Chapter 5 — Proteomics

What is Proteomics?

Proteomics is the comprehensive study of the proteome: the entire complement of proteins, including their variations and modifications, produced by an organism or system¹. This field complements genomics and transcriptomics to probe the central dogma of biology at the protein level. Understanding the complex interactions of proteins, especially in the context of gene regulation, opens opportunities to understand health and disease and improve disease diagnosis, treatment, and prevention.

Proteins are fundamental to biological systems by performing functions essential for life, including catalyzing metabolic reactions, replicating DNA, responding to stimuli, and transporting molecules from one location to another. Proteins are involved in virtually every process within a cell²: structural proteins, like collagen and keratin, provide scaffolding to organize cells and tissues; protein hormones, cytokines, and chemokines carry messages between cells and organs using protein receptors to receive these signals; and antibodies protect the body from pathogens by binding to specific targets. Proteins also regulate gene expression and the activity of other proteins, demonstrating their crucial role in homeostasis and gene regulation.

How Do We Study Proteomics?

Proteomics uses a variety of sophisticated technologies. These technologies are pivotal for analyzing the structure, function, and interactions of proteins, providing insights that are crucial for understanding cellular processes and disease mechanisms. Here are some key technologies that have played crucial roles in gaining valuable insights from proteomics:

Mass spectrometry

Mass spectrometry is the cornerstone of proteomics^3,4, enabling the identification and quantification of proteins from biological samples. It works by fragmenting and ionizing proteins, and then measuring the mass-to-charge ratio of ionized particles to determine the molecular weight of peptides and proteins. A typical workflow involves digesting proteins into peptides, which would then have a characteristic mass spectrum “fingerprint.” There are three main types of mass spectrometry:

• MALDI-TOF (Matrix-Assisted Laser Desorption/Ionization Time-of-Flight): Uses a laser to ionize peptide fragments on a substrate, creating ionized peptide fragments that reach a detector that uses the time of flight to measure mass-to-charge ratio (lower mass-to-charge ratio has a shorter time of flight)⁵. MALDI-TOF is useful for rapid identification of proteins by generating peptide mass fingerprints without interference by other biomolecules or substrates.
• ESI-MS (Electrospray Ionization Mass Spectrometry): Often coupled with liquid chromatography (LC-ESI-MS/MS), it provides detailed information about peptide sequences and post-translational modifications because it does not fragment the analyte into smaller ions⁶.
• Tandem MS (MS/MS): Allows for peptide sequencing by fragmenting peptides and deducing the amino acid sequence by analyzing the resulting fragments⁷.

Chromatography

Chromatography techniques are used to separate proteins and peptides before analysis, which often includes MS. These include:

• Liquid Chromatography (LC)⁴: Particularly useful for its ability to separate thousands of proteins and peptides in a single run.
• Ion Exchange Chromatography⁸: Separates proteins based on their charge.
• Affinity Chromatography⁹: Targets specific proteins and peptides with high specificity using ligands that bind to a particular protein structure or motif.

Imaging mass spectrometry

Imaging mass spectrometry (IMS) allows the visualization of the spatial distribution of proteins and peptides directly in biological tissues¹⁰. Variations, such as nano secondary ion MS (NanoSIMS) offer subcellular localization of metabolites¹¹. This technique has important applications in clinical research, such as mapping tumor margins or understanding disease pathology at a molecular level¹².

Two-dimensional gel electrophoresis (2D-GE)

2D-GE is a powerful technique that separates proteins in a gel matrix based on their isoelectric point and molecular weight¹³. This method is useful for visualizing protein expression differences in various samples, such as comparing diseased tissue against healthy controls. 2D Difference Gel Electrophoresis (2D-DIGE) is a variation of 2D-GE that adds fluorescent labels to two or more samples to provide an inexpensive method to compare proteomes^14,15. Points of interest in the DIGE gels can be excised and further characterized using mass spectrometry.

Protein microarrays

Protein microarrays immobilize proteins on a solid surface, allowing parallel analysis of many interactions¹⁶. These arrays can be used for antibody characterization, enzyme-substrate identification, and analyzing protein-protein interactions.

Experimental Considerations

These technologies offer unique insights into proteome under various physiological and pathological conditions. The choice of techniques will depend on the biological question, as their cost, resolution, sensitivity, preparation, and analysis requirements will vary. Techniques can be combined to manage the depth and breadth of insights. For example, mass spectrometry is frequently used to identify interactors of a specific protein, or global changes in response to a perturbation.

However, MS requires large amounts of starting protein, and requires bioinformatics expertise to run and analyze. 2D gel electrophoresis, on the other hand, can be used to quickly and inexpensively get a snapshot of the proteome; comparing two proteomes using DIGE can help focus on specific differences, but would not be sensitive to low abundance proteins. IMS is a fairly new technique that is continuously advancing, now able to provide organelle-localized analysis of cellular metabolites, allowing deeper probing of the cell’s inner workings.

Modern Uses of Proteomics in Multiomics: Case Studies

Metabolon has worked with several customers and collaborators to perform a variety of multiomics research studies across a wide range of topics. Below we discuss some key case studies showcasing the use of proteomics and metabolomics along with other omics data.

Multiomics data integration reveals metabolome as the top predictor of the cervicovaginal microenvironment

In a large-scale multiomics investigation combining microbiomics, metabolomics, and immunoproteomics data, researchers explored the complex interactions occurring within the cervicovaginal microenvironment in women with cervical neoplasia¹⁷. Integrating these datasets, they developed predictive models for vaginal health. Their work, published in PLoS Computational Biology, used cervicovaginal samples collected from 72 women, with invasive cervical carcinoma (ICC [n=8]), low and high-grade intraepithelial lesions (HSIL [n=27], LSIL [n=10]), and HPV-positive (n=18) and HPV-negative (n=9) controls (Figure 1).

Machine learning techniques (neural networks and random forest classifiers) were used to analyze microbiome, metabolome, proteome, and vaginal pH data to identify potential cancer and immune predictive biomarkers. Among their results, they found lipids (e.g. sphingolipids and long-chain unsaturated fatty acids) to be strong predictors of genital inflammation. In contrast, amino acid metabolites were predictors of vaginal microbiome composition and vaginal pH. Key immune proteins (IL-6, IL-10, MIP-1α) were strongly associated with vaginal microbiota, vaginal pH (MIF), and genital inflammation.

Integrating microbiome, metabolome, and immunoproteome data generally offered more accurate models than a single omics modality, underscoring the benefit of comprehensive multiomics. This comprehensive approach provided a holistic view of the cervicovaginal microenvironment, revealing complex interconnections between different biological layers that can serve as biomarkers for cervical cancer diagnosis and treatment.

Proteomic and metabolomic characterization of COVID-19 patient sera

In a study published in Cell, researchers combined proteomics and metabolomics data to better understand COVID-19 disease severity (Figure 2)¹⁸. Collecting sera samples from 65 patients with severe (n=28) or non-severe (n=37) COVID-19, patients afflicted with non-COVID-19 respiratory illnesses (n=25), and healthy controls (n=28) for proteomics and metabolomics analysis, the researchers identified dysregulation of macrophage function, platelet degranulation, and the complement system, and metabolic suppression in patients with COVID-19. Additionally, 93 proteins and 204 metabolites were correlated with disease severity.

The researchers trained a random forest machine learning model on proteomic and metabolic data (22 proteins and 7 metabolites) from 31 COVID-19 patients, 13 of which had severe COVID-19 infection. Validation of this model on independent cohorts revealed that it could correctly stratify patients based on disease severity in almost 85% of test cases. The authors hypothesized that many of these biomarkers are involved in the immune system or are indicative of liver damage that may precede overt symptoms of severe COVID-19. The authors also found that these biomarkers were prescient of severe COVID-19 infections up to four days before severe symptoms manifested¹⁰.

Development of a multiomics model for identification of predictive biomarkers for COVID-19 severity: a retrospective cohort study

In a study published in Lancet Digital Health, researchers combined metabolomics, lipidomics, and proteomics to identify predictive biomarkers of COVID-19-associated complications¹⁹. In a cohort of 455 COVID-19 patients and 182 controls, the researchers retrospectively profiled circulating cytokines and other proteins, lipids, and metabolites.

Using advanced machine learning models on the data, the researchers identified 102 biomarkers that effectively predicted severe clinical outcomes in COVID-19. Some, like C-type lectin domain family 6 member A (CLEC6A), ether phosphatidylethanolamine (P-18:1/18:1), and 2-hydroxydecanoate (Figure 3), are novel biomarkers that have not previously been associated with COVID-19 severity. These discoveries emphasize the potential of multiomics approaches to reveal intricate details about disease mechanisms that traditional cytokine panels cannot provide.

Not only did integration of multiple omics datasets enable the identification of interconnected biological pathways and processes, but the application of machine learning algorithms led to the creation of predictive models that outperformed those based on traditional markers alone, illustrating the power of combining diverse biological data as multiomics analyses for enhanced predictive accuracy and deeper biological insights.

Proteomics in Multiomics Research

These case studies demonstrated the importance of proteomics as an essential component of multiomics research to provide a holistic understanding of biological systems and disease processes. Multiomics already plays a key role in advancing personalized medicine, and will likely continue doing so, especially as technology improves.

References

1. Al-Amrami S, Al-Jabri Z, Al-Zaabi A, et al. Proteomics: Concepts and applications in human medicine. World J Biol Chem. 2021;12(5): 57–69. doi: 10.4331/wjbc.v12.i5.57
2. Alberts B, Johnson A, Lewis J, et al. Molecular Biology of the Cell. 4th edition. New York: Garland Science; 2002. Analyzing Protein Structure and Function. Available from: https://www.ncbi.nlm.nih.gov/books/NBK26820/
3. Han X, Aslanian A, and Yates III JR. Mass Spectrometry for Proteomics. Curr Opin Chem Biol. 2008;12(5): 483–490. doi: 10.1016/j.cbpa.2008.07.024
4. Parker CE, Warren MR, Mocanu V. Mass Spectrometry for Proteomics. In: Alzate O, editor. Neuroproteomics. Boca Raton (FL): CRC Press/Taylor & Francis; 2010. Chapter 5. Available from: https://www.ncbi.nlm.nih.gov/books/NBK56011/
5. Li D, Yi J, Han G, Qiao L. MALDI-TOF mass spectrometry in clinical analysis and research. ACS Meas Sci Au. 2022;2(5):385-404. doi:10.1021/acsmeasuresciau.2c00019
6. Pitt JJ. Principles and applications of liquid chromatography-mass spectrometry in clinical biochemistry. Clin Biochem Rev. 2009;30(1):19-34.
7. Neagu AN, Jayathirtha M, Baxter E, Donnelly M, Petre BA, Darie CC. Applications of tandem mass spectrometry (MS/MS) in protein analysis for biomedical research. Molecules. 2022;27(8):2411. doi:10.3390/molecules27082411
8. Jungbauer A and Hahn R. Ion-exchange chromatography. Methods Enzymol. 2009:463:349-71. doi: 10.1016/S0076-6879(09)63022-6
9. Urh M, Simpson D, and Zhao K. Affinity chromatography: general methods. Methods Enzymol. 2009:463:417-38. doi: 10.1016/S0076-6879(09)63026-3
10. McDonnell LA, Heeren RMA. Imaging mass spectrometry. Mass Spectrometry Reviews. 2007;26(4):606-643. doi:10.1002/mas.20124
11. Siuzdak G. Subcellular quantitative imaging of metabolites at the organelle level. Nat Metab. 2023;5(9):1446-1448. doi:10.1038/s42255-023-00882-z
12. Arentz G, Mittal P, Zhang C, et al. Applications of mass spectrometry imaging to cancer. In: Advances in Cancer Research. Vol 134. Elsevier; 2017:27-66. doi:10.1016/bs.acr.2016.11.002
13. Adams LD and Gallagher SR. Two-dimensional gel electrophoresis. Curr Protoc Mol Biol. 2004 Sep:Chapter 10:Unit 10.4. doi: 10.1002/0471142727.mb1004s67
14. Kondo T. Cancer biomarker development and two-dimensional difference gel electrophoresis (2D-DIGE). Biochim Biophys Acta Proteins Proteom. 2019;1867(1):2-8. doi:10.1016/j.bbapap.2018.07.002
15. Meleady P. Two-Dimensional Gel Electrophoresis and 2D-DIGE. Methods Mol Biol. 2023;2596:3-15. doi:10.1007/978-1-0716-2831-7_1
16. Reymond Sutandy FX, Qian J, Chen C-S, et al. Overview of Protein Microarrays. Curr Protoc Protein Sci. 2013; 72(1): 2711–27116. doi: 10.1002/0471140864.ps2701s72
17. Bokulich NA, Łaniewski P, Adamov A, Chase DM, Caporaso JG, Herbst-Kralovetz MM. Multi-omics data integration reveals metabolome as the top predictor of the cervicovaginal microenvironment. Segata N, ed. PLoS Comput Biol. 2022;18(2):e1009876. doi:10.1371/journal.pcbi.1009876
18. Shen B, Yi X, Sun Y, et al. Proteomic and metabolomic characterization of COVID-19 patient sera. Cell. 2020;182(1):59-72.e15. doi:10.1016/j.cell.2020.05.032
19. Byeon SK, Madugundu AK, Garapati K, et al. Development of a multiomics model for identification of predictive biomarkers for COVID-19 severity: a retrospective cohort study. The Lancet Digital Health. 2022;4(9):e632-e645. doi:10.1016/S2589-7500(22)00112-1