Multiscale modeling and empirical studies of normal and pathological brain aging

Boston University Medical Campus, Boston MA

Investigators

Abstract

Project Summary/Abstract Progressive myelin dystrophy, loss of synapses, and neuronal hyperexcitability are hallmark changes in the neocortex common to both normal aging and early-stage sporadic Alzheimer’s Disease (AD), a disease of aging. The precise relationships of these changes and how they drive age- and AD-related cognitive impairments remain elusive. In the rhesus monkey, a species unaffected by clinical AD, these sub-lethal age-related changes in the dorsolateral prefrontal cortex (dlPFC) correlate with degree of working memory (WM) impairment. Given that myelin dystrophy is an early and profound phenotype, it is likely that long-range heavily myelinated axons or those that have high myelin turnover may be most vulnerable. In monkey dlPFC, layer 3 (L3) pyramidal cells (PCs) extend horizontally oriented axon collaterals from the main axon to form periodically spaced columnar axonal plexus clusters. These clusters are implicated in sustained neuronal activity during WM and are the anatomical expression of interconnected L3 PCs forming excitatory reverberating circuits modulated by local inhibitory inputs. The parent L3 PC axons project to three targets: local intrinsic targets within the same cluster, long-range intrinsic targets in adjacent clusters, and long-range extrinsic targets in other cortical areas. Our preliminary data suggest that these target-specific axons differ with regard to their structural features including degree of myelination. We hypothesize that heavily myelinated axons/collaterals and their downstream targets are more vulnerable to the effects of normal and pathological aging than are lightly myelinated or unmyelinated axons/collaterals and this underlies specific WM dysfunction. Further, we hypothesize that remyelination of affected fibers is capable of rescuing neuron and circuit function in computational models. The organizational features of dlPFC circuits provide a rigorous framework for modeling and empirical validation across many spatial and temporal scales. Aim 1 will model the effects of selective myelin loss and its contribution to/interaction with hyperexcitability on action potential conduction of parent axons and their local and long-range collaterals, and how these changes impact synaptic communication in small circuits of neurons analogous to the clusters and gaps seen in dlPFC. Aim 2 will employ large-scale models of WM to answer the important question of how progressive changes in myelin, synaptic connections and neuronal excitability interact to effect WM and how the addition of neuron loss further impacts network function. Both Aims 1 and 2 will test the effect of variable and site specific remyelination on model function. Aim 3 will use quantitative anatomical and physiological approaches in rhesus monkey dlPFC across the adult lifespan to provide validation of model predictions and also help constrain models in an iterative manner. Results will be invaluable for the future development of biophysically realistic computational models of WM that incorporate multiple brain areas, layers, and cell types in the context of normal aging and AD. These insights, combined with precise techniques to perturb and rehabilitate neurons and networks, are essential for the development of therapeutics for cognitive decline in normal aging and in AD.