Neuronal diversity can lead to efficiency, new research shows

Neurons have been stained with a dye so they can be viewed. Groups of neurons are hypothesized to be diversified, so that redundancy is eliminated and efficiency throughout the brain is increased. (credit: Courtesy of Wikimedia Commons) Neurons have been stained with a dye so they can be viewed. Groups of neurons are hypothesized to be diversified, so that redundancy is eliminated and efficiency throughout the brain is increased. (credit: Courtesy of Wikimedia Commons)

For many years, neuroscientists have believed that the characteristics differentiating neurons from one another were just biological flaws, but this concept may actually be false. Nathan N. Urban, the head of the department of biological sciences at Carnegie Mellon, has focused his research on neuronal processes within the olfactory system in animal models — specifically mice. Urban’s current research involves testing whether or not diversity within neuron networks is significant to complex brain computations.

Urban completed his undergraduate degree at the University of Pittsburgh in neuroscience, mathematics, and philosophy. Upon graduation, he received a Rhodes Scholarship and spent two years at the University of Oxford studying math and philosophy. Urban then completed his Ph.D. in neuroscience at Pitt and a postdoctoral appointment at the Max Planck Institute, a scientific research-based society in Germany, for three and a half years. He has been with the Mellon College of Science faculty since January 2002.

Urban and his team, which includes postdoctoral student Krishnan Padmanabhan, published a paper about neuronal diversity, titled “Intrinsic biophysical diversity decorrelates neuronal firing while increasing information content,” in Nature Neuroscience last August. The team found that groups of diverse neurons were able to transmit twice as much information as groups of relatively similar neurons.

The research project, funded by the National Institutes of Health, began when Padmanabhan observed slight differences between the responses of seemingly identical neurons to electrical stimuli. When groups of identical neurons were subjected to a constant computer-generated electrical current, Padmanabhan found that no two responses were the same. This prompted the question: Would the brain be better if all neurons were identical, as the team originally thought they were, akin to parts in an assembly line, or is there a natural advantage to neuronal diversity?

When a mouse’s brain first detects a smell, receptors in the olfactory system of the animal send a signal to an initial group of neurons in the brain. These 10 million neurons then pass on the signal to a second group of neurons, one synapse deeper in the brain’s neural network. The second group of neurons, which only number 50,000, are the focus of Urban’s experiments. Urban is already hypothesizing how the brain compensates for this decrease in connections. “You are going to have ... all this information that is available in these 10 million neurons, and now you are going to reduce it to 50,000 neurons. In order to do that well, you have to be really efficient and make sure there is not any redundancy or overlap in terms of what those neurons are doing. There is no room for redundancy when you have this huge reduction.”

To perform experiments, Urban and his team manipulated glass capillary tubing to a fine point and ruptured a neuronal membrane in a specific area of a single neuron. Using a method called whole-cell recording, Urban used computer equipment to generate electrical currents of varying wavelengths to stimulate the cell. A response from a neuron was categorized as a reaction from the neuron’s ion channels. Ion channels allow molecules to move in and out of the cell, so differences in the amount of movement between different neurons can be recorded. According to Padmanabhan and Urban’s paper, “In these regions and populations of neurons, differences in the expression of ion channels and morphology result in the marked heterogeneity of the intrinsic properties of [inhibitory and excitatory] cells, and the responses are therefore diverse even when similar inputs are delivered.” In short, identical electric currents passed through neurons thought to be identical resulted in different reactions within each neuron’s ion channels, suggesting neural diversity.

“We think we have some evidence of a particular ion channel, which is one of the reasons why we see the heterogeneity in this particular population,” Urban said. “Some neurons express more of this channel; some express less of this channel — that’s an important component in why they may be responding differently to these stimuli.”

Urban also saw plans for future experimentation: “The ultimate prediction for us would be if this diversity was helpful, we would like to find ways of testing that by either making different populations of neurons more homogeneous or more heterogeneous. If I can make populations more homogeneous, then that would have a behavioral effect, and animals would be less able to perform some task that depended on that group of neurons.” For Urban, understanding the key differences in these neuron ion channels would allow him to manipulate different mechanisms to find out more information.

One question Urban and his team are trying to solve is how homogeneous groups of neurons affect memory. If neuronal connections were to degrade over time, would this affect your mental abilities or just your mental capacity? Heterogeneous neurons, if damaged or severed, could have different consequences on memory than homogeneous ones, which could help scientists understand how “robust” the brain is, or how it deals with different types of injuries.

For Urban, this new research is just the beginning to uncovering the intricacies of neuronal processes. His ideal five-year plan would include developing a better understanding of mechanistic properties of neurons and what chemical properties or reactions distinguishes one neuron from the next. Although this research is still in early stages, information garnered can shed light on brain disorders such as Alzheimer’s disease and Parksinson’s disease.

“At this point, we just want to find out mechanistic things,” Urban said.