Abstract
Proteins function in the context of their environment, so an understanding of cellular processes requires a knowledge of protein localization. Thul et al. used immunofluorescence microscopy to map 12,003 human proteins at a single-cell level into 30 cellular compartments and substructures (see the Perspective by Horwitz and Johnson). They validated their results by mass spectroscopy and used them to model and refine protein-protein interaction networks. The cellular proteome is highly spatiotemporally regulated. Many proteins localize to multiple compartments, and many show cell-to-cell variation in their expression patterns. Presented as an interactive database called the Cell Atlas, this work provides an important resource for ongoing efforts to understand human biology.Science, this issue p. eaal3321; see also p. 806INTRODUCTIONA complete view of human biology can only be achieved by studying the molecular components of its smallest functional unit, the cell. Cells are internally organized into compartments called organelles. The spatial partitioning provided by organelles creates an enclosed environment or surface for chemical reactions tailored to fulfill specific functions. These functions are tightly linked to a specific set of proteins. Therefore, resolving the subcellular location of the human proteome provides information about the function of the organelle and its underlying cellular mechanisms. We present a subcellular map of the human proteome, called the Cell Atlas, to facilitate functional exploration of individual proteins and their role in human biology and disease.RATIONALEImmunofluorescence (IF) microscopy was used to systematically resolve the spatial distribution of human proteins in cultivated cell lines and map them to cellular compartments and substructures with single-cell resolution. This approach allowed definition of the precise location of a majority of the human proteins in their cellular context and exploration of single-cell variations in protein expression patterns. The proteome-wide information about protein spatial distribution was validated with an orthogonal proteomics method, and the results were integrated into existing network models of protein-protein interactions for increased accuracy.RESULTSWe report a high-resolution characterization of the spatial subcellular distribution of the human proteome based on more than 80,000 confocal IF images. A total of 12,003 proteins targeted by 13,993 antibodies were classified into one or several of 30 cellular compartments and substructures, altogether defining the proteomes of 13 major organelles. The organelles with the largest proteomes were the nucleus and its substructures (6245 proteins), such as bodies and speckles, and the cytosol (4279 proteins). However, smaller organelles such as the midbody, rods and rings, and nucleoli also showed a larger diversity than previously recognized. Intriguingly, about half of all proteins were localized to multiple compartments, showing that there is a shared pool of proteins even among functionally unrelated organelles. Single-cell analysis revealed 1855 proteins with variation in their expression pattern, either in terms of expression levels or spatial distribution. Last, the spatial information was used to refine biological networks. Our location-pruned network that restricts protein interaction to the same organelle improved the accuracy of the human interactome model. The analysis also included transcriptomics data for all putative protein-coding genes (19,628) in 56 human cell lines of various origins. On average, cell lines expressed 11,490 genes, with half of them (6295) being expressed across all samples, suggesting a “housekeeping” role.CONCLUSIONThe cellular proteome is compartmentalized and spatiotemporally regulated to a high degree. The high-resolution subcellular map of the human proteome that we provide describes this cellular complexity, with many multilocalizing proteins and single-cell variations. The map is presented as an interactive database called the Cell Atlas, part of the Human Protein Atlas (www.proteinatlas.org). The Cell Atlas constitutes a key resource for a holistic understanding of the human cell and its complex underlying molecular machinery, as well as a major step toward modeling the human cell.Creation of an image-based map of the human subcellular proteome.The subcellular locations of 12,003 proteins were determined by IF microscopy in cell lines of various origins. High-resolution IF images such as those shown above enabled mapping of proteins to distinct subcellular structures. This resulted in the definition of the proteomes of 13 major cellular organelles, revealing multilocalizing proteins, as well as expression variability on a single-cell level.Resolving the spatial distribution of the human proteome at a subcellular level can greatly increase our understanding of human biology and disease. Here we present a comprehensive image-based map of subcellular protein distribution, the Cell Atlas, built by integrating transcriptomics and antibody-based immunofluorescence microscopy with validation by mass spectrometry. Mapping the in situ localization of 12,003 human proteins at a single-cell level to 30 subcellular structures enabled the definition of the proteomes of 13 major organelles. Exploration of the proteomes revealed single-cell variations in abundance or spatial distribution and localization of about half of the proteins to multiple compartments. This subcellular map can be used to refine existing protein-protein interaction networks and provides an important resource to deconvolute the highly complex architecture of the human cell.
