Glypican-2 levels in cerebrospinal fluid predict the status of adult hippocampal neurogenesis

Adult hippocampal neurogenesis is a remarkable form of brain plasticity through which new neurons are generated throughout life. Despite its important roles in cognition and emotion and its modulation in various preclinical disease models, the functional importance of adult hippocampal neurogenesis in human health has not been revealed because of a lack of tools for monitoring adult neurogenesis in vivo. Therefore, we performed an unbiased proteomics screen to identify novel proteins expressed during neuronal differentiation using a human neural stem cell model, and we identified the proteoglycan Glypican-2 (Gpc2) as a putative secreted marker of immature neurons. Exogenous Gpc2 binds to FGF2 and inhibits FGF2-induced neural progenitor cell proliferation. Gpc2 is enriched in neurogenic regions of the adult brain. Its expression is increased by physiological stimuli that increase hippocampal neurogenesis and decreased in transgenic models in which neurogenesis is selectively ablated. Changes in neurogenesis also result in changes in Gpc2 protein level in cerebrospinal fluid (CSF). Gpc2 is detectable in adult human CSF, and first pilot experiments with a longitudinal cohort indicate a decrease over time. Thus, Gpc2 may serve as a potential marker to monitor adult neurogenesis in both animal and human physiology and disease, warranting future studies.


Discovery proteomics of neural progenitor cells (NPCs).
NPCs were cultivated in three biological batches using the standardized protocol described below. NPCs were first grown to confluence in an undifferentiated state (approximately 10-15x10 6 cells/T-175). At t = 0, NPCs were seeded in a fresh T-175 at a density of 50,000 cells/cm 2 (approximately 8.7x10 6 cells) and grown in differentiation medium (depleted of FGF and EGF) for the indicated times. The control t = 0 culture was also transferred to a fresh T-175 but was cultivated in presence of growth factors for an additional 24 h before collection. Thus, for each biological replicate, each time point (t = 0, 1, 2, 4, 7, 14, 21 and 28 days) represented a separate cell culture (in a separate T-175 dish) sharing the same biological origin.
At each time point, cells were collected from a given T-175 dish, rinsed with PBS, pelleted at low speed and stored at -80°C until processed.
NPCs collected from one T-175 dish (approximately 10-20x10 6 cells) were lysed by sonication in 1 ml of 40 mM Tris and 600 mM NaCl, pH 8.8, supplemented with Protease inhibitor complete (Roche) and 3 l of Benzonase (Sigma). The lysates were then subjected to ultracentrifugation at 100,000 g for 1 h at 10°C. The insoluble membrane fraction was then washed twice with 1 M KCl and then twice with 0.1 M Na 2 CO 3 (each wash followed by a 1 h centrifugation step at 50,000 g at 4°C) to remove the bulk of the contaminating soluble proteins. The crude membrane protein fraction was then solubilized in 8 M urea, 1% (w/v) CHAPS and 50 mM Tris/HCl, pH 8.8, and the total protein concentration was estimated using the Bradford assay (Bio-Rad). After proteins were reduced and alkylated with DTE and iodoacetamide, respectively, 10 g of total protein was applied onto two duplicate 4-12% Bis-Tris NuPAGE gels (MES buffer system, Novex), with each gel encompassing one complete biological experiment (one time point per lane). In addition, for each biological experiment, 10 g of a sample consisting of an equivalent amount of protein pooled from each time point was also analyzed separately on a separate SDS-PAGE gel (2 lanes/biological replicate). After the proteins were stained with Colloidal Blue (Novex), each lane was then gridded into 12 bands each (using the molecular mass markers as a reference) that were subsequently subjected to in-gel digestion as described below. Each sample was destained with 50 mM ammonium bicarbonate and 30% (v/v) acetonitrile and then briefly dried in a Speed-Vac concentrator before being re-suspended in 20 l of 50 mM ammonium bicarbonate containing 10 ng trypsin (Promega) overnight at room temperature. Peptides were successively extracted using 100 l of acetonitrile, 100 l of 1% (v/v) formic acid and 100 l of acetonitrile, and the three fractions were pooled and dried using a Speed-Vac concentrator. Samples were stored at 4°C until use. bars. Each peptide sample was dissolved in 20 L of buffer A (2% can and 0.5% acetic acid), from which 5 L (diluted 1:5 in the same buffer) was loaded at a flow rate of 450 nL/min in 100% buffer A for 12 min. After loading, the flow was decreased to 250 nL/min, and the peptides were eluted from the reverse phase column as follows: 12-14 min, 0-5% buffer B (80% ACN and 0.5% acetic acid); 14-30 min, 5-30% buffer B using curve 4 (slightly concave) of the Chromeleon software (Dionex); and 30-90 min, 30-55% Buffer B using curve 6 (slightly convex) of the Chromeleon software. The column was then washed with 100% buffer B for 15 min at a rate of 350 nL/min and re-equilibrated in 100% buffer A at a rate of 350 nL/min for 25 min.
Peptides were analyzed by tandem mass spectrometry using the following standard operating parameters: the electrospray voltage was set to 2.0 kV, and the capillary temperature was set to 170°C. Survey scans (scanning range m/z 400- Raw data were processed using the SEQUEST search algorithm 27 (SEQUEST version 27.0, revision 12, Thermo Electron). Searches were performed against the UniProtKB/Swiss-Prot protein knowledgebase database (version 2011_07) filtered for "Homo sapiens" (58,772 sequences concatenated with their decoy entries). The data were searched with a mass tolerance of +/-5 ppm for parent ions and +/-1.0 Da for the fragment ions. Methionines (reduced/oxidized; +15.9949 Da) were considered differential modifications, and cysteines were considered fully carbamidomethylated (+57.0199 Da). Only fully trypsinized peptides with no more than one missed cleavage were considered for data analysis. The False Discovery Rate (FDR) of the proteins identified in this study was evaluated using the Roche algorithm "Clotho" (1), and only proteins with an FDR ≤1% were examined further. When possible, identifications were reduced to a single entry per protein family by keeping the canonical UniProtKB/Swiss-Prot naming convention if equivalent isoforms were reported.

Glypican-2 interaction profiling by immunocompetitive capture mass spectrometry
Immunocompetitive capture mass spectrometry was performed essentially as previously described (1). Briefly, 21-day differentiated NPCs were harvested and lysed in Dulbecco's phosphate-buffered saline containing CaCl 2 , MgCl 2 (Gibco), 1% NP40 (Calbiochem) and protease inhibitor, complete, EDTA-free (Roche). After a 15min incubation on ice, lysates were cleared by centrifugation at 14,000 rpm for 10 min at 4°C, and protein concentrations were estimated using the bicinchoninic acid protein assay kit (Pierce). In-house produced anti-GPC2 or anti-FGF2 antibodies (Abcam ab126861) were coupled to Affi-Gel 10 agarose beads (Bio-Rad) for 3 h at 4°C. Unreacted binding sites were blocked by the addition of 0.2 M ethanolamine, and the beads were washed four times with cold PBS and stored at 4°C.
Beads (20 l) were incubated with cell lysates (850 µg total protein per condition) for 1 h at 4°C, washed four times with lysis buffer and eluted with SDS sample buffer.
For the competition experiments, cell lysates were pre-incubated with increasing amounts of free anti-GPC2 antibody (0, 100, 250, 500 and 1,000 ng) in triplicate for 1 h before incubation with the immobilized anti-GPC2 antibody. Eluates were separated on a 4-20% Tris-Glycine SDS-PAGE gel, and proteins were stained with Colloidal Blue (Novex). Four bands ranging from 20 to 120 kDa were excised and subjected to in-gel digestion with trypsin using standard procedures.
Samples were analyzed with a nanoflow Easy-nLC system (Proxeon) connected to an LTQ-Orbitrap Velos (Thermo Fischer Scientific). Peptides were loaded onto an AQUA C18 trap (100 m x 10 mm, Phenomenex) and separated on a ReproSil-PurC18-AQ (75 m x 200 mm, 3 m particle size, 120 Å, Dr. Maisch GmbH) analytical column using a 50-min gradient of 0-35% acetonitrile (with 0.6% acetic acid) at a rate of 250 nL/min. Full-scan MS spectra were acquired in the Orbitrap with a resolution of r = 60,000 (at m/z 400). The detected ions were recalibrated on the fly using the ambient air polysiloxane at m/z 445.120024 as the lock-mass. The 10 most intense ions with a charge state >1+ were selected for collision-induced dissociation in the linear ion trap and then excluded for 30 s. Raw MS files were converted into .dta files using Extract-MSn (version 1.0.0.8) and searched against a database consisting of the human part of the SwissProt database (August 2011, 35 106 entries, including splice variants) using Sequest (version 27.0, revision 12, both Thermo Fisher Scientific). Only fully trypsinized peptides with no more than one missed cleavage were considered, and mass tolerances of 5 ppm and 10 ppm were used for the precursor and fragment ions, respectively. Oxidized methionine (+15.9949 Da) was set as a differential modification, and carbamidomethylated cysteines (+57.0215 Da) were set as a static modification. The spectral false discovery rate (specFDR) was restricted to 1.0% by performing a target-decoy search using a concatenated-decoy database.
The top ranked GPC2-interacting protein candidates FGF2, ECHA and ECHB were confirmed by Western blotting. The proteins were detected with Abcam anti-ECHA (ab54477), anti-ECHB (ab88256) and anti-FGF2 (ab126861) antibodies; the latter was also used for the co-immunoprecipitation experiment.