With a $90,000 grant from FRAXA Research Foundation, Dr. Terrence Sejnowski and Dr. Cian O’Donnell, through their own computational models of neural networks, will take the results of previous FRAXA-funded projects and incorporate them into their advanced computer models of brain function. This promises to help us to understand how the Fragile X brain responds to drugs and many other interventions.
Many cognitive processes like thinking, decision making, seeing, and hearing involve electrical signals in a part of the brain called the neocortex. During development, neurons in the neocortex get wired up to each other. Correct electrical signaling in adulthood depends on correct wiring during development.
In recent decades, researchers have found that this developmental wiring process depends heavily on ‘spontaneous’ activity in the brain. This activity is called spontaneous because it is not driven from events in the outside world, but instead happens when animals or humans are resting or sleeping. Research has also lead to the idea that this spontaneous activity during development might be altered in Fragile X syndrome (FXS). If true, this might give us a clue as to how cortical neurons are wired differently in FXS adults, and potentially how to reverse or reduce the effects.
Unfortunately from a treatment point of view there is an enormous number of possible ways in which altered spontaneous activity might cause altered cortical wiring (and vice versa). Teasing these possibilities apart experimentally would require a large trial-and-error search for the mechanisms. In our project we aim to use detailed computational models of Fragile X cortical circuits to try to speed up this process. Our aim is to come up with a reduced shortlist of possible candidate mechanisms for future experiments. The advantages of a computational approach over the traditional lab experiments in this instance are: 1) we can quickly test many different possibilities in a much shorter time frame, 2) we can simulate experiments that might be difficult or impossible to conduct in the lab, and 3) we can make quantitative predictions for future experiments, which is often hard to do using mental reasoning alone.
Finally, once we have ideas about possible mechanisms, we can use the computer models to screen candidate interventions by simulating the actions of certain drugs. The simulations also allow for a rapid dose-response test for any given drug. This information could help speed up future drugs trials in both animals in the lab and patients in the clinic.
Cian O’Donnell, PhD
FRAXA Postdoctural Fellow
