Skip to Navigation Skip to Page Content

From left, Ashwin Mohan, Sandeep Pendyam and Guoshi Li, all doctoral students working in Professor Satish Nair's Computational Neurobiology lab. Li has received post-doc offers from Cornell, the University of California-Irvine and Rutgers and will be going to Cornell. Photo by L.G. Patterson

Electrochemical communications in the human brain, responsible for all mental and physical functions, occur at synaptic connections between brain cells, also known as neurons. Upward of 30 thousand of a human’s 100 billion neurons will fit on the head of a pin, and each one of those cells may potentially make as many as 200 thousand synaptic connections.

Very little brainpower is required to look at those numbers and realize why the study of the brain, known as neurobiology, poses such a challenge.
New light is being shed on the work undertaken by neuroscientists through collaborations with electrical engineers such as Satish Nair, professor of electrical and computer engineering at the University of Missouri, and his Computational Neurobiology Group. Two such collaborations have proven so fruitful that three of Nair’s doctoral students have been first authors on papers that recently have appeared in prestigious peer-reviewed publications including the Journal of Neurophysiology, Neuroscience and Psychiatric Annals, as well as taking the form of a chapter in book by Nova Science Publishers titled, “New Research on Neuronal Networks.”

Guoshi Li has been working through Nair’s collaboration with Gregory Quirk, a neuroscientist in the University of Puerto Rico’s School of Medicine, computationally modeling the fear circuit in mammals. Disorders in these circuits can lead to a multitude of anxiety orders such as post-traumatic stress disorder (PTSD) and agoraphobia.

Two additional doctoral candidates in the Computational Neurology Group are working in collaboration with Nair and Peter Kalivas, professor and chair of the Department of Neurosciences at the Medical University of South Carolina. Ashwin Mohan and Sandeep Pendyam are modeling neuroplasticity in the brain in regard to cocaine addition.

Fear circuits

“An auditory fear-conditioning experiment is most commonly used to study fear,” said Li, explaining that when a tone is paired with a shock, the subject learns to associate the tone with the shock and learns fear. “In experiments, a rat’s heart rate and blood pressure increase each time it hears the tone.  Neurons in the amygdala—almond-shaped masses inside cerebral hemispheres—encode the fear memory.”

“But when the tone is played repeatedly in the absence of shock, the fear behavior eventually subsides, due to the formation of a new ‘extinction’ memory. Furthermore, fear can spontaneously recover after extinction. So, fear memory is not lost, but is overcome by extinction memory.”

“However, the neuromechanism for the formation of fear and extinction memories is not well understood,” said Li. “Simple models previously developed cannot provide a look into the deeper mechanisms as those which occur in the pathology of PTSD where the fear circuit is disrupted. Due to this disruption, PTSD patients cannot retrieve the extinction memory, although it exists in their brain, so the fear memory dominates every time the patient experiences the fear cue.”

Li explained that the brain functions as a dynamic system and can be studied through the use of computational modeling of synaptic connections using experimental data from rat brains provided by Quirk’s lab. It has been shown that memory occurs in the synapses, and modeling allows Li to make novel predictions, which can then be tested experimentally.

As techniques to study neuroscience improve, so does the potential for computational modeling. Mathematical equations allow researchers to combine information about many functions within an overall system to make predictions.

“Many neurons together can perform a complicated function. It is essentially the same as looking at a complex system such as an airplane. Through the use of system concepts, including such things as math, circuits and physics, we come up with a model and then use the model to make predictions and devise new experiments,” Li said.

Treatment for disorders such as PTSD depends on where fear extinction is stored, which connections are involved and which circuits misfire. Identifying where these problems occur may lead to the suggestions of new drugs to treat the problem. Computational modeling can give better understanding of brain mechanisms and circuits and, in fact, Li’s modeling has shown that the fear memory isn’t completely erased by extinction.

“Dr. Quirk has learned a lot from our model and is eager to test it,” said Li. “Before coming to MU, I never did anything in biology and neuroscience. It’s very exciting to discover these new principles.”

Neuroplasticity in cocaine addiction

“So many things happen simultaneously in the brain with drug addiction that it is not easy to fit them together,” said Mohan, explaining one of the most challenging aspects of applying computational power to the problem of studying the brain’s neuroadaptations in cocaine addicts.

“In normal people, the brain responds to the prospect of food,” said Mohan. “But if given a choice between food and cocaine, an addict will choose cocaine.”

In their research, Mohan and Pendyam were able to show that the brain structure with its synaptic connections actually changes in people who are addicted to cocaine.

Yet it is also a known fact that people have positively responded to drug therapies in rehabilitation, demonstrating neuronal elasticity.

“A drug called N-acetycysteine is being tested on addicts to see if it could help them to recover,” said Pendyam.
The eventual aim of the research is to discover how these drugs actually work by devising a model of the fundamental cellular/molecular level workings of an addicted brain.

In order to build a model to study this, the pair did extensive literature research for all known parameters of the neuronal system affected by addiction, including the data from Kavalis’ lab.

“It is non-linear research in which ‘A’ has no direct correlation to ‘B’. We know what changes, and computationally we can model the process and make predictions,” Mohan said.

Based on their modeling, the MU researchers concluded that in the cocaine-addicted brain’s response to the drug, parameters beyond those provided by Kavalis’ research must undergo alteration. They returned their results to their collaborator and he was able to prove that their predictions were correct.
The pair explained that glutamate is the major chemical released in electrochemical synaptic transmissions, and the amount released defines the strength of the synapse, so the amount must be regulated. Once the chemicals have done their job, they must be recycled; some neurons produce the chemicals, and some—glial cells—absorb them. A pathology occurs anywhere there is a mismatch in this process.

“In an addict’s limbic system, the brain’s pleasure center, we found excessive production of glutamate,” said Pendyam.

“Our model showed that the glial glutamate transporters are almost 40 percent less functional after chronic cocaine use,” said Mohan. “The sensors begin to fail and there is no way for the system to regulate itself.”

“The neuroplasticity is irreparable and there is permanent damage. Your brain has retuned itself and this pathology is addiction,” Pendyam said. “It is not just a behavioral change, but a disease, and so must be treated as such.”

“We used a systems approach to look at problems in biology, and using that approach we found a key parameter they had missed,” said Mohan. “It requires learning about neurobiology, and takes lots of patience, but it’s exciting to solve health-related problems using my education and background in engineering.”

“It’s experimental research, but as engineers we are able to look at the mechanisms and predict responses,” Pendyam adds. “It all depends on the collaboration where the sum is much greater than the parts. Neither lab can do this on their own.”

Categorized as:

Magazine Issue: