bioRxiv Science⌕ Search

Biology subjects

Yoo, O.

Publications and source records attributed to Yoo, O..

3 recordsLinked to original sources

Inter-individual variation of cellular and gene-expression properties of the human striatum

The human brain varies from person to person in ways that shape behaviors and vulnerabilities, yet the cellular and molecular bases for inter-individual variation are largely unknown. Here we describe an analysis of cellular and gene-expression variation in four key structures of the striatum complex - the caudate, putamen, nucleus accumbens, and internal capsule - as well as the prefrontal cortex, from single-nucleus RNA-seq analysis of 3.9 million nuclei from 178 adult brain donors. We found that people with more astrocytes in any one brain region tended to have this property in all brain regions sampled; the same was true of striatal interneurons, microglia, and oligodendrocyte precursor cells (OPCs). OPCs showed attrition with age, declining in numbers by approximately 40% between age 30 and age 80 in both gray matter and white matter regions. We identified thousands of age-associated (but few sex-associated) variations in gene expression; the vast majority of these effects of age were cell-type-specific. Aging most strongly affected gene expression in projection neurons - especially striatal medium spiny neurons (MSNs/SPNs) - and had a much smaller effect on gene expression in interneurons. Individuals ages could be predicted to within about five years based on RNA-expression patterns from any of the striatal cell types. Common genetic variants detectably affected the expression levels of some ten thousand genes; the great majority of these effects were cell-type-specific. These data will provide a foundation for exploring natural inter-individual variation, aging, and tissue-based studies of human brain vulnerabilities.

neuroscience↗

Mesoscale molecular architecture of the human striatum across cell types and lifespan

The human striatum is a central hub for a diverse array of motor, cognitive, and affective behaviors, yet it lacks obvious cytoarchitectural boundaries that define functional territories. Here, we uncover a robust and molecularly defined mesoscale architecture in the human striatum. Using Slide-tags, a scalable single-nucleus spatial transcriptomics technology, we profiled 1.1 million cells across the full span of the anterior striatum of 19 postmortem donors, spatially mapping all striatal populations. Our data uncover a natural subdivision of the striatum into six zones, each defined by molecularly distinct populations of medium spiny neurons, and featuring spatially coordinated neuron-astrocyte signaling. Relative to MSNs in ventral zones, MSNs in dorsal zones exhibit higher expression of genes for synaptic remodeling and plasticity via ephrin and TGF-beta, while the ventral zone is defined by greater expression of semaphorin, protein chaperone, and hedgehog signaling pathways. By imputing zonal identities onto a larger single-nucleus RNA-seq cohort of 131 donors, we find that the dorsal zones exhibit greater age-related transcriptional changes, and that overall, the gene-expression differences that define spatial zonation patterns are attenuated with advancing age. This atlas provides a mesoscale molecular definition of human striatal anatomy, linking cell type identity to functional specialization and aging susceptibility.

neuroscience↗

ortho_seqs: A Python tool for sequence analysis and higher order sequence-phenotype mapping

MotivationAn important goal in sequence analysis is to understand how parts of DNA, RNA, or protein sequences interact with each other and to predict how these interactions result in given phenotypes. Mapping phenotypes onto underlying sequence space at first- and higher order levels in order to independently quantify the impact of given nucleotides or residues along a sequence is critical to understanding sequence-phenotype relationships. ResultsWe developed a Python software tool, ortho_seqs, that quantifies higher order sequence-phenotype interactions based on our previously published method of applying multivariate tensor-based orthogonal polynomials to biological sequences. Using this method, nucleotide or amino acid sequence information is converted to vectors, which are then used to build and compute the first- and higher order tensor-based orthogonal polynomials. We derived a more complete version of the mathematical method that includes projections that not only quantify effects of given nucleotides at a particular site, but also identify the effects of nucleotide substitutions. We show proof of concept of this method, provide a use case example as applied to synthetic antibody sequences, and demonstrate the application of ortho_seqs to other other sequence-phenotype datasets. Availabilityhttps://github.com/snafees/ortho_seqs & documentation https://ortho-seqs.readthedocs.io/

bioinformatics↗