Cellular prion protein (PrP) is prominently expressed in brain, in differentiated neurons but also in neural stem/precursor cells (NPCs). The misfolding of PrP is a central event in prion diseases, yet the physiological function of PrP is insufficiently understood. Although PrP has been reported to associate with the neural cell adhesion molecule (NCAM), the consequences of concerted PrP-NCAM action in NPC physiology are unknown. Here, we generated NPCs from the subventricular zone (SVZ) of postnatal day 5 wild-type and PrP null (−/−) mice and observed that PrP is essential for proper NPC proliferation and neuronal differentiation. Moreover, we found that PrP is required for the NPC response to NCAM-induced neuronal differentiation. In the absence of PrP, NCAM not only fails to promote neuronal differentiation but also induces an accumulation of doublecortin-positive neuronal progenitors at the proliferation stage. In agreement, we noted an increase in cycling neuronal progenitors in the SVZ of PrP−/− mice compared with PrP+/+ mice, as evidenced by double labeling for the proliferation marker Ki67 and doublecortin as well as by 5-bromo-2′-deoxyuridine incorporation experiments. Additionally, fewer newly born neurons were detected in the rostral migratory stream of PrP−/− mice. Analysis of the migration of SVZ cells in microexplant cultures from wild-type and PrP−/− mice revealed no differences between genotypes or a role for NCAM in this process. Our data demonstrate that PrP plays a critical role in neuronal differentiation of NPCs and suggest that this function is, at least in part, NCAM-dependent. Stem Cells2014;32:1674–1687
Cellular prion protein (PrP) is a cell surface glycoprotein prominently expressed in brain that has been studied intensively due to its direct involvement in the pathogenesis of prion diseases [1-3]. Yet, elucidation of the physiological function of PrP has lagged behind partly because PrP-null mice (PrP−/−) show no overt phenotype. Nevertheless, it is now being appreciated that PrP is a pivotal molecule with diverse roles in brain development and in neural plasticity in the adult [4-12]. These studies have demonstrated that PrP has both positive and negative effects on axon growth, while it also influences myelination, synapse formation, and long-term potentiation. PrP functions are mediated via homophilic binding  or heterophilic interactions with molecules of the extracellular matrix or with other molecules present on the cell surface of adjacent cells [14-18]. PrP also associates with the neural cell adhesion molecule (NCAM) [19, 20], which plays important roles during nervous system development and regeneration [21-23]. NCAM mediates cell–cell interactions and modulates neuronal migration, neurite extension and fasciculation, as well as synaptogenesis via cis- or trans-homophilic and heterophilic binding [24-27].
NCAM may also be important in neural stem cell biology, influencing neural stem cell proliferation, migration, and differentiation, both during development and in the adult, as it is abundantly expressed along the rostral migratory pathway extending from the subventricular zone (SVZ) to the olfactory bulb [28, 29]. The SVZ along with the hippocampal subgranular layer comprise the two main areas in brain where new neurons are continuously born throughout life from resident neural stem/precursor cells (NPCs) [30-32]. Neurogenesis is a complex process depending on the coordinate regulation of cell cycle exit and differentiation of neuronal progenitors [33-38]. Of relevance, it has been reported that NCAM inhibits proliferation and promotes neuronal differentiation of NPCs [21, 39-41]. Interestingly, PrP has been recently implicated in NPC proliferation and differentiation events [9, 42, 43]. Therefore, we hypothesized and here examined the possibility that PrP and NCAM act in conjunction to influence NPC physiology, particularly their proliferation and differentiation.
Materials and Methods
All animals were handled according to European legislation for animal rights (86/906/EEC) along with Institutional guidelines. PrP-null mice (ZrchI Prnp0/0), kindly provided by J. Collinge (MRC Prion Unit and Department of Neurodegenerative Disease, Institute of Neurology, Queen Square, London WC1N 3BG, UK), were of the original “Zurich I” background fully backcrossed onto an Friend Virus B NIH Jackson (FVB) background [44, 45]. The PrP−/− colony is maintained in the heterozygous state and homozygous PrP−/− postnatal animals are generated by intercrossing heterozygous FVB mice. Genotyping was performed by polymerase chain reaction using previously published primers .
Production of Chimeric Molecules and Coupling to Fluospheres
NCAM-Fc fusion protein comprising the extracellular part of NCAM fused with the Fc part of human IgG, the negative control MUC18-Fc chimera (MUC-Fc), and N-cadherin-Fc chimera (NCAD-Fc) were produced as described previously [46-48]. Briefly, HEK 293T cells (ATCC, Molsheim Cedex, France; http://www.lgcstandards-atcc.org/?geo_country=gr) were transiently transfected with pIG1 expression plasmids encoding NCAM-Fc or MUC-Fc or with the pCECH plasmid encoding NCAD-Fc (generously provided by Dr R. M. Mège) following the calcium phosphate precipitation method. Culture supernatants were harvested 60 hours after transfection and chimeric molecules were purified on protein A-Sepharose. For coupling to fluospheres, NCAM-Fc, MUC-Fc, or NCAD-Fc chimeras were incubated overnight at 4°C with red fluorescing microspheres (FluoSpheres Sulfate Microspheres, 1.0 µm, Red Fluorescent (580/605), Molecular Probes, Life Technologies Corporation, Carlsbad, CA; http://www.lifetechnologies.com/gr/en/home/brands/molecular-probes.html) previously conjugated with anti-human Fc . Coupled FluoSpheres were resuspended in bovine serum albumin/phosphate-buffered saline and sonicated before they were added (dilution 1:1000) to dissociated neurosphere cultures. After incubation for 1 hour at 37°C, cultures were washed three times, fixed with 4% w/v paraformaldehyde and immunolabeled with chicken polyclonal anti-nestin (1:1000; Novus Biochemicals, LLC, Littleton, CO; http://www.novusbio.com/).
Neurosphere cultures were prepared from the SVZ area of littermate postnatal day 5 (P5) PrP−/− and wild-type FVB PrP+/+ mice, as described previously [49, 50], and were maintained in a serum-free medium [1 : 1 mixture of Dulbecco's modified Eagle's medium and F-12 with penicillin (100 units/ml) and streptomycin (100 μg/ml)] containing B-27 supplement (2%; Invitrogen, Life Technologies Corporation, Carlsbad, CA; http://www.lifetechnologies.com/gr/en/home/brands/invitrogen.html), HEPES buffer (8 mM; Invitrogen), recombinant human epidermal growth factor (EGF; 20 ng/ml, R&D Systems, Minneapolis, MN; http://www.rndsystems.com/), and basic fibroblast growth factor (bFGF; 20 ng/ml, R&D Systems). After 5–7 days in culture, floating neurospheres were trypsin-dissociated and allowed to re-form spheres at least three times before further use. For proliferation studies, neurospheres were dissociated to single cells, plated onto poly-l-lysine coated coverslips in 24-well plates, at a density of 30,000 cells per well and further cultured for 3 days in the presence of EGF and bFGF. At the end of the third day, 5-bromo-2′-deoxyuridine (BrdU, 10 μM, Sigma-Aldrich, St Louis, MO; http://www.sigmaaldrich.com/european-export.html) was added for 20 minutes and immunolabeled cells were quantified. For differentiation, neurospheres were dissociated to single cells, plated onto poly-l-lysine-coated coverslips in 24-well plates at a density of 60,000–70,000 cells per well and cultured for 3 days in the absence of growth factors. For evaluation of the effect of NCAM on differentiation, NCAM-Fc fusion protein comprising the extracellular part of NCAM fused with the Fc part of human IgG was added in the culture medium at a concentration of 5 μg/ml, while MUC-Fc at an equal concentration served as negative control . MUC18 or melanoma cell adhesion molecule  is an integral membrane glycoprotein, also of the immunoglobulin superfamily, serving as negative control for NCAM. In some experiments, additional controls with NCAD-Fc and a Fc fragment (Jackson ImmunoResearch, West Grove, PA; http://www.jacksonimmuno.com/home.asp) were also included. For assessment of proliferating neuronal precursors, cells were plated onto poly-l-lysine-coated coverslips in 24-well plates at a density of 40,000 cells per well and cultured for 2 days in the presence of bFGF alone. The medium was then replaced with fresh containing bFGF and 5 μg/ml recombinant NCAM-Fc or MUC-Fc and the cells remained in culture for another 2 days.
Protein Extraction and Immunoblotting, Immunocytochemical and Immunohistochemical Procedures, SVZ Explant Migration Assay, Morphological Analysis of βIII-Tubulin-Positive NPC-Derived Neurons, and Statistical Analysis
Detailed methods are presented in Supporting Information Materials and Methods.
Proliferation and Multipotentiality of NPCs Derived from PrP−/− and Wild-Type Mice
Neurosphere cultures generated from littermate P5 knock-out (PrP−/−) or wild-type (PrP+/+) mice could be maintained for several passages in the presence of EGF and bFGF without apparent loss of self-renewal (Fig. 1A, 1B) or multipotentiality as revealed by their ability to differentiate into neurons, astrocytes, and oligodendrocytes (Fig. 1H–1W) after withdrawal of growth factors. Nevertheless, upon close observation, it was revealed that free-floating PrP−/− neurospheres derived from cells plated at clonal density were consistently smaller than PrP+/+ neurospheres (Fig. 1A–1C; mean neurosphere diameter 68.6 ± 2.9 μm in PrP−/− vs. 87.5 ± 2.8 μm in PrP+/+ cultures; n = 6, p = 0.0008). In accordance, estimation of BrdU incorporation in single-cell-dissociated cultures maintained for 3 days in the presence of EGF and bFGF, revealed an approximately 20% reduction in the percentage of BrdU+ cells in PrP−/− cultures when compared with PrP+/+ cultures (Fig. 1D–1F; 36.6 ± 1.0% vs. 46.4 ± 0.8%; n = 5, p = 0.002). We then asked whether there is a difference between genotypes in NPC maintenance and self-renewal. To address this question, neurospheres were dissociated and plated at clonal density in the presence of EGF and bFGF and after 3 days the number of secondary neurospheres was counted. Significantly fewer neurospheres were present in PrP−/− than in PrP+/+ cultures (Figs. 3A, 3B, and quantification in 3G: 445 ± 25 vs. 278±13 neurospheres in Prp+/+ vs. PrP−/− cultures, p = 0.00008).These results suggest that PrP is essential for NPC proliferation and self-renewal, in agreement with previous reports [14, 42].
Double immunofluorescence labeling showed that PrP is expressed in nestin+ NPCs in cultures maintained in the presence of growth factors (Fig. 1H–1J) as well as in TUJ1+ neurons (Fig. 1L–1N), glial fibrillary acidic protein (GFAP+) astrocytes (Fig. 1P–1R; see also Fig. 4E, 4G), and O4+ oligodendrocytes (Fig. 1T–1V) derived from NPCs upon growth factor withdrawal. The specificity of PrP staining was ascertained by absence of PrP fluorescence in cultures from PrP−/− mice (Fig. 1K, 1O, 1S, 1W). Further, PrP was detected by immunoblotting in wild-type NPCs but not in NPCs derived from PrP−/− mice (Fig. 1X). On the other hand, NCAM was present in both genotypes without appreciable differences in either proliferating cultures or in differentiating cultures maintained in the absence of EGF and bFGF (Fig. 1X; Supporting Information Fig. S1A, S1B). Notably, a marked reduction in NCAM expression was apparent in PrP−/− cells cultured in the presence of bFGF only, a condition that favours the generation of neuroblasts (Supporting Information Fig. S1C). For comparison, the expression of PrP and NCAM is shown in protein extracts from brains of young postnatal wild-type and PrP knock-out mice (Fig. 1X; Supporting Information Fig. S1D).
PrP is Essential for the Response of NPCs to NCAM-Induced Neuronal Differentiation
We next investigated the involvement of PrP in neuronal differentiation of NPCs and the potential interplay of PrP with NCAM in this process. For this purpose, we first examined the ability of NCAM-Fc to bind onto NPCs before initiation of their differentiation. We observed significant binding of NCAM-Fc coupled FluoSpheres, but not MUC18-Fc coupled FluoSpheres, onto nestin+ NPCs derived from either PrP+/+ or PrP−/− mice (Fig. 2A–2H and quantification in Fig. 2I). This indicates presence of NCAM binding sites οn NPCs from both genotypes, reflecting the ability of NCAM to associate in trans with other molecules of NCAM via homophilic interactions (see Fig. 1X for NCAM expression in NPCs) or to bind to its multiple receptors [reviewed in ] via heterophilic interactions.
We then analyzed neuronal differentiation of NPCs in single-cell dissociated cultures from PrP+/+ or PrP−/− mice, maintained for 3 days without growth factors, but in the presence of NCAM-Fc or the negative control MUC-Fc (Fig. 3A–3D). We found a small but significant reduction in the percentage of βIII-tubulin+ neurons in PrP−/− cultures when compared with PrP+/+ cultures, as estimated by immunofluorescence labeling with the TUJ1 antibody (Fig. 3I; 21.7 ± 0.2% vs. 25 ± 0.1% βIII-tubulin+ neurons; n = 7, p = 0.01). Interestingly, addition of NCAM-Fc resulted in a 23% increase in βIII-tubulin+ neurons in PrP+/+ cultures (Fig. 3A, 3B, 3I) while NPCs that lack PrP did not respond to NCAM-induced neuronal differentiation (Fig. 3C, 3D, 3I). Astrocytic differentiation, as estimated by the percentage of cells double-labeled for GFAP and S100 protein to distinguish GFAP+ precursors from differentiated double labeled cells , did not vary significantly between the two genotypes and was not affected by exposure to NCAM-Fc in either PrP+/+ or PrP−/− cultures (Fig. 3E–3H, 3M). To verify the specificity of the NCAM-Fc effect on NPC-derived βIII-tubulin+ neurons, we used two additional controls: N-cadherin-Fc chimera (NCAD-Fc;  and an Fc fragment. NCAD-Fc binds to NPCs of both genotypes (Supporting Information Fig. S2A–S2E) and its effect on the generation of NPC-derived βIII-tubulin+ neurons does not differ from that of MUC-Fc when added to PrP+/+ or PrP−/− cultures (Supporting Information Fig. S2F). Similarly, with the addition of the Fc fragment (Supporting Information Fig. S2F). Therefore, we used MUC-Fc as control in subsequent experiments.
Next, we analyzed morphological features of the βIII-tubulin+ neurons in PrP+/+ and PrP−/− cultures in the absence or presence of NCAM-Fc, as a measure of their degree of maturation. We examined the following parameters: (a) neurite outgrowth, by measuring the length of the longest neurite per neuron; (b) the total number of neurites per neuron; and (c) neurite branching, by estimating the total number of branches in all neurites per neuron. We found that neurite length did not differ significantly between genotypes or in the presence or absence of NCAM-Fc (Fig. 3J). By contrast, the total number of branches per neuron was severely affected in PrP−/− cultures when compared with PrP+/+ cultures in the presence of MUC-Fc (Fig. 3K; for PrP+/+ vs. PrP−/− 0.73 ± 0.06 vs. 0.44 ± 0.06; n = 7, p = 0.0005; see also prominent neurite branching of DCX+ cells in PrP+/+ cultures as opposed to PrP−/− cultures in Fig. 4Aii–4Dii). Nevertheless, the total number of branches was not altered within each genotype by exposure to NCAM-Fc (Fig. 3K; for PrP+/+ in the presence of MUC-Fc vs. NCAM-Fc 0.73 ± 0.06 vs. 0.8 ± 0.08; n = 7, p = 0.52 and for PrP−/− in the presence of MUC-Fc vs. NCAM-Fc 0.44 ± 0.06 vs. 0.32 ± 0.04; n = 7, p = 0.11). On the other hand, the number of total neurites per neuron was similar between genotypes, but surprisingly it was significantly reduced only in PrP−/− cultures upon exposure to NCAM-Fc (Fig. 3L; for PrP+/+ in the presence of MUC-Fc vs. NCAM-Fc 1.6 ± 0.07 vs. 1.6 ± 0.08, n = 7, p = 0.93 and for PrP−/− in the presence of MUC-Fc vs. NCAM-Fc 1.63 ± 0.08 vs.1.08 ± 0.08; n = 7, p = 0.000018).
To test whether the effects of NCAM-Fc might be consolidated in longer term cultures, we examined the morphological features of NPC-derived neurons after 10 days in differentiating conditions, following addition of NCAM-Fc at the initiation of differentiation. Again, neurite length did not differ significantly between genotypes in the absence of NCAM-Fc. However, at this time point, there was a statistically significant increase in neurite outgrowth in the presence of NCAM-Fc in PrP+/+ cultures but not in PrP−/− cultures, which remained unresponsive (Supporting Information Fig. S3A). Similarly, neurite number and branching were not different between genotypes, but PrP+/+ cultures responded to NCAM-Fc by presenting an increase in these parameters in contrast to PrP−/− cultures (Supporting Information Fig. S3B, S3C). These data indicate that PrP is essential for the response of NPCs to NCAM-induced neuronal differentiation.
To check whether exposure of NPCs to NCAM-Fc activates an intracellular MAPK cascade, we assessed by immunoblotting the phosphorylation status of Erk1/2 that is known to be activated downstream of NCAM signaling during neuronal differentiation [53, 54]. In the absence of NCAM-Fc, a higher basal phosphorylation level of Erk1/2 was noted in PrP−/− NPCs when compared with wild-type cells (Fig. 3N and quantification in Fig. 3O), in agreement with previous observations in a variety of neuronal cells [55-57]. Interestingly, an increase in Erk1/2 phosphorylation was observed in PrP+/+ cultures upon treatment with NCAM-Fc but not in PrP−/− cultures (Fig. 3N, 3O).
PrP−/− Neuronal Progenitors Accumulate at the Proliferation Stage in Response to NCAM and Do Not Proceed to Differentiation
Given the involvement of PrP in proliferation and differentiation of NPCs, we asked if PrP acts in concert, or not, with NCAM on the generation of proliferating neuroblasts following progressive removal of mitogens. To answer this question NPCs cultured in the presence of EGF/bFGF were transferred and maintained for 4 days in medium-containing bFGF alone, a condition known to allow the emergence of proliferating as well as nonproliferating doublecortin positive neuroblasts . Cells were cultured in the presence of NCAM-Fc or the negative control MUC18-Fc during the last 2 days of bFGF treatment and were then double labeled for the proliferation marker Ki67 and DCX for the detection of precursors and neuroblasts (Fig. 4A–4D). In control conditions (MUC-Fc treated cultures), the total percentage of Ki67+ proliferating cells was slightly higher in PrP+/+ cultures when compared with PrP−/− cultures (Fig. 4E; 16.2 ± 1.4% vs. 13.7 ± 1.0%, p < 0.05), while this percentage decreased in response to NCAM-Fc in PrP+/+ cultures (Fig. 4E; 16.2 ± 1.4% vs. 11.8 ± 1.9, p < 0.05), but not in PrP−/− cultures (Fig. 4E; 13.7 ± 1.0% vs. 12.4 ± 0.9, p = 0.36). Additionally, we found that in control conditions and in agreement with the βIII-tubulin data, the total fraction of DCX+ cells was significantly smaller in PrP−/− cultures when compared with PrP+/+ cultures (Fig. 4F; 7.5 ± 1% vs. 10.3 + 0.85%, n = 10, p = 0.0497). Accordingly, the percentage of DCX+ cells increased by 53% in response to NCAM-Fc in the PrP+/+ cultures (Fig. 4F; 15.7 ± 1% vs.10.3 ± 0.85%; n = 10, p = 0.00009), while it remained unaffected in the PrP−/− cultures (Fig. 4F; 7.5 ± 1% in the presence of MUC-Fc vs. 5.8 ± 1% in the presence of NCAM-Fc, respectively; n = 10, p = 0.23). This further supports an NCAM-dependent positive influence of PrP in neuronal differentiation.
We then assessed the involvement of PrP and NCAM in the generation of cycling neuronal progenitors by estimating the fraction of DCX+/Ki67+ double-labeled cells (Fig. 4G). We found that the percentage of cycling neuroblasts was severely compromised in PrP−/− cultures when compared with PrP+/+ cultures when these were maintained under control conditions, in the presence of MUC-Fc (DCX+/Ki67+ as a fraction of total cells: 0.9 ± 0.1% in PrP+/+ cultures vs. 0.4 ± 0.07% in PrP−/− cultures, n = 10, p = 0.001; and in Fig. 4G: DCX+/Ki67+ as a fraction of DCX+ cells: 9.7 ± 1.8% in PrP+/+ cultures vs. 5 ± 0.7%, in PrP−/− cultures, n = 10, p = 0.03). Interestingly, when treated with NCAM-Fc, the fraction of PrP+/+ cycling neuroblasts was drastically reduced by 62% (Fig. 4G; 9.7 ± 1.8% in MUC-Fc treated PrP+/+ cultures vs. 3.7 ± 0.47% in NCAM-Fc treated PrP+/+ cultures, n = 10, p = 0.007), supporting that the concerted action of PrP and NCAM drives neuronal progenitors to exit the cell cycle and differentiate. By contrast, in the case of PrP−/− cultures exposure to NCAM-Fc resulted in a significant increase in the percentage of DCX+/Ki67+ cycling neuroblasts, indicating that in the absence of PrP, NCAM not only fails to promote neuronal differentiation, but also induces an accumulation of neuronal progenitors at the proliferation stage (Fig 4G; 60% increase in the index of DCX+/Ki67+ cells: 5.0 ± 0.7% in MUC-Fc treated PrP−/− cultures vs. 8.0 ± 1.0% in NCAM-Fc treated PrP−/− cultures, n = 10, p = 0.04).
As documented in Supporting Information Fig. S1C, NCAM expression is compromised in PrP−/− cultures grown the presence of bFGF alone, when compared with PrP+/+ cultures. Therefore, the phenotypes observed are the combined result of PrP absence and the reduced endogenous NCAM expression.
Abnormal Generation of Cycling Neuronal Progenitors in the SVZ of Adult PrP−/− Mice
To investigate whether such an accumulation of cycling neuronal progenitors also occurs in vivo in mice lacking PrP, we examined the behavior of neuronal progenitors in the SVZ, which represents the main neurogenic region in the adult brain. To this end, we double-labeled coronal brain sections at the level of the lateral ventricle containing the SVZ with Ki67 and DCX (Fig. 5A–5C). Quantification revealed that although the total number of Ki67+ proliferating cells in the SVZ was similar between genotypes, the number of double-labeled Ki67+/DCX+ cycling neuroblasts was higher in PrP−/− mice, raising their percentage among total proliferating cells from 38.0 ± 2.5% in wild-type mice to 52.0 ± 3.3% in PrP−/− mice (Fig. 5F; number of Ki67+/DCX+ cells per section: 27 ± 2.5 in PrP−/− mice vs. 17.4 ± 2.8 in PrP+/+ mice, n = 3, p = 0.03).
Slow-proliferating neural stem cells (GFAP+) residing in the SVZ give rise to transit-amplifying neuronal progenitors (Mash1+) and cycling neuroblasts (DCX+), which in turn exit the cell cycle and generate post-mitotic neurons that migrate out of the SVZ along the rostral migratory stream (RMS) to reach the olfactory bulb where they integrate into local circuits [29, 58]. As documented in Fig. 5F, concomitant with the increase in the fraction of Ki67+/DCX+ cells in PrP−/− mice, there was a decrease in the fraction of Ki67+/DCX- cells. Therefore, we investigated whether PrP ablation also affects the properties of earlier progenitors, namely the GFAP+ neural stem cells and the Mash1+ neuronal progenitors present in the SVZ. In accordance with their low proliferation rate, very few Ki67+/GFAP+ double-labeled cells were observed in the SVZ of both genotypes (not shown), yet a reduction was noted in PrP−/− mice (number of Ki67+/GFAP+ cells per section: 2.25 ± 0.26 in PrP−/− mice vs. 4.7 ± 0.47 in PrP+/+ mice, n = 3, p = 0.01). Additionally, the total number of Mash1+ cells as well as the number of cycling Ki67+/Mash1+ cells was reduced in PrP−/− mice (Fig. 5D, 5E, 5G; number of total Mash1+ cells per section: 16.3 ± 0.75 in PrP−/− mice vs. 21.7 ± 0.82 in PrP+/+ mice, n = 3, p = 0.02; number of Ki67+/Mash1+ cells per section: 12 ± 0.93 in PrP−/− mice vs. 17.5 ± 1.16 in PrP+/+ mice, n = 3, p = 0.02;). Taken together, our results suggest the presence of fewer cycling neural stem cells and Mash1+ progenitors in the PrP−/− SVZ, but more cycling DCX+ neuroblasts presumably stalled at the proliferation stage.
Accumulation of BrdU+/DCX+ Neuroblasts in the Adult SVZ of PrP−/− Mice
To further assess the proliferation/differentiation properties of the DCX+ cells, we analyzed BrdU incorporation in SVZ cells of wild-type and PrP−/− mice, 2 and 24 hours after a single dose of BrdU injection. For this purpose, longitudinal brain sections at the level of the SVZ/RMS were double labeled with antibodies against BrdU and DCX (Fig. 5H–5N). We observed that the number of BrdU+/DCX+ neuroblasts remained constant in the SVZ of PrP+/+ mice over a period of 24 hours (Fig. 5K, 5M, 5O; at 2 hours 20.4 ± 2.1 and at 24 hours 17.1 ± 2.9 cells per section, n = 3, p = 0.42), but more than doubled in PrP−/− mice (Fig. 5L, 5N, 5O; at 2 hours 17.7 ± 2.5 and at 24 hours 38.2 ± 2.2 cells per section, n = 3, p = 0.003). These results further illustrate abnormal precursor cell proliferation in the SVZ of PrP−/− mice and reveal an accumulation over time of cycling neuroblasts. In accordance, at 24 hours less BrdU+/DCX+ cells were detected in the RMS of PrP−/− mice when compared with PrP+/+ mice (Fig. 6A–6E; 43.7 ± 2 BrdU+/DCX+ cells per section in PrP+/+ vs. 35.4 ± 1.3 in PrP−/− mice, respectively, n = 3, p = 0.03). As this result could be due to the presence of less proliferating neuroblasts in the RMS of PrP−/− mice, we also counted the numbers of BrdU+/DCX+ cells in the RMS 2 hours post-BrdU injection, but found no statistically significant differences between genotypes (25.6 ± 1.7 BrdU+/DCX+ cells per section in PrP+/+ mice compared with 23.3 ± 3.4 cells in PrP−/− mice, p = 0.59).
To test whether the observed accumulation of cycling DCX+ precursors in the SVZ of PrP−/− mice reflects a proliferation defect and/or a compromise in their migration properties, we analyzed cell migration in an in vitro system of SVZ microexplant cultures from postnatal day 5 littermate PrP−/− or PrP+/+ mice grown on Matrigel and monitored by phase-contrast microscopy [59, 60]. The assay was performed in the presence of NCAM-Fc or the negative control MUC-Fc to also determine the potential involvement of NCAM in this process. As shown in Fig. 7, SVZ microexplants obtained from wild-type mice exhibited a radial pattern of outward cell migration characterized by chain formation and expression of PSA-NCAM. Therefore, this in vitro system is a faithful proxy of the in vivo cell migration in the RMS, where newly born neurons also express PSA-NCAM and undergo chain migration [22, 23, 61]. Quantification showed no significant differences in SVZ cell migration between genotypes or in the presence/absence of NCAM-Fc over a 24-hour period (Fig. 7).
Regulation of cell cycle progression/exit and differentiation of neuronal precursors is critical for the generation of appropriate numbers of neurons at the right time and place. It is now recognized that during nervous system development, key regulators of cell cycle progression influence neural cell fate and differentiation and conversely, cell fate determinants and differentiation-inducing proteins regulate the cell cycle [reviewed in [62-65] and references therein]. In this work, we illustrated the requirement of PrP for normal proliferation and differentiation of SVZ precursors in vitro and in vivo and we were able to distinguish for the first time the differential consequences of PrP ablation in GFAP+ neural stem cells or Mash1+ transit-amplifying precursors on the one hand and DCX+ neuroblasts on the other. Moreover, we identified a critical interplay between the cellular prion protein PrP and the cell adhesion molecule NCAM which affects cell cycle progression/exit and differentiation of neural stem/precursor cells in vitro. We found that PrP is essential for proper NPC proliferation and differentiation towards a neuronal phenotype as well as for the NPC response to NCAM-induced neuronal differentiation. Further, we noted that in the absence of PrP, NCAM not only fails to promote neuronal differentiation, but also induces an accumulation of cycling neuroblasts. Taken together, our results highlight PrP together with NCAM as important regulators of neural stem/precursor cell physiology.
PrP is expressed in neural stem/progenitor cells as well as in differentiated neurons and glial cells both in vitro and in vivo in the developing and adult brain (this study and [14, 42, 66, 67]). Interestingly, previous studies have shown that PrP is present in the two main neurogenic regions of the adult brain, namely the SVZ of the lateral wall of the lateral ventricle and the subgranular layer of the hippocampus, suggesting that it may be involved in neural stem cell biology. Indeed, PrP overexpression has been reported to act positively on NPC proliferation in the embryonic and postnatal CNS . In addition there is emerging evidence that PrP may have possible implications in embryogenesis  and tissue regeneration . Notably, PrP expression is increased during spontaneous embryonic stem cell differentiation, and a positive correlation between PrP and nestin expression is observed after induction of differentiation, suggesting that PrP could be involved in neural fate decisions [43, 70].
Here we analyzed wild-type and PrP-null mice and obtained evidence that PrP is essential for proper NPC proliferation and differentiation, both in vitro and in vivo. PrP appears to influence the number of neurosphere-founder cells and the growth of their progeny in vitro, in line with in vivo data showing reduced numbers of slow proliferating neural stem cells and transit amplifying progenitors in the adult SVZ of PrP−/− mice. PrP also has a positive effect in the percentage of betaIII-tubulin+ neurons generated from NPCs in vitro, while its absence delays neurite branching. Of relevance, it has been recently reported by Loubet et al. that PrP is involved in neuritogenesis via modulation of integrin-extracellular matrix interactions . These authors have postulated that integrin aggregation caused by PrP knock-down impairs neurite formation by increasing the stability of actin microfilaments. Moreover, replication of abnormal prion aggregates has been recently shown to occur in endogenous adult NPCs resulting in their reduced neuronal differentiation , which in light of our results, may be interpreted as loss of function of prion protein in these cells during the course of prion disease.
The in vitro system of neurosphere cultures was instrumental for delineating for the first time a critical crosstalk between PrP and NCAM in NPCs. PrP and NCAM have been previously shown to associate in a complex  capable to enhance the differentiation of primary neurons in culture . Both molecules have been independently implicated in nervous system development and credited with a role in neural stem cell function [21, 42]. However, their possible cross-talk in NPCs has not been previously investigated. In this work we showed that NCAM is expressed in NPCs and demonstrated that the interplay between PrP and NCAM regulates cell cycle exit and differentiation of neuronal precursors. Interestingly, it has been previously shown that NCAM function in neuronal precursors does not involve homophilic binding or heterophilic signaling through its well-known partner, the FGF-receptor, but rather an interaction with a hitherto unknown heterophilic ligand [21, 39, 41]. According to our results, it is possible that this ligand may be PrP, although an indirect interaction between the two molecules cannot be precluded.
The in vitro data communicated here validate our hypothesis for a joint function of cellular prion protein and NCAM in NPC proliferation and differentiation. We observed that addition of recombinant NCAM-Fc in wild-type NPCs results in decrease in proliferating neuronal progenitors and a simultaneous increase in the generation of differentiated neurons. Interestingly, NPCs derived from PrP−/− mice do not respond to the NCAM-induced inhibition of proliferation or the induction of neuronal differentiation. By contrast, upon NCAM-Fc addition in PrP−/− NPCs, neuronal progenitors accumulate at the proliferation stage without proceeding to differentiation. This could be due to PrP ablation, but also to the decreased endogenous levels of NCAM observed in proliferating PrP−/− neuroblasts. In any case, our data support the idea that NCAM signaling creates a shift towards neuronal differentiation by driving neuronal progenitors out of the cell cycle to acquire a differentiated neuronal phenotype. This result is in agreement with the global inhibitory effect of NCAM on proliferation and its described role in modulating neuronal differentiation [21, 72, 73]. Upon PrP ablation, NCAM on one side fails to enhance differentiation while, at the same time, neuronal progenitors are halted at the proliferation stage, suggesting that the PrP-NCAM crosstalk is crucial for the coordinated regulation of cell cycle progression/exit and the differentiation of neuronal precursors towards a neuronal phenotype. Moreover, the generation of neurons with reduced neurite number, as an early response to NCAM in the absence of PrP, coincides with the recorded arrest of neuronal progenitors at the proliferation stage under these conditions, and further supports the notion that lack of PrP disrupts the differentiation program initiated by NCAM.
Using BrdU incorporation experiments we marked a cohort of fast proliferating neuronal progenitors in vivo in wild-type and PrP-null mice, which we followed for 2 and 24 hours, post BrdU injection. In this manner we confirmed that PrP is essential for proliferation and timely cell cycle exit and differentiation of neuronal precursors in vivo. The evidence comes from the observation that, in the absence of PrP, there is an accumulation of DCX+ cycling neuroblasts in the SVZ of PrP−/− mice. Interestingly, this finding is in line with our in vitro observations where exposure of NPCs to NCAM results in an increase in cycling DCX+ neuronal progenitors in PrP−/− cultures when compared with PrP+/+ cultures. Combined, our data suggest that NCAM may also act in concert with PrP in vivo to induce cell cycle exit and differentiation of neuronal progenitors in the SVZ. Taking into consideration that SVZ explants from PrP−/− mice show normal chain migration in vitro, it is unlikely that the accumulation of progenitors that lack PrP expression in the SVZ may be attributed to altered migratory ability per se. It is rather the consequence of abnormal proliferation versus differentiation of neuronal progenitors in the SVZ of PrP−/− mice, which may compromise their ability to leave the niche.
In the context of our study it is noteworthy that adult SVZ neurogenesis is enhanced by olfactory experience and, vice versa, neurogenesis optimizes odor discrimination and learning [74, 75]. In line, NCAM knockout mice in which postnatal and adult SVZ neurogenesis are compromised, have deficits in spontaneous odor discrimination . Interestingly, it has been reported that PrP−/− mice also show altered behavioural phenotype in two independent odor-guided tasks , confirmed by examination of three PrP knockout lines in different genetic backgrounds. It therefore appears that PrP has profound effects on structural and functional neuronal plasticity, partly mediated by its interplay with NCAM.
The data we and others have acquired [9, 14, 42, 43] suggest that PrP has a multilevel function in controlling neural stem/precursor cell proliferation and differentiation by interacting with a number of other proteins during development and in the adult. PrP therefore participates in the complex molecular networks and signaling mechanisms, both cell intrinsic and cell extrinsic that influence cell fate and differentiation. Our results highlight the cooperation between PrP and NCAM in these processes. The interaction of these two molecules in vivo and their relation to specific signaling pathways during postnatal and adult neurogenesis merits further investigation, not the least because the physiological role of PrP may provide novel insights to the neuropathology of prion diseases.
We thank Prof. John Collinge (MRC Prion Unit, University College London Institute of Neurology) for providing PrP knock-out mice and Dr. René Marc Mège (Institut Jacques Monod, Université Paris Diderot/CNRS) for NCAD-Fc plasmid. This work was supported by the E.U. FP7 REGPOT Project 264083 NEUROSIGN, the Hellenic General Secretariat for Research and Technology Grant SYNERGASIA-NOISEPLUS 09SΥΝ-21–969, the Foundation BNP Paribas and the framework action KRIPIS to the Hellenic Pasteur Institute: “Development proposals of Research Institutions” co-funded by the European Regional Development Fund and national resources. K.P. was a recipient of post-doctoral scholarship from the National Scholarships Foundation (IKY).
K.P.: conception and design, collection and assembly of data, data analysis and interpretation, manuscript writing; F.P.: experimental design, collection and assembly of data, data analysis and interpretation; T.S: provision of study material and financial support; R.M.: conception and design, financial support, data analysis and interpretation, manuscript writing.
Disclosure of Potential Conflicts of Interest
The authors indicate no potential conflicts of interest.