Prediction of high-responding peptides for targeted protein assays by mass spectrometry. Nat Biotechnol. 2009;27:190-198. [PMID: 19169245 DOI: 10.1038/nbt.1524]

Is a gene-centric human proteome project the best way for proteomics to serve biology?

With the recent developments in proteomic technologies, a complete human proteome project (HPP) appears feasible for the first time. However, there is still debate as to how it should be designed and what it should encompass. In "proteomics speak", the debate revolves around the central question as to whether a gene-centric or a protein-centric proteomics approach is the most appropriate way forward. In this paper, we try to shed light on what these definitions mean, how large-scale proteomics such as a HPP can insert into the larger omics chorus, and what we can reasonably expect from a HPP in the way it has been proposed so far.

Combination of improved (18)O incorporation and multiple reaction monitoring: a universal strategy for absolute quantitative verification of serum candidate biomarkers of liver cancer

Stable isotope dilution-multiple reaction monitoring-mass spectrometry (SID-MRM-MS), which is an alternative to immunoassay methods such as ELISA and Western blotting, has been used to alleviate the bottlenecks of high-throughput verification of biomarker candidates recently. However, the inconvenience and high isotope consumption required to obtain stably labeled peptide impedes the broad application of this method. In our study, the (18)O-labeling method was introduced to generate stable isotope-labeled peptides instead of the Fmoc chemical synthesis and Qconcat recombinant protein synthesis methods. To make (18)O-labeling suitable for absolute quantification, we have added the following procedures: (1) RapiGest SF and microwave heating were added to increase the labeling efficiency; (2) trypsin was deactivated completely by chemical modification using tris(2-carboxyethyl)phosphine (TCEP) and iodoacetamide (IAA) to prevent back-exchange of (18)O to (16)O, and (3) MRM parameters were optimized to maximize specificity and better distinguish between (18)O-labeled and unlabeled peptides. As a result, the (18)O-labeled peptides can be prepared in less than 1 h with satisfactory efficiency (>97%) and remained stable for 1 week, compared to traditional protocols that require 5 h for labeling with poor stability. Excellent separation of (18)O-labeled and unlabeled peptides was achieved by the MRM-MS spectrum. Finally, through the combined improvement in (18)O-labeling with multiple reaction monitoring, an absolute quantification strategy was developed to quantitatively verify hepatocellular carcinoma-related biomarker candidates, namely, vitronectin and clusterin, in undepleted serum samples. Sample preparation and capillary-HPLC analysis were optimized for high-throughput applications. The reliability of this strategy was further evaluated by method validation, with accuracy (%RE) and precision (%RSD) of less than 20% and good linearity (r(2) > 0.99), and clinical validation, which were consistent with previously reported results. In summary, our strategy can promote broader application of SID-MRM-MS for biomarkers from discovery to verification regarding the significant advantages of the convenient and flexible generation of internal standards, the reduction in the sample labeling steps, and the simple transition.

The necessity of adjusting tests of protein category enrichment in discovery proteomics

MOTIVATION

Enrichment tests are used in high-throughput experimentation to measure the association between gene or protein expression and membership in groups or pathways. The Fisher's exact test is commonly used. We specifically examined the associations produced by the Fisher test between protein identification by mass spectrometry discovery proteomics, and their Gene Ontology (GO) term assignments in a large yeast dataset. We found that direct application of the Fisher test is misleading in proteomics due to the bias in mass spectrometry to preferentially identify proteins based on their biochemical properties. False inference about associations can be made if this bias is not corrected. Our method adjusts Fisher tests for these biases and produces associations more directly attributable to protein expression rather than experimental bias.

RESULTS

Using logistic regression, we modeled the association between protein identification and GO term assignments while adjusting for identification bias in mass spectrometry. The model accounts for five biochemical properties of peptides: (i) hydrophobicity, (ii) molecular weight, (iii) transfer energy, (iv) beta turn frequency and (v) isoelectric point. The model was fit on 181 060 peptides from 2678 proteins identified in 24 yeast proteomics datasets with a 1% false discovery rate. In analyzing the association between protein identification and their GO term assignments, we found that 25% (134 out of 544) of Fisher tests that showed significant association (q-value ≤0.05) were non-significant after adjustment using our model. Simulations generating yeast protein sets enriched for identification propensity show that unadjusted enrichment tests were biased while our approach worked well.

Highly reproducible label free quantitative proteomic analysis of RNA polymerase complexes

The use of quantitative proteomics methods to study protein complexes has the potential to provide in-depth information on the abundance of different protein components as well as their modification state in various cellular conditions. To interrogate protein complex quantitation using shotgun proteomic methods, we have focused on the analysis of protein complexes using label-free multidimensional protein identification technology and studied the reproducibility of biological replicates. For these studies, we focused on three highly related and essential multi-protein enzymes, RNA polymerase I, II, and III from Saccharomyces cerevisiae. We found that label-free quantitation using spectral counting is highly reproducible at the protein and peptide level when analyzing RNA polymerase I, II, and III. In addition, we show that peptide sampling does not follow a random sampling model, and we show the need for advanced computational models to predict peptide detection probabilities. In order to address these issues, we used the APEX protocol to model the expected peptide detectability based on whole cell lysate acquired using the same multidimensional protein identification technology analysis used for the protein complexes. Neither method was able to predict the peptide sampling levels that we observed using replicate multidimensional protein identification technology analyses. In addition to the analysis of the RNA polymerase complexes, our analysis provides quantitative information about several RNAP associated proteins including the RNAPII elongation factor complexes DSIF and TFIIF. Our data shows that DSIF and TFIIF are the most highly enriched RNAP accessory factors in Rpb3-TAP purifications and demonstrate our ability to measure low level associated protein abundance across biological replicates. In addition, our quantitative data supports a model in which DSIF and TFIIF interact with RNAPII in a dynamic fashion in agreement with previously published reports.

Immunoaffinity enrichment and liquid chromatography-selected reaction monitoring mass spectrometry for quantitation of carbonic anhydrase 12 in cultured renal carcinoma cells

Liquid chromatography-selected reaction monitoring (LC-SRM) is a highly specific and sensitive mass spectrometry (MS) technique that is widely being applied to selectively qualify and validate candidate markers within complex biological samples. However, in order for LC-SRM methods to take on these attributes, target-specific optimization of sample processing is required, in order to reduce analyte complexity, prior to LC-SRM. In this study, we have developed a targeted platform consisting of protein immunoaffinity enrichment on magnetic beads and LC-SRM for measuring carbonic anhydrase 12 (CA12) protein in a renal cell carcinoma (RCC) cell line (PRC3), a candidate biomarker for RCC whose expression at the protein level has not been previously reported. Sample processing and LC-SRM assay were optimized for signature peptides selected as surrogate markers of CA12 protein. Using LC-SRM coupled with stable isotope dilution, we achieved limits of quantitation in the low fmol range sufficient for measuring clinically relevant biomarkers with good intra- and interassay accuracy and precision (≤17%). Our results show that using a quantitative immunoaffinity capture approach provides specific, accurate, and robust assays amenable to high-throughput verification of potential biomarkers.

Targeted proteomics by selected reaction monitoring mass spectrometry: applications to systems biology and biomarker discovery

Mass Spectrometry-based proteomics is now considered a relatively established strategy for protein analysis, ranging from global expression profiling to the identification of protein complexes and specific post-translational modifications. Recently, Selected Reaction Monitoring Mass Spectrometry (SRM-MS) has become increasingly popular in proteome research for the targeted quantification of proteins and post-translational modifications. Using triple quadrupole instrumentation (QqQ), specific analyte molecules are targeted in a data-directed mode. Used routinely for the quantitative analysis of small molecular compounds for at least three decades, the technology is now experiencing broadened application in the proteomics community. In the current review, we will provide a detailed summary of current developments in targeted proteomics, including some of the recent applications to biological research and biomarker discovery.

Targeted data acquisition for improved reproducibility and robustness of proteomic mass spectrometry assays

Quantitative mass spectrometry-based proteomic assays often suffer from a lack of robustness and reproducibility. We here describe a targeted mass spectrometric data acquisition strategy for affinity enriched subproteomes-in our case the kinome-that enables a substantially improved reproducibility of detection, and improved quantification via isobaric tags. Inclusion mass lists containing m/z, charge state, and retention time were created based on a set of 80 shotgun-type experiments performed under identical experimental conditions. For each target protein, peptides were selected according to their frequency of observation and isobaric tag for relative and absolute quantitation (iTRAQ) reporter ion quality. Retention times of selected peptides were aligned using similarity driven pairwise alignment strategy yielding <1 min standard deviation for 4 h gradients. Multiple fragmentation of the same peptides resulted in better statistics and more precise reporter ion based quantification without any loss in coverage. Overall, 24% more target proteins were quantified using the targeted data acquisition approach, and precision of quantification improved by >1.5-fold. We also show that a combination of higher energy collisional dissociation (HCD) with collisional induced dissociation (CID) outperformed pulsed-Q-dissociation (PQD) on the OrbitrapXL. With the CID/HCD based targeted data acquisition approach 10% more quantifiable target proteins were identified and a 2-fold increase in quantification precision was achieved. We have observed excellent reproducibility between different instruments, underlining the robustness of the approach.

Quantitation of the ribosomal protein autoregulatory network using mass spectrometry

Relative levels of ribosomal proteins were quantified in crude cell lysates using mass spectrometry. A method for quantifying cellular protein levels using macromolecular standards is presented that does not require complex sample separation, identification of high-responding peptides, affinity purification, or postgrowth modifications. Perturbations in ribosomal protein levels by overexpression of individual proteins correlate to known autoregulatory mechanisms and extend the network of ribosomal protein regulation.

Reducing the babel in plant volatile communication: using the forest to see the trees

While plants of a single species emit a diversity of volatile organic compounds (VOCs) to attract or repel interacting organisms, these specific messages may be lost in the midst of the hundreds of VOCs produced by sympatric plants of different species, many of which may have no signal content. Receivers must be able to reduce the babel or noise in these VOCs in order to correctly identify the message. For chemical ecologists faced with vast amounts of data on volatile signatures of plants in different ecological contexts, it is imperative to employ accurate methods of classifying messages, so that suitable bioassays may then be designed to understand message content. We demonstrate the utility of 'Random Forests' (RF), a machine-learning algorithm, for the task of classifying volatile signatures and choosing the minimum set of volatiles for accurate discrimination, using data from sympatric Ficus species as a case study. We demonstrate the advantages of RF over conventional classification methods such as principal component analysis (PCA), as well as data-mining algorithms such as support vector machines (SVM), diagonal linear discriminant analysis (DLDA) and k-nearest neighbour (KNN) analysis. We show why a tree-building method such as RF, which is increasingly being used by the bioinformatics, food technology and medical community, is particularly advantageous for the study of plant communication using volatiles, dealing, as it must, with abundant noise.

Synthetic peptide arrays for pathway-level protein monitoring by liquid chromatography-tandem mass spectrometry

Effective methods to detect and quantify functionally linked regulatory proteins in complex biological samples are essential for investigating mammalian signaling pathways. Traditional immunoassays depend on proprietary reagents that are difficult to generate and multiplex, whereas global proteomic profiling can be tedious and can miss low abundance proteins. Here, we report a target-driven liquid chromatography-tandem mass spectrometry (LC-MS/MS) strategy for selectively examining the levels of multiple low abundance components of signaling pathways which are refractory to standard shotgun screening procedures and hence appear limited in current MS/MS repositories. Our stepwise approach consists of: (i) synthesizing microscale peptide arrays, including heavy isotope-labeled internal standards, for use as high quality references to (ii) build empirically validated high density LC-MS/MS detection assays with a retention time scheduling system that can be used to (iii) identify and quantify endogenous low abundance protein targets in complex biological mixtures with high accuracy by correlation to a spectral database using new software tools. The method offers a flexible, rapid, and cost-effective means for routine proteomic exploration of biological systems including "label-free" quantification, while minimizing spurious interferences. As proof-of-concept, we have examined the abundance of transcription factors and protein kinases mediating pluripotency and self-renewal in embryonic stem cell populations.

Quantitative analysis of low-abundance peptides in HeLa cell cytoplasm by targeted liquid chromatography/mass spectrometry and stable isotope dilution: emphasising the distinction between peptide detection and peptide identification

We present the application of a targeted liquid chromatography/mass spectrometry (LC/MS) approach developed on a linear ion trap for the evaluation of the abundance of cytoplasmic proteins from a HeLa cell extract. Using a standard data-dependent approach, we identified some specific peptides from this extract which were also commercially available in their AQUA form (use for absolute quantitation). For some of the peptides, we observed a non-linear response between the intensity and the added quantity which was then fitted using a quadratic fit. All AQUA peptides spiked into a mix of 3 microg of the HeLa cell digest extract were detected down to 16 fmol. We placed an emphasis on peptide detection which, in this study, is performed using a combination of properties such as three specific Q3-like ion signatures (for a given Q1-like selection) and co-elution with the AQUA peptide counterparts. Detecting a peptide without necessarily identifying it using a search engine imposes less constraint in terms of tandem mass (MS/MS) spectra purity. An example is shown where a peptide is detected using those criteria but could not be identified by Mascot due to its lower abundance. To complement this observation, we used a cross-correlation analysis approach in order to separate two populations of MS/MS fragments based on differences in their elution patterns. Such an approach opens the door to new strategies to analyse lower intensity peptide fragments. An in silico analysis of the human trypsinosome allows the evaluation of how unique are the sets of features that we are using for peptide detection.

Proteomics data repositories

The field of proteomics, particularly the application of MS analysis to protein samples, is well established and growing rapidly. Proteomic studies generate large volumes of raw experimental data and inferred biological results. To facilitate the dissemination of these data, centralized data repositories have been developed that make the data and results accessible to proteomic researchers and biologists alike. This review of proteomics data repositories focuses exclusively on freely available, centralized data resources that disseminate or store experimental MS data and results. The resources chosen reflect a current "snapshot" of the state of resources available with an emphasis placed on resources that may be of particular interest to yeast researchers. Resources are described in terms of their intended purpose and the features and functionality provided to users.

Basic design of MRM assays for peptide quantification

With the recent availability and accessibility of mass spectrometry for basic and clinical research, the requirement for stable, sensitive, and reproducible assays to specifically detect proteins of interest has increased. Multiple reaction monitoring (MRM) or selective reaction monitoring (SRM) is a highly selective, sensitive, and robust assay to monitor the presence and amount of biomolecules. Until recently, MRM was typically used for the detection of drugs and other biomolecules from body fluids. With increased focus on biomarkers and systems biology approaches, researchers in the proteomics field have taken advantage of this approach. In this chapter, we will introduce the reader to the basic principle of designing and optimizing an MRM workflow. We provide examples of MRM workflows for standard proteomic samples and provide suggestions for the reader who is interested in using MRM for quantification.

Overview of quantitative LC-MS techniques for proteomics and activitomics

LC-MS is a useful technique for protein and peptide quantification. In addition, as a powerful tool for systems biology research, LC-MS can also be used to quantify post-translational modifications and metabolites that reflect biochemical pathway activity. This review discusses the different analytical techniques that use LC-MS for the quantification of proteins, their modifications and activities in a multiplex manner.

The bottleneck in the cancer biomarker pipeline and protein quantification through mass spectrometry-based approaches: current strategies for candidate verification

BACKGROUND

Although robust discovery-phase platforms have resulted in the generation of large numbers of candidate cancer biomarkers, a comparable system for subsequent quantitative assessment and verification of all candidates is lacking. Established immunoassays and available antibodies permit analysis of small subsets of candidates; however, the lack of commercially available reagents, coupled with high costs and lengthy production and purification times, have rendered the large majority of candidates untestable.

CONTENT

Mass spectrometry (MS), and in particular multiple reaction monitoring (MRM)-MS, has emerged as an alternative technology to immunoassays for quantification of target proteins. Novel biomarkers are expected to be present in serum in the low (microg/L-ng/L) range, but analysis of complex serum or plasma digests by MS has yielded milligram per liter limits of detection at best. The coupling of prior sample purification strategies such as enrichment of target analytes, depletion of high-abundance proteins, and prefractionation, has enabled reliable penetration into the low microgram per liter range. This review highlights prospects for candidate verification through MS-based methods. We first outline the biomarker discovery pipeline and its existing bottleneck; we then discuss various MRM-based strategies for targeted protein quantification, the applicability of such methods for candidate verification, and points of concern.

SUMMARY

Although it is unlikely that MS-based protein quantification will replace immunoassays in the near future, with the expected improvements in limits of detection and specificity in instrumentation, MRM-based approaches show great promise for alleviating the existing bottleneck to discovery.

Perspectives of targeted mass spectrometry for protein biomarker verification

The identification of specific biomarkers will improve the early diagnosis of disease, facilitate the development of targeted therapies, and provide an accurate method to monitor treatment response. A major challenge in the process of verifying biomarker candidates in blood plasma is the complexity and high dynamic range of proteins. This article reviews the current, targeted proteomic strategies that are capable of quantifying biomarker candidates at concentration ranges where biomarkers are expected in plasma (i.e. at the ng/ml level). In addition, a workflow is presented that allows the fast and definitive generation of targeted mass spectrometry-based assays for most biomarker candidate proteins. These assays are stored in publicly accessible databases and have the potential to greatly impact the throughput of biomarker verification studies.

Mass spectrometric investigation of molecular variability of grass pollen group 1 allergens

Natural grass pollen allergens exhibit a wide variety of isoforms. Precise characterization of such microheterogeneity is essential to improve diagnosis and design appropriate immunotherapies. Moreover, standardization of allergen vaccine production is a prerequisite for product safety and efficiency. Both qualitative and quantitative analytical methods are thus required to monitor and control the huge natural variability of pollens, as well as final product quality. A proteomic approach has been set up to investigate in depth the structural variability of five group 1 allergens originating from distinct grass species (Ant o 1, Dac g 1, Lol p 1, Phl p 1, and Poa p 1). Whereas group 1 is the most conserved grass pollen allergen, great variations were shown between the various isoforms found in these five species using mass spectrometry, with many amino acid exchanges, as well as variations in proline hydroxylation level and in main N-glycan motifs. The presence of O-linked pentose residues was also demonstrated, with up to three consecutive units on the first hydroxyproline of Ant o 1. In addition, species-specific peptides were identified that might be used for product authentication or individual allergen quantification. Lastly, natural or process-induced modifications (deamidation, oxidation, glycation) were evidenced, which might constitute useful indicators of product degradation.

Discovery and validation of colonic tumor-associated proteins via metabolic labeling and stable isotopic dilution

The unique biology of a neoplasm is reflected by its distinct molecular profile compared with normal tissue. To understand tumor development better, we have undertaken a quantitative proteomic search for abnormally expressed proteins in colonic tumors from Apc(Min/+) (Min) mice. By raising pairs of Min and wild-type mice on diets derived from natural-abundance or (15)N-labeled algae, we used metabolic labeling to compare protein levels in colonic tumor versus normal tissue. Because metabolic labeling allows internal control throughout sample preparation and analysis, technical error is minimized as compared with in vitro labeling. Several proteins displayed altered expression, and a subset was validated via stable isotopic dilution using synthetic peptide standards. We also compared gene and protein expression among tumor and nontumor tissue, revealing limited correlation. This divergence was especially pronounced for species showing biological change, highlighting the complementary perspectives provided by transcriptomics and proteomics. Our work demonstrates the power of metabolic labeling combined with stable isotopic dilution as an integrated strategy for the identification and validation of differentially expressed proteins using rodent models of human disease.

MaRiMba: a software application for spectral library-based MRM transition list assembly

Multiple reaction monitoring mass spectrometry (MRM-MS) is a targeted analysis method that has been increasingly viewed as an avenue to explore proteomes with unprecedented sensitivity and throughput. We have developed a software tool, called MaRiMba, to automate the creation of explicitly defined MRM transition lists required to program triple quadrupole mass spectrometers in such analyses. MaRiMba creates MRM transition lists from downloaded or custom-built spectral libraries, restricts output to specified proteins or peptides, and filters based on precursor peptide and product ion properties. MaRiMba can also create MRM lists containing corresponding transitions for isotopically heavy peptides, for which the precursor and product ions are adjusted according to user specifications. This open-source application is operated through a graphical user interface incorporated into the Trans-Proteomic Pipeline, and it outputs the final MRM list to a text file for upload to MS instruments. To illustrate the use of MaRiMba, we used the tool to design and execute an MRM-MS experiment in which we targeted the proteins of a well-defined and previously published standard mixture.

The validation and assessment of machine learning: a game of prediction from high-dimensional data

In applied statistics, tools from machine learning are popular for analyzing complex and high-dimensional data. However, few theoretical results are available that could guide to the appropriate machine learning tool in a new application. Initial development of an overall strategy thus often implies that multiple methods are tested and compared on the same set of data. This is particularly difficult in situations that are prone to over-fitting where the number of subjects is low compared to the number of potential predictors. The article presents a game which provides some grounds for conducting a fair model comparison. Each player selects a modeling strategy for predicting individual response from potential predictors. A strictly proper scoring rule, bootstrap cross-validation, and a set of rules are used to make the results obtained with different strategies comparable. To illustrate the ideas, the game is applied to data from the Nugenob Study where the aim is to predict the fat oxidation capacity based on conventional factors and high-dimensional metabolomics data. Three players have chosen to use support vector machines, LASSO, and random forests, respectively.