Clinical Chemistry
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Clinical Chemistry 52: 1501-1509, 2006. First published June 8, 2006; 10.1373/clinchem.2006.069294
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow 069294.Supplemental data
Right arrow All Versions of this Article:
clinchem.2006.069294v1
52/8/1501    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
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 Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de Bont, J. M.
Right arrow Articles by Pieters, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by de Bont, J. M.
Right arrow Articles by Pieters, R.
Related Collections
Right arrow Proteomics and Protein Markers
(Clinical Chemistry. 2006;52:1501-1509.)
© 2006 American Association for Clinical Chemistry, Inc.

Proteomics and Protein Markers

Identification of Apolipoprotein A-II in Cerebrospinal Fluid of Pediatric Brain Tumor Patients by Protein Expression Profiling

Judith M. de Bont1, Monique L. den Boer1,a, Roel E. Reddingius1, Jaap Jansen2, Monique Passier1, Ron H.N. van Schaik3, Johan M. Kros4, Peter A.E. Sillevis Smitt5, Theo H. Luider5 and Rob Pieters1

1 Erasmus MC—Sophia Children’s Hospital—University Medical Center Rotterdam, Department of Pediatric Oncology and Hematology, Rotterdam, The Netherlands.
2 Ciphergen Biosystems Inc., Fremont, CA.
Erasmus MC—University Medical Center Rotterdam, Departments of3 Clinical Chemistry,4 Pathology, and5 Neuro-oncology, Rotterdam, The Netherlands.

aAddress correspondence to this author at: Erasmus MC—Sophia Children’s Hospital, Department of Pediatric Oncology and Hematology, Dr. Molewaterplein 60, 3015 GJ Rotterdam, The Netherlands. Fax 31-104089433; e-mail m.l.denboer{at}erasmusmc.nl.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: Our aim was to detect differences in protein expression profiles of cerebrospinal fluid (CSF) from pediatric patients with and without brain tumors.

Methods: We used surface-enhanced laser desorption/ionization time-of-flight (SELDI-TOF) mass spectrometry and Q10 ProteinChip arrays to compare protein expression profiles of CSF from 32 pediatric brain tumor patients and 70 pediatric control patients. A protein with high discriminatory power was isolated and identified by subsequent anion-exchange and reversed-phase fractionation, gel electrophoresis, and mass spectrometry. The identity of the protein was confirmed by Western blotting and immunohistochemistry.

Results: Of the 247 detected protein peak clusters, 123 were differentially expressed between brain tumor and control patients with a false discovery rate of 1%. Double-loop classification analysis gave a mean prediction accuracy of 88% in discriminating brain tumor patients from control patients. From the 123 clusters, a highly overexpressed protein peak cluster in CSF from brain tumor patients was selected for further analysis and identified as apolipoprotein A-II. Apolipoprotein A-II expression in CSF was correlated with the CSF albumin concentration, suggesting that the overexpression of apolipoprotein A-II is related to a disrupted blood–brain barrier.

Conclusions: SELDI-TOF mass spectrometry can be successfully used to find differentially expressed proteins in CSF of pediatric brain tumor and control patients. Apolipoprotein A-II is highly overexpressed in CSF of pediatric brain tumor patients, which most likely is related to a disrupted blood–brain barrier. Ongoing studies are aimed at finding subtype specific proteins in larger groups of pediatric brain tumor patients.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
In contrast to hematologic malignancies in childhood, knowledge of the basic biological characteristics of pediatric brain tumors is limited. Studies based on molecular analyses and gene expression profiling are now evolving and may provide possible clues about pathogenetic and prognostic factors in, for example, primitive neuroectodermal tumors (1). Despite the power of genomic and transcriptomic technologies, a major shortcoming of these approaches is that gene expression does not always correlate with protein expression. Because proteins actually change the functional state of the cell, proteins are the most relevant markers of function. Early proteomic approaches used 2-dimensional gel electrophoresis followed by identification of differentially expressed proteins by mass spectrometry. Novel proteomic approaches aiming at identifying novel biomarkers for cancer diagnosis and staging are rapidly developing. Surface-enhanced laser desorption/ionization time-of-flight (SELDI-TOF) 1 mass spectrometry is based on 2 techniques: chromatography and mass spectrometry (2). Proteins from complex protein mixtures are retained on specific chromatographic surfaces and subsequently analyzed by a linear TOF mass spectrometer to detect differences in protein expression profiles between groups of patients. The SELDI approach has recently been used successfully to identify biomarkers in biological fluids in various malignancies (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15).

It has been shown that disease-related proteins can be detected in cerebrospinal fluid (CSF) of patients with a primary brain tumor by 2-dimensional gel electrophoresis (16) and in Alzheimer patients by SELDI-TOF mass spectrometry (17). The objective of this study was to determine whether protein expression profiles obtained by SELDI-TOF mass spectrometry of CSF can be used to distinguish pediatric brain tumor patients from control patients.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
study population and samples
A total of 32 lumbar CSF samples from newly diagnosed and untreated pediatric brain tumor patients (17 males and 15 females; median age, 7.2 years) and 70 pediatric control patients (39 males and 31 females; median age, 8.5 years) were analyzed in this study. All CSF samples were collected at the Erasmus MC—University Medical Center—Sophia Children’s Hospital in Rotterdam, The Netherlands under uniform and standardized conditions. The brain tumor group consisted of 16 patients with medulloblastoma and 16 brain tumor patients with other histologic subtypes (high-grade glioma, n = 7; atypical rhabdoid tumor, n = 2; pilocytic astrocytoma, n = 2; plexus carcinoma, n = 2; anaplastic ependymoma, n = 2; germ cell tumor, n = 1), which will be referred to as "other tumors".

The 70 lumbar control CSF samples were obtained from pediatric patients 1 year after the end of treatment for acute lymphoblastic leukemia (n = 47), pediatric patients without a malignancy (infection, n = 6; hematologic disease, n = 4; autoimmune disease, n = 2, idiopathic intracranial hypertension, n = 1) and pediatric patients with a malignancy outside the central nervous system (Hodgkin disease, n = 6; neuroblastoma, n = 4). Each patient or the patient’s parents gave informed consent before enrollment.

Cytologic analysis of the CSF samples showed malignant cells in 10 medulloblastoma CSF samples. CSF from patients with other tumors and from control patients did not show any cytologic abnormalities.

sample preparation
All CSF samples were immediately centrifuged at 250g for 5 min to remove cellular debris, and cytospins were made to evaluate cytology. Supernatants were aliquoted and stored at –80 °C until analysis. Only CSF samples without macroscopic blood contamination were included in the study.

Before SELDI protein profiling, the protein concentration of the CSF samples was determined by the bicinchoninic acid protein assay (Pierce).

seldi protein profiling
Q10 ProteinChip® arrays (strong anion-exchange chromatographic surface; Ciphergen Biosystems Inc.) were equilibrated with 50 mmol/L Tris (pH 8.5) containing 1 mL/L Triton (binding buffer). Crude CSF samples, each containing 5 µg of protein, were loaded on the arrays in duplicate together with binding buffer and incubated for 1 h at room temperature. Arrays were washed sequentially with 50 mmol/L Tris (pH 8.5) containing 1 mL/L Triton, 50 mmol/L Tris (pH 8.5), and deionized water. After the arrays were air-dried at room temperature, energy-absorbing matrix sinapinic acid in an aqueous solution containing 500 mL/L acetonitrile and 5 mL/L trifluoroacetic acid was added to each spot. Mass analysis of the bound proteins was performed with a PBS IIc instrument (Ciphergen) in positive operating mode. Mass spectra were collected by the accumulation of 65 laser shots at a laser intensity of 175 and a detector sensitivity of 8. The highest mass to acquire was set at m/z 200 000 with an optimization range between m/z 2000 and 30 000. The instrument was calibrated by use of Protein Calibration Standard I (Bruker Daltonics). Peak detection and clustering were performed with the ProteinChip Biomarker WizardTM module embedded in the ProteinChip software, Ver. 3.1 (Ciphergen).

All spectra were compiled, aligned to a common calibrant, and normalized to the total ion current. The Biomarker Wizard automatically detected peaks with a signal-to-noise ratio >5. Subsequently, peaks were clustered by use of a 0.3% mass window and a second-pass peak selection with a signal-to-noise ratio >2. The part of the spectrum with m/z values <2000 was not used for analysis because the signal for the energy-absorbing matrix generally interferes with peak detection in this area.

We used the nonparametric Mann–Whitney U-test implemented in the Biomarker Wizard software to statistically compare differences in intensity data for the various protein peak clusters after normalization between patients with brain tumor and control patients. We corrected for multiple testing errors by applying the false discovery rate (FDR) described by Benjamini and Hochberg (18). We used a FDR of 1%, which means that 99% of the selected protein peaks are expected to be true positive and not found by chance. We used the 123 protein peak clusters differentially expressed between brain tumor and control patients to perform a double-loop classification analysis with Biomarker PatternsTM software (Ciphergen). Biomarker Patterns software is a tool for tree-structured data analysis, in which classification and regression trees (CART) methodology (19) is implemented. The data were randomly split into a training set (two thirds of the data; ncontrol = 47; nbrain tumor = 21) and a test set (one third of the data; ncontrol = 23; nbrain tumor = 11). The training set was used to generate the optimal classification tree in the Biomarker Patterns Software by use of a 10-fold internal cross-validation procedure. The optimal classification tree was tested for accuracy by the test set. The prediction accuracy was calculated as the ratio of correctly classified brain tumor samples to the total number of brain tumor samples. Specificity was calculated as the ratio of correctly classified control samples to the total number of control samples. This procedure was repeated 3 times with randomly chosen training (two thirds of samples) and test (one third of samples) sets. Mean values of prediction accuracy and specificity of the 3 repeated classification analyses are reported.

purification and identification of differentially expressed proteins
CSF samples were fractionated by use of Q ceramic HyperD® F columns (Ciphergen), with 50 mmol/L Tris pH 9.0, pH 7.0, pH 5.0, pH 4.0, pH 3.0, and organic solvent washes. Subfractionation was done by use of reversed-phase chromatography (RPC) beads (30 nm; Polymer Laboratories Ltd.) with increasing concentrations of acetonitrile in 1 mL/L trifluoroacetic acid. Fractions containing the protein peaks of interest [monitored on NP20 (normal-phase chromatographic surface) ProteinChip arrays] were subjected to nonreducing sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) on 16% Tris-glycine gels. The gel was stained with Coomassie blue (Invitrogen), and gel bands were cut. After destaining in 500 mL/L acetonitrile–500 mL/L 50 mmol/L ammonium bicarbonate and dehydration in acetonitrile, bands were dried in a SpeedVac. Proteins were passively eluted out of the gel bands by rehydration in 450 mL/L formic acid–300 mL/L acetonitrile—100 mL/L isopropanol and sonication for 3.5 h at room temperature. The extracts containing the markers (tested on NP20 arrays) were dried in a SpeedVac, resuspended in 20 µg/L modified trypsin (Roche Applied Science) dissolved in 100 mmol/L ammonium bicarbonate–100 mL/L acetonitrile, and incubated for 4 h at 37 °C. Peptide mass spectra obtained in the ProteinChip System Series 4000 instrument (Ciphergen) were used for peptide mass fingerprinting (ProFound database). Tandem mass spectrometry (MS/MS) of selected peptides was done with a Micromass Q-TOFII equipped with a PCI 1000 ProteinChip Tandem MS Interface (Ciphergen).

western blots
CSF samples (9.7 µg of protein) were subjected to nonreducing SDS-PAGE on 15% Tris-glycine gels. After proteins were transferred to a nitrocellulose membrane at 140 V for 90 min at 4 °C in a wet transfer system (Bio-Rad) and blocking with block buffer [50 g/L nonfat milk in 0.01 mmol/L phosphate-buffered saline (pH 7.4) containing 2 mL/L Tween 20] for 30 min, the membrane was incubated overnight at 4 °C with mouse monoclonal antibodies against apolipoprotein A-II (APO-A-II; Biodesign International) diluted (1:400) in blocking buffer. The membrane was washed with 2 mL/L Tween 20 in 0.01 mmol/L phosphate-buffered saline (pH 7.4) and incubated for 1 h at room temperature with antimouse IgG–horseradish peroxidase conjugate (DakoCytomation) diluted (1:500) in blocking buffer containing 20 mL/L pooled human serum. Detection solution (SuperSignal West Femto Maximum Sensitivity Substrate; Pierce) was prepared according to the manufacturer’s instructions. Chemiluminescence was measured with the Syngene chemigenius.

csf/blood albumin quotient
The albumin concentration was analyzed in paired CSF and EDTA-plasma samples from 22 brain tumor patients by an immunochemical albumin assay on the Immage 800 system (Beckman Coulter). The CSF/blood albumin quotient (Qalb) was calculated and used as an estimation of the integrity of the blood–brain barrier. In 20 CSF samples from control patients, the albumin concentration was determined by the same assay. As no paired plasma samples for these control patients were available, we determined the plasma albumin concentration reference range with this assay in a separate set of 27 healthy pediatric control patients (median, 32.3 g/L; range, 11.3–41.4 g/L). The Qalb in the control patients was then calculated by use of the median plasma albumin concentration from the separate control set.

immunohistochemistry
We sectioned formalin-fixed, paraffin-embedded tissues from the medulloblastoma, ependymoma, high-grade glioma, control cerebellum, and prostate carcinoma into 4-µm-thick sections and stained them with hematoxylin and eosin for histologic examination. Prostate carcinoma tissue sections showed specific staining after incubation with the primary antibodies and were used as positive control [also described by Malik et al. (6)]. Tissue sections incubated with only the secondary antibody were used as negative controls. Sections were deparaffinized and rehydrated through a graded xylene–ethanol series. Antigen retrieval, achieved by boiling the slides for 5 min in 0.01 mol/L citric acid (pH 6.0), was followed by a 30-min incubation in 30 mL/L hydrogen peroxide in 0.01 mmol/L phosphate-buffered saline (pH 7.4) to inhibit endogenous peroxidases. We then added 30 mL/L goat serum or 30 mL/L horse serum (Vector Laboratories) as a protein block for the monoclonal and polyclonal antibodies, respectively. Tissue sections were incubated with either a monoclonal (1:2000; Biodesign International) or a polyclonal (1:2000; Rockland Immunochemicals) anti-APO A-II antibody overnight at 4 °C. The sections were then incubated for 1 h at room temperature with 1:1000 dilutions of biotinylated goat antimouse IgG for the monoclonal antibody and biotinylated horse antirabbit IgG (Vector Laboratories) for the polyclonal antibody. Incubation with the secondary antibody was followed by a 1-h incubation with avidin-biotin peroxidase complex (1:400 dilution; Vectastain ABC Kit; Vector Laboratories). Staining was performed with a solution containing 0.5 g/L 3,3'-diaminobenzidine and 0.3 mL/L hydrogen peroxide in 30 mmol/L imidazole/1 mmol/L EDTA (pH 7.0). Tissue section were slightly counterstained with hematoxylin, dehydrated, and mounted.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
reproducibility
The CV for the protein peak intensities obtained by SELDI-TOF mass spectrometry was determined on 12 protein peak clusters present in both a brain tumor and a control sample, which were measured in 8 independent experiments. The CV of the peak intensities for the 12 peak clusters ranged from 6% to 24% and did not differ statistically between the brain tumor and control CSF sample.

seldi mass spectrometry
Mass spectrometric analysis of the proteins from the CSF of brain tumor and control patients that were bound to the affinity tags of the Q10 ProteinChip arrays (Ciphergen) revealed 247 protein peak clusters between m/z 2000 and 200 000. When we used an FDR of 1%, 123 protein peak clusters were significantly different between brain tumor and control patients. These protein peak clusters were used in a double-loop classification analysis. The most accurate classification models had a mean prediction accuracy of 88% (range, 82%–100%) and a mean specificity of 88% (range, 78%–96%) in discriminating CSF of brain tumor patients from CSF of control patients.

None of the 247 detected protein peak clusters was found to be differentially expressed between the different histologic brain tumor subtypes or between medulloblastoma patients with and without CSF cytologic abnormalities.

identification of protein markers
From the 123 differentially expressed protein peak clusters, we selected a protein peak cluster at Mr 17 000, consisting of an m/z 17 248 and an m/z 17 369 protein peak, that was highly abundant in brain tumor patients but virtually absent in control patients (Fig. 1 ). To identify this protein cluster, we subjected CSF from a brain tumor patient sequentially to 3 fractionation methods. On anion-exchange Q ceramic HyperD F columns, the m/z 17 000 protein cluster eluted in the pH 5.0 and pH 4.0 fractions (Fig. 2A ). When the of the pooled pH 4.0 and pH 5.0 fractions were further fractionated by use of RPC beads, the of the m/z 17 000 peaks eluted in the 500 mL/L acetonitrile wash fraction (Fig. 2B ), which was subsequently subjected to nonreducing SDS-PAGE on 16% Tris-glycine gels (Fig. 2C ). Nine gel bands were cut from the gel (Fig. 2C ) and processed for passive elution, which revealed that the m/z 17 000 protein cluster was present in gel band 7 (Fig. 2D ).


Figure 1
View larger version (19K):
[in this window]
[in a new window]
  		Figure 1. Differential expression of the Mr 17 000 protein peak cluster in CSF of brain tumor and control patients.

(A), representative SELDI expression profiles of control and brain tumor CSF samples, showing significantly higher expression of a protein cluster at m/z 17 000 in CSF of brain tumor patients compared with control patients. (B), scatter plots of relative intensities of the m/z 17 248 and m/z 17 369 protein peaks (see panel A) as detected by SELDI-TOF analysis of 70 control and 32 brain tumor CSF samples. The lines represent the median value within each subgroup.


Figure 2
View larger version (31K):
[in this window]
[in a new window]
  		Figure 2. Purification and identification of the m/z 17 000 protein peak cluster.

(A), CSF was subjected to anion-exchange fractionation on Q Hyper D F columns, and eluted protein fractions were tested on NP20 ProteinChips for the presence of the Mr 17 000 protein cluster. The protein peak cluster at m/z 17 000 eluted in the pH 4.0 and pH 5.0 fractions. (B), the pH 4.0 and 5.0 fractions were pooled and subfractionated by use of RPC beads. (C), the 50% acetonitrile RPC (ACN 50% in B) subfraction, which contained the m/z 17 000 protein cluster, was run on a 16% Tris-glycine SDS-PAGE gel, which was stained with colloidal Coomassie blue. (D), check of the passive elution of gel bands 1–9 on NP20 arrays showed that the m/z 17 000 protein cluster was present in gel band 7. To illustrate the differences between the analyzed gel bands, the passive elution from band 1 is also displayed. (E), peptide mass fingerprinting of the trypsin-digested protein extract from band 7 identified the m/z 17 000 protein cluster as APO A-II. MS/MS analysis of the peptide peaks indicated with * (m/z 1156, 1199, and 2513) confirmed the identification of the m/z 17 000 cluster as APO A-II [also see supplementary information I-III (in the online Data Supplement) for MS/MS data of the 3 peptide peaks].

Subsequent peptide mass fingerprinting (ProFound) after tryptic digestion of the passive elution from band 7 gave a significant hit (Z = 2.39) for APO A-II (sequence coverage, 88%). MS/MS analysis of 3 peptide peaks in the peptide mass spectrum (Fig. 2EUp , peaks indicated by *) confirmed the identification of APO A-II [also see supplementary information I-III (in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol52/issue8/) for MS/MS data on the 3 peptide peaks].

Western blot analysis revealed staining of 2 protein bands at approximately Mr 17 000 and Mr 34 000 by anti-APO A-II antibody. These proteins were abundantly present in CSF of brain tumor patients compared with control cases (Fig. 3 ). This confirmed the identification and differential expression of APO A-II in CSF of control and brain tumor patients.


Figure 3
View larger version (84K):
[in this window]
[in a new window]
  		Figure 3. Differential expression of APO A-II in CSF from control and brain tumor patients.

Western blot analysis of CSF from control and brain tumor patients showed differential staining by anti-APO A-II antibody of a protein band at approximately the same mass as the band from which APO A-II was identified by MS/MS (Fig. 2Up ), confirming the identity of APO A-II.

disrupted blood–brain barrier
The observed increase in APO A-II in CSF of brain tumor patients might be the result of protein leakage from the blood to the CSF attributable to a disrupted blood–brain barrier. To test this hypothesis, we measured the albumin concentration in CSF from brain tumor patients and control patients as well as the blood albumin concentration in brain tumor patients to calculate the CSF/blood albumin ratio, which serves as a good estimate of the integrity of the blood–brain barrier (20).

The median albumin concentration in CSF of brain tumor patients was significantly higher than that in CSF of control patients (0.19 and 0.10 g/L, respectively; P = 0.0001). The median Qalb in brain tumor and control patients were 8.61 x 10–3 and 3.04 x 10–3 (P <0.0001), respectively (Fig. 4 ), which suggests a disrupted blood–brain barrier in brain tumor patients, based on the reference Qalb values in children (upper limit for patients under 15 years, 5.0 x 10–3) (21)(22)(23). The albumin concentration in CSF was significantly (P <0.0001) correlated with the peak intensities of the 2 protein peaks in the m/z 17 000 APO A-II protein cluster in our SELDI experiments (Fig. 5 ). These data indicate that the increased APO A-II concentration in CSF of brain tumor patients might be explained by a disrupted blood–brain barrier.


Figure 4
View larger version (15K):
[in this window]
[in a new window]
  		Figure 4. CSF/Blood albumin ratio in control and brain tumor patients.

The CSF/blood albumin ratio was calculated in control and brain tumor patients as representation of the integrity of the blood brain barrier. The CSF/blood albumin ratio was significantly higher in brain tumor patients than in control patients, suggesting a disruption of the blood–brain barrier in the brain tumor patients. The lines represent the median value within each subgroup.


Figure 5
View larger version (13K):
[in this window]
[in a new window]
  		Figure 5. Correlation between CSF albumin concentration and relative intensity of m/z 17 000 protein peak cluster in protein expression profiles.

The CSF albumin concentration was significantly correlated with the peak intensity of both proteins in the Mr 17 000 protein cluster observed in the protein expression profiles. {blacksquare}, m/z 17 248 peak (Spearman correlation coefficient = 0.75; P <0.0001); {circ}, m/z 17 369 peak (Spearman correlation coefficient = 0.91; P <0.0001).

Immunohistochemistry did not show specific cellular APO A-II staining in pediatric brain tumor (see supplementary information IV in the online Data Supplement) or healthy cerebellum tissue sections.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
SELDI ProteinChip technology has been used successfully to compare protein expression profiles in body fluids in various malignancies (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15). We analyzed SELDI protein expression profiles of CSF from pediatric brain tumor and control patients and observed 123 protein peak clusters that were differentially expressed with an FDR of 1%. Double-loop classification analysis gave a mean prediction accuracy of 88% and a mean specificity of 88% in discriminating brain tumor patients from control patients.

Probably because of the relatively small number of each brain tumor subtype in our study, we were not able to differentiate between different brain tumor subtypes based on the protein expression profiles of CSF. However, our analysis did identify a protein marker in CSF, which is common to all brain tumors. A protein cluster at m/z 17 000 that was highly discriminative between tumor and control patients was identified as APO A-II by purification and subsequent peptide mass fingerprinting and MS/MS. Western blotting confirmed the identification and differential expression of APO A-II in CSF from brain tumor and control patients.

In most mammals, APO A-II is present in a monomeric form, whereas in humans, APO A-II also exists as a dimer of 2 identical chains, cross-linked by a single disulfide bond at cysteine-6 at the amino terminus of the protein (24)(25). Because the monomeric form of APO A-II is Mr 8700, the Mr 17 000 form identified in our study most likely represents the dimeric form of APO A-II. The masses of the 2 protein peaks (m/z 17 248 and 17 369) in the m/z 17 000 protein peak cluster in our SELDI spectra were in concordance with previously described masses of isoforms of APO A-II (25)(26). Multiple isoforms of APO A-II have been described in humans (25)(27)(28)(29) and are the result of posttranslational modifications, such as oxidization or truncation of the C-terminal glutamine residue (25)(26)(27)(28).

APO A-II is the second most abundant human HDL apolipoprotein and is synthesized predominantly in the liver (30). Despite the high abundance of APO A-II, little is known about its biological functions. APO A-II is thought to influence the metabolism of HDL and glucose (31)(32)(33). Genetic variations in APO A-II appear to be involved in senile amyloidogenesis in mice and humans (30)(34)(35), and APO A-II has recently been linked to malignancies. In contrast to the increased APO A-II concentrations in CSF from brain tumor patients in our study, a study in mice indicated a decrease in serum APO A-II in mice that had received injections of a B-cell lymphoma (36). A study in prostatic disease indicated that a monomeric isoform of APO A-II (Mr 8900) is specifically overexpressed in serum of patients with prostate cancer (6).

Our data suggest that the overexpression of APO A-II in CSF from brain tumor patients might have resulted from protein leakage from the blood through a disrupted blood–brain barrier. APO E and APO A-I are the major apolipoproteins in CSF, being present in HDL. Whereas evidence exists that, for example, APO E can be produced by astrocytes and microglia in the central nervous system (37), no data are available that indicate APO A-II production in the central nervous system. In patients with neuroborreliosis, CSF APO A-II concentrations were correlated with albumin concentrations regardless of cytology (38). Because an increased Qalb is indicative of a loss of integrity of the blood–brain barrier, increased APO A-II together with increased CSF albumin might reflect a disrupted blood–brain barrier. The significant difference in CSF albumin concentration between brain tumor patients and control patients, the increased Qalb in brain tumor patients, and the significant correlation between the CSF albumin concentration and the intensity of the APO A-II peak in our SELDI experiments may indicate that a disrupted blood–brain barrier indeed might be the source of the increased APO A-II in CSF of brain tumor patients. Because none of the CSF samples in our study had macroscopic blood contamination, we excluded a traumatic lumbar puncture as the cause of the increased APO A-II in CSF. Additional evidence that the relative abundance of APO A-II in CSF may be attributable to a disrupted blood–brain barrier is our finding of the m/z 17 000 APO A-II cluster in the protein expression profiles of 2 control patients with infected ventriculo-peritoneal drains, in whom ventricular instead of lumbar CSF was analyzed (data not shown).

Malik et al. (6) suggested that APO A-II is a potentially specific marker for prostatic disease. Several other apolipoproteins, such as APO D, E, and J, have also been linked to proliferation and cell growth in various malignancies (39)(40)(41)(42)(43), which may suggest that apolipoproteins, including APO A-II, are more general cancer markers, possibly being involved in the regulation of proliferation and growth of tumor cells. However, we observed no specific cellular APO A-II staining in our brain tumor tissue sections. This observation further strengthens the assumption that the increased APO A-II in the CSF of brain tumor patients might be related to a secondary phenomenon, such as a disrupted blood–brain barrier.

In conclusion, our study indicates that SELDI-TOF mass spectrometry can be successfully used to detect proteins that are differentially expressed in CSF of pediatric patients with and without brain tumors. Of the 123 differentially expressed protein peak clusters, we identified a highly overexpressed m/z 17 000 peak cluster in CSF from brain tumor patients as APO A-II. The overexpression of APO A-II in CSF from brain tumor patients was correlated with an increased albumin concentration in CSF, which suggests a relationship with a disrupted blood–brain barrier. Ongoing studies are aimed at detecting subtype-specific proteins in larger groups of pediatric brain tumor patients.


   Acknowledgments
 
We acknowledge Vladimir Podust, PhD, and Nathan Harris (Ciphergen Biosystems, Inc.) for excellent suggestions regarding the protein purification procedure and providing us with the MS/MS data.


   Footnotes
 
1 Nonstandard abbreviations: SELDI-TOF, surface-enhanced laser desorption/ionization time of flight; CSF, cerebrospinal fluid; FDR, false discovery rate; RPC, reversed-phase chromatography; SDS-PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; MS/MS, tandem mass spectrometry; and APO, apolipoprotein.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Pomeroy SL, Tamayo P, Gaasenbeek M, Sturla LM, Angelo M, McLaughlin ME, et al. Prediction of central nervous system embryonal tumour outcome based on gene expression. Nature 2002;415:436-442.[CrossRef][Medline] [Order article via Infotrieve]
  2. Tang N, Tornatore P, Weinberger SR. Current developments in SELDI affinity technology. Mass Spectrom Rev 2004;23:34-44.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  3. Wright GL, Jr, Cazares LH, Leung S-M, Nasim S, Adam B-L, Yip T-T, et al. Proteinchip surface enhanced laser desorption/ionization (SELDI) mass spectrometry: a novel protein biochip technology for detection of prostate cancer biomarkers in complex protein mixtures. Prostate Cancer Prostatic Dis 2000;2:264-276.
  4. Lehrer S, Roboz J, Ding H, Zhao S, Diamond EJ, Holland JF, et al. Putative protein markers in the sera of men with prostatic neoplasms. BJU Int 2003;92:223-225.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  5. Gretzer MB, Chan DW, van Rootselaar CL, Rosenzweig JM, Dalrymple S, Mangold LA, et al. Proteomic analysis of dunning prostate cancer cell lines with variable metastatic potential using SELDI-TOF. Prostate 2004;60:325-331.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  6. Malik G, Ward MD, Gupta SK, Trosset MW, Grizzle WE, Adam BL, et al. Serum levels of an isoform of apolipoprotein A-II as a potential marker for prostate cancer. Clin Cancer Res 2005;11:1073-1085.[Abstract/Free Full Text]
  7. Wong YF, Cheung TH, Lo KW, Wang VW, Chan CS, Ng TB, et al. Protein profiling of cervical cancer by protein-biochips: proteomic scoring to discriminate cervical cancer from normal cervix. Cancer Lett 2004;211:227-234.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  8. Petricoin EF, Ardekani AM, Hitt BA, Levine PJ, Fusaro VA, Steinberg SM. Use of proteomic patterns in serum to identify ovarian cancer. Lancet 2002;359:572-577.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  9. Paweletz CP, Trock B, Pennanen M, Tsangaris T, Magnant C, Liotta LA, et al. Proteomic patterns of nipple aspirate fluids obtained by SELDI-TOF: potential for new biomarkers to aid in the diagnosis of breast cancer. Dis Markers 2001;17:301-307.[Web of Science][Medline] [Order article via Infotrieve]
  10. Won Y, Song HJ, Kang TW, Kim JJ, Han BD, Lee SW. Pattern analysis of serum proteome distinguishes renal cell carcinoma from other urologic diseases and healthy persons. Proteomics 2003;3:2310-2316.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  11. Wadsworth JT, Somers KD, Stack BC, Jr, Cazares L, Malik G, Adam BL, et al. Identification of patients with head and neck cancer using serum protein profiles. Arch Otolaryngol Head Neck Surg 2004;130:98-104.[Abstract/Free Full Text]
  12. Zhukov TA, Johanson RA, Cantor AB, Clark RA, Tockman MS. Discovery of distinct protein profiles specific for lung tumors and pre-malignant lung lesions by SELDI mass spectrometry. Lung Cancer 2003;40:267-279.[Web of Science][Medline] [Order article via Infotrieve]
  13. Wilson LL, Tran L, Morton DL, Hoon DS. Detection of differentially expressed proteins in early-stage melanoma patients using SELDI-TOF mass spectrometry. Ann N Y Acad Sci 2004;1022:317-322.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  14. Koopmann J, Zhang Z, White N, Rosenzweig J, Fedarko N, Jagannath S, et al. Serum diagnosis of pancreatic adenocarcinoma using surface-enhanced laser desorption and ionization mass spectrometry. Clin Cancer Res 2004;10:860-868.[Abstract/Free Full Text]
  15. Rosty C, Christa L, Kuzdzal S, Baldwin WM, Zahurak ML, Carnot F, et al. Identification of hepatocarcinoma-intestine-pancreas/pancreatitis-associated protein I as a biomarker for pancreatic ductal adenocarcinoma by protein biochip technology. Cancer Res 2002;62:1868-1875.[Abstract/Free Full Text]
  16. Zheng P, Luider TM, Pieters R, Avezaat CJJ, Sillevis Smitt PAE, Kros JM. Identification of tumor-related proteins by proteomic analysis of cerebrospinal fluid from patients with primary brain tumors. J Neuropathol Exp Neurol 2003;62:855-862.[Web of Science][Medline] [Order article via Infotrieve]
  17. Carrette O, Demalte I, Scherl A, Yalkinoglu O, Corthals G, Burkhard P, et al. A panel of cerebrospinal fluid potential biomarkers for the diagnosis of Alzheimer’s disease. Proteomics 2003;3:1486-1494.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  18. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc B 1995;57:289-300.
  19. Breiman L, Friedman JH, Olshen RA, Stone CL. Classification and Regression Trees 1984:368 Wadsworth International Group Belmont, CA. .
  20. Reiber H. External quality assessment in clinical neurochemistry: survey of analysis for cerebrospinal fluid (CSF) proteins based on CSF/serum quotients. Clin Chem 1995;41:256-263.[Abstract/Free Full Text]
  21. Reiber H, Otto M, Trendelenburg C, Wormek A. Reporting cerebrospinal fluid data: knowledge base and interpretation software. Clin Chem Lab Med 2001;39:324-332.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  22. Brettschneider J, Claus A, Kassubek J, Tumani H. Isolated blood-cerebrospinal fluid barrier dysfunction: prevalence and associated diseases. J Neurol 2005;252:1067-1073.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  23. Tibbling G, Link H, Ohman S. Principles of albumin and IgG analyses in neurological disorders. I. Establishment of reference values. Scand J Clin Lab Invest 1977;37:385-390.[Web of Science][Medline] [Order article via Infotrieve]
  24. Puppione DL, Fischer WH, Park M, Whitelegge JP, Schumaker VN, Golfeiz S, et al. Sequence of horse (Equus caballus) apoA-II. Another example of a dimer forming apolipoprotein. Comp Biochem Physiol B Biochem Mol Biol 2004;138:213-220.
  25. Deterding LJ, Cutalo JM, Khaledi M, Tomer KB. Separation and characterization of human high-density apolipoproteins using a nonaqueous modifier in capillary electrophoresis-mass spectrometry. Electrophoresis 2002;23:2296-2305.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  26. Bondarenko PV, Farwig ZN, McNeal CJ, Macfarlane RD. MALDI- and ESI-MS of the HDL apolipoproteins; new isoforms of APO A-I, II. Int J Mass Spectrom 2002;219:671-680.[CrossRef]
  27. Schmitz G, Ilsemann K, Melnik B, Assmann G. Isoproteins of human apolipoprotein A-II: isolation and characterization. J Lipid Res 1983;24:1021-1029.[Abstract]
  28. Lackner KJ, Edge SB, Gregg RE, Hoeg JM, Brewer HBJ. Isoforms of apolipoprotein A-II in human plasma and thoracic duct lymph. Identification of proapolipoprotein A-II and sialic acid-containing isoforms. J Biol Chem 1985;260:703-706.[Abstract/Free Full Text]
  29. Niederkofler EE, Tubbs KA, Kiernan UA, Nedelkov D, Nelson RW. Novel mass spectrometric immunoassays for the rapid structural characterization of plasma apolipoproteins. J Lipid Res 2003;44:630-639.[Abstract/Free Full Text]
  30. Martin-Campos JM, Escola-Gil JC, Ribas V, Blanco-Vaca F. Apolipoprotein A-II, genetic variation on chromosome 1q21–q24, and disease susceptibility. Curr Opin Lipidol 2004;15:247-253.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  31. Tailleux A, Duriez P, Fruchart J-C, Clavey V. Apolipoprotein A-II, HDL metabolism and atherosclerosis. Atherosclerosis 2002;164:1-13.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  32. Dayal B, Ertel NH. ProteinChip technology: a new and facile method for identification and measurement of high-density lipoproteins apoA-I and apoA-II and their glycosylated products in patients with diabetes and cardiovascular disease. J Proteome Res 2002;1:375-380.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  33. Warden CH, Daluiski A, Bu X. Evidence for linkage of the apolipoprotein A-II locus to plasma apolipoprotein A-II and free fatty acid levels in mice and humans. Proc Natl Acad Sci U S A 1993;90:10886-10890.[Abstract/Free Full Text]
  34. Higuchi K, Kitagawa K, Naiki H. Polymorphism of apolipoprotein A-II (apoA-II) among inbred strains of mice: relationship between the molecular type of apoA-II and mouse senile amyloidosis. Biochem J 1991;279:427-433.
  35. Yazaki M, Lieprieks JJ, Barats MS. Hereditary systemic amyloidosis associated with a new apolipoprotein AII stop codon mutation Stop 78 Arg. Kidney Int 2003;64:11-16.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  36. Chertov O, Biragyn A, Kwak LW, Simpson JT, Boronina T, Hoang VM, et al. Organic solvent extraction of proteins and peptides from serum as an effective sample preparation for detection and identification of biomarkers by mass spectrometry. Proteomics 2004;4:1195-1203.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  37. Boyles JK, Pitas RE, Wilson E, Mahley RW, Taylor JM. Apolipoprotein E associated with astrocytic glia of the central nervous system and with non-myelinating glia of the peripheral nervous system. J Clin Invest 1985;76:1501-1513.[Web of Science][Medline] [Order article via Infotrieve]
  38. Taborsky L, Adam P, Sobek O, Dostal M, Dvorakova J, Dubska L. Levels of apolipoprotein A-II in cerebrospinal fluid in patients with neuroborreliosis are associated with lipophagocytosis. Folia Microbiol (Praha) 2003;48:849-855.
  39. Niemi M, Häkkinen T, Karttunen TJ, Eskelinen S, Kervinen K, Savolainen MJ, et al. Apolipoprotein E and colon cancer; expression in normal and malignant human intestine and effect on cultured human colonic adenocarcinoma cells. Eur J Intern Med 2002;13:37-43.[CrossRef][Medline] [Order article via Infotrieve]
  40. Hunter S, Young A, Olson J, Brat DJ, Bowers G, Wilcox JN, et al. Differential expression between pilocytic and anaplastic astrocytomas: identification of apolipoprotein D as a marker for low-grade, non-infiltrating primary CNS neoplasms. J Neuropathol Exp Neurol 2002;61:275-281.[Web of Science][Medline] [Order article via Infotrieve]
  41. Trougakos IP, Lourda M, Agiostratidou G, Kletsas D, Gonos ES. Differential effects of clusterin/apolipoprotein J on cellular growth and survival. Free Radic Biol Med 2005;38:436-449.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  42. Alvarez ML, Barbon JJ, Gonzalez LO, Abelairas J, Boto A, Vizoso FJ. Apolipoprotein D expression in retinoblastoma. Ophthalmic Res 2003;35:111-116.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  43. Chen YC, Pohl G, Wang TL, Morin PJ, Risberg B, Kristensen GB, et al. Apolipoprotein E is required for cell proliferation and survival in ovarian cancer. Cancer Res 2005;65:331-337.[Abstract/Free Full Text]



The following articles in journals at HighWire Press have cited this article:


Home page
 Clin. Chem.Home page
H. Dihazi, G. A. Muller, S. Lindner, M. Meyer, A. R. Asif, M. Oellerich, and F. Strutz
Characterization of Diabetic Nephropathy by Urinary Proteomic Analysis: Identification of a Processed Ubiquitin Form as a Differentially Excreted Protein in Diabetic Nephropathy Patients
Clin. Chem., September 1, 2007; 53(9): 1636 - 1645.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow 069294.Supplemental data
Right arrow All Versions of this Article:
clinchem.2006.069294v1
52/8/1501    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
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 Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by de Bont, J. M.
Right arrow Articles by Pieters, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by de Bont, J. M.
Right arrow Articles by Pieters, R.
Related Collections
Right arrow Proteomics and Protein Markers

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS