Classification of mouse B cell types using surfaceome proteotype maps

Marc Van Oostrum,Antonius Rolink,Maik Müller,Fabian Klein,Bernd Wollscheid,Patrick G A Pedrioli,Panagiotis Tsapogas,Lukas Reiter,Hui Zhang,Roland Bruderer

doi:10.1038/s41467-019-13418-5

Abstract

System-wide quantification of the cell surface proteotype and identification of extracellular glycosylation sites is challenging when samples are limited. Here, we miniaturize and automate the previously described Cell Surface Capture (CSC) technology, increasing sensitivity, reproducibility and throughput. We use this technology, which we call autoCSC, to create population-specific surfaceome maps of developing mouse B cells and use targeted flow cytometry to uncover developmental cell subpopulations.

Highlights

System-wide quantification of the cell surface proteotype and identification of extracellular glycosylation sites is challenging when samples are limited
Compared to the number of surface proteins identified from a sample of 30 × 106 cells, we recovered a median of 62% with 1 × 106, the variance increased below 5 × 106 (Supplementary Fig. 3)
In summary, miniaturised and automated Cell Surface Capture (CSC) technology was validated and used to analyse cancer cell lines and developing primary B cells. mRNA abundance is not sufficient to infer protein levels in many scenarios[16], and to predict genotype–phenotype relationships it is necessary to go beyond global protein levels and determine proteoform abundances within the spatial domains where activity is required for function[17]

Summary

Results

Automation and miniaturisation of cell surface capture. Miniaturisation and automation have proven beneficial in phosphopeptide[8], hydrazide[9] and antibody-based[10] enrichment processes. Compared to the number of surface proteins identified from a sample of 30 × 106 cells, we recovered a median of 62% with 1 × 106, the variance increased below 5 × 106 (Supplementary Fig. 3). With this limitation in mind, we set out to de novo map and quantitatively compare the surfaceomes of developing B cells using autoCSC. With autoCSC, we quantified 248 unique glycosylated asparagines located within proteotypic peptides and grouped them into 147 protein groups with a median of two unique sites per protein group (Fig. 2b and Supplementary Data 2). Few proteins we found exclusively on a particular population, for example CD80 and CD130 on B1 cells (Supplementary Fig. 5) (Fig. 2b)

SB LB PB

Discussion

Methods

Code availability