Genetic alterations in autism spectrum disorder (ASD) ultimately manifest at the protein level. Because of the strongest discordance between RNA expression and protein abundance specific to the brain cortex and the functional convergence of the ostensibly heterogeneous genomic mutations in ASD onto specific protein interaction pathways, Jingjing Li, Arnold Kriegstein, Michael Snyder and Mohan Babu therefore propose to generate high-resolution quantitative proteome and interactome maps in the developing cerebral cortex to uncover mutationally convergent pathways in ASD.

The research team has established a collection of 110 fresh, frozen, typically developing human cerebral cortical samples from the first, second and third trimesters of gestation as well as from infant, child and adolescent age groups. These age groups represent key brain developmental stages before, during and after the onset of ASD. The team plans to leverage state-of-the-art, high-precision mass spectrometers (i.e., the Orbitrap Eclipse Tribrid platform, which enables single-cell proteomic profiling) to deeply profile these samples and, for the first time, generate a panoramic view of the proteome and interactome landscapes in the developing human cerebral cortex.

Preliminary experiments based on prenatal cortical samples from the second trimester have demonstrated the precision and clinical relevance of the proteomic data that the team is able to collect. In the current project, the group plan to use tandem mass tags (TMT) and mass spectrometry (MS3) methods to measure proteome-wide protein abundance in the developing human cerebral cortex across the developmental stages mentioned above, followed by single-cell proteomic profiling targeting the second trimester samples (different cortical layers), which represents a critical developmental stage in ASD. They also plan to perform bulk and single-cell transcriptomic profiling in the same tissue samples, enabling direct comparisons between transcriptomes and proteomes at bulk and single-cell resolutions.

The concordance and discordance between RNA expression and protein abundance in each cell type from different cortical layers and at varying developmental stages will be determined. In parallel, they also plan to perform co-fractionation/mass spectrometry (CF/MS) to identify protein assemblies at a subcellular resolution (targeting nuclear, mitochondrial and cytosolic compartments) in the cortical samples across the different stages, constructing cortex-specific and developmental-stage-specific protein interactomes and revealing key protein interactions or complexes involved in cortical development. Most notably, the unprecedented subcellular resolution in the developing brain will, for the first time, enable functional investigations of the emerging role of mitochondrial dysfunction in ASD.

The team also plans to use machine learning to integrate the generated proteome and interactome atlas data with large-scale SFARI datasets of ASD whole-exome and genomes (from the Simons Simplex Collection and SPARK) to identify mutationally convergent pathways, as well as tissue(s) of origin, in ASD pathophysiology. Upon the project’s completion, all the generated data will be made available to the science community.

Extending from recent work on charting proteomic maps of the human body, this study will reveal the abundance, composition and topology of protein complexes in discrete regions of the developing brain in a native physiological context. Results from this study are expected to advance our understanding of the molecular etiologies of ASD and pave the way for the development of novel therapeutics.