Conceived and designed the experiments: LS DHG JPW KFF HIK. Performed the experiments: LS NMO. Analyzed the data: LS NMO HIK. Contributed reagents/materials/analysis tools: JPW KFF. Wrote the paper: LS HIK.
Current address: Department of Neurosurgery, Stanford Medical School, Stanford, California, United States of America
The authors have declared that no competing interests exist.
The central nervous system (CNS) develops from a heterogeneous pool of neural stem and progenitor cells (NSPC), the underlying differences among which are poorly understood. The study of NSPC would be greatly facilitated by the identification of additional proteins that mediate their function and that would distinguish amongst different progenitor populations.
To identify membrane and membrane-associated proteins expressed by NSPC, we used a proteomics approach to profile NSPC cultured as neurospheres (NS) isolated from the murine cortex during a period of neurogenesis (embryonic day 11.5, E11.5), as compared to NSPC isolated at a peak of gliogenesis (postnatal day 1, P0) and to differentiated E11.5 NS. 54 proteins were identified with high expression in E11.5 NS, including the TrkC receptor, several heterotrimeric G proteins, and the Neogenin receptor. 24 proteins were identified with similar expression in E11.5 and P0 NS over differentiated E11.5 NS, and 13 proteins were identified with high expression specifically in P0 NS compared to E11.5 NS. To illustrate the potential relevance of these identified proteins to neural stem cell biology, the function of Neogenin was further studied. Using Fluorescence Activated Cell Sorting (FACS) analysis, expression of Neogenin was associated with a self-renewing population present in both E11.5 and adult subventricular zone (SVZ) NS but not in P0 NS. E11.5 NS expressed a putative Neogenin ligand, RGMa, and underwent apoptosis when exposed to a ligand-blocking antibody.
There are fundamental differences between the continuously self-renewing and more limited progenitors of the developing cortex. We identified a subset of differentially expressed proteins that serve not only as a set of functionally important proteins, but as a useful set of markers for the subsequent analysis of NSPC. Neogenin is associated with the continuously self-renewing and neurogenic cells present in E11.5 cortical and adult SVZ NS, and the Neogenin/RGMa receptor/ligand pair may regulate cell survival during development.
The central nervous system (CNS) develops from a population of neural stem and progenitor cells (NSPC) in a spatially and temporally defined manner, with prenatal neurogenesis followed by a wave of postnatal gliogenesis, to generate the appropriate architecture, and types and number of cells of which the mature CNS is compromised
Gene expression analysis has identified transcriptional differences that exist amongst various populations of NSPC and several candidate stem and progenitor genes have been identified
To identify proteins that may define subpopulations of NSPC, we chose to compare membrane and membrane-associated protein expression profiles of cortical neurospheres (NS) generated during a highly neurogenic period (embryonic day 11.5, E11.5) and during a gliogenic period (postnatal day 1, P0). The potential and longevity of these NS cultures was characterized, with E11.5 NS reflective of a more stem cell-like population, and the P0 NS, of a more restricted progenitor. Using protein expression analysis, we identified differences in membrane and membrane-associated proteins expressed by these populations of NS, including the receptor, Neogenin, which may have different functions as development proceeds and which may be a marker for an early embryonic cortical NSPC. These experiments demonstrate fundamental differences between embryonic and postnatal cortical NSPC, and provides a list of candidate membrane and membrane-associated proteins expressed by NSPC.
To validate the cell source for the subsequent proteomics experiments, NS cultures from E11.5 and P0 cortex were characterized according to proliferation, multipotentiality and longevity in culture. To examine the proliferation of E11.5 and P0 NS cultures with time, low density cultures (1,000 cells/ml) were generated from three separate
Representative phase contrast images of NS generated from E11.5 (A–C) and from P0 (D–F) cultures, started on D1 (A, D), D7 (B, E) and on D14 (C, F). Scale bar, 50 µm. G.) NS produced as a percentage of the original plating density from E11.5 and P0 cultures as a function of culture start day. The data represent n = 5 independent experiments. Results are presented as the mean % NS produced +/− standard error of the mean (SEM). * p≤0.05 * * p≤0.01
To determine if there were also differences in the ability to generate the three main cell types of the CNS, differentiated NS were scored for immunoreactivity for markers of neurons (Tuj1), oligodendrocytes (O4) and astrocytes (GFAP). E11.5 cultures from D1, D7 and D14 predominately and consistently produced tripotent NS containing all three major cell types of the CNS upon differentiation (
Representative fluorescent images of neurons from differentiated E11.5 NS are shown in A–C and P0 NS in D–F (neurons in green in all panels, nuclei in red in panels B, C, D, E, and astrocytes in red in panels A and C). E11.5 (G) and P0 (H) individual differentiated NS derived from D1, D7 and D14 cultures were scored for presence of Tuj1-, O4-, and GFAP-immunoreactive cells: ‘A’, astrocytes only; ‘N/A’ - both neurons and astrocytes; ‘O/A’ – both oligodendrocytes and astrocytes; ‘N/O/A’ – presence of all three cell types. Significant differences in the generation of di- and multi-potent NS existed between E11.5 (G) and P0 (H) NS cultures, including an increase in P0 NS containing only O/A at Day 7 (p<0.05), and decreases in tripotent P0 NS at Day 7 (p<0.05) and at Day 14 (p<0.01). Results are presented as the % of total NS +/− SEM. The percent of neurons (I), astrocytes (J), and oligodendrocytes (K) produced by E11.5 (shaded symbols) and P0 NS (open symbols) as a function of time in culture was evaluated by scoring random fields. Results are presented as the mean % of total cells +/− SEM. Note different y-axis scale for I, J, and K. * p≤0.05 * * p≤0.01
To further reveal differences in the neurogenic and gliogenic capacity of the embryonic and postnatal cultures, the percent of each cell type produced upon differentiation was determined. Low density E11.5 D1, D7 and D14 cultures produced all three neural cell lineages, with no significant differences in the percent of cell type produced over time (
To identify proteins that define neurogenic E11.5 NS, membrane enriched fractions from three groups of cells were compared by 2DGE (
SyproRuby-stained representative 2D gels (A) of membrane-enriched fractions from E11.5 NS, P0 NS, and E11.5 differentiated NS. The Venn diagram (B) summarizes the number of identified proteins with differential expression between E11.5 NS and P0 NS. Overlap between the two sets indicates those proteins expressed by both E11.5 and P0, yet increased over E11.5 differentiated NS. The predicted sub-cellular locations of these proteins according to 1 WoLF PSORT (wolfpsort.org) are listed in (C).
Protein Name | Gene Symbol | UniProtKB/TrEMBL Number |
Voltage-dependent anion channel 2 | Vdac2 | Q99L98 |
Neogenin | Neo1 | P97798 |
Seizure related gene 6 | Sez6 | Q7TSK2 |
Sodium-calcium exchanger | Slc8a1 | O35157 |
Neurotrophic tyrosine kinase receptor | TrkC | Q6VNS1 |
TGF-beta receptor type III | Tgfbr3 | O88393 |
Guanine nucleotide-binding protein | Gnao1 | P18872 |
Guanine nucleotide-binding protein | Gna12/Gna13 | P27600 |
GTP-binding protein REM 1 | Rem1 | O35929 |
Guanine nucleotide binding protein | Gnb2-rs1 | Q5NCC6 |
Guanine nucleotide-binding protein | Gnb2l1 | Q9CSQ0 |
Similar to interleukin 17 receptor E | Il17re | Q6AZ51 |
Hypothetical protein C6B12.02c | SPAC6B12.02c | O14207 |
Receptor tyrosine-protein kinase | Erbb2 | Q6ZPE0 |
Alpha 3 catenin | Ctnna3 | Q8C0N3 |
FK506 binding protein 9 | Fkbp9 | Q80ZZ6 |
Cyclic nucleotide gated channel alpha 2 | Cnga2 | Q80XH6 |
Selectin P | SELP | Q5TI45 |
ATPase, H+ transporting | Atp6v0d1 | Q921S5 |
Arsenical pump-driving ATPase | Asna1 | O54984 |
(smad8/smad9) Mothers against decapentaplegic homolog 9 | Smad9 | Q9JIW5 |
Down-regulated by Ctnnb1, a | Drctnnb1a | Q6P9N1 |
Tumor rejection antigen gp96 | Tra1 | Q8CCY5 |
Early endosome antigen 1 | Eea1 | Q8BL66 |
Neogenin was highly expressed in E11.5 NS, as confirmed by western blot analysis of membrane-enriched fractions from E11.5 NS, E11.5 Differentiated NS and P0 NS (
Neogenin protein was expressed at high levels in the membrane enriched fraction of E11.5 NS over P0 NS, and decreased upon differentiation (A), while mRNA levels appeared to be expressed at equal levels (B). Equal amounts of protein and of RNA were loaded into each well, with GAPDH used as a control for the semi-quantitative PCR.
To characterize the Neogenin-expressing cells, E11.5 and P0 NS were examined by immunocytochemistry for co-expression of Nestin, a putative marker for neural stem and progenitor cells, and Neogenin (
Using an antibody to the extracellular portion of Neogenin that does not block ligand binding
Mature secondary NS from both P0 (B) and E11.5 (A) cortex were dissociated and sorted according to levels of Neogenin expression. Cells were grown at 5,000 cells/ml (Passage 1–2 in C, D), followed by passages at 1,000 cells/ml (Passage 3–5 in C, D). FACS experiments were replicated with a second culture with similar results. Mature tertiary adult SVZ NS were sorted according to levels of Neogenin expression, and cultured for 14 d (E). Experiments were performed in triplicate. Results are presented as the % NS generated +/− standard deviation. Addition of ligand-blocking antibody resulted in increased activated Caspase3/7 in E11.5 NS (F). Anti-Neogenin antibody was added at a concentration of 5 µg/ml and cells were exposed to antibody for 24 h. Results are presented as the percent of control activated Caspase3/7, using hydrogen peroxide as a positive control treatment, and PBS and mouse IgG as negative control treatments. *p≤0.05.
Based on previous work that suggests Neogenin may function as a dependence receptor, as well as by the identification of an active Caspase3 cleavage site in the intracellular region of the protein
RGMa, a GPI-linked protein, has been previously identified as a high affinity ligand for the Neogenin receptor
RGMa protein expression appeared to be similar across E11.5 and P0 NS and E11.5 differentiated NS (A), as is RGMa mRNA (B). Western blot analysis of concentrated E11.5 and P0 NS conditioned medium (CM) demonstrate NS do not release endogenous RGMa
By immunocytochemistry, approximately 41% of the dissociated cells in E11.5 NS expressed RGMa while only 22% of the cells from P0 NS expressed RGMa (
In these experiments, embryonic (E11.5) and postnatal (P0) murine cortex-derived NS were used as a source of NSPC for the analysis of membrane and membrane-associated proteins. These cultures were characterized to reveal that E11.5 NS contain cells that were highly neurogenic and proliferative as compared to P0 NS. We identified membrane and membrane-associated proteins highly expressed by E11.5 NS, as compared to P0 NS and differentiated E11.5 NS. To demonstrate the relevance of these proteins to stem cell biology, Neogenin, one of the proteins highly expressed by E11.5 NS, was further studied, and while additional experiments are warranted, the protein may play a role in NSPC survival.
While the NS culture model has its limitations, the use of acutely isolated cortical tissue has its own limitations, specifically that the tissue itself will be a heterogeneous mix of cells (including postmitotic cells) and it is difficult to precisely ascertain the proliferation and potential of this heterogeneous population of cells. The NS culture model attempts to reduce this heterogeneity, as most if not all postmitotic cells are eliminated in the initial passage. We chose the NS culture model, in part because we were able to address the issue of obtaining sufficient starting material, but also because we were able to comprehensively and directly test the proliferative abilities and the potential of the cultures which were being further interrogated by protein expression. The NS-generating cells isolated from E11.5 cortex were able to self-renew, to produce the major cell lineages of the CNS, and to do both with longevity
To further characterize the proliferative and neurogenic cells in the E11.5 NS, highly expressed membrane and membrane-associated proteins specific to this population were identified, over both differentiated E11.5 NS and the more restricted P0 NS. Of those identified, 54 were highly expressed in E11.5 NS over both differentiated E11.5 NS and P0 NS, 24 proteins were comparably expressed in E11.5 and P0 NS over differentiated E11.5 NS, and 13 proteins were highly expressed specifically in P0 NS compared to E11.5 NS. The analysis of hydrophobic membrane proteins continues to present significant challenges to the field of proteomics. No one separation technique has emerged to meet all needs, with both liquid- and gel- based separation techniques presenting specific issues relating to solubility of membrane proteins. In our gel-based approach, we maximized membrane protein representation by utilizing the zwitterionic sulfobetaine detergent, amidosulfobetaine-14 (ASB14), to solubilize the membrane-enriched pellet, a detergent which is also compatible with isoelectric focusing, in addition to overnight isoelectric strip rehydration. In the membrane-enriched fraction, 31% of the identified proteins were classified as membrane or membrane-associated. The remaining identified proteins likely reside in other compartments or in the membranes of other organelles; however, as published sub-proteomics analyses increase, proteins hitherto thought to have predictable sub-cellular locations are increasingly being found in other compartments. Heat shock protein 90 (HSP90) for instance, which has increased expression in E11.5 NS, has been observed in the cytoplasm and the outer membrane, with distinct location-specific functions, and is currently a clinically relevant target in tumour metastasis
Several heterotrimeric G proteins were also highly expressed in E11.5 NS, including guanine nucleotide-binding protein G (o) (alpha subunit 1), guanine nucleotide-binding protein (alpha-12, alpha-13 subunit), and Receptor for Activated C Kinase 1 (RACK1). While the precise function of these proteins in this developmental context is not clear, G proteins are crucial in asymmetric cell division in normal development during neuroblast divisions of
The tyrosine kinase receptor, TrkC, was, surprisingly, highly expressed in E11.5 NS. The TrkC receptor, and its preferred ligand neurotrophin-3 (NT-3), has been shown to regulate neuronal differentiation and survival
Neogenin was highly expressed in a population of NSPC from E11.5 NS, as compared to the more restricted P0 NS, with overall expression of Neogenin decreasing upon differentiation. To demonstrate the biological relevance of the proteins identified in this study, Neogenin was studied in further detail to examine its role in NSPC biology.
Neogenin has been shown previously to have developmentally distinct functions and has been diversely described as an axonal guidance receptor, as a stabilizer of the mammary gland progenitor cell niche, and in the formation of the neural tube at the earliest points in development
The putative GPI-anchored ligand for Neogenin, RGMa, was also expressed by E11.5 and P0 NS cultures. RGMa is a member of a family of Repulsive Guidance proteins, originally isolated as a chemorepulsive molecule
Despite what is known, the question remains as to the precise role of the Neogenin and RGMa receptor-ligand pair in neural stem cell biology. Previous reports support the development-dependent function of these proteins. At the 2 cell stage in development in
Taken together, these data suggest that Neogenin and RGMa may have different functions during embryonic and postnatal development in NSPC
In summary, while the embryonic and early postnatal cortex contains a heterogeneous pool of progenitors as reflected in the NS culture model, characterization of these cultures according to their ability to proliferate, to generate the major cell types of the CNS and to do so with longevity, has revealed important differences between NS generated at E11.5 and at P0. These studies have identified membrane and membrane-associated proteins highly expressed by proliferative and neurogenic E11.5 NS, as compared to more restricted P0 NS. The identified proteins are likely candidates for further interrogation, as demonstrated by the studies on the function of Neogenin, and has provided some important insight into the broad protein signature of NSPC. These proteins may also serve as markers of NSPC, perhaps most successfully in a combinatorial approach, as has been recently demonstrated
Tissue culture reagents were obtained from GIBCO-Invitrogen. Basic fibroblast growth factor (bFGF) was obtained from Peprotech. Heparin and protease inhibitor cocktail were obtained from Sigma-Aldrich. The following antibodies were used: anti-O4 (1∶20, Chemicon), anti-Tuj1 (1∶500, Covance), anti-Nestin (1∶50, Rat 401 DSHB-University of Iowa), anti-GFAP (1∶1000, DAKO), anti-HSP90 (1∶1000, Cell Signaling), anti-RACK1 (1∶2500, BD Transduction Laboratories), anti-TrkC (1∶1000, R&D Systems), anti-actin (1∶500, Sigma), anti-Neogenin cytoplasmic (1∶1000, R&D Systems), anti-Neogenin monoclonal extracellular (5 µg/ml, R&D Systems), αRGMa (1∶1000, R&D Systems), and Alexa-conjugated secondary antibodies (1∶2000, Molecular Probes). Propidium iodide (Molecular Probes) was used at 1∶3000. The detergent, ASB14, was purchased from Calbiochem.
NS were grown from E11.5 and P0 cortex (including ventricular and subventricular zone, but with minimal ventral cortex) from CD-1 mice (Charles River), as previously described
Low density cultures were analyzed for differentiation potential and longevity. As outlined in
SVZ tissue was isolated from adult CD-1 mice, followed by mechanical trituration with TripLExpress (Invitrogen). Cells were grown at 50,000 cells/ml medium (DMEM/F12 containing B27, 20 ng/ml bFGF, 50 ng/ml EGF and penicillin-streptomycin), and passaged every 7 d. Sorting experiments were performed on tertiary SVZ NS.
Immunocytochemistry was performed as previously described
As outlined in
Protein expression was confirmed by western blot analysis using total cell lysate or the P3 fraction. Protein concentration was determined using the Bradford assay (BioRad) and equal amounts of protein were loaded. To determine presence of released RGMa, medium was collected from each culture (24–48 h post-transfection for exogenously produced RGMa), centrifuged, concentrated, and 30 µl of each was analyzed by Western blot.
Semi-quantitative RT-PCR was used to evaluate mRNA levels, with GAPDH as a control. The following primer sets were used to examine the expression of various transcripts: Neogenin sense ggg tca aga atg ggg atg tgg tta, antisense ctc tcc tgg ctg gct ggt att ctc; RGMa sense tct tcg acc tcc tca cga ct, antisense atg gtg cca agg aga atc tg.
Flow cytometry was performed in the UCLA Jonsson Comprehensive Cancer Center Flow Cytometry Facility using the Becton Dickinson FACSVantage SE and FACSAriaII High-Speed Cell Sorter Flow Cytometers. E11.5 and P0 NS were grown at 5,000 to 10,000 cells/ml, while adult SVZ NS were grown at 200,000 cells/ml. Cells were exposed to 10 µg/ml anti-Neogenin antibody (extracellular, non-ligand blocking, gift of Dr. H. Cooper) for 1 h followed by exposure to Alexa 488 secondary antibody for 1 h. Sorted E11.5 and P0 cells were grown at 5,000 cells/ml in complete Neurobasal medium for two passages, followed by passages at low density (1,000 cells/ml). NS were counted in each culture prior to passaging. Adult SVZ NS were sorted into a 96-well plate containing medium at 20 cells per well and allowed to grow for 2 weeks before assessing sphere formation. Both antibodies to the extracellular portion of Neogenin labeled the same cells
E11.5 and P0 NS were grown at 2,500 to 10,000 cells/ml. Anti-Neogenin antibody was added at 2.5 to 10 µg/ml of medium and cells were grown for 24 h. Hydrogen peroxide, anti-mouse IgG, and PBS were used as positive and negative controls for the apoptosis assay. Caspase3/7 activities were measured using the Caspase-Glo 3/7 Assay (Promega), according to the manufacturer's protocol, and luminescence was detected using the Analyst HT Microplate Reader (LJL Biosystems).
Partial list of proteins with higher expression in both E11.5 and P0 NS as compared to differentiated E11.5 NS.Proteins were identified by μLC-MS/MS. Detailed information regarding protein identification can be found in
(0.03 MB DOC)
List of all proteins identified by μLC-MS/MS. One of two programs was used to establish protein identification: Mascot or Sonar. The respective scores for each protein are listed, as well as the identified peptide sequence. LCQ-DECA MS/MS data sets were analyzed by Sonar MS/MS software (Genomic Solutions, Version 2004.01.15.01), and QSTAR MS/MS data sets were analyzed by Mascot software (Matrix Science), with reference to databases from NCBI, SIB and EBI. An Expect score of greater than 1×10-2 was considered a positive identification. Search parameters included +/− 2 Da precursor, +/− 0.4 Da fragment, 3 missed cleavages by trypsin, carboxyamidomethylation of cysteines, and oxidized methionines.
(0.24 MB PDF)
Neurosphere culture design. Acutely dissociated telencephalon (E11.5) or cortex (P0) from CD1 mice was grown at both high density (50,000 cells/ml) and low density (1,000 cells/ml, designated D1) as described. Subsequent D7, D14, and D21 low density cultures were derived from fully mature high density 1°, 2°, or 3° NS, respectively. High density cultures were passaged every 7d, and low density cultures every 14d.
(0.34 MB TIF)
Proteomics workflow for identification of differentially expressed proteins in NS by 2DGE. Additional details regarding the methodology can be found in the
(4.00 MB TIF)
Representative total ion chromatograph and MS/MS spectrum. A total ion chromatograph (inset, B) and MS/MS spectrum (A) of a tryptic peptide of guanine nucleotide-binding protein beta subunit 2-like 1, a heterotrimeric G protein. The MS/MS spectrum shown is focused in the mass range where the strongest b and y ions are present.
(2.32 MB TIF)
Protein expression for several identified proteins. Western blot analysis of the membrane-enriched preparation for TrkC, Rack1, and HSP90 of E11.5 NS (lane 1), E11.5 Differentiated NS (lane 2), P0 NS (lane 3). The loading control, Actin, was used for total cell lysate. Protein loading was equal in all lanes and measured by Bradford analysis.
(0.10 MB TIF)
Neogenin is highly expressed in E11.5 NS and is co-expressed with Nestin in a sub-population of cells. By immunocytochemistry (A), Neogenin (green) and Nestin (blue) expression was evident in both E11.5 and P0 NS, although there was higher expression of both in E11.5 NS (top panel, A). The percent of cells expressing these proteins is shown in (B). Cell nuclei are shown in red (propidium iodide).
(0.09 MB TIF)
Incubation with ligand-blocking anti-Neogenin antibody increases percent of tryphan-blue positive cells in E11.5 cells. E11.5 and P0 cells were incubated with either the ligand-blocking anti-Neogenin antibody or the cell sorting anti-Neogenin antibody for 3h and tryphan blue positive cells were counted.
(0.06 MB TIF)
To Dr. H. Cooper for the provision of the extracellular, non-ligand blocking anti-Neogenin antibody.