Purpose: Proteins not only comprise nearly all drug targets, but complex protein networks also drive biomedical system function. For example, protein kinases, phosphatases and their substrates are frequently dysregulated in cancer, giving rise to the concept of “oncogene products”. These are important proteins that are also dysregulated in other diseases, driving proliferation, differentiation, drug resistance, resistance to apoptosis, metastatic behaviors and more. There are now effective therapeutic approaches that affect protein phosphorylation and other post-translational modifications (PTMs) of proteins. In spite of progress, many of the underlying mechanisms controlling health and disease remain unclear. The measurement of protein abundance and PTMs, to identify the most relevant protein biomarkers, has been hindered by limited availability of suitable reagents for antibody based-ligand binding assays (LBA). We describe herein the use of modern, high sensitivity, high accuracy liquid chromatography-tandem mass spectrometry- (LC-MS/MS) based proteomic analyses. Protein abundance and PTMs are directly measured, thus revealing biomedical mechanisms and biomarkers. Unbiased proteomic discovery enables proteins and their PTMs to be identified and quantified without preconceptions of which proteins are most relevant to biomedical system behavior.
Methods: Successful studies begin with well-planned and executed sample preparation from any biological material of interest. Total proteomes, including PTMs can be analyzed (Figure 1A). Alternatively, for a more narrow focus, antibody-based enrichment of proteins (Figure 1B), or purification of single proteins (Figure 1C) can be performed prior to proteomic analysis. Effective proteolytic digestion is required for quantitative accuracy and reproducibility of datasets (Figures 1D, 1D’). For optimal results, mixtures of peptides comprising total proteomes (Figures 1E, 1A’) require a first dimension of separation before LC-MS/MS analysis. PTMs frequently exist at a low abundance, so if PTM analysis is a goal, specific PTM enrichment should be included when available (Figure 1F). To generate raw proteomic data, high sensitivity, nanoflow-reversed-phase electrospray ionization (nano-RP ESI) high-resolutionaccurate mass spectrometry (HRAMS; Figure 1G) is also required for optimal accuracy and to minimize the false negative rate. Raw MS/MS data is searched against a protein database of interest, e.g. human, which contains all known and predicted proteins encoded by the genome (Figure 1H). Search algorithms match raw MS/MS spectra against theoretical MS/MS spectra of peptides contained within the protein database. The goodness of fit of the peptide-spectrum matches is computed, lower confidence identifications are removed, and peptides are annotated within their parent proteins. Subsequently, individual proteins can be qualified or validated as directed by discovery (Figure 1I).
Data from a focused analysis of in-vitro phosphorylation of a “favorite protein” (workflow comprising Figures 1A, 1C, 1D’, 1F, 1G and1H) by the protein kinase ATM is shown in Table 1. Some of the phosphorylation sites on the transcription factor MEF2D (i.e. phosphoserine, shown as “pS”, on blue background) were found in the negative controls at a similar abundance as in the ATM kinase treatment. In contrast, a phosphothreonine residue (shown as “pT”) was specifically phosphorylated by ATM (yellow background; Table 1).
Hybrid LBA-LC-HRAMS analyses are used to analyze protein complexes (workflow illustrated by Figures 1A, 1B, 1D’, 1G and 1H). Enrichment of PTMs (Figure 1F) can be included. For example, transcription factor complexes, and their phosphorylation sites, differed between muscle stem cells and more differentiated derivatives (data not shown). Many examples of hybrid LBA-LC-HRAMS analyses have been reported.
An improved understanding of pluripotent human embryonic stem cells (hESCs) and their multipotent neural stem cell derivatives (hNSCs) at the earliest stages of neural commitment required unbiased, semi-quantitative discovery of these (phospho)proteomes (Figure 2; Singec et al., 2016, Stem Cell Reports Vol. 7, p. 527).
Stephen Rundlett– Associate Director, Biomarker Program, Intertek Pharmaceutical Services, San Diego, California