Budayeva and Kirkpatrick argue in their review Monitoring protein communities and their responses to therapeutics (Nature Reviews Drug Discovery, 2020) that mass spectrometry help us decipher complex relations between protein communities and understand how disease and drug influence them.

Protein communities

I understand the term ‘protein communities’ as a generalized term of gene-gene interaction, such as direct protein-protein interaction, two proteins being co-regulated, or simply co-localized. The authors define a protein community as a set of proteins within a cell that are interconnected through one or more physical, spatial, co-regulatory or functional relationships. Though apparently extracellular protein-protein interactions may be considered as well.

Experimental approaches to study proteomics

Mass spectrometry has been used in drug discovery mainly for following purposes:

  • Global proteome profiling: estimating abundance of all proteins that can be measured.
  • PTM profiling: understanding post-translational modifications (PTMs), for instance to understand the mechanism of drug resistance or to elucidate signalling events following target engagement.
  • Activity-based protein profiling (ABPP): a technology that uses chemical probes that react with enzymes. The probe consists of two elements most of the time: a reactive group (known as a ‘warhead’), and a tag. Sometimes the probe may contain a binding group that enhances selectivity. The warhead contains a electrophile which becomes covalently linked to a nucleophilic residue in the active site of an active enzyme. If the enzyme is inhibited or post-translationally modified on the residue targetted by the probe, it will not react with the probe. The technique, when used in combination with mass spectrometry, can reveal for instance the selectivity of a kinase inhibitor.
  • Chemoproteomics studies: proteomics studies using chemical probes to reveal small-molecule-protein interactions.

The authors highlighted a few approaches for disease understanding and drug discovery. They include affinity purification, proximity labelling, organelle proteome profiling, PTM profiling, chemoaffinity enrichment, and thermal proteome profiling.

  • Affinity purification, also known as immunoaffinity purification, identifies protein interactions by antibody-mediated isolation of a target protein with its interaction partners.
  • Proximity labelling, as its name suggests, labels a protein with a reactive substrate so that proteins in proximity are captured by affinity purification. It can capture transient and functional interactions.
  • Organelle proteome profiling identifies spatial protein distribution and expression patterns of proteins enriched in organelle fractions.
  • Post-transcriptional modification (PTM) profiling enriches and quantifies modified peptides (for instance phosphorylated peptides). It offers information of physical interaction, spatial interaction, and co-regulatory interaction.
  • Chemoaffinity enrichment uses a chemical tool compound (for instance a kinase inhibitor) to profile biochemical activities of proteins in the presence and absence of a drug (for instance another kinase inhibitor of interest). The competitive binding of two compounds to proteins offers information about expression levels, activity, and affinity of an isolated enzyme for the drug.
  • Thermal proteome profiling uses the fact that protein thermal stability changes upon configuration change induced by binding to specific binding partners such as small molecules. It can be used for target identification.

Practical considerations

The authors listed many examples, such as ERBB/EGFR receptors, GPCRs, and PD-L1, to illustrate the use of proteomics in disease understanding for drug mechanism-of-aciton investigation.

The authors discussed common methods of quantification in mass spectrometry-based proteomics. They also suggested that it is necessary to optimize time courses and compound concentration of proteomics experiment on the basis of the model system and on the biological activity of the drug. For RTK (receptor tyrosine kinase)-mediated phosphorylation events, detection can be done seconds after stimulation, but typically profiling is done within 2-10 minutes after ligand binding. Lysine acetylation and ubiquitylation can also be observed in the range of minutes. TNF-mediated NF-kB translocation happen in waves after 1, 3 an 6h.

Outlook

The authors believe that future progress is expected in in vivo proteomics, automation, stoichiometric relationships between proteins, and improve proteomic methods for low-level clinical samples. For drug discovery, connecting protein communities with phenotypes is important, for instance by using proteomics to understand the full sequence of events following target engagement and the connections of these events to efficacy, adverse effects, and resistance.

Conclusions

The paper summarizes diverse applications of mass-spectrometry proteomics in disease understanding, and possibilities of using them to elucidate mechanisms of action of drugs on the protein level.

The paper is already comphrehensive. Nevertheless, I would love to see more industrial examples of using proteomics to guide leade identification, optimization, and selection. In addition, probably due to the length limitation, the review does discuss analysis of preoteomics data in detail. Interested users may find the review Bioinformatic analysis of proteomics data by Schmidt et al. and the publication An Assessment of Software Solutions for the Analysis of Mass Spectrometry Based Quantitative Proteomics Data by Müller et al. informative and complementary to this review.

P.S. Some interesting statistics:

  • A single human cell contains an estimated 10 billion (10 to the power of 10) individual proteins.
  • Protein abundance ranges from one to upward of 10 million copies per cell.