Assessment of Ultra-Deep Mass Spectrometry Based Proteomics Compared to Flow Cytometry and RNA-Based Methods for the Discovery and Validation of Therapeutic Targets in Immune Cells

Abstract

Immuno-oncology (IO) has substantially improved the survival of cancer patients over the past several years encouraging the discovery of novel IO targets which are typically proteins expressed on the surface of immune cells. Sensitive quantification of proteins in complex biological samples is routinely achieved by immunoassays that use antibodies specific to target proteins. Such approaches can be a limitation in IO drug discovery since the development of de novo antibodies is associated with long lead times, high costs, and high failure rates, often resulting in the substitution of the protein readout by a surrogate transcript readout using RNA sequencing methods. Protein quantification using mass spectrometry (MS) is agnostic to any capture agent and removes the barriers of availability or specificity of antibody-based assays. Furthermore, MS proteomics workflows can support large scale discovery studies with wide proteome coverage, and also represent an attractive antibody-free alternative in targeted quantitative studies for protein measurement. The main purpose of this work was primarily to assess the sensitivity of DIA-based MS-proteomics platform for the deep proteome profiling of human primary immune cells by measuring the maximum protein coverage obtained across cell inputs ranging from 2 million down to single cells ( performance assessment). The secondary purpose of this work was to assess the accuracy of the quantification of targeted MS proteomics workflow in comparison with the gold standard QIFI® flow cytometry approach, and other methods using mRNA readouts in immortalized human hematological cell lines ( accuracy assessment). We found that our MS-based proteomics workflows achieve high sensitivity in detection of immune cell markers down to few immune cells as input. Additionally, we report a strong correlation of the quantitative data derived from our MS-based proteomics workflows compared to the flow cytometry gold standard (r 2= 0.7354). Notably, we observed that this correlation was significantly higher than using transcriptomics approaches, emphasizing the benefit of substituting immunoassays for MS-based proteomics platforms in IO target discovery and validation.