Requirement for Neurogenesis to Proceed through the Division of Neuronal Progenitors following Differentiation of Epidermal Growth Factor an
http://www.100md.com
《干细胞学杂志》
a Centre for Brain Repair, University of Cambridge, Forvie Site, Cambridge, United Kingdom;
b Waisman Center Stem Cell Research Program, Departments of Anatomy and Neurology, University of Wisconsin-Madison, Madison, Wisconsin, USA
Key Words. Human stem cell ? Mitogen ? Neurogenesis ? Progenitor ? Epidermal growth factor ? Fibroblast growth factor-2
Correspondence: Clive N. Svendsen, Ph.D., Waisman Center Stem Cell Research Program, Departments of Anatomy and Neurology, University of Wisconsin-Madison, 1500 Highland Avenue, Madison, Wisconsin 53705-2280, USA. Telephone: 608-265-8668; Fax: 608-265-4103; e-mail: svendsen@waisman.wisc.edu
ABSTRACT
Neural precursor cells (NPCs), encompassing both multipotent neural stem cells and more restricted progenitors, have been shown to reside within several regions of the human central nervous system (CNS). They can be isolated from human fetal tissue and induced to propagate ex vivo as aggregates of undifferentiated cells termed neurospheres . Such cells retain the potential for glial and neuronal differentiation both in vitro and after transplantation into the developing or adult CNS . Apart from providing insight into the mechanisms of neural development, an expandable source of human precursors also affords opportunities for the screening of novel CNS pharmaceuticals in relevant cell-based assay systems and raises the potential for achieving neuro-reconstruction and therapeutic gene delivery after neural transplantation. Presently, however, such applications are limited by the outstanding requirement to characterize neural precursor cultures at a biological level with respect to such features as (a) the cellular composition of neurospheres, (b) the relevance of different mitogenic growth conditions to cell-cycle regulation, proliferative profile, and differentiation potential, and (c) the mechanisms underlying stem cell–derived neurogenesis.
To isolate neural stem cells from the general neural precursor pool and other cell types found in the developing human brain, different strategies are being developed. These include fluorescence-activated cell sorting of AD-133–positive cells or genetically tagging cells expressing putative stem cell markers, such as nestin . However, although enriching the stem cell population from primary tissue is of great interest, an alternative strategy would be to select for responsive cells by continual exposure to the relevant mitogens in culture. Under these conditions, nonresponsive cells will be lost from the culture system because they will not divide. Using this approach, some researchers have suggested that long-term human neural stem cell cultures require simultaneous supplementation with both epidermal growth factor (EGF) and fibroblast growth factor-2 (FGF-2) , whereas others have indicated that FGF-2 combined with leukemia inhibitory factor (LIF) is sufficient to sustain stem cell growth . It is notable that both reports have described the mechanical dissociation of neurospheres at sequential passages. We have previously described a method for neurosphere passaging that does not require their dissociation, thereby allowing continual cell/cell contact . Using this technique, we have shown that exposure to either EGF or FGF-2 is sufficient for the growth of human neurospheres for up to 150 days—equivalent to approximately 30 population doublings . More recently, we have shown that growth of long-term EGF-responsive cells beyond this period requires the addition of LIF to the culture medium .
The generation of new cortical neurons during development is believed to occur through the proliferation of ventricular zone neuroepithelial cells, which then assume a neuroblast phenotype and migrate along radial glia to an appropriate neural layer . More recently, it has been suggested that radial glia themselves may divide asymmetrically to generate new neurons . In some regions of the adult brain, neural stem cells may have an astrocytic phenotype, perhaps originating from radial glial precursors . The relationships between stem cells present in the fetal and adult human CNS and the mechanisms by which they undergo neurogenesis have not yet been established. The availability of human neurospheres may provide a convenient system for modeling at least some aspects of neural development. We show here that after 10 weeks of growth, human neurospheres contain a relatively homogeneous population of neural precursors that respond uniformly to high concentrations of either EGF or FGF-2 by undergoing self-renewal. After mitogen withdrawal, the overwhelming majority of these cells were unable to undergo neuronal differentiation directly. However, if they were allowed to undergo cell division while migrating from the neurosphere in the absence of high mitogen concentrations, then neurogenesis could proceed through the generation of neuronal progenitors.
MATERIALS AND METHODS
EGF and FGF-2–Responsive Neurospheres Generate Similar Numbers of Neuronal Progenitors
Evidence from human studies suggests that cells responding to EGF alone appear relatively late (after the first trimester) during neural development , whereas the successful proliferation of cells isolated at earlier time points requires alternative or additional growth factors, such as FGF-2 , insulin-like growth factor-1 , or factors present in serum . The findings may reflect the late expression of the EGF-receptor and are consistent with previous rodent studies, which have demonstrated that the acquisition of EGF responsiveness by E14 forebrain progenitors can be promoted by prior exposure to FGF-2 . Accordingly, for the present studies, neurosphere cultures derived from the cortices of three different first-trimester embryos were initiated by priming the freshly isolated progenitors with a combination of both EGF and FGF-2 for 2 weeks. To investigate whether subsequent differences in the mitogenic growth conditions might affect neurosphere differentiation potential, cultures were then supplemented with either EGF, FGF-2, or EGF + FGF-2 for a further 8 weeks. After mitogen withdrawal and plating of neurospheres onto poly-L-lysine/laminin–coated coverslips, neurons and astroglia were seen to emerge from the sphere (Figs. 1A–1C). Very few cells were found to express the oligodendrocyte marker Gal-C. The proportion of cells adopting a neuronal phenotype was found to vary between the three cultures (two-way ANOVA; effect of embryo type, p < .001; Fig. 1D), which may reflect the different developmental ages of the embryos from which the cells were derived. However, the type of mitogen used was not found to significantly affect neuronal differentiation (two-way ANOVA; effect of mitogen, p > .05), and there was no significant interaction between the type of mitogen used and the embryo used (two-way ANOVA; mitogen type versus embryo, p > .05; Fig. 1D). When the data from the three embryos were combined, there was no significant difference between the neuronal proportions emerging from neurospheres grown in EGF, FGF-2, or EGF + FGF-2 (24.1 ± 5.0%, 28.5 ± 4.5%, 29.3 ± 5.3%, respectively; one-way ANOVA; effect of mitogen, p > .05; Fig. 1E). Similarly, the total number of cells emerging from the neurospheres was unaffected by the choice of mitogenic growth conditions (two-way ANOVA; effect of mitogen, p > .05; Fig. 1F).
Figure 1. EGF-responsive and FGF-2–responsive neurospheres give rise to comparable numbers of neurons. Photomicrographs show time course of emergence of differentiating neural progenitors from whole neurospheres (KO52) plated onto poly-L-lysine/laminin under serum-free conditions for between 1 and 7 days (A–C). The initial appearance of GFAP-positive cells and radiating processes (green) was soon followed by the emergence of neuronal progenitors expressing b-tubulin-III (red). Neuronal emergence was found to vary significantly among different embryo sources (two-way ANOVA; effect of embryo, F2,18 = 20.7; p < .001) (D), although there was no significant effect of the mitogenic growth condition (effect of mitogen, F2,18 = 2.2; p > .05; mitogen versus embryo, F4,18 =;0.05; p > .05). For data averaged across all three embryos, mitogenic growth condition had no significant influence either on neuronal emergence (one-wayANOVA; F2,6 = 0.3; p> .05) (E) or on the total numbers of migrating cells (two-way ANOVA; effect of mitogen, F2,12 = 0.25; p> .05; phenotype versus mitogen, F2,12 = 0.1; p> .05) (F). Data are means ± standard error of the mean for three experiments conducted for each embryo. Scale bars = 20 μm in (A–C). Abbreviations:ANOVA, analysis of variance; EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2.
Although the proportion of neuronal progenitors to emerge after plating was similar for all three growth conditions, it was still conceivable that the mitogens were having differential effects on the division rates of progenitor cells within the sphere. To test for this possibility, cortical neurospheres (KO52) grown under the different mitogenic conditions were prepulsed with BrdU (0.1 mM) for 24 hours before plating and differentiation in the absence of any mitogen. After 7 days, the proportions of BrdU-labeled cells to have migrated out from the neurosphere were 46.9 ± 6.7%, 47.1 ± 8.5%, and 51.1 ± 4.9% for neurospheres grown in EGF, FGF-2, and EGF + FGF-2, respectively (Fig. 2). The proportions of BrdU-co-labeled neurons were 6.5 ± 1.5%, 8.1 ± 1.7%, and 8.4 ± 2.5% for EGF, FGF-2, and EGF + FGF-2, respectively (Fig. 2). These proportions did not differ significantly across the three growth conditions (one-way ANOVA, p > .05). These data would collectively support the view that EGF and FGF-2 are acting in such a way that they are supporting the proliferation of the same (or a similar) population of cells having the same differentiation potentials after mitogen withdrawal.
Figure 2. EGF and FGF-2 have comparable effects on BrdU incorporation before plating. Data are for neurospheres (K052) grown in EGF + FGF-2 (14 days) and then either EGF, FGF-2, or EGF + FGF-2 (56 days). Spheres were pulsed with BrdU (0.1 μM) for 24 hours before plating and differentiation in the absence of mitogen. (A–C): ?-tubulin-III–positive neurons (red) colabeled (arrows) for BrdU (green) at 7 days after plating of neurospheres grown using the different mitogenic conditions . The proportions of (D) BrdU-labeled cells or (E) BrdU-labeled neuronal progenitors to emerge from plated neurospheres did not differ significantly between the different growth conditions . Abbreviations: BrdU, bromodeoxyuridine; EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2.
Phospho-CREB Analysis Identifies Human Neural Precursors Responding to Both EGF and FGF-2
To investigate further whether EGF and FGF-2 act on the same or distinct populations of human neural progenitors within these neurospheres, the immunocytochemical detection of phospho-CREB was used as a functional readout for cell responsiveness to the mitogens . Neurospheres (KO53) grown for 2 weeks in EGF and FGF-2 were dissociated and plated onto poly-L-lysine/laminin-coated coverslips for 24 hours. Under control conditions, fewer than 1% of the cells were immunoreactive for phospho-CREB (Fig. 3A). However, after stimulation with EGF or FGF-2 or a combination of both growth factors, more than 40% of cells expressed phospho-CREB-immunoreactivity (Figs. 3B–3D). If two separate populations of cells were present, EGF and FGF-2 combined would be expected to stimulate significantly more cells than either alone. However, there was no significant difference across the three treatment conditions (ANOVA, post hoc tests; significantly different from control cells, p < .001; Fig. 3E). This strongly suggests that EGF and FGF-2 act on the same cell population within the neurosphere.
Figure 3. Phospho-CREB analysis identifies cells responding to both EGF and FGF-2. Neurospheres (KO53) were derived from 9-week human embryonic cortex and grown in EGF + FGF-2 before dissociation and plating under mitogen-free conditions for 24 hours. Fluorescent photomicrographs show the precursors under control (unstimulated) conditions (A) and after stimulation with EGF (B), FGF-2 (C), or EGF + FGF-2 (D) for 7 minutes. Cells immunoreactive for phospho-CREB (red) are indicated under high power . Nuclei are stained with Hoechst (blue). Quantitative analysis (E) revealed similar proportions of cells responding to the three mitogenic conditions (analysis of variance, post hoc tests; significant difference from controls, p< .001). Scale bars = 15 μm (A–D), 10 μm . Abbreviations: EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2.
Cumulative BrdU Labeling and Cell-Cycle Analysis Defines a Single Population of Progenitors Responding to EGF and FGF-2
To further address the nature of the cells responding to EGF and FGF-2 after neurosphere plating, a cumulative BrdU-labeling method was undertaken for cell-cycle analysis. Neurospheres (KO53) grown for 2 weeks in EGF and FGF-2 and then for 8 weeks in either EGF, FGF-2, or a combination of both factors were plated onto poly-L-lysine/laminin-coated coverslips under serum-free conditions and in the continued presence of the same mitogen conditions. At 48 hours after plating, BrdU was added for 6, 12, 24, or 48 hours. The overwhelming majority of cells migrating out from the spheres were nestin-positive, and many incorporated BrdU over the various time periods, suggesting that these cells were undifferentiated neural precursors undergoing proliferation (Fig. 4). Graphical plots of BrdU labeling index (the fraction of BrdU-labeled cells) against time (duration of BrdU pulse) were found to be linear, suggesting that most proliferating cells comprised a single homogeneous population of progenitors with the same Tc and Ts (Fig. 5). If there had been more than one population of proliferating progenitors (with different values for Tc and Ts), then the slope of the best fit to the data points would not have been linear. A least-squares fit to the present data using a two-population model did not produce a significantly better fit than the one-population model. The linear relationship held true for each of the mitogenic conditions used for inducing cell proliferation. Indeed, there was no significant difference between the regression lines for the three growth conditions (ANOVA, p > .05), and the derived values for Ts and Tc fell within comparable ranges (Fig. 5). The data therefore suggest that under the different mitogenic conditions, the same cell population was undergoing proliferation after plating. These findings are entirely consistent with the view that EGF and FGF-2 are acting on the same cells, both within growing neurospheres and after their plating in the continual presence of mitogens.
Figure 4. Fluorescent photomicrographs showing cumulative BrdU labeling of undifferentiated human neural progenitors emerging from plated neurospheres. Neurospheres (KO53) were derived from 9-week human embryonic cortex and grown in EGF + FGF-2 (2 weeks) and then EGF, FGF-2, or EGF + FGF-2 (8 weeks) before plating onto poly-L-lysine/laminin in the presence of the following mitogens: EGF (A–D), FGF-2 (E–H), or EGF + FGF-2 (I–L). After 48 hours, plated cells were then pulsed with BrdU (0.1 μM) for 6 (A, E, I), 12 (B, F, J), 24 (C, G, K), or 48 hours (D, H, L) before fixation and staining for nestin (red) and BrdU (green). Cells in S-phase are BrdU-positive. Some dividing cells are indicated (arrows). Nuclei are stained with Hoechst (blue). For quantitative data, see Figure 5. Scale bars = 20 μm. Abbreviations: BrdU, bromodeoxyuridine; EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2.
Figure 5. Cell-cycle analysis using the cumulative BrdU-labeling method for plated human neural progenitors proliferating in response to EGF, FGF-2, or EGF + FGF-2 (see text for details). For each of the three growth conditions, plots of BrdU-labeling index (L) versus duration of BrdU pulse were found to be linear, consistent with a single population of proliferating cells. The tabulated values for Tc and Ts were derived from the gradient (m) and the L axis intercept (n), in which Tc = 1/m and Ts = n/m. Experimental errors are quoted as 95% confidence intervals. There was no significant difference between the regression lines for the different growth conditions (analysis of covariance, p> .05). Abbreviations: BrdU, bromodeoxyuridine; EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2; Tc, cell-cycle time; Ts, S-phase time.
Neuronal Progenitors Continue to Divide after Mitogen Withdrawal and Neurosphere Differentiation
We have demonstrated that approximately 50% of human neural precursors can be expected to undergo division during a given 24-hour period under the specified growth conditions (Figs. 2, 5). In this context, however, it was difficult to reconcile the observation that only 6%–8% of neurons generated from plated human neurospheres over a 7-day period could be accounted for by precursors undergoing division during the 24 hours before plating (Fig. 2). This raised the possibility that many new neurons arise from cells that continue to divide after plating in the absence of exogenous mitogens. To test for this hypothesis, groups of neurospheres (KO52) were pulsed with BrdU for sequential 24-hour periods from day 1 through 7 after plating and subsequently fixed immediately after the pulsation period. All cultures were then costained for BrdU and the neuronal marker ?-tubulin-III. For cultures fixed on days 1, 3, 5, and 7, the proportion of emerging cells seen to adopt a neuronal phenotype increased from 6.6 ± 3.2% (day 1) to 21.5 ± 4.9% (day 7), as expected (Fig. 6A). This was associated with a decrease in the overall proportion of emerging cells incorporating BrdU during the preceding 24 hours from 27.8 ± 6.6 % (day 1) to 1.1 ± 0.1% (day 7) (Fig. 6B), suggesting that overall cell division decreases with time after plating. However, based on BrdU incorporation, the proportion of new neurons born after plating increased from day 1 (2.5 ± 0.6%) to day 3 (5.5 ± 0.3%), with maximum neurogenesis being observed at approximately day 5 (6.7 ± 1.3%; one-way ANOVA; significant difference between the groups, p < .05; Fig. 6C), suggesting that there was a peak of neurogenesis occurring among the migrating cells after neurosphere plating.
Figure 6. Neurogenesis peaks at 3–4 days after plating in the absence of mitogens. (A–C): Groups of differentiating neurospheres (KO52) were pulsed with BrdU (0.1 μM) for sequential 24-hour periods and then fixed immediately. Graphs show (A) an increasing proportion of neurons and (B) a decreasing proportion of proliferating cells over a 7-day differentiation period. (C): However, many migrating neuronal progenitors were found to incorporate BrdU during the preceding 24 hours, indicating a peak of neurogenesis occurring around day 5 (one-way ANOVA; significant difference between groups, F3,12 = 7.5, p < .05). In sister cultures (D–F), groups of differentiating neurospheres were pulsed with BrdU (0.1 mM) for sequential 24-hour periods and then washed immediately with fresh medium and fixed after 7 days to allow the newly derived cells to mature. (D): Similar proportions of neurons were found on day 7 irrespective of the timing of the BrdU pulse. (E):The proportions of cells incorporating BrdU diminished from days 1–7. (F):The overwhelming majority of migrating neuronal progenitors showed a peak of neurogenesis at approximately day 3 (one-way ANOVA; significant difference between the groups, F3,12 = 14.1, p < .05). Fluorescent photomicrographs (G–I) showing cultures pulsed with BrdUfor 24 hours before day 1, 3, or 7 and fixed on day 7. ?-tubulin-III (red) and BrdU (green). Neuronal progenitors incorporating BrdU during the preceding 24 hours are indicated (arrows). Data are means ± standard error of the mean for n = 4 neurospheres per group. Scale bars = 15 μm. Abbreviations:ANOVA, analysis of variance; BrdU, bromodeoxyuridine.
Additional BrdU pulse-chase experiments, designed to capture later stages of neuronal emergence and maturation, were performed on sister neurosphere cultures. Rather than being fixed immediately after sequential 24-hour BrdU pulse periods, cultures were washed with fresh plating medium and then maintained in culture before fixation at the end of 7 days. The proportions of cells expressing a neuronal phenotype by day 7 were unaffected by the timing of the BrdU pulse, showing that there was no overt toxicity from BrdU at this concentration (0.1 mM; one-way ANOVA; p > .05; Fig. 6D). Of the cells to have emerged by day 7, the proportions incorporating BrdU from days 1 through 7 diminished from 24.2 ± 5.0% to 1.1 ± 0.2%, respectively (Fig. 6E). However, the respective proportions of neurons (seen on day 7) incorporating BrdU on days 1, 3, 5, and 7 were 23.4 ± 2.9%, 39.4 ± 7.5%, 17.6 ± 0.6%, and 2.6 ± 0.6% (one-way ANOVA; significant difference between the groups, p < .05) (Figs. 6F–6I), again indicating that there was a peak of neurogenesis occurring among migrating cells after sphere plating.
Neurogenesis Is Blocked by Cytosine-?-Arabinofuranoside or Addition of FGF-2 or EGF to Plated Neurospheres
To additionally establish that new neurons were generated from dividing cells in the absence of mitogens, we used the antimitotic drug cytosine-?-arabinofuranoside (Ara-C). Because the peak of neurogenesis among migrating progenitors was seen to occur between days 3 and 5, Ara-C (0.2–2.0 μM) was administered during this specific period. BrdU was simultaneously added to the medium from days 3 to 5 to label proliferating cells. Cultures (KO52) were fixed on day 7 for BrdU and ?-tubulin-III colabeling. At all concentrations of Ara-C, there was almost complete inhibition of cell division among the migrating neural progenitors between days 3 and 5, as indicated by the absence of total BrdU-labeled nuclei on day 7 (Figs. 7B, 7C). Similarly, it was not possible to detect any BrdU-labeled neurons on day 7 (Figs. 7B, 7D), showing that Ara-C was completely effective at preventing the division of migrating neuronal progenitors. As a consequence of inhibited cell division and neurogenesis between days 3 and 5, the absolute numbers of Hoechst-positive nuclei and ?-tubulin-III–positive neurons (Figs. 7B, 7E, 7F) seen outside the sphere on day 7 were significantly reduced by approximately 50% in the Ara-C–treated cultures (one-way ANOVA, post hoc tests; control versus Ara-C groups, p < .01). Although it is possible that exposure to Ara-C could also reduce the availability of proneural factors from dividing glia, these findings are also consistent with the view that there may be an obligatory requirement for the migrating neuronal progenitors to divide in order for neurogenesis to proceed.
Figure 7. Ara-C blocks neurogenesis after neurosphere plating. Photomicrographs show neuronal progenitors emerging from neurospheres (KO52) after plating under control conditions (A) or after treatment with Ara-C (1 μM) for 72 hours (days 3–5) (B).All cultures were simultaneously pulsed with BrdU (0.1 μM) for 72 hours. Cultures were stained for both ?-tubulin-III (red) and BrdU (green) on day 7. Arrows indicate double-labeled neurons arising from progenitors dividing on days 3–5. There was an absence of BrdU incorporation by cells treated with Ara-C at all concentrations (B–D) associated with an approximate 50% reduction in the total number of emergent cells (E) and neurons (F) . Scale bars in (A, B) = 20 μm.Abbreviations:Ara-C, cytosine-?-arabinofuranoside; BrdU, bromodeoxyuridine.
We next asked whether neurogenesis could be prevented by driving the progenitors with the mitogens EGF and FGF-2 after plating. BrdU pulse-chase experiments were conducted using neurospheres (KO52) plated in the presence of EGF, FGF-2, or EGF with FGF-2. The respective proportions of cells expressing a neuronal phenotype on day 7 were 0.7 ± 0.1%, 1.2 ± 0.5%, and 0.5 ± 0.1%, which was significantly lower than seen for spheres plated in the absence of mitogens (21.5 ± 4.9%; Fig. 8A, one-way ANOVA, post hoc tests; mitogen versus control groups, p < .001). This would indicate that the emergence of neuronal progenitors from human neurospheres is a function of mitogen withdrawal rather than neurosphere plating per se. Moreover, in the presence of mitogens, a significantly higher proportion of the migrating cells on day 7 incorporated BrdU during the preceding 24-hour period (1.1 ± 0.1%, 64 ± 2.7%, 52.7 ± 1.5%, 51.2 ± 1.0% for control, EGF, FGF-2, and EGF with FGF-2, respectively; Fig. 8B, one-way ANOVA, post hoc tests; mitogen versus control groups, p < .001), thus suggesting that the mitogens act to prevent neuronal differentiation by maintaining cells in a proliferative undifferentiated state.
Figure 8. Effects of mitogen treatments on neuronal differentiation and the proliferation of cells migrating from plated neurospheres. (A): Plated neurospheres treated with EGF, FGF-2, or EGF + FGF-2 gave rise to significantly fewer differentiating neurons among the migrating neural progenitors on day 7 (one-way analysis of variance, post hoc tests; control versus mitogen groups, F3,12 = 17.5, p < .001). (B): When mitogen-treated neurospheres were pulsed with BrdU (0.1 μM) for 24 hours on day 7, more than 50% of the migrating cells incorporated the marker. Abbreviations: BrdU, bromodeoxyuridine; EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2.
DISCUSSION
The authors gratefully acknowledge the award of fellowships from the Wellcome Trust (to T.O. and C.N.S.) and the Raymond and Beverly Sackler Trust (to T.O.) in support of these investigations.
REFERENCES
Svendsen CN, ter Borg MG, Armstrong RJE et al. A new method for the rapid and long term growth of human neural precursor cells. J Neurosci Methods 1998;85:141–152.
Svendsen CN, Caldwell MA, Ostenfeld T. Human neural stem cells: isolation, expansion and transplantation. Brain Pathol 1999;9:499–513.
Carpenter MK, Cui X, Hu ZY et al. In vitro expansion of a multipotent population of human neural progenitor cells. Exp Neurol 1999;158:265–278.
Quinn SM, Walters WM, Vescovi AL et al. Lineage restriction of neuroepithelial precursor cells from fetal human spinal cord. J Neurosci Res 1999;57:590–602.
Vescovi AL, Parati EA, Gritti A et al. Isolation and cloning of multipotential stem cells from the embryonic human CNS and establishment of transplantable human neural stem cell lines by epigenetic stimulation. Exp Neurol 1999; 156:71–83.
Palm K, Salin-Nordstrom T, Levesque MF et al. Fetal and adult human CNS stem cells have similar molecular characteristics and developmental potential. Brain Res Mol Brain Res 2000;78:192–195.
Piper DR, Mujtaba T, Keyoung H et al. Identification and characterization of neuronal precursors and their progeny from human fetal tissue. J Neurosci Res 2001;66:356–368.
Caldwell MA, He X, Wilkie N et al. Growth factors regulate the survival and fate of cells derived from human neurospheres. Nat Biotechnol 2001;19:475–479.
Ostenfeld T, Svendsen CN. Recent advances in stem cell neurobiology. Adv Tech Stand Neurosurg 2003;28:3–89.
Fricker RA, Carpenter MK, Winkler C et al. Site-specific migration and neuronal differentiation of human neural progenitor cells after transplantation in the adult rat brain. J Neurosci 1999;19:5990–6005.
Flax JD, Aurora S,Yang C et al. Engraftable human neural stem cells respond to developmental cues, replace neurons, and express foreign proteins. Nat Biotechnol 1998;16: 1033–1039.
Vescovi AL, Gritti A, Galli R et al. Isolation and intracerebral grafting of nontransformed multipotential embryonic human CNS stem cells. J Neurotrauma 1999;16:689–693.
Ostenfeld T, Caldwell MA, Prowse KR et al. Human neural precursor cells express low levels of telomerase in vitro and show diminishing cell proliferation with extensive axonal outgrowth following transplantation. Exp Neurol 2000; 164:215–226.
Englund U, Fricker-Gates RA, Lundberg C et al. Transplantation of human neural progenitor cells into the neonatal rat brain: extensive migration and differentiation with long-distance axonal projections. Exp Neurol 2002;173:1–21.
Uchida N, Buck DW, He D et al. Direct isolation of human central nervous system stem cells. Proc Natl Acad Sci U S A 2000;97:14720–14725.
Keyoung HM, Roy NS, Benraiss A et al. High-yield selection and extraction of two promoter-defined phenotypes of neural stem cells from the fetal human brain. Nat Biotechnol 2001;19:843–850.
Wright LS, Li J, Caldwell MA et al. Gene expression in human neural stem cells: effects of leukemia inhibitory factor. J Neurochem 2003;86:179–195.
Rakic P. Mode of cell migration to the superficial layers of fetal monkey neocortex. J Comp Neurol 1972;145:61–83.
Nadarajah B, Parnavelas JG. Modes of neuronal migration in the developing cerebral cortex. Nat Rev Neurosci 2002;3: 423–432.
Noctor SC, Flint AC, Weissman TA et al. Neurons derived from radial glial cells establish radial units in neocortex. Nature 2001;409:714–720.
Noctor SC, Flint AC,Weissman TA et al. Dividing precursor cells of the embryonic cortical ventricular zone have morphological and molecular characteristics of radial glia. J Neurosci 2002;22:3161–3173.
Parnavelas JG, Nadarajah B. Radial glial cells: are they really glia? Neuron 2001;31:881–884.
Tamamaki N, Nakamura K, Okamoto K et al. Radial glia is a progenitor of neocortical neurons in the developing cerebral cortex. Neurosci Res 2001;41:51–60.
Alvarez-Buylla A, Garcia-Verdugo JM, Tramontin AD. A unified hypothesis on the lineage of neural stem cells. Nat Rev Neurosci 2001;2:287–293.
Doetsch F, Caille I, Lim DA, Garcia-Verdugo JM et al. Subventricular zone astrocytes are neural stem cells in the adult mammalian brain. Cell 1999;97:703–716.
Svendsen CN, Fawcett JW, Bentlage C et al. Increased survival of rat EGF-generated CNS precursor cells using B27 supplemented medium. Exp Brain Res 1995;102:407–414.
Caldwell MA, Svendsen CN. Heparin, but not other proteoglycans potentiates the mitogenic effects of FGF-2 on mesencephalic precursor cells. Exp Neurol 1998;152:1–10.
Ostenfeld T, Joly E, Tai YT et al. Regional specification of rodent and human neurospheres. Brain Res Dev Brain Res 2002;134:43–55.
Ciccolini F, Svendsen CN. Fibroblast growth factor 2 (FGF-2) promotes acquisition of epidermal growth factor (EGF) responsiveness in mouse striatal precursor cells: identification of neural precursors responding to both EGF and FGF-2. J Neurosci 1998;18:7869–7880.
Nowakowski RS, Lewin SB, Miller MW. Bromodeoxyuridine immunohistochemical determination of the lengths of the cell cycle and the DNA-synthetic phase for an anatomically defined population. J Neurocytol 1989;18:311–318.
van Heyningen P, Calver AR, Richardson WD. Control of progenitor cell number by mitogen supply and demand. Curr Biol 2001;11:232–241.
Svendsen CN, Clarke DJ, Rosser AE et al. Survival and differentiation of rat and human epidermal growth factor-responsive precursor cells following grafting into the lesioned adult central nervous system. Exp Neurol 1996; 137:376–388.
Buc-Caron MH. Neuroepithelial progenitor cells explanted from human fetal brain proliferate and differentiate in vitro. Neurobiol Dis 1995;2:37–47.
Chalmers-Redman RM, Priestley T, Kemp JA et al. In vitro propagation and inducible differentiation of multipotential progenitor cells from human fetal brain. Neuroscience 1997;76:1121–1128.
Bjorklund A, Lindvall O. Cell replacement therapies for central nervous system disorders. Nat Neurosci 2000;3: 537–544.
Snyder EY, Vescovi AL. The possibilities/perplexities of stem cells. Nat Biotechnol 2000;18:827–828.
Bayatti N, Engele J. Cyclic AMP differentially regulates the expression of fibroblast growth factor and epidermal growth factor receptors in cultured cortical astroglia. Neuroscience 2002;114:81–89.
Burrows RC, Lillien L, Levitt P. Mechanisms of progenitor maturation are conserved in the striatum and cortex. Dev Neurosci 2000;22:7–15.
Kuhn HG, Winkler J, Kempermann G et al. Epidermal growth factor and fibroblast growth factor-2 have different effects on neural progenitors in the adult rat brain. J Neurosci 1997;17:5820–5829.
Santa-Olalla J, Covarrubias L. Basic fibroblast growth factor promotes epidermal growth factor responsiveness and survival of mesencephalic neural precursor cells. J Neurobiol 1999;40:14–27.
Gritti A, Frolichsthal-Schoeller P, Galli R et al. Epidermal and fibroblast growth factors behave as mitogenic regulators for a single multipotent stem cell-like population from the subventricular region of the adult mouse forebrain. J Neurosci 1999;19:3287–3297.
Burrows RC, Wancio D, Levitt P et al. Response diversity and the timing of progenitor cell maturation are regulated by developmental changes in EGFR expression in the cortex. Neuron 1997;19:251–267.
Chiasson BJ, Tropepe V, Morsehead CM et al. Adult mammalian forebrain ependymal and subependymal cells demonstrate proliferative potential, but only subependymal cells have neural stem cell characteristics. J Neurosci 1999;19:4462–4471.
Wang YJ, Shahrokh Z,Vemuri S et al. Characterization, stability, and formulations of basic fibroblast growth factor. Pharm Biotechnol 1996;9:141–180.
Qian X, Davis AA, Goderie SK et al. FGF2 concentration regulates the generation of neurons and glia from multipotent cortical stem cells. Neuron 1997;18:81–93.
Klagsbrun M, Baird A. A dual receptor system is required for basic fibroblast growth factor activity. Cell 1991;67: 229–231.
Caldwell MA, He X, Svendsen CN. Heparin modulates the effects of EGF on neural precursor cells. Soc Neurosci Abs 2001;132:11.
Anderson SA, Eisenstat DD, Shi L et al. Interneuron migration from basal forebrain to neocortex: dependence on Dlx genes. Science 1997;278:474–476.
Lavdas AA, Grigoriou M, Pachnis V et al. The medial ganglionic eminence gives rise to a population of early neurons in the developing cerebral cortex. J Neurosci 1999;19: 7881–7888.
Qian X, Goderie SK, Shen Q et al. Intrinsic programs pf patterned cell lineages in isolated vertebrate CNS ventricular zone cells. Development 1998;125:3143–3152.
Mayer-Proschel M, Kalyani AJ, Mujtaba T et al. Isolation of lineage-restricted neuronal precursors from multipotent neuroepithelial stem cells. Neuron 1997;19:773–785.
Lo L, Anderson DJ. Postmigratory neural crest cells expressing c-RET display restricted developmental and proliferative capacities. Neuron 1995;15:527–539.(Thor Ostenfelda, Clive N.)
b Waisman Center Stem Cell Research Program, Departments of Anatomy and Neurology, University of Wisconsin-Madison, Madison, Wisconsin, USA
Key Words. Human stem cell ? Mitogen ? Neurogenesis ? Progenitor ? Epidermal growth factor ? Fibroblast growth factor-2
Correspondence: Clive N. Svendsen, Ph.D., Waisman Center Stem Cell Research Program, Departments of Anatomy and Neurology, University of Wisconsin-Madison, 1500 Highland Avenue, Madison, Wisconsin 53705-2280, USA. Telephone: 608-265-8668; Fax: 608-265-4103; e-mail: svendsen@waisman.wisc.edu
ABSTRACT
Neural precursor cells (NPCs), encompassing both multipotent neural stem cells and more restricted progenitors, have been shown to reside within several regions of the human central nervous system (CNS). They can be isolated from human fetal tissue and induced to propagate ex vivo as aggregates of undifferentiated cells termed neurospheres . Such cells retain the potential for glial and neuronal differentiation both in vitro and after transplantation into the developing or adult CNS . Apart from providing insight into the mechanisms of neural development, an expandable source of human precursors also affords opportunities for the screening of novel CNS pharmaceuticals in relevant cell-based assay systems and raises the potential for achieving neuro-reconstruction and therapeutic gene delivery after neural transplantation. Presently, however, such applications are limited by the outstanding requirement to characterize neural precursor cultures at a biological level with respect to such features as (a) the cellular composition of neurospheres, (b) the relevance of different mitogenic growth conditions to cell-cycle regulation, proliferative profile, and differentiation potential, and (c) the mechanisms underlying stem cell–derived neurogenesis.
To isolate neural stem cells from the general neural precursor pool and other cell types found in the developing human brain, different strategies are being developed. These include fluorescence-activated cell sorting of AD-133–positive cells or genetically tagging cells expressing putative stem cell markers, such as nestin . However, although enriching the stem cell population from primary tissue is of great interest, an alternative strategy would be to select for responsive cells by continual exposure to the relevant mitogens in culture. Under these conditions, nonresponsive cells will be lost from the culture system because they will not divide. Using this approach, some researchers have suggested that long-term human neural stem cell cultures require simultaneous supplementation with both epidermal growth factor (EGF) and fibroblast growth factor-2 (FGF-2) , whereas others have indicated that FGF-2 combined with leukemia inhibitory factor (LIF) is sufficient to sustain stem cell growth . It is notable that both reports have described the mechanical dissociation of neurospheres at sequential passages. We have previously described a method for neurosphere passaging that does not require their dissociation, thereby allowing continual cell/cell contact . Using this technique, we have shown that exposure to either EGF or FGF-2 is sufficient for the growth of human neurospheres for up to 150 days—equivalent to approximately 30 population doublings . More recently, we have shown that growth of long-term EGF-responsive cells beyond this period requires the addition of LIF to the culture medium .
The generation of new cortical neurons during development is believed to occur through the proliferation of ventricular zone neuroepithelial cells, which then assume a neuroblast phenotype and migrate along radial glia to an appropriate neural layer . More recently, it has been suggested that radial glia themselves may divide asymmetrically to generate new neurons . In some regions of the adult brain, neural stem cells may have an astrocytic phenotype, perhaps originating from radial glial precursors . The relationships between stem cells present in the fetal and adult human CNS and the mechanisms by which they undergo neurogenesis have not yet been established. The availability of human neurospheres may provide a convenient system for modeling at least some aspects of neural development. We show here that after 10 weeks of growth, human neurospheres contain a relatively homogeneous population of neural precursors that respond uniformly to high concentrations of either EGF or FGF-2 by undergoing self-renewal. After mitogen withdrawal, the overwhelming majority of these cells were unable to undergo neuronal differentiation directly. However, if they were allowed to undergo cell division while migrating from the neurosphere in the absence of high mitogen concentrations, then neurogenesis could proceed through the generation of neuronal progenitors.
MATERIALS AND METHODS
EGF and FGF-2–Responsive Neurospheres Generate Similar Numbers of Neuronal Progenitors
Evidence from human studies suggests that cells responding to EGF alone appear relatively late (after the first trimester) during neural development , whereas the successful proliferation of cells isolated at earlier time points requires alternative or additional growth factors, such as FGF-2 , insulin-like growth factor-1 , or factors present in serum . The findings may reflect the late expression of the EGF-receptor and are consistent with previous rodent studies, which have demonstrated that the acquisition of EGF responsiveness by E14 forebrain progenitors can be promoted by prior exposure to FGF-2 . Accordingly, for the present studies, neurosphere cultures derived from the cortices of three different first-trimester embryos were initiated by priming the freshly isolated progenitors with a combination of both EGF and FGF-2 for 2 weeks. To investigate whether subsequent differences in the mitogenic growth conditions might affect neurosphere differentiation potential, cultures were then supplemented with either EGF, FGF-2, or EGF + FGF-2 for a further 8 weeks. After mitogen withdrawal and plating of neurospheres onto poly-L-lysine/laminin–coated coverslips, neurons and astroglia were seen to emerge from the sphere (Figs. 1A–1C). Very few cells were found to express the oligodendrocyte marker Gal-C. The proportion of cells adopting a neuronal phenotype was found to vary between the three cultures (two-way ANOVA; effect of embryo type, p < .001; Fig. 1D), which may reflect the different developmental ages of the embryos from which the cells were derived. However, the type of mitogen used was not found to significantly affect neuronal differentiation (two-way ANOVA; effect of mitogen, p > .05), and there was no significant interaction between the type of mitogen used and the embryo used (two-way ANOVA; mitogen type versus embryo, p > .05; Fig. 1D). When the data from the three embryos were combined, there was no significant difference between the neuronal proportions emerging from neurospheres grown in EGF, FGF-2, or EGF + FGF-2 (24.1 ± 5.0%, 28.5 ± 4.5%, 29.3 ± 5.3%, respectively; one-way ANOVA; effect of mitogen, p > .05; Fig. 1E). Similarly, the total number of cells emerging from the neurospheres was unaffected by the choice of mitogenic growth conditions (two-way ANOVA; effect of mitogen, p > .05; Fig. 1F).
Figure 1. EGF-responsive and FGF-2–responsive neurospheres give rise to comparable numbers of neurons. Photomicrographs show time course of emergence of differentiating neural progenitors from whole neurospheres (KO52) plated onto poly-L-lysine/laminin under serum-free conditions for between 1 and 7 days (A–C). The initial appearance of GFAP-positive cells and radiating processes (green) was soon followed by the emergence of neuronal progenitors expressing b-tubulin-III (red). Neuronal emergence was found to vary significantly among different embryo sources (two-way ANOVA; effect of embryo, F2,18 = 20.7; p < .001) (D), although there was no significant effect of the mitogenic growth condition (effect of mitogen, F2,18 = 2.2; p > .05; mitogen versus embryo, F4,18 =;0.05; p > .05). For data averaged across all three embryos, mitogenic growth condition had no significant influence either on neuronal emergence (one-wayANOVA; F2,6 = 0.3; p> .05) (E) or on the total numbers of migrating cells (two-way ANOVA; effect of mitogen, F2,12 = 0.25; p> .05; phenotype versus mitogen, F2,12 = 0.1; p> .05) (F). Data are means ± standard error of the mean for three experiments conducted for each embryo. Scale bars = 20 μm in (A–C). Abbreviations:ANOVA, analysis of variance; EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2.
Although the proportion of neuronal progenitors to emerge after plating was similar for all three growth conditions, it was still conceivable that the mitogens were having differential effects on the division rates of progenitor cells within the sphere. To test for this possibility, cortical neurospheres (KO52) grown under the different mitogenic conditions were prepulsed with BrdU (0.1 mM) for 24 hours before plating and differentiation in the absence of any mitogen. After 7 days, the proportions of BrdU-labeled cells to have migrated out from the neurosphere were 46.9 ± 6.7%, 47.1 ± 8.5%, and 51.1 ± 4.9% for neurospheres grown in EGF, FGF-2, and EGF + FGF-2, respectively (Fig. 2). The proportions of BrdU-co-labeled neurons were 6.5 ± 1.5%, 8.1 ± 1.7%, and 8.4 ± 2.5% for EGF, FGF-2, and EGF + FGF-2, respectively (Fig. 2). These proportions did not differ significantly across the three growth conditions (one-way ANOVA, p > .05). These data would collectively support the view that EGF and FGF-2 are acting in such a way that they are supporting the proliferation of the same (or a similar) population of cells having the same differentiation potentials after mitogen withdrawal.
Figure 2. EGF and FGF-2 have comparable effects on BrdU incorporation before plating. Data are for neurospheres (K052) grown in EGF + FGF-2 (14 days) and then either EGF, FGF-2, or EGF + FGF-2 (56 days). Spheres were pulsed with BrdU (0.1 μM) for 24 hours before plating and differentiation in the absence of mitogen. (A–C): ?-tubulin-III–positive neurons (red) colabeled (arrows) for BrdU (green) at 7 days after plating of neurospheres grown using the different mitogenic conditions . The proportions of (D) BrdU-labeled cells or (E) BrdU-labeled neuronal progenitors to emerge from plated neurospheres did not differ significantly between the different growth conditions . Abbreviations: BrdU, bromodeoxyuridine; EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2.
Phospho-CREB Analysis Identifies Human Neural Precursors Responding to Both EGF and FGF-2
To investigate further whether EGF and FGF-2 act on the same or distinct populations of human neural progenitors within these neurospheres, the immunocytochemical detection of phospho-CREB was used as a functional readout for cell responsiveness to the mitogens . Neurospheres (KO53) grown for 2 weeks in EGF and FGF-2 were dissociated and plated onto poly-L-lysine/laminin-coated coverslips for 24 hours. Under control conditions, fewer than 1% of the cells were immunoreactive for phospho-CREB (Fig. 3A). However, after stimulation with EGF or FGF-2 or a combination of both growth factors, more than 40% of cells expressed phospho-CREB-immunoreactivity (Figs. 3B–3D). If two separate populations of cells were present, EGF and FGF-2 combined would be expected to stimulate significantly more cells than either alone. However, there was no significant difference across the three treatment conditions (ANOVA, post hoc tests; significantly different from control cells, p < .001; Fig. 3E). This strongly suggests that EGF and FGF-2 act on the same cell population within the neurosphere.
Figure 3. Phospho-CREB analysis identifies cells responding to both EGF and FGF-2. Neurospheres (KO53) were derived from 9-week human embryonic cortex and grown in EGF + FGF-2 before dissociation and plating under mitogen-free conditions for 24 hours. Fluorescent photomicrographs show the precursors under control (unstimulated) conditions (A) and after stimulation with EGF (B), FGF-2 (C), or EGF + FGF-2 (D) for 7 minutes. Cells immunoreactive for phospho-CREB (red) are indicated under high power . Nuclei are stained with Hoechst (blue). Quantitative analysis (E) revealed similar proportions of cells responding to the three mitogenic conditions (analysis of variance, post hoc tests; significant difference from controls, p< .001). Scale bars = 15 μm (A–D), 10 μm . Abbreviations: EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2.
Cumulative BrdU Labeling and Cell-Cycle Analysis Defines a Single Population of Progenitors Responding to EGF and FGF-2
To further address the nature of the cells responding to EGF and FGF-2 after neurosphere plating, a cumulative BrdU-labeling method was undertaken for cell-cycle analysis. Neurospheres (KO53) grown for 2 weeks in EGF and FGF-2 and then for 8 weeks in either EGF, FGF-2, or a combination of both factors were plated onto poly-L-lysine/laminin-coated coverslips under serum-free conditions and in the continued presence of the same mitogen conditions. At 48 hours after plating, BrdU was added for 6, 12, 24, or 48 hours. The overwhelming majority of cells migrating out from the spheres were nestin-positive, and many incorporated BrdU over the various time periods, suggesting that these cells were undifferentiated neural precursors undergoing proliferation (Fig. 4). Graphical plots of BrdU labeling index (the fraction of BrdU-labeled cells) against time (duration of BrdU pulse) were found to be linear, suggesting that most proliferating cells comprised a single homogeneous population of progenitors with the same Tc and Ts (Fig. 5). If there had been more than one population of proliferating progenitors (with different values for Tc and Ts), then the slope of the best fit to the data points would not have been linear. A least-squares fit to the present data using a two-population model did not produce a significantly better fit than the one-population model. The linear relationship held true for each of the mitogenic conditions used for inducing cell proliferation. Indeed, there was no significant difference between the regression lines for the three growth conditions (ANOVA, p > .05), and the derived values for Ts and Tc fell within comparable ranges (Fig. 5). The data therefore suggest that under the different mitogenic conditions, the same cell population was undergoing proliferation after plating. These findings are entirely consistent with the view that EGF and FGF-2 are acting on the same cells, both within growing neurospheres and after their plating in the continual presence of mitogens.
Figure 4. Fluorescent photomicrographs showing cumulative BrdU labeling of undifferentiated human neural progenitors emerging from plated neurospheres. Neurospheres (KO53) were derived from 9-week human embryonic cortex and grown in EGF + FGF-2 (2 weeks) and then EGF, FGF-2, or EGF + FGF-2 (8 weeks) before plating onto poly-L-lysine/laminin in the presence of the following mitogens: EGF (A–D), FGF-2 (E–H), or EGF + FGF-2 (I–L). After 48 hours, plated cells were then pulsed with BrdU (0.1 μM) for 6 (A, E, I), 12 (B, F, J), 24 (C, G, K), or 48 hours (D, H, L) before fixation and staining for nestin (red) and BrdU (green). Cells in S-phase are BrdU-positive. Some dividing cells are indicated (arrows). Nuclei are stained with Hoechst (blue). For quantitative data, see Figure 5. Scale bars = 20 μm. Abbreviations: BrdU, bromodeoxyuridine; EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2.
Figure 5. Cell-cycle analysis using the cumulative BrdU-labeling method for plated human neural progenitors proliferating in response to EGF, FGF-2, or EGF + FGF-2 (see text for details). For each of the three growth conditions, plots of BrdU-labeling index (L) versus duration of BrdU pulse were found to be linear, consistent with a single population of proliferating cells. The tabulated values for Tc and Ts were derived from the gradient (m) and the L axis intercept (n), in which Tc = 1/m and Ts = n/m. Experimental errors are quoted as 95% confidence intervals. There was no significant difference between the regression lines for the different growth conditions (analysis of covariance, p> .05). Abbreviations: BrdU, bromodeoxyuridine; EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2; Tc, cell-cycle time; Ts, S-phase time.
Neuronal Progenitors Continue to Divide after Mitogen Withdrawal and Neurosphere Differentiation
We have demonstrated that approximately 50% of human neural precursors can be expected to undergo division during a given 24-hour period under the specified growth conditions (Figs. 2, 5). In this context, however, it was difficult to reconcile the observation that only 6%–8% of neurons generated from plated human neurospheres over a 7-day period could be accounted for by precursors undergoing division during the 24 hours before plating (Fig. 2). This raised the possibility that many new neurons arise from cells that continue to divide after plating in the absence of exogenous mitogens. To test for this hypothesis, groups of neurospheres (KO52) were pulsed with BrdU for sequential 24-hour periods from day 1 through 7 after plating and subsequently fixed immediately after the pulsation period. All cultures were then costained for BrdU and the neuronal marker ?-tubulin-III. For cultures fixed on days 1, 3, 5, and 7, the proportion of emerging cells seen to adopt a neuronal phenotype increased from 6.6 ± 3.2% (day 1) to 21.5 ± 4.9% (day 7), as expected (Fig. 6A). This was associated with a decrease in the overall proportion of emerging cells incorporating BrdU during the preceding 24 hours from 27.8 ± 6.6 % (day 1) to 1.1 ± 0.1% (day 7) (Fig. 6B), suggesting that overall cell division decreases with time after plating. However, based on BrdU incorporation, the proportion of new neurons born after plating increased from day 1 (2.5 ± 0.6%) to day 3 (5.5 ± 0.3%), with maximum neurogenesis being observed at approximately day 5 (6.7 ± 1.3%; one-way ANOVA; significant difference between the groups, p < .05; Fig. 6C), suggesting that there was a peak of neurogenesis occurring among the migrating cells after neurosphere plating.
Figure 6. Neurogenesis peaks at 3–4 days after plating in the absence of mitogens. (A–C): Groups of differentiating neurospheres (KO52) were pulsed with BrdU (0.1 μM) for sequential 24-hour periods and then fixed immediately. Graphs show (A) an increasing proportion of neurons and (B) a decreasing proportion of proliferating cells over a 7-day differentiation period. (C): However, many migrating neuronal progenitors were found to incorporate BrdU during the preceding 24 hours, indicating a peak of neurogenesis occurring around day 5 (one-way ANOVA; significant difference between groups, F3,12 = 7.5, p < .05). In sister cultures (D–F), groups of differentiating neurospheres were pulsed with BrdU (0.1 mM) for sequential 24-hour periods and then washed immediately with fresh medium and fixed after 7 days to allow the newly derived cells to mature. (D): Similar proportions of neurons were found on day 7 irrespective of the timing of the BrdU pulse. (E):The proportions of cells incorporating BrdU diminished from days 1–7. (F):The overwhelming majority of migrating neuronal progenitors showed a peak of neurogenesis at approximately day 3 (one-way ANOVA; significant difference between the groups, F3,12 = 14.1, p < .05). Fluorescent photomicrographs (G–I) showing cultures pulsed with BrdUfor 24 hours before day 1, 3, or 7 and fixed on day 7. ?-tubulin-III (red) and BrdU (green). Neuronal progenitors incorporating BrdU during the preceding 24 hours are indicated (arrows). Data are means ± standard error of the mean for n = 4 neurospheres per group. Scale bars = 15 μm. Abbreviations:ANOVA, analysis of variance; BrdU, bromodeoxyuridine.
Additional BrdU pulse-chase experiments, designed to capture later stages of neuronal emergence and maturation, were performed on sister neurosphere cultures. Rather than being fixed immediately after sequential 24-hour BrdU pulse periods, cultures were washed with fresh plating medium and then maintained in culture before fixation at the end of 7 days. The proportions of cells expressing a neuronal phenotype by day 7 were unaffected by the timing of the BrdU pulse, showing that there was no overt toxicity from BrdU at this concentration (0.1 mM; one-way ANOVA; p > .05; Fig. 6D). Of the cells to have emerged by day 7, the proportions incorporating BrdU from days 1 through 7 diminished from 24.2 ± 5.0% to 1.1 ± 0.2%, respectively (Fig. 6E). However, the respective proportions of neurons (seen on day 7) incorporating BrdU on days 1, 3, 5, and 7 were 23.4 ± 2.9%, 39.4 ± 7.5%, 17.6 ± 0.6%, and 2.6 ± 0.6% (one-way ANOVA; significant difference between the groups, p < .05) (Figs. 6F–6I), again indicating that there was a peak of neurogenesis occurring among migrating cells after sphere plating.
Neurogenesis Is Blocked by Cytosine-?-Arabinofuranoside or Addition of FGF-2 or EGF to Plated Neurospheres
To additionally establish that new neurons were generated from dividing cells in the absence of mitogens, we used the antimitotic drug cytosine-?-arabinofuranoside (Ara-C). Because the peak of neurogenesis among migrating progenitors was seen to occur between days 3 and 5, Ara-C (0.2–2.0 μM) was administered during this specific period. BrdU was simultaneously added to the medium from days 3 to 5 to label proliferating cells. Cultures (KO52) were fixed on day 7 for BrdU and ?-tubulin-III colabeling. At all concentrations of Ara-C, there was almost complete inhibition of cell division among the migrating neural progenitors between days 3 and 5, as indicated by the absence of total BrdU-labeled nuclei on day 7 (Figs. 7B, 7C). Similarly, it was not possible to detect any BrdU-labeled neurons on day 7 (Figs. 7B, 7D), showing that Ara-C was completely effective at preventing the division of migrating neuronal progenitors. As a consequence of inhibited cell division and neurogenesis between days 3 and 5, the absolute numbers of Hoechst-positive nuclei and ?-tubulin-III–positive neurons (Figs. 7B, 7E, 7F) seen outside the sphere on day 7 were significantly reduced by approximately 50% in the Ara-C–treated cultures (one-way ANOVA, post hoc tests; control versus Ara-C groups, p < .01). Although it is possible that exposure to Ara-C could also reduce the availability of proneural factors from dividing glia, these findings are also consistent with the view that there may be an obligatory requirement for the migrating neuronal progenitors to divide in order for neurogenesis to proceed.
Figure 7. Ara-C blocks neurogenesis after neurosphere plating. Photomicrographs show neuronal progenitors emerging from neurospheres (KO52) after plating under control conditions (A) or after treatment with Ara-C (1 μM) for 72 hours (days 3–5) (B).All cultures were simultaneously pulsed with BrdU (0.1 μM) for 72 hours. Cultures were stained for both ?-tubulin-III (red) and BrdU (green) on day 7. Arrows indicate double-labeled neurons arising from progenitors dividing on days 3–5. There was an absence of BrdU incorporation by cells treated with Ara-C at all concentrations (B–D) associated with an approximate 50% reduction in the total number of emergent cells (E) and neurons (F) . Scale bars in (A, B) = 20 μm.Abbreviations:Ara-C, cytosine-?-arabinofuranoside; BrdU, bromodeoxyuridine.
We next asked whether neurogenesis could be prevented by driving the progenitors with the mitogens EGF and FGF-2 after plating. BrdU pulse-chase experiments were conducted using neurospheres (KO52) plated in the presence of EGF, FGF-2, or EGF with FGF-2. The respective proportions of cells expressing a neuronal phenotype on day 7 were 0.7 ± 0.1%, 1.2 ± 0.5%, and 0.5 ± 0.1%, which was significantly lower than seen for spheres plated in the absence of mitogens (21.5 ± 4.9%; Fig. 8A, one-way ANOVA, post hoc tests; mitogen versus control groups, p < .001). This would indicate that the emergence of neuronal progenitors from human neurospheres is a function of mitogen withdrawal rather than neurosphere plating per se. Moreover, in the presence of mitogens, a significantly higher proportion of the migrating cells on day 7 incorporated BrdU during the preceding 24-hour period (1.1 ± 0.1%, 64 ± 2.7%, 52.7 ± 1.5%, 51.2 ± 1.0% for control, EGF, FGF-2, and EGF with FGF-2, respectively; Fig. 8B, one-way ANOVA, post hoc tests; mitogen versus control groups, p < .001), thus suggesting that the mitogens act to prevent neuronal differentiation by maintaining cells in a proliferative undifferentiated state.
Figure 8. Effects of mitogen treatments on neuronal differentiation and the proliferation of cells migrating from plated neurospheres. (A): Plated neurospheres treated with EGF, FGF-2, or EGF + FGF-2 gave rise to significantly fewer differentiating neurons among the migrating neural progenitors on day 7 (one-way analysis of variance, post hoc tests; control versus mitogen groups, F3,12 = 17.5, p < .001). (B): When mitogen-treated neurospheres were pulsed with BrdU (0.1 μM) for 24 hours on day 7, more than 50% of the migrating cells incorporated the marker. Abbreviations: BrdU, bromodeoxyuridine; EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2.
DISCUSSION
The authors gratefully acknowledge the award of fellowships from the Wellcome Trust (to T.O. and C.N.S.) and the Raymond and Beverly Sackler Trust (to T.O.) in support of these investigations.
REFERENCES
Svendsen CN, ter Borg MG, Armstrong RJE et al. A new method for the rapid and long term growth of human neural precursor cells. J Neurosci Methods 1998;85:141–152.
Svendsen CN, Caldwell MA, Ostenfeld T. Human neural stem cells: isolation, expansion and transplantation. Brain Pathol 1999;9:499–513.
Carpenter MK, Cui X, Hu ZY et al. In vitro expansion of a multipotent population of human neural progenitor cells. Exp Neurol 1999;158:265–278.
Quinn SM, Walters WM, Vescovi AL et al. Lineage restriction of neuroepithelial precursor cells from fetal human spinal cord. J Neurosci Res 1999;57:590–602.
Vescovi AL, Parati EA, Gritti A et al. Isolation and cloning of multipotential stem cells from the embryonic human CNS and establishment of transplantable human neural stem cell lines by epigenetic stimulation. Exp Neurol 1999; 156:71–83.
Palm K, Salin-Nordstrom T, Levesque MF et al. Fetal and adult human CNS stem cells have similar molecular characteristics and developmental potential. Brain Res Mol Brain Res 2000;78:192–195.
Piper DR, Mujtaba T, Keyoung H et al. Identification and characterization of neuronal precursors and their progeny from human fetal tissue. J Neurosci Res 2001;66:356–368.
Caldwell MA, He X, Wilkie N et al. Growth factors regulate the survival and fate of cells derived from human neurospheres. Nat Biotechnol 2001;19:475–479.
Ostenfeld T, Svendsen CN. Recent advances in stem cell neurobiology. Adv Tech Stand Neurosurg 2003;28:3–89.
Fricker RA, Carpenter MK, Winkler C et al. Site-specific migration and neuronal differentiation of human neural progenitor cells after transplantation in the adult rat brain. J Neurosci 1999;19:5990–6005.
Flax JD, Aurora S,Yang C et al. Engraftable human neural stem cells respond to developmental cues, replace neurons, and express foreign proteins. Nat Biotechnol 1998;16: 1033–1039.
Vescovi AL, Gritti A, Galli R et al. Isolation and intracerebral grafting of nontransformed multipotential embryonic human CNS stem cells. J Neurotrauma 1999;16:689–693.
Ostenfeld T, Caldwell MA, Prowse KR et al. Human neural precursor cells express low levels of telomerase in vitro and show diminishing cell proliferation with extensive axonal outgrowth following transplantation. Exp Neurol 2000; 164:215–226.
Englund U, Fricker-Gates RA, Lundberg C et al. Transplantation of human neural progenitor cells into the neonatal rat brain: extensive migration and differentiation with long-distance axonal projections. Exp Neurol 2002;173:1–21.
Uchida N, Buck DW, He D et al. Direct isolation of human central nervous system stem cells. Proc Natl Acad Sci U S A 2000;97:14720–14725.
Keyoung HM, Roy NS, Benraiss A et al. High-yield selection and extraction of two promoter-defined phenotypes of neural stem cells from the fetal human brain. Nat Biotechnol 2001;19:843–850.
Wright LS, Li J, Caldwell MA et al. Gene expression in human neural stem cells: effects of leukemia inhibitory factor. J Neurochem 2003;86:179–195.
Rakic P. Mode of cell migration to the superficial layers of fetal monkey neocortex. J Comp Neurol 1972;145:61–83.
Nadarajah B, Parnavelas JG. Modes of neuronal migration in the developing cerebral cortex. Nat Rev Neurosci 2002;3: 423–432.
Noctor SC, Flint AC, Weissman TA et al. Neurons derived from radial glial cells establish radial units in neocortex. Nature 2001;409:714–720.
Noctor SC, Flint AC,Weissman TA et al. Dividing precursor cells of the embryonic cortical ventricular zone have morphological and molecular characteristics of radial glia. J Neurosci 2002;22:3161–3173.
Parnavelas JG, Nadarajah B. Radial glial cells: are they really glia? Neuron 2001;31:881–884.
Tamamaki N, Nakamura K, Okamoto K et al. Radial glia is a progenitor of neocortical neurons in the developing cerebral cortex. Neurosci Res 2001;41:51–60.
Alvarez-Buylla A, Garcia-Verdugo JM, Tramontin AD. A unified hypothesis on the lineage of neural stem cells. Nat Rev Neurosci 2001;2:287–293.
Doetsch F, Caille I, Lim DA, Garcia-Verdugo JM et al. Subventricular zone astrocytes are neural stem cells in the adult mammalian brain. Cell 1999;97:703–716.
Svendsen CN, Fawcett JW, Bentlage C et al. Increased survival of rat EGF-generated CNS precursor cells using B27 supplemented medium. Exp Brain Res 1995;102:407–414.
Caldwell MA, Svendsen CN. Heparin, but not other proteoglycans potentiates the mitogenic effects of FGF-2 on mesencephalic precursor cells. Exp Neurol 1998;152:1–10.
Ostenfeld T, Joly E, Tai YT et al. Regional specification of rodent and human neurospheres. Brain Res Dev Brain Res 2002;134:43–55.
Ciccolini F, Svendsen CN. Fibroblast growth factor 2 (FGF-2) promotes acquisition of epidermal growth factor (EGF) responsiveness in mouse striatal precursor cells: identification of neural precursors responding to both EGF and FGF-2. J Neurosci 1998;18:7869–7880.
Nowakowski RS, Lewin SB, Miller MW. Bromodeoxyuridine immunohistochemical determination of the lengths of the cell cycle and the DNA-synthetic phase for an anatomically defined population. J Neurocytol 1989;18:311–318.
van Heyningen P, Calver AR, Richardson WD. Control of progenitor cell number by mitogen supply and demand. Curr Biol 2001;11:232–241.
Svendsen CN, Clarke DJ, Rosser AE et al. Survival and differentiation of rat and human epidermal growth factor-responsive precursor cells following grafting into the lesioned adult central nervous system. Exp Neurol 1996; 137:376–388.
Buc-Caron MH. Neuroepithelial progenitor cells explanted from human fetal brain proliferate and differentiate in vitro. Neurobiol Dis 1995;2:37–47.
Chalmers-Redman RM, Priestley T, Kemp JA et al. In vitro propagation and inducible differentiation of multipotential progenitor cells from human fetal brain. Neuroscience 1997;76:1121–1128.
Bjorklund A, Lindvall O. Cell replacement therapies for central nervous system disorders. Nat Neurosci 2000;3: 537–544.
Snyder EY, Vescovi AL. The possibilities/perplexities of stem cells. Nat Biotechnol 2000;18:827–828.
Bayatti N, Engele J. Cyclic AMP differentially regulates the expression of fibroblast growth factor and epidermal growth factor receptors in cultured cortical astroglia. Neuroscience 2002;114:81–89.
Burrows RC, Lillien L, Levitt P. Mechanisms of progenitor maturation are conserved in the striatum and cortex. Dev Neurosci 2000;22:7–15.
Kuhn HG, Winkler J, Kempermann G et al. Epidermal growth factor and fibroblast growth factor-2 have different effects on neural progenitors in the adult rat brain. J Neurosci 1997;17:5820–5829.
Santa-Olalla J, Covarrubias L. Basic fibroblast growth factor promotes epidermal growth factor responsiveness and survival of mesencephalic neural precursor cells. J Neurobiol 1999;40:14–27.
Gritti A, Frolichsthal-Schoeller P, Galli R et al. Epidermal and fibroblast growth factors behave as mitogenic regulators for a single multipotent stem cell-like population from the subventricular region of the adult mouse forebrain. J Neurosci 1999;19:3287–3297.
Burrows RC, Wancio D, Levitt P et al. Response diversity and the timing of progenitor cell maturation are regulated by developmental changes in EGFR expression in the cortex. Neuron 1997;19:251–267.
Chiasson BJ, Tropepe V, Morsehead CM et al. Adult mammalian forebrain ependymal and subependymal cells demonstrate proliferative potential, but only subependymal cells have neural stem cell characteristics. J Neurosci 1999;19:4462–4471.
Wang YJ, Shahrokh Z,Vemuri S et al. Characterization, stability, and formulations of basic fibroblast growth factor. Pharm Biotechnol 1996;9:141–180.
Qian X, Davis AA, Goderie SK et al. FGF2 concentration regulates the generation of neurons and glia from multipotent cortical stem cells. Neuron 1997;18:81–93.
Klagsbrun M, Baird A. A dual receptor system is required for basic fibroblast growth factor activity. Cell 1991;67: 229–231.
Caldwell MA, He X, Svendsen CN. Heparin modulates the effects of EGF on neural precursor cells. Soc Neurosci Abs 2001;132:11.
Anderson SA, Eisenstat DD, Shi L et al. Interneuron migration from basal forebrain to neocortex: dependence on Dlx genes. Science 1997;278:474–476.
Lavdas AA, Grigoriou M, Pachnis V et al. The medial ganglionic eminence gives rise to a population of early neurons in the developing cerebral cortex. J Neurosci 1999;19: 7881–7888.
Qian X, Goderie SK, Shen Q et al. Intrinsic programs pf patterned cell lineages in isolated vertebrate CNS ventricular zone cells. Development 1998;125:3143–3152.
Mayer-Proschel M, Kalyani AJ, Mujtaba T et al. Isolation of lineage-restricted neuronal precursors from multipotent neuroepithelial stem cells. Neuron 1997;19:773–785.
Lo L, Anderson DJ. Postmigratory neural crest cells expressing c-RET display restricted developmental and proliferative capacities. Neuron 1995;15:527–539.(Thor Ostenfelda, Clive N.)