Megavariate data analysis of mass spectrometric proteomics data using latent variable projection method

There are many data mining techniques for processing and general learning of multivariate data. However, we believe the wavelet transformation and latent variable projection method are particularly useful for spectroscopic and chromatographic data. Projection based methods are designed to handle hugely multivariate nature of such data effectively. For the actual analysis of the data we have used latent variable projection methods such as principal component analysis (PCA) and partial least squares projection to latent structures based discriminant analysis (PLS-DA) to analyze the raw data presented to the participants of the First Duke Proteomics Data Mining Conference. PCA was used to solve problem #1 (clustering problem) and the PLS-DA was used to solve problem #2 (classification problem). The idea of internal and external cross-validation was used to validate the model obtained from the classification analysis. The simple two-component PLS-DA model obtained from the analysis performed well. The model has completely separated the two groups from all the data. The same model applied on two-thirds of the data showed good performance by external validation with independent test set of remaining 13 specimens obtained by setting aside the spectra of every third specimen (accuracy of 85%).     

Sequence analysis of peptide mixtures by automated integration of Edman and mass spectrometric data.

A computer algorithm is described that utilizes both Edman and mass spectrometric data for simultaneous determination of the amino acid sequences of several peptides in a mixture. Gas phase sequencing of a peptide mixture results in a list of observed amino acids for each cycle of Edman degradation, which by itself may not be informative and typically requires reanalysis following additional chromatographic steps. Tandem mass spectrometry, on the other hand, has a proven ability to analyze sequences of peptides present in mixtures. However, mass spectrometric data may lack a complete set of sequence-defining fragment ions, so that more than one possible sequence may account for the observed fragment ions. A combination of the two types of data reduces the ambiguity inherent in each. The algorithm first utilizes the Edman data to determine all hypothetical sequences with a calculated mass equal to the observed mass of one of the peptides present in the mixture. These sequences are then assigned figures of merit according to how well each of them accounts for the fragment ions in the tandem mass spectrum of that peptide. The program was tested on tryptic and chymotryptic peptides from hen lysozyme, and the results are compared with those of another computer program that uses only mass spectral data for peptide sequencing. In order to assess the utility of this method the program is tested using simulated mixtures of varying complexity and tandem mass spectra of varying quality.     

Integrating 'top-down" and "bottom-up" mass spectrometric approaches for proteomic analysis of Shewanella oneidensis

Here we present a comprehensive method for proteome analysis that integrates both intact protein measurement ("top-down") and proteolytic fragment characterization ("bottom-up") mass spectrometric approaches, capitalizing on the unique capabilities of each method. This integrated approach was applied in a preliminary proteomic analysis of Shewanella oneidensis, a metal-reducing microbe of potential importance to the field of bioremediation. Cellular lysates were examined directly by the "bottom-up" approach as well as fractionated via anion-exchange liquid chromatography for integrated studies. A portion of each fraction was proteolytically digested, with the resulting peptides characterized by on-line liquid chromatography/tandem mass spectrometry. The remaining portion of each fraction containing the intact proteins was examined by high-resolution Fourier transform mass spectrometry. This "top-down" technique provided direct measurement of the molecular masses for the intact proteins and thereby enabled confirmation of post-translational modifications, signal peptides, and gene start sites of proteins detected in the "bottom-up" experiments. A total of 868 proteins from virtually every functional class, including hypotheticals, were identified from this organism.     

Comprehensive proteomic mass spectrometric characterization of human cannabinoid CB2 receptor

The CB1 and CB2 cannabinoid receptors belong to the GPCR superfamily and are associated with a variety of physiological and pathophysiological processes. Both receptors, with several lead compounds at different phases of development, are potentially useful targets for drug discovery. For this reason, fully elucidating the structural features of these membrane-associated proteins would be extremely valuable in designing more selective, novel therapeutic drug molecules. As a first step toward obtaining information on the structural features of the drug-receptor complex, we describe the full mass spectrometric (MS) analysis of the recombinant human cannabinoid CB2 receptor. This first complete proteomic characterization of a GPCR protein beyond rhodopsin was accomplished by a combination of several LC/MS approaches involving nanocapillary liquid chromatography, coupled with either a quadrupole-linear ion trap or linear ion trap-FTICR mass spectrometer. The CB2 receptor, with incorporated N-terminal FLAG and C-terminal HIS6 epitope tags, was functionally expressed in baculovirus cells and purified using a single step of anti-FLAG M2 affinity chromatography. To overcome the difficulties involved with in-gel digestion, due to the highly hydrophobic nature of this membrane-associated protein, we conducted in-solution trypsin and chymotrypsin digestions of purified and desalted samples in the presence of a low concentration of CYMAL5. This was followed by nanoLC peptide separation and analysis using a nanospray ESI source operated in the positive mode. The results can be reported confidently, based on the overlapping sequence data obtained using the highly mass accurate LTQ-FT and the 4000 Q-Trap mass spectrometers. Both instruments gave very similar patterns of identified peptides, with full coverage of all transmembrane helices, resulting in the complete characterization of the cannabinoid CB2 receptor. Mass spectrometric identification of all amino acid residues in the cannabinoid CB2 receptor is a key step toward the "Ligand Based Structural Biology" approach developed in our laboratory for characterizing ligand binding sites in GPCRs using a variety of covalent cannabinergic ligands.     

Analysis and validation of proteomic data generated by tandem mass spectrometry

The analysis of the large amount of data generated in mass spectrometry-based proteomics experiments represents a significant challenge and is currently a bottleneck in many proteomics projects. In this review we discuss critical issues related to data processing and analysis in proteomics and describe available methods and tools. We place special emphasis on the elaboration of results that are supported by sound statistical arguments.     

Algorithms for alignment of mass spectrometry proteomic data

MOTIVATION: The analysis of biological samples with high-throughput mass spectrometers has increased greatly in recent years. As larger datasets are processed, it is important that the spectra are aligned to ensure that the same protein intensities are correctly identified in each sample. Without such an alignment procedure it is possible to make errors in identifying the signals from peptides with similar molecular weight. Two algorithms are provided that can improve the alignment among samples. One algorithm is designed to work with SELDI data produced from a Ciphergen instrument, and the other can be used with data in a more general format. RESULTS: The two algorithms were applied to samples drawn from a common pool of reference serum. The results indicate substantial improvement in consistently identifying peptide signals in different samples.     

Protein identification by in-gel digestion and mass spectrometric analysis

This protocol details a method for the identification of proteins that have been separated by gel electrophoresis. In-gel digestion of the protein bands with trypsin followed by quadrupole ion-trap or other triple quadrupole mass spectrometry techniques is described. The proteins can be identified by database searching of the mass fingerprint of the intact peptides and of the characteristic fragment masses produced by tandem mass spectrometry.     

Structure elucidation of metabolite x17299 by interpretation of mass spectrometric data

Introduction

A major bottleneck in metabolomic studies is metabolite identification from accurate mass spectrometric data. Metabolite x17299 was identified in plasma as an unknown in a metabolomic study using a compound-centric approach where the associated ion features of the compound were used to determine the true molecular mass.

Objectives

The aim of this work is to elucidate the chemical structure of x17299, a new compound by de novo interpretation of mass spectrometric data.

Methods

An Orbitrap Elite mass spectrometer was used for acquisition of mass spectra up to MS⁴ at high resolution. Synthetic standards of N,N,N-trimethyl-l-alanyl-l-proline betaine (l,l-TMAP), a diastereomer, and an enantiomer were chemically prepared.

Results

The planar structure of x17299 was successfully proposed by de novo mechanistic interpretation of mass spectrometric data without any laborious purification and nuclear magnetic resonance spectroscopic analysis. The proposed structure was verified by deuterium exchanged mass spectrometric analysis and confirmed by comparison to a synthetic standard. Relative configuration of x17299 was determined by direct chromatographic comparison to a pair of synthetic diastereomers. Absolute configuration was assigned after derivatization of x17299 with a chiral auxiliary group followed by its chromatographic comparison to a pair of synthetic standards.

Conclusion

The chemical structure of metabolite x17299 was determined to be l,l-TMAP.

     

Gas chromatographic/mass spectrometric analysis of specific isomers of polychlorodibenzofurans

The gas chromatographic/mass spectrometric characteristics of 26 congeners of polychlorinated dibenzofurans, previously characterized by specific synthetic routes and by standard spectroscopic techniques, have been evaluated. The electron impact mass spectra are not particularly isomer-specific, though 2,3,7,8-tetrachlorodibenzofuran is distinguishable on this basis from the three other tetrachloro isomers investigated in this work. Positive ion methane chemical ionization mass spectra do show a greater degree of isomer distinction, and are reasonably reproducible. Electron attachment negative ion spectral characteristics are also presented. Preliminary results on negative ion chemical ionization mass spectra, obtained using methane plus small amounts of oxygen as reagent gas, are reported.     

Artificial neural network analysis of pyrolysis mass spectrometric data in the identification of Streptomyces strains

Abstract Sixteen representatives of three morphologically distinct groups of streptomycetes were recovered from soil using selective isolation procedures. Duplicated batches of the test strains were examined by Curie-point pyrolysis mass spectrometry and the first data set used for conventional multivariate statistical analyses and as a training set for an artificial neural network. The second set of data was used for 'operational fingerprinting' and for testing the artificial neural network. All of the test strains were correctly identified using the artificial neural network whereas only fifteen of the sixteen strains were assigned to the correct group using the conventional operational fingerprinting procedure. Artificial neural network analysis of pyrolysis mass spectrometric data provides a rapid, cost-effective and reproducible way of identifying and typing large numbers of microorganisms.     

Improved mass spectrometric proteomic profiling of the secretome of rat vascular endothelial cells

Serum albumin contamination of cells cultured in vitro significantly impedes the mass spectrometric analysis of proteins secreted by the cells. Here we report a novel washing and culturing technique for rat vascular endothelial cells that considerably reduces the concentration of the commonly used additive for cell culture, bovine serum albumin (BSA), in the secretome of these cells. Cells are rinsed stringently and cultured for 24 h in serum-free media without appreciably impeding cell growth or viability. The percentage of BSA scans identified by tandem mass spectrometry (LC-MS/MS) in stringently rinsed cells (average 13.2%) was significantly lower than either the moderately rinsed or no rinse cell treatments (average 35.2% and 45.2% respectively). Furthermore, the stringent wash treatment allowed the confident identification of a larger portion of the secretome of rat endothelial cells by LC-MS/MS.     

The effect of using an inappropriate protein database for proteomic data analysis

A recent study by Bromenshenk et al., published in PLoS One (2010), used proteomic analysis to identify peptides purportedly of Iridovirus and Nosema origin; however the validity of this finding is controversial. We show here through re-analysis of a subset of this data that many of the spectra identified by Bromenshenk et al. as deriving from Iridovirus and Nosema proteins are actually products from Apis mellifera honey bee proteins. We find no reliable evidence that proteins from Iridovirus and Nosema are present in the samples that were re-analyzed. This article is also intended as a learning exercise for illustrating some of the potential pitfalls of analysis of mass spectrometry proteomic data and to encourage authors to observe MS/MS data reporting guidelines that would facilitate recognition of analysis problems during the review process.     

Identification of naturally processed ligands in the C57BL/6 mouse using large-scale mass spectrometric peptide sequencing and bioinformatics prediction

Most major histocompatibility complex (MHC) class I–peptide-binding motifs are currently defined on the basis of quantitative in vitro MHC–peptide-binding assays. This information is used to develop bioinformatics-based tools to predict the binding of peptides to MHC class I molecules. To date few studies have analyzed the performance of these bioinformatics tools to predict the binding of peptides determined by sequencing of naturally processed peptides eluted directly from MHC class I molecules. In this study, we performed large-scale sequencing of endogenous peptides eluted from H2K^b and H2D^b molecules expressed in spleens of C57BL/6 mice. Using sequence data from 281 peptides, we identified novel preferred anchor residues located in H2K^b and H2D^b-associated peptides that refine our knowledge of these H2 class I peptide-binding motifs. The analysis comparing the performance of three bioinformatics methods to predict the binding of these peptides, including artificial neural network, stabilized matrix method, and average relative binding, revealed that 61% to 94% of peptides eluted from H2K^b and H2D^b molecules were correctly classified as binders by the three algorithms. These results suggest that bioinformatics tools are reliable and efficient methods for binding prediction of naturally processed MHC class I ligands. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Spatially-correlated mass spectrometric analysis of microbe--mineral interactions

A new methodology for examining the interactions of microbes with heterogeneous minerals is presented. Imaging laser desorption Fourier transform mass spectrometry was used to examine the colonization patterns of Burkholderia vietnamiensis G4 (previously Burkholderia cepacia G4) on a heterogeneous basalt sample. Depth-profile imaging found that the bacterium preferentially colonized the plagioclase mineral phase within the basalt.     

Ion-spray mass spectrometric analysis of glycosaminoglycan oligosaccharides

Oligosaccharides from hyaluronic acid and chondroitin 6-sulfate were prepared by digestion with testicular hyaluronidase and separated according to their degree of polymerization by gel-permeation chromatography. These materials were successively analyzed by negative-mode ion-spray mass spectrometry with an atmospheric-pressure ion source. An ion-spray interface was used to produce ions via the ion evaporation process, producing mass spectra containing a series of molecular species carrying multiple charges. Using two adjacent multiply charged molecular ions, the exact molecular weights up to the tetradecasaccharide were calculated with a precision of ±1 dalton. This type of mass spectrometry was also demonstrated to be feasible for the analysis of mixtures of oligosaccharides, including tetra-, hexa-, octa- and decasaccharides, from hyaluronic acid or chondroitin 6-sulfate without separation. Ion-spray mass spectrometry was thus shown to be applicable to the structural analysis of oligosaccharides from glycosaminoglycans.Abbreviations HA hyaluronic acid - Ch6S chondroitin 6-sulfate - GAG glycosaminoglycan - GlcA d-glucuronic acid - GlcNAc 2-acetamido-2-deoxy-d-glucose - GalNAc 2-acetamido-2-deoxy-d-galactose.     

Enrichment of integral membrane proteins for proteomic analysis using liquid chromatography-tandem mass spectrometry

An increasing number of proteomic strategies rely on liquid chromatography-tandem mass spectrometry (LC-MS/MS) to detect and identify constituent peptides of enzymatically digested proteins obtained from various organisms and cell types. However, sample preparation methods for isolating membrane proteins typically involve the use of detergents and chaotropes that often interfere with chromatographic separation and/or electrospray ionization. To address this problem, a sample preparation method combining carbonate extraction, surfactant-free organic solvent-assisted solubilization, and proteolysis was developed and demonstrated to target the membrane subproteome of Deinococcus radiodurans. Out of 503 proteins identified, 135 were recognized as hydrophobic on the basis of their calculated hydropathy values (GRAVY index), corresponding to coverage of 15% of the predicted hydrophobic proteome. Using the PSORT algorithm, 53 of the proteins identified were classified as integral outer membrane proteins and 215 were classified as integral cytoplasmic membrane proteins. All identified integral cytoplasmic membrane proteins had from 1 to 16 mapped transmembrane domains (TMDs), and 65% of those containing four or more mapped TMDs were identified by at least one hydrophobic membrane spanning peptide. The extensive coverage of the membrane subproteome (24%) by identification of highly hydrophobic proteins containing multiple TMDs validates the efficacy of the described sample preparation technique to isolate and solubilize hydrophobic integral membrane proteins from complex protein mixtures.     

High-throughput proteomic analysis of formalin-fixed paraffin-embedded tissue microarrays using MALDI imaging mass spectrometry

A novel method for high-throughput proteomic analysis of formalin-fixed paraffin-embedded (FFPE) tissue microarrays (TMA) is described using on-tissue tryptic digestion followed by MALDI imaging MS. A TMA section containing 112 needle core biopsies from lung-tumor patients was analyzed using MS and the data were correlated to a serial hematoxylin and eosin (H&E)-stained section having various histological regions marked, including cancer, non-cancer, and normal ones. By correlating each mass spectrum to a defined histological region, statistical classification models were generated that can sufficiently distinguish biopsies from adenocarcinoma from squamous cell carcinoma biopsies. These classification models were built using a training set of biopsies in the TMA and were then validated on the remaining biopsies. Peptide markers of interest were identified directly from the TMA section using MALDI MS/MS sequence analysis. The ability to detect and characterize tumor marker proteins for a large cohort of FFPE samples in a high-throughput approach will be of significant benefit not only to investigators studying tumor biology, but also to clinicians for diagnostic and prognostic purposes.     

Biomarker discovery for arsenic exposure using functional data. Analysis and feature learning of mass spectrometry proteomic data

Plasma biomarkers of exposure to environmental contaminants play an important role in early detection of disease. The emerging field of proteomics presents an attractive opportunity for candidate biomarker discovery, as it simultaneously measures and analyzes a large number of proteins. This article presents a case study for measuring arsenic concentrations in a population residing in an As-endemic region of Bangladesh using plasma protein expressions measured by SELDI-TOF mass spectrometry. We analyze the data using a unified statistical method based on functional learning to preprocess mass spectra and extract mass spectrometry (MS) features and to associate the selected MS features with arsenic exposure measurements. The task is challenging due to several factors, the high dimensionality of mass spectrometry data, complicated error structures, and a multiple comparison problem. We use nonparametric functional regression techniques for MS modeling, peak detection based on the significant zero-downcrossing method, and peak alignment using a warping algorithm. Our results show significant associations of arsenic exposure to either under- or overexpressions of 20 proteins.     

Ab initio prediction of metabolic networks using Fourier transform mass spectrometry data

Fourier transform mass spectrometry has recently been introduced into the field of metabolomics as a technique that enables the mass separation of complex mixtures at very high resolution and with ultra high mass accuracy. Here we show that this enhanced mass accuracy can be exploited to predict large metabolic networks ab initio, based only on the observed metabolites without recourse to predictions based on the literature. The resulting networks are highly information-rich and clearly non-random. They can be used to infer the chemical identity of metabolites and to obtain a global picture of the structure of cellular metabolic networks. This represents the first reconstruction of metabolic networks based on unbiased metabolomic data and offers a breakthrough in the systems-wide analysis of cellular metabolism.