Proteomic identification of biomarkers in the cerebrospinal fluid (CSF) of astrocytoma patients

Abstract

The monitoring of changes in the protein composition of the cerebrospinal fluid (CSF) can be used as a sensitive indicator of central nervous system (CNS) pathology, yet its systematic application to analysis of CNS neoplasia has been limited. There is a pressing need for both a better understanding of gliomagenesis and the development of reliable biomarkers of the disease. In this report, we used two proteomic techniques, two-dimensional gel electrophoresis (2-DE), and cleavable Isotope-Coded Affinity Tag (cICAT) to compare CSF proteomes to identify tumor- and grade-specific biomarkers in patients bearing brain tumors of differing histologies and grades. Retrospective analyses were performed on 60 samples derived from astrocytomas WHO grade II, III, and IV, schwannomas, metastastic brain tumors, inflammatory samples, and non-neoplastic controls. We identified 103 potential tumor-specific markers of which 20 were high-grade astrocytoma-specific. These investigations allowed us to identify a spectrum of signature proteins that could be used to distinguish CSF derived from control patients versus those with low- (AII) or high-grade (AIV) astrocytoma. These proteins may represent new diagnostic, prognostic, and disease follow-up markers when used alone or in combination. These candidate biomarkers may also have functional properties that play a critical role in the development and malignant progression of human astrocytomas, thus possibly representing novel therapeutic targets for this highly lethal disease.

Figure 1. Representative 2-DE gels of CSF

Albumin and IgG depleted and TCA precipitated CSF samples from patients with A) non-tumoral control (#5) B) astrocytoma WHO grade IV (#27) were analyzed by 2-DE analysis using IEF strips pH 4–7 and 12.5% SDS-PAGE. Spots were visualized by silver staining and spot patterns were analyzed using ImageMaster software. Each sample was run in triplicate.

Figure 2. Analysis of CSF using proteomic methods

A) Average number of proteins found in CSF samples from each study group. Error bars indicate range of proteins found within the group; B) Functional categories and number of proteins identified within each functional category in the CSF samples analyzed by 2-DE and ICAT analysis; Low-grade = AII and high-grade = AIV CSF, and C) Ven diagram showing differentially expressed proteins found by 2-DE (dark grey); cICAT (light grey) and both (white) in the CSF from glioma patients compared to non-tumoral controls. D) Ven diagram showing distribution of the 130 differentially expressed proteins found in control and each astrocytoma grade samples.

Figure 3. Examples of differentially expressed proteins found by 2-DE and identified by MS/MS analysis

A) Differential expression of and FGF-14 on enlarged sections of representative 2-DE gel of control and AIV CSF. Solid circles highlight detected proteins by silver stain. B) Representative spectrum from one of the 15 peptides found from silverstained 2-DE gel spots for PEDF identified by MS/MS. The spectrum shown, identifies a tryptic peptide (LQSLFDSPDFSK) for PEDF.

Figure 4. Verification of 2-DE and cICAT results by western blot analysis

Differential expression of SPARC, beta-2-microglobulin (β2m), beta-defensin-6 (β-def-6), beta-defensin-7 (β-def-7), Attractin (Atrn), FGF-14L (35 kd), FGF-14S (27 kd), VEGF-B and Tau were verified in the CSF from AII- AIV patients compared to non-tumoral controls. Transthyretin (TTR) was used as a loading control. Representative results (A, B, C) are shown on different patient samples of a total of 10 controls, 6 AII, 7 AIII and 19 AIV samples analyzed. Sample numbers are indicated above each lane.

Figure 5. Potential biomarker panels differentiate between astrocytoma grades using CSF

A) Panel of 32 proteins that can be used to differentiate between AII-IV patients, non-tumoral controls, and other forms of brain tumors. Black = no expression, yellow = low expression, orange = moderate expression and red = high expression level as determined from ImageMaster analysis of 2-DE gels. C = Biomarker class; white = control specific, light blue = AII, green = AIII and pink = AIV biomarker. B) Restricted panel containing top 13 proteins from panel A. white = control, light blue = AII, green = AIII and purple = AIV biomarkers. Vertical axis represents relative level (0–4) of each protein as estimated from 2-DE.