The neuroanatomy of autism - a developmental perspective

Abstract

Autism Spectrum Disorders (ASDs) are a heterogeneous group of neurodevelopmental disorders that are diagnosed solely on the basis of behaviour. A large body of work has reported neuroanatomical differences between individuals with ASD and neurotypical controls. Despite the huge clinical and genetic heterogeneity that typifies autism, some of these anatomical features appear to be either present in most cases or so dramatically altered in some that their presence is now reasonably well replicated in a number of studies. One such finding is the tendency towards overgrowth of the frontal cortex during the early postnatal period. Although these reports have been focused primarily on the presumed pathological anatomy, they are providing us with important insights into normal brain anatomy and are stimulating new ideas and hypotheses about the normal trajectory of brain development and the function of specific anatomical brain structures. The use of model systems that include genetic model organisms such as the mouse and, more recently, human induced pluripotent stem cell-derived brain organoids to model normal and pathological human cortical development, is proving particularly informative. Here we review some of the neuroanatomical alterations reported in autism, with a particular focus on well-validated findings and recent advances in the field, and ask what these observations can tell us about normal and abnormal brain development.

Keywords: autism; cerebellum; development; dysplasia; frontal cortex; minicolumns; neuroanatomy; overgrowth; regionalisation.

Figure 1

Hypothesis of potential mechanisms underlying focal disorganisation and altered minicolumn distribution in ASD cortex. Normal: During normal cortical development (middle panel), radial glial NSCs (black) located at the ventricular zone, with radial processes extending to the pial surface of the neural epithelium, can undergo symmetrical (to produce more radial glial NSCs) or asymmetric divisions to maintain the NSC pool and generate committed neural progenitors (brown) (A). The latter migrate towards the pial surface on the radial scaffold, where they differentiate into projection neurons. This process continues with different classes of neurons (pink, green), being produced in successive waves over developmental time (B), so that different neuronal types are grouped together in different cortical layers (C). GABAergic interneurons (orange) migrate into the neuroepithelium (B), and integrate into the circuitry (C). At the end of this process, projection neurons are organised in radial minicolumns, a reminder of their developmental origin. Focal disorganisation: We hypothesise that focal disruptions may be the result of early genetic or epigenetic changes in NSCs (i), such that the resulting neuronal progenitors have defects in differentiation (ii) or migration (iii) such that they adopt the wrong fate (iv), or end up in an in appropriate position in the brain (v). The end result is a cortex with apparent disorganisation of neuronal cell types and perhaps even distorted cell densities. Minicolumn width: We hypothesise that alterations in symmetric NSC divisions such that the early density of radial scaffolds is increased, will result in increased density of minicolumns. Alternatively, the migration and integration of GABAergic interneurons and other independent mechanisms might affect the spacing of minicolumns.