Age-Related Changes in Axon Density and Myelin Thickness Are Altered in ASD

Background: Brain imaging studies have linked autism spectrum disorder (ASD) symptomatology to impairments in neuronal communication and cortical synchrony, with the prevailing hypothesis being over-connectivity of local brain regions and under-connectivity of distant regions. However, the cellular developmental trajectory that underlies alterations in neuronal connectivity remains understudied. Postmortem human brain tissue provides an invaluable opportunity to evaluate axonal ultrastructure in brain regions implicated in ASD. Electron microscopy of frontal lobe white matter reveals a disproportionate density of short- and long-range axons in ASD that varies with age (Zikopoulos et al 2018). Although there is extensive evidence of aberrant temporal cortical and subcortical development in ASD, the microscopic ultrastructure of axons in temporal white matter remains unknown.

Objectives: Our goal is to identify the neurodevelopmental changes in cortical axonal ultrastructure in white matter below the superior temporal gyrus (STG) and fusiform gyrus (FG) in neurotypical (NT) human brains, and determine if this trajectory is altered in ASD.

Methods: Postmortem human brain samples from 30 male cases (15 ASD, 15 age-matched NT controls) ranging from 2-44 years of age were micro-dissected from superficial and deep white matter of the inferior longitudinal fasciculus underlying STG in the dorsal region and FG in the ventral region. Tissue was prepared for ultrastructural analysis by electron microscopy (Liu and Schumann 2014). Ultrathin (70nm) sections were imaged at high resolution (8,400x magnification) to manually measure axon density, size, and myelination thickness. Small/medium axons (inner-diameter <0.7µM) in superficial white matter are presumed to be more short-ranging whereas larger axons (>0.7µM) in deeper white matter are presumed to be long-ranging.

Results: The density of smaller size axons decreases with age in NT superficial white matter underlying FG (p<0.01) and STG (trend p<0.08). This age-related NT decrease in axon density does not occur in ASD in either region, and in fact slightly increases, leading to an overall increase in the density of smaller axons in ASD relative to NT in STG superficial white matter (p <0.01). In contrast, there is a reduction in small axon density in FG deep white matter (p<0.05). The thickness of myelin in larger axons in superficial white matter underlying STG increases with age in NT (p<0.01), but not in ASD, leading to an overall decrease of myelin thickness in ASD in both STG and FG relative to NT (p<0.01).

Conclusions: The increase in density of smaller axons underlying STG is consistent with the hypothesis of over-connectivity of short-range local connections. The density of smaller axons decreases with age in neurotypical development, likely attributed to continual refinement of connections and synaptic pruning. However, this pattern of fine-tuning local temporal lobe connectivity is not as evident in ASD. Myelin thickness of larger, presumably longer-range, axons increases in neurotypical development with age, however this increase does not occur in ASD. In fact, myelin thickness of larger axons is reduced in ASD, which may have a dramatic impact on neuronal communication.