A perspective on proteomics in cell biology

Abstract

During the past 15 years mass spectrometry (MS)-based analyses have become established as the method of choice for direct protein identification and measurement. Owing to the remarkable improvements in the sensitivity and resolution of MS instruments, this technology has revolutionised the opportunities available for the system-wide characterisation of proteins, with wide applications across virtually the whole of cell biology. In this article we provide a perspective on the current state of the art and discuss how the future of cell biology research may benefit from further developments and applications in the field of MS and proteomics, highlighting the major challenges ahead for the community in organising the effective sharing and integration of the resulting data mountain.

Figure 1

Bottom-up proteomics workflow. ‘Proteomics’ is used throughout as an umbrella term for the large-scale identification and analysis of proteins. We focus here specifically on the analysis of proteins by mass spectrometry (MS), because this has emerged as by far the most widely used and efficient current technology. The standard ‘bottom-up’ proteomics workflow illustrated involves isolating proteins from either cells or tissues, digesting them to peptides using one or more proteases (e.g., trypsin), then separating the resulting peptide mixtures by nano-LC (nano-liquid chromatography) and identifying the peptides in a mass spectrometer. The resulting peptide identifications are subsequently mapped to proteins, using genomic information to identify open reading frames (ORFs) that encode these peptides. Although the MS analysis actually measures peptides, most subsequent data analysis in cell biology experiments interprets the results in terms of the inferred protein identifications, and the quality of the data can vary according to how many peptides were detected for each protein. Usually a minimum of at least two separate peptides are required to confirm protein identification.

Figure 2

Isotope labelling strategies. Isotope labelling methods such as SILAC (stable isotope labelling with amino acids in cell culture) and iTRAQ (isobaric tag for relative and absolute quantitation) provide a convenient approach for the quantitative proteomic comparison of two or more experimental variables by introducing tags that can be discriminated in the mass spectrometer to distinguish and measure the proteins from each separate condition or cell sample. (A) Isotope labelling is a highly flexible strategy that can be adapted to identify and compare protein interaction partners, subcellular protein localisation, drug treatment, viral infection, and effects of genotype etc. (B) A multispectral image (MSI) spectrum of a peptide selected in a triple SILAC experiment. In the illustrated example the m/z signal from the peptide (x axis) is separated in the spectrum into three clusters of signals, corresponding to the ‘light’, ‘medium’, and ‘heavy’ isotopic forms. The measured ion intensities (y axis) for each isotopic form of the peptide are then compared, and these reflect the corresponding property of the protein from which the peptide was derived in each cell state. In most cases, data are subsequently represented by averaging the separate values measured for all peptides identified for each protein.

Figure 3

Approaches to protein interaction analysis. (A) Isotope labelling (e.g., SILAC) has been widely used as a method to discriminate reliably between specific and non-specific protein interaction partners in immunoprecipitation and affinity pull-down experiments. By measuring the ratio of light (control – e.g., non-specific Ig or no bait protein) to heavy (e.g., specific Ab or tagged bait protein) isotope-labelled proteins that are copurified and detected by MS analysis, proteins that bind non-specifically will typically have 1:1 H/L ratios whereas specific interaction partners will have high H/L ratios. Using a triple-labelling strategy this can be extended, for example, by using the comparison of M/L and H/L isotope ratios to compare specific binding either to different protein isoforms or mutants, or to compare binding in the presence or absence of an inhibitor, etc. (B) A high-throughput proteomics strategy for analysing protein complexes by (i) first separating protein complexes present in cell extracts using a chromatography method such as SEC, then (ii) detecting essentially all of the proteins in each resultant SEC fraction following protease digestion and mass spectrometry. This highlights candidate protein components of complexes based upon coelution profiles across the SEC fractions, and can potentially distinguish complexes containing either specific protein isoforms and/or PTMs. It can be applied to the system-wide comparison of differences in cellular protein complexes under different growth conditions, or following drug treatments or other perturbations. (C) An example of a protein, TRXR1_HUMAN, which has multiple isoforms showing differing profiles of interaction. Abbreviations: Ab, antibody; H/L, heavy/light; Ig, immunoglobulin; M/L, medium/light; PTMs, post-translational modifications; SEC, size exclusion chromatography; SILAC, stable isotope labelling with amino acids in cell culture.

Figure 4

Data integration and resources for online sharing. The figure illustrates the need to link highly annotated, multi-dimensional proteomics data with information from large-scale, genomic and transcriptomic sequences and associated literature. Four publicly available online tools for sharing data related to proteomics are illustrated – STRING (Search Tool for the Retrieval of Interacting Genes), DAVID (Database for Annotation, Visualisation and Integrated Discovery), The Human Protein Atlas, and the Encyclopedia of Proteome Dynamics.