GABAergic immunoreactivity is predominant in neurons derived from expanded human neural precursor cells in vitro

Available online at www.sciencedirect.com R

Experimental Neurology 182 (2003) 113–123

www.elsevier.com/locate/yexnr

GABAergic immunoreactivity is predominant in neurons derived from expanded human neural precursor cells in vitro Meena Jain,a,* Richard J.E. Armstrong,a Pamela Tyers,a Roger A. Barker,a,b,1 and Anne E. Rosserc,d,1 a

Cambridge University Centre for Brain Repair, Forvie Site, Robinson Way, Cambridge CB2 2PY, UK b Department of Neurology, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 2QQ, UK c School of Biosciences, Museum Avenue Box 911, Cardiff University, Cardiff CF10 3US, UK d Department of Neurology and Genetics, University of Wales College of Medicine, Health Park, Cardiff CF14 4XW, UK Received 8 July 2002; revised 29 October 2002; accepted 6 December 2002

Abstract Neural precursor cells have been previously isolated from the developing human nervous system and their properties studied both in vitro and in transplantation paradigms in vivo. However, their ability to differentiate into neurons of different neurochemical phenotypes remains poorly defined. In this study, the default in vitro neuronal differentiation of hENPs derived from five different regions of the human embryonic brain (cerebral cortex, striatum, cerebellum, ventral mesencephalon, and spinal cord) was studied after varying periods of time in culture. The results were directly compared to those from similarly prepared murine ENPs. hENPs prepared from all five regions showed a significant reduction in the number of neurons generated at each passage, such that by passage 4 only between 5 and 10% of cells spontaneously adopted a neuronal phenotype after differentiation in vitro. A similar observation was obtained with murine ENPs. hENPs prepared from more caudal parts of the developing neuroaxis generated fewer neurons compared to the more rostral regions. The only neuronal phenotype identified in these cultures was GABA, with 15– 60% of the neurons immunopositive for this neurotransmitter. Thus there appears to be important differences between hENPs dependent on region of origin and time in vitro under standard culture conditions, forming decreasing numbers of neurons with increasing time in culture and more caudal sites of harvest, and with the major identifiable neurotransmitter being GABA. Such characterisation is important in the process of learning how to manipulate the neuronal phenotype of these cells. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Human neural precursor cells; Differentiation; In vitro; GABAergic neurons; Neural stem cells

Introduction Neural stem and precursor cells that show the properties of self-renewal and multipotentiality, that is, the ability to generate both neuronal and glial cell types, have been isolated from the neuraxis of the embryos of a variety of species, including the mouse, rat, pig, and more recently human, and in some regions are thought to persist into adulthood (Armstrong et al., 2001; Ciccolini and Svendsen, 1998; Davis and Temple, 1994; Morshead et al., 1994; Piper et al., 2001; Reynolds et al., 1992; * Corresponding author. Fax: ⫹44-1223-331174. E-mail address: [email protected] (M. Jain). 1 Joint senior authorship.

Reynolds and Weiss, 1996; Rosser et al., 1997; Roy et al., 2000; Svendsen et al., 1997; Uchida et al., 2000; Vescovi et al., 1993). In addition cells have been isolated and compared from a variety of regions of both adult and fetal derived brains from such species (Hitoshi et al., 2002; Kukekov et al., 1999; Laywell et al., 1999; Ostenfeld et al., 2002; Quinn et al., 1999; Weiss et al., 1996). Human-derived cells are of particular interest due to their potential for use in therapeutic interventions such as neural transplantation for neurodegenerative conditions (Shihabuddin et al., 1999; Svendsen and Caldwell, 2000; Svendsen et al., 1999). It has previously been shown that precursor cells isolated from the human embryonic brain can be propagated for extended periods of time in vitro in the presence of epider-

0014-4886/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0014-4886(03)00055-4

114

M. Jain et al. / Experimental Neurology 182 (2003) 113–123

mal and fibroblast growth factors (EGF and FGF2) (Carpenter et al., 1999; Svendsen et al., 1998) and that, following mitogen withdrawal, these cells show the capacity to develop into neurons, astrocytes, and oligodendrocytes. However, the specific neuronal phenotype of neurons developing from these cells following expansion in culture is an important issue, since this would reflect the intrinsic fate of these cells. In other words, are they programmed to give rise to a particular subset of neurons depending on their region of origin (i.e., regionally specified), or are they not influenced by their region of origin once they are placed in an in vitro environment. Recent evidence suggests that neural stem cells are regionally specified as defined by expression of molecular markers in vitro (Hitoshi et al., 2002), but it remains to be resolved whether neuronal phenotypic differentiation reflects the region from which the ENPs originated or whether the cells are capable of generating a wide range of phenotypes regardless of the site from which the tissue is harvested. In addition, on a practical level, if human expanded neural precursor cells (hENPs) are to be useful for neural transplantation therapy in specific neurodegenerative disorders, then the particular neuronal phenotype differentiating from these cells is critical. Studies using rodent precursor cells have demonstrated occasional mature neurons on differentiation that display multiple neurotransmitter phenotypes in culture, such as substance P, glycine, glutamate, dopamine, and acetylcholine (Kalyani et al., 1998; Reynolds et al., 1992; Vicario-Abejon et al., 2000). However, in these studies, the vast majority of neurons that expressed any phenotypic marker were GABAergic, and other neurochemical phenotypes in vitro were very infrequent and usually only seen when the cells had not been subjected to multiple passages (Ahmed et al., 1995; Reynolds and Weiss, 1998). Studies on ENPs derived from embryonic human tissue also suggest that the principal phenotype is GABA (Carpenter et al., 1999), although this has not been systematically documented by region and time in culture. Indeed it is unclear whether the nature of neuronal differentiation from hENPs is modified by the length of time in culture and whether this is influenced by the region of origin of the ENPs. The aim of this study was to therefore explicitly investigate these aspects of embryonic human ENP behaviour, namely to delineate the default neuronal differentiation in vitro of neurons originating from such cells isolated from different regions of the human embryonic brain and followed over time in culture, and to compare this to that found in the better characterised murine system.

Materials and methods Tissue collection Human fetal tissue was collected locally from routine elective termination of pregnancy, in accordance with

guidelines established by the U.K. Department of Health, the Polkinghorne Committee, and following Local Research Ethics Committee approval. Five individual fetuses (postconceptional ages 49, 57, 68, 68 and 72 days) were collected for use in these experiments with the gestational age being determined by preoperative ultrasound. For the murine studies, time-mated CD-1 mice (n⫽3, Harlan) were purchased and embryos were taken on embryonic day 13 using standard procedures. For all experiments, dissected brain tissue derived from embryos from a single dam was pooled. All animal experiments were carried out in full compliance with the U.K. Animals (Scientific Procedures) Act 1986. Cell culture Cultures of human expanded neural precursors were prepared as described previously (Svendsen et al., 1998). Briefly, regional dissection of the embryonic brain was performed in ice-cold sterile phosphate-buffered saline (PBS, pH 7.4) supplemented with 0.6% glucose. Tissue isolated from the embryonic cerebral cortex, ganglionic eminences (embryonic striatum), ventral mesencephalon (VM), rhombencephalic lip (embryonic cerebellum), and spinal cord was incubated in 0.1% trypsin (Worthington) for 20 min at 37°C and then washed in DNAse (Sigma) and trypsin inhibitor (Sigma). Tissue was triturated to a cell suspension and the density and viability of an aliquot was checked using trypan blue exclusion with a haemocytometer. Viable cells were seeded at a density of 200,000/ml into substrate-free tissue culture flasks. The growth medium consisted of DMEM/Hams F12 (3:1, Gibco), penicillin G/streptomycin sulphate/amphotericin B (1%, Gibco), B27 (2% Gibco), human recombinant FGF-2 and EGF (both at 20 ng/ml, R&D Systems), and heparin (5 ␮g/ml, Sigma). The proliferating neural precursors were fed every 4 days by replacing half the medium with fresh. Once the resulting neurospheres had reached a size of 300 – 400 ␮m in diameter, they were passaged by chopping into quarters, as described previously (Svendsen et al., 1998). The chopped spheres were then resuspended in fresh growth medium as above, substituting B27 with N2 growth supplement (1%, Gibco). To obtain mouse-derived spheres, embryos from a pregnant dam were removed and the cerebral cortex and ganglionic eminences (embryonic striatum) were dissected and pooled. Only forebrain regions were used due to unfavourable growth profiles (M. Jain, P. Tyers, unpublished observations) and limited neuronal differentiation of ENPs derived from more caudal regions of the brain (Hitoshi et al., 2002). Cells were grown as above but they were passaged every 7 as opposed to 14 days (since murine cells proliferate faster in culture than human cells) using trypsin and mechanical dissociation. The chopping method of passaging, which has been shown to promote the long-term expansion of human spheres (Svendsen et al., 1998), was not used for

M. Jain et al. / Experimental Neurology 182 (2003) 113–123

the murine neurospheres in favour of more standard methods. In essence the cultures were exposed to 0.1% trypsin for 20 min at 37°C followed by washing with DNAse and trypsin inhibitor as above and the spheres were dissociated using a flame-polished Pasteur pipette in order to obtain a single cell suspension. For human ENPs, an aliquot of the primary cell suspension (P0) and of the chopped spheres was taken and dissociated to a single cell suspension using a trypsin digest and mechanical trituration at passages 2 (approximately 6 weeks in culture) and 4 (approximately 10 weeks in culture). Cell counts were obtained using the trypan blue exclusion method and extrapolations of total cell counts were used to calculate the expansion ratios at these passage times (expansion ratio ⫽ total number of cells at given passage time/ total number of cells initially seeded into flask). Cells were then plated onto 13-mm poly-L-lysine-coated glass coverslips in 24-well plates, at a density of 50,000 cells in 30 ␮l of differentiation medium consisting of DMEM, 2% B27, 1% PSF, and 1% fetal calf serum (Harlan). After 30 – 60 min, to allow the cells to attach to the substrate, they were then flooded with 500 ␮l of differentiation medium and placed in an incubator at 37°C with 5% CO2. Cells were fed after 3 days in culture by replacing half the medium with fresh. After 7 days, the cells were fixed with 10% formalin for 20 min followed by three washes in PBS. The whole experiment was repeated five times using tissue from separate human embryonic brain samples. In a similar way, aliquots of the primary cell suspension and of the cell suspensions obtained at passages 2 and 4 of the murine ENPs were taken, plated onto poly-L-lysine-coated glass coverslips, and cultured as described above. Experiments using murine cultures were repeated using three different litters. Immunocytochemistry Coverslips were incubated with primary antibodies directed against ␤-III- tubulin (TuJ1;monoclonal;1:500,Sigma); GFAP (polyclonal;1:500;DAKO); GABA (polyclonal; 1:500,Sigma); glutamate (polyclonal;1:500;Sigma); calbindin-D28k (polyclonal,1:500;Swiss Antibodies), tyrosine hydroxylase (TH; monoclonal; 1:400; Chemicon); and dopamine and cyclic AMP-related phosphoprotein (DARPP-32; monoclonal; 1:300;gift from Dr. Greengard). Normal goat serum (NGS;5%) diluted in PBS/0.16% Triton X-100 (PBS/T) was added to cells for 30 min to block nonspecific binding. This was removed, and primary antibody, diluted in PBS/T with 1% NGS to the appropriate concentration, was added and incubated overnight at 4°C. Following three washes in PBS, appropriate secondary antibodies (all 1:200) were then added for 2 h: biotinylated anti-mouse Ig (Serotec) for ␤-III- tubulin, TH, and DARPP-32 and FITC-conjugated anti-rabbit Ig (Chemicon) for GFAP, GABA, glutamate, and calbindin-D28k. Following three more washes in PBS, cells were incubated for a

115

further 2 h with Hoechst no. 33258 (1:5000; Sigma) and RITC- conjugated streptavidin (1:200; Serotec). Coverslips were finally washed in PBS and mounted in PBS/glycerol(1: 1). For the DARPP-32 antibody only, following incubation with the biotinylated secondary antibody and three washes with Tris-buffered saline (TBS), streptavidin– biotin-conjugated complex (1:200; Dako) was added to cells for 1 h at room temperature. This was followed by two washes with TBS and one wash with Tris nonsaline (TNS). Staining was visualized using diaminobenzidine (DAB; 0.5 mg/ml) in 0.03% of 30% hydrogen peroxide and TNS. Coverslips were washed three times in TNS following incubation in DAB for 1–5 min and were then mounted in PBS/glycerol as described above. Staining was visualized on a LeicaLeitz DMRB microscope and cell counts performed at 40⫻ magnification in 5 fields on two duplicate coverslips (total 10 fields) per region at all three different passage times. Statistical analysis The analysis of the numbers of different types of cells as a function of time in culture and region of harvest was undertaken with two-way ANOVA and that of the cumulative expansion data was with repeated-measures ANOVA, both using Genstat Release 3.2 [Lawes Agricultural Trust (Rothamsted Experimental Station), UK]. Post hoc analysis for both human and mouse tissue was performed using the Newman–Keuls multiple comparison test.

Results Expansion Cumulative expansion rates from the five different human fetuses showed the effect of region of origin on the growth kinetics of the hENPs (Fig. 1A). There were significant differences in the expansion rates of hENPs derived from different regions (region: F(4,14) ⫽ 4.11, p ⬍ 0.05), and this varied with the number of passages (region ⫻ passages: F(8,25) ⫽ 2.71, p ⬍ 0.01). Post hoc analysis revealed that differences in expansion between the regions became significant at passage 4 with cortically derived hENPs showing significantly greater expansion compared to those derived from the VM and spinal cord (p ⬍ 0.01 for both). The expansion ratios of hENPs derived from both the striatum and the cerebellum at this passage time are also higher than those for the VM and spinal cord (p ⬍ 0.05 for both), whilst the ratios for ENPs derived from the cortex, striatum, and cerebellum did not differ significantly from each other. Therefore, overall the greatest expansion is seen with more rostrally derived hENPs, in particular from the cortex, and the least with ventral mesencephalic- and spinal cord-derived hENPs, namely the more caudal regions of the neuraxis. The expansion ratios of the murine cortical and striatal

116

M. Jain et al. / Experimental Neurology 182 (2003) 113–123

regions of the developing human CNS at all time points studied. In all regions, neurons adopted a similar variety of morphologies, ranging from simple unipolar and bipolar to more complex multipolar neurons. There did not appear to be any regional specificity related to the origin of the ENPs, although this was not explicitly quantified (Fig. 3A–E). There was a significant overall decrease in the percentage of neurons produced over time in culture from hENPs, irrespective of region of origin. The region of origin did, however, affect the number of neurons derived at different passage times (region ⫻ passage: F(8,12) ⫽ 13.82, p ⱕ 0.001, Fig. 2A), and post-hoc analysis revealed that, at passage 2, the number of neurons derived from the striatum and cortex was significantly higher (P ⱕ 0.01 for striatum, p ⱕ 0.05 for cortex) than that derived from the cerebellum, VM, and spinal cord, although by passage 4 this difference was lost. Thus it appears that cells derived from the forebrain region of the cerebral cortex and striatum show a much more gradual decrease in the percentage of neurons obtained over time compared to those derived from more caudal regions of the brain, which show a much steeper decline. This effect of region of origin was not reflected in the murine ENPs (region ⫻ passage: F(2,4) ⫽ 2.32, n.s., Fig. 2C), both of which were derived from forebrain regions. This finding is in accordance with the data derived from hENPs, in which there was no significant difference between the cortically and striatally derived ENPs. Neurochemical phenotype differentiation

Fig. 1. (A) Cumulative expansion ratios of ENPs isolated from the cerebral cortex, striatum, ventral mesencephalon, cerebellum, and spinal cord of the human embryonic brain. Results are pooled from a total of five embryos aged between 49 and 72 days. (B) Cumulative expansion ratios of ENPs isolated from the cerebral cortex and striatum of the mouse embryonic brain. Results are pooled from a total of three embryonic litters aged E13. All values are expressed as the mean ⫹/⫺ SEM.

ENPs are shown in Fig. 1B and were comparable to those seen for equivalent human tissue with no significant difference between the two forebrain regions (F(1,2) ⫽ 0.811, n.s.). Neuronal differentiation The number and percentage of ␤-III-tubulin-positive neurons (as a proportion of total cells) generated from hENPs decreased significantly at each passage (F(2,7) ⫽ 71.17, p ⬍ 0.001, Fig. 2A). A similar pattern was seen in murine ENPs derived from the striatum and cortex (Fig. 2C), although it just failed to reach statistical significance (F(2,4) ⫽ 0.074, n.s.). ␤-III-Tubulin-positive neurons were generated from all

The number of GABAergic neurons derived from hENPs varied between 15 and 60% of total neurons (Fig. 2B). There was no significant effect of passage time (passage: F(2,5) ⫽ 0.963, n.s.) or region of origin (region: F(4,9) ⫽ 0.699, n.s.) on the proportion of GABAergic neurons emerging from the hENPs, although the actual number decreased with time because of the decline in the overall number of neurons with increasing passage times (Fig. 4). This pattern was also reflected in the neurons developing from murine ENPs derived from the cortex and striatum (passage: F(1,1) ⫽ 0.177, n.s.; region: F(1,2) ⫽ 0.515 n.s., Fig. 2D). Immunoreactivity for glutamate was observed in both primary and expanded cells from all the human CNS regions of interest, although interpretation of these data is difficult given that glutamate is an intermediate in the biosynthetic pathway to GABA (White, 1981) and that, in some neuronal populations, these two neurotransmitters are known to be colocalised (Hill et al., 2000; Somogyi and Llewellyn-Smith, 2001). However some inferences can be made. The average percentage of glutamatergic neurons as a proportion of total cells was 40.5⫹/⫺19.4% at P0, 32.9⫹/ ⫺5.0% at P2, and 4.7%⫹/⫺3.2% at P4, respectively, for human ENPs irrespective of region of origin. The population of GABAergic neurons as a proportion of total cells

M. Jain et al. / Experimental Neurology 182 (2003) 113–123

117

Fig. 2. (A and C) Bar graph to illustrate the percentage of ␤-III-tubulin–positive cells differentiating from hENPs and mouse ENPs, respectively, as a proportion of total cells. (B and D) Bar graph to illustrate the percentage of GABA-positive cells as a proportion of total neurons from hENPs and a murine ENPs, respectively. All results are expressed as the mean ⫹ SEM of 10 fields from five independent cultures for hENPs and from three different cultures for murine ENPs. Ctx, cortex, Str, striatum; VM, ventral mesenchephalon; Cbm, cerebellum; SC, spinal cord; P0, primary; P2, passage 2; P4, passage 4.

was consistently higher overall in the ENP populations (35.5⫹/⫺6.7% at P2 and 26.9⫹/⫺3.8% at P4) and this implies to some extent that the two populations of GABAergic and glutamatergic neurons are induced separately within the ENP populations and that the GABAergic neurochemical phenotype is the predominant one. Glutamatergic neurons were also present in neurons derived from cortical and striatal murine neurospheres, as well as being present in the primary tissue, and did not show any variation in morphology despite time in culture or region of origin (data not shown). Despite positive immunoreactivity for the DARPP-32 in parallel cultures of primary human striatal cells (Fig. 5A), no immunoreactivity was found in the expanded cells from any region. This was also true for rate limiting enzyme in dopamine synthesis, TH, which was present in parallel primary cultures of human VM (Fig. 5B) but not in expanded cells derived from this or any other region (Table 1). Calbindin-D28k was expressed in primary tissue obtained from

all regions (Fig. 5C), but positive immunoreactivity was seen only rarely in neurons from ENPs, and thus could not be quantified. Analysis of the murine neurons derived from cortical and striatal cells showed positive calbindin D28k immunoreactivity in the primary cells, although again only very occasional neurons derived from expanded cells expressed this protein and these tended to be present in the passage 2 cultures only. DARPP-32 was present in the primary striatal cells but not in expanded ones, and TH was not expressed in any cells as examined immunocytochemically (data not shown). Thus the pattern of results is very similar to that obtained for the human cells. Astrocytic differentiation Analysis of GFAP immunoreactivity from hENPs revealed a highly significant interaction between the region of

118

M. Jain et al. / Experimental Neurology 182 (2003) 113–123

M. Jain et al. / Experimental Neurology 182 (2003) 113–123

119

Table 1 Expression of neurotransmitters and calbindin-D28k in primary and expanded neural precursor cells obtained from the different regions of the human embryonic brain Stain

GABA Glutamate TH DARPP-32 Calbindin D28k

Primary

hENPs

Cortex

Striatum

VM

Cbm

Spinal cord

All regions, (P2, P4)

⫹ ⫹ ⫺ ⫺ ⫹/⫺

⫹ ⫹ ⫺ ⫹ ⫹/⫺

⫹ ⫹ ⫹ ⫺ ⫹/⫺

⫹ ⫹ ⫺ ⫺ ⫹/⫺

⫹ ⫹ ⫺ ⫺ ⫹/⫺

⫹ ⫹ ⫺ ⫺ ⫹/⫺

Note. ⫹, positive immunocytochemistry for the particular neurotransmitter; ⫺, negative immunocytochemistry; ⫹/⫺, occasional/variable positive immunocytochemistry; VM, ventral mesencephalon; Cbm, cerebellum; TH, tyrosine hydroxylase; DARPP-32, dopamine and cyclic AMP-regulated phosphoprotein; ENP, expanded neural precursor cells.

origin and the number of astrocytes obtained at each passage time (region ⫻ passage: F(8,12) ⫽ 7.17, p⫽0.001, Fig. 6B). Post hoc analysis revealed that, at each sequential passage time, the number of astrocytes derived from hENPs increased significantly regardless of region of origin (p ⬍ 0.01 in all cases, Fig. 6A), except for the VM-derived ENPs in which numbers increased by passage 2 but then did not increase further at passage 4. In addition, the number of astrocytes produced by cortical ENPs at P4 is significantly higher than those produced by any other region (p ⬍ 0.01 in all cases). However, there were no significant differences between the other regions at any of the passage times.

studies of human ENPs derived from the spinal cord (Quinn et al., 1999) and that neuronal differentiation declined when ENPs were derived from more caudal as opposed to rostral forebrain sites. Fourth, concomitant with this decrease in the proportion of neurons differentiating from ENPs, the number of astrocytes that emerge spontaneously increases significantly with time spent in culture, although this did not seem to show any rostrocaudal polarity. Oligodendrocyte differentiation was not explicitly studied but appeared to be extremely limited from the hENPs (data not shown), again consistent with other studies (Carpenter et al., 1999; Ostenfeld et al., 2002; Vescovi et al., 1999). Finally forebrainderived human ENPs appear to behave in a similar fashion to forebrain murine-derived ENPs.

Discussion Forebrain hENPs show the greatest level of expansion In this study, we have investigated the properties of ENPs derived from human tissue over time in culture to delineate their default neuronal differentiation and how this varies according to region of origin of the cells. We have shown a number of findings. First, under our culture conditions, hENPs derived from rostral areas of the brain, in particular the cortex, show greater expansion than their more caudally derived counterparts, such as the VM and spinal cord. Second, and most importantly, the predominant identified neurochemical phenotype in this study was GABA. Third, the absolute number of neurons emerging declines with time spent in culture, in agreement with earlier

The regional differences in the expansion rates may reflect the later and longer periods of neurogenesis in the rostral as opposed to the caudal regions of the brain (Bayes et al., 1993) (Fig. 4). However, ENPs derived from the cerebellum were found to have expansion rates similar to those of the striatum. Thus, there may be a critical embryonic age at which to isolate ENPs in order to obtain optimal expansion that varies according to the region of interest. Data from the murine cultures showed similar expansion rates for ENPs derived from the two forebrain regions, namely the cortex and striatum, and, in addition, we have

Fig. 3. Photomicrographs showing the morphology of ␤-III-tubulin-positive neurons differentiating from P2 human ENPs derived from the cortex (A), striatum (B), VM (C), cerebellum (D), and spinal cord (E). The neurons are similar in appearance regardless of the region of origin and typically exhibit unior bipolar morphologies, although more complex appearances are also occasionally seen from cells from all regions. Bar represents 20 ␮m. Fig. 4. Photomicrographs showing the appearance of GABA-positive cells derived from human cortical ENPs at the three time points studied: P0 (A), P2 (B), and P4 (C). Bar represents 20 ␮m. GABA-positive cells, green, ␤-III-tubulin-positive cells, red; cell nuclei, blue. Fig. 5. Photomicrographs displaying positive immunoreactivity for the neuronal phenotypic markers DARPP-32 (brown) (A), TH (red) (B) calbindin D28k (green) and ␤-III, tubulin-positive cells (red) (C) in primary (P0) human striatum, ventral mesencephalon, and cortex, respectively. No immunoreactivity for DARPP-32 or TH was found in any neurons derived from ENPs, but occasional calbindin D28k- positive cells were observed. Bar represents 20 ␮m. In all cases, cell nuclei were labeled with Hoechst (blue). Fig. 6. (A) Photomicrographs showing the proportion of neurons and astrocytes derived from cortical hENPs at the three studied times in culture. There is a clear decrease in the number of neurons and an increase in the number of astrocytes with increasing time in vitro. (B). Bar chart to illustrate the percentage of GFAP- positive cells differentiating from hENPs as a proportion of total cells. Results are expressed as the mean ⫹ SEM of 10 fields from five independent cultures. Bar represents 50 ␮m. P0, primary; P2, passage 2; P4, passage 4. ␤-III-tubulin, red; GFAP, green; Hoescht, blue.

120

M. Jain et al. / Experimental Neurology 182 (2003) 113–123

Fig. 7. Approximate times of neurogenesis for the regions of the human brain studied in this experiment. The cortex and striatum show a longer period of neurogenesis than the more caudal regions such as the VM and spinal cord, thus this may explain why ENPs from the more rostral regions show a greater ability to generate neurons for longer in culture. However neurogenesis within the developing cerebellum extends throughout developmental life from week 5 onward and into early postnatal life, and yet in culture they quickly lose their ability to generate neurons. VM, ventral mesencephalon.

made the same observation from similarly prepared rat cultures (P. Tyers, A.E. Rosser, unpublished observations). Our data also suggested that the rodent precursors showed a declining expansion rate by passage 4 compared to the murine precursors, which is entirely consistent with previous observations (Svendsen et al., 1997). Neuronal numbers from human ENPs decline with time in culture There are a number of possible explanations for the decline in neuronal number from hENPs over time in culture. One possibility is that some aspect of the passaging procedure preferentially selects for hENPs that are less committed to a neuronal lineage and favours a more glial lineage. Although this cannot be excluded, the fact that a different passaging technique was used for the murine cultures but still yielded the same result argues against this. Another possibility is that as cells continue to divide in vitro, an intrinsic developmental programme is followed in which neuronal progenitors are gradually replaced by glial progenitors in accordance with normal development in which neurons are generated ahead of glial cells (Qian et al., 1997, 2000; Williams and Price, 1995), an explanation that would be consistent with our data (see Fig. 6). Alternatively, the continued culture of hENPs in vitro for prolonged periods may alter their response characteristics so that additional signals are required to achieve neuronal differentiation. In this respect, long-term expanded hENPs have been shown to retain their ability to differentiate into neurons when exposed to the complex environmental signals present in vivo, particularly in the developing brain (Englund et al., 2002; Fricker et al., 1999). A final possibility is that the potential of ENPs may differ according to the age of the embryo from which they were obtained, such that cells

obtained from younger embryos may differentiate more readily into neurons and produce a wider range of neuronal phenotypes than those from older embryos. This study utilized ENPs from embryos of different ages and the results were pooled, so the influence of the gestational age could not be analysed. This aspect of hENP behaviour is currently being investigated, although the study is complicated by the difficulties of obtaining large amounts of accurately staged human embryonic tissue of varying ages. The decline in the number of neurons generated from ENPs in a rostrocaudal direction may relate to the fact that, embryologically, the telencephalon has a longer period of neurogenesis than the more caudal areas. This conclusion is somewhat weakened though by the fact that, whilst the Purkinje cells of the cerebellum undergo neurogenesis between weeks 5–7 of human embryonic life, the majority of interneurons begin to develop from week 19 postfertilization and into early postnatal life (Bayer et al., 1993). Therefore, cerebellar ENPs ought to show the greatest ability for neuronal differentiation at later passages (Fig. 7), which was not the case in this study. An alternative possibility is that the ENPs that reside in the rostral brain areas may retain their ability to differentiate into neurons for longer periods of time than those in caudal areas, in agreement with the recent study by Ostenfeld and colleagues (2002) involving both rodent and human ENPs. In this latter study, it was shown that cortical and striatal ENPs generated significantly more neurons than the mesencephalon, thalamus, and cerebellum, which also tended to yield smaller neurons than their more caudal counterparts. Whilst these differences exist in the ability of different region to yield neurons from hENPs, this ceases to be significant at passage 4. This may relate to the selection of a common multipotent “stem” cell, which is not seen at earlier times in culture because more regionally specified committed progenitors dominate the cell population at these

M. Jain et al. / Experimental Neurology 182 (2003) 113–123

times. However, as longer passage times were not studied, this interpretation remains conjectural.

121

However, in neither study was the effect of time in culture on the extent of neuronal development nor on the type of neuronal phenotypic differentiation investigated.

The predominant immunochemical phenotype in hENP-derived neurons is GABAergic

Concluding remarks

The propensity for hENP-derived neurons to adopt a GABA phenotype is in agreement with the study by Carpenter et al. (1999) and is not surprising given this phenotype is most numerous in the mature CNS. The proportion of these neurons did not change over time in culture but it should be stressed that the majority of neurons derived from hENPs after 7 days in culture did not express GABA or any other specific neurotransmitter. This may relate to the limited repertoire of markers used in our study or may be an indication that the majority of the neurons remained immature under these conditions, although previously it has been shown that immature neurons within the developing brain are capable of expressing a range of neurotransmitter phenotypes (Bayer et al., 1993; Konig et al., 1988; Schambra et al., 1989; Wallace and Lauder, 1983). The predominance of GABA and the lack of any other neurotransmitter may suggest that GABA is the default pathway for neurons derived from hENPs, at least in those that are capable of full differentiation. However, the in vitro environment is vastly different from the in vivo situation and does not provide the necessary extrinsic cues in order to allow ENPs to generate the wide range of neurochemical phenotyes that is needed during development. Additional instructive signals are needed for this to occur, as has been shown in a number of studies. For example, VM-derived precursors have previously been shown to generate dopaminergic neuroblasts but these stop dividing after a short period of time in vitro (Bouvier and Mytilineou, 1995; Caldwell and Svendsen, 1998) and require some external stimulus in order to promote the production of tyrosine hydroxylase-positive neurons. In vivo studies lend support to this in that it has been shown that a proportion of hENPs are capable of responding appropriately to developmental cues on transplantation into the developing brain (Brustle et al., 1998; Englund et al., 2002; Flax et al., 1998; Rosser et al., 2000). For example, human cells derived from an 8-week-old fetus and propagated with EGF and FGF-2 transplanted into the embryonic rat telencephalic ventricle displayed neuronal, astrocytic, and oligodrendoglial differentiation. Neurons derived from such cells displayed morphologies appropriate to the site of integration, although no expression of neurochemical phenotype was reported (Brustle et al., 1998). Similarly a report by Rosser et al., (2000) showed appropriate morphological, but not neurochemical neuronal differentiation when human cell populations were implanted into different regions of the neonatal rat brain. In both these studies, it is possible that phenotypic neuronal markers may have developed if survival times had been longer, given the prolonged developmental time course of human cells when grafted (Belkadi et al., 1997).

The present study highlights the fact that the derivation of a range of neuronal phenotypes from hENPs of clinical use remains a major challenge, and manipulations of the standard culture system will be necessary to generate more neurons and still further to generate specific neuronal phenotypes. Enhancement of neuronal number has been achieved in some recent studies using both genetic and epigenetic approaches. For example, Caldwell and colleagues (2001) have used neurotrophic factors 3 and 4 (NT3 and 4) and platelet derived growth factor to significantly increase the numbers of neurons derived from human ENPs, from 8 to ⬎60%, although the predominant phenotype was again GABA. An understanding of molecular factors, such as Notch signaling, the HLH family of transcription factors and, the Shc family of adaptor proteins, which have all been shown to play a role in neuronal differentiation, may also prove to be useful in this regard (Bartlett et al., 1998; Cattaneo and Pelicci, 1998; Conti et al., 2001; Kabos et al., 2002; Kageyama and Ohtsuka, 1999; Nakamura et al., 2000). However, further work is required in order to manipulate ENPs to generate more significant numbers of specific neuronal phenotypes before these cells could be considered for use in a clinical transplantation programme. Finally, given the recent evidence that astrocytes derived from particular areas of the brain, such as the hippocampus, may be able to instruct neuronal stem cells to develop a neuronal phenotype (Song et al., 2002), it is possible that the regional specificity that we, and others, have observed in populations of ENPs to some degree is controlled by regionally specific astrocytes. Thus the possibility remains that exposure of ENPs from one particular region to astrocytes from another may be able to influence phenotypic differentiation. In summary, this study has demonstrated that neuronal differentiation from human ENPs significantly decreases over time and in a rostrocaudal gradient and that the predominant neurochemical phenotype identified immunohistochemically is GABA under standard culture conditions. The findings presented here provide further evidence that ENPs derived from different areas of the brain retain some regional specification, as demonstrated by the differences in expansion rates and the numbers of neurons generated, but that neuronal phenotypic differentiation remains relatively consistent regardless of region of origin. In this respect, this is the first study that has systematically assessed the default neuronal differentiation of human ENPs over time in culture from different regions of the brain and as such provides a basis for further research into the manipulation of these cells, with the eventual aim to provide clinically useful phenotypes.

122

M. Jain et al. / Experimental Neurology 182 (2003) 113–123

Acknowledgments The authors thank Dr. P. Greengard for the DARPP-32 antibody and Dr. S. Elneil for providing the human tissue. This work was supported by Merck, Sharp, and Dohme and the Parkinson’s Disease Society. A.E.R. is a Lister Institute Clinical Research Fellow.

References Ahmed, S., Reynolds, B.A., Weiss, S., 1995. BDNF enhances the differentiation but not the survival of CNS stem cell-derived neuronal precursors. J. Neurosci. 15, 5765–5778. Armstrong, R.J., Harrower, T.P., Hurelbrink, C.B., McLaughin, M., Ratcliffe, E.L., Tyers, P., Richards, A., Dunnett, S.B., Rosser, A.E., Barker, R.A., 2001. Porcine neural xenografts in the immunocompetent rat: immune response following grafting of expanded neural precursor cells. Neuroscience 106, 201–216. Bartlett, P.F., Brooker, G.J., Faux, C.H., Dutton, R., Murphy, M., Turnley, A., Kilpatrick, T.J., 1998. Regulation of neural stem cell differentiation in the forebrain. Immunol. Cell Biol. 76, 414 – 418. Bayer, S.A., Altman, J., Russo, R.J., Zhang, X., 1993. Timetables of neurogenesis in the human brain based on experimentally determined patterns in the rat. Neurotoxicology 14, 83–144. Belkadi, A.M., Geny, C., Naimi, S., Jeny, R., Peschanski, M., Riche, D., 1997. Maturation of fetal human neural xenografts in the adult rat brain. Exp. Neurol. 144, 369 –380. Bouvier, M.M., Mytilineou, C., 1995. Basic fibroblast growth factor increases division and delays differentiation of dopamine precursors in vitro. J. Neurosci. 15, 7141–7149. Brustle, O., Choudhary, K., Karram, K., Huttner, A., Murray, K., DuboisDalcq, M., Mckay, R.D., 1998. Chimeric brains generated by intraventricular transplantation of fetal human brain cells into embryonic rats. Nat. Biotechnol. 16, 1040 –1044. Caldwell, M.A., He, X., Wilkie, N., Pollack, S., Marshall, G., Wafford, K.A., Svendsen, C.N., 2001. Growth factors regulate the survival and fate of cells derived from human neurospheres. Nat. Biotechnol. 19, 475– 479. Caldwell, M.A., Svendsen, C.N., 1998. Heparin, but not other proteoglycans potentiates the mitogenic effects of FGF-2 on mesencephalic precursor cells. Exp. Neurol. 152, 1–10. Carpenter, M.K., Cui, X., Hu, Z.Y., Jackson, J., Sherman, S., Seiger, A., Wahlberg, L.U., 1999. In vitro expansion of a multipotent population of human neural progenitor cells. Exp. Neurol. 158, 265–278. Cattaneo, E., Pelicci, P.G., 1998. Emerging roles for SH2/PTB-containing Shc adaptor proteins in the developing mammalian brain. Trends Neurosci. 21, 476 – 481. Ciccolini, F., Svendsen, C.N., 1998. 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. 18, 7869 –7880. Conti, L., Sipione, S., Magrassi, L., Bonfanti, L., Rigamonti, D., Pettirossi, V., Peschanski, M., Peschanski, M., Haddad, B., Pelicci, P., Milanesi, G., Pelicci, G., Cattaneo, E., 2001. Shc signaling in differentiating neural progenitor cells. Nat. Neurosci. 4, 579 –586. Davis, A.A., Temple, S., 1994. A self-renewing multipotential stem cell in embryonic rat cerebral cortex. Nature 372, 263–266. Englund, U., Fricker-Gates, R.A., Lundberg, C., Bjorklund, A., Wictorin, K., 2002. Transplantation of human neural progenitor cells into the neonatal rat brain: extensive migration and differentiation with longdistance axonal projections. Exp. Neurol. 173, 1–21. Flax, J.D., Aurora, S., Yang, C., Simonin, C., Willis, A.M., Billinghurst, L.L., Jendoubi, M., Sidman, R.L., Wolfe, J.H., Kim, S.U., Snyder, E.Y., 1998. Engraftable human neural stem cells respond to develop-

mental cues, replace neurons, and express foreign genes. Nat. Biotechnol. 16, 1033–1039. Fricker, R.A., Carpenter, M.K., Winkler, C., Greco, C., Gates, M.A., Bjorklund, A., 1999. Site-specific migration and neuronal differentiation of human neural progenitor cells after transplantation in the adult rat brain. J. Neurosci. 19, 5990 – 6005. Hill, E., Kalloniatis, M., Tan, S.S., 2000. Glutamate, GABA and precursor amino acids in adult mouse neocortex: cellular diversity revealed by quantitative immunocytochemistry. Cereb. Cortex 10, 1132–1142. Hitoshi, S., Tropepe, V., Ekker, M., Van Der Kooy, D., 2002. Neural stem cell lineages are regionally specified, but not committed, within distinct compartments of the developing brain. Development 129, 233–244. Kabos, P., Kabosova, A., Neuman, T., 2002. Blocking HES1 expression initiates GABAergic differentiation and induces the expression of p21 CIP1/WAF1 in human neural stem cells. J. Biol. Chem. 277, 8763– 8766. Kageyama, R., Ohtsuka, T., 1999. The Notch-Hes pathway in mammalian neural development. Cell Res. 9, 179 –188. Kalyani, A.J., Piper, D., Mujtaba, T., Lucero, M.T., Rao, M.S., 1998. Spinal cord neuronal precursors generate multiple neuronal phenotypes in culture. J. Neurosci. 18, 7856 –7868. Konig, N., Wilkie, M.B., Lauder, J.M., 1988. Tyrosine hydroxylase and serotonin containing cells in embryonic rat rhombencephalon: a wholemount immunocytochemical study. J. Neurosci. Res. 20, 212–223. Kukekov, V.G., Laywell, E.D., Suslov, O., Davies, K., Scheffler, B., Thomas, L.B., O’Brien, T.F., Kusakabe, M., Steindler, D.A., 1999. Multipotent stem/progenitor cells with similar properties arise from two neurogenic regions of adult human brain. Exp. Neurol. 156, 333– 344. Laywell, E.D., Kukekov, V.G., Steindler, D.A., 1999. Multipotent neurospheres can be derived from forebrain subependymal zone and spinal cord of adult mice after protracted postmortem intervals. Exp. Neurol. 156, 430 – 433. Morshead, C.M., Reynolds, B.A., Craig, C.G., McBurney, M.W., Staines, W.A., Morassutti, D., Weiss, S., Van Der, K.D., 1994. Neural stem cells in the adult mammalian forebrain: a relatively quiescent subpopulation of subependymal cells. Neuron 13, 1071–1082. Nakamura, Y., Sakakibara, S., Miyata, T., Ogawa, M., Shimazaki, T., Weiss, S., Kageyama, R., Okano, H., 2000. The bHLH gene hes1 as a repressor of the neuronal commitment of CNS stem cells. J. Neurosci. 20, 283–293. Ostenfeld, T., Joly, E., Tai, Y.T., Peters, A., Caldwell, M., Jauniaux, E., Svendsen, C.N., 2002. Regional specification of rodent and human neurospheres. Brain Res. Dev. Brain Res. 134, 43–55. Piper, D.R., Mujtaba, T., Keyoung, H., Roy, N.S., Goldman, S.A., Rao, M.S., Lucero, M.T., 2001. Identification and characterization of neuronal precursors and their progeny from human fetal tissue. J. Neurosci. Res. 66, 356 –368. Qian, X., Davis, A.A., Goderie, S.K., Temple, S., 1997. FGF2 concentration regulates the generation of neurons and glia from multipotent cortical stem cells. Neuron 18, 81–93. Qian, X., Shen, Q., Goderie, S.K., He, W., Capela, A., Davis, A.A., Temple, S., 2000. Timing of CNS cell generation: a programmed sequence of neuron and glial cell production from isolated murine cortical stem cells. Neuron 28, 69 – 80. Quinn, S.M., Walters, W.M., Vescovi, A.L., Whittemore, S.R., 1999. Lineage restriction of neuroepithelial precursor cells from fetal human spinal cord. J. Neurosci. Res. 57, 590 – 602. Reynolds, B.A., Tetzlaff, W., Weiss, S., 1992. A multipotent EGF-responsive striatal embryonic progenitor cell produces neurons and astrocytes. J. Neurosci. 12, 4565– 4574. Reynolds, B.A., Weiss, S., 1992. Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 255, 1707–1710. Reynolds, B.A., Weiss, S., 1996. Clonal and population analyses demonstrate that an EGF-responsive mammalian embryonic CNS precursor is a stem cell. Dev. Biol. 175, 1–13.

M. Jain et al. / Experimental Neurology 182 (2003) 113–123 Rosser, A.E., Tyers, P., Dunnett, S.B., 2000. The morphological development of neurons derived from EGF- and FGF-2-driven human CNS precursors depends on their site of integration in the neonatal rat brain. Eur. J. Neurosci. 12, 2405–2413. Rosser, A.E., Tyers, P., Ter Borg, M., Dunnett, S.B., Svendsen, C.N., 1997. Co-expression of MAP-2 and GFAP in cells developing from rat EGF responsive precursor cells. Brain Res. Dev. Brain Res. 98, 291–295. Roy, N.S., Wang, S., Jiang, L., Kang, J., Benraiss, A., Harrison-Restelli, C., Fraser, R.A., Couldwell, W.T., Kawaguchi, A., Okano, H., Nedergaard, M., Goldman, S.A., 2000. In vitro neurogenesis by progenitor cells isolated from the adult human hippocampus. Nat. Med. 6, 271– 277. Schambra, U.B., Sulik, K.K., Petrusz, P., Lauder, J.M., 1989. Ontogeny of cholinergic neurons in the mouse forebrain. J. Comp. Neurol. 288, 101–122. Shihabuddin, L.S., Palmer, T.D., Gage, F.H., 1999. The search for neural progenitor cells: prospects for the therapy of neurodegenerative disease. Mol. Med. Today 5, 474 – 480. Somogyi, J., Llewellyn-Smith, I.J., 2001. Patterns of colocalization of GABA, glutamate and glycine immunoreactivities in terminals that synapse on dendrites of noradrenergic neurons in rat locus coeruleus. Eur. J. Neurosci. 14, 219 –228. Song, H., Stevens, C., Gage, F., 2002. Astroglia induce neurogenesis from adult neural stem cells. Nature 417, 39 – 44. Svendsen, C.N., Caldwell, M.A., 2000. Neural stem cells in the developing central nervous system: implications for cell therapy through transplantation. Prog. Brain Res. 127, 13–34. Svendsen, C.N., Caldwell, M.A., Ostenfeld, T., 1999. Human neural stem cells: isolation, expansion and transplantation. Brain Pathol. 9, 499 – 513. Svendsen, C.N., Skepper, J., Rosser, A.E., Ter Borg, M.G., Tyres, P.,

123

Ryken, T., 1997. Restricted growth potential of rat neural precursors as compared to mouse. Brain Res. Dev. Brain Res. 99, 253–258. Svendsen, C.N., Ter Borg, M.G., Armstrong, R.J., Rosser, A.E., Chandran, S., Ostenfeld, T., Caldwell, M.A., 1998. A new method for the rapid and long term growth of human neural precursor cells. J. Neurosci. Methods 85, 141–152. Uchida, N., Buck, D.W., He, D., Reitsma, M.J., Masek, M., Phan, T.V., Tsukamoto, A.S., Gage, F.H., Weissman, I.L., 2000. Direct isolation of human central nervous system stem cells. Proc. Natl. Acad. Sci. USA 97, 14720 –14725. Vescovi, A.L., Gritti, A., Galli, R., Parati, E.A., 1999. Isolation and intracerebral grafting of nontransformed multipotential embryonic human CNS stem cells. J. Neurotrauma 16, 689 – 693. Vescovi, A.L., Reynolds, B.A., Fraser, D.D., Weiss, S., 1993. bFGF regulates the proliferative fate of unipotent (neuronal) and bipotent (neuronal/astroglial) EGF-generated CNS progenitor cells. Neuron 11, 951–966. Vicario-Abejon, C., Collin, C., Tsoulfas, P., McKay, R.D., 2000. Hippocampal stem cells differentiate into excitatory and inhibitory neurons. Eur. J. Neurosci. 12, 677– 688. Wallace, J.A., Lauder, J.M., 1983. Development of the serotonergic system in the rat embryo: an immunocytochemical study. Brain Res. Bull. 10, 459 – 479. Weiss, S., Dunne, C., Hewson, J., Wohl, C., Wheatley, M., Peterson, A.C., Reynolds, B.A., 1996. Multipotent CNS stem cells are present in the adult mammalian spinal cord and ventricular neuroaxis. J. Neurosci. 16, 7599 –7609. White, H.L., 1981. Glutamate as a precursor of GABA in rat brain and peripheral tissues. Mol. Cell. Biochem. 39, 253–259. Williams, B.P., Price, J., 1995. Evidence for multiple precursor cell types in the embryonic rat cerebral cortex. Neuron 14, 1181–1188.