Improvement of the performance and reliability of a cell-surface untargeted proteomics workflow in the context of multiple myeloma antigen discovery

Disease management of multiple myeloma (MM) is challenging owing to patients’ multiple relapses and resistance to standard treatments. Therefore, the introduction of innovative cell therapies such as chimeric antigen receptor T-cells (CAR-T cells) opened new horizons in the outcome of severe refractory patients affected by aggressive forms of the disease. However, one of the limitations of the expansion of those treatments is the lack of specific MM tumor-associated antigens that could be targeted by current immunotherapies. Indeed, effectiveness of current CAR-T cells treatments could be hampered due to possible antigen-evasion strategies. Therefore, the discovery of new cell surface antigens could be an interesting approach to avoid occurrence of resistance that could lead to treatment failures.

For this purpose, a sensitive and robust cell surface mass spectrometry (MS)-based proteomics workflow was implemented and improved to increase identification rate of proteins. In the analytical point of view, a sensitive microfluidic liquid chromatography-based chip coupled to high resolution MS was used in this study to analyze protein digests from a human MM cell line. Moreover, data-independent acquisition mediated by ion mobility was also applied to maximize proteome coverage.

The performance of sample preparation is a critical part of the workflow that influences the output in terms of protein identification. Indeed, different steps of this procedure including biotinylation, elution, digestion and peptide clean-up were optimized using one million cells from a human MM cell line to yield the highest number of identification. Furthermore, the workflow was also applied to the analysis of cell line samples containing decreasing number of cells from one million to fifty thousands. The aim was to evaluate the performance of the workflow to analyze low number of cells due to the complexity to obtain primary cells from patients containing high number of cells.

Performing the whole workflow, including sample preparation, MS analysis and data treatment, is time-consuming. The initial sample preparation protocol contains more than 50 steps of manual handling of the samples which emphasizes the complexity of the process. In order to ensure the reliability of the sample preparation procedure, the use of quality control (QC) metrics is relevant. Therefore, an approach was implemented in our workflow to remove samples that generated low quality data. This post-analysis method is based on the identification of “all process” QC protein of different intensities to determine the reliability of the sample preparation. This approach was developed using samples containing one million cells but was also evaluated for lower number of cells until fifty thousands. Moreover, the analysis of patient samples was performed using the developed workflow and the identification of QC proteins allowed to discard low quality samples from further analysis.

In conclusion, a untargeted MS-based proteomics workflow to analyze MM cell lines and primary cells was successfully developed through the optimization of the sample preparation workflow. Additionally, the implementation of QC proteins allowed to increase the reliability of sample preparation.