FASEB J.
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Haqqani, A. S.
Right arrow Articles by Stanimirovic, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Haqqani, A. S.
Right arrow Articles by Stanimirovic, D.
(The FASEB Journal. 2005;19:1809-1821.)
© 2005 FASEB

Characterization of vascular protein expression patterns in cerebral ischemia/reperfusion using laser capture microdissection and ICAT-nanoLC-MS/MS

Arsalan S. Haqqani, Momir Nesic, Ed Preston, Ewa Baumann, John Kelly and Danica Stanimirovic1

Institute for Biological Sciences, National Research Council of Canada, Ottawa, Ontario, Canada

1 Correspondence: Institute for Biological Sciences, National Research Council of Canada, 1200 Montreal Road, Bldg. M-54, Ottawa, ON, K1A 0R6, Canada. E-mail: danica.stanimirovic{at}nrc-cnrc.gc.ca


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Cerebral ischemia rapidly initiates structural and functional changes in brain vessels, including blood-brain barrier disruption, inflammation, and angiogenesis. Molecular events that accompany these changes were investigated in brain microvessels extracted using laser-capture microdissection (LCM) from Sprague-Dawley rats subjected to a 20 min transient global cerebral ischemia followed by 1, 6, or 24 h reperfusion. Proteins extracted from ~300 LCM captured microvessels (20–100 µm) were ICAT-labeled and analyzed by nanoLC-MS. In-house software was used to identify paired ICAT peaks, which were then sequenced by nanoLC-MS/MS. Pattern analyses using k-means clustering method classified 57 differentially expressed proteins in 7 distinct dynamic patterns. Protein function was assigned using Panther Classification system. Early reperfusion (1 h) was characterized by down-regulation of ion pumps, nutrient transporters, and cell structure/motility proteins, and up-regulation of transcription factors, signal transduction molecules and proteins involved in carbohydrate metabolism. The up-regulation of inflammatory cytokines and proteins involved in the extracellular matrix remodeling and anti-oxidative defense was observed in late reperfusion (6–24 h). The up-regulation of IL-1ß and TGF-1ß in ischemic brain vessels was confirmed by ELISA, quantitative PCR, and/or immunohistochemistry. A biphasic postischemic (1 and 24 h) BBB opening for 3H-sucrose was evident in the same model. Differentially expressed proteins identified in brain vessels during reperfusion are likely involved in orchestrating functional vascular responses to ischemia, including the observed BBB disruption.—Characterization of vascular protein expression patterns in cerebral ischemia/reperfusion using laser capture microdissection and ICAT-nanoLC-MS/MS. Haqqani, A. S., Nesic, M., Preston, E., Baumann, E., Kelly, J., Stanimirovic, D.


Key Words: blood-brain barrier • protein extraction • peptides


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
THE NEUROVASCULAR(NV) unit, broadly defined as a segment of brain vasculature, integrates three principal functionalities: the blood-brain barrier (BBB), regulation of the cerebral blood flow, and neuroimmune interfacing (1) . The NV unit is composed of functionally integrated cellular (including brain endothelial cells, astrocytes, pericytes, and smooth muscle cells) and acellular elements that form the basement membrane. Current understanding of the NV unit underscores the importance of multicellular interplay in the regulation of BBB function (2) .

Brain insults (e.g., stroke, trauma) or conditions (e.g., brain tumors) characterized by tissue hypoxia induce dynamic structural and functional changes in the NV unit, including BBB breakdown, vasoparalysis and hypoperfusion, leukocyte adhesion and infiltration, prothrombotic conversion, angio- and vasculogenesis (2 , 3) . At the microanatomy level, alterations include disruption of inter-endothelial tight junctions, retraction of pericytes from the abluminal surface of the capillary, breakdown of the basal lamina with transudation of plasma, endothelial cell migration, and proliferation (2 , 3) . This remodeling requires de novo synthesis of genes and proteins that enable dynamic changes in cell communication with other cells as well as extracellular matrix (ECM) (4) .

Vascular expression of various mediators during ischemic brain disease has been studied using "classical" approaches of in situ hybridization or immunohistochemistry (5 6 7 8 9) . However, the complexity of the NV unit remodeling in response to hypoxic or ischemic states necessitates integration of global molecular analyses enabled by genomics and proteomics technologies with correlative functional outcome measures. Recently, a study using targeted microarray of predominantly vascular genes implicated in angiogenesis demonstrated that most of the well-known angiogenic and vessel-stabilizing factors such as thrombospondins increased as early as 1 h after ischemia in peri-infarct brain tissues (7) . However, attempts to selectively analyze vascular molecular fingerprints have been hampered by use of whole brain tissue where specific vascular changes are easily masked by abundant parenchyma cells.

To examine molecular changes that occur specifically in cerebral vessels, we recently a developed laser capture microdissection (LCM) microscopy technique to extract microvessels from brain sections (10 , 11) . Microarray gene expression analyses of LCM-captured tissues require the development of nucleic acid amplification protocols where the quantitative relationship between different transcripts is preserved (12) . Global protein analyses of such samples would provide a greater value in understanding changes in translated protein effectors of biological functions. However, isolation of a large number of cells (>20,000) by LCM is usually required to extract enough protein for analysis by gel-based proteomic methods (13) ; this method is not feasible for analyses of multiple samples. In this study, we have combined for the first time LCM isolation of brain microvessels with a non-gel, ICAT-based proteomic method to examine the biologically significant vascular protein expression changes in the animal model of global cerebral ischemia/reperfusion. This is the first report that describes the dynamic patterns of vascular protein expression that accompany and may participate in events linked to the BBB disruption, inflammatory endothelial activation, and initiation of cell cycle and angiogenesis.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Cerebral ischemia/reperfusion
The global cerebral ischemia/reperfusion model was chosen for this study since it was expected that a large number of brain vessels would be required for a reliable proteomic analyses and, in contrast to focal ischemia models, an ischemic injury to the vessels would be more uniform and controllable in forebrain regions supplied by carotid arteries.

Sprague-Dawley rats of both sexes weighing 300–350 g were obtained from Charles River Laboratories, Inc. (Wilmington, MA, USA). Rats were cage acclimated for at least 6 days prior to surgery and housed in pairs with free access to Purina rat chow and water. Bilateral cerebral ischemia using the two-vessel occlusion (2VO) model (14) was carried out under sodium pentobarbital anesthesia (65 mg/kg i.p.) and mechanical ventilation with a 30:70 O2:N2 mixture. Physiological monitoring and appropriate adjustments to ventilation and warming devices were carried out to ensure presence of acceptable values for tympanic and rectal temperature (37.5°–38.0°C), blood pH (7.35–7.45), pCO2 (35–45 mm Hg), and pO2 (>90 mm Hg). Bilateral carotid clamping for 20 min was combined with blood withdrawal through a tail artery cannula to maintain arterial pressure at 42–47 mm Hg. After blood reinfusion, clamp removal, and wound closure, rats were taken off ventilation, maintained normothermic (37.0°–37.5°C) until recovery from anesthetic, then returned to housing. Control animals were sham operated; their arteries were not clamped and blood pressure was not lowered. After surgery, the animals were allowed to recover for 1, 6, or 24 h, decapitated and brains were rapidly removed, wrapped in aluminum foil, and placed on dry ice. Separate groups of animals subjected to the above protocols were used to measure BBB permeability and for proteomics analyses of brain vessels. Experiments were carried out at the Animal Care and Veterinary Services of the National Research Council (Ottawa, ON, Canada) in accordance with guidelines of the Canadian Council on Animal Care.

Blood-brain barrier permeability measurements
At 1, 6, or 24 h after 2VO ischemia, rats were anesthetized with sodium pentobarbital, a femoral artery cannula inserted, and direct femoral i.v. injections of 0.3 and 0.5 mL saline were delivered containing, respectively, sodium heparin (300 U) followed by 30 Ci [3H]sucrose (NET 341, Perkin Elmer Life Sciences Inc., Wellesley, MA, USA). A syringe pump was immediately started (time 0) to withdraw a total 0.5 mL arterial blood sample over the next 30 min. At 0+25 min the right carotid artery was cannulated; at 0+30 min the heart was stopped with direct pentobarbital injection, jugular veins severed, and the head vasculature cleared of blood by carotid perfusion of 25 mL saline. Weighed samples of entire cerebral cortex and 50 µL volumes of plasma were prepared for liquid scintillation counting. The concentration of each tracer was determined in brain parenchyma (Cparen, dpm/g) and plasma (Cplasma, dpm/mL). The latter value was corrected for dead space saline dilution and multiplied by the circulation time (1800 s) to provide the time-integrated plasma concentration (0 {int} 1800Cplasma dt, dpm·s·mL–1). Transfer constants (Kis) were calculated from the following relationship based on a two-compartment model for BBB tracer permeation: Ki = Cparen/0 {int} 1800Cplasmadt (15 16 17) .

Tissue sectioning and vessel staining
Frozen rat brains were embedded in Tissue-Tec freezing medium (Miles Laboratories, Inc., Elkhart, IN, USA) and sectioned in a cryostat (Jung CM3000, Leica, Richmond Hill, ON, Canada) at 8 µm thickness. Sections were placed on Superfrost Plus microscope slides (Fisher Scientific, Nepean, ON, Canada) and kept at –80°C until use. The sections were fixed in 75% ethanol for 30 s, rinsed in water, stained with Ricinus Communis agglutinin I (RCA I) for 2 min to detect brain vessels as described elsewhere (10) , rinsed in water, and dehydrated by a sequential exposures to 70%, 96%, and 100% ethanol for 30 s each. Slides were air-dried for 1 min and used for laser-capture microdissection of vessels. The tissue fixation, staining, and dehydration procedures used are fast and mild, and result in excellent preservation of RNA (10 , 11) .

Laser capture microdissection of brain vessels
LCM of cortical vessels was performed using a PixCell IIe Laser-capture Microscope (Arcturus, Mountain View, CA, USA), and high-sensitivity (HS) caps (CapSure LCM Caps, Arcturus, CA, USA), and ethylene vinyl acetate (EVA) membranes. Complete details of the protocol have been described elsewhere (10 , 11) . Briefly, LCM of cortical brain vessels was performed using a 7.5 µm laser spot size and pulse power of 20–40 mW. Approximately 50–60 cortical microvessels ranging in size from 10 to 50 µm were captured onto each cap and at least 5 caps were captured (one per each of 5 sections) for each animal. The captured microvessel fraction contained capillaries, arterioles, and venules. Upon completion of microdissection, each cap was placed in an individual 0.5 mL Eppendorf tube (Brinkmann Instruments Ltd., Mississauga, ON, Canada) and frozen at –80°C until protein extraction. Vessels captured using these procedures are enriched in endothelial and blood-brain barrier markers (10) , but also have minimal GFAP contamination by RT-PCR (10) and likely contain pericytes and basement membrane fragments.

Experimental design
Since there were no other examples in published literature of ICAT applications to time series experiments, the experimental design was based on parallels between microarray and ICAT analyses: both involve labeling of molecules from two different samples and their mixing to compare relative expression levels of these molecules in a single experiment.

A schematic of the experimental protocol is shown in Fig. 1 . All protein extracts from the 12 sham-operated animals were pooled to generate a universal reference sample (Fig. 1) , which was then labeled with light ICAT reagent. Extracts from the stroke-induced animals were kept separate during the labeling process. Each stroke sample was then mixed with the pooled reference group and processed as described in Fig. 1 . Pooling control samples (sham) minimizes the differences due to subject-to-subject variation, making differential expression easier to find. This is often desirable when primary interest is not on the individual, but rather on characteristics of the population from which individuals are obtained, such as identifying biomarkers or expression patterns common across individuals. The common reference design has not been found to perform worse than nonpooled designs (18) .



View larger version (40K):
[in this window]
[in a new window]
  		Figure 1. A schematic of the experimental design and protocols used in the study. Animals were subjected to a 20 min transient global cerebral ischemia and reperfused for 1, 6, or 24 h; sham-operated controls were treated as described in Materials and Methods. Cerebral vessels were extracted by LCM from several brain sections of each animal. Protein extracts from sham-operated animals were pooled to create a universal reference against which each ischemic animal was compared. ICAT labeling, ICAT peak analyses, peptide sequencing and data analyses protocols are as described in Materials and Methods.

Protein extraction, ICAT labeling, and purification
EVA membranes from each cap were carefully removed and placed in protein extraction buffer (50 mM Tris-HCl, pH 8.5, 0.1% SDS). These were incubated, with shaking, at 65°C for 1 h to extract proteins. After extraction, samples originating from the same animals were pooled (5 caps per animal). All the proteins from the 12 sham-operated animals were also pooled and divided back into 12 equivalent fractions. The protein samples from the 12 ischemia-induced animals were kept separate. Protein yield was ~1 µg from 300 LCM-captured microvessels (collected from 5 sections in each animal).

Cleavable ICAT reagents (Applied Biosystems, Foster City, CA, USA) were used for this work. The reagents consist of a cleavable biotin group, an isotopic tag (heavy or light) and a cysteine-reactive group. The light reagent contains nine 12C atoms and the heavy reagent contains nine 13C atoms; the 9 Da difference allows the peptides labeled with light or heavy reagent to be readily distinguished on a nanoLC-MS spectrum (see Fig. 3B ). Protein samples from 12 ischemia-induced animals were separately labeled with an isotopically heavy ICAT reagent while the 12 equivalent samples from sham-operated animals were separately labeled with an isotopically light ICAT reagent. The labeling was carried out in denaturing buffer (50 mM Tris-HCl, pH 8.5, 0.1% SDS, 1 mM TCEP) as recommended by the manufacturer. The labeling reaction involves alkylation of cysteine residues in proteins by excess amounts of the ICAT reagent. According to the manufacturer’s manual, the labeling efficiency of cleavable ICAT reagent is >98%. The conditions used were previously optimized by Aebersold and co-workers (19) such that the labeling is complete and efficient at various protein concentrations.



View larger version (40K):
[in this window]
[in a new window]
  		Figure 3. Standardization, quantification, and normalization of ICAT protocols. A) Scatter plots showing intensities of ICAT pairs quantified by nanoLC-MS analysis. Left panel: Bovine serum albumin (BSA) was labeled with either heavy or light ICAT reagent and mixed at 1:1, 2:1 or 1:2 H:L ratios. Right panel: Protein from brain endothelial cells were labeled with either heavy or light reagent and mixed at a 1:1 ratio. B) A typical spectrum of an ICAT-labeled LCM sample. Proteins from ischemic vessels were labeled with heavy ICAT reagent and those from controls were labeled with light reagent. Three examples of ICAT pairs (2+ charge) from the spectrum are shown in more details (arrows) that indicate a 1.8-fold down-regulation, a no change (1.0) or a 5.7-fold up-regulation in ischemic samples compared with control. Normalized values are in brackets. C) Log-log scatter plots of nanoLC-MS-quantified intensities of ICAT pairs from one of the 12 samples before and after global median normalization. Log10 of peak intensities of heavy vs. light ICAT-labeled pairs were plotted. Factorlines showing 1.5-fold (dotted) and 2.0-fold (solid) up- or down-regulation are also shown. An M vs. A plot of the normalized data to visualized any skewness of the data. Log2(HL)/2 vs. Log2(H/L) were plotted, where H and L are normalized peak intensities of heavy and light ICAT-labeled pairs, respectively.

After labeling, each heavy ICAT labeled ischemic sample was mixed with an equivalent amount of the light ICAT-labeled pooled control sample and digested with 30 µg of trypsin (Promega, Madison WI, USA) at 37°C for 16 h. The resulting peptides were desalted using strong cation exchange chromatography (Applied Biosystems), the biotinylated ICAT-labeled peptides were affinity purified on an Avidin column (Applied Biosystems), and the biotin group was cleaved off using trifluoroacetic acid, all according to the manufacturer’s protocol.

Mass spectrometric (MS) analysis
A hybrid quadrupole time-of-flight MS (Q-TOFTM Ultima, Waters, Millford, MA, USA) with an electrospray ionization source (ESI) and an online reverse phase nanoflow liquid chromatography column (nanoLC, 0.3 mm x 15 cm PepMap C18 capillary column, Dionex/LC-Packings, San Francisco, CA, USA) was used for all analyses. The ICAT-labeled samples were separated on the nanoLC column using a gradient of 5–95% acetonitrile 0.2% formic acid in 50 min, 3.5 µL/min supplied by a CapLC HPLC pump (Waters). Analysis of each of the 12 ICAT-labeled samples was done in two steps. In the first step, 5% of the samples were analyzed by nanoLC-MS in a survey (MS-only) mode to determine the ratio of all the heavy and light ICAT pairs and to identify differentially expressed pairs as described in the data analysis section. The differentially expressed pairs were included in a "target list." In the second step, the samples were reinjected (5%) into the mass spectrometer and only the peptides included in the target list were sequenced in a nanoLC-MS/MS mode. MS/MS spectra were obtained only on 2+, 3+, and 4+ ions. These were then submitted to Mascot® search engine (Matrix Science Ltd., London, UK) (20) to search against a NCBI nonredundant, trypsin-digested (allowing 1 missed cleavage) rodentia database. Initial searches were restricted to cysteine-containing peptides containing a variable modification with either heavy or light ICAT. Individual ion scores indicating identity or extensive homology (P<0.05) were considered significant (20) . The searches were also done without any restrictions to confirm that an ion identified as an ICAT peptide in a restricted search is not identified as another peptide with a significant score in an unrestricted search. The MS/MS spectrum of each identified ICAT-labeled peptides was manually examined and confirmed.

Data analysis
From the nanoLC-MS analysis of each ICAT-labeled sample, the peaks in all MS scans were background-subtracted, Savitzky-Golay-based smoothed and centroided using MasslynxTM software v4.0 (Waters), and exported as a list of text files (~2500–3000 files per nanoLC-MS analysis). Each file contained mass/charge (m/z) and intensity information for all the peaks in a scan. The text files were analyzed as follows using Perl-based software that was developed in-house. The charge state of each peak was determined and the peaks were deisotoped. Elution range (single ion chromatogram) for each 2+, 3+, and 4+ ion was determined and an integrated intensity was calculated. ICAT pairs were identified as all the coeluting peptide pairs separated by 9 Da (one cysteine) or 18 Da (two cysteine). It was necessary to normalize the intensities of the heavy and light peaks to correct for possible unequal mixing of the two samples. This correction was necessary in LCM samples since the starting amount of protein was very small and could not be reliably quantified using protein assays. A global median normalization method was used to correct for this. The ratio of corrected intensities of heavy and light pairs corresponding to fold change was calculated (ratio=IntensityHeavy/IntensityLight) and graphed on a scatter plot to visualize the distribution of the overall expression. The median H:L ratio in all 12 ICAT samples was between 0.85 and 1.2, with the expected ratio being 1, suggesting that the difference in total protein levels between sham and experimental (ischemic) samples was < 20%. ICAT pairs showing differential expression (up- or down-regulated by 1.5-fold) were manually confirmed and included in a target list for subsequent sequencing by nanoLC-MS/MS analysis as described above.

Statistical significance (Pvalue) of each differential expression was calculated using a statistical method described for detecting biologically significant expression changes in microarray data (21) using Statomics v0.3 sofware with R language for statistical computing and graphics (v1.9.1). Once a list of differentially expressed proteins were identified in the 1, 6, and 24 h groups, their expression patterns over time were clustered using k-means clustering method in Matlab 7.0 software. The proteins were also categorized according to their functional groups using Panther Classification system (panther.celera.com).

ELISA
In separate experiments, 600 microvessels captured from control or ischemic animals were used to determine IL-1ß by ELISA. The protein was extracted in 75 µL of modified Salter’s buffer (1.5% NP-40, O.75% DOC, and 0.3% SDS) by vortexing inverted LCM tubes for 1 h at 4°C. The protein extract was then sonicated for 3 intermittent 1 min intervals with a 1 min sample cooling on ice between pulses. Protein concentration in each sample was determined using DC protein assay (Biorad, Mississauga, ON). IL-1ß levels in 50 µL of LCM-captured vessel extracts were determined using ELISA kit (Biosource International, Camarillo, California, USA) as recommended by manufacturer.

Real-time quantitative PCR (Q-PCR)
SYBR® Green PCR Core Reagent Kits and an ABI Prism 7700 Sequence Detector System (Applied Biosystems Inc., Foster city, CA, USA) were used to conduct Q-PCR analyses in LCM-collected vessels using described previously methods (11) . The Q-PCR primers for TGF-1ß (fwd-5'-GCTGCTGACCCCCACTGAT; rev-5'-TGCCGGACAACTCCAGTGA) and ß-actin (fwd-5'-TGTCCACCTTCCAGCAGATGT; rev-5'-AGTCCGCCTAGAAGCATTTGC) were designed according to the published sequences from GenBank and synthesized by Applied Biosystems.

Immunohistochemistry
Brain sections used for LCM-assisted vessel extractions were incubated in methanol (EM Science, Gibbstown, NJ, USA) for 10 min at 4°C and washed three times using 0.2 M phosphate buffer. The sections were blocked with Universal Blocking Solution (Dako Diagnostics, Mississauga, ON, Canada) for 30 min at room temperature. The blocking solution was removed without washing and sections were incubated with 5 µg/mL of the primary polyclonal anti-TGFß-1 antibody (Cedarlane Laboratories, Hornby, Canada) diluted in antibody diluting buffer (Dako Diagnostics) for 24 h at 4°C. Sections were then washed three times in 0.2 M phosphate buffer (pH 7.3; 5 min/wash), and exposed to 3.74 µg/mL secondary antibody, Alexa Fluor goat anti rabbit IgG (Molecular Probes, Eugene, OR, USA) diluted in the antibody diluting buffer (Dako Diagnostics) for 30 min at room temperature. The slides were dehydrated by a sequential exposure to increasing concentrations of ethanol (30 s in 70% ethanol, 30 s in 96% ethanol, and 60 s in 100% ethanol), followed by a 5 min treatment with 99.98% xylenes (Anachemia Canada Inc., Montreal, QC). The slides were air-dried an additional 6 min. The sections were then double-stained with the 5.0 mg/mL FITC-tagged lectin RCA I diluted at 1:30 in DEPC water for 3 min. RCA I binds carbohydrate residues on endothelial cell surface and selectively stains rat brain vessels (11) . The sections were rapidly washed five times (10 s/wash) in a DEPC water, dehydrated by a sequential exposure to increasing concentrations of ethanol (Commercial Alcohols Inc., Brampton, ON), and dehydrates as above. The slides were air-dried for 6 min and observed under the Zeiss Axiovert 200 fluorescent microscope (Carl Zeiss, Maple Groove, MN, USA). The numbers of both RCA I- and TGFß-1-positive cortical vessels were counted and analyzed in three brain sections of each animal using Northern Eclipse v.5.0 software by someone blinded to the experimental design; the ratio of TGFß-1 immunopositive vessels to all RCA I-stained vessels was calculated for each animal.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Blood-brain barrier permeability after transient global cerebral ischemia
A 20 min transient forebrain ischemia induced by two vessel occlusion and hypotension produces a "typical" delayed neuronal damage in CA1 region of the hippocampus, and a diffuse neuronal loss in the neocortex (22 , 23 ). Histological analyses 24 h after ischemia show shrunken, eosinophilic neurons with pyknotic nuclei and increased mitochondrial injury (22) . The BBB disruption for both sucrose (15 , 17) and horseradish peroxidase (24) in the same ischemia model was more pronounced in the cerebral cortex then in the subcortical brain regions. Therefore, both BBB permeability and vascular proteomic studies were performed in cortical vessels.

In experiments in five nonischemic rats baseline transfer constants for the diffusion of sucrose across the intact cortical BBB ranged from 1.48 to 2.52 nL·g–1·s–1 (Fig. 2 ). Ki values for sucrose measured at different reperfusion times after a 20 min global cerebral ischemia (Fig. 2) showed a biphasic response, with values around 10 nL·g–1·s–1 at 1 h of reperfusion, followed by a transient reduction at 6 h reperfusion and a dramatic secondary increase at 24 h of reperfusion reaching values over 25 nL·g–1·s–1 (Fig. 2) .



View larger version (10K):
[in this window]
[in a new window]
  		Figure 2. Changes in brain transfer constants (Ki) for 3H-sucrose after a transient 20 min global cerebral ischemia. Forebrain ischemia followed by 1, 6, or 24 h of reperfusion was induced in rats and the blood-brain barrier permeability for 3H-sucrose was measured as described in Materials and Methods. Each bar is mean ± SD from 5 animals. *Significant (P<0.05; 1-way ANOVA) difference from sham-operated control (C).

ICAT standardization
To analyze protein expression changes in the neurovascular unit that accompany the observed BBB disruption after forebrain ischemia, we have combined recently developed methods for staining and isolation of multicellular microvascular structures from human (11) and rat brain using LCM (10) with a gel-free ICAT proteomics technique. ICAT labeling allows mass spectrometry-based quantification of relative amounts of protein in two samples, usually a diseased and a control (25) . To facilitate the quantification of multiple samples using ICAT, a universal reference experimental design was used.

Before analyzing the LCM-captured vessels, we examined whether ICAT-based proteomics is capable of accurate and reproducible quantification. In an initial experiment, the ability to accurately quantify different ratios of a standard protein was examined. Bovine serum albumin protein was labeled with either heavy (H) or light (L) ICAT reagent and mixed at 1:1, 2:1 or 1:2 H:L ratios. Each mixture was trypsin-digested and processed as described in Materials and Methods and analyzed by nanoLC-MS. Figure 3A (left panel) shows the scatter plot of the ICAT pairs identified in each mixture. The observed mean H:L ratios of the ICAT pairs were 0.999 ± 0.118, 1.93 ± 0.17 and 0.535 ± 0.167 for samples that were originally mixed at 1:1, 2:1 and 1:2 H:L ratios, respectively. In another experiment, the variability in the quantitative data was demonstrated using a biological sample. Proteins extracted from brain endothelial cells were labeled with either H or L reagent and mixed at 1:1 ratio. A scatter plot of the ICAT pairs identified in the sample is shown in Fig. 3A (right panel). After normalization, the average H:L ratio was of 1.00 ± 0.13 (mean±SD for 895 pairs; 99% confidence interval of ±0.011). More than 98% of the pairs (P=880) had ratios between –1.5 and 1.5. In subsequent analyses, we used a fold change of >1.5 or < –1.5 to define differential expression.

MS analysis of the LCM-captured vessels was done in two stages: 1) determination of relative peptide abundance and 2 ) MS/MS sequencing of differentially expressed peptides. For quantification, each of the 12 ICAT samples was analyzed by nanoLC-MS, and all heavy-light pairs of ICAT peaks (Fig. 3B ) were identified using in-house developed software. A total of 3109 ± 401, 2907 ± 194, and 3434 ± 648 different ICAT pairs were identified for 1, 6, and 24 h reperfusion groups, respectively (mean±SD of 4 animals). In addition, 114 ± 18, 77 ± 13, and 129 ± 22 single (unpaired) peaks were also identified for each group, respectively, which were potentially either heavy-alone or light-alone ICAT peaks.

Data normalization and detection of differentially expressed peptides
Normalization of data and determination of biologically significant differences are key processes in the analysis of large datasets. Once all of the ICAT peptide pairs and the potential unpaired peaks were identified in each sample, their intensities were visualized using Scatter plots (Log2 H vs. Log2 L) to assess the quality of the data (see Fig. 3C for rat 13 data). The intensities of the peaks were corrected using the global median normalization method (Fig. 3C ), which also allowed animal-to-animal comparisons. Skewness of the data, usually seen with microarray data analysis (26) , was not observed with the ICAT data as examined using M vs. A plots (0.5Log2 HL vs. Log2 H/L; Fig. 3C ). Fold change ratios (H/L) were then calculated for each sample. The in-house software was also able to compare the intensity ratios of each peptide pair in every sample since the same control was used for all. From this analysis, the mean coefficient of variation (COV) among the biological replicates was found to be 24 ± 5%, 26 ± 5%, and 27 ± 9% for the 1, 6, and 24 h group, respectively. ICAT pairs showing differential expression after normalization, defined as up or down-regulation by 1.5-fold, were included in a target list for sequencing analysis by nanoLC-MS/MS. Potentially unpaired ICAT peaks were also included in the list. Statistical significance (P-value) of each differentially expressed protein was calculated using a recently described statistical method for detecting biologically significant expression differences in microarray data (21) .

Sequencing of differentially expressed peptides from target list using nanoLC-MS/MS
The peptide extracts were reanalyzed by nanoLC-MS/MS and only those peptides on the target lists were selected for MS/MS sequencing. Approximately 50% of the peptide ions in these lists gave good quality MS/MS spectra that could be used for database searching and protein identification. Unpaired single peaks identified as ICAT peptides were given a fold change (H/L) value of +5 or –5 depending on the type of the label (H or L, respectively). From the three time groups, a total of 57 unique proteins were identified as being differentially expressed in cerebral vessels after global ischemia.

Temporal changes of identified proteins
Several proteins showed differential expression at one or two time points only. To study the kinetics of each protein, the expression values in the time points that did not show differential expression were identified by reverting back to the original nanoLC-MS data using the in-house software. The k-means pattern analysis, applied then on a complete dataset, grouped the proteins in 7 distinct dynamic patterns (Fig. 4 , Table 1 ).Proteins in group A and group B (Fig. 4) showed "reverse" patterns in that they were either transiently up-regulated or down-regulated, respectively, at 1 h reperfusion. Proteins in group C showed a transient down-regulation at 1 h and 6 h reperfusion (Fig. 4) . Proteins in group D were up-regulated from 1 to 24 h of reperfusion, whereas proteins in group E were up-regulated only at 24 h reperfusion (Fig. 4) . Proteins in groups F and G showed oscillatory patterns, with proteins in group F up-regulated at 6 h and proteins in group G at both 6 and 24 h reperfusion (Fig. 4) .



View larger version (28K):
[in this window]
[in a new window]
  		Figure 4. K-means clustering into 7 groups (A–G) of differentially expressed proteins identified by ICAT at various times of reperfusion after ischemia in LCM-isolated vessels. Proteins belonging to each group are listed in Table 1 .


View this table:
[in this window]
[in a new window]
  		Table 1. List of proteins showing differential expression at 1, 6, or 24 h in rat cerebral vessels after a transient global cerebral ischemiaa

Table 1 shows a list of all differentially expressed proteins assigned into dynamic pattern groups with appropriate identifiers that correspond to protein identity in Fig. 4 .

Functional protein categories
Differentially expressed proteins were categorized into 11 categories (Table 1 , Fig. 5 ) according to their participation in biological processes using Panther Classification System (panther.celera.com). Proteins belonging to some functional categories exhibited coordinate temporal dynamics (Fig. 5) . Cell structure and motility proteins were down-regulated at 1 h of reperfusion and returned to control levels at 6 and 24 h. Immediate early genes and inflammation proteins were up-regulated at all reperfusion time points (Fig. 5) , whereas proteins involved in the extracellular matrix remodeling exhibited prominent up-regulation only at 24 h reperfusion (Fig. 5) . Proteins assigned to each functional category in Fig. 5 were identified in Table 1 .



View larger version (20K):
[in this window]
[in a new window]
  		Figure 5. Functional classification of differentially expressed proteins identified by ICAT at times 1, 6, and 24 h of reperfusion after ischemia in LCM-isolated vessels into 11 functional categories. Proteins were classified using Panther Classification system (panther.celera.com). Each spot represents a fold change of one protein. The gray area represents a fold change of < 1.5. The functional category number to which each protein is classified is given in Table 1 .

Validation of observed changes with alternative techniques
To validate changes in some proteins observed by proteomic profiling, we selected IL-1ß and TGFß-1, classified into inflammatory and signal transduction functional categories. The up-regulation of IL-1ß in has been shown in cultured brain endothelial cells and astrocytes in response to hypoxic stress in vitro (27) , but not in vessels in vivo. A 2-fold up-regulation of IL-1ß protein in ischemic vessels at 6 h reperfusion was observed by both ICAT and ELISA (Fig. 6 ). TGFß-1 up-regulation in human and animal brain tissue after stroke has been linked to the pathogenesis of the angiogenic response (28 , 29) . TGFß-1 mRNA determined by Q-PCR was significantly up-regulated by 6 h of reperfusion and increased further to 75% above control levels by 24 h of reperfusion (Fig. 7 A). Consistent with TGFß-1 mRNA expression, ICAT proteomic analyses showed an initial increase of TFGß-1 protein at 6 h and a 3-fold up-regulation at 24 h (Fig. 7B ). Immunohistochemistry for TGFß-1 was performed in sections adjacent to those used for LCM capture of vessels and the ratio between number of vessels stained with lectin RCA I (green) and those showing TGFß-1 immunorectivity (red) was determined. An increased number (Fig. 7C ) of RCA I-stained vascular structures immunoreactive for TGFß-1 was detected at 24 h of reperfusion (Fig. 7G-I ) compared with sham-operated animals (Fig. 7D-F ). Increased TGFß-1 immunoreactivity was also seen in the brain parenchyma of ischemic animals (Fig. 7E, H ).



View larger version (13K):
[in this window]
[in a new window]
  		Figure 6. The expression of IL-1ß in LCM-captured brain vessels by ICAT (A) and ELISA (B). The analyses were performed as described in Materials and Method. Each bar is mean ± SE from 4 different animals. Asterisks indicate a significant (*P<0.05; **P<0.01; 1-way ANOVA) difference from sham-operated control.



View larger version (35K):
[in this window]
[in a new window]
  		Figure 7. Expression of TGFß-1 in LCM-captured brain vessels by Q-PCR (A) and ICAT (B) or in brain sections by immunohistochemistry (C–I). The analyses were performed as described in Materials and Methods. A, B) Each bar is mean ± SE from 4 different animals. *A significant (P<0.05; 1-way ANOVA) difference from sham-operated control. C–I) Double staining of brain vessels for RCA-1 (D ,G) and TGFß-1 (E, H) and their colocalization (F, I) in brain sections of sham-operated animals (D–F) and 24 h after global cerebral ischemia (G–I). Double staining procedures were as described in Materials and Methods. Vessel counts are shown in panel C. Each bar is mean ± SE of the number of vessels positive for either RCA I or TGFß-1 counted in 3 sections of each animal from 4 different animals.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
This study reports the first comprehensive analyses of dynamic protein expression changes in brain vessels affected by ischemia/reperfusion. Ischemic brain disease, a hallmark of various brain pathologies, including stroke, brain trauma, and expansive intracranial processes, is either caused by vascular pathology (e.g., thrombosis, aneurism) or aggravated by vascular changes that result in brain edema, vasoparalysis, or microcirculatory stasis (30) . "Targeting" vascular changes in these pathologies for either therapeutic intervention or imaging applications requires mapping of temporal molecular changes in brain vessels that participate in aberrant function(s). In this report, we describe the development and application of such methods for molecular mapping and discovery of vascular target proteins implicated in ischemic brain pathology.

Proteomic analyses of diseased tissues are usually performed using the whole tissue, where heterogeneous cell composition readily masks changes in specific tissue compartments (31) . The LCM-assisted extraction of defined cell types, while addressing cell selectivity, requires capture of large cell numbers (>20,000) for gel-based proteomic analyses (13) . To analyze ischemia-induced changes in brain vessels, a method to link LCM-assisted vessel capture with an emerging MS-based ICAT proteomic technology (13 , 25 , 32) was developed. Recently, two other reports described an application of either16O/18O isotopic sample labeling of LCM-extracted breast cancer cells (10,000 cells; 1–5 µg protein) (33) or cICAT labeling of 50,000–100,000 LCM-dissected hepatocellular carcinoma cells (34) coupled to LC-IT-MS/MS or 2D-LC-MS/MS analyses, respectively. Both studies were essentially conducted on two (control and disease) samples and did not report biological variations or reproducibility of the method. In this study, a quantitative comparison of relative protein levels in multiple LCM-procured samples by ICAT-LC-MS/MS was achieved using a universal pooled reference experimental design (35) . This experimental design has often been used for microarray analyses of time-series datasets (35) . The use of a pooled control provided a standard for reducing intra-experimental variation and allowed a more reliable comparison of data between experiments. Although the protein levels in LCM-procured samples (1 µg protein from 300 vessels) in this study were comparable or lower from those reported in two other studies (33 , 34) , the separation of nanoLC-MS (MS-only) survey of paired peaks from a nanoLC-MS/MS sequencing step achieved a similar protein coverage and identification of proteins in a pI range form 3 to 10, including numerous transmembrane and unknown proteins.

Analyses of > 50 identified differentially expressed proteins demonstrated a sequential, orchestrated program of microvascular protein regulation over the course of postischemic reperfusion that temporally correlated with the appearance of a major vascular pathology including BBB disruption. Although LCM-procured brain vessels are highly enriched in endothelial cell markers (10) due to the anatomical organization of the neurovascular unit, some identified proteins may also originate from other cellular (e.g., astrocytes, pericytes) or acellular (e.g., basement membrane) components of the brain vascular compartment.

Early reperfusion after global cerebral ischemia was characterized by increased BBB leakiness for 3H-sucrose, loss of vascular proteins involved in cytoskeletal and cellular integrity, loss of energy-dependent sodium pump and amino acid transporters, and up-regulation of the sodium channel and sodium bicarbonate cotransporter, all consistent with described previously losses of endothelial cell viability (36) , disintegration of tight junction structures (36) , and increased sodium and water fluxes (37) in similar models of global cerebral ischemia. The induction of proteins involved in carbohydrate metabolism (aldolase A and malate dehydrogenase), immediate early genes (hsp 70 and hsp 60), transcription factors (IRF-2 and DDX-5), and signal transduction molecules (FGF-11, Gna12, and cyclophilin A), pointed to an early adaptive response of surviving cellular NV unit components to reperfusion injury. IRFs regulate gene programs of host defense, cytokine signaling, and cell growth (38) through interactions with other transcription factors that control inflammatory response such as NF-{kappa}B (39) , whereas DEAD box helicase (DDX) regulates transcription programs of cell proliferation and VEGF expression (40) . Early up-regulation of these transcription factors suggests that signals involved in an EC switch into proliferative, inflammatory and potentially angiogenic phenotype are set in motion as early as 1 h after brain ischemia. Coincident with these changes, early up-regulation of ADAM 23, a brain-specific member of disintegrin protein family (41) that interacts with alphavbeta3 integrin to promote cell attachment and adhesion (42) indicates concomitant alterations in cell-ECM interactions required for subsequent vascular remodeling.

The vascular changes in later reperfusion were characterized by a transient attenuation and subsequent dramatic increase in BBB permeability at 6 and 24 h of reperfusion, respectively. A majority of protein changes seen in early reperfusion reverted to control levels at 6 h, with only a limited number of proteins remaining differentially expressed, including an initial up-regulation of proteins involved in vascular inflammation. In contrast, 24 h reperfusion was characterized with a new wave of protein up-regulation in ischemic brain vessels. The up-regulation of transcription factors HIF-1{alpha}, SOX18 and IRF-2, inflammatory cytokine IL-1ß and chemokine MIP-2, and ECM-degrading metalloprotease MMP9, and down-regulation of the basement membrane proteoglycan agrin coincided with a marked BBB leakiness and vasogenic brain edema observed in the same model (16 , 17) . Vascular inflammation (43) , loss of basal lamina antigens (44) , and the activation of proteolytic enzymes MMP-2 and MMP-9 (44 , 45) have been described in rodent and primate brain after focal ischemia and are considered key molecular effectors responsible for BBB disruption and vascular remodeling (3 , 6 , 8) . Temporal and spatial (46) correlation between MMP-9 expression and Evans blue extravasation in brain microvessels has been demonstrated after a transient focal cerebral ischemia. Use of knockout animals has confirmed the importance of some of these effectors, including MMP-9 (47) and IL-1ß (48) , for vascular and neuronal outcomes after cerebral ischemia.

The expression changes of some of these effectors, such as chemokine MIP-2, in the brain vascular compartment was shown for the first time in this report. In parallel with increased expression of proinflammatory proteins, the up-regulation of annexin 1, a glucocorticoid-inducible protein that inhibits the expression and activity of inflammatory enzymes, including phospholipase A2, iNOS and inducible COX, and has a profound inhibitory effect on neutrophil and monocyte migration during inflammation (49) has been observed. Coincidentally, in global cerebral ischemia models, neutrophil recruitment into the brain is minimal despite up-regulation of recruiting adhesion molecules in brain vessels at 24 h of reperfusion (unpublished observation).

Two proteins involved in anti-oxidative defense, SOD-2 (MnSOD) and metallothionein 4, were highly expressed in brain vessels at 24 h of reperfusion. The up-regulation of SOD-2 and heme-oxygenase has been described in brain endothelial cells in vitro subjected to sublethal oxidative stress (50) . This is the first demonstration of a similar response in intact brain vessels after ischemia/reperfusion.

A hallmark of pathologies characterized by angiogenesis is tissue hypoxia. HIF-1{alpha}, known to transcriptionally regulate genes involved in an endothelial "angiogenic switch" (51) , was up-regulated in vessels 24 h after reperfusion. Since HIF-1{alpha} is rapidly degraded in the presence of oxygen, this suggests that in late reperfusion vessels undergo secondary hypoxic injury potentially caused by perivascular edema. HIF-1-regulated TGFß-1 was also elevated at both mRNA and protein levels with increased number of TGFß-1-positive brain vessels in late reperfusion. TGFß-1 and hypoxia synergistically cooperate to induce VEGF expression (52) , and up-regulation of TGFß-1 in the brain has previously been linked to the angiogenic response after stroke (28) . Angiogenic and proliferative vessels after stroke have been shown to lack BBB phenotype (53) and this likely accounts for the observed dramatic secondary increase in BBB permeability.

ICAT is arguably the most popular proteomic method involving differential isotopic labeling and has a number of distinct advantages that made it attractive for this analysis. However, the technique does have its limitations. ICAT selectively targets cysteine residues and therefore ~3% of mammalian proteins lacking cysteine residues cannot be analyzed. In addition, some cysteines are blocked or are inaccessible to the labeling reagent. Furthermore, not every ICAT labeled peptide is compatible with the nanoLC-MS and –MS/MS process. Though ICAT is better at identifying low abundant molecules than gel-based methods, the identification of very low abundant peptides by MS/MS still remains a challenge. These technical limitations could explain why differential expression of several proteins that might have been expected from the literature, such as ICAM-1, has not been observed in this study. Furthermore, a nonuniform expression of certain proteins in individual vessels is "averaged" in populations of 300 vessels captured from each animal. The technique does not provide information on protein post-translational modifications, an issue less relevant for this study, where temporal profile of protein changes was the main objective.

The neurological outcome of cerebral ischemia is often determined by vascular events. This study describes innovative methods to associate dynamics of vascular protein expression with vascular pathology observed after cerebral ischemia. Identified protein markers can be used to develop molecular imaging or therapeutic approaches to monitor or treat vascular pathology in ischemic brain disease.


   ACKNOWLEDGMENTS
 
These studies are supported by NRC-Genomics and Health Initiative, Heart and Stroke Foundation of Ontario grant #T 5099, and NRC-Helmholtz Collaborative Research Program funding. We thank Philippe Valade, Wen Ding, Luc Tessier and Kenneth Chan for their technical assistance.

Received for publication February 11, 2005. Accepted for publication July 6, 2005.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Dirnagl, U. (1997) Metabolic aspects of neurovascular coupling. Adv. Exp. Med. Biol. 413,155-159[Medline]
  2. Abbott, N. J. (2002) Astrocyte-endothelial interactions and blood-brain barrier permeability. J. Anat. 200,629-638[CrossRef][Medline]
  3. del Zoppo, G. J., Mabuchi, T. (2003) Cerebral microvessel responses to focal ischemia. J. Cereb. Blood Flow Metab. 23,879-894[CrossRef][Medline]
  4. Stanimirovic, D. B. (2001) Inflammatory activation of brain cells by hypoxia: Transcription factors and signaling pathways. Feuerstein, G. eds. Inflammation and Stroke ,101-111 Birkhauser Verlag Basel.
  5. Abumiya, T., Lucero, J., Heo, J. H., Tagaya, M., Koziol, J. A., Copeland, B. R., del Zoppo, G. J. (1999) Activated microvessels express vascular endothelial growth factor and integrin alpha(v)beta3 during focal cerebral ischemia. J. Cereb. Blood Flow Metab. 19,1038-1050[CrossRef][Medline]
  6. Ellison, J. A., Barone, F. C., Feuerstein, G. Z. (1999) Matrix remodeling after stroke. De novo expression of matrix proteins and integrin receptors. Ann. N.Y. Acad. Sci. 890,204-222[Abstract/Free Full Text]
  7. Hayashi, T., Noshita, N., Sugawara, T., Chan, P. H. (2003) Temporal profile of angiogenesis and expression of related genes in the brain after ischemia. J. Cereb. Blood Flow Metab. 23,166-180[CrossRef][Medline]
  8. Mun-Bryce, S., Rosenberg, G. A. (1998) Matrix metalloproteinases in cerebrovascular disease. J. Cereb. Blood Flow Metab. 18,1163-1172[CrossRef][Medline]
  9. Stanimirovic, D., Satoh, K. (2000) Inflammatory mediators of cerebral endothelium: a role in ischemic brain inflammation. Brain Pathol. 10,113-126[Medline]
  10. Mojsilovic-Petrovic, J., Nesic, M., Pen, A., Zhang, W., Stanimirovic, D. (2004) Development of rapid staining protocols for laser-capture microdissection of brain vessels from human and rat coupled to gene expression analyses. J. Neurosci. Methods 133,39-48[CrossRef][Medline]
  11. Zhang, W., Mojsilovic-Petrovic, J., Andrade, M. F., Zhang, H., Ball, M., Stanimirovic, D. B. (2003) The expression and functional characterization of ABCG2 in brain endothelial cells and vessels. FASEB J. 17,2085-2087[Abstract/Free Full Text]
  12. Singh, R., Maganti, R. J., Jabba, S. V., Wang, M., Deng, G., Heath, J. D., Kurn, N., Wangemann, P. (2005) Microarray-based comparison of three amplification methods for nanogram amounts of total RNA. Am. J. Physiol. 288,C1179-C1189
  13. Shekouh, A. R., Thompson, C. C., Prime, W., Campbell, F., Hamlett, J., Herrington, C. S., Lemoine, N. R., Crnogorac-Jurcevic, T., Buechler, M. W., Friess, H., et al (2003) Application of laser capture microdissection combined with two-dimensional electrophoresis for the discovery of differentially regulated proteins in pancreatic ductal adenocarcinoma. Proteomics 3,1988-2001[CrossRef][Medline]
  14. Smith, M. L., Auer, R. N., Siesjo, B. K. (1984) The density and distribution of ischemic brain injury in the rat following 2–10 min of forebrain ischemia. Acta Neuropathol. (Berlin) 64,319-332[CrossRef][Medline]
  15. Preston, E., Allen, M., Haas, N. (1983) A modified method for measurement of radiotracer permeation across the rat blood–brain barrier: the problem of correcting brain uptake for intravascular tracer. J. Neurosci. Methods 9,45-55[CrossRef][Medline]
  16. Preston, E., Webster, J. (2004) A two-hour window for hypothermic modulation of early events that impact delayed opening of the rat blood-brain barrier after ischemia. Acta Neuropathol. (Berlin) 108,406-412[CrossRef][Medline]
  17. Preston, E., Sutherland, G., Finsten, A. (1993) Three openings of the blood-brain barrier produced by forebrain ischemia in the rat. Neurosci. Lett. 149,75-78[CrossRef][Medline]
  18. Kendziorski, C., Irizarry, R. A., Chen, K. S., Haag, J. D., Gould, M. N. (2005) On the utility of pooling biological samples in microarray experiments. Proc. Natl. Acad. Sci. USA 102,4252-4257[Abstract/Free Full Text]
  19. Smolka, M. B., Zhou, H., Purkayastha, S., Aebersold, R. (2001) Optimization of the isotope-coded affinity tag-labeling procedure for quantitative proteome analysis. Anal. Biochem. 297,25-31[CrossRef][Medline]
  20. Hirosawa, M., Hoshida, M., Ishikawa, M., Toya, T. (1993) MASCOT: multiple alignment system for protein sequences based on three-way dynamic programming. Comput. Appl. Biosci. 9,161-167[Abstract/Free Full Text]
  21. Bickel, D. R. (2004) Degrees of differential gene expression: detecting biologically significant expression differences and estimating their magnitudes. Bioinformatics 20,682-688[Abstract/Free Full Text]
  22. Preston, E., Webster, J. (2002) Differential passage of [14C]sucrose and [3H]insulin across rat blood-brain barrier after cerebral ischemia. Acta Neuropathol. (Berlin) 103,237-242[CrossRef][Medline]
  23. Ginsberg, M. D., Busto, R. (1989) Rodent models of cerebral ischemia. Stroke 20,1627-1642[Abstract/Free Full Text]
  24. Dietrich, W. D., Alonso, O., Busto, R. (1993) Moderate hyperglycemia worsens acute blood-brain barrier injury after forebrain ischemia in rats. Stroke 24,111-116[Abstract/Free Full Text]
  25. Gygi, S. P., Rist, B., Gerber, S. A., Turecek, F., Gelb, M. H., Aebersold, R. (1999) Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat. Biotechnol. 17,994-999[CrossRef][Medline]
  26. Smyth, G. K., Speed, T. (2003) Normalization of cDNA microarray data. Methods 31,265-273[CrossRef][Medline]
  27. Zhang, W., Smith, C., Howlett, C., Stanimirovic, D. (2000) Inflammatory activation of human brain endothelial cells by hypoxic astrocytes in vitro is mediated by IL-1beta. J. Cereb. Blood Flow Metab. 20,967-978[Medline]
  28. Krupinski, J., Kumar, P., Kumar, S., Kaluza, J. (1996) Increased expression of TGF-beta 1 in brain tissue after ischemic stroke in humans. Stroke 27,852-857[Abstract/Free Full Text]
  29. Yamashita, K., Gerken, U., Vogel, P., Hossmann, K., Wiessner, C. (1999) Biphasic expression of TGF-beta1 mRNA in the rat brain following permanent occlusion of the middle cerebral artery. Brain Res. 836,139-145[CrossRef][Medline]
  30. Schaller, B., Graf, R. (2004) Cerebral ischemia and reperfusion: the pathophysiologic concept as a basis for clinical therapy. J. Cereb. Blood Flow Metab. 24,351-371[CrossRef][Medline]
  31. Aldred, S., Grant, M. M., Griffiths, H. R. (2004) The use of proteomics for the assessment of clinical samples in research. Clin. Biochem. 37,943-952[CrossRef][Medline]
  32. Shekouh, A. R., Thompson, C. C., Prime, W., Campbell, F., Hamlett, J., Herrington, C. S., Lemoine, N. R., Crnogorac-Jurcevic, T., Buechler, M. W., Friess, H., et al (2003) Application of laser capture microdissection combined with two-dimensional electrophoresis for the discovery of differentially regulated proteins in pancreatic ductal adenocarcinoma. Proteomics 3,1988-2001
  33. Zang, L., Toy, D. P., Hancock, W. S., Sgroi, D. C., Karger, B. L. (2004) Proteomic analysis of ductal carcinoma of the breast using laser capture microdissection, LC-MS, and 16O/18O isotopic labeling. J. Proteome Res. 3,604-612[CrossRef][Medline]
  34. Li, C., Hong, Y., Tan, Y. X., Zhou, H., Ai, J. H., Li, S. J., Zhang, L., Xia, Q. C., Wu, J. R., Wang, H. Y., et al (2004) Accurate qualitative and quantitative proteomic analysis of clinical hepatocellular carcinoma using laser capture microdissection coupled with isotope-coded affinity tag and two-dimensional liquid chromatography mass spectrometry. Mol. Cell. Proteomics 3,399-409[Abstract/Free Full Text]
  35. Novoradovskaya, N., Whitfield, M. L., Basehore, L. S., Novoradovsky, A., Pesich, R., Usary, J., Karaca, M., Wong, W. K., Aprelikova, O., Fero, M., et al (2004) Universal Reference RNA as a standard for microarray experiments. BMC Genomics 5,20-[CrossRef][Medline]
  36. Pomfy, M., Huska, J. (1992) The state of the microcirculatory bed after total ischaemia of the brain. An experimental ultrastructural study. Funct. Dev. Morphol. 2,253-258[Medline]
  37. Mori, K., Miyazaki, M., Iwase, H., Maeda, M. (2002) Temporal profile of changes in brain tissue extracellular space and extracellular ion (Na(+), K(+)) concentrations after cerebral ischemia and the effects of mild cerebral hypothermia. J. Neurotrauma 19,1261-1270[CrossRef][Medline]
  38. Mamane, Y., Heylbroeck, C., Genin, P., Algarte, M., Servant, M. J., LePage, C., DeLuca, C., Kwon, H., Lin, R., Hiscott, J. (1999) Interferon regulatory factors: the next generation. Gene 237,1-14[CrossRef][Medline]
  39. Drew, P. D., Franzoso, G., Carlson, L. M., Biddison, W. E., Siebenlist, U., Ozato, K. (1995) Interferon regulatory factor-2 physically interacts with NF-kappa B in vitro and inhibits NF-kappa B induction of major histocompatibility class I and beta 2-microglobulin gene expression in transfected human neuroblastoma cells. J. Neuroimmunol. 63,157-162[CrossRef][Medline]
  40. Kahlina, K., Goren, I., Pfeilschifter, J., Frank, S. (2004) p68 DEAD box RNA helicase expression in keratinocytes. Regulation, nucleolar localization, and functional connection to proliferation and vascular endothelial growth factor gene expression. J. Biol. Chem. 279,44872-44882[Abstract/Free Full Text]
  41. Goldsmith, A. P., Gossage, S. J., ffrench-Constant, C. (2004) ADAM23 is a cell-surface glycoprotein expressed by central nervous system neurons. J. Neurosci. Res. 78,647-658[CrossRef][Medline]
  42. Cal, S., Freije, J. M., Lopez, J. M., Takada, Y., Lopez-Otin, C. (2000) ADAM 23/MDC3, a human disintegrin that promotes cell adhesion via interaction with the alphavbeta3 integrin through an RGD-independent mechanism. Mol. Biol. Cell 11,1457-1469[Abstract/Free Full Text]
  43. Ishikawa, M., Zhang, J. H., Nanda, A., Granger, D. N. (2004) Inflammatory responses to ischemia and reperfusion in the cerebral microcirculation. Front. Biosci. 9,1339-1347[Medline]
  44. Fukuda, S., Fini, C. A., Mabuchi, T., Koziol, J. A., Eggleston, L. L., Jr, del Zoppo, G. J. (2004) Focal cerebral ischemia induces active proteases that degrade microvascular matrix. Stroke 35,998-1004[Abstract/Free Full Text]
  45. Romanic, A. M., White, R. F., Arleth, A. J., Ohlstein, E. H., Barone, F. C. (1998) Matrix metalloproteinase expression increases after cerebral focal ischemia in rats: inhibition of matrix metalloproteinase-9 reduces infarct size. Stroke 29,1020-1030[Abstract/Free Full Text]
  46. Maier, C. M., Hsieh, L., Yu, F., Bracci, P., Chan, P. H. (2004) Matrix metalloproteinase-9 and myeloperoxidase expression: quantitative analysis by antigen immunohistochemistry in a model of transient focal cerebral ischemia. Stroke 35,1169-1174[Abstract/Free Full Text]
  47. Asahi, M., Asahi, K., Jung, J. C., del Zoppo, G. J., Fini, M. E., Lo, E. H. (2000) Role for matrix metalloproteinase 9 after focal cerebral ischemia: effects of gene knockout and enzyme inhibition with BB-94. J. Cereb. Blood Flow Metab. 20,1681-1689[CrossRef][Medline]
  48. Ohtaki, H., Funahashi, H., Dohi, K., Oguro, T., Horai, R., Asano, M., Iwakura, Y., Yin, L., Matsunaga, M., Goto, N., et al (2003) Suppression of oxidative neuronal damage after transient middle cerebral artery occlusion in mice lacking interleukin-1. Neurosci. Res. 45,313-324[CrossRef][Medline]
  49. Parente, L., Solito, E. (2004) Annexin 1: more than an anti-phospholipase protein. Inflamm. Res. 53,125-132[CrossRef][Medline]
  50. Methy, D., Bertrand, N., Prigent-Tessier, A., Stanimirovic, D., Beley, A., Marie, C. (2004) Differential MnSOD and HO-1 expression in cerebral endothelial cells in response to sublethal oxidative stress. Brain Res. 1003,151-158[CrossRef][Medline]
  51. Carmeliet, P. (2000) Mechanisms of angiogenesis and arteriogenesis. Nat. Med. 6,389-395[CrossRef][Medline]
  52. Sanchez-Elsner, T., Botella, L. M., Velasco, B., Corbi, A., Attisano, L., Bernabeu, C. (2001) Synergistic cooperation between hypoxia and transforming growth factor-beta pathways on human vascular endothelial growth factor gene expression. J. Biol. Chem. 276,38527-38535[Abstract/Free Full Text]
  53. Zhang, Z. G., Zhang, L., Tsang, W., Soltanian-Zadeh, H., Morris, D., Zhang, R., Goussev, A., Powers, C., Yeich, T., Chopp, M. (2002) Correlation of VEGF and angiopoietin expression with disruption of blood-brain barrier and angiogenesis after focal cerebral ischemia. J. Cereb. Blood Flow Metab. 22,379-392[Medline]



This article has been cited by other articles:


Home page
 Genome ResHome page
D. B. McClatchy, L. Liao, S. K. Park, J. D. Venable, and J. R. Yates
Quantification of the synaptosomal proteome of the rat cerebellum during post-natal development
Genome Res., September 1, 2007; 17(9): 1378 - 1388.
[Abstract]