Three Subtypes of Autism Spectrum Disorder with Transcriptomic Signatures Derived from Morphometric Similarity Networks
Hongxiu Jiang,Raul Rodriguez-Cruces,Ke Xie,Valeria Kebets,Yezhou Wang,Clara F. Weber,Ying He,Jonah Kember,Jean-Baptiste Poline,Danilo Bzdok,Seok-Jun Hong,Boris Bernhardt,Xiaoqian Chai
DOI: https://doi.org/10.1101/2024.09.06.611561
2024-09-10
Abstract:Autism spectrum disorder (ASD) is a prevalent and highly heterogeneous neurodevelopmental disorder. Previous studies have attempted to identify ASD subgroups by analyzing isolated cortical structural features. However, these studies have not considered the relationship between multiple structural features, which provide information on the structural organization of the brain. Morphometric similarity network (MSN), a structural brain network contributed by multiple anatomical features (gray matter volume, mean cortical thickness, surface area, mean curvature, Gaussian curvature, curvature index, and fold index), strongly relates to cytoarchitectonic and genomic measures of histological similarity between cortical areas. We applied K-means clustering on MSN from 236 individuals with ASD and identified three subtypes. Subtype-1 showed relatively similar MSN values with typically developmental individuals (TD). Subtype-2 showed higher morphometric similarities in the lateral frontal and temporal cortical regions and lower in anterior prefrontal and occipital regions compared to TD. These patterns were the opposite in subtype-3. Behaviorally, subtype-3 had more severe verbal and social deficits compared to subtype-2. The weaker resting-state functional connectivity (rs-FC) between the language and salience networks was observed between subtype-2 and TD. Subtype-3 had stronger rs-FC between salience and default mode networks (DMN), between frontoparietal and visual networks, and between language and dorsal attention networks, while weaker rs-FC within DMN, within sensorimotor, and within salience networks. In addition, genes with expression patterns associated with regional MS changes in ASD subtypes were functionally enriched in neuron-specific biological processes related to nervous system development, synaptic signaling and chromatin organization. These genes were particularly enriched in GABAergic neurons, glutamatergic neurons, astrocytes and microglia. Taken together, our findings suggest the existence of different neuroanatomical subtypes based on multiple anatomical features in ASD with distinct transcriptomic signatures and functional connectome patterns.
Neuroscience