Magnus Wetterhall

Temporally resolved differential proteomic analysis of human ventricular CSF for monitoring traumatic brain injury biomarker candidates

A shotgun proteomic approach based on nanoflow liquid chromatography (nanoLC) in conjunction with matrix assisted laser desorption/ionization time of flight tandem mass spectrometry (MALDI TOF MS/MS) was utilized to quantitatively analyze the protein content of consecutive ventricular cerebrospinal fluid (CSF) samples of severe traumatic brain injury (TBI) patients on an individual basis. CSF was acquired from the lateral ventricle 1-9 days after the TBI incident by canula drain to investigate temporally resolved protein changes in three patients that required intracranial pressure monitoring during neurointensive care. The samples were subjected to at once tryptic digestion followed by isobaric tag labeling before multiplexed peptide separation and MS analysis. By using this approach, we were able to follow characteristic changes in protein concentrations over time allowing new conclusions to be drawn about ongoing pathological processes during TBI. Certain suggested protein-biomarker candidates for TBI, like acute phase reactants (APRs), fibrinogens (FIB), cystatin C (CC) or more brain specific proteins like glial fibrillary acid protein (GFAP) and neuron-specific enolase (NSE) were found to be significantly up-regulated which is in strong consistence with previously reported results. This methodology appears to be a promising tool for studying candidate biomarkers of neurovascular and traumatic brain injuries in the neurointensive care setting.

Rapid capillary electrophoresis time-of-flight mass spectrometry separations of peptides and proteins using a monoquaternarized piperazine compound (M7C4I) for capillary coatings.

A monoquaternarized piperazine, 1‐(4‐iodobutyl) 4‐aza‐1‐azoniabicyclo[2,2,2] octane iodide (M7C4I), has been evaluated as a surface derivatization reagent for CE in combination with TOF MS for the analysis of proteins, peptides, and protein digests. The M7C4I piperazine, at alkaline pH, forms a covalent bond via alkylation of the ionized silanols producing a cationic surface with a highly stable and reversed EOF. The obtained surface yields rapid separations (less than 5 min) of peptides and proteins at acidic pH with high separation efficiencies (up to 1.1×106 plates/m for peptides and up to 1.8×106 plates/m for proteins) and no observed bleeding of the coating reagent into the mass spectrometer. The simplicity of the coating procedure also enables fast (2 min) regeneration of the surface, if necessary. This is useful in the analysis of complex samples in order to prevent possible memory effects. The potential of using M7C4I‐coated capillaries for MS analysis of complex samples is demonstrated by the separation of peptides, proteins, and protein digests. Even more, the spectacular thing in which large intact proteins with molecular masses over 0.5 MDa could be separated. The coating showed good ability to handle these large proteins with high efficiency and retained peak shape as demonstrated by separation of IgG1 (150 kDa) and thyroglobulin (669 kDa).

Quantification of the Brain Proteome in Alzheimer’s Disease Using Multiplexed Mass Spectrometry

We have compared the brain proteome in the temporal neocortex between Alzheimers disease (AD) patients and non-AD individuals by using shotgun mass spectrometry based on a stable isotope dimethyl labeling. A total of 827 unique proteins were identified and quantitated. Of these, 227 proteins were found in at least 9 out of 10 AD/control pairs and were further subjected to statistical analysis. A total of 69 proteins showed different levels (p-value < 0.05) in AD versus control brain samples. Of these proteins, 37 were increased and 32 were decreased as compared to the non-AD subjects. Twenty-three proteins comprise novel proteins that have not previously been reported as related to AD, e.g., neuronal-specific septin-3, septin-2, septin-5, dihydropteridine reductase, and clathrin heavy chain 1. The proteins with altered levels in the AD brain represent a wide variety of pathways suggested to be involved in the disease pathogenesis, including energy metabolism, glycolysis, oxidative stress, apoptosis, signal transduction, and synaptic functioning. Apart from leading to new insights into the molecular mechanisms in AD, the findings provide us with possible novel candidates for future diagnostic and prognostic disease markers.

Methodological Aspects on Microdialysis Protein Sampling and Quantification in Biological Fluids: An In Vitro Study on Human Ventricular CSF

There is growing interest in sampling of protein biomarkers from the interstitial compartment of the brain and other organs using high molecular cutoff membrane microdialysis (MD) catheters. However, recent data suggest that protein sampling across such MD membranes is a highly complex process that needs to be further studied. Here, we report three major improvements for microdialysis sampling of proteins in complex biological matrixes. The improvements in this in vitro study using human ventricular cerebrospinal fluid as the sample matrix include increased fluid recovery control, decreased protein adsorption on the microdialysis membrane and materials, and novel quantitative mass spectrometry analysis. Dextrans in different concentrations and sizes were added to the perfusion fluid. It was found that dextrans with molecular mass 250 and 500 kDa provided a fluid recovery close to 100%. An improved fluid recovery precision could be obtained by self-assembly triblock polymer surface modification of the MD catheters. The modified catheters also delivered a significantly increased extraction efficiency for some of the investigated proteins. The final improvement was to analyze the dialysates with isobaric tagged (iTRAQ) proteomics, followed by tandem mass spectrometric analysis. By using this technique, 48 proteins could be quantified and analyzed with respect to their extraction efficiencies. The novel aspects of microdialysis protein sampling, detection, and quantification in biological fluids presented in this study should be considered as a first step toward better understanding and handling of the challenges associated with microdialysis sampling of proteins. The next step is to optimize the developed methodology in vivo.

CE MALDI-TOF/TOF MS for multiplexed quantification of proteins in human ventricular cerebrospinal fluid

CE, interfaced off‐line to MALDI‐TOF/TOF MS, was for the first time used to quantitatively monitor the protein content in complex biological and clinical samples with iTRAQ™ labeling. The usefulness and advantage of iTRAQ™ labeling, in combination, with CE MALDI‐TOF/TOF MS is demonstrated on mixtures of protein standards and by a case study on human ventricular cerebrospinal fluid samples collected from a patient with traumatic brain injury during patient recovery. Mixtures of five standard proteins were initially analyzed to optimize the experimental conditions for the CE MALDI‐MS and MS/MS analysis. The interactions of proteins and peptides with the capillary inner wall during CE separation were minimized using PolyE‐323 modified capillaries. The analysis of the ventricular cerebrospinal fluid samples yielded 43 significantly (p<0.05 MudPIT scoring) identified proteins that could be quantitatively monitored over time. The identified changes in protein levels for several of these proteins are well in line with the reports from previous studies on protein patterns that could be related to the post‐traumatic processes of traumatic brain injury. This study shows that the presented approach, combining isobaric tags with CE MALDI‐TOF/TOF MS, is a useful choice for quantitative proteomic analysis.

Optimization of capillary electrophoresis conditions for coupling to a mass spectrometer via a sheathless interface

When optimizing a capillary electrophoresis/electrospray ionization mass spectrometry (CE/ESI-MS) system, consideration has to be given not only to the separation but also to the electrospray stability. Methods developed for CE/UV analysis of drugs and peptides were considered and modified to be suitable for a CE/MS system with a robust sheathless interface. Different concentrations of the organic modifiers acetonitrile, methanol and 2-propanol were used in the separation buffer. The type and concentrations of these modifiers were also compared with reference to electrospray stability, sensitivity and time of analysis. In addition, different ionic strengths in the buffers were evaluated with reference to electrospray stability. The repeatability was used for the estimation of electrospray stability. The degree to which these parameters influenced the separation and the ESI stability was studied using a nine-peptide standard mixture and the antibiotic drugs bacampicillin and ampicillin as test substances. The analysis time and resolution were used as measures of the efficiency of the separation. A time-of-flight MS analyzer was used since it has the potential advantages of becoming a better fit for integration of CE with MS owing to the speed and sensitivity of this mass analyzer. The detection limit, i.e. 1 microM, for bacampicillin was comparable to what could be achieved with CE/MS on a quadrupole instrument using selected ion monitoring and sheath flow ESI.

Mining ventricular cerebrospinal fluid from patients with traumatic brain injury using hexapeptide ligand libraries to search for trauma biomarkers

Traumatic brain injury (TBI) is an acute event resulting from external force to the brain and is a major cause of death and disability associated with high health care costs in the western world. Additional injuries, originating from the secondary molecular events after the initial intensive care, may be limited by the use of objective biomarkers to provide the best treatment and patient prediction outcome. In this study, hexapeptide ligand libraries (HLL) have been used for the enrichment of suggested protein biomarkers for TBI in cerebrospinal fluid (CSF). HLL have the potential to enrich low abundant proteins and simultaneously reduce the high abundant proteins, rendering a sample with significantly reduced dynamic range. The CSF proteome from two TBI inflicted patients have been extensively mapped using a large initial sample volume obtained by extraventricular drainage. Shotgun proteomics, in combination with isoelectric focusing (IEF) and nano-LC-MS/MS, identified 339 unique proteins (MudPIT scoring p < or = 0.05) with a protein overlap of 130 between the patients. As much as 45% of the proteins reported in the literature to be associated with degenerative/regenerative processes occurring after a trauma to the head were identified. Out of the most prominent potential protein biomarkers, such as neuron specific enolase, glial fibrillary acidic protein, myelin basic protein, creatine kinase B-type and S-100beta, all except myelin basic protein were detected in the study. This study shows the possibility of using HLL as a tool for screening of low abundant protein biomarkers in human CSF.

Effect of sequence length, sequence frequency, and data acquisition rate on the performance of a Hadamard transform time-of-flight mass spectrometer

Various factors influencing the performance of a Hadamard transform time-of-flight mass spectrometer (HT-TOFMS) have been investigated. Using a nitrogen corona discharge to produce an ion stream of N2+, N3+, and N4+, it is found for spectra containing only N4+ that the signal-to-noise ratio (SNR) closely approaches the value calculated from the ion background by assuming that the ion background follows a Poisson distribution. In contrast, for a more intense beam containing N2+, N3+, and N4+, the SNR is less than its theoretical value because of the appearance of discrete spikes in the mass spectrum caused by deviations in the actual modulation sequence from the ideal one. These spikes can be reduced, however, by decreasing the modulation voltage. Under these optimized conditions, the pseudo-random sequence length is varied to understand how it alters SNR, mass resolution, and scan speed. When the length of the pseudo-random sequence is doubled, the SNR increases by √2 while the time necessary to record a mass spectrum also doubles. Mass resolution can be varied between 500 and 1200 at m/z = 609 as the sequence length, modulation speed (10 MHz, 25 MHz), and acquisition rate (up to 50 MHz) are changed. Scan speeds of 6000 passes per s can be obtained using a sequence containing 4095 elements modulated at 25 MHz. The capability to tailor the HT-TOFMS to increase the scan speed and resolution with a constant 50% duty cycle makes the technique extremely appealing as a mass analyzer for measuring rapid changes in the composition of an ion stream.

Comparative study of label and label-free techniques using shotgun proteomics for relative protein quantification.

The analytical performance of three different strategies, iTRAQ (isobaric tag for relative and absolute quantification), dimethyl labeling (DML) and label free (LF) for relative protein quantification using shotgun proteomics have been evaluated. The methods have been explored using samples containing (i) Bovine proteins in known ratios and (ii) Bovine proteins in known ratios spiked into Escherichia coli. The latter case mimics the actual conditions in a typical biological sample with a few differentially expressed proteins and a bulk of proteins with unchanged ratios. Additionally, the evaluation was performed on both QStar and LTQ-FTICR mass spectrometers. LF LTQ-FTICR was found to have the highest proteome coverage while the highest accuracy based on the artificially regulated proteins was found for DML LTQ-FTICR (54%). A varying linearity (k: 0.55-1.16, r(2): 0.61-0.96) was shown for all methods within selected dynamic ranges. All methods were found to consistently underestimate Bovine protein ratios when matrix proteins were added. However, LF LTQ-FTICR was more tolerant toward a compression effect. A single peptide was demonstrated to be sufficient for a reliable quantification using iTRAQ. A ranking system utilizing several parameters important for quantitative proteomics demonstrated that the overall performance of the five different methods was; DML LTQ-FTICR>iTRAQ QStar>LF LTQ-FTICR>DML QStar>LF QStar.

Analysis of membrane and hydrophilic proteins simultaneously derived from the mouse brain using cloud-point extraction

In this study, a temperature-induced phase fractionation known as cloud-point extraction (CPE) with the non-ionic surfactant Triton X-114 was used to simultaneously extract, concentrate, and fractionate hydrophobic and hydrophilic proteins from mouse brain tissue. Two bottom-up proteomic techniques were used to comprehensively identify the extracted proteins. The first “shotgun”-based approach included tryptic digestion of the proteins followed by reversed-phase nanoliquid chromatography (RP-nanoLC) in combination with electrospray ionization (ESI) tandem mass spectrometry (MS/MS). In the second approach, the extracted intact proteins were first separated by one-dimensional (1D) gel electrophoresis and then in-gel digested with trypsin and analyzed with nanoLC-MS/MS. In total, 1,825 proteins were unambiguously identified and the percentage of membrane proteins was 26% which is at the reported genome expression levels of 20–30%. The protein overlap between the two approaches was high. The majority (77%) of the identifications in the first approach was also found by the second method. The protein overlap between the CPE-extracted hydrophilic and hydrophobic fractions was rather small (16–23%) for both methods, which indicates a good phase separation. A quantitative evaluation of the CPE with iTRAQ labeling and nanoLC-ESI-MS/MS analysis gave iTRAQ ratios at the expected levels and an overall variation of the entire method at 17–31%. The results indicate very reproducible sample preparation and analysis methods that readily can be applied on large-scale sample sets.