Single-cell mass spectrometry profiling of the developing human brain reveals widespread mRNA-protein discordance that is particularly pronounced in neurodevelopmental disorder-associated genes, with proteins exhibiting markedly higher cell-type specificity than transcripts and proteomic trajectory analysis identifying the intermediate progenitor-to-excitatory neuron transition as a critical vulnerability window for autism spectrum disorders characterized by a highly conserved, co-regulated protein module enriched for ASD-associated de novo mutations.

Proteins are the functional effectors of virtually all biological processes, and accurately measuring their abundance and dynamics is essential for understanding development and disease. Although mRNA levels have historically been used as proxies for protein expression, growing evidence, especially from studies of the human cerebral cortex, has revealed widespread discordance between transcript and protein abundance. To directly address this limitation, we developed a rigorously optimized workflow that combines single-cell mass spectrometry with precise sample preparation to resolve, for the first time, quantitative proteomes of individual cells from the developing human brain. Our platform achieved deep proteomic coverage (∼800 proteins per cell) even in immature prenatal human neurons (5–10 μm diameter, ∼100 pg of protein per cell), capturing major brain cell types and enabling proteome-wide characterization at single-cell resolution. This approach revealed extensive transcriptome–proteome discordance across cell types, with particularly strong discrepancies in genes associated with neurodevelopmental disorders, a finding validated through orthogonal experiments. Proteins exhibited markedly higher cell-type specificity than their mRNA counterparts, underscoring the importance of proteomic-level analysis for resolving cellular identity and function. By reconstructing developmental trajectories from radial glia to excitatory neurons at the proteomic level, we identified dynamic stage-specific protein co-expression modules and pinpointed the intermediate progenitor-to-neuron transition as a molecularly vulnerable phase linked to autism. Altogether, by enabling single cell proteomics, this study establishes a foundational resource and technological advance for developmental neuroscience. It demonstrates that single-cell proteomics can capture critical developmental events and disease mechanisms that are undetectable at the transcript level. As this technology continues to improve in sensitivity and scalability, single-cell proteomics will become an indispensable tool for uncovering the molecular logic of brain development and for illuminating pathophysiological processes underlying neurodevelopmental disorders.