SEARCH

SEARCH BY CITATION

Keywords:

  • Prion protein;
  • Stress inducible protein 1;
  • Neurospheres;
  • Self-renewal;
  • Proliferation

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES

Prion protein (PrPC), when associated with the secreted form of the stress-inducible protein 1 (STI1), plays an important role in neural survival, neuritogenesis, and memory formation. However, the role of the PrPC-STI1 complex in the physiology of neural progenitor/stem cells is unknown. In this article, we observed that neurospheres cultured from fetal forebrain of wild-type (Prnp+/+) and PrPC-null (Prnp0/0) mice were maintained for several passages without the loss of self-renewal or multipotentiality, as assessed by their continued capacity to generate neurons, astrocytes, and oligodendrocytes. The homogeneous expression and colocalization of STI1 and PrPC suggest that they may associate and function as a complex in neurosphere-derived stem cells. The formation of neurospheres from Prnp0/0 mice was reduced significantly when compared with their wild-type counterparts. In addition, blockade of secreted STI1, and its cell surface ligand, PrPC, with specific antibodies, impaired Prnp+/+ neurosphere formation without further impairing the formation of Prnp0/0 neurospheres. Alternatively, neurosphere formation was enhanced by recombinant STI1 application in cells expressing PrPC but not in cells from Prnp0/0 mice. The STI1-PrPC interaction was able to stimulate cell proliferation in the neurosphere-forming assay, while no effect on cell survival or the expression of neural markers was observed. These data suggest that the STI1-PrPC complex may play a critical role in neural progenitor/stem cells self-renewal via the modulation of cell proliferation, leading to the control of the stemness capacity of these cells during nervous system development. STEM CELLS 2011;29:1126–1136


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES

Prion protein (PrPC) is a cell surface glycoprotein that has been studied intensively due to its role in transmissible spongiform encephalopathies [1–3]. PrPC has also been shown to have physiological functions governing nervous system development [4]. During the development of the central nervous system (CNS), multipotent neural progenitor/stem cells generate neurons or glial cells in a finely tuned process modulated by both intrinsic factors and extracellular molecules that act as positive or negative regulators [5]. For example, soluble factors such as epidermal growth factor (EGF), brain-derived neural growth factor [6, 7], and/or the extracellular matrix (ECM) proteins, laminin and heparan sulfate, are able to control cell fate status, including the self-renewal, proliferation, survival, migration, and differentiation of neural progenitor cells [8].

PrPC has been considered to be a pivotal molecule in the brain, able to orchestrate a wide range of neurotrophic signaling events, with roles in brain development as well as neural plasticity in the adult brain [4]. The ability of PrPC to bind ECM proteins and support their proper organization and to modulate the activity of ion channels, G protein-coupling receptors, and soluble factors is likely to be the basis of its neurotrophic functions [9, 10]. The ECM constitutes an essential source of instructive signals capable of regulating the behavior of progenitor/stem cells in the developing CNS [11]. At least three PrPC partners, namely laminin [12], vitronectin [13], and heparan sulfate proteoglycan (HSPG) [14] have been described as important modulators of neural stem/progenitor cell biology [15–17]. The laminin/β1 integrin complex, a well known component of the dynamic stem cell niche in many tissues, contributes to neural progenitor/stem cells maintenance by controlling cell proliferation and survival [8, 11, 15, 18]. The role of vitronectin and HSPG in neural progenitor/stem cell behavior involves modulation of the activity of soluble factors. For example, vitronectin, when associated with the secreted signaling protein sonic hedgehog, is able to stimulate motor neuron differentiation from neuroepithelial cells [16]. In addition, HSPG binds to basic fibroblast growth factor (bFGF) and acts as a coreceptor for bFGF, a molecule crucial for the normal proliferation and differentiation of neural stem cells in the developing cerebral cortex [17].

PrPC also associates with the soluble form of stress-inducible protein 1 (STI1), a cochaperone protein with neurotrophic properties [10, 19]. STI1 binds to PrPC through specific domains that have been mapped to residues 113–128 for PrPC and 230–245 for STI1 [19]. STI1 is a ubiquitous protein with abundant expression beginning as early as embryonic day 8 (E8) in mouse nervous system development [20]. STI1 presents a spatial-temporal expression pattern similar to its ligand, PrPC, which itself has been described to be expressed as early as E7.5 [13, 21, 22]. The STI1 protein is normally found either free or complexed with heat shock proteins (Hsp) in the cytoplasm, with a small fraction localized to the nucleus [23]. The secreted form of STI1 has been observed in various cell types [10, 24, 25] and tumor samples [26]. Soluble STI1 has autocrine/paracrine activity via its binding to PrPC, which triggers mechanisms involved in development and neural plasticity, such as neuritogenesis, neuroprotection against staurosporine-induced cell death, and memory formation [10, 19, 25, 27–29]. Recently, it has been shown that STI1-PrPC-related processes require α7 nicotinic acetylcholine receptor (α7nAchR) activation for the transduction of extracellular signals [30].

Remarkably, a growing number of reports are mentioning the potential involvement of PrPC in regulation of stem cell self-renewal and/or proliferation. For example, PrPC not only is expressed on the surface of long-term repopulating hematopoietic stem cells (LT-HSCs) but also is necessary for HSC self-renewal capacity [31]. On the other hand, PrPC expression in proliferating regions of the adult brain was found to be restricted to postmitotic neurons, suggesting that it may have an indirect effect on the proliferation of the underlying mitotic precursors [32]. Interestingly, treatment of human embryonic stem cells (hESCs) with recombinant PrPC was able to delay spontaneous differentiation, contributing to maintenance of the high proliferative status observed in these cells [33].

Given the importance of PrPC in nervous system development and evidence implicating its partner STI1 as a novel neurotrophic factor with functions in neural plasticity mechanisms, we designed this study to investigate the role of PrPC-STI1 interaction in cell fate-related mechanisms. To this end, we used neurosphere cultures as a model to study neural progenitor/stem cell physiology.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES

Reagents

Mouse recombinant STI1 was purified as described previously [19]. EGF and bFGF were purchased from Sigma (St. Louis, MO). STI1 peptide (230-ELGNDAYKKKDFDKAL-245) was synthesized by Genescript (Piscataway, NJ). Monoclonal PrPC antibody (6H4) was purchased from Prionics (Zurich, Switzerland); polyclonal antibodies against STI1 (anti-STI1) and PrPC (anti-PrPC) were previously characterized [19, 27]. Anti-bromodeoxyuridine (anti-BrdU), anti–glyceraldehyde 3-phosphate dehydrogenase (anti-GAPDH), and anticleaved caspase-3 were purchased from Chemicon (Temecula, CA), Ambion (Austin, TX), and Cell Signaling (Danvers, MA), respectively. Anti-Ki67 and anti–proliferating cell nuclear antigen (anti-PCNA) were from DAKO (Glostrup, Denmark) and Zymed (San Francisco, CA), respectively. The neural markers anti-nestin, anti-glial fibrillary acidic protein (anti-GFAP), and anti-βIII-tubulin were from BD (Franklin Lakes, NJ), Chemicon, and DAKO, respectively.

Animals

PrPC-null mice (ZrchI Prnp0/0) was provided by Dr. Weissmann [34]. Prnp+/+ mice were generated by crossing F1 descendants from 129/SV and C57BL/6J matings. The animals were treated in accordance with the Canadian Council on Animal Care (CCAC) Guide (1999). The experimental procedures were approved by the Fundação Antônio Prudente Ethics Committee for animal research (Process# 025/08).

Neurosphere Primary Culture

Primary cultures of neurospheres were isolated from E14 forebrain of wild-type and Prnp0/0 mice. Forebrain was dissected in Hanks' balanced saline solution (HBSS) (Invitrogen, Carlsbad, CA) and treated with trypsin (0.25%) for 20 minutes at 37°C. Trypsin was washed out and cells were mechanically dissociated in Dulbecco's modified Eagle's medium (DMEM)-F12 medium containing B-27 (Invitrogen), Glutamine (2 mM; Invitrogen), penicillin (100 IU), and streptomycin (100 μg/ml; Invitrogen). Cells were cultured in the presence of 20 ng/ml each of EGF and bFGF at 37°C and 5% CO2. After 7 days in vitro (DIV 7), neurospheres were dissociated for use in all assays. In neurosphere cloning assay, 200 cells per well were plated on 96-well plates and incubated for 7 days at 37°C and 5% CO2. Every 2 days, cells were treated with STI1 or antibodies. Neurospheres were imaged and the number and diameter were measured using ImageJ software. Neurosphere diameter was estimated by drawing a line across the center of the sphere and the length of which was recorded in micrometers.

Immunofluorescence

Whole neurospheres were fixed with 4% paraformaldehyde and embedded in paraffin. Sections (3 μm) were deparaffinized, rehydrated, and submitted to epitope retrieval by microwave in 10 mM citrate buffer (pH 6.4) and treated with 50 mM glycine. Sections were blocked with phosphate-buffered saline (PBS) containing 0.2% Triton-X100 and 20% goat serum at room temperature (RT) for 1 hour. Sections were incubated at RT for 16 hours with anti-PrPC (1:250), anti-STI1 (1:100), anti-βIII-tubulin, anti-GFAP, anti-nestin, anti-BrdU (1:100), anti-Ki67 (1:50), or anti-PCNA (1:100) in TBS 0.1% Triton-X100 with 1% goat serum. Following washes, anti-mouse Alexa-568 or anti-rabbit Alexa-488 (1:3,000, Molecular Probes, Eugene, OR) was incubated for 1 hour at RT followed by DAPI (Sigma) staining.

Neurospheres were dissociated with trypsin (0.25%) in HBSS for 20 minutes at 37°C. Trypsin was washed out, and cells were mechanically dissociated in DMEM-F12. Cells (1 × 105 cells) were plated on cover slips coated with poly-L-lysine and treated with retinoic acid (100 nM), bone morphogenetic protein 2 (BMP-2; 50 ng/ml), or insulin-like growth factor 1 (IGF-1; 500 ng/ml) for 48 hours at 37°C. Cells were washed with PBS and fixed for 20 minutes. After blocking, cells were incubated at RT for 16 hours with anti-βIII-tubulin, anti-GFAP, or anti-nestin. After washing, secondary antibodies fluorescent-conjugated plus DAPI were incubated as described above. Cells were imaged using a Leica (Cambridge, United Kingdom) TCS SP5 II laser scanning confocal system.

Immunoblotting Analysis

Protein extracts or conditioned medium (CM) from neurospheres were analyzed. CM was centrifuged (10,000g), filtered (20 μm), and 50X concentrated (Minicon, Millipore, Molsheim, France). Protein extracts (40 μg) or CM was subject to 10% sodium dodecyl sulfate (SDS)–polyacrylamide gel electrophoresis followed by immunoblotting with anti-STI1 (1:10,000), anti-PrPC (1:1,000), anti-Nestin (1:1,000), anti-GFAP (1:1,000), or anti-βIII-tubulin (1:1,000) antibodies. Anti-actin antibodies (1:200, Sigma) were used as protein loading controls for cell extracts. Anti-GAPDH antibody was used as a cell lysis control.

Flow Cytometry

Nonpermeabilized Cells

Cells (105) were dissociated, washed twice with PBS, and incubated with anti-PrPC (1:100) in PBS containing 5% bovine serum albumin (BSA) for 30 minutes at 4°C. After washes, samples were incubated with anti-mouse Alexa-568 at 1:3,000 for 30 minutes at 4°C.

Permeabilized Cells

Cells (105) were dissociated, fixed, and permeabilized using BD Cytofix/Cytoperm Plus kit according to the manufacture's instructions. The double labeling was performed by combinations of anti-nestin, anti-GFAP, or anti-βIII-tubulin in PBS containing 5% BSA for 30 minutes on ice. Anti-mouse Alexa-568 or anti-rabbit Alexa-488 (1:3,000) was incubated for 30 minutes at 4°C. Flow cytometry was performed using a FACSCalibur (Becton Dickinson, Franklin Lakes, NJ) instrument.

Enzyme-Linked Immunosorbent Assay

Wells were coated with CM (50 μl) overnight at 4°C, washed with PBS plus 0.3% Triton X-100, blocked with PBS containing 5% nonfat milk, and incubated for 2 hours at 37°C. Anti-STI1 (7 μg/ml) was incubated for 2 hours at 37°C. After washing, anti-rabbit IgG–horseradish peroxidase (1:2,000, GE Healthcare, Piscataway, NJ) was incubated for 1 hour at 37°C followed by the addition of orthophenylenediamine solution (0.33 mg/ml in 0.5 M citrate buffer, pH 5.2, and 0.4% hydrogen peroxide) for 5 minutes at RT. The reaction was halted by the addition of 4 M sulfuric acid. Absorbance (490 nm) was measured using a Bio-Rad Benchmark (Hercules, CA) microplate reader.

Cell Death Assay

Dissociated cells (2 × 105) were grown on poly-L-lysine coated cover slips and incubated with STI1 (1 or 2 μM) or control buffer (TBS) for 24 hours at 37°C and 5% CO2. Staurosporine (50 nM) was used as positive control. Cell cultures were fixed for 20 minutes at RT and incubated at RT for 1 hour with anticleaved caspase-3 (1:100) diluted in PBS containing 1% BSA. Anti-rabbit IgG Alexa 488 (1:3,000) plus DAPI was incubated for 1 hour. The percentage of cleaved caspase-3 positive cells in the total population (DAPI counterstained) was quantified.

Proliferation Assays

Thymidine Incorporation

Neurospheres were treated as described above and pulsed with 0.4 μCi [3H]-thymidine per well for 16 hours at 37°C and 5% CO2. Neurospheres were imaged and counted. Cells were washed with PBS and lysed with 1% SDS for 15 minutes at RT and [3H]-thymidine incorporation was measured in cell lysates. The ratio between the number of spheres formed per counts per minute was plotted.

BrdU Incorporation

Neurospheres treated 16 hours with STI1 (0.5 or 1 μM) at 37°C and 5% CO2 were pulsed with 30 μM BrdU for 2 hours at 37°C. After dissociation, cells were washed and BrdU was immunostained using BD BrdU Flow kit according to the manufacturer's instructions (BD Biosciences).

Statistical Analysis

Results are represented as mean ± SE. Data were compared by one-way analysis of variance followed by Tukey's multiple comparison test or single mean Student's t test. In all cases, p < .05 were considered statistically significant.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES

Neurosphere Culture Characterization

Neurosphere cultures were obtained from the E14 forebrains of wild-type (Prnp+/+) and knockout (Prnp0/0) mice. Prnp+/+ cells were maintained for several passages in the presence of EGF and bFGF without apparent loss of either self-renewal or multipotentiality, as reflected by their continuous capacity to differentiate into neurons (Fig. 1A, left), astrocytes (center), or oligodendrocytes (right) upon exposure to differentiation protocols using retinoic acid, BMP-2, or IGF-1, respectively (Fig. 1A). Similar pattern was observed in Prnp0/0 cells (data not shown). These results confirm that sphere-forming cells represent typical neural progenitor/stem cells.

thumbnail image

Figure 1. Neural marker expression in Prnp+/+ and Prnp0/0 neurospheres. (A): Dissociated Prnp+/+ neurospheres differentiate into neurons (βIII-tubulin), astrocytes (GFAP), and oligodendrocytes (RIP) after induction with retinoic acid (100 nM) plus forskolin (5 μM), BMP-2 (50 ng/ml), or IGF (500 ng/ml), respectively. Nuclei were stained with DAPI (100 ng/ml). (B): Cell extracts from neurospheres were submitted to sodium dodecyl sulfate–polyacrylamide gel electrophoresis and the content of neural markers and actin (loading control) in neurospheres under basal conditions was analyzed by western blotting. (C): Representative dot-plots from flow cytometry experiments demonstrating Prnp+/+ cells labeled with neural markers (βIII-tubulin/GFAP or βIII-tubulin/Nestin) in basal conditions at 7 days in vitro. (D): Comparison of neural markers expression between Prnp+/+ and Prnp0/0 cells. Values represent the mean ± SE of seven independent experiments, p > .05 for all comparisons. (E): Neurosphere suspensions were harvested, fixed, paraffin-embedded, and double-labeled with antibodies against βIII-tubulin, GFAP, or nestin. Nuclei were stained with DAPI. (F): Immunofluorescence in paraffin-embedded neurosphere sections with anti-BrdU and (G) anti-Ki67 or anti-PCNA antibodies. Nuclei were stained with DAPI. Scale bars = 20 μm, except in RIP+ (Fig. 1A, right), which represents 10 μm. Abbreviations: DAPI, 4′6-diamidino-2-phenylindole; GFAP, glial fibrillary acidic protein; PCNA, proliferating cell nuclear antigen; RIP, oligodendrocyte marker.

Download figure to PowerPoint

To evaluate the profile of neural markers expression in neurospheres, undifferentiated cells, astrocyte-committed cells, and newborn neurons were identified using antibodies against nestin, GFAP, and βIII-tubulin, respectively. Neurospheres derived from both Prnp+/+ and Prnp0/0 mouse embryos expressed the three neural markers at similar levels, as assessed by semiquantitative analysis (Fig. 1B) and by immunophenotyping of the cell population using flow cytometry (Fig. 1C, 1D). The majority of cells were found to be double-positive, indicating that neurosphere cells express neuronal, glial, or undifferentiated markers (Fig. 1C, 1D). In addition, in situ immunofluorescence of neurosphere sections confirmed the expression of the three neural markers in cells derived from wild-type and Prnp0/0 mice (Fig. 1E). Remarkably, the pattern of immunoreactivity was uniform for βIII-tubulin and GFAP, whereas nestin expression presented a peculiar staining, with most of the positive cells being located at the neurosphere periphery (Fig. 1E). These findings suggest that neurosphere borders could represent a niche for undifferentiated cells, probably due to the continuous exposure of these cells to mitogenic factors present in the medium.

To address whether the nestin-positive cells represent dividing cells, the cells proliferative status was assessed by BrdU incorporation. In agreement with the observed nestin distribution, BrdU-positive cells were preferentially located at the neurosphere periphery (Fig. 1F), indicating that neurosphere borders represent a region conducive to cell proliferation. This finding cannot be attributed to improper BrdU infiltration into the neurosphere mass, as the same staining pattern was observed using endogenous nuclear markers of proliferation such as Ki67 and PCNA (Fig. 1G).

Expression and Distribution of PrPC and STI1 in Neurospheres

STI1 and PrPC expression patterns were examined in whole neurosphere cultures using confocal microscopy. Neurosphere sections derived from Prnp+/+ cells showed strong and unvarying immunoreactivity for both PrPC and STI1 (Fig. 2A, upper panels). No signal was observed with the respective control antibodies (data not shown). Immunoreactivity for PrPC and STI1 colocalizes in Prnp+/+ neurospheres (Fig. 2A, middle panels, merge), suggesting that these proteins may interact in progenitor cells. High expression levels were also observed for STI1 in Prnp0/0 neurospheres despite the absence of PrPC expression (Fig. 2A, lower panels).

thumbnail image

Figure 2. Stress-inducible protein 1 (STI1) and prion protein (PrPC) expression in Prnp+/+ and Prnp0/0 neurospheres. (A):Prnp+/+ and Prnp0/0 neurospheres were cultured as low-density cell suspensions in the presence of epidermal growth factor and basic fibroblast growth factor. Neurospheres were fixed, paraffin-embedded, and sectioned (3 μm). Immunofluorescence was performed to detect STI1 (green) and PrPC (red), and the nuclei were stained with DAPI (blue). Middle panels represent high magnification images from Prnp+/+ neurospheres (demarcated area in panel above). (B): Western blotting analysis for PrPC, STI1, and actin protein expression in extracts from neurospheres. (C): FACS analysis of Prnp+/+ cells (purple curve) labeled with anti-PrPC antibody. Abbreviations: DAPI, 4′6-diamidino-2-phenylindole; PrPC, Prion protein; STI1, stress-inducible protein 1.

Download figure to PowerPoint

Levels of STI1 and PrPC protein were also examined by immunoblotting in homogenates from Prnp+/+ and Prnp0/0 neurospheres. Figure 2B shows that STI1 protein levels are similar in neurospheres from both genotypes, indicating that STI1 expression is independent of PrPC expression. To evaluate whether PrPC is correctly targeted to the plasma membrane, dissociated nonpermeabilized neurosphere cells were analyzed by flow cytometry. As shown in Figure 2C, PrPC was detected at the cell surface, demonstrating its proper targeting.

PrPC Plays a Role in Neurosphere Formation

Based on the PrPC expression observed in progenitor cells, and previous reports showing that PrPC expression may be a prerequisite in neurogenesis [32], we asked whether PrPC may be involved in neural progenitor/stem cell maintenance. To address this question, we tested the self-renewal capacity of Prnp+/+ and Prnp0/0 progenitor cells. Neurospheres were dissociated and plated in clonal density in the presence of EGF and bFGF and, at DIV 7, the number of secondary neurospheres was counted. Cultures of Prnp0/0 cells presented fewer neurospheres than wild-type cultures (Fig. 3A). Representative images of Prnp+/+ and Prnp0/0 neurospheres are shown in Figure 3B, 3C. We also tested whether the absence of PrPC expression influences neurosphere size. Most of the neurospheres formed presented similar diameters (30–120 μm) in cultures from both genotypes (Fig. 3D). These data indicate that the ablation of PrPC reduces the number of neurospheres but does not influence sphere diameter. Accordingly, the blockage of PrPC on the cell surface with PrPC antibody (6H4) led to a significant decrease in the number of neurospheres when compared with control IgG treatment (Fig. 3E). Remarkably, the number of wild-type neurospheres formed in the presence of 6H4 antibody was equivalent to that observed in Prnp0/0 cultures (Fig. 3A vs. 3E) indicating that PrPC blockade is able to mimic the phenotype observed for Prnp0/0 neurospheres. Using flow cytometry, we have shown that 6H4 antibody binds to PrPC on the cell surface without affecting the protein internalization, leading us to consider a blocking effect for this molecule in our cell cultures (data not shown). Together, these results suggest that PrPC is a positive regulator of neural progenitor/stem cells self-renewal.

thumbnail image

Figure 3. Prion protein (PrPC) ablation impairs neurosphere formation. Dissociated primary neurospheres from Prnp+/+ and Prnp0/0 mice were cultured at a low density for 7 days in the presence of epidermal growth factor and basic fibroblast growth factor. The (A) number and the (D) diameter of Prnp+/+ and Prnp0/0 neurospheres were measured. Values represent the mean ± SE (vertical bars) of four independent experiments. Representative images of (B)Prnp+/+ and (C)Prnp0/0 neurosphere cultures. (E): Cultures were treated with anti-PrPC antibody (6H4, 6 μg/ml) or control mouse IgG and the number of neurospheres was counted. Values represent the mean ± SE of three independent experiments. In (A) and (E), single-mean Student's t test was used to compare Prnp+/+ versus Prnp0/0 neurospheres or 6H4 versus control IgG. In (D), analysis of variance followed by Tukey's multiple comparison test was used. *, p < .01. Scale bars = 200 μm.

Download figure to PowerPoint

The Interaction of STI1 with PrPC Enhances Neurosphere Formation

The soluble form of STI1 secreted by astrocytes has been shown to have trophic properties both in neurons and glia [10, 25]. Therefore, we decided to evaluate whether neurospheres are also able to secrete STI1 into the extracellular milieu. The presence of secreted STI1 in the CM of Prnp0/0 and Prnp+/+ neurosphere was measured by enzyme-linked immunosorbent assay (Fig. 4A) and immunoblotting (Fig. 4B) and did not differ between the genotypes. To confirm that soluble STI1 in CM is not a result of cell lysis, GAPDH immunostaining was performed. A very weak signal was observed for GAPDH, indicating negligible cell disruption (Fig. 4B).

thumbnail image

Figure 4. Stress-inducible protein 1 (STI1) treatment increases neurosphere formation in a prion protein-dependent manner. (A): Enzyme-linked immunosorbent assay quantification of STI1 content in conditioned medium (CM) from Prnp+/+ and Prnp0/0 neurosphere cultures (n = 3). (B): Immunoblotting of CM from neurospheres Prnp+/+ and Prnp0/0 developed with anti-STI1 and anti-GAPH (lysis control) antibodies. (C): Cells were treated with antibodies against STI1 (6 μg/ml) or rabbit preimmune IgG, and the number of neurospheres was counted (n = 3). (D): The number of neurospheres treated with increasing concentrations of recombinant STI1 (0.05–2 μM) or 8 μM of the STI1230–245 peptide. (E): Range of neurosphere diameters following exposure to STI1 (1 μM). Values represent the mean ± SE of five independent experiments. *, p < .05 (analysis of variance followed by Tukey's multiple comparison test). Abbreviations: GAPDH, glyceraldehyde 3-phosphate dehydrogenase; pep, peptide; STI1, stress-inducible protein 1.

Download figure to PowerPoint

Considering that secreted STI1 can act as an autocrine factor through binding to PrPC [10], we asked whether the disruption of this interaction with specific antibodies could impair neurosphere formation. In a neurosphere clonal assay, anti-STI1 significantly reduced the number of neurospheres in Prnp+/+ cultures, while no effect was observed in Prnp0/0 cultures (Fig. 4C). Remarkably, the number of neurospheres formed from Prnp+/+ cultures treated with an anti-STI1 was equivalent to that obtained when cultures were treated with anti-PrPC (Fig. 3E) or to that from Prnp0/0 cultures (Fig. 4C). Conversely, increasing the concentration of STI1 (0.5–2.0 μM) significantly improved the formation of Prnp+/+, but not Prnp0/0, neurospheres (Fig. 4D).

Our previous findings showed that a STI1 peptide (STI1230–245), which contains the PrPC binding site, is able to mimic the effects of full-length STI1 protein in neuritogenesis, neuroprotection, and memory consolidation [28, 29]. However, in neurosphere forming assays, the STI1230–245 peptide was unable to reproduce the effects of STI1 (Fig. 4D), suggesting that other STI1 domains are required for its proper activity on neurospheres. We further assessed whether STI1 treatment affected the size of neurospheres. As indicated in Figure 4E, STI1 treatment increased the percentage of smaller neurospheres (10–30 μm) in Prnp+/+ and Prnp0/0 populations when compared with untreated cells (Fig. 3D). Together, these results support the concept that secretable STI1, via its interaction with PrPC, is important for neurosphere formation.

STI1-PrPC Complex Formation Enhances the Proliferation, But Not the Survival, of Neural Progenitor/Stem Cells

To confirm the effects of STI1 on neural progenitor/stem cells self-renewal, we conducted cell proliferation assays using [3H]-thymidine incorporation. STI1 treatment significantly increases the ratio of [3H]-thymidine incorporation (counts per minute) per number of Prnp+/+ neurospheres (Fig. 5A). In accordance with our previous results, STI1 treatment had no effect on Prnp0/0-derived neurospheres. The positive effect of STI1-PrPC binding on the proliferation of neurospheres was also confirmed by BrdU incorporation assays (Fig. 5B).

thumbnail image

Figure 5. Stress-inducible protein 1 (STI1) promotes the proliferation of neurosphere cells. Dissociated Prnp+/+ and Prnp0/0 neurosphere cells were treated with STI1 for 16 hours. (A): Neurospheres were pulsed with [3H]-thymidine (1 mCi/ml) for 16 hours. The number of neurospheres was quantified and then harvested for radioactivity measurements. Values represent the ratio between counts per minute versus the number of neurospheres (n = 7). (B): Neurospheres were pulsed with BrdU (30 μM) in the last 2 hours of treatment and submitted to flow cytometry. Comparison between Prnp+/+ and Prnp0/0 cells treated with STI1 (0.5 or 1.0 μM; n = 3). (C): Neurospheres were dissociated and plated on poly-L-lysine coated cover slips for 24 hours in the presence of STI1, and cell death was evaluated by the expression of cleaved caspase-3 using specific antibody. Values represent the mean ± SE of four independent experiments. *, p < .05, analysis of variance followed by Tukey's multiple comparison test. Abbreviations: BrdU, bromodeoxyuridine; casp 3+, caspase-3+; STI1, stress-inducible protein 1.

Download figure to PowerPoint

As our previous studies reported the neuroprotective effect promoted by the STI1-PrPC interaction [19, 27, 29], here we addressed whether STI1-PrPC affected neurosphere survival. The presence of cleaved caspase-3 was evaluated in dissociated neurospheres treated with STI1. As depicted in Figure 5C, the activity of caspase-3 was not altered. Thus, STI1 treatment has no effect on cell survival, even at a high STI1 concentration (2 μM). These results indicate that enhancement in the number of neurospheres that results from the PrPC-STI1 association is caused by the increased proliferation of neural progenitors/stem cells instead of enhanced cell survival.

The Pattern of Neural Markers Is Not Altered by STI1

To further characterize whether the continuous exposure of neurosphere to STI1 modifies the expression pattern of neural markers, we conducted flow cytometry analysis using antibodies against specific neural markers. As shown in Figure 6A, no significant differences were found in the expression levels of neuronal, glial, or undifferentiated markers between wild-type and Prnp0/0 neurospheres. Representative histograms from Prnp+/+ and Prnp0/0 cells following STI1 treatment are shown in Figure 6B. These results demonstrate that STI1 treatment does not modify the expression pattern of committed versus uncommitted cells.

thumbnail image

Figure 6. Stress-inducible protein 1 treatment does not alter the expression profile of neural markers in neurospheres. (A): Flow cytometry analysis was performed for neural markers in Prnp+/+ and Prnp0/0 neurospheres following STI1 treatment. Values represent the mean ± SE of five independent experiments. There was no statistical difference between groups. (B): Representative histograms of Prnp+/+ and Prnp0/0 neurosphere cultures with and without (CT) STI1 treatment. Abbreviations: CT, control; GFAP, glial fibrillary acidic protein; STI1, stress-inducible protein 1; Un, unlabeled cells; βIII-Tub, βIII-tubulin.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES

In this study, we demonstrate that the interaction between PrPC and STI1 promotes neural progenitor/stem cell self-renewal. The major observations supporting this conclusion are as follows: (a) PrPC and STI1 are homogenously expressed and colocalize in neurospheres; (b) Prnp0/0 progenitor/stem cells generate relatively fewer neurospheres than do wild-type cells; (c) STI1 is secreted by neurospheres, confirming its previous characterization as an autocrine neurotrophic factor; (d) neutralizing antibodies against STI1 or PrPC impair neurosphere formation; and (e) the PrPC-STI1 interaction potentiates neurosphere proliferation, but has no effect on neural progenitor/stem cell fate or survival.

Observations that PrPC is expressed in a variety of neurons and glial cells during brain development suggest that this protein may be involved in neural stem cell-fate mechanisms [4]. High PrPC-expressing cells have been reported to regulate neural precursor proliferation in the prenatal and postnatal CNS [32]. In mouse embryonic stem cells (ESCs), recombinant PrP is able to control self-renewal capacity and differentiation status [33]. Moreover, PrPC expression is increased during spontaneous ESC differentiation, and a positive correlation between PrPC and nestin expression is observed after the induction of differentiation, suggesting that PrPC could be involved in determining neural fate [34].

Several studies have shown that PrPC is involved in neurotrophic signal transduction via interactions with a number of partners [4, 9]. At least two important transmembrane proteins with well-established neurotrophic activities have been identified as PrPC ligand, namely, neural cell adhesion molecule (NCAM) [35] and 37 kDa/67 kDa laminin receptor precursor (37LRP/67LRP) [36, 37]. NCAM plays important roles in the developing and the adult brain, including pivotal functions in neuronal differentiation and neurite outgrowth [38]. PrPC interacts directly with NCAM, leading to the stabilization of NCAM in lipid rafts and the activation of p59fyn to induce NCAM-dependent neuritogenesis [39].

PrPC can also modulate indirectly 37LRP/67LRP-dependent neurotrophic activities, including cell adhesion and signal transduction, through its binding to this receptor [40, 41]. In addition, PrPC is able to associate with ECM proteins, such as laminin and vitronectin. These interactions modulate neurite outgrowth in hippocampal neurons and axonogenesis in peripheral neurons, which may be relevant to the neurotrophic activity of PrPC [12, 13]. Interestingly, the integrins, the major ECM receptors, also regulate fundamental processes in neural stem cells, such as progenitor proliferation and survival, thus contributing to their maintenance [15, 18]. Strikingly, integrin αvβ3 activity is upregulated in PrPC-vitronectin mediated axonogenesis [13]. This interaction is the only compensatory mechanism proposed, so far, to balance the absence of PrPC in knockout mice, which in turn may explain the lack of altered phenotype in these animals [34].

Our group recently described the association of PrPC with the type I metabotropic glutamate receptor (mGluR1/5) when it is engaged with laminin [42], strengthening the idea that PrPC is able to assemble different signaling platforms, in this case coupling to metabotropic glutamate receptors, which are also involved in neural precursor differentiation and proliferation [43]. Nonetheless, the functional implications of PrPC binding to laminin as well as vitronectin, NCAM, and 37LRP/67LRP on neural progenitor/stem cell self-renewal remain to be elucidated.

One of the most studied PrPC-interacting partners is soluble STI1 [19, 44]. As the PrPC-STI1 complex was first characterized, several reports have implicated STI1 as a necessary component for the induction of PrPC-mediated neurotrophic effects [10, 19, 25, 27, 29]. STI1 can be released by cultured astrocytes and, once in the extracellular milieu, it can induce internalization of its receptor, PrPC, which modulates distinct events in brain development, such as differentiation, and protects against cellular insults [10, 25, 45]. In neurons, STI1 transduces neuroprotective signals through the activation of cyclic AMP-dependent protein kinase and also promotes extracellular signal–regulated kinase (Erk) and PI3-K/mTOR activity that is involved in neuritogenesis [29, 46]. We have also identified α7nAchR as the transmembrane protein responsible for the signal transduction of STI1-PrPC, which promotes calcium influx that is necessary to trigger neuronal survival and differentiation [30]. These receptors are expressed during development from undifferentiated precursors to fully committed neural cells [47] and participate in several neural processes such as neurite outgrowth and neuroprotection via their Ca2+ permeability [48, 49]. Given evidence showing that PrPC can organize signaling platforms, we are actively investigating the participation of α7nAchR in the signal transduction mediated by the STI1-PrPC interaction in neural progenitor/stem cell self-renewal.

Recently, the human homologue of STI1, also known as Hop (Hsp70/Hsp90 organizing protein), was shown to facilitate the phosphorylation and nuclear translocation of signal transducer and activator of transcription 3 in ESCs, implying a role for the Hsp70/Hsp90 chaperone heterocomplex machinery in pluripotency signaling [50]. In this case, it is reasonable to believe that the cytoplasmic/nuclear translocation of STI1 and its role in pluripotency may not be correlated with PrPC binding, as PrPC is localized predominantly to the cell surface [51]. Interestingly, the main partners of STI1, the chaperones Hsp70 and Hsp90, have been associated with key cellular mechanisms in neural stem/progenitor cells. Hsp70 is able to mediate neuroprotection and increase the survival of adult neuronal precursor cells following focal cerebral ischemia in mice [52]. In addition, Hsp90 is involved in the regulation of the hypoxia-driven proliferation of embryonic neural progenitor/stem cells [53].

Given the above evidence, this study assessed the role of the STI1-PrPC complex in neural stem cell biology using a well-established model of neural stem cell culture. We observed that neurospheres from wild-type and Prnp0/0 mice presented similar expression patterns and levels of GFAP, βIII-tubulin, and nestin under sphere-forming conditions (in the presence of growth factors). Using different approaches or cells types, some reports have shown significant changes in cells' differentiation status (uncommitted vs. committed) as assessed by nestin expression in wild-type and Prnp0/0 cells [32, 54]. Steele et al. [32] showed that PrPC-expressing precursors differentiate from their multipotent state (after withdrawal of bFGF) faster than those from Prnp0/0, indicating that PrPC levels are related to neural precursor differentiation. On the other hand, in our culture conditions, in which the growth factors were maintained, there were no changes in the neural marker distribution when PrPC-expressing cells were treated with STI1, suggesting that the PrPC-STI1 interaction is unable to modulate neural differentiation. Peralta et al. [54] showed that PrPC expression in ESCs is associated with differentiation from pluripotent to multipotent neural status, an event which also correlates with nestin expression. In this work, we demonstrated that nestin expression is not affected by PrPC expression in neural-committed cells. Thus, it is possible that the PrPC activity that accompanies nestin expression is restricted to differentiation from a pluripotent to a multipotent neural type. Furthermore, previous data from our laboratory have shown that cerebral cortex and hippocampus from Prnp0/0 E17 mouse brains have higher contents of nestin and lower GFAP levels than those of wild-type animals, indicating that PrPC plays a role in cellular differentiation in the developing brain [25]. In fact, these findings indicate that PrPC expression accompanies the neural differentiation process in distinct cell contexts, although its absence could be partially compensated for by other molecules, leading to a slower differentiation process.

We have also presented the evidence that PrPC participates in neural precursor self-renewal as Prnp0/0 cells generated less neurospheres than wild-type cells. Indeed, LT-HSCs from Prnp0/0 bone marrow exhibits impaired self-renewal in serial transplantation of lethally irradiated mouse recipients, which is a hallmark of LT-HSCs with respect to their capacity for reconstituting bone marrow [31]. One proposed mechanism that has been advanced to explain how PrPC could sustain self-renewal refers its property of having the ability to protect cells from apoptosis-related signals, such as oxidative stress [55], bax-mediated cell death [56–58], staurosporine-induced cell death [27, 29, 30, 59, 60], and hypoxic-ischemic insults [61–63]. Remarkably, we have demonstrated that PrPC-STI1 engagement protects postmitotic neuroblastic cells in retina explants and dissociated hippocampal neurons subjected to induced cell death [27, 29]. Importantly, in all cases, the STI1 neuroprotective effect has been mimicked by its PrPC-binding peptide, which indicates that the PrPC binding domain on STI1 is sufficient to activate the signaling pathways involved in PrPC-STI1 protective properties. Herein, the cytoprotective signals triggered by PrPC-STI1 were not observed in neural progenitor cells measured by the presence of cleaved caspase-3, even when the STI1 peptide was used, which suggests that in this case STI1 may not be involved in neuroprotection. In fact, it is well known that most viable cells die after neurosphere dissociation because the culture conditions do not support the survival of the majority of the cells [64]. Interestingly, Koyanagi et al. [65] showed that after dissociation, neural precursors cells may undergo anoikis (a form of apoptosis triggered by detachment from the surrounding ECM), which can be regulated by Rho/Rho-associated kinase signaling. Indeed, the massive postdissociation cell death observed in our experiments could also be explained by a lack of cell-cell contact, which can be crucial for cell survival prior of cell division in low-density cultures. Hence, the lack of a protective effect of PrPC-STI1 engagement in neuronal progenitors could be explained by a different cell death scenario presented here relative to the previously described protection against drug-inducible cell death.

Conversely, STI1, in a PrPC dependent manner, was able to regulate cell proliferation, which in turn promotes the enhancement of neural stem cell self-renewal. Some reports have shown participation of PrPC in cell proliferation. In the adult CNS, PrPC expression levels correlate with neural precursor proliferation rates in the subventricular zone and dentate gyrus in vivo [32]. Lee and Baskakov [33] showed that the addition of recombinant PrPC helps to maintain high proliferation activity during spontaneous differentiation of hESCs. Interestingly, we have shown that engagement of secreted STI1 with cell surface attached-PrPC promotes proliferation via PI3K and Erk1/2 pathways in glioblastoma-derived cells [66]. Moreover, the participation of PrPC in proliferation also involves the transactivation of cyclin D, an important player in G1/S progression, after activation of the PI3K/Akt pathway in gastric cancer cells, providing a link between mitogen signaling and cell cycle machinery [67]. Thus, our findings corroborate data in previous reports associating PrPC expression with cell proliferation.

Buttressed by prior findings, the current data provide additional support for the idea that STI1 can be considered a neurotrophin owing to its indubitable properties in neural progenitors/stem cell behavior in complex with PrPC during brain development. In addition, it will be important to define the role of the PrPC-STI1complex in adult neural progenitor/stem cell self-renewal, which may help build the rationale for novel therapeutic strategies for the treatment of acute brain injuries, such as ischemia and trauma, as well as for the treatment of chronic neurodegenerative illnesses, such as Alzheimer's disease and prion diseases.

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES

The current findings support the concept that PrPC and STI1 are necessary for the continued maintenance of neural stem cells in vitro. The capacity for self-renewal is vital because it allows a stem cell population to perpetuate. In addition, the stem cells must integrate proliferation control with the maintenance of an undifferentiated state. It is possible that the PrPC-STI1-dependent regulation of neural progenitor/stem cell self-renewal and proliferation also operates in vivo during normal development and/or following injury to the CNS.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES

We are thankful to Carlos F. Nascimento for technical assistance. This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP grant numbers: 07/08410-2, 09/14027-2, and 03/13189-2), Programa Institutos Nacionais de Ciência e Tecnologia, do Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq/MCT), and Ludwig Institute for Cancer Research. Fellowships from FAPESP to T.G.S., I.R.S., and B.C.-S. are gratefully acknowledged. V.R.M. is an International Scholar of the Howard Hughes Medical Institute.

DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES

The authors indicate no potential conflicts of interest.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES