Proteomics and Mass Spectrometry
of Biomolecules

Developments

PTMs play a major role in protein function and in cellular processes regulation. More than two hundred types of protein modifications have been described, most important ones including phosphorylation, glycosylation, acetylation, and ubiquitylation. PTMs affect protein physicochemical properties, which can modulate protein activity, stability, localization, association to other molecules, and thus protein function. PTMs are usually covalent modifications of specific amino acids, presenting a wide range of molecular mass and stability. Moreover, they are often present in substoichiometric quantities. The protein mass change resulting from the addition or deletion of a PTM can be measured using mass spectrometry (MS) and the localization of the PTM in the protein sequence can be determined from tandem MS (MS/MS) analyses of modified peptides after enzymatic proteolysis. In the last decade, MS has been very successful in characterizing PTMs. The sensitivity, mass accuracy, and high sequencing speed of modern mass spectrometers further improve today the ability of MS to characterize PTMs. Nevertheless, challenges still remain, that are obtaining a full protein sequence coverage, localizing a labile modification (phosphorylation, glycosylation), detecting a small fraction of modified peptide mixed with a large excess of corresponding unmodified peptide, quantitatively comparing samples to address dynamic issues, analyzing complex samples corresponding to a class of post-translationally modified proteins, like a phosphoproteome.

It is now clear that proteins mainly act in association with other proteins in a dynamic way to reach a tight control of cellular processes. One of the most important challenges for proteomics in the analysis of protein-protein interactions is thus to determine the variations in protein composition and abundance of protein complexes either in different cell types or after inducing a protein expression change, or over time after applying a stimulus. The purification or enrichment of protein complexes of interest is a critical step before MS analysis to obtain meaningful data, one of the main difficulties being to distinguish true interactors from contaminants. In this project, different strategies have been successfully applied to the analysis of protein complexes from various organisms.

Part 1: Strategies using an affinity purification of endogeneous protein complexes
The affinity purification of endogeneous protein complexes requires the use of efficient and specific antibodies. The choice of an appropriate control experiment in these studies is then very important to properly identify specific partners, as cross reactivity can occur with the antibody. This strategy allowed the study of several protein complexes, mainly in cancer cells, that are involved in different cellular processes. New protein interactions have been identified by MS and validated by other biochemical approaches.

Part 2: Strategies using a tandem purification method with tagged proteins
This strategy can be applied to study proteins in cells, where a tag can be introduced at the end of the protein of interest. This tag is then used to affinity purify very specifically the protein and its partners. To increase further the specificity, two affinity purification steps using two different tags can be conducted. The tandem purification is particularly efficient in yeast and allowed the characterization of several ribosomal complexes, leading to a better understanding of these protein machines. This strategy has also been applied to membrane proteins in mammalian cells and allowed for the first time the purification of G protein-coupled receptor-associated proteins under native conditions.

Part 3: Strategies based on BIA-MS coupling
Biomolecular Interactions Analysis (BIA) can be used to detect the interactions occurring between a molecule immobilized on a chip and proteins contained in a solution flowing through the chip. To identify the proteins associated with the bait on the chip, we optimized their recovery and their subsequent MS analysis. The main advantage of this strategy is the low amount of material required. It has then been successfully applied to the identification of proteins interacting with several types of receptors.

On-going and future studies of protein complexes will make use of quantitative proteomic approaches and will also require the development of new strategies to better characterize these complexes

Even though considerable progress have been made in the analysis of very complex mixtures by nanoLC-MS/MS based proteomic approaches, low-abundance proteins remain very difficult to detect. Moreover, these low-abundance proteins are often the ones of most biological interest. Their identification is mainly limited by the acquisition speed and the dynamic range of mass spectrometers. To overcome these limitations, extensive sample fractionation and reduction of the sample dynamic range can be performed. Clinical applications of proteomics for biomarker discovery in body fluids add another difficulty, as the vast majority of the protein content is represented by only one or a few proteins, which will hinder the detection of other proteins in the sample.

Part 1: The equalization strategy
A new strategy based on the use of beads coated with thousands of peptides obtained from combinatorial chemistry (Proteominer beads developed by BioRad) has been evaluated. It allows the interaction of all proteins in the sample with the beads in similar proportions and thus dramatically decreases the dynamic range of protein abundances in the sample. This strategy has been efficiently used for capturing and revealing a very large population of previously undetected proteins in human erythrocytes, containing large amounts of haemoglobins. The equalization strategy thus appears as a powerful tool to go much deeper into the proteome characterization and to increase very significantly the number of proteins identified by nanoLC-MS/MS analyses.

Part 2: Clinical applications for biomarker discovery in body fluids
On-going and future programs using the equalization strategy will aim at identifying biomarkers in body fluids, like urine and cerebrospinal fluid (CSF). Further development of the equalization strategy will be needed to handle low amounts of CSF. These studies will also lead to the analysis and comparison of numerous biological samples, which will require the development of specific analytical strategies associated to dedicated bioinformatic tools.

Recent progress in proteomics has generated a huge amount of MS and MS/MS data, which have to be handled for protein identification and quantification. This can only be performed automatically in several steps by appropriate bioinformatic tools. Protein identification requires stringent validation criteria and protein quantification has to fit all existing MS-based quantitative approaches as well as to evolve rapidly to accommodate new analytical strategies.

Part 1: Strategies based on isotopic labeling
Quantitative proteomic strategies based on isotopic labeling (ICAT, SILAC, 14N/15N, iTRAQ) are the most robust MS-based methods developed so far. They allow the fine quantification of protein abundance variations between two or a few samples. Thanks to the introduction of light or heavy isotopes in the samples, these can be distinguished after being mixed. Samples are then analyzed by nanoLC-MS/MS in a single run, allowing a precise quantification. We have developed a new open-source software able to handle this type of MS data, MFPaQ. Being initially compatible with one data format and two quantitative methods, it has been improved to handle several types of isotopic labels and data acquisition on different MS instruments.

Part 2: Label-free approaches
The comparison of numerous samples, like required in clinical applications, favored the development of label-free approaches. Here, each sample is analyzed independently and protein abundance variations are evaluated by peak intensity measurements from run to run. This requires a very high repeatability of the analytical procedure and dedicated tools to handle the comparison of all MS data generated. Preliminary results obtained with modern high resolution instruments are promising and we will develop these approaches, including the necessary bioinformatic tools.