Neuroepithelial Stem Cells from the Embryonic Spinal Cord: Isolation, Characterization, and Clonal Analysis

DEVELOPMENTAL BIOLOGY 186, 202–223 (1997) ARTICLE NO. DB978592

Neuroepithelial Stem Cells from the Embryonic Spinal Cord: Isolation, Characterization, and Clonal Analysis Anjali Kalyani, Kristin Hobson, and Mahendra S. Rao1 Department of Neurobiology and Anatomy, University of Utah School of Medicine, 50 North Medical Drive, Salt Lake City, Utah 84132

Adherent cultures of E10.5 rat neuroepithelial cells (NEP cells) from the caudal neural tube require FGF (fibroblast growth factor) and CEE (chick embryo extract) to proliferate and maintain an undifferentiated phenotype in culture. Epidermal growth factor (EGF) does not support E10.5 NEP cells in adherent culture and NEP cells do not form EGF-dependent neurospheres. NEP cells, however, can be grown as FGF-dependent neurospheres. NEP cells express nestin and lack all lineage-specific markers for neuronal and glial sublineages, retain their pleuripotent character over multiple passages, and can differentiate into neurons, astrocytes, and oligodendrocytes when plated on laminin in the absence of CEE. In clonal culture, NEP cells undergo self-renewal and generate colonies that vary in size from single cells to several thousand cells. With the exception of a few single-cell clones, all other NEP-derived clones contain more than one identified phenotype, with over 40% of the colonies containing A2B5, b-111 tubulin, and GFAP-immunoreactive cells. Thus, NEP cells are multipotent and capable of generating multiple neural derivatives. NEP cells also differentiate into motoneurons immunoreactive for choline acetyl transferase (ChAT) and the low-affinity neurotrophin receptor (p75) in both mass and clonal culture. Double labeling of clones for ChAT and glial, neuronal, or oligodendrocytic lineage markers shows that motoneurons always arose in mixed cultures with other differentiated cells. Thus, NEP cells represent a common progenitor for motoneurons and other spinal cord cells. The relationship of NEP cells with other neural stem cells is discussed. q 1997 Academic Press

INTRODUCTION During development neuroepithelial cells of the caudal neural tube differentiate into neurons and glia. Neurons arise from neuroepithelial precursors first and eventually develop unique phenotypes defined by their trophic requirements, morphology, and function. Motoneurons are among the first neurons to develop (Hamburger, 1948; Nornes and Das, 1974; Altman and Bayer, 1984; Phelps et al., 1988, 1990) and can be distinguished from other neurons in the spinal cord by their position and the expression of a number of specific antigens (Chen and Chiu, 1992). Tag-1 (Dodd et al., 1988), islet-1 (Erickson et al., 1992), and p75 (Camu et al., 1992) are expressed uniquely on rat and chick motoneurons early in their development but are not detectable on other spinal cord cells and therefore may serve to distinguish motoneurons from other neural tube cells. Astrocytes characterized by glial fibrillery acid protein (GFAP) immu1

To whom correspondence should be addressed.

noreactivity appear soon after. GFAP staining is seen at E16 (Hirano and Goldman, 1988, and references therein), and astrocytic cells proliferate and populate the gray and white matter of the spinal cord. Both type 1 and type 2 astrocytes have been identified in the spinal cord (Warf et al., 1991). Oligodendrocytes appear later and are first detected around birth, although oligodendrocyte precursors may be present as early as E14 based on PDGFRA expression and culture assays (Pringle and Richardson, 1993; and Warf et al., 1991, respectively). The temporal sequence of differentiated cell development has raised the possibility that lineage-specific precursors may exist. Evidence for the presence of O2A progenitor cells (Warf et al., 1991; Pringle and Richardson, 1993; Ono et al., 1995; Timsit et al., 1995), neuroblasts, which can give rise to motor neurons as well as other neurons (Ray and Gage, 1994), and committed astrocytes (Miller and Szigeti, 1991) suggests the presence of committed precursors in the developing spinal cord. Alternatively, a common precursor may give rise to different phenotypes under different environmental signals (see reviews by Anderson, 1989; McConnell, 0012-1606/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

202

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

203

NEP Cell Differentiation

1991), or a common precursor may divide in a strict temporal sequence to give rise to committed precursor cells (see review Raff et al., 1989). No detailed lineage experiments have been performed in the mouse or rat spinal cord that could distinguish between these alternatives. Retroviral labeling in chick spinal cord, however, suggests that a common multipotent precursor may give rise to both motoneurons and nonneuronal cells (Leber et al., 1990). Neuroepithelial cell culture assays (Reynolds et al., 1992, 1996; Vescovi et al., 1993; Kilpatrick and Bartlett, 1993, 1995; Davis and Temple, 1994; Temple and Davis, 1994) and lineage analysis experiments in vivo and in vitro have suggested the presence of multipotent precursors (Price and Thurlow, 1987; Cepko, 1988; Luskin et al., 1988; Sanes, 1988; Gallileo et al., 1990; Leber et al., 1990; Williams, 1995; Williams and Price, 1995) in other neural areas. These experiments strongly suggest that cortical, cerebellar, and tectal development involve a multipotent precursor differentiating into a more restricted precursor which then gives rise to even more restricted phenotypes. By analogy, one may expect that a similar multipotent precursor present in the developing caudal neural tube will give rise to motoneurons and other neural tube cells. This assumption, while likely to be true, nevertheless needs to be formally tested. This is particularly true in the light of recent evidence for heterogeneity in precursor populations. Culture conditions for maintaining neuroepithelial cells from embryonic neural tube appear different from those required for cortical stem cell cultures. For example, epidermal growth factor (EGF) is required for cortical neuroepithelial stem cell cultures (Reynolds et al., 1992, 1996; Vescovi et al., 1993), but EGF will not maintain neural tube neuroepithelial cells in culture (Groves, 1992; Kilpatrick and Bartlett, 1993, 1995; see Results). Cortical stem cells grow in suspension cultures, while neuroepithelial cells grow on fibronectin (Reynolds et al., 1992, 1995; Groves, 1992). Similarly, the development potential of spinal neuroepithelial cells includes the generation of neural crest stem cells (Bronner-Fraser and Fraser, 1988, 1989; Artinger et al., 1995). Cortical cells do not produce neural crest, and cortical stem cells do not give rise to crest-like derivatives (Reynolds et al., 1992; Vescovi et al., 1993; our unpublished results). In this manuscript we describe the dissociation and maintenance of neuroepithelial cells in culture, their antigenic properties, and the conditions required to promote differentiation. We show by clonal analysis that individual neuroepithelial stem cells are multipotent and are capable of selfrenewal. We present evidence that motoneurons arise from a stem cell that also generates other neurons, glia, and astrocytes and discuss the relationship of the neuroepithelial (NEP) stem cells with other characterized stem cells.

RESULTS Embryonic NEP Cells Require FGF but Not EGF for Survival and Proliferation The neural tube undergoes closure at Embryonic Day 10 in rats (Hamburger, 1948) and earliest differentiation occurs

a day later (Hamburger, 1948; Nornes and Das, 1974; Altman and Bayer, 1984). E10.5 therefore represents the earliest time when a large number of undifferentiated NEP cells may be isolated easily. Sections through E10.5 neural tubes showed that the majority of cells were nestin positive as previously described (Chen and Chiu 1992; Dahlstrand et al., 1995) and did not express any differentiation markers (data not shown). E10.5 rat NEP cells were dissociated and plated on fibronectin-coated dishes and their development monitored under several different conditions. Fibronectin was chosen as a growth substrate since preliminary experiments showed that NEP cells will not adhere to collagen or poly-L-lysine and adhere only poorly to laminin (data not shown). All subsequent experiments to maintain NEP cells in culture were performed on fibronectin-coated dishes. NEP cells plated on fibronectin at low density required FGF for survival (Fig. 1). All NEP cells cultured in the absence of FGF were dead within 48 hr of plating. In contrast, cells grown in medium supplemented with acidic or basic FGF survived (Figs. 1B and 1C) in culture. EGF was not a survival factor for NEP cells (Fig. 2) either in adherent cultures (Fig. 1) or in suspension cultures (data not shown) at any dose tested (1–100 ng/ml). Thus, NEP cells differed from neurospheres (Reynolds et al., 1992, 1996) in their growth characteristics and trophic requirements. While NEP cells survived in acidic or basic FGF, growth was poor and cells differentiated rapidly (Fig. 2A). When the medium was supplemented with chick embryo extract (CEE) cells proliferated and appeared homogenous (Fig. 2B). Virtually all cells had divided and incorporated BRDU (Fig. 3) after 5 days in culture and dividing cells continued to express nestin (Fig. 3). CEE alone was not a survival factor since NEP cells did not survive in medium supplemented with CEE in the absence of exogenously added FGF and the effect of CEE could not be mimicked by the addition of EGF (data not shown). Thus, CEE contains a component distinct from EGF, that in concert with FGF, maintains NEP cells in an undifferentiated state in culture.

FGF and CEE Are Sufficient to Maintain Neuroepithelial Cells in an Undifferentiated State in Culture NEP cells grown in FGF and CEE on fibronectin appear homogenous and continue to divide in culture (see above). To determine their antigenic phenotype, NEP cells grown in culture for a period of 5 days were tested for differentiation using a variety of antigenic markers (summarized in Table 1). NEP cells expressed nestin but did not express any other marker tested. NEP cells passaged at a 1:3 dilution every fifth day as adherent cultures could be maintained as nestin-immunoreactive cells that did not express any markers characteristic of differentiated cells (Fig. 4 and data not shown) for at least three passages. Subsequent passaging over 3 months maintained nestin-immunoreactive, lineagenegative cells, but a small percentage of GFAP-immunoreactive cells (1–5%) could also be detected (data not shown).

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

204

Kalyani, Hobson, and Rao

FIG. 1. FGF but not EGF supports NEP cells in culture. E10.5 rat neural tube cells were dissociated and plated at low density and grown in NEP medium with either aFGF (25 ng/ml), bFGF (25 ng/ml), EGF (50 ng/ml), or no added factor for 48 hr. Culture dishes were fixed and examined by phase-contrast microscopy. Cells grown in aFGF or bFGF survived and increased in density. In contrast, no surviving cells were seen in cultures grown without added factor or with 50 ng/ml EGF.

Subsequent analysis was therefore limited to passage 1–3 cells. Thus, NEP cells could be passaged and their numbers amplified when grown under nondifferentiating conditions.

Cultured Neuroepithelial Cells Can Differentiate into Neurons, Astrocytes, and Oligodendrocytes The CNS consists of three major phenotypes: neurons, oligodendrocytes, and astrocytes, each of which expresses characteristic antigenic markers. To determine if undifferentiated, cultured NEP cells could differentiate into CNS neurons and glia, NEP cells were replated on laminin in neuroepithelial culture medium (without the addition of CEE). Omission of CEE was used to promote differentiation as previous experiments had shown that NEP cells spontaneously differentiate in the absence of CEE. Laminin was used as a substrate instead of fibronectin because laminin has been shown to promote proliferation and neuronal differentiation (Drago et al., 1991). Under these conditions, NEP cells rapidly differentiated, as characterized by alter-

ations in morphology and the expression of lineage-specific antigenic markers. Small, phase-bright cells with small processes could be seen as early as 48 hr after replating onto laminin in the absence of CEE. Cells with this morphology expressed b111 tubulin immunoreractivity and a subset of the b-111 tubulin-immunoreactive cells also expressed neurofilament 160 (NF160) immunoreactivity (Fig. 5). b-111 tubulin immunoreactive, NF1600 cells were also seen (data not shown). These cells likely represent immature neurons (Memberg and Hall, 1994). The number of b-111 tubulin immunoreactive cells increased in culture over 5 days at which time they represented 20 { 4% (mean of two independent experiments) of the total number of cells. In addition to the small, phase-bright b-111 tubulin immunoreactive cells, cells with a larger cell soma and more elaborate processes were seen. These cells were p75, NF160, and choline acetyltransferase (ChAT) immunoreactive and were seen both as single cells and as clusters (Fig. 6). In the developing neural tube, p75 and ChAT immunoreactivity

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

205

NEP Cell Differentiation

FIG. 2. CEE is required to maintain NEP cells in an undifferentiated state. E10.5 rat neural tube cells were dissociated and equal numbers of cells were plated at low density in a 35-mm dish and grown in NEP medium containing bFGF (25 ng/ml) with 10% CEE (B) or without CEE (A) for 5 days. Cells grown in the absence of CEE grew slowly and some cells appeared rounded and phase bright. Cells grown in the presence of 10% CEE appeared more homogenous and had proliferated to form a confluent monolayer (compare A and B).

is characteristic of motoneurons (Camu et al., 1992). The p75/ChAT-immunoreactive cells will be referred to as motoneurons in subsequent sections. ChAT-immunoreactive cells represented a small proportion (4 { 2% SEM) of the

total number of cells (mean from two independent experiments). The largest proportion of differentiated cells were immunoreactive for GFAP. After 5 days in culture, GFAP-immu-

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

206

Kalyani, Hobson, and Rao

FIG. 3. NEP cells proliferate in vitro in the presence of FGF and 10% CEE. E10.5 rat neural tube cells were dissociated and plated at low density and grown in NEP medium with FGF and 10% CEE for 5 days. BRDU was added to cultured cells at Day 2. Cells were fixed and labeled for nestin and sister wells were tested for BRDU. All NEP cells were nestin immunoreactive (compare A and B) and most of the cells had divided and incorporated BRDU over a 3-day period (C and D).

noreactive cells constituted 73 { 6% SEM of all cells present. Two characteristic morphologies could be identified (Fig. 7). A flattened pancake-shaped cell with small or absent processes and a smaller more fibroblastic cell with long elaborate processes were seen. Neither population of GFAPimmunreactive cells expressed A2B5, p75, or b-111 immunoreactivity, indicating that these cells were most likely type 1 astrocytes (Fig. 8). No type 2 astrocytes, as defined by coexpression of A2B5 and GFAP (Raff, 1989; Lillien and Raff, 1990), were identified, although such type 2 astrocytes have been generated from NEP cells under other culture

conditions (Mayer-Proschel and Rao, 1996, Society of Neuroscience Abstract 217.13). Differentiating NEP cultures were also labeled with markers previously identified as being expressed on oligodendrocytes and their precursors. Three days after replating NEP cells, a subset of the cells began to express A2B5-immunoreactivity. A2B5-immunoreactive cells initially did not express detectable levels of Gal-C, O4, and O1 immunoreactivity (data not shown and Mayer-Proschel and Rao, unpublished results). After an additional 3 days in culture, Gal-C-immunoreactive cells could be seen (Fig. 9), which also expressed A2B5-immuno-

FIG. 4. NEP cells do not express differentiation markers when grown with CEE and FGF. Passage 3. NEP cells were grown on fibronectin in NEP medium containing FGF (25 ng/ml) and 10% CEE for a period of 5 days. Sister plates were fixed and double labeled for nestin (A, C, E, and G) and one of the following lineage markers: p75 (B), GFAP (D), b-111 tubulin (F), or A2B5 (H). Cultured, passaged, nestinimmunoreactive NEP cells do not express p75 (B), b-111 tubulin (F), GFAP (D), or A2B5 (H) immunoreactivity.

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

207

AID

DB 8592

/

6x25$$8592

06-04-97 22:03:06

dba

FIG. 5. Neuronal differentiation of NEP cells. E10.5 NEP cells grown on fibronectin in NEP medium for 5 days were harvested by trypsinization and replated onto laminin-coated dishes in NEP medium without CEE. Replated cells were grown for an additional 5 days and then fixed and processed for b-111 tubulin and neurofilament immunoreactivity. DAPI histochemistry was used to localize nuclei (C and D). b-111 tubulinimmunoreactive cells (A and B) differentiate after 5 days in culture. Double labeling with b-111 tubulin (C, green) and neurofilament (D, red) antibodies shows that b-111 tubulin immunoreactive cells also express neurofilament immunoreactivity.

208

AID

Kalyani, Hobson, and Rao

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

DB 8592

/

6x25$$8592

06-04-97 22:03:06

dba

209

NEP Cell Differentiation

FIG. 6. A subset of the neurons generated from NEP cultures express motoneuron markers. E10.5 NEP cells grown on fibronectin for 5 days were harvested by trypsinization and replated onto laminin-coated dishes in NEP medium without CEE for a period of 5 days. Culture dishes were fixed and processed for ChAT (B, D, and F), b-111 tubulin (C and D), and p75 immunoreactivity (E and F). A small number of ChAT-immunoreactive cells (compare A and B) differentiate after 5 days in culture. Double labeling with b-111 tubulin (C) and p75 (E) shows that ChAT-immunoreactive cells coexpress b-111 tubulin and p75 immunoreactivity.

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$8592

06-04-97 22:03:06

dba

210

Kalyani, Hobson, and Rao

FIG. 7. NEP cells differentiate into GFAP immunoreactive cells. E10.5 NEP cells grown on fibronectin for 5 days were harvested by trypsinization and replated onto fibronectin-coated dishes in NEP medium without CEE, but with the addition of 10% FBS for a period of 5 days. Culture dishes were fixed and labeled for glial markers using fluorescent secondaries. NEP cells differentiate into GFAPimmunoreactive cells with two distinct morphologies: a spindle Schwann-cell-like morphology (C and D) and a flattened morphology (A and B).

reactivity. These cells appeared flattened and did not have the characteristic morphology of O2A progenitors or mature oligodendrocytes. Longer periods in culture, however, allowed more mature-looking oligodendrocytes with a small cell body and extensive processes to develop (Fig. 10). These cells expressed O1 and Gal-C immunoreactivity, markers characteristic of differentiated oligodendrocytes (Fig. 10). Thus, NEP cells can generate oligodendrocytes that mature over 10 days in culture. The pattern of antigen expression further suggests the existence of a dividing oligodendrocyte precursor that subsequently generates oligodendrocytes, as has been described from spinal cord cultures from older embryos (Warf et al., 1991; Miller and Szigeti, 1991). In summary, NEP cells grown in culture can generate neurons, astrocytes, and oligodendrocytes when replated on laminin in the absence of CEE. This culture condition, while suboptimal for any particular phenotype, is sufficient

to generate differentiated progeny and has been used to assess differentiation in subsequent experiments.

Individual NEP Cells Are Multipotent and Undergo Self Renewal in Culture NEP cells grown in culture could either be a homogenous population of cells where each cell could differentiate into all phenotypes or a heterogeneous population with a variety of differentiation potentials. To distinguish between these possibilities, cultured NEP cells were grown at clonal density, individual cells were circled, and their development was followed for a period of 15 days. Clones were analyzed for differentiation by triple labeling using GFAP, b-111 tubulin, and A2B5 as markers for astrocytes, neurons, and oligodendrocyte precursors. Figure 11 shows an example of one such clone, which contained all three phenotypes.

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

FIG. 8. NEP cell-derived GFAP immunoreactive cells do not express stem cell, Schwann cell, or neuronal markers. E10.5 NEP cells grown on fibronectin for 5 days were harvested by trypsinization and replated onto fibronectin coated dishes in NEP medium without CEE, but with the addition of 10% FBS for a period of 5 days. Culture dishes were fixed and double labeled for GFAP (green, A – D), p75 (red, A), nestin (red, B), b111 tubulin (red, C), and A2B5 (red, D) using fluorescent secondaries. GFAP-immunoreactive cells do not express nestin, A2B5, p75, or b-111 tubulin immunoreactivity and are thus distinct from other lineages generated from NEP cells.

NEP Cell Differentiation

AID

DB 8592

/

211

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

6x25$$8592

06-04-97 22:03:06

dba

212

AID

DB 8592

/

6x25$$8592

06-04-97 22:03:06

dba

213

NEP Cell Differentiation

FIG. 10. NEP cells differentiate into oligodendrocytes. E10.5 NEP cells grown on fibronectin for 5 days were harvested by trypsinization and replated onto laminin-coated dishes in NEP medium without CEE for a period of 10 days. Culture dishes were fixed and labeled for oligodendroglial markers using fluorescent secondaries. (A) A2B5-immunoreactive cells that have acquired a typical oligodendrocytic morphology in culture. Double labeling of cells shows Gal-C-positive cells (B) that are O1 (C) immunoreactive.

Thus, at least some NEP cells are capable of generating neurons, astrocytes, and oligodendrocytes. To confirm that A2B5-immunoreactive cells represent oligodendrocytes, some clones were restained with O1 or Gal-C (data not shown). Several hundred clones were analyzed by triple labeling and the results are summarized in Table 1. All clones analyzed contained more than one phenotype. Neuron and oligodendrocyte clones, as well as neuron and astrocyte clones, were identified (Table 1) A significant proportion of NEP cells generated colonies containing all three pheno-

types of cells (Table 1). In all cases, when clones were studied carefully, it was possible to identify cells that did not express any of the markers tested, suggesting that precursor cells were still present. Furthermore, no clones that contained only one cell type could be identified, suggesting that at this stage no fully committed precursors were present in culture. To determine if multipotent stem cells underwent selfrenewal, NEP cells were plated at low density and single cells were followed for 10 days. Clones at this stage varied

FIG. 9. NEP cells differentiate into cells of the oligodendrocyte lineage. E10.5 NEP cells grown on fibronectin for 5 days were harvested by trypsinization and replated onto laminin-coated dishes in NEP medium without CEE for a period of 5 to 10 days. A significant number of NEP cells grown in culture (A and B) for 5 days expressed A2B5 immunoreactivity. A2B5 immunoreactive cells (D) also expressed nestin (C). Sister plates allowed to grow for 10 days were double labeled for A2B5 (E) and Gal-C (F), or for O4 (G) and Gal-C (H). Note the presence of A2B5/Gal-C-immunoreactive (arrows in E and F) and Gal-C /O1-immunoreactive cells (arrows in G and H).

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

FIG. 11. NEP cells form multipotent colonies in culture. E10.5 NEP cells grown on fibronectin for 5 days were harvested by trypsinization and replated onto fibronectin-coated 35-mm dishes at clonal density in NEP medium with 10% CEE. Single isolated cells were circled and followed for a period of 15 days. Clones were then triple labeled for A2B5, b-111 tubulin, and GFAP expression. (A) Phase picture of one such clone. (C) The same clone double labeled for b-111 tubulin (red) and A2B5 (green). (B) GFAP-immunoreactive cells in the same clone. (D) A bright field picture of the same clone at higher magnification to show the relationship of the GFAP-immunoreactive cells and large neuronal cells.

214

AID

Kalyani, Hobson, and Rao

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

DB 8592

/

6x25$$8592

06-04-97 22:03:06

dba

215

NEP Cell Differentiation

TABLE 1 NEP Cells Form Colonies Containing Neurons, Astrocytes, and Oligodendrocytes Antigen expressed

% of clones

A2B5 / b-111 tubulin* A2B5 / GFAP* A2B5 / b-111 tubulin / GFAP* GFAP / b-111 tubulin* GFAP alone b-111 tubulin alone A2B5 alone

13 { 2 28 { 2 42 { 3 17 { 1 None None None

Total number of colonies counted

256 (100%)

Note. E10.5 NEP cells grown on fibronectin for 5 days were harvested by trypsinization and replated onto fibronectin-coated 35mm dishes at clonal density (50–100 cells/35-mm dish) in NEP medium with CEE. Single isolated cells were circled and followed for a period of 15 days. Clones were then triple labeled for A2B5, b-111 tubulin, and GFAP expression. The number of clones expressing one or more markers is listed. The numbers represent the total number of colonies from three independent clonal assays. * Nonoverlapping staining.

Neurospheres generated in FGF-containing medium could be replated onto laminin and would subsequently adhere and differentiate into neurons, astrocytes, and oligodendrocytes (Fig. 12). Neurospheres could also generate secondary and tertiary neurospheres (data not shown) as has been described for EGF generated neurospheres (Reynolds et al., 1992, 1996; Vescovi et al., 1993). Neurospheres, when replated on fibronectin-coated plates, grew as undifferentiated nestin-immunoreactive cells (see Fig. 12) and appeared morphologically similar to NEP cells that had been generated from neural tube dissociation (compare Figs. 4 and 12). Adherent NEP cells proliferated and could subsequently be passaged and used to generate additional neurospheres (data not shown). Thus, NEP cells and FGF-dependent neurospheres represent identical cells grown under adherent or nonadherent culture conditions but are distinct from the EGF-dependent neurospheres generated from older embryos (Reynolds et al., 1992, 1996; Vescovi et al., 1993).

Motoneurons Arise from a Common Multipotent NEP Precursor Cell Motoneurons are the earliest cell type to differentiate from caudal neuroepithelium (Hamburger, 1948; Nornes

in size from hundreds to several thousand cells. The largest clones were isolated and a subset of the cells (50–100 cells) were replated at clonal density and individual clones were followed (Table 2). Out of 15 clones that were followed, each clone contained between 1 and 15 daughter clones (3– 50% of replated cells) that had differentiated into neurons, astrocytes, and oligodendrocytes. Thus, individual NEP cells are capable of at least limited self-renewal and can generate multipotent daughter cells.

NEP Cells Can Form FGF-Dependent Neurospheres Stem cells that undergo self-renewal and retain their ability to differentiate into multiple phenotypes have been previously described (Reynolds et al., 1992, 1996; Vescovi et al., 1993; Kilpatrick and Bartlett, 1993, 1995; Davis and Temple, 1994; Temple and Davis, 1994). One such stem cell is the neurosphere isolated from cortical ventricular zone, which can be maintained in an undifferentiated state over multiple passages (Reynolds et al., 1992, 1996; Vescovi et al., 1993) in defined medium in the presence of EGF. To determine if NEP cells could be grown as neurospheres, cells grown as adherent cultures were trypsinized, pelleted, and grown in bacterial plates (non-tissue culture treated) as suspension cultures at a density of 100–300 cells. The medium used was identical to the NEP medium used for adherent cultures. Most cells did not survive the replating. On average, 2.5 { 1.0 cells (1.2%) formed neurospheres (mean from three independent experiments). No neurospheres were obtained when cells were grown in EGF (50 ng/ml) rather than FGF (20 ng/ml) containing medium (data not shown).

TABLE 2 NEP Stem Cells Undergo Self-Renewal in Culture Clone number

Number of cells followed

Number of multipotent daughter clones

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

40 34 38 36 42 22 21 13 17 21 19 23 41 16 37

15 3 12 8 9 2 3 1 8 7 3 4 13 7 9

Note. NEP cells were grown in clonal culture for a period of 10 days. Large clones were identified and harvested by trypsinization, and a subset of the cells was replated on fibronectin in NEP medium. Individual cells from each parent clone (1–50) were circled and followed in culture. Fifteen days after replating, clones were triple labeled for O1, b-111 tubulin, and GFAP expression. The number of daughter clones that expressed all three markers is listed. Results are the sum of three independent experiments. Note that the percentage of multipotent daughter clones varied widely, but all clones followed generated some multipotent daughter cells.

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

216

Kalyani, Hobson, and Rao

FIG. 12. FGF dependent neurospheres can be generated from NEP cell cultures. E10.5 NEP cells grown on fibronectin for 5 days were harvested by trypsinization and replated onto nonadherent dishes at clonal density in NEP medium with FGF and 10% CEE. A sister plate was processed for nestin expression (A). Single isolated cells (B) were followed for a period of 15 days. Spheres that formed were then replated onto either fibronectin in nondifferentiating medium (D) or on laminin in the absence of CEE (E). Spheres grown on fibronectin were labeled with BRDU (red) and nestin (green) to show that the majority of cells are nestin-immunoreactive dividing cells (D). Spheres grown on laminin were triple labeled for O1 (red), b-111 tubulin (green), and GFAP (blue) expression. Note: a neurosphere generated from NEP cells grown in culture can differentiate into neurons, astrocytes, and oligodendrocytes or be maintained as undifferentiated nestin immunoreactive dividing cells.

and Das, 1974; Altman and Bayer, 1984; Phelps et al., 1988, 1990). N-CAM- and p75-immunoreactive neurons are seen in vivo (Chen and Chiu 1992; Camu and Henderson, 1990) and in vitro (Camu and Henderson, 1990; Rao and Anderson, unpublished results) within 12 hr of the time when neural tubes are isolated and NEP cells placed in culture. It is therefore possible that a committed motoneuron precursor is already present at the time NEP cells were placed in culture. To determine if such a precursor existed, we analyzed NEP clonal cultures with motoneuron- and other lineage-specific markers. NEP cells isolated from caudal neural tubes were grown in clonal culture for 10-15 days. Cultures were then labeled with antibodies to ChAT, GalC, and GFAP. Clones containing ChAT immunoreactive cells were scored on the basis of the other markers expressed (Table 3). A representative clone showing that motoneurons

arise from a common progenitor is shown (Fig. 13). Of 87 clones analyzed, no clone containing motoneurons alone was seen. Most motoneuron-containing clones also contained astrocytes, other neurons, and/or oligodendrocytes. Thus, a common progenitor exists which can generate motoneurons and other spinal cord cells.

DISCUSSION This study describes an embryonic spinal cord stem cell (termed NEP cell) derived from caudal neuroepithelium that requires FGF and CEE to proliferate and self-renew. NEP cells are characterized by the expression of nestin and the absence of lineage-specific markers and can be maintained in an undifferentiated state in culture. Under appropriate

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

217

NEP Cell Differentiation

TABLE 3 Motoneurons Arise in Colonies Containing Other Neural Derivatives Antigen expressed

Number of clones

ChAT / b-111 tubulin ChAT / GFAP ChAT / A2B5

93% (26/28) 94% (30/32) 89% (24/27)

Total number of clones

100% (87)

Note. E10.5 NEP cells grown on fibronectin for 5 days were harvested by trypsinization and replated onto fibronectin-coated 35mm dishes at clonal density in NEP medium with CEE. Single isolated cells were circled and followed for a period of 21 days. Clones were then double labeled for ChAT and b-111 tubulin, GFAP, or A2B5. The number of clones expressing both markers is summarized. Note no clones contained only ChAT-immunoreative cells.

environmental conditions, NEP cells differentiate into the three principal cell types: neurons, astrocytes, and oligodendrocytes. Based on our culture assays and previously published data (Miller and Szigeti, 1991; Warf et al., 1991; Pringle and Richardson 1993; Ray and Gage, 1994) we propose a model for spinal cord NEP cell differentiation (Fig. 14). This model is similar to that proposed for hematopoiesis and for differentiation of neural crest (see review by Anderson, 1989) in that we postulate the existence of multipotent cells that undergo progressive restriction in their developmental potential. We propose that NEP cells represent an apparently homogenous population of cells in the caudal neural tube that express nestin but no other lineage marker. These cells divide and self-renew in culture and generate differentiated phenotypes. While our present experiments do not allow us to determine if intermediate dividing precursors with a more restricted potential are present, previous data (Miller and Szigeti, 1991; Warf et al., 1991; Pringle and Richardson, 1993; Ray and Gage, 1994) suggest that such restricted precursors exist. These include precursors that generate oligodendrocytes and type 2 astrocytes (Miller and Szigeti, 1991; Warf et al., 1991), bipotent (astrocyte and neuronal) precursors (Murphy et al., 1990), as well as neuronal progenitors (Ray and Gage, 1994) that generate several kinds of neurons. The model therefore suggests that multipotent precursors generate differentiated cells through intermediate more restricted precursors. Consistent with the idea of intermediate precursors are our clonal assay results which suggest the existence of cells with a more restricted proliferative potential (see Results, see also MayerProschel and Rao, 1996, Society for Neuroscience Abstract 217.13). To demonstrate the relationship between NEP cells and more restricted progenitors, however, additional experiments are required. The model proposed is consistent with cortical stem cell culture results described previously (Reynolds et al., 1992;

Vescovi et al., 1993; Kilpatrick and Bartlett, 1993; Temple and Davis, 1994) but contrasts with conclusions based on retroviral labeling assays in vivo and in vitro. Retroviral studies in mouse cortex have failed to find multipotent clones that generate all three major CNS phenotypes; rather, predominantly unipotent or bipotent clones have been characterized (Price, 1987; Cepko, 1988; Sanes, 1989). Retroviral studies, however, utilized older animals than those used for culture assays, suggesting that at later stages more committed precursors may be present. Retroviral labeling at earlier stages may reveal multipotent clones. We and others (see, for example, Temple and Davis, 1994) have

FIG. 13. Motoneurons arise in colonies containing other neural derivatives. E10.5 NEP cells grown on fibronectin for 5 days were harvested by trypsinization and replated onto fibronectin-coated 35-mm dishes at clonal density in NEP medium with CEE. Single isolated cells were circled and followed for a period of 21 days. Clones were then triple labeled for ChAT (brown), b-111 tubulin (red), and A2B5 (green) expression. (A) A bright-field picture of one such clone. Note ChAT-immunoreactive cells with long processes. (B) A fluorescent image of the same clone containing b-111 tubulinimmunoreactive cells, as well as A2B5-immunoreactive cells. Note: a subset of the b-111 tubulin-immunoreactive cells coexpress ChAT immunoreactivity (compare A and B).

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

218

Kalyani, Hobson, and Rao

FIG. 14. A model for NEP cell differentiation in culture. A summary figure showing the antigenic properties of NEP cells and their progeny. Question marks against the neural crest and type 2 astrocyte lineage indicate that these lineages were not identified in this article. Possible transdifferentiation of A2B5 and GFAP immunoreactive cells into other lineages is discussed in the text.

noted that the frequency of multipotent clones falls rapidly with age. It should be noted also that retroviral studies in the chick spinal cord at sufficiently early time points have identified multipotent clones (Leber et al., 1990).

motoneuron differentiation involves a multipotent precursor undergoing progressive stages of commitment.

Motoneurons Arise from a Common NEP Precursor

NEP cells grow on fibronectin and require FGF and an uncharacterized component present in the chick embryo extract preparation to proliferate and maintain an undifferentiated phenotype in culture. These requirements are different from those required for stem cell neurospheres isolated from E14.5 cortical ventricular zone cells (Reynolds et al., 1992; Reynolds and Weiss, 1996). Neurospheres grow in suspension culture, do not require CEE or FGF, but are dependent on EGF for survival (Reynolds et al., 1992; Reynolds and Weiss, 1996). FGF itself is not sufficient for longterm growth, although FGF may support their growth transiently (Vescovi et al., 1993). The difference in the growth and trophic requirements of NEP cells and neurospheres raises the possibility that these cells represent either distinct precursors (representative of different brain regions), or different developmental stages of the same precursor. Our results suggest that FGF-dependent NEP cells grow as adherent or suspension cultures depending on the substrate selected. NEP cells grown in suspension or as adherent cultures are antigenically similar (lineage negative), can switch between adherent and suspension growth, and are capable

Our clonal assays show that p75-immunoreactive ChATpositive cells arise in mixed cultures along with other cells of the spinal cord. No clone consisting exclusively of p75/ ChAT-immunoreactive cells was identified, indicating that at the age single cells were plated, committed motoneuron precursors were not present. Our observation that motoneurons arise from a common NEP precursor are consistent with results obtained in chick neural tube experiments (Bronner-Fraser and Fraser, 1988, 1989; Leber et al., 1990; Artinger et al., 1995). These authors have shown that single neural tube precursor cells can generate motoneurons, other spinal cord cells, as well as neural crest derivatives. Our data do not exclude the presence of an intermediate proliferating precursor that generates either exclusively motoneurons or motoneurons and other neurons. Indeed, an intermediate neuron-only precursor (that generates motoneurons) although not a motoneuron-specific precursor has been defined earlier (Ray and Gage, 1994). Our results, together with previous observations, therefore, suggest that

Relationship of NEP Cells to Neurospheres and Other Stem Cells

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

219

NEP Cell Differentiation

of differentiating into neurons, oligodendrocytes, and astrocytes. Thus, NEP derived FGF-dependent neurospheres which proliferate, self-renew, and can be dissociated and passaged (see results) are similar to EGF-dependent neurospheres. Since NEP cells were isolated from animals 4 days younger than the animals from which neurospheres are grown, it seems possible that FGF-dependent NEP cells are precursors to the EGF-dependent neurosphere cells. Such a possibility is consistent with the temporal distribution of FGF (Nurcombe et al., 1992), EGF (Lazar and Blum, 1992), and EGF receptors (Adamson and Meek, 1984) and the observation that EGF-dependent neurospheres may be supported by FGF (Gritti et al., 1995; Vescovi et al., 1993), albeit transiently. We have, however, been unable to document a transition from FGF to EGF dependence in either adherent or suspension culture (unpublished observations). Furthermore, FGF-dependent neurospheres have been isolated from adult brains using culture conditions (substituting FGF for EGF) identical to those described for generating EGF-dependent neurospheres (Gritti et al., 1995), raising the possibility that two independent precursor populations exist. Consistent with this possibility are the recent results of Santa-Olalla and Covarrubias (1995) and Kilpatrick and Bartlett (1995), who have shown that two distinct populations of embryonic precursors, one EGF dependent and the other EGF/FGF dependent, are present in discrete regions of the developing brain. Additional experiments, however, are required to clarify the relationship between EGF-dependent neurospheres and FGF-dependent NEP cells. NEP cells may represent a multipotent precursor characteristic of the brain stem and spinal cord while neurospheres may represent a stem cell more characteristic of the cortex. NEP cells appear most similar to the neuroepithelial cultures from the myelencephalon and telencephalon (Murphy et al., 1990; Drago et al., 1991; Kilpatrick and Bartlett, 1993). Like those cells, NEP cells are FGF dependent, grow as adherent cells, and require an uncharacterized component present in CEE and/or serum (Kilpatrick and Bartlett, 1993). Cells isolated by Kilpatrick and Bartlett (1993) do not form neurospheres and do not survive in EGF-containing medium (in the absence of FGF). Thus, brain stem and spinal precursor cells (present results; Kilpatrick and Bartlett, 1993, and references therein) may be distinct from cortical precursors. While brain stem and spinal cord NEP cells share many characteristics, differences also exist. Brain stem neuroepithelial cells are bipotent and have not been shown to differentiate into oligodendrocytes. Further, spinal cord NEP cells rapidly differentiate into astrocytes in the presence of serum (see Results). In contrast, brain stem NEP cells remain in an undifferentiated state in the presence of serum (Kilpatrick and Bartlett, 1993 and references therein). Thus, significant differences exist which may be due to the differing culture conditions, species differences, or differences in the precursor population. Direct comparisons of the cell types under identical conditions may resolve this issue.

Relationship of NEP Cells to Neural Crest Stem Cells NEP cells differ from neural crest stem cells in their morphology and antigenic profile. Crest cells are more fibroblastic, and they tend to be migratory and avoid cell contact (Boisseau et al., 1991; Bannerman and Pleasure, 1993). NEP cells appear more flattened and epitheliod and tend to grow as tightly packed monolayers (Fig. 3; Kilpatrick and Bartlett, 1993). Unlike rat neural crest cells, NEP cells do not express immunoreactivity for the low-affinity neurotrophin receptor (Fig. 4). The progeny of the NEP cells in our culture conditions differ from neural crest cell derivatives. The GFAP-immunoreactive cells from NEP cultures do not express detectable nestin and p75 immunoreactivity (Fig. 8). Schwann cells, in contrast, express high levels of both p75 and GFAP in culture (Stemple and Anderson, 1992). Schwann cells differentiating from neural crest also express myelin markers such as O4 and P0 (Stemple and Anderson, 1992; Rao and Anderson, 1996). Neither of these markers is expressed by GFAP-immunoreactive cells derived from NEP cultures (see Results). NEP cultures contain A2B5immunoreactive cells which subsequently express O4, GalC, and O1. Cells with this pattern of antigen expression are not seen as derivatives of neural crest. Furthermore, while crest-derived parasympathetic neurons express ChAT immunoreactivity in vivo, such neurons have yet to be described from neural crest cultures (see review by Anderson, 1989). NEP cells, however, readily differentiate to generate large numbers of neurons coexpressing p75 and ChAT. Thus, NEP cells and neural crest stem cells are morphologically and antigenically distinct and generate differentiated progeny that are phenotypically different and therefore represent distinct stem cells. Neural crest stem cells which give rise to the PNS likely derive from neuroepithelial stem cells. In the chick spinal cord, for example, it has been shown that a common precursor can give rise to both motoneurons and other spinal cord cells, as well as neural crest cells (Bronner-Fraser and Fraser, 1988, 1989; Artinger et al., 1995). By analogy, one may expect a similar common CNS–PNS progenitor in mice and rats. The available data does not, however, allow us to determine whether the NEP stem cell represents this common CNS–PNS progenitor. We have not yet identified cells with a neural crest morphology and antigenic profile in our NEP cultures. Our observation in mass cultures and previous results have been that neural crest cells do not differentiate from dissociated neural tubes (Rao and Anderson, unpublished results). It should be noted, however, that our culture medium and substrate conditions differ from those considered optimal for neural crest derivatives. Peripheral neurons and Schwann cells differentiate from precursors with the addition of dibutryl cyclic AMP and serum in the presence of EGF, FGF, and NGF (Stemple and Anderson, 1992; Rao and Anderson, 1996). It is therefore possible that manipulating culture conditions would promote neural crest differentiation from NEP cells, and that NEP cells represent a com-

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

220

Kalyani, Hobson, and Rao

mon neural progenitor that can generate all CNS and PNS derivatives. Further experimentation is required to determine the relationship between NEP cells and neural crest cells. In summary, we have characterized a stem cell, the NEP cell, from the developing spinal cord and have compared its properties with those of other nervous system stem cells. NEP cells represent a mutliptoent precursor cell population that undergoes self renewal and differentiates into all major CNS phenotypes in culture. These cells differ from previously identified stem cells in their culture conditions and proliferative potential. NEP cell cultures provide a large source of transient cells that can be sorted to obtain differentiated cell types for further analysis, providing a valuable tool for dissecting the molecular events underlying differentiation.

MATERIALS AND METHODS

Hanks’ balanced salt solutions (HBSS) (Gibco/BRL). The trunk segments of the embryos (last 10 somites) were dissected using tungsten needles, rinsed, and then transferred to fresh HBSS. Trunk segments were incubated at 47C in 1% trypsin solution (Gibco/ BRL) for a period of 10 to 12 min. The trypsin solution was replaced with fresh HBSS containing 10% fetal bovine serum (Gibco/BRL). The segments were gently triturated with a Pasteur pipet to release neural tubes free from surrounding somites and connective tissue. Isolated neural tubes were transferred to a 0.05% trypsin/EDTA solution (Gibco/BRL) for a additional period of 10 min. Cells were dissociated by trituration and plated in 35-mm dishes at high density. Cells were maintained at 377C in 5% C02/95% air. Cells were replated at low density 1 to 3 days after plating. Cells from several dishes were then harvested by trypsinization (0.05% trypsin/EDTA solution for 2 min). Cells were pelleted, resuspended in a small volume, and counted. About 5000 cells were plated in a 35-mm dish (Corning, supplied by Intermountain Scientific, Bountiful, UT). For clonal analysis, cells harvested by trypsinization were plated at a density of 50–100 cells per 35-mm dish. Individual cells were identified and located on the dish by marking the position with a grease pencil. Cells were grown in DMEM/F12 with additives as described for a period ranging from 10 to 15 days.

NEP Cell Culture Medium The basal medium used in all experiments is a chemically defined medium modified from that described by Stemple and Anderson (1992). The medium consists of DMEM–F12 supplemented with additives described by Bottenstein and Sato (1979) and further supplemented with FGF (5–20 ng/ml) and CEE extract (10%) prepared as described previously (Stemple and Anderson, 1992; Rao and Anderson, 1994; Sommers et al., 1996). The recipe is detailed below: to DMEM–F12 (Gibco/BRL, Gaithersberg, MD) add 100 mg/ ml transferrin (Calbiochem, San Diego, CA), 5 mg/ml insulin (Sigma), 16 mg/ml putrescine (Sigma), 20 nM progesterone (Sigma), 30 nM selenious acid (Sigma), 1 mg/ml bovine serum albumin (Gibco/BRL), plus B27 additives (Gibco/BRL), FGF (25 ng/ml), and 10% CEE. In general, the additives were stored as 1001 concentrates at 0207C until used. In general, 200 ml of NEP medium was prepared with all additives except CEE and used within 2 weeks of preparation. CEE was added to the NEP medium at the time of feeding cultured cells.

Substrate Preparation Fibronectin (New York Blood Center, New York, NY, or Sigma, St. Louis, MO) was resuspended to a stock concentration of 10 mg/ ml and stored at 0807C until used. The fibronectin stock was diluted to a concentration of 250 mg/ml in D-PBS (Gibco/BRL). The fibronectin solution was applied to tissue culture dishes and immediately withdrawn. Laminin (Gibco/BRL or Sigma) was used at a concentration of 50–250 mg/ml and dishes were coated overnight. In some cases plates were precoated with pDL (30–70 kDa) (Biomedical Technologies Inc., Stoughton, MA). pDL was dissolved in distilled water and applied to tissue culture plates for 1 hr. Excess pDL was withdrawn and the plates were allowed to air dry. Plates were rinsed with water and allowed to dry again. pDL plates were then coated with laminin as described above.

Neuroepithelial Cell Cultures Sprague–Dawley rat embryos were removed at Embryonic Day 10.5 (13–22 somites) and placed in a dish containing Ca/Mg-free

Clonal Cultures of Neuroepithelial Cells Cells were trypsinized and plated in 35-mm dishes coated with fibronectin at a dilution of 50 cells/dish. In some experiments, cells were plated at 10 cells/dish. Cells were allowed to settle for a period of 4 hr and then single cells were circled and their development followed in culture. In most experiments, clonal cultures were terminated after 12 days. In experiments to demonstrate oligodendrocyte development, clones were followed for 18–21 days. For replating individual clones, a glass cloning ring (Fisher Scientific, Pittsburg, PA) was placed around each clone and the cells were isolated by trypsinization for 1–2 min with 100 ml of trypsin/ EDTA solution. Cells were resuspended in fresh medium and an aliquot of cells (50–100) was replated onto fibronectin-coated culture dishes. Single cells were identified and circled with a grease pencil, and their development was followed as described above. Primary or replated clonal culture plates were usually triple labeled with the cell surface antigen and the appropriate secondary used in live cell cultures. Clones were then fixed in 4% paraformaldehyde for 10 min and processed sequentially for the other antigens. The diamino benzidine (DAB, Sigma) reaction to detect horseradish peroxidase-labeled secondaries was always performed after all other staining had been completed, as we observed reduced staining with some antigens if the clones were processed for DAB histochemistry first.

Generation of FGF-Dependent Neurospheres Neuroepithelial cells cultured in nondifferentiating conditions for a period of 5 days were harvested by trypsinization and replated onto bacterial dishes (Intermountain Scientific, Bountiful, UT) in neuroepithelial culture medium (see above) at a density of 5000 cells/dish. After 7 to 10 days, cells had divided and formed floating aggregates. Individual neurospheres were harvested and replated onto laminin-coated dishes in the absence of CEE. After 10 days in differentiation conditions, neurospheres were analyzed by triplelabel histochemistry as described above. To grow neurospheres as adherent cells, neurospheres were har-

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

221

NEP Cell Differentiation

TABLE 4

Antibody/kind Rat 401 (nestin) Anti-NCAM Anti-b-111 tubulin Anti-Neurofilament Anti-ChAT Anti-p75 (IgG192) Anti-GFAP Anti A2B5 Anti-Gal-C Anti-O4 Anti-O1

Dilution/solution

Mouse (IgG) Mouse (IgG) Mouse (IgG1) Mouse (IgG2) Goat (IgG) Mouse (IgG) Rabbit (IgG) Mouse (IgM) Mouse (IgG) Mouse (IgM) Mouse (IgM)

1:1 1:3 1:100 1:100 1:100 1:3 1:500 1:3 1:3 1:1 1:3

DSHB DSHB Sigma Sigma Chemicon DSHB Accurate BMB BMB BMB BMB

Antigen recognized

Cell type recognized

Intermediate filament Polysialated N-CAM Intermediate filament Neurofilament 68 Choline acetyl transferase Neurotrophin receptor p75 Glial fibrillary acid Ganglioside Galactocerebroside Sulfatide Galactocerebroside

Stern cells, oligo precursors Neurons Neurons Neurons Motoneurons Motoneurons Astrocytes Oligodendrocytes and Precursors Oligodendrocytes and Precursors Oligodendrocytes Oligodendrocytes

Note. The antibodies, their source, and the epitope they recognize are listed. Column four lists the cell types recognized by each of the antibodies in dissociated spinal cord culture. Note that A2B5 has been shown to recognize neurons under some culture conditions, nestin has been shown to label some astrocytes (Dahlstrand et al., 1995), and p75 may label a subclass of astrocytes (Hutton et al., 1992). We, however, do not detect such labeling in our cultures.

vested and replated onto fibronectin-coated dishes in nondifferentiating conditions (with FGF and CEE). Adherent cells were processed for immunocytochemistry as described. To generate secondary neurospheres, large individual neurospheres were harvested and dissociated by trituration using a fire-polished Pasteur pipet. Cells were counted, 100 cells were plated into a bacterial dish, and spheres were allowed to develop for 10 to 14 days.

body in blocking buffer for another hour. Cultures were rinsed with three changes of PBS. Double labeling experiments were performed by simultaneously incubating cells in appropriate combinations of primary antibodies followed by non-cross-reactive secondary antibodies. In some experiments, cultures were counter stained with DAPI (Hoechst dye 33258) at a concentration of 5 mg/ml to stain the nuclei of all cells.

Generation of Neurons, Oligodendrocytes, and Astrocytes

ACKNOWLEDGMENTS

Neuroepithelial cells cultured under nondifferentiating conditions for a period of 5 days were harvested by trypsinization and replated onto dishes sequentially coated with poly-D-lysine (1 mg/ ml) and laminin (0.05 mg/ml) or laminin (0.25 mg/ml) alone in neuroepithelial culture medium. For neuronal and oligodendrocyte differentiation, the medium consisted of neuroepithelial culture medium described above with the omission of 10% CEE. For glial differentiation, cells were plated on laminin, and 10% fetal calf serum was substituted for 10% CEE. Differentiation was assayed 5 days or 9 days after replating (as detailed under Results).

This work was supported by a grant from the Muscular Dystrophy Association and a NIH FIRST award to M.S.R. A.K. was supported by a graduate student fellowship from the Department of Neurobiology and Anatomy. The authors thank Drs. Margot-Mayer Proschel, M. Noble, Sheryl Scott, and T. N. Parks for their critical reading of the manuscript. M.S.R. gratefully acknowledges the constant support of Dr. S. Rao through all phases of this project.

Immunocytochemistry

Adamson, E. D., and Meek, J. (1984). The ontogeny of epidermal growth factor receptors during mouse development. Dev. Biol. 103, 62–70. Altman, J., and Bayer, S. (1984). The development of the rat spinal cord. Adv. Anat. Embryol. Cell Biol. 85, 32–46. Anderson, D. J. (1989). The neural crest lineage problem: Neuropoiesis? Neuron 3, 1–12. Artinger, K. B., Fraser, S., and Bronner-Fraser, M. (1995). Dorsal and ventral types can arise from common neural tube progenitors. Dev. Biol. 172, 591–601. Bannerman, P. G., and Pleasure, D. (1993). Protein growth factor requirements of rat neural crest cells. J. Neurosci. Res. 36, 46– 57. Bartlett, P., Reid, H., Bailey, K., and Bernard, O. (1988). Immortalization of mouse neural precursor cells by the c-myc oncogene. Proc. Natl. Acad. Sci. USA 85, 3255–3259.

The antisera used and their concentrations are summarized in Table 4. All secondary antibodies were obtained from Jackson Immunologicals (Westgrove, PA) and used as per manufacturers recommendations. Staining procedures were as described in Stemple and Anderson (1992). In brief, staining for cell surface antigens was carried out in cultures of living cells. For neurofilament proteins, GFAP, and b-111 tubulin cells were fixed with acid-ethanol. For other intracellular antigens, cultures were fixed in 4% formaldehyde for 15 min. For BRDU immunocytochemistry, cells were further permeabilized, as described previously (Memberg and Hall, 1995). Cultures were incubated with the primary antibody in blocking buffer (PBS, 1 mg/ml BSA, 0.5% Triton X-100, and 1% goat serum) for a period of 1 hr, rinsed with phosphate-buffered saline (PBS), and incubated with a species specific secondary anti-

REFERENCES

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

222

Kalyani, Hobson, and Rao

Bottenstein, J. E., and Sato, G. (1979). Growth of a rat neuroblastoma cell line in serum free supplemented medium. Proc. Natl. Acad. Sci. USA 76, 514–517. Boisseau, S., Poirier, V., Semonin, O., Leconte, L., Santha, M., Poujeol, C., Rougon, G., and Simonneau, M. (1991). A mammalian in vitro model to study gangliogenesis from neural crest cells. J. Physiol. 85, 117–122. Bronner-Fraser, M., and Fraser, S. E. (1988). Cell lineage analysis shows multipotentiality of some avian neural crest cells. Nature 355, 161–164. Bronner-Fraser, M., and Fraser, S. E. (1989). Developmental potentials of avian trunk neuralcrest cells in situ. Neuron 3, 755–766. Camu, W., and Henderson, C. E. (1992). Purification of embryonic rat motoneurons by panning on a monoclonal antibody to the low-affinity NGF receptor. J. Neurosci. Methods 44, 59–70. Cepko, C. l. (1988). Immortalization of neural cells via retrovirus mediated oncogengene transduction. Annu. Rev. Neurosci. 12, 47–65. Chen, E. W., and Chiu, A. Y. (1992). Early stages in the development of spinal motor neurons. J. Comp. Neuro. 320, 291–303. Dahlstrand, J., Lardelli, M., and Lendahl, U. (1995). Nestin mRNA expression correlates with the central nervous system progenitor state in many, but not all, regions of the developing nervous system. Dev. Brain. Res. 84, 109–129. Davis, A. A., and Temple, S. (1994). A self renewing multipotential stem cell in embryonic rat cerebral cortex. Nature 372, 263–266. Dodd, J., Morton, S. B., Karagogeos, D., Yamamoto, M., and Jessell, T. M. (1988). Spatial regulation of axonal glycoprotein expression on subsets of embryonic spinal neurons. Neuron 1, 105–116. Drago, J., Murphy, M., Carroll, S. M., Harvey, R. P., and Bartlett, P. F. (1991). Fibroblast growth factor-mediated proliferation of central nervous system precursors depends on endogenous production of insulin-like growth factor 1. Proc. Natl. Acad. Sci. USA 88, 2199–2203. Ericson, J., Thor, S., Edlund, T., Jessell, T. M., and Yamada, T. (1992). Early stages of motor neuron differentiation revealed by expression of homeobox gene islet-1. Science 256, 1555–1559. Ford, M. D., Bartlett, P. F., and Nurcombe, V. (1994). Co-localization of FGF-2 and a novel heparan sulphate proteoglycan in embryonic mouse brain. Neuroreport 5, 565–568. Galileo, D. S., Grey, G. E., Owens, G. C., Majors, J., and Sanes, J. R. (1990). Neurons and glia arise from a common progenitor in chick optic tectum: demonstration with two retroviruses and cell type specific antibodies. Proc. Natl. Acad. Sci. USA 87, 458–462. Groves, E. A., Williams, B. P., Li, D., Freidrich, A., and Price, J. (1993). Multiple restricted lineages in the embryonic rat cerebral cortex. Development 117, 553–561. Groves, A. K. (1992). The Derivation and Characterization of Conditionally Immortal Cell Lines from the Murine Central Nervous System. Ph.D. Thesis, Council for National Academic Awards, Great Britain. Gritti, A., Parati, E. A., Cova, L., Frolichsthal, P., Galli, R., Wanke, E., Faravelli, L., Morassutti, D. J., Roisen, F., Nickel, D. D., and Vescovi, A. (1996). Multipotential stem cells from the adult mouse brain proliferate and self-renew in response to basic fibroblast growth factor. J. Neurosci. 16, 1091–1100. Hamburger, V. (1948). The mitotic patterns in the spinal cord of the chick embryo and their relationship to the histogenetic process. J. Comp. Neurol. 88, 221–284. Hirano, M., and Goldman, J. E. (1988). Gliogenesis in the rat spinal cord: Evidence for the origin of astrocytes and oligodendrocytes from radial precursors. J. Neurosci. Res. 21, 155–167.

Hutton, L. A., de, Vallis, J., and Perez-Polo, J. R. (1992). Expression of p75 NGF-R, TrkA and TrkB mRNA in rat C6 Glioma and type 1 astrocyte cultures. J. Neurosci. Res. 32, 375–383. Kilpatrick, T. J., and Bartlett, B. F. (1993). Cloning and growth of multipotential neural precursors: Requirements for proliferation and differentiation. Neuron 10, 255–265. Kilpatrick, T. J., and Bartlett, B. F. (1995). Cloned multipotential precursors from the mouse cerebrum require FGF-2 whereas glial restricted precursors are stimulated by either FGF-2 or EGF. J. Neurosci. 15, 3653–3661. Kitchens, D. L., Snyder, E. Y., and Gottlieb, D. I. (1994). FGF and EGF are mitogens for immortalized neural progenitors. J. Neurobiol. 25, 797–807. Lazaar, L. M., and Blum, M. (1992). Regional distribution and developmental expression of epidermal growth factor and transforming growth factor-a mRNA in mouse brain by quantitative nuclease protection assay. J. Neurosci. 12, 1688–1697. Leber, S. M., Breedlove, S. M., and Sanes, J. R. (1990). Lineage, arrangement, and death of clonally related motoneurons in chick spinal cord. J. Neurosci. 10, 2451–2462. Lillien, L. E., and Raff, M. C. (1990). Analysis of the cell-cell interactions that control type-2 astrocyte development in vitro. Neuron 4, 525–534. Luskin, M. B., Pearlman, A. L., and Sanes, J. R. (1988). Cell lineage in the cerebral cortex of the mouse studied in vivo and in vitro with a recombinant retrovirus. Neuron 1, 635–647. McConnell, S. K. (1991). The generation of neuronal diversity in the central nervous system. Annu. Rev. Neurosci. 525, 269–300. Memberg, S. P., and Hall, A. K. (1995). Dividing neuron precursors express neuron-specific tubulin. J. Neurobiol. 27, 26–43. Miller, R. H., and Szigeti, V. (1991). Clonal analysis of astrocyte diversity in neonatal spinal cord cultures. Development 115, 133–142. Murphy, M., Drago, J., and Bartlett, P. F. (1990). Fibroblast growth factor stimulates the proliferation and differentiation of neural precursor cells in vitro. J. Neurosci. Res. 25, 463–475. Nornes, H. O., and Das, G. D. (1974). Temporal pattern of neurogenesis in the spinal cord of rat.1. Time and sites of origin and migration and settling patterns of neuroblasts. Brain Res. 73, 121–138. Nurcombe, V., Ford, M. D., and Wildshut, J. A. (1993). Developmental regulation of neural response to to FGF-1 and FGF-2 by heparan sulfate proteoglycan. Science 260, 103–106. Ono, K., Bansal, R., Payne, J., Rutishauser, U., and Miller, R. H. (1995). Early development and dispersal of oligodendrocytes precursors in the embryonic chick spinal cord. Development 121, 1743–1754. Phelps, P. E., Barber, R. P., and Vaughn, J. E. (1988). Generation patterns of four groups of cholinergic neurons in rat cervical spinal cord: A combined tritiated thymidine autoradiographic and choline acetyltransferase immunocytochemical study. J. Comp. Neurol. 273, 459–472. Phelps, P. E., Barber, R. P., Brennan, L. A., Maines, V. M., Salvaterra, P. M., and Vaughn, J. E. (1990). Embryonic development of four subsets of cholinergic neurons in rat cervical spinal cord. J. Comp. Neurol. 291, 9–26. Price, J., and Thurlow, L. (1988). Cell lineage in the rat cerebral cortex: A study using retroviral mediated gene transfer. Development 104, 473–482. Pringle, N. P., and Richardson, W. D. (1993). A singularity of PDGF alpha-receptor expression in the dorsoventral axis of the neural

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba

223

NEP Cell Differentiation

tube may define the origin of the oligodendrocyte lineage. Development 117, 525–533. Raff, M. (1989). Glial cell diversification in the rat optic nerve. Science 243, 1450–1455. Rao, M. S., and Anderson, D. J. (1996). The immortalization of a neural crest stem cell. J. Neurobiol., in press. Ray, J., and Gage, F. (1994). Spinal cord neuroblasts proliferate in response to basic fibroblast growth factor. J. Neurosci. 14, 3548– 3564. Reynold, B. A., Tetzlaff, W., and Weiss, S. (1992). A multipotent EGF-responsive striatal embryonic progenitor cell produces neurans and astrocytes. J. Neurosci. 12, 4565–4574. Reynold, B. A., and Weiss, S. (1996). Clonal and population analysis demonstrate that an EGF-responsive mammalian embryonic CNS precursor is a stem cell. Dev. Biol. 175, 1–13. Sanes, J. R. (1989). Analysing cell lineage with a recombinant retrovirus. Trends Neurosci. 12, 21–28. Santa-Ollala, J., and Covarrubias, L. (1995). Epidermal growth factor (EGF), Transforming growth factor alpha (TGF-alpha) and basic fibroblast growth factor (bFGF0 differentially influence neural precursor cells of mouse embryonic mesencephalon. J. Neurosci. Res. 42, 172–183. Stemple, D. L., and Anderson, D. J. (1992). Isolation of a stem cell for neurons and glia from the mamalian neural crest. Cell 71, 973–985. Sommers, L., Shah, N., Rao, M. S., and Anderson, D. J. (1995). Cellular function of the bHLH transcription factor MASH1 in mammalian neurogenesis. Neuron 15, 1245–1258.

Temple, S., and Davis, A. (1994). Isolated rat cortical progenitor cells are maintained in division in vitro by membrans associated factors. Development 120, 999–1008. Timsit, S., Martinez, S., Allinquant, B., Peyron, F., Puelles, L., and Zalc, B. (1995). Oligodendrocytes originate in a restricted zone of the embryonic ventral tube defined by DM-20 mRNA expression. J. Neurosci. 12, 1012–1024. Turner, D. L., and Cepko, C. (1987). Cell lineage in the retina: A common progenitor for neurons and glia persists late in development. Nature 328, 131–136. Vescovi, A. L., Reynold, B. A., Fraser, D. D., and Weiss, S. (1993). bFGF regulates the proliferative fate of unipotent (neuronal) and bipotent (neuron/astroglial) EGF-generated CS progenitor cells. Neuron 11, 951–966. Warf, B. C., Fok-Seang, J., and Miller, R. H. (1991). Evidence for the ventral origin of oligodendrocytic precursors in the rat spinal cord. J. Neurosci. 11, 2477–2488. Williams, B. P. (1995). Precursor cell types in the germinal zone of the cerebal cortex. Bioessays 17, 391–393. Williams, B. P., and Price, J. (1995). Evidence for multiple precursor cell types in the embryonic rat cerebral cortex. Neuron 14, 1181– 1188. Yamada, T., Pfaff, S., Edlund, Y., and Jessell, T. (1993). Control of cell pattern in the neural tube: Motor neuron induction by diffusible factors from notochord and floor plate. Cell 73, 673–686. Received for publication December 6, 1996 Accepted April 16, 1997

Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.

AID

DB 8592

/

6x25$$$141

06-04-97 22:03:06

dba