Adding Color to Mass Spectra of Biopolymers: Charge Determination Analysis (CHARDA) Assigns Charge State to Every Ion Peak.

Traditionally, mass spectrometry (MS) output is the ion abundance plotted versus the ionic mass-to-charge ratio m/z. While employing only commercially available equipment, Charge Determination Analysis (CHARDA) adds a third dimension to MS, estimating for individual peaks their charge states z starting from z = 1 and color coding z in m/z spectra. CHARDA combines the analysis of ion signal decay rates in the time-domain data (transients) in Fourier transform (FT) MS with the interrogation of mass defects (fractional mass) of biopolymers. Being applied to individual isotopic peaks in a complex protein tandem (MS/MS) data set, CHARDA aids peptide mass spectra interpretation by facilitating charge-state deconvolution of large ionic species in crowded regions, estimating z even in the absence of an isotopic distribution (e.g., for monoisotopic mass spectra). CHARDA is fast, robust, and consistent with conventional FTMS and FTMS/MS data acquisition procedures. An effective charge-state resolution Rz ≥ 6 is obtained with the potential for further improvements.

Profiling the Metabolism of Human Cells by Deep 13C Labeling.

Studying metabolic activities in living cells is crucial for understanding human metabolism, but facile methods for profiling metabolic activities in an unbiased, hypothesis-free manner are still lacking. To address this need, we here introduce the deep-labeling method, which combines a custom 13C medium with high-resolution mass spectrometry. A proof-of-principle study on human cancer cells demonstrates that deep labeling can identify hundreds of endogenous metabolites as well as active and inactive pathways. For example, protein and nucleic acids were almost exclusively de novo synthesized, while lipids were partly derived from serum; synthesis of cysteine, carnitine, and creatine was absent, suggesting metabolic dependencies; and branched-chain keto acids (BCKAs) were formed and metabolized to short-chain acylcarnitines, but did not enter the tricarboxylic acid cycle. Remarkably, BCKAs could substitute for essential amino acids to support growth. The deep-labeling method may prove useful to map metabolic phenotypes across a range of cell types and conditions.

A Web Service Framework for Interactive Analysis of Metabolomics Data.

Analyzing mass spectrometry-based metabolomics data presents a major challenge to metabolism researchers, as it requires downloading and processing large data volumes through complex "pipelines", even in cases where only a single metabolite or peak is of interest. This presents a significant hurdle for data sharing, reanalysis, or meta-analysis of existing data sets, whether locally stored or available from public repositories. Here we introduce mzAccess, a software system that provides interactive, online access to primary mass spectrometry data in real-time via a Web service protocol, circumventing the need for bulk data processing. mzAccess allows querying instrument data for spectra, chromatograms, or two-dimensional MZ-RT areas in either profile or centroid modes through a simple, uniform interface that is independent of vendor or instrument type. Using a cache mechanism, mzAccess achieves response times in the millisecond range for typical liquid chromatography-mass spectrometry (LC-MS) peaks, enabling real-time browsing of large data sets with hundreds or even thousands of samples. By simplifying access to metabolite data, we hope that this system will help enable data sharing and reanalysis in the metabolomics field.

Prognostic polypeptide blood plasma biomarkers of Alzheimer's disease progression.

BACKGROUND: Patients with mild cognitive impairment (MCI) have varying risks of progression to Alzheimer's disease (AD). OBJECTIVE: To test the utility of the relative abundances of blood plasma polypeptides for predicting the risk of AD progression. METHODS: 119 blood plasma samples of patients with MCI with different outcomes (stable MCI and progressive MCI) were analyzed by untargeted, label-free shotgun proteomics. Predictive biomarkers of progressive MCI were selected by multivariate analysis, followed by cross-validation of the predictive model. RESULTS: The best model demonstrated the accuracy of ca. 79% in predicting progressive MCI. Sex differences of the predictive biomarkers were also assessed. We have identified some sex-specific protein biomarkers, e.g., alpha-2-macrogloblin (A2M), which strongly correlates with female AD progression but not with males. CONCLUSION: Significant sex bias in AD-specific biomarkers underscores the necessity of selecting sex-balanced cohort in AD biomarker studies, or using sex-specific models. Blood protein biomarkers are found to be promising for predicting AD progression in clinical settings.

Blood plasma IgG Fc glycans are significantly altered in Alzheimer's disease and progressive mild cognitive impairment.

Blood-based anti-amyloid-ß (Aß) immunoglobulins (IgGs) and peripheral inflammation are factors correlating with development of Alzheimer's disease (AD). IgG functionality can drastically change from anti- to pro-inflammatory via alterations in the IgG-Fc N-glycan structure. Herein, we tested if IgG-Fc glycosylation in plasma is indeed altered during the development of AD. Samples from age-matched subjects of 23 controls, 58 patients with stable mild cognitive impairment (SMCI), 34 patients with progressive (P)MCI, and 31 patients with AD were investigated. Label-free shotgun proteomics was applied without glycoprotein enrichment. Glycans on peptides EEQYNSTYR (IgG1) and EEQFNSTFR (IgG2) were quantified, and their abundances were normalized to total IgGn glycoform abundance. Univariate and multivariate statistics were employed to investigate the correlations between the patients groups and the abundances of the IgG glycoforms as well as those of inflammatory mediating proteins. Significant differences (p ≤ 0.05) were found, with a lower abundance of complex galactosylated and sialylated forms in AD. For females, a decline in glycoform complexity correlated with disease progress but an inverse change was found in males prior to the onset of AD. Principal component analysis (PCA; Males: R(2)X(cum) = 0.65, Q(2)(cum) = 0.34; Females: R(2)X(cum) = 0.62, Q(2)(cum) = 0.36), confirmed the gender similarities (for controls, SMCI and AD) as well as differences (for PMCI), and showed a close correlation between pro-inflammatory protein markers, AD, female PMCI, and truncated IgG-Fc glycans. The differences observed between genders prior to the onset of AD may indicate a lower ability in females to suppress peripheral inflammation, which may lead to exacerbated disease progression.

In silico instrumental response correction improves precision of label-free proteomics and accuracy of proteomics-based predictive models.

In the analysis of proteome changes arising during the early stages of a biological process (e.g. disease or drug treatment) or from the indirect influence of an important factor, the biological variations of interest are often small (∼10%). The corresponding requirements for the precision of proteomics analysis are high, and this often poses a challenge, especially when employing label-free quantification. One of the main contributors to the inaccuracy of label-free proteomics experiments is the variability of the instrumental response during LC-MS/MS runs. Such variability might include fluctuations in the electrospray current, transmission efficiency from the air-vacuum interface to the detector, and detection sensitivity. We have developed an in silico post-processing method of reducing these variations, and have thus significantly improved the precision of label-free proteomics analysis. For abundant blood plasma proteins, a coefficient of variation of approximately 1% was achieved, which allowed for sex differentiation in pooled samples and ≈90% accurate differentiation of individual samples by means of a single LC-MS/MS analysis. This method improves the precision of measurements and increases the accuracy of predictive models based on the measurements. The post-acquisition nature of the correction technique and its generality promise its widespread application in LC-MS/MS-based methods such as proteomics and metabolomics.

The effects of 5-fluorouracil on the proteome of colon cancer cells.

The pyrimidine analogue 5-fluorouracil (5FU) is used as a treatment for solid tumors, but its mechanism of action is not fully understood. We have used mass spectrometry to study the mechanism of action of 5FU, and we have measured the effects of this drug on the composition and on the turnover of the proteome of RKO cancer cells. We have identified novel potential targets of 5FU that are affected after very short exposure times. We have also shown that 5FU has a massive effect on the proteins involved in RNA metabolism. After only 1 h of treatment, 5FU causes a post-transcriptional reduction in the abundance of components of the translation machinery (mostly ribosomal proteins), and this reduction is accompanied by a down-regulation of the translational capacity of the cells. Neither rapamycin nor raltitrexed, two drugs that also block cell proliferation, reduce the abundances of ribosomal proteins as 5FU does, which suggests that the down-regulation of ribosomal proteins is coupled to the mechanism of action of 5FU. Some of our observations conflict with previous reports based on RNA quantification. This shows how important it is to complement RNA profiling studies with analyses of drug toxicity at the protein level.

Proteomic analysis of urinary biomarker candidates for nonmuscle invasive bladder cancer.

Nonmuscle invasive tumors of the bladder often recur and thereby bladder cancer patients need regular re-examinations which are invasive, unpleasant, and expensive. A noninvasive and less expensive method, e.g. a urine dipstick test, for monitoring recurrence would thus be advantageous. In this study, the complementary techniques mass spectrometry (MS) and Western blotting (WB)/dot blot (DB) were used to screen the urine samples from bladder cancer patients. High resolving MS was used to analyze and quantify the urinary proteome and 29 proteins had a significantly higher abundance (p<0.05) in bladder cancer samples compared with control urine samples. The increased abundance found in urine from bladder cancer patients compared with controls was confirmed with Western blot for four selected proteins; fibrinogen ß chain precursor, apolipoprotein E, α-1-antitrypsin, and leucine-rich α-2-glycoprotein 1. Dot blot analysis of an independent urine sample set pointed out fibrinogen ß chain and α-1-antitrypsin as most interesting biomarkers having sensitivity and specificity values in the range of 66-85%. Exploring the Human Protein Atlas (HPA) also revealed that bladder cancer tumors are the likely source of these proteins. They have the potential of being useful in diagnosis, monitoring of recurrence and thus may improve the treatment of bladder tumors, especially nonmuscle invasive tumors.

Heme binding in gas-phase holo-myoglobin cations: distal becomes proximal?

His64 and His93 are the two well-known sites of heme binding in water-dissolved holo-myoglobin, with His93 being a proximal, strongly binding partner, while the distal His64 weakly coordinates to the heme through a small-molecule ligand, e.g., water or O(2). The heme bonding scheme in a water-free environment is as yet unclear. Here we employed electron transfer dissociation tandem mass spectrometry to study the preferential attachment site of the ferri-heme (Fe(3+)) in electrospray-produced 12+, 14+, and 16+ holo-myoglobin ions. Contrary to expectations, in lower-charge complexes that should have a structure resembling that in solution, the heme seems to be preferentially attached to the "distal" histidine. In contrast, in the highest studied charge state, the "proximal" histidine is the site of preferential attachment; the 14+ charge state is an intermediate case. This surprising finding raises a question of heme coordination in proteins transferred to water-free environment, as well as the effect of the protonation sites on heme bonding.

Alzheimer's disease and mild cognitive impairment are associated with elevated levels of isoaspartyl residues in blood plasma proteins.

Increased levels of isoaspartyl residues (isoAsp) have previously been found in proteins of Alzheimer's disease (AD) brains and in blood proteins of patients suffering from uremia, the disease sharing common pathological features with AD. One can hypothesize that higher levels of isoAsp should be present in blood proteins of AD patients. Also, because of higher AD prevalence in females, they can be expected to have higher level of isoAsp than males. Here we modified our recently developed proteome-wide isoAsp analysis approach for testing these hypotheses. Eight blood plasma samples pooled from 218 individuals suffering from either mild cognitive impairment (MCI) or AD were analyzed by tandem mass spectrometry using electron transfer dissociation. Based on specific fragmentation pattern of isoAsp, the healthy controls were found to contain lower level of isoAsp compared with age-matched MCI and AD patients (p = 0.03). This result was further validated (p = 0.05) by 96 individual sample analyses, giving the combined value of p ≈ 0.01. Female pooled samples were found to contain higher level of isoAsp than male in both pooled and individual samples, with overall p ≈ 0.01. These findings verify the above hypotheses, and provide protein candidates for further investigation of the link between isoAsp and AD.

Calibration function for the Orbitrap FTMS accounting for the space charge effect.

Ion storage in an electrostatic trap has been implemented with the introduction of the Orbitrap Fourier transform mass spectrometer (FTMS), which demonstrates performance similar to high-field ion cyclotron resonance MS. High mass spectral characteristics resulted in rapid acceptance of the Orbitrap FTMS for Life Sciences applications. The basics of Orbitrap operation are well documented; however, like in any ion trap MS technology, its performance is limited by interactions between the ion clouds. These interactions result in ion cloud couplings, systematic errors in measured masses, interference between ion clouds of different size yet with close m/z ratios, etc. In this work, we have characterized the space-charge effect on the measured frequency for the Orbitrap FTMS, looking for the possibility to achieve sub-ppm levels of mass measurement accuracy (MMA) for peptides in a wide range of total ion population. As a result of this characterization, we proposed an m/z calibration law for the Orbitrap FTMS that accounts for the total ion population present in the trap during a data acquisition event. Using this law, we were able to achieve a zero-space charge MMA limit of 80 ppb for the commercial Orbitrap FTMS system and sub-ppm level of MMA over a wide range of total ion populations with the automatic gain control values varying from 10 to 10(7).