Identifying metabolites by integrating metabolome databases with mass spectrometry cheminformatics.

Novel metabolites distinct from canonical pathways can be identified through the integration of three cheminformatics tools: BinVestigate, which queries the BinBase gas chromatography-mass spectrometry (GC-MS) metabolome database to match unknowns with biological metadata across over 110,000 samples; MS-DIAL 2.0, a software tool for chromatographic deconvolution of high-resolution GC-MS or liquid chromatography-mass spectrometry (LC-MS); and MS-FINDER 2.0, a structure-elucidation program that uses a combination of 14 metabolome databases in addition to an enzyme promiscuity library. We showcase our workflow by annotating N-methyl-uridine monophosphate (UMP), lysomonogalactosyl-monopalmitin, N-methylalanine, and two propofol derivatives.

MRMPROBS suite for metabolomics using large-scale MRM assays.

UNLABELLED: We developed new software environment for the metabolome analysis of large-scale multiple reaction monitoring (MRM) assays. It supports the data format of four major mass spectrometer vendors and mzML common data format. This program provides a process pipeline from the raw-format import to high-dimensional statistical analyses. The novel aspect is graphical user interface-based visualization to perform peak quantification, to interpolate missing values and to normalize peaks interactively based on quality control samples. Together with the software platform, the MRM standard library of 301 metabolites with 775 transitions is also available, which contributes to the reliable peak identification by using retention time and ion abundances. AVAILABILITY AND IMPLEMENTATION: MRMPROBS is available for Windows OS under the creative-commons by-attribution license at http://prime.psc.riken.jp.

MRMPROBS: a data assessment and metabolite identification tool for large-scale multiple reaction monitoring based widely targeted metabolomics.

We developed a new software program, MRMPROBS, for widely targeted metabolomics by using the large-scale multiple reaction monitoring (MRM) mode. The strategy became increasingly popular for the simultaneous analysis of up to several hundred metabolites at high sensitivity, selectivity, and quantitative capability. However, the traditional method of assessing measured metabolomics data without probabilistic criteria is not only time-consuming but is often subjective and makeshift work. Our program overcomes these problems by detecting and identifying metabolites automatically, by separating isomeric metabolites, and by removing background noise using a probabilistic score defined as the odds ratio from an optimized multivariate logistic regression model. Our software program also provides a user-friendly graphical interface to curate and organize data matrices and to apply principal component analyses and statistical tests. For a demonstration, we conducted a widely targeted metabolome analysis (152 metabolites) of propagating Saccharomyces cerevisiae measured at 15 time points by gas and liquid chromatography coupled to triple quadrupole mass spectrometry. MRMPROBS is a useful and practical tool for the assessment of large-scale MRM data available to any instrument or any experimental condition.

A computational drug metabolite detection using the stable isotopic mass-shift filtering with high resolution mass spectrometry in pioglitazone and flurbiprofen.

The identification of metabolites in drug discovery is important. At present, radioisotopes and mass spectrometry are both widely used. However, rapid and comprehensive identification is still laborious and difficult. In this study, we developed new analytical software and employed a stable isotope as a tool to identify drug metabolites using mass spectrometry. A deuterium-labeled compound and non-labeled compound were both metabolized in human liver microsomes and analyzed by liquid chromatography/time-of-flight mass spectrometry (LC-TOF-MS). We computationally aligned two different MS data sets and filtered ions having a specific mass-shift equal to masses of labeled isotopes between those data using our own software. For pioglitazone and flurbiprofen, eight and four metabolites, respectively, were identified with calculations of mass and formulas and chemical structural fragmentation analysis. With high resolution MS, the approach became more accurate. The approach detected two unexpected metabolites in pioglitazone, i.e., the hydroxypropanamide form and the aldehyde hydrolysis form, which other approaches such as metabolite-biotransformation list matching and mass defect filtering could not detect. We demonstrated that the approach using computational alignment and stable isotopic mass-shift filtering has the ability to identify drug metabolites and is useful in drug discovery.

Identification by differential tissue proteome analysis of talin-1 as a novel molecular marker of progression of hepatocellular carcinoma.

OBJECTIVE: Hepatocellular carcinoma (HCC) is characterized by a multistage process of tumor progression. This study addressed its molecular features to identify novel protein candidates involved in HCC progression. METHODS: Using liquid chromatography-tandem mass spectrometry, proteomes of 4 early HCCs and 4 non-HCC tissues derived from 2 cases of liver transplant surgery were compared with respect to the separation profiles of their tryptic peptides. Immunohistochemistry was performed on 106 HCC nodules to confirm the results of the proteomic analysis. RESULTS: Statistical analysis of the profiles selected the peptide peaks differentiating HCC from non-HCC. A database search of the tandem mass spectrometry data from those peptide peaks identified 61 proteins, including a cytoskeletal protein, talin-1, as upregulated in HCC. Talin-1 expression levels in HCC nodules were significantly associated with the dedifferentiation of HCC (p = 0.001). A follow-up survey of the examined clinical cases revealed a correlation between talin-1 upregulation and a shorter time to recurrence after resection (p = 0.039), which may be related to the higher rate of portal vein invasion in HCCs with talin-1 up-regulation (p = 0.029). CONCLUSIONS: Proteomic analysis led to identification of talin-1 as a promising HCC marker. Talin-1 upregulation is associated with HCC progression and may serve as a prognostic marker.

i-RUBY: a novel software for quantitative analysis of highly accurate shotgun-proteomics liquid chromatography/tandem mass spectrometry data obtained without stable-isotope labeling of proteins.

We developed a novel software named i-RUBY (identification-Related qUantification-Based strategY algorithm for liquid chromatography/tandem mass spectrometry (LC/MS/MS) data) that enables us to perform fully automatic ion current-based spectral feature analysis of highly accurate data obtained by LC/MS/MS. At the 1st step, this software utilizes accurate peptide/protein identification information for peak detection and peak matching among measurements. Then, at the 2nd step, it picks yet unidentified peaks and matches them to the peaks identified at the 1st step by a linear interpolation algorithm. The analysis of human plasma externally spiked with a known amount of yeast alcohol dehydrogenase 1 showed a good linear relationship between the amount of protein spiked and the quantitative values obtained by i-RUBY analysis. Experiment using human plasma digests spiked with a mixture of known amounts of synthetic peptides derived from two yeast proteins, alcohol dehydrogenase 1 and glucose-6-phospate isomerase, showed the expansion by the 2nd step of i-RUBY of the lower quantification limits to 1/10 to 1/1000 of those reached only by identified peaks at the 1st step. Good correlations between the i-RUBY results and the amount of proteins were confirmed by the analysis of real samples, i.e., sera of normal subjects and cancer patients, by comparing quantitative values of acute-phase proteins obtained by i-RUBY analysis of LC/MS/MS data with those obtained by an immunological method using Bio-Plex. These results taken together show that i-RUBY is a useful tool for obtaining dependable quantitative information from highly accurate shotgun-proteomics LC/MS/MS data.

Synuclein-gamma is closely involved in perineural invasion and distant metastasis in mouse models and is a novel prognostic factor in pancreatic cancer.

PURPOSE: Perineural invasion is associated with the high incidence of local recurrence and a dismal prognosis in pancreatic cancer. We previously reported a novel perineural invasion model and distinguished high- and low-perineural invasion groups in pancreatic cancer cell lines. This study aimed to elucidate the molecular mechanism of perineural invasion. EXPERIMENTAL DESIGN: To identify key biological markers involved in perineural invasion, differentially expressed molecules were investigated by proteomics and transcriptomics. Synuclein-gamma emerged as the only up-regulated molecule in high-perineural invasion group by both analyses. The clinical significance and the biological property of synuclein-gamma were examined in 62 resected cases of pancreatic cancer and mouse models. RESULTS: Synuclein-gamma overexpression was observed in 38 (61%) cases and correlated with major invasive parameters, including perineural invasion and lymph node metastasis (P < 0.05). Multivariate analyses revealed synuclein-gamma overexpression as the only independent predictor of diminished overall survival [hazard ratio, 3.4 (95% confidence interval, 1.51-7.51)] and the strongest negative indicator of disease-free survival [2.8 (1.26-6.02)]. In mouse perineural invasion and orthotopic transplantation models, stable synuclein-gamma suppression by short hairpin RNA significantly reduced the incidence of perineural invasion (P = 0.009) and liver/lymph node metastasis (P = 0.019 and P = 0.020, respectively) compared with the control. CONCLUSIONS: This is the first study to provide in vivo evidence that synuclein-gamma is closely involved in perineural invasion/distant metastasis and is a significant prognostic factor in pancreatic cancer. Synuclein-gamma may serve as a promising molecular target of early diagnosis and anticancer therapy.

De novo peptide sequencing using ion peak intensity and amino acid cleavage intensity ratio.

MOTIVATION: Peptide-sequencing methods by mass spectrum use the following two approaches: database searching and de novo sequencing. The database-searching approach is convenient; however, in cases wherein the corresponding sequences are not included in the databases, the exact identification is difficult. On the other hand, in the case of de novo sequencing, no preliminary information is necessary; however, continuous amino acid sequence peaks and the differentiation of these peaks are required. It is, however, very difficult to obtain and differentiate the peaks of all amino acids by using an actual spectrum. We propose a novel de novo sequencing approach using not only mass-to-charge ratio but also ion peak intensity and amino acid cleavage intensity ratio (CIR). RESULTS: Our method compensates for any undetectable amino acid peak intervals by estimating the amino acid set and the probability of peak expression based on amino acid CIR. It provides more accurate identification of sequences than the existing methods, by which it is usually difficult to sequence.

Disease proteomics toward bedside reality.

The human genome has been sequenced, and investigation of its products has become possible in a sequence-based framework. More than 200,000 protein species are expressed in the body from approximately 30000 human genes. The term proteome, coined as a linguistic equivalent to the concept of genome, is used to describe the complete set of proteins that is expressed, and modified following expression, by the entire genome in a cell at any one time. Protein types and amounts expressed in a body vary greatly depending upon whether it is healthy or ill. Therefore, proteomics is attracting an increasing interest in its application to better understanding of disease processes, to development of new biomarkers for diagnosis and early detection of disease, and to accelerate drug development. There are numerous opportunities for medicine, although it is quite challenging to meet the needs for high sensitivity and high throughput required for disease-related investigations.

MRM-DIFF: data processing strategy for differential analysis in large scale MRM-based lipidomics studies.

Based on theoretically calculated comprehensive lipid libraries, in lipidomics as many as 1000 multiple reaction monitoring (MRM) transitions can be monitored for each single run. On the other hand, lipid analysis from each MRM chromatogram requires tremendous manual efforts to identify and quantify lipid species. Isotopic peaks differing by up to a few atomic masses further complicate analysis. To accelerate the identification and quantification process we developed novel software, MRM-DIFF, for the differential analysis of large-scale MRM assays. It supports a correlation optimized warping (COW) algorithm to align MRM chromatograms and utilizes quality control (QC) sample datasets to automatically adjust the alignment parameters. Moreover, user-defined reference libraries that include the molecular formula, retention time, and MRM transition can be used to identify target lipids and to correct peak abundances by considering isotopic peaks. Here, we demonstrate the software pipeline and introduce key points for MRM-based lipidomics research to reduce the mis-identification and overestimation of lipid profiles. The MRM-DIFF program, example data set and the tutorials are downloadable at the "Standalone software" section of the PRIMe (Platform for RIKEN Metabolomics, http://prime.psc.riken.jp/) database website.

Nested case control study of proteomic biomarkers for interstitial lung disease in Japanese patients with non-small-cell lung cancer treated with erlotinib: a multicenter phase IV study (JO21661).

BACKGROUND: Interstitial lung disease (ILD) is a serious adverse drug reaction associated with epidermal growth factor receptor tyrosine-kinase inhibitors (EGFR TKIs). Its risk factors are yet to be fully elucidated. We sought to identify proteomic biomarkers associated with ILD development in erlotinib-treated Japanese patients with non-small-cell lung cancer (NSCLC) to build predictive models. PATIENTS AND METHODS: We conducted a nested case-control study. The participants were patients with NSCLC enrolled in a phase IV study of erlotinib in whom ILD developed within 120 days after erlotinib administration. The controls were randomly selected patients without ILD from the overall study cohort who were also treated with erlotinib. Serum samples were obtained before the first administration of erlotinib and were assayed by liquid chromatography-mass spectroscopy/mass spectroscopy (LC-MS/MS). Logistic regression analysis was performed to identify the peptide and proteins associated with ILD. RESULTS: A total of 645 patients were enrolled in the cohort; 15 case patients and 64 controls were analyzed. When multiplicity was taken into account, we were unable to statistically verify any genuine association between individual markers and ILD. Investigation of the predictive power based on leave-one-out cross-validation (LOOCV) showed that the area under the receiver operating characteristic curve was 0.73 at a maximum. Additional analysis suggested that 3 proteins (C3, C4A/C4B, and APOA1) have a stronger association with ILD than do the other proteins tested. CONCLUSION: We were unable to demonstrate predictive serum protein markers for ILD development. However, C3, C4A/C4B, and APOA1 are worthy of further investigation.

Proteomic biomarkers for acute interstitial lung disease in gefitinib-treated Japanese lung cancer patients.

Interstitial lung disease (ILD) events have been reported in Japanese non-small-cell lung cancer (NSCLC) patients receiving EGFR tyrosine kinase inhibitors. We investigated proteomic biomarkers for mechanistic insights and improved prediction of ILD. Blood plasma was collected from 43 gefitinib-treated NSCLC patients developing acute ILD (confirmed by blinded diagnostic review) and 123 randomly selected controls in a nested case-control study within a pharmacoepidemiological cohort study in Japan. We generated ∼7 million tandem mass spectrometry (MS/MS) measurements with extensive quality control and validation, producing one of the largest proteomic lung cancer datasets to date, incorporating rigorous study design, phenotype definition, and evaluation of sample processing. After alignment, scaling, and measurement batch adjustment, we identified 41 peptide peaks representing 29 proteins best predicting ILD. Multivariate peptide, protein, and pathway modeling achieved ILD prediction comparable to previously identified clinical variables; combining the two provided some improvement. The acute phase response pathway was strongly represented (17 of 29 proteins, p = 1.0×10(-25)), suggesting a key role with potential utility as a marker for increased risk of acute ILD events. Validation by Western blotting showed correlation for identified proteins, confirming that robust results can be generated from an MS/MS platform implementing strict quality control.

Personalized medicine and proteomics: lessons from non-small cell lung cancer.

Personalized medicine allows the selection of treatments best suited to an individual patient and disease phenotype. To implement personalized medicine, effective tests predictive of response to treatment or susceptibility to adverse events are needed, and to develop a personalized medicine test, both high quality samples and reliable data are required. We review key features of state-of-the-art proteomic profiling and introduce further analytic developments to build a proteomic toolkit for use in personalized medicine approaches. The combination of novel analytical approaches in proteomic data generation, alignment and comparison permit translation of identified biomarkers into practical assays. We further propose an expanded statistical analysis to understand the sources of variability between individuals in terms of both protein expression and clinical variables and utilize this understanding in a predictive test.