks,” he explains—with a predetermined number, type and position regulating how the neuron fires.
More recently, physiological experiments have suggested that cells with wildly different ion channel compositions could have similar firing patterns. But researchers have consistently attributed such variability to experimental error, clinging to the notion of a “platonic ideal” of a neuron with an unchangeable firing pattern.
To look at the problem more closely, De Schutter and Pablo Achard, PhD, a postdoc in his lab, computationally modeled the Purkinje cell, an especially complex type of neuron best known for forming as many as 200,000 synapses. The team’s model specified 10 types of ion channels and broke the cell into four different regions that the channels could occupy. Using a mathematical approach called the phase-plane method, the model neurons were permitted to evolve their ion channel densities to produce all four firing patterns that Purkinje cells display. To the researchers’ surprise, about 20 possible combinations of ion-channel densities fit the bill. The results were published in the July 2006 issue of PloS Computational Biology.
The work suggests, De Schutter says, that neurons are preprogrammed for a particular type of firing pattern. Each cell then decides locally how to distribute ion channels to achieve its signaling goal. How well the model predicts real biological properties is not clear and is difficult to test experimentally. Neurophysiologists can record impulses from just a few neurons at a time; the sheer quantity of recordings needed for a thorough comparison with the model doesn’t yet exist. “My estimate is you’d need about two people working full time for a year for that data,” De Schutter says.
Regardless, the model provides a more nuanced version of how physiologists should conceive of their cells, says Eve Marder, PhD, a professor of biology at Brandeis University who studies both physiology and modeling in neurons. “The assumption has been that there was a single solution and the variance was your fault,” Marder says. “This paper is a beautiful example showing that we shouldn’t be thinking about a single solution to capture what a neuron is doing, but a family of solutions.”