Publication Summary and Abstract
Tomkins, A., Humphries, M. D., Vasilaki, E., & Gurney, K. N. (2011), How Degrading Networks Can Increase Select Cognitive Functions, Frontiers in Computational Neuroscience, Presented at the BC11 : Computational Neuroscience & Neurotechnology Bernstein Conference & Neurex Annual Meeting, Freiburg, Germany.
It has recently been shown that despite the severe cell atrophy associated with manifest Huntington's Disease (HD), increased proficiency can be demonstrated in select behavioural tasks (Beste et al., 2008). Using computational models of the striatal micro-circuit we show possible mechanisms underlying the phenomena as a result of the excitotoxicity (Fan and Raymond 2007) and the competitive and selective dynamics of signal selection in the striatum. The increase in sensitivity to endogenous glutamate associated with HD could increase the performance of NMDA dependant tasks such as those involving auditory sensory memory (Kujala et al. 2007), in spite of the severe neural deterioration. Using Mismatch Negativity (MMN), studies by Beste et al. (2008) show that patients with symptomatic HD out perform pre-symptomatic HD (pHD) and a healthy control group, exhibiting lower reaction times and error rates for standard and deviant stimuli. We propose that this phenomena is a feature of the network structure of striatal neurons when combined with the excitotoxicity hypothesis, occurring due to the connectivity of the striatum. Using the biologically inspired model of the striatal microcircuit taken from Humphries et al. (2010), we show competitive dynamics within the network, demonstrating an inherent competitive selective signal behaviour. By modelling two separate cortical inputs, representing two distinct actions, to sub-populations in the striatal model, we show that the network displays preferential selectivity to a strong input signal, suppressing the activity of the weaker signal at the onset of the stronger. The network displays a further reactive selectivity: the removal of the stronger signal subsequently boosts the population response to the weaker signal, possibly encouraging behavioural switching back to the ongoing signal. Defining these two elements of selectivity allows us to investigate how this selectivity evolves as the network degrades as in manifest HD. Mimicking the excitotoxicity. we up-regulate the NMDA conductivity versus the AMPA, progressively eliminating over-excitable neurons from the network. Using a new measure for signal selectivity, our model show that excitotoxic cell atrophy, combined with increased excitability of the surviving cells, can encourage signal selectivity. Suggesting that a pathogenic increase in a transmitter sensitivity, while encouraging cell atrophy, can mediate network level behaviour, affecting the competitive signal selection dynamics due to increased activity. Thus, our results suggest that the paradoxical cognitive enhancement of the manifest HD patients is due to their selectively enhanced signal selection in the striatum. We suggest that the striatal model can be a useful tool to shed light onto the ramifications of hypotheses such as excitotoxicity, and by examining changes in network dynamics, may provide the basis for a select few counter-intuitively improved cognitive functions in HD patients.
Article via Digital Object Identifer (DOI) - article retrieved is probably subject to publisher's copyright