Transplantation of embryonic neuroectodermal progenitor cells into the site of a photochemical lesion: immunohistochemical and electrophysiological analysis.

Abstract

GFP labeled/NE-4C neural progenitor cells cloned from primary neuroectodermal cultures of p53- mouse embryos give rise to neurons when exposed to retinoic acid in vitro. To study their survival and differentiation in vivo, cells were transplanted into the cortex of 6-week-old rats, 1 week after the induction of a photochemical lesion or into noninjured cortex. The electrophysiological properties of GFP/NE-4C cells were studied in vitro (8-10 days after differentiation induction) and 4 weeks after transplantation using the whole-cell patch-clamp technique, and immunohistochemical analyses were carried out. After transplantation into a photochemical lesion, a large number of cells survived, some of which expressed the astrocytic marker GFAP. GFP/GFAP-positive cells, with an average resting membrane potential (Vrest) of -71.9 mV, displayed passive time- and voltage-independent K+ currents and, additionally, voltage-dependent A-type K+ currents (KA) and/or delayed outwardly rectifying K+ currents (KDR). Numerous GFP-positive cells expressed NeuN, betaIII-tubulin, or 68 kD neurofilaments. GFP/betaIII-tubulin-positive cells, with an average Vrest of -61.6 mV, were characterized by the expression of KA and KDR currents and tetrodotoxin-sensitive Na+ currents. GFP/NE-4C cells also gave rise to oligodendrocytes, based on the detection of oligodendrocyte-specific markers. Our results indicate that GFP/NE-4C neural progenitors transplanted into the site of a photochemical lesion give rise to neurons and astrocytes with membrane properties comparable to those transplanted into noninjured cortex. Therefore, GFP/NE-4C cells provide a suitable model for studying neuro- and gliogenesis in vivo. Further, our results suggest that embryonic neuroectodermal progenitor cells may hold considerable promise for the repair of ischemic brain lesions.

Full text

Page 1

Transplantation of Embryonic Neuroectodermal
Progenitor Cells into the Site of a Photochemical
Lesion: Immunohistochemical and
Electrophysiological Analysis
Miroslava Ande ˇrova ´,1,3Sˇa ´rka Kubinova ´,1,3Marti Jelitai,4Helena Nepras ˇova ´,1–3
Kater ˇina Glogarova ´,1–3Iva Prajerova ´,1,2Lucie Urdzı ´kova ´,1,3
Alexandr Chva ´tal,1–3Eva Sykova ´,1–3
1Institute of Experimental Medicine, Academy of Sciences of the Czech Republic, Prague, Czech Republic
2Department of Neuroscience, Charles University, Second Medical Faculty, Prague, Czech Republic
3Center for Cell Therapy and Tissue Repair, Charles University, Prague, Czech Republic
4Department of Neural Cell Biology, Institute of Experimental Medicine of the Hungarian
Academy of Sciences, Budapest, Hungary
Received 14 October 2005; accepted 24 January 2006; accepted 14 February 2006
ABSTRACT:
cells cloned from primary neuroectodermal cultures of
p53–mouse embryos give rise to neurons when exposed
to retinoic acid in vitro. To study their survival and differ-
entiation in vivo, cells were transplanted into the cortex of
6-week-old rats, 1 week after the induction of a photo-
chemical lesion or into noninjured cortex. The electro-
physiological properties of GFP/NE-4C cells were studied
in vitro (8–10 days after differentiation induction) and
4 weeks after transplantation using the whole-cell patch-
clamp technique, and immunohistochemical analyses
were carried out. After transplantation into a photochem-
ical lesion, a large number of cells survived, some of which
expressed the astrocytic marker GFAP. GFP/GFAP-posi-
GFP labeled/NE-4C neural progenitor
tive cells, with an average resting membrane potential
(Vrest) of –71.9 mV, displayed passive time- and voltage-in-
dependent K+currents and, additionally, voltage-depend-
ent A-type K+currents (KA) and/or delayed outwardly
rectifying K+currents (KDR). Numerous GFP-positive
cells expressed NeuN, bIII-tubulin, or 68 kD neurofila-
ments. GFP/bIII-tubulin-positive cells, with an average
Vrestof –61.6 mV, were characterized by the expression
of KAand KDRcurrents and tetrodotoxin-sensitive Na+
currents. GFP/NE-4C cells also gave rise to oligodendro-
cytes, based on the detection of oligodendrocyte-specific
markers. Our results indicate that GFP/NE-4C neural
progenitors transplanted into the site of a photochemical
lesion give rise to neurons and astrocytes with membrane
properties comparable to those transplanted into nonin-
jured cortex. Therefore, GFP/NE-4C cells provide a suita-
ble model for studying neuro- and gliogenesis in vivo. Fur-
ther, our results suggest that embryonic neuroectodermal
progenitor cells may hold considerable promise for the
repair of ischemic brain lesions.
' 2006 Wiley Periodicals, Inc.
J Neurobiol 66: 1084–1100, 2006
Keywords: ischemia; astrocytes; neurons; oligodendro-
cytes; potassium and sodium currents; membrane pro-
perties; patch clamp
Correspondence to: M. Ande ˇrova ´ (anderova@biomed.cas.cz).
' 2006 Wiley Periodicals, Inc.
Published online 12 June 2006 in Wiley InterScience (www.
interscience.wiley. com).
DOI 10.1002/neu.20278
Grant sponsor: Grant Agency of the Czech Republic; Grant num-
ber: 305/03/1172, 305/04/1293.
Grant sponsor: Grant Agency of Czech Academy of Sciences;
Grant number: KJB5039401.
Grant sponsor: Ministry of Education, Youth and Sports of the
Czech Republic; Grant number: AVOZ 50390512, 1M0021620803,
LC554.
1084

Page 2

The limited ability of the adult central nervous sys-
tem (CNS) to generate new neurons and glial cells is
a constraining factor in its self-repair capacity in de-
generative diseases, after stroke and brain or spinal
cord injury. Embryonic stem (ES) cells can replace
damaged cells. ES cells are genetically normal, pluri-
potent, capable of indefinite replication, and have
been derived from several vertebrate species, includ-
ing mice and humans. They are capable of differenti-
ating into neural cells in vitro when induced by reti-
noic acid (RA) (Schlett and Madarasz, 1997; Schlett
et al., 1997; Liu et al., 2000) or fibroblast growth fac-
tor and further treated with brain-derived neurotro-
phic factor (BDNF) and/or transforming growth fac-
tor ? (TGF-?; Park et al., 2004). Several members of
the transforming growth factor ? (TGF-?) superfam-
ily have been reported to promote astroglial differen-
tiation from CNS precursor cells (Satoh et al., 2000).
It was also shown that it is possible to influence the
development of ES cells towards neurons by co-cul-
ture with stromal cells (Yoshizaki et al., 2004) or
astrocytes (Nakayama et al., 2003). ES cells are able
to survive after transplantation into the injured or
damaged brain (Englund et al., 2002; Chu et al.,
2003; Yoshizaki et al., 2004) or spinal cord (Liu
et al., 2000), and after intravenous injection ES cells
can migrate into brain or spinal cord lesions (Chu
et al., 2003; Fujiwara et al., 2004; Jendelova et al,
2004) and differentiate into neurons, astrocytes, and
oligodendrocytes. There has been an increasing num-
ber of functional studies on ES cell–derived neurons
in vitro (Piper et al., 2000; Mistry et al., 2002; Ben-
ninger et al., 2003; Calhoun et al., 2003; Balasubra-
maniyan et al., 2004); however, there have been very
few studies on the functional properties of ES cells
after transplantation into noninjured CNS tissue
(Englund et al., 2002; Kim et al., 2002; Wernig et al.,
2004; Uchida et al., 2005). In these studies grafted
cells gave rise to neurons that were able to generate
action potentials and received functional excitatory
and inhibitory synaptic inputs from neighboring cells.
Nonetheless, studies addressing the functional prop-
erties of ES cell–derived progeny after transplanta-
tion into the adult CNS under pathological conditions
are scarce (Ruschenschmidt et al., 2005; Hayashi
et al., 2006).
NE-4C neural stem cells cloned from primary neu-
roectodermal cultures of p53?mouse embryos and la-
beled with GFP give rise to neurons and astrocytes
when exposed to retinoic acid in vitro (Schlett and
Madarasz, 1997; Schlett et al., 1997). These noncom-
mitted, proliferating neuroectodermal progenitors can
integrate into the brain tissue at the time and site of
tissue genesis (Demeter et al., 2004). In this study we
examined the differentiation and survival of NE-4C
cells induced first in vitro by RA and then trans-
planted either into the noninjured cortex or into the
site of a photochemical lesion, a model of ischemic
brain injury. Their electrophysiological properties
were studied using the patch-clamp technique, and
immunohistochemical analysis was used to identify
the cells during differentiation.
METHODS
Cell Culture
The NE-4C cell line was cloned from primary neuroecto-
dermal cultures of p53?mouse embryos and labeled with
green fluorescent protein (GFP/NE-4C; Schlett et al., 1997;
Schlett and Madarasz, 1997; Demeter et al., 2004). Cells
were grown in a humidified atmosphere with 5% CO2at
378C in minimum essential medium (Sigma-Aldrich),
supplemented with 10% fetal calf serum (Gibco, Paisley,
Scotland), 4 mM glutamine, 40 ?g/mL gentamycine, and
400 ?g/mL Geneticine (Sigma-Aldrich). Subconfluent cul-
tures were split by trypsinization with 0.05% trypsin in
phosphate-buffered saline (PBS) and transferred into poly-
L-lysine (PLL, Sigma-Aldrich) coated dishes at a cell den-
sity of 104cells/cm2. After one or two passages, neural dif-
ferentiation was induced.
Differentiation of GFP/NE-4C
Cells In Vitro
GFP/NE-4C cells were plated at a density of 2.5 ? 104
cells/cm2onto PLL-coated cover slips and cultured in 5%
fetal calf serum medium with RA (10?7M) for 2 days after
plating. Ten days after the induction of differentiation, the
cells were analyzed for Kþand Naþcurrent expression
using the patch-clamp method; after fixation, the cells were
analyzed for the expression of astrocyte-, oligodendrocyte-,
and neuron-specific markers.
Photochemical Lesion
A model of thrombotic stroke, the photochemical lesion
was chosen as a model of cortical injury suitable for trans-
plantation studies (Watson et al., 1985). This model is
based on a photochemical reaction inducing thrombosis
in vivo and leading to a cerebral infarction. Wistar rats
(150–200 g) were anesthetized with isoflurane (2% isoflur-
ane in air; Forane, Abbott Laboratories Ltd, UK). Rose
Bengal (Sigma-Aldrich), a potent photosensitizing dye, was
injected intravenously into the femoral vein (8 mg/100 g).
The scalp was incised to expose the skull surface, and
the area of the skull above the right somatosensory
cortex (bregma, 2 mm; lateral, 2.5 mm) was exposed to a
laser (Melles Griot, 1.5 mW, 543.5 nm) for 10 min. The
skin overlying the cranium was then sutured. The rats
were left to recover and returned to their cages. The area of
Neural Progenitors in Cortical Lesion1085
Journal of Neurobiology. DOI 10.1002/neu

Page 3

the infarction extended to approximately 1.0 to 1.5 mm in
diameter.
Cell Transplantation
After 2 days of treatment with retinoic acid, the cells were
harvested by trypsinization, centrifuged, diluted to their
final concentration (1–2 ? 105/?L) in PBS, and immedi-
ately used for transplantation into noninjured (control) and
lesioned cortex. The transplantation was carried out 7 days
after the induction of the photochemical lesion. Rats were
anesthetized with isoflurane and mounted in a stereotaxic
frame. With aseptic technique, a small hole was drilled into
the skull above the lesion and 3 ?L of a cell suspension was
slowly injected over a 15-min period into the lesion using a
Hamilton syringe. For immunosuppression, Depo-Medrol
(methylprednisolone; Pharmacia, Puurs, Belgium) was
administered at a single dose of 2 mg/1 week. Four weeks
after transplantation, brain slices were prepared and used in
patch-clamp experiments in situ, or the rats were used for
immunohistochemical analysis.
Preparation of Acute Brain Slices
The rats were sacrificed by decapitation under isoflurane
anesthesia at 4 weeks after cell transplantation. The brain
was quickly dissected and placed in artificial cerebrospinal
fluid (ACF) at 6 to 88C. For patch-clamp recording, the
brain was hemisected and glued with tissue adhesive (Elec-
tron Microscopy Science) to a Teflon plate. Transverse
250?m-thick slices were made using an automatic oscillat-
ing tissue slicer (HM 650 V, Microm Int. GmbH, Walldorf,
Germany). The slices were kept at 22 to 258C for up to 6 h
in ACF containing (in mM): NaCl 117.0, KCl 3.0, CaCl2
1.5, MgCl21.3, Na2HPO41.25, NaHCO435.0, D-glucose
10.0, osmolarity 300 mmol/kg. The solution was continu-
ously gassed with a mixture of 95% O2and 5% CO2to
maintain a final pH of 7.4. Osmolarity was measured using
a vapor pressure osmometer (Vapro 5520, Wescor Inc.).
Patch-Clamp Recordings
Cell membrane currents were recorded with the patch-
clamp technique in the whole-cell configuration (Hamill
et al., 1981). Recording pipettes with a tip resistance of 4 to
6 MO were made from borosilicate capillaries (Ru ¨ckl &
Sons, Otvovice, Czech Republic) using a Brown-Flaming
micropipette puller (P-97, Sutter Instruments Company).
Electrodes were filled with a solution containing (in mM):
KCl 130.0, CaCl20.5, MgCl22.0, EGTA 5.0, HEPES 10.0.
The pH was adjusted with KOH to 7.2. For immunohisto-
chemical identification after patch-clamp measurements,
the recorded cells were filled with either Lucifer Yellow
(LY, Sigma-Aldrich) or Alexa-Fluor hydrazid 594 (Molec-
ular Probes, Eugene, OR) by dialyzing the cytoplasm with
the patch pipette solution. All recordings were made in sli-
ces perfused with ACF at a temperature of 22 to 258C. The
slices were placed in a chamber mounted on the stage of a
fluorescence
Germany) and fixed using a U-shaped platinum wire with a
grid of nylon threads. The cells were approached by the
patch electrode using an INFRAPATCH system (Luigs &
Neumann, Ratingen, Germany). The cells and the recording
electrodes were imaged with a digital camera (Axiocam
HRc, Carl Zeiss, Germany). Current signals were amplified
with an EPC-9 amplifier (HEKA Elektronik, Lambrecht/
Pfalz, Germany), lowpass-filtered at 3 kHz and sampled at
5 kHz by an interface connected to an AT-compatible com-
puter system, which also served as a stimulus generator.
Data acquisition, storage, and analysis were performed with
TIDA (HEKA Elektronik, Lambrecht/Pfalz, Germany).
microscope (Axioskop FX,CarlZeiss,
Electrophysiological Measurements
and Protocols
Resting membrane potential (Vrest) was measured by
switching the EPC-9 amplifier to the current-clamp mode.
The holding potential was ?70 mV.
Membrane capacitance (Cm) was determined from the
current transients elicited by a 10 mV test pulse depolariz-
ing the cell membrane from ?70 mV to ?60 mV.
Current patterns were obtained by clamping the cell
membrane from a holding potential of either ?70 mV or
?50 mV to values ranging from ?160 mV to þ20 mV and
?140 mV to þ40 mV, respectively, at intervals of 10 mV.
Pulse duration was 50 ms. In order to isolate voltage-gated
KDRand KIRcurrent components, the voltage step from
?70 mV to ?60 mV was used to subtract the time- and
voltage-independent currents. To activate delayed out-
wardly rectifying Kþcurrents the cells were held at ?50
mV, and the amplitude of the KDRcurrent was measured at
þ40 mV at the end of the pulse. The A-type Kþcurrent
component was isolated by subtracting current traces
clamped at ?110 mV from those clamped at ?50 mV, and
its amplitude was measured at the peak value. The ampli-
tudes of inwardly rectifying Kþcurrents were measured at
?160 mV at the end of the pulse. Tetrodotoxin (TTX)-sen-
sitive Naþcurrents were isolated by subtracting the current
traces measured in 1 ?M TTX-containing solution from
those measured under control conditions. Naþcurrent
amplitudes were measured at the peak value.
Immunohistochemistry
Cell Culture. GFP/NE-4C cells attached to PLL-coated
cover slips were fixed postrecording in 4% paraformalde-
hyde in 0.1M phosphate buffer (PB, pH 7.4) for 30 min,
washed and kept in 0.1M PB at 58C for further processing.
Brain Slices. At 4 weeks after transplantation, the animals
were anesthetized (100 mg/kg sodium pentobarbital) and
perfused transcardially with 100 mL of saline, followed by
200 mL of 4% paraformaldehyde in 0.1M PB (pH 7.4).
Brains were dissected out and postfixed in paraformalde-
hyde solution overnight, then placed in 30% sucrose in
0.1M PB and allowed to sink for cryoprotection. Coronal
1086Ande ˇrova ´ et al.
Journal of Neurobiology. DOI 10.1002/neu

Page 4

slices, 40 ?m thick, were prepared using a microtome (HM
400, Microm Int. GmbH, Waldorf, Germany). For postre-
cording cell identification, 200 ?m slices were fixed with
4% paraformaldehyde in 0.1M PB (pH 7.4) overnight at
58C, then washed and kept in 0.1M PB at 58C. Free-floating
sections were incubated with specific antibodies for astro-
cytes, neurons, or oligodendrocytes.
To identify GFP/NE-4C cells as astrocytes, an antibody
directed against glial fibrillary acidic protein (GFAP) was
used. Slices were incubated overnight at 58C with a mouse
monoclonal antibody directed against GFAP and conjugated
with Cy3 (Sigma-Aldrich), diluted 1:200 in PBS containing
1% bovine serum albumin (BSA, Sigma-Aldrich) and 0.5%
Triton X-100 (Sigma-Aldrich). To identify cells as astrocytes
or their precursors, slices were incubated overnight at 58C
with a mouse monoclonal antibody directed against the ?-
subunit of calcium binding protein (S100?, Sigma-Aldrich),
diluted 1:200 in PBS containing 1% BSA and 0.5% Triton X-
100, followed by incubation with the secondary antibody,
goat anti-mouse IgG conjugated with Alexa-Fluor 594 or 633
(Molecular Probes). To identify newly generated or reactive
astrocytes, antibodies directed against GFAP (conjugated
with Cy3) and nestin (Chemicon, Temecula, CA) were
employed; nestin reactivity was visualized using a secondary
antibody, goat anti-mouse IgG, conjugated with Alexa-Fluor
594 or 633 (Molecular Probes).
To identify neurons, antibodies directed against neuron-
specific nuclear protein (NeuN), microtubule-associated
protein (Map2; Chemicon), ?III-tubulin (Exbio, Prague,
Czech Republic), or neurofilaments 68 kD (NF-68, Sigma-
Aldrich) were used. Primary antibodies were diluted as fol-
lows: anti-?III-tubulin 1/1000, anti-NeuN 1/2000, anti-
Map2 1/1000, and anti-NF-68 1/1000. To identify the for-
mation of presynaptic terminals, an antibody directed
against synaptophysin was used (dilution 1/200). The cul-
ture or tissue slices were incubated with primary antibodies
overnight at 58C followed by incubation with the secondary
antibody, goat anti-mouse IgG conjugated with Alexa-Fluor
594 or 633 (Molecular Probes).
To identify oligodendrocytes, antibodies directed against
myelin-oligodendrocyte specific protein (MOSP), RIP, O1,
and O4 (Chemicon) were employed. Primary antibodies
were diluted as follows: anti-MOSP 1/1000, anti-RIP 1/
2000, and anti-O1 and anti-O4 1/1000. The culture or the
tissue slices were incubated with primary antibodies over-
night at 58C followed by incubation with the secondary
antibody, goat anti-mouse IgG conjugated with Alexa-Fluor
594 or 633 (Molecular Probes) or goat anti-mouse IgM con-
jugated with Cy5 (Chemicon). After immunostaining the
slices were mounted using Vectashield mounting medium
(Vector Laboratories, Burlingame, CA), and were examined
using a LEICA TCS SP system spectral confocal micro-
scope equipped with an Ar/HeNe laser.
Statistical Analysis
The results are expressed as the mean 6 SEM. Statistical
analysis of the differences between groups was evaluated
using a t test. Values of p < 0.05 were considered signifi-
cant.
RESULTS
Membrane Properties of GFP/NE-4C
Cells after the Induction of Differentiation
In Vitro
Within 2 days after plating, nondifferentiated, nestin-
positive GFP/NE-4C cells adhered to the PLL sub-
strate and started to form extensions. From day 3 on,
neurons could be identified by the expression of ?III-
tubulin [Fig. 1(A)] or NF-68 (not shown). These cells
developed on the top of apparently nondifferentiated
(nestin-positive) cells and showed a clear neuronal
phenotype, with a small cell body (<20 ?m) and long
thin processes that made contact with other neurons.
Using the patch-clamp method in the whole-cell con-
figuration, we found that from day 6 on, these cells
displayed outwardly rectifying Kþcurrents, including
fast activating and inactivating A-type (KA) and
delayed outwardly rectifying currents [KDR, Fig.
1(B)], and moreover they expressed TTX-sensitive
Naþcurrents [INa, Fig. 1(C)]. The following mem-
brane parameters are summarized in Table 1: resting
membrane potential (Vrest), input resistance (IR),
membrane capacitance (Cm), and KDR, KA, and INa
current densities. In 19 (from a total number of 35)
GFP/?III-tubulin–positive cells, the application of
100 ?M GABA elicited a large inward current [Fig.
1(D)] sensitive to bicuculline (data not shown). No
response to 100 ?M glutamate was observed. Besides
GFP/?III-tubulin–positive cells, we also found GFP-
positive cells displaying passive, symmetrical nonde-
caying as well as decaying currents, but we were not
able to identify them either as astrocytes or oligoden-
drocytes based on GFAP- or MOSP-positive staining.
These data demonstrate that 6 to 10 days after the
induction of differentiation in vitro, functional KA,
KDR, and INachannels are present in GFP/NE-4C/
?III-tubulin- or NF-68-positive neurons and ? 50%
of these cells express functional GABAAreceptors.
No astrocyte or oligodendrocytes were identified.
Transplantation of GFP/NE-4C cells into
Unlesioned Cortex
The survival and differentiation of GFP/NE-4C cells
were studied in cortical tissue that was not subjected
to photochemical lesioning (control). Immunohisto-
chemical analysis of GFP/NE-4C cells transplanted
into noninjured cortex was performed in a total of
Neural Progenitors in Cortical Lesion1087
Journal of Neurobiology. DOI 10.1002/neu

Page 5

Figure 1
induction of differentiation by retinoic acid. (A) Photomicrographs of GFP-labeled cells (A, left),
after 10 days of differentiation in vitro, immunostained for ?III-tubulin (A, middle). Yellow color
indicates similar patterns in overlay (A, right). (B) Typical current profiles obtained by hyper- and
depolarizing the cell from ?140 to þ40 mV from a holding potential of ?50 mV (B1) and those
obtained by hyper- and depolarizing the cell from ?140 to þ40 mV from a holding potential of
?50 mV after a hyperpolarizing prepulse to ?110 mV (B2). The fast activating A-type Kþcurrent
shown in B3was isolated by subtracting the current traces in B1 from those shown in B2. The
resulting current/voltage (I/V) relationships for delayed rectifying Kþ(filled squares) and fast acti-
vating A-type Kþcurrents (filled circles) are shown in B4. (C) The membrane current pattern of
GFP/?III-tubulin–positive cells prior to the application of tetrodotoxin (TTX, control, C1) and in
the presence of 1 ?M TTX (C2). The TTX-sensitive current was obtained by subtracting the cur-
rents after TTX application from control currents (C3). The resulting I/V relationship for the TTX-
sensitive current is shown in C4(filled triangles). (D) GABA-evoked currents recorded at a holding
potential of ?70 mV (left). The membrane currents were obtained by clamping the cell membrane
potential to different values, by rectangular voltage steps, from the holding potential of ?70 mV to
potentials of ?140, ?105, ?35, 0, and þ35 mV (see the inset for voltage protocol). The resulting I/
V relationships for the control traces (filled circles) and the traces in the presence of GABA (empty
circles) are shown in D (right).
Neuroectodermal stem cells express a typical neuronal current pattern in vitro after the

Page 6

7 rats. The animals were sacrificed 4 weeks after
transplantation. A large number of GFP-positive cells
survived, and they filled the entire track produced by
the injection. No migration of GFP-positive cells into
the cortex was observed, but when the transplanted
cells reached the corpus callosum, they started to
migrate from the site of injection. The site of injec-
tion was surrounded by endogenous reactive astro-
cytes and partially filled with GFP/GFAP-positive
cells [Fig. 2(A,B)]. GFP/NE-4C–derived neurons
were identified by NeuN–, Map2–, ?III-tubulin–, and
NF-68–positive staining [Fig. 2(C)]. Their morphol-
ogy was characterized by a rounded cell body with a
diameter of 15 to 20 ?m and one or two short pro-
cesses, and they expressed a typical neuronal current
pattern: TTX-sensitive Naþcurrents (data not shown)
and KDRand KAcurrents (Table 1). They were able
to fire action potentials, expressed as repetitive Naþ
channel opening during either a depolarizing pulse or
during depolarization evoked by 20 mM Kþapplica-
tion [Fig. 2(E,F)]. Their membrane properties are
summarized in Table 1. Based on GFAP and S100?
immunoreactivity, a large number of GFP cells dif-
ferentiated into astrocytes and displayed the morphol-
ogy typical of reactive astrocytes [Fig. 2(B)]. They
expressed a time- and voltage-independent Kþcur-
rent component, and additionally they displayed ei-
ther KAand KDRcurrents (further termed A1 astro-
cytes, n ¼ 7) or KDRcurrents only [A2 astrocytes, n
¼ 6, Fig. 2(D)]. No KIR current component was
detected in these cells. Their membrane properties
are summarized in Table 2. A small number of GFP/
NE-4C cells differentiated into oligodendrocytes
based on MOSP-positive staining [Fig. 2(B)], but no
GFP/O1- or GFP/RIP-positive cells were identified.
No GFP-positive cells were identified postrecording
as oligodendrocytes.
Fate of GFP/NE-4C Cells after
Transplantation into the Site of a
Photochemical Lesion
In order to study the development of photochemical
lesions without subsequent cell transplantation over
the same time course as transplanted photochemical
lesions and furthermore, to show that there is no sig-
nificant endogenous neuro- and gliogenesis in the
cortex that could replace damaged cells, a photo-
chemical lesion was induced in the cortex of 4 rats
that did not undergo subsequent cell transplantation
(termed PCL rats). These rats were sacrificed at 4 to 5
weeks after lesion induction.
In an additional 11 rats, we studied the fate of
transplanted GFP/NE-4C cells using immunohisto-
chemical and electrophysiological analysis. These
animals were subjected to a photochemical lesion,
and 1 week later they were transplanted with RA-
induced GFP/NE-4C cells. The rats were sacrificed 4
to 5 weeks after lesion induction, and the results were
compared with those of PCL rats and controls.
Neurons. In PCL rats, close to the photochemical
lesion the cortex showed typical staining for NeuN,
?III-tubulin and NF-68 as observed in noninjured ani-
mals. At the site of the photochemical lesion, there
was a cavity partially filled with necrotic tissue, and
no positive staining for NeuN, ?III-tubulin, or NF-68
was observed inside the photochemical lesion [Fig.
Table 1
Differentiation In Vitro (by Retinoic Acid) and 4 Weeks after Transplantation into Noninjured Cortex
(Control) or into the Site of a Photochemical Lesion (Lesion)
Membrane Properties of GFP/bIII-Tubulin–Positive Cells 8 Days after the Induction of
GFP/?III-Tubulin–
Positive CellsDifferentiated in vitro
ControlLesion
Vrest[mV]
Cm[pF]
IR [MO]
KDR/Cm[pA/pF]
KA/Cm[pA/pF]
INa/Cm[pA/pF]
n
?40.6 6 1.6
55.7 6 3.9
812.1 6 47.2
12.7 6 3.7
12.8 6 3.5
17.2 6 2.9
55
?51.6 6 5.6*
22.5 6 5.2**
435.4 6 65.7**
39.6 6 9.6*
46.5 6 13.2*
53.9 6 18.5***
8
?61.6 6 4.0***
35.1 6 6.9*
460.6 6 38.1***
35.8 6 7.2**
46.2 6 10.6***
50.9 6 10.1***
18
Vrest, resting membrane potential; Cm, membrane capacitance; IR, input resistance; KDR/Cm, KA/Cm, and INa/Cm, KDR, KA, and INacurrent den-
sities; n, number of cells.
Astericks indicate significant differences between neurons derived in vitro and those derived in noninjured cortex or in lesioned cortex.
Note that there were no significant differences between GFP/?III-tubulin–positive cells in control and lesioned cortex.
*p < 0.05.
**p < 0.01.
***p < 0.001.
Neural Progenitors in Cortical Lesion1089
Journal of Neurobiology. DOI 10.1002/neu

Page 7

Figure 2
weeks after transplantation into noninjured cortex. (A) Coronal section of noninjured rat cortex
transplanted with GFP/NE-4C cells and immunostained for GFAP 4 weeks after transplantation.
(B) Higher magnification photomicrographs illustrate details of GFP/NE-4C cell immunoreactivity
for GFAP (top), S100? (middle), and MOSP (bottom) in the cortex. (C) Higher magnification pho-
tomicrographs illustrate details of GFP/NE-4C cell immunoreactivity (from the top) for NeuN,
?III-tubulin, Map2, and NF-68 in the cortex. Yellow color indicates double-stained cells. [D (top)]
The membrane current patterns of GFP/GFAP-positive cells were measured in response to voltage
steps from a holding potential of ?70 mV. To activate the currents, the membrane was clamped for
50 ms from the holding potential to de- and hyperpolarizing potentials ranging from ?160 to þ20
mV, in 10 mV increments (D, left, A1 astrocyte; right, A2 astrocyte). [D (middle)] The A-type Kþ
current was isolated by subtracting the current traces without a hyperpolarizing prepulse from those
with the prepulse (left), and the delayed outward rectifying Kþcurrent was isolated by passive cur-
rent subtraction (right). [D (bottom)] The morphology and immunohistochemical identification of
the recorded A1 and A2 astrocytes showing positive staining for GFAP. Arrowheads indicate
recorded cells. [E (top)] Membrane current patterns of GFP/?III-tubulin–positive cells (E, left)
were measured in response to voltage steps from a holding potential of ?70 mV. To activate the
currents, the membrane was clamped for 50 ms from the holding potential to increasing de- and
hyperpolarizing potentials ranging from ?160 to þ20 mV, in 10 mV increments. The morphology
and immunohistochemical identification (E, right) of a recorded GFP/NE-4C cell showing positive
staining for ?III-tubulin. Arrowheads indicate the recorded cell. (F) Typical current evoked in a
GFP/?III-tubulin–positive cell by the application of 20 mM Kþat a holding potential of ?70 mV.
The membrane currents were obtained by clamping the cell membrane potential to different values,
by rectangular voltage steps, from the holding potential to potentials of ?140, ?105, ?35, 0, and
þ35 mV (see the inset). Note the increased firing rate evoked by depolarization.
The GFP/NE-4C cell line gives rise to astrocytes, oligodendrocytes, and neurons 4

Page 8

3(A)]. Four weeks after cell transplantation, a large
number of GFP-positive cells survived, and they
filled the entire photochemical lesion. No migration
of GFP-positive cells outside of the photochemical
lesion was observed. Surprisingly, there was a larger
number of NeuN-positive cells inside of the lesion
compared to the number of GFP-positive cells [Fig.
3(B,C)]. GFP-positive cells were identified as neu-
rons according to their positive staining for NeuN,
?III-tubulin, or NF-68 [Fig. 3(C)]. These cells were
characterized by a rounded cell body with a diameter
of 15 to 20 ?m and by one or two, occasionally more,
short processes. Patch-clamp recording revealed that
GFP/?III-tubulin– or GFP/NF-68–positive cells dis-
played a typical neuronal membrane current pattern;
the expression of tetrodotoxin-sensitive Naþcurrents
[Fig. 4(A–C)] and KDRand KAcurrents [Fig. 4(D)].
Average Vrest and IR were ?61.6 6 4.0 mV and
460.6 6 38.1 MO (n ¼ 18), respectively. Furthermore,
12 GFP/?III-tubulin– or GFP/NF-68–positive cells
were able to fire action potentials, expressed as the re-
petitive opening of Naþchannels during a depolarizing
pulse or during depolarization evoked by the applica-
tion of 20 mM Kþ[Fig. 4(D–F)]. Six cells showed only
large TTX-sensitive Naþcurrents at the beginning of
the depolarizing pulse [Fig. 4(A,B)]. To show that
newly derived neurons are able to form synaptic con-
nections, we used an antibody against synaptophysin
[Fig. 5(A–C)]. No synaptophysin-positive staining was
observed inside of the photochemical lesions of PCL
rats [Fig. 5(A)]. However, the lesions in rats trans-
planted with GFP/NE-4C cells showed synaptophysin-
positive staining in the areas that were occupied by
transplanted cells [Fig. 5(B,C)].
Astrocytes. In PCL rats as well as in grafted rats, im-
munohistochemical analysis of the photochemical
lesions showed increased GFAP, S100?, and nestin
staining in the vicinity of the lesion, a pattern typical
of astrogliosis [Fig. 6(A), left]. In PCL rats, we found
no GFAP-, S100?-, or nestin-positive cells inside of
the lesion [Fig. 6(A)], while in grafted animals we
found numerous GFP cells that were GFAP- or
S100?- or nestin-positive [Fig. 6(B,C)]. GFP-positive
astrocytes showed the morphology of either reactive
or normal astrocytes. Similarly as in control animals,
patch-clamp recordings revealed two electrophysio-
logically different types of GFP/GFAP-positive cells
after transplantation into the site of the photochemi-
cal lesion (Table 2). Twelve cells (from a total num-
ber of 23) displayed a large time- and voltage-inde-
pendent Kþcurrent component, and additionally they
displayed KA, KDR, and KIR currents that became
most apparent following passive current subtraction
[A1 astrocytes, Fig. 7(A,C,D)]. Eleven cells dis-
played KDRcurrents in addition to time- and voltage-
independent Kþcurrents [A2 astrocytes, n ¼ 11, Fig.
7(B,C,E)]. No voltage-dependent Naþcurrents were
detected in these cells. Despite the fact that two elec-
trophysiologically distinct types of astrocytes were
found, their passive membrane properties were not
significantly different. The average value of astrocyte
Vrestand IR were ?71.9 6 2.5 mV and 144.4 6 13.0
MO, respectively (n ¼ 23).
Oligodendrocytes. Similarly to the astrocytic and
neuronal markers, no positive staining for the oligo-
dendrocyte markers MOSP, RIP, O1 [Fig. 8(A)], and
O4 (not shown) was observed inside of the photo-
Table 2
Cortex (Control) or into the Site of a Photochemical Lesion (Lesion)
Membrane Properties of GFP/GFAP-positive Cells 4 Weeks after Transplantation into Noninjured
GFP/GFAP-
Positive Cells
A1A2
Control LesionControlLesion
Vrest[mV]
Cm[pF]
IR [MO]
KDR/Cm[pA/pF]
KA/Cm[pA/pF]
KIR/Cm[pA/pF]
n
?72.4 6 5.5
38.9 6 4.3
239.3 6 43.7
12.9 6 2.9
19.3 6 6.4
0
7
?70.3 6 3.9
57.8 6 6.3*
156.8 6 19.7
2.8 6 0.6**
3.9 6 0.4**
0.9 6 0.2*
12
?62.8 6 5.6
59.7 6 13.4
146.2 6 24.0
3.8 6 0.6
0
0
6
?73.5 6 3.5
71.6 6 10.8
130.8 6 18.4
5.2 6 3.2
0
0
11
Vrest, resting membrane potential; Cm, membrane capacitance; IR, input resistance; KDR/Cmand KA/Cm, KDRand KAcurrent densities; n,
number of cells.
A1 astrocytes displayed a large time- and voltage-independent Kþcurrent component, but additionally they displayed KA, KDR, and KIRcur-
rents that became most apparent following passive current subtraction; A2 astrocytes displayed KDRcurrents in addition to time- and voltage-inde-
pendent Kþcurrents. Asterisks indicate significant differences between GFP/GFAP-positive cells in control and those derived in lesioned cortex.
*p < 0.05.
**p < 0.01.
***p < 0.001.
Neural Progenitors in Cortical Lesion 1091
Journal of Neurobiology. DOI 10.1002/neu

Page 9

chemical lesion in PCL rats. In the case of trans-
planted animals, we found only a small number of
GFP/MOSP-, RIP- or O1-positive cells in the lesion
[Fig. 8(B,C)], and we found no GFP/O4-positive
cells. The GFP/MOSP-positive cells displayed the
morphology of oligodendrocyte precursors with small
cell bodies and short processes, while GFP/RIP- and
O1-positive cells showed a few parallel processes,
reflecting the morphology of more mature oligoden-
drocytes. Only one GFP-positive cell was identified
postrecording as MOSP-positive. This cell expressed
decaying passive Kþcurrents and large KDRcurrents,
the current pattern being similar to that shown by
astrocytes [Fig. 7(C)]. Vrestand IR were ?68 mV and
128 MO, respectively.
DISCUSSION
In the present study, we show that in vitro derived
neurons, originated from neuroectodermal stem cells
(GFP/NE-4C clone) after RA-induced differentiation,
express outwardly rectifying, fast activating and inac-
tivating A-type Kþcurrents, delayed outwardly recti-
fying Kþcurrents and inwardly rectifying Naþcur-
rents. Immunohistochemistry demonstrated that RA-
pretreated GFP/NE-4C cells give rise to neurons,
astrocytes, and oligodendrocytes when transplanted
into noninjured cortex as well as to the site of a pho-
tochemical lesion. Further, we describe the passive
membrane properties and current patterns of GFP/
NE-4C–derived neurons and astrocytes.
The NE-4C cell line was shown to be a suitable
model for studying neurogenesis in vitro (Schlett and
Madarasz 1997; Schlett et al., 1997, 2000; Jelitai
et al., 2004) as well as in vivo (Demeter et al., 2004)
when transplanted into adult, newborn, or embryonic
brains without RA predifferentiation. We demon-
strate that this cell line is, after RA predifferentiation,
a suitable model for studying neuro- and gliogenesis
in vivo after transplantation into either injured or
noninjured brain. Similarly as in previous studies,
undifferentiated GFP/NE-4C cells expressed nestin
(Schlett and Madarasz, 1997), a marker for early em-
bryonic neural stem cells (Lendahl et al., 1990); after
RA induction, GFP/NE-4C cells started to form pro-
cesses and developed a neuronal phenotype, as also
reported previously (Schlett and Madarasz 1997;
Schlett et al., 1997, 2000). From day 3 onwards, neu-
rons could be identified by the expression of the neu-
ron-specific markers ?III-tubulin and NF-68 and were
found on top of nondifferentiated cells (Schlett et al.,
1997; Demeter et al., 2004). NE-4C cells were also
reported to develop into neurons during co-culture
with neonatal astrocytes (Kornyei et al., 2005). It was
suggested that ES cells are unable to develop sponta-
neously into fully mature neurons in vitro and that the
addition of extrinsic factors is crucial for inducing the
transcription process leading to the formation of func-
tional voltage dependent Naþchannels (Balasubra-
maniyan et al., 2004). After RA or Sonic Hedgehog
agonist treatment, motoneurons derived from ES cells
in vitro expressed fast-inactivating Naþchannels,
KDRchannels and Ca2þchannels and were able to fire
Figure 3
weeks after transplantation into the site of a photochemical
lesion. (A) Coronal sections of rat cortex with a nontrans-
planted photochemical lesion, immunostained for NeuN,
?III-tubulin, and NF-68 5 weeks after the injury. Enlarge-
ments of the tissue section shown on the left illustrate
immunoreactivity for NeuN, ?III-tubulin, and NF-68 in the
vicinity of the lesion. Note that there is no NeuN, ?III-tubu-
lin, or NF-68 immunoreactivity inside the lesion. (B) Coro-
nal section of rat cortex with a photochemical lesion and
transplanted GFP/NE-4C cells immunostained for NeuN 4
weeks after transplantation. (C) Higher magnification pho-
tomicrographs illustrate details of GFP/NE-4C cell immu-
noreactivity for NeuN (top), ?III-tubulin (middle), and NF-
68 (bottom) inside the lesion. Yellow color indicates dou-
ble-stained cells.
GFP/NE-4C cell line gives rise to neurons 4
1092Ande ˇrova ´ et al.
Journal of Neurobiology. DOI 10.1002/neu

Page 10

action potentials and form functional synapses with
muscle cells (Miles et al., 2004). Stem cells from the
adult human brain have also been reported to develop
into firing neurons in vitro after differentiation in-
duced by the addition of fetal calf serum and mitogen
removal (Westerlund et al., 2003).
Using whole-cell patch-clamp recordings, we have
demonstrated that GFP/?III-tubulin– and GFP/NF-
Figure 4
the rat cortex 4 weeks after transplantation. (A) The membrane current pattern of a GFP/?III-tubu-
lin–positive cell prior to the application of TTX (control, A, left) and in the presence of 1 ?M TTX
(A, right). Note that the voltage dependent Naþcurrent was completely blocked by TTX. Mem-
brane current patterns and the current/voltage relationship were measured in response to voltage
steps from a holding potential of ?70 mV. To activate the currents, the membrane was clamped for
50 ms from the holding potential to increasing de- and hyperpolarizing potentials ranging from
?160 to þ20 mV, in 10 mV increments. (B) The TTX-sensitive current was obtained by subtract-
ing the currents after TTX application from control currents (B, left), and the resulting current/volt-
age (I/V) relationship for the TTX-sensitive current is shown in B (right, filled triangels). (C) The
morphology (C, left) and immunohistochemical identification (C, right) of a recorded GFP/NE-4C
cell (indicated by arrowheads) showing positive staining for ?III-tubulin. (D) The membrane cur-
rent pattern of a GFP/NF-68–positive cell (D, left) was obtained as described in (A), and the current
traces in D (right) show the activation of an A-type Kþcurrent in response to a hyperpolarizing pre-
pulse to ?110 mV. Note the typical neuronal current pattern: the repetitive opening of Naþchan-
nels during the depolarizing pulse. (E) Current evoked by the application of 20 mM Kþat a holding
potential of ?70 mV. The membrane currents were obtained by clamping the cell membrane poten-
tial to different values, by rectangular voltage steps, from the holding potential to potentials of
?140, ?105, ?35, 0, and þ35 mV (see the inset). Note the increased firing rate evoked by depola-
rization. (F) The morphology (F, left) and immunohistochemical identification (F, right) of a
recorded GFP/NE-4C cell (indicated by arrowheads) showing positive staining for NF-68.
Typical membrane current patterns of GFP/?III-tubulin– and NF-68–positive cells in
Neural Progenitors in Cortical Lesion1093
Journal of Neurobiology. DOI 10.1002/neu

Page 11

68–positive cells display passive membrane proper-
ties in vitro that are within the normal range of rodent
neurons and express KA, KDR, and Naþcurrents, a
typical neuronal current pattern. The amplitudes of
TTX-sensitive Naþcurrents as well as KAand KDR
currents in these cells are significantly smaller than
those of fully differentiated neurons (Benninger et al.,
2003; Balasubramaniyan et al., 2004; Miles et al.,
2004), probably due to the low number or low perme-
ability of channels on the cell membrane. Although
these cells expressed functional Naþchannels, they
were unable to fire action potentials in vitro when
exposed to 20 mM Kþ. Our data are with agreement
with those of Herberth and colleagues (2002), who
described Naþcurrents on the fourth day of induc-
tion, although in their study KCl treatment was
applied only during the first 2 days of in vitro differ-
entiation and not at the later stages of differentiation.
The low Naþcurrent density or an immature form of
this current might be the reason why these cells fail
to induce action potential firing when exposed to ele-
vated Kþin vitro.
Furthermore, from day 8 onwards most of the
GFP/?III-tubulin–positive cells expressed functional
GABAA receptors. The expression of functional
GABAAand glutamate receptors has been shown in
hippocampal neural progenitors cultured for 21 days
(Mistry et al., 2002). Our finding that ?III-tubulin–
positive NE-4C cells predominantly express GABAA
Figure 5
tochemical lesion 4 weeks after cell transplantation. (A)
Coronal section of rat cortex with a nontransplanted photo-
chemical lesion, immunostained for synaptophysin (A, left).
Enlargement of the tissue section (A, right) illustrates syn-
aptophysin immunoreactivity in the vicinity of the lesion.
Note that there is no synaptophysin immunoreactivity
inside the lesion. (B, C) Coronal sections of rat cortex with
a photochemical lesion and transplanted GFP/NE-4C cells
immunostained for synaptophysin 4 weeks after transplan-
tation. The higher magnification photomicrographs (C)
show the details of synaptophysin expression in a photo-
chemical lesion occupied by GFP/NE-4C cells.
The expression of synaptophysin inside a pho-
Figure 6
cytes 4 weeks after transplantation into the site of a photo-
chemical lesion. (A) Lower magnification photomicrograph
of a rat cortex coronal section with a nontransplanted
photochemical lesion, immunostained for GFAP (A, left).
The higher magnification photomicrographs (A) illustrate
immunoreactivity for GFAP, S100?, and nestin in the vi-
cinity of the lesion. Note the typical astrogliosis around the
lesion, but no GFAP, S100?, or nestin immunoreactivity
inside the lesion. (B) Coronal section of rat cortex with a
photochemical lesion and transplanted GFP/NE-4C cells
immunostained for GFAP 4 weeks after transplantation. (C)
Higher magnification photomicrographs illustrate details of
GFP/NE-4C cell immunoreactivity for GFAP (top), S100?
(middle), and nestin (bottom) inside the lesion. Yellow
color indicates double-stained cells.
The GFP/NE-4C cell line gives rise to astro-
1094Ande ˇrova ´ et al.
Journal of Neurobiology. DOI 10.1002/neu

Page 12

receptors is in agreement with the proposed role of
GABA in neurogenesis (Wang et al., 2003, 2005;
Jelitai et al., 2004), and also with the fact that neurons
generated from expanded populations of neural stem
or progenitor cells are, to a large extent, GABAergic
(Gritti et al., 1996; Johe et al., 1996). The absence of
a glutamate response in GFP/NE-4C cells in vitro
could be explained by the lack of a functional gluta-
mate receptor. NMDA receptors are heteromeric
complexes, containing at least one NR1 and one or
more NR2 subunits (Mori and Mishina, 1995). NR1
subunit protein was detected in both nondifferentiated
and differentiated NE-4C cells, while NR2B subunit
protein, essential for receptor activity, initially
appeared 6 days after the induction of differentiation
by retinoic acid; after 9 days the majority of the pro-
tein was localized in the membrane (Jelitai et al.,
2002). Because the NR2B subunit protein has been
proposed to be a limiting factor in the co-assembly
and cell surface targeting of the receptor (McIIhinney
et al., 1996, 1998), the lack of the mature form of
NR2B protein might explain the lack of response to
glutamate. Furthermore, the NMDA receptors already
present in the cell membrane can be affected by gly-
cosylation and phosphorylation (Clark et al., 1998).
The expression of functional glutamate receptors
Figure 7
after transplantation. (A, B) The membrane current patterns of GFP/GFAP-positive cells and the
current/voltage (I/V) relationship were measured in response to voltage steps from a holding poten-
tial of ?50 mV. To activate the currents, the membrane was clamped for 50 ms from the holding
potential to de- and hyperpolarizing potentials ranging from ?140 to þ40 mV, in 10 mV incre-
ments (A, left, A1 astrocyte; B, left, A2 astrocyte) and to de- and hyperpolarizing potentials ranging
from ?140 to þ40 mV after the hyperpolarizing prepulse (A, middle). The A-type Kþcurrent was
isolated by subtracting the current traces without a hyperpolarizing prepulse from those with the
prepulse (A, right), and the delayed outward rectifying Kþcurrent was isolated by passive current
subtraction (B, right). (C) The resulting I/V relationship for the current traces shown in A (left;
filled triangles) and in B (left; filled circles) and for the isolated A-type Kþcurrent shown in A
(right, empty triangles) and the isolated delayed outwardly rectifying Kþcurrent shown in B(right;
empty circles). (D, E) The morphology (D, E top) and immunohistochemical identification (D, E
bottom) of the recorded GFP/NE-4C cells showing positive staining for GFAP.
The membrane current patterns of GFP/GFAP-positive cells in the rat cortex 4 weeks
Neural Progenitors in Cortical Lesion1095
Journal of Neurobiology. DOI 10.1002/neu

Page 13

might also occur at later stages of differentiation
when astrocytes and oligodendrocytes are derived;
however, further electrophysiological studies are
needed to confirm this hypothesis. Although NE-4C
cells have been described as capable of developing
into astrocytes and oligodendrocyte following RA
treatment in vitro (Schlett et al., 1997), we found no
GFAP-, S100?-, or MOSP-positive cells during 10
days of GFP/NE-4C cell culture, but we did find cells
expressing NG2-proteoglycan, a marker of oligoden-
drocyte progenitors or NG2 glia (unpublished data).
GFAP expression has been observed only at later
stages of differentiation, from day 12 onwards
(Schlett et al., 1997).
Because ES cells and neural progenitors represent
a promising tool for cell-replacement therapy, it is
important to examine their survival potential after
transplantation into the adult brain and also into mod-
els of damaged adult neural tissue (Bjorklund et al.,
2002; Takagi et al., 2005), where the conditions differ
from those found in developing neural tissue. The
integration of neural precursors differentiated from
ES cells in vitro into the fetal brain (Wernig et al.,
2002, 2004) or fetal neural tube (Plachta et al., 2004)
has already been shown. Also, immortalized neural
progenitors have the ability to integrate and differen-
tiate after transplantation into the noninjured, devel-
oping or adult brain, without any sign of tissue dis-
ruption or tumor formation, but the phenotypic fate of
the cells is at least quantitatively different from that
of endogenous neural progenitors (Martı ´nez-Serranoa
and Bjorklund 1997). In vitro expanded neural stem/
progenitor cells can undergo region-specific differen-
tiation after transplantation into the developing or
adult brain, and display morphologies and markers
characteristic of mature neurons (Englund et al.,
2002).
RA-untreated GFP/NE-4C cells do not spontane-
ously develop into neurons, and they have a low pro-
liferation rate when transplanted into the neonatal or
adult rat brain (Demeter et al., 2004). In our study,
RA-pretreated cells were immunohistochemically
identified as neurons, astrocytes, and oligodendro-
cytes in noninjured cortex as well as in photochemi-
cal lesions 4 weeks posttransplantation. The passive
membrane properties of GFP/?III-tubulin– or GFP/
NF68–positive cells were comparable with those
observed in endogenous mature neurons in the cortex
(unpublished data). In noninjured cortex as well as in
lesions, Vrestof in vivo derived neurons shifted to
more negative values, IR decreased and KDR?, KA?,
and Naþcurrent densities increased compared to
in vitro derived neurons (Table 1). Moreover, in vivo
derived neurons were able to fire action potentials,
based on repetitive Naþchannel opening during a
depolarizing pulse or the application of 20 mM Kþ.
Despite the fact that the Naþand Kþcurrent densities
of newly derived neurons from noninjured and
lesioned cortex were comparable (Table 1), the neu-
rons in the lesions displayed a lower firing rate during
exposure to 20 mM Kþthan those from noninjured
cortex. Although their Naþcurrent densities were
approximately three times greater then those ob-
served in vitro [Fig. 1(C) and 4(B)], in vivo the Naþ
Figure 8
dendrocytes 4 weeks after transplantation into the site of a
photochemical lesion. (A) Lower magnification photomi-
crograph of a rat cortex coronal section with a nontrans-
planted photochemical lesion, immunostained for RIP (A,
left). The higher magnification microphotographs illustrate
immunoreactivity for RIP, MOSP, and O1 in the vicinity of
the lesion. Note that there is no RIP, MOSP, and O1 immu-
noreactivity inside of the lesion. (B) Coronal section of rat
cortex with a photochemical lesion and transplanted GFP/
NE4C cells was immunostained for MOSP 4 weeks after
transplantation. (C) Higher magnification microphoto-
graphs illustrate details of the GFP/NE-4C cell immuno-
reactivity for RIP (top), O1 (middle), and MOSP (bottom)
inside the lesion. Yellow color indicates double-stained
cells.
The GFP/NE-4C cell line gives rise to oligo-
1096 Ande ˇrova ´ et al.
Journal of Neurobiology. DOI 10.1002/neu

Page 14

current densities did not reach the values of fully
differentiated neurons ((Englund et al., 2002; Balasu-
bramaniyan et al., 2004). The appearance of synapto-
physin inside the lesion suggests the presence of pre-
synaptic terminals adjacent to the membranes of
GFP/NE-4C cells and roughly corresponds to the
morphological development of synapses in the brain
(Jelitai et al., 2002; Mizoguchi et al., 2002).
In both noninjured and lesioned cortex, two types
of GFP/GFAP–positive cells (A1 and A2 astrocytes)
were identified according to their current patterns,
although their passive membrane properties were not
significantly different. This finding could be ex-
plained either by the existence of a single type of
astrocyte at different stages of differentiation, or by
the existence of two types of astrocytes, as proposed
by many authors (Seifert et al., 1997; Bergles and
Jahr, 1997, 1998; Zhou and Kimelberg, 2001; Mat-
thias et al., 2003). Because a number of GFAP-posi-
tive cells also showed nestin-positive staining, these
cells might be reactive astrocytes, as described previ-
ously (Frisen et al., 1995; Anderova et al., 2004). In
the lesioned cortex A1 astrocytes showed lower KDR
and KAcurrent densities and increased KIRcurrent
density when compared to A1 astrocytes from nonin-
jured cortex. Because it has been shown that passive
Kþcurrents are upregulated during astrocyte matura-
tion (Kressin et al., 1995), all of these differences, to-
gether with the higher IR seen in unlesioned tissue,
might reflect the delayed maturation of A1 astrocytes
in noninjured cortex. In vivo derived GFP astrocytes
showed significantly smaller KDR, KA, and KIRcur-
rent densities and higher IR values compared to
mature endogenous astrocytes in the cortex. The Cm
of GFP/GFAP-positive astrocytes was not signifi-
cantly different from the Cmof mature astrocytes in
the cortex (Bordey et al., 2001; Anderova et al.,
2004).
Interestingly, immunohistochemical staining re-
vealed a large number of NeuN-positive/GFP-nega-
tive cells [Fig. 2, 3(C)] as well as increased NG2-pro-
teoglycan immunoreactivity inside the lesion 4 weeks
after transplantation (unpublished data). These cells
likely do not represent surviving endogenous neurons
or NG2-expressing glia, as there was no positive
staining for NeuN or NG2 in photochemical lesions
without transplanted cells. These cells might repre-
sent NE-4C cells that do not express GFP or endoge-
nous stem cells or neural progenitors that migrated
into the lesion and differentiated into neurons or NG2
glia. NG2-proteoglycan expressing progenitors may
be multipotential precursors capable of differentiating
into astrocytes, oligodendrocytes and perhaps even
into neurons, as proposed by Lin and Bergles (2002)
and Berry and colleagues (2002). Moreover, the par-
ticipation of NG2 glia in the early events leading to
astrogliosis and in tissue repair was demonstrated af-
ter a cortical stab wound (Alonso et al., 2005). An
increased proliferation of NG2 glia was observed af-
ter spreading depression (Tamura et al., 2004). Thus,
endogenous stem cells could be attracted by factors
released by the transplanted neuroectodermal cells.
As endogenous stem cells are not activated solely by
the presence of a brain lesion, the administration of
mitogens, differentiating agents, and survival agents
has been used to explore endogenous stem cell partic-
ipation in the behavioral and biochemical improve-
ment of cholinergic function (Calza et al., 2003).
Therefore, it is likely that endogenous cells could
have been attracted by factors released by the im-
planted GFP/NE-4C cells.
In conclusion, we have demonstrated that embry-
onic neuroectodermal progenitor cells (GFP/NE-4C
clone) pretreated with RA give rise to neurons
in vitro and neurons, astrocytes, and oligodendrocytes
in vivo after transplantation into noninjured cortex
and to the site of a photochemical lesion. These cells
display passive membrane properties within the nor-
mal range of rodent neurons and astrocytes and
express typical neuronal and astrocytic current pat-
terns. Based on this data, the GFP/NE-4C cell line
transplanted into the site of an injury can provide an
appropriate model for studying neuro- and gliogene-
sis in vivo.
We would like to thank Hana Hronova for immunohisto-
chemical staining.
REFERENCES
Alonso G. 2005. NG2 proteoglycan-expressing cells of the
adult rat brain: possible involvement in the formation of
glial scar astrocytes following stab wound. Glia 49:318–
338.
Anderova M, Antonova T, Petrik D, Neprasova H, Chvatal
A, Sykova E. 2004. Voltage-dependent potassium cur-
rents in hypertrophied rat astrocytes after a cortical stab
wound. Glia 48:311–326.
Balasubramaniyan V, de Haas AH, Bakels R, Koper A,
Boddeke HW, Copray JC. 2004. Functionally deficient
neuronal differentiation of mouse embryonic neural stem
cells in vitro. Neurosci Res 49:261–265.
Benninger F, Beck H, Wernig M, Tucker KL, Brustle O,
Scheffler B. 2003. Functional integration of embryonic
stem cell-derived neurons in hippocampal slice cultures.
J Neurosci 23:7075–7083.
Bergles DE, Jahr CE. 1997. Synaptic activation of gluta-
mate transporters in hippocampal astrocytes. Neuron
19:1297–1308.
Neural Progenitors in Cortical Lesion 1097
Journal of Neurobiology. DOI 10.1002/neu

Page 15

Bergles DE, Jahr CE. 1998. Glial contribution to glutamate
uptake at Schaffer collateral-commissural synapses in the
hippocampus. J Neurosci 18:7709–7716.
Berry M, Hubbard P, Butt AM. 2002. Cytology and lineage
of NG2-positive glia. J Neurocytol 31:457–467.
Bjorklund LM, Sanchez-Pernaute R, Chung S, Andersson
T, Chen IY, McNaught KS, Brownell AL, Jenkins BG,
Wahlestedt C, Kim KS, Isacson O. 2002. Embryonic
stem cells develop into functional dopaminergic neurons
after transplantation in a Parkinson rat model. Proc Natl
Acad Sci U S A 99:2344–2349.
Bordey A, Lyons SA, Hablitz JJ, Sontheimer H. 2001. Elec-
trophysiological characteristics of reactive astrocytes in
experimental cortical dysplasia. J Neurophysiol 85:1719–
1731.
Calhoun JD, Lambert NA, Mitalipova MM, Noggle SA,
Lyons I, Condie BG, Stice SL. 2003. Differentiation
of rhesus embryonic stem cells to neural progenitors
and neurons. Biochem Biophys Res Commun 306:191–
197.
Calza L, Giuliani A, Fernandez M, Pirondi S, D’Intino
G, Aloe L, Giardino L. 2003. Neural stem cells and
cholinergic neurons: regulation by immunolesion and
treatment with mitogens, retinoic acid, and nerve
growth factor. Proc Natl Acad Sci U S A 100:7325–
7330.
Clark RA, Gurd JW, Bissoon N, Tricaud N, Molnar E,
Zamze SE, Dwek RA, McIlhinney RA, andWing DR.
1998. Identification of lectin-purified neural glyco-pro-
teins, GPs 180, 116, and 110, with NMDA and AMPA
receptor subunits: conservation of glycosylation at the
synapse. J Neurochem 70:2594–2605.
Chu K, Kim M, Jeong SW, Kim SU, Yoon BW. 2003.
Human neural stem cells can migrate, differentiate, and
integrate after intravenous transplantation in adult rats
with transient forebrain ischemia. Neurosci Lett 343:
129–133.
Demeter K, Herberth B, Duda E, Domonkos A, Jaffredo T,
Herman JP, Madarasz E. 2004. Fate of cloned embryonic
neuroectodermal cells implanted into the adult, newborn
and embryonic forebrain. Exp Neurol 188:254–267.
Englund U, Bjorklund A, Wictorin K, Lindvall O, Kokaia
M. 2002. Grafted neural stem cells develop into func-
tional pyramidal neurons and integrate into host corti-
cal circuitry. Proc Natl Acad Sci U S A 99:17089–
17094.
Frisen J, Johansson CB, Torok C, Risling M, Lendahl U.
1995. Rapid, widespread, and longlasting induction of
nestin contributes to the generation of glial scar tissue af-
ter CNS injury. J Cell Biol 131:453–464.
Fujiwara Y, Tanaka N, Ishida O, Fujimoto Y, Murakami T,
Kajihara H, Yasunaga Y, Ochi M. 2004. Intravenously
injected neural progenitor cells of transgenic rats can
migrate to the injured spinal cord and differentiate into
neurons, astrocytes and oligodendrocytes. Neurosci Lett
366:287–291.
Gritti A, Parati EA, Cova L, Frolichsthal P, Galli R,
Wanke E, Faravelli L, Morassutti DJ, Roisen F, Nickel
DD, Vescovi AL. 1996. Multipotential stem cells from
the adult mouse brain proliferate and self-renew in re-
sponse to basic fibroblast growth factor. J Neurosci 16:
1091–1100.
Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ.
1981. Improved patch-clamp techniques for high-resolu-
tion current recording from cells and cell-free membrane
patches. Pflugers Arch 391:85–100.
Hayashi J, Takagi Y, Fukuda H, Imazato T, Nishimura M,
Fujimoto M, Takahashi J, Hashimoto N, Nozaki K. 2006.
Primate embryonic stem cell-derived neuronal progeni-
tors transplanted into ischemic brain. J Cereb Blood Flow
Metab. In press.
Herberth B, Pataki A, Jelitai M, Schlett , Deak , Spat A,
Madarasz E. 2002. Changes of KCl sensitivity of prolif-
erating neural progenitors during in vitro neurogenesis.
J Neurosci Res, 67:574–582.
Jelitai M, Anderova M, Marko K, Kekesi K, Koncz P,
Sykova E, Madarasz E. 2004. Role of gamma-aminobu-
tyric acid in early neuronal development: studies with an
embryonic neuroectodermal stem cell clone. J Neurosci
Res 76:801–811.
Jelitai M, Schlett K, Varju P, Eisel U, Madarasz E. 2002.
Regulated appearance of NMDA receptor subunits and
channel functions during in vitro neuronal differentiation.
J Neurobiol 51:54–65.
Jendelova P, Herynek V, Urdzikova L, Glogarova K, Krou-
pova J, Andersson B, Bryja ,V, Burian M, Hajek M,
Sykova E. 2004. Magnetic resonance tracking of trans-
planted bone marrow and embryonic stem cells labeled
by iron oxide nanoparticles in rat brain and spinal cord.
J Neurosci Res 76:232–243.
Johe KK, Hazal TG, Muller T, Dugich-Djordjevic MM,
McKay RDG. 1996. Single factors direct the differentia-
tion of stem cells from fetal and adult nervous system.
Genes Dev 10:3129–3140.
Kamei N, Oishi Y, Tanaka N, Ishida O, Fujiwara Y, Ochi
M. 2004. Neural progenitor cells promote corticospinal
axon growth in organotypic co-cultures. Neuroreport
15:2579–2583.
Kim JH, Auerbach JM, Rodriguez-Gomez JA, Velasco I,
Gavin D, Lumelsky N, Lee SH, Nguyen J, Sanchez-Per-
naute R, Bankiewicz K, McKay R. 2002. Dopamine
neurons derived from embryonic stem cells function in
an animal model of Parkinson’s disease. Nature 418:50–
56.
Kornyei Z, Szlavik V, Szabo B, Gocza E, Czirok A, Madar-
asz E. 2005. Humoral and contact interactions in astro-
glia/stem cell co-cultures in the course of glia-induced
neurogenesis. Glia 49:430–444.
Kressin K, Kuprijanova E, Jabs R, Seifert G, Steinhauser C.
1995. Developmental regulation of Naþand Kþconduc-
tances in glial cells of mouse hippocampal brain slices.
Glia 15:173–187.
Lendahl U, Zimmerman LB, McKay RD. 1990. CNS stem
cells express a new class of intermediate filament protein.
Cell 60:585–595.
Lin SC, Bergles DE. 2002. Physiological characteristics
of NG2-expressing glial cells. J Neurocytol 31:537–
549.
1098 Ande ˇrova ´ et al.
Journal of Neurobiology. DOI 10.1002/neu