Modelling Cortical and Thalamocortical Synaptic Loss and Compensation Mechanisms in Alzheimer’s Disease

Abstract

Confirming that synaptic loss is directly related to cognitive deficit in Alzheimer's disease (AD) has been the focus of many studies. Compensation mechanisms counteract synaptic loss and prevent the catastrophic amnesia induced by synaptic loss via maintaining the activity levels of neural circuits. In this chapter we investigate the interplay between various synaptic degeneration and compensation mechanisms, and abnormal cortical and thalamocortical oscillations based on two studies involving different implementations of a large-scale neural network model. Study 1 involves a large scale cortical model (C-model) and includes 100,000 neurons exhibiting several cortical firing patterns, 8.5 million synapses, short-term plasticity, axonal delays and receptor kinetics. The structure of the model is inspired by the anatomy of the cerebral cortex. Study 2 involves an extended model, a thalamocortical network model which oscillates within the alpha frequency band (8-13 Hz) as recorded in the wakeful relaxed state with closed eyes. The thalamocortical network model (TC-model) includes different types of cortical excitatory and inhibitory neurons recurrently connected to thalamic and reticular thalamic regions with the ratios and distances derived from the mammalian thalamocortical system. The results of Study 1 suggest that cortical oscillations respond differently to compensation mechanisms. Local compensation preserves the baseline activity of theta (5-7 Hz) and alpha (8-12 Hz) oscillations whereas delta (1-4 Hz) and beta (13-30 Hz) oscillations are maintained via global compensation. Applying compensation mechanisms independently shows greater effects than combining both compensation mechanisms in one model and applying them in parallel. Consequently, it can be speculated that enhancing local compensation might recover the neural processes and cognitive functions that are associated with theta and alpha oscillations whereas inducing global compensation might contribute to the repair of neural (cognitive) processes which are associated with delta and beta band activity. Study 2 focuses on investigating the impacts of four types of connectivity loss on the model's spectral dynamics, namely degeneration of corticocortical, thalamocortical, corticothalamic and corticoreticular couplings, with an emphasis on the influence of each modelled case on the spectral output of the model. Synaptic compensation has been included in each model to examine the interplay between synaptic deletion and compensation mechanisms, and the oscillatory activity of the network. The results of power spectra and event related desynchronisation/synchronisation (ERD/S) analyses show that the dynamics of the thalamic and cortical oscillations are significantly influenced by corticocortical synaptic loss. Interestingly, the patterns of changes in thalamic spectral activity are correlated with those in the cortical model. Similarly, the thalamic oscillatory activity is diminished after partial corticothalamic denervation. Given the results, it can be speculated that thalamic atrophy is a secondary pathology to cortical shrinkage in Alzheimer's disease. In addition, this study finds that the inhibition from neurons in the thalamic reticular nucleus (RTN) to thalamic relay (TCR) neurons plays a key role in regulating thalamic oscillations; disinhibition disrupts thalamic oscillatory activity even though TCR neurons are more depolarized after being released from RTN inhibition. Both study 1 and study 2 indicate that compensation mechanisms may vary across cortical regions and the activation of inappropriate compensation mechanism in a particular region may fail to recover network dynamics and/or induce secondary pathological changes in the network. Both studies provide a better understanding on the neural causes of abnormal oscillatory activity in neurodegeneration.