Acta Neuropathol (2010) 120:85–96
DOI 10.1007/s00401-010-0677-y
ORIGINAL PAPER
Balloon cells in human cortical dysplasia and tuberous sclerosis:
isolation of a pathological progenitor-like cell
Shireena A. Yasin • Kate Latak • Francesca Becherini • Anita Ganapathi • Khadijah Miller •
Oliver Campos • Simon R. Picker • Nelly Bier • Martin Smith • Maria Thom • Glenn Anderson
J. Helen Cross • William Harkness • Brian Harding • Thomas S. Jacques
•
Received: 9 January 2010 / Revised: 19 March 2010 / Accepted: 19 March 2010 / Published online: 30 March 2010
Ó Springer-Verlag 2010
Abstract Neural stem cells are present in the human
post-natal brain and are important in the development of
brain tumours. However, their contribution to non-neoplastic human disease is less clear. We have tested the
hypothesis that malformations of cortical development
contain abnormal (pathological) stem cells. Such malformations are a major cause of epilepsy. Two of the most
common malformations [focal cortical dysplasia (FCD)
Electronic supplementary material The online version of this
article (doi:10.1007/s00401-010-0677-y) contains supplementary
material, which is available to authorized users.
S. A. Yasin K. Latak F. Becherini A. Ganapathi K. Miller O. Campos S. R. Picker N. Bier G. Anderson B. Harding T. S. Jacques
Department of Histopathology, Great Ormond Street Hospital,
Great Ormond Street, London WC1N 3JH, UK
M. Smith J. Helen Cross
Department of Neurology, Great Ormond Street Hospital,
Great Ormond Street, London WC1N 3JH, UK
W. Harkness
Department of Neurosurgery, Great Ormond Street Hospital,
Great Ormond Street, London WC1N 3JH, UK
S. A. Yasin K. Latak A. Ganapathi K. Miller O. Campos S. R. Picker T. S. Jacques (&)
Neural Development Unit, UCL-Institute of Child Health,
30 Guilford Street, London WC1N 1EH, UK
e-mail: t.jacques@ich.ucl.ac.uk
J. Helen Cross B. Harding
The Neuroscience Unit, UCL-Institute of Child Health,
30 Guilford Street, London WC1N 1EH, UK
M. Thom
UCL-Institute of Neurology, Queen Square,
London WC1N 3BG, UK
and cortical tubers] are characterised by the presence of a
population of abnormal cells known as balloon cells. The
identity of these cells is unknown but one hypothesis is that
they are an abnormal stem cell that contributes to the
pathogenesis of the malformation. We have characterised
in tissue, and isolated in culture, an undifferentiated population of balloon cells from surgical resections of FCD
and cortical tubers. We show that b1-integrin labels a subpopulation of balloon cells with a stem cell phenotype and
show for the first time that these cells can be isolated in
vitro. We have characterised the immunohistochemical,
morphological and ultrastructural features of these cells.
This is the first isolation of an abnormal cell with features
of a progenitor/stem cell from a non-neoplastic disease of
the brain.
Keywords Cortical dysplasia Tuberous sclerosis Stem cell Progenitor cell Epilepsy
Introduction
Neural stem cells are present in the human brain throughout life and abnormal (pathological) stem cells are present
within tumours and drive tumour formation (e.g. [7, 14,
20]). Normal endogenous neural stem cells respond to a
wide variety of insults in the brain [14]. However, an
abnormal stem or progenitor cell has not been isolated from
a non-neoplastic disease in humans.
Malformations of cortical development are an important
cause of drug-resistant epilepsy and are the most common
structural cause of epilepsy in young children [4]. The
most frequent type is focal cortical dysplasia (FCD) [4],
which is characterised by focal abnormalities of cortical
cytology and architecture. Children with this disease
123
86
develop frequent seizures often during infancy. Treatment
with anti-epileptic drugs is usually ineffective and the
children may require surgical removal of the affected part
of the brain. FCD also represents a unique opportunity to
study the mechanisms of cortical development in humans.
However, the pathogenesis of FCD is not understood.
The commonest form of FCD in children is FCD Type
IIb [15], a form that contains a unique population of
abnormal cells known as balloon cells. Balloon cells are so
named because of their characteristic ‘‘balloon-like’’
enlarged cytoplasm. They are morphologically indistinguishable from the giant cells seen in the cortical pathology
(cortical tubers) of tuberous sclerosis (TSC). For clarity the
term ‘balloon cell’ is used here for both cell populations
[24]. The identity and origin of these cells in FCD and TSC
is unknown. A common view is that these cells are glia or
neurones and this has been supported by a large number of
studies indicating that subsets of these cells express
markers of differentiated cells in vivo including Neurofilament, NeuN, TuJ1, glutamate receptors and GFAP [23].
However, some balloon cells express stem cell/progenitorrelated markers (CD133, CD34 and Nestin) in paraffin
tissue sections [22, 23, 26]. However, the results of these
studies have been varied in the number of positive cells that
are present with several reports identifying a very small
population of positive cells. These discrepancies may
represent heterogeneity of the cells.
We hypothesised that there is a population of balloon
cells that are abnormal progenitor cells. In the first part of
this paper we demonstrate that this population can be
identified by the expression of b1 integrin and in the second part, we describe the isolation and characterisation of
this population in vitro.
Materials and methods
Acta Neuropathol (2010) 120:85–96
clinical features of cases used for balloon cell culture are
given in Supplementary material 1.
Immunohistochemical staining of paraffin sections
Paraffin-embedded sections were cut at 4–7 lm, dewaxed
in xylene (10 min), rehydrated through a graded alcohol
series, blocked in hydrogen peroxide (10% hydrogen
peroxide in PBS) and then rinsed in distilled water.
Antigen retrieval was performed either by proteinase
digestion (0.02% bacterial proteinase type XXIV at 37°C)
[b1 integrin (1:50, SC-9970, Santa Cruz)-10 min, and
GFAP (1:1,000, DAKO)-5 min] or by pressure cooking in
either EDTA-citrate buffer pH 6.2 [ILK (1:500, Chemicon), pAKT (1:50), p4E-BP1 (1:400) pS6 (1:100) (all
from Cell Signalling Technology), MCM2 (1:900 BM28,
BD Biosciences), Neurofilament (1:120, ICN Pharmaceuticals), RT97 (1:50, Novocastra), CD34 (1:500,
DAKO), Ki67 (1:100, DAKO), aV integrin (1:50, SC9969, Santa Cruz), SOX2 (1:1000, Chemicon), Nestin
(1:100, Chemicon), Vimentin (1:200, DAKO) and Musashi-1 (1:250, a kind gift from Professor Hideyuki Okano,
Keio University School of Medicine, Japan)] or pH 9.6
(CD133, 1:25, Miltenyi Biotec AC133). Sections were
first pre-blocked with 3% BSA for 1 h. Primary antibodies
were incubated at room temperature for 60 min. Detection
was performed with DAKO Chemmate Envision kit (b1
integrin, ILK, pAKT, RT97, CD34, Ki67, aV integrin),
DAKO REAL Envision kit (SOX2, CD133, Vimentin,
MCM2, p4E-BP1 and pS6), Vectorstain ABC kit and
0.5% diaminobenzidine for 5 min (Musashi-1) or by
extravidin biotin-peroxidase (Nestin, GFAP and Neurofilament). The sections were counterstained with Mayer’s
haematoxylin. For immunofluorescence, detection was
performed using TSA (Tyramide Signal Amplification)
plus Fluorescence system [PerkinElmer LAS Ltd Buckinghamshire (UK)].
Subjects
Enzyme histochemistry
All subjects had been referred to the supra-regional service
for complex epilepsy. Minimum investigations included
prolonged video-EEG recordings, epilepsy protocol brain
MRI, and psychometric and psychiatric assessments. Many
subjects were also investigated using SPECT, PET and
invasive EEG monitoring. All cases were evaluated by a
multi-disciplinary team including paediatric specialists in
epilepsy, neurosurgery, neuroradiology, neurophysiology,
psychiatry and psychology to thoroughly evaluate the
likelihood of success in achieving seizure control, risk of
possible adverse events including new neurological morbidity, and also to understand the child’s and family’s
perspective on the aims of surgery. Summaries of the
123
Cytochrome oxidase (COX)
10 lm thick cryostat sections of fresh frozen tissue were
incubated in the following solution: 3,30 -diaminobenzidine,
tetrahydrochloride (DAB) (BDH, 0.5 mg/ml), Catalase
(BDH, 20 lg/ml), Cytochrome C (type III Sigma, 1 mg/
ml), 0.05 M phosphate buffer pH 7.4, for 1 h at 37°C.
Sections were then rinsed with tap water then incubated in
1% osmium tetroxide (Johnson Mathey) for 10 min. Sections were then counterstained lightly with Carazzi
haematoxylin then dehydrated, cleared and mounted in
DPX (BDH).
Acta Neuropathol (2010) 120:85–96
NADH-TR
10 lm thick sections were incubated in analar acetone
(BDH) for 5 min at 4°C to de-fat sections, then rapidly
dried using compressed air. Sections were then incubated
in the following solution for 2 h at 37°C: MTT 1 mg/ml
(Sigma), Tris–HCL 0.05 M pH 7.4 (BDH), cobalt chloride
0.5 M (BDH), NADH 2 mg/ml (BDH), in distilled water
pH 7.8. Sections were rinsed with tap water then mounted
in Aquatex (BDH).
Morphometry and statistics
A systematic random sampling technique was used to identify fields of view as we and others have previously described
[6]. In brief, the first field of view is sampled on a grid at
random position and subsequent fields of view are sampled at
a fixed distance from the first field (the distance determined
for the total surface area and the total number of fields to be
sampled). The number of fields of view required to obtain a
stable running mean was used to determine the number of
fields sampled. Balloon cells were identified using standard
morphological criteria and counting was performed using the
forbidden line rule. Adjacent sections were stained for b1
integrin and for GFAP, Nestin, Neurofilament, RT97, CD34
or Ki67. The fields of view were aligned using medium-sized
vessels and the same balloon cells were identified on adjacent sections. Less than 5% of cells were uninformative due
to not being identifiable on both sections.
Cell culture
Fresh surgical resections were sampled for areas of macroscopic abnormality or from regions of the grey–white
matter junction and collected in Minimal Essential Medium
(Sigma). Tissue was dissociated using the Papain Dissociation System (Worthington Biochemical Corporation,
NJ). At the end of dissociation, the cells were either
re-suspended and frozen in 10% DMSO in fetal bovine
serum, or were cultured in a defined medium [DMEM/F12
(Sigma) containing B27 supplement (Invitrogen), 100 units/ml
penicillin and 0.1 mg/ml streptomycin (Sigma), 20 ng/ml
FGF2 and 20 ng/ml EGF (Peprotech) in uncoated tissue
culture flasks (TPP)] at 37°C in 5% CO2. In each case, a
small piece of tissue was sampled from the region immediately adjacent to that sampled for culture, and this piece
was formalin fixed and paraffin embedded.
Live cell staining
Following thawing, the viability of cultured cells was
assessed by their ability to exclude propidium iodide (PI,
Sigma 5 lg/ml in media) and by staining of the nuclei for
87
Hoechst 33342 (Cambrex Bioscience, 10 lg/ml in media).
The mitochondria of balloon cells were labelled by incubating them in MitoTracker dye (200 nM in media,
Molecular Probes) for 15 min at 37°C. Cells were then
washed and re-suspended in fresh culture media.
BrdU incorporation
Thawed cells were allowed to settle overnight at 37°C.
They were either plated the day after thawing onto 5 lg/ml
poly-D-lysine-coated 1 N HCL-treated 13 mm coverslips
or kept in suspension. The plated cells were then pulsed
every 2 days for 9 days with 10 lL BrdU (Roche) and
either 0.5 lg/ml rapamycin (Autogen Bioclear) or 1% FBS.
The cells in suspension were kept in defined media (see
above) and were pulsed every 2 days with BrdU and then
plated onto coverslips following 10 days. The coverslips
were washed with PBS and incubated with 400 ll cold
methanol at -20°C for 1 min then washed again with PBS.
Cells were then incubated in blocking solution containing
10% sheep serum, 0.1% Triton X-100 and 10 lg/ml
DNaseI (Roche) for 1 h at room temperature. They were
then incubated with anti-BrdU antibody (1:500, Abcam)
diluted in 1% sheep serum/PBS overnight at 4°C.
Immunocytochemistry on cultured cells
Cultured cells were plated following 1 or 7 days in vitro
(DIV) onto 5 lg/ml poly-D-lysine-coated 1 N HCL-treated
13 mm coverslips for 3 h and then fixed (4% paraformaldehyde for 15 min at room temperature). Coverslips were
washed with PBS and incubated in 10% sheep serum in
PBS containing 0.1% Triton X-100 for 1 h. They were then
incubated with primary antibodies diluted in 1% sheep
serum/PBS overnight at 4°C. Primary antibodies used
were: Nestin (1:2,000, Chemicon), Musashi-1 (1:250, a
kind gift from Professor Hideyuki Okano, Keio University
School of Medicine, Japan), SOX2 (1:200 Chemicon),
CD133 (1:200, Abcam), Vimentin (1:200, DAKO), b1
integrin (1:100, Santa Cruz SC-9970), Neurofilament
(1:200, ICN Pharmaceuticals) and GFAP (1:600, DAKO).
The following day, coverslips were incubated with either,
anti-mouse, rabbit or rat Alexa secondary antibodies
(1:400, Molecular Probes) for 1.5 h at room temperature.
They were washed with PBS and mounted in Prolong Gold
Anti-Fade reagent containing DAPI (Molecular Probes).
Cells for MCM2 staining were fixed with 50/50 cold
methanol/ethanol at -20°C for 10 min. Coverslips were
then washed with PBS and incubated in 10% sheep serum
in PBS containing 0.1% Triton X-100 for 1 h. Cells were
then incubated with anti-MCM2 antibody (1:100, BM28,
BD Biosciences) diluted in 1% sheep serum/PBS overnight
at 4°C. Secondary antibodies were as above.
123
88
Transmission electron microscopy
Cultured balloon cells were plated onto 5 lg/ml poly-Dlysine-coated eight-well glass chamber slides (Lab TekII,
Nalgene Nunc) in a 100 ll drop of media and incubated at
37°C for 1 h. The culture media was topped up to 400 ll
per well and cells incubated for another hour at 37°C. Cells
were fixed in 3% glutaraldehyde, 0.1 M sodium cacodylate, 5 mM calcium chloride pH 7.4 and processed for resin
embedding.
Results
An undifferentiated balloon cell can be identified
in vivo by expression of b1 integrin
In common with previous reports, we found that balloon
cells express neural stem cell markers, e.g. Nestin and
Fig. 1 Balloon cells express the neural stem/progenitor cell markers,
SOX2 (a), CD133 (b), Musashi (Msi-1) (c) and b1-integrin (d, e). b1
integrin was not detectable in normally formed neocortex (NCx, f),
cortex with polymicrogyria (PMG, g) or FCD lacking balloon cells
123
Acta Neuropathol (2010) 120:85–96
CD133 [6, 8]. In addition, we found expression of the
progenitor/stem cell markers, SOX2, Musashi-1 and b1
integrin (Fig. 1a–e; Supplementary material 2). We
focused on the b1 integrin subunit because it is expressed
at higher levels on normal neural stem cells compared to
more committed cells, it has an important role in regulating normal neural stem cell behaviour and it is essential
for normal cortical development [4, 9, 10, 12, 16]. We
found positive staining for b1 integrin in FCD with balloon cells (FCDIIb; n = 10 cases) and tuberous sclerosis
(TSC n = 8 cases). Immunoreactivity was limited to
balloon cells and the surrounding neuropil, and was
expressed on 87.43% of balloon cells in FCD (±1.15%,
95% confidence interval, 3 cases). b1 integrin did not
stain normal neurones, glial cells, dysplastic/dysmorphic
neurones or giant neurones in the surrounding tissue.
Positive staining of non-parenchymal tissues (endothelium, vascular smooth muscle and meningothelial cells)
was seen in most cases.
(FCDIIa, h). Balloon cells in both FCD (i, j) and TSC (k, l) express
the integrin-linked kinase (ILK, i, k) and the phosphorylated form of
one of its downstream targets (AKT) (pAKT j, l). Scale bar 25 lm
Acta Neuropathol (2010) 120:85–96
In contrast, cerebral cortex from patients with histologically confirmed hippocampal sclerosis but no cortical
dysplasia (n = 5) (Fig. 1f), polymicrogyria (n = 5) (Fig. 1g)
and patients with cortical dysplasia without balloon cells
(FCDIIa: n = 9, FCDIa: n = 2, MCD: n = 2) (Fig. 1h)
were negative for parenchymal b1 integrin staining.
Interestingly, a variable degree of weaker fibrillary
staining was seen in a subpial pattern in all cases (irrespective of diagnosis) and which appeared to correspond
with the subpial (Chaslin’s) gliosis, commonly seen in
patients with longstanding epilepsy. We conclude that b1
integrin expression is a useful and specific marker of
balloon cells.
High levels of b1 expression are associated with
undifferentiated stem cells [4, 10] and therefore we
hypothesised that b1-expressing balloon cells may represent an undifferentiated subset of balloon cells. In order to
test this hypothesis, we stained adjacent sections for b1integrin and for various differentiation markers. We found
that none of the b1-positive balloon cells expressed the
neuronal marker, Neurofilament and only a very small
proportion of b1-positive cells stained for the glial marker
GFAP (3.1%) (Fig. 2; Supplementary material 3). In
contrast, nearly half of b1-negative balloon cells expressed GFAP (45.5%) (Supplementary material 3, P \ 0.05,
v2 interactions between b1 expression and all differentiation markers, 3 independent cases of FCD). These data
confirm that b1 integrin is a marker of a balloon cell and
does not identify reactive glia or neurones in paraffin
sections. It also supports the hypothesis that the b1positive cells are an undifferentiated subset of balloon
cells.
If integrin expression is of functional significance in
balloon cells, we would predict that additional components
of integrin signalling pathways would be present in balloon
cells. Indeed, we found that balloon cells expressed the
integrin-related signalling molecule, integrin-linked kinase
(ILK) (Fig. 1i, k). One of the downstream targets of ILK is
the Serine473 residue of AKT [16], a serine/threonine
kinase that is part of the mTOR pathway [1, 8, 12, 19].
Indeed, balloon cells in FCD and giant cells in TSC showed
immunoreactivity for the phosphorylated (Ser473) form of
AKT (Fig. 1j, l) in keeping with previous reports [19].
AKT is a part of pathway that connects the insulin growth
factor receptor, AKT, the TSC gene products (tuberin and
hamartin) and mTOR. This pathway has been implicated in
FCD and TSC by a number of studies [1, 8, 12, 19] (see
also below). Our data indicate that balloon cells contain the
necessary components to link integrin signalling to this
established aberrant pathway. While this does not reveal
the function of the integrin signalling pathway, it does
indicate that integrins are good candidate molecules for
regulating balloon cells.
89
Fig. 2 b1-integrin positive balloon cells lack markers of mature
neurones or glial cells. The left-hand panel (a, c, e, g) shows
immunohistochemistry for b1-integrin. The right-hand panel shows
an adjacent section from the same region stained for lineage markers;
Nestin (b), GFAP (d), Neurofilament (f cocktail, h RT97). Arrows in
a, c, e and g indicate b1 integrin positive balloon cells, the
corresponding balloon cell expressing Nestin (b), or negative for
GFAP, NF or RT-97 (d, f, h, respectively). Scale bar 25 lm
Patterns of integrin expression distinguish subsets
of balloon cells in FCD from those in tuberous sclerosis
The b1 integrin subunit forms the greatest number of
different integrin dimers [9]. In order to determine the
specificity of this expression pattern in balloon cells, we
also examined the aV integrin subunit, which forms the
second largest family of integrin dimers. In contrast to b1
integrin expression, the aV integrin subunit was expressed
only in cases of cortical tuber (5 out of 8 cases) and in no
cases of FCDIIb (n = 10, P \ 0.01 Fisher’s exact test)
(Fig. 3). The aV integrin subunit was expressed only on a
small subpopulation of balloon cells in TSC and staining of
adjacent sections showed that [80% of aV-positive balloon cells were positive for b1 integrin. These results
suggest that the pattern of integrin subunit expression
123
90
Acta Neuropathol (2010) 120:85–96
Fig. 3 Expression of aV
integrin differentiates FCD (b)
from cortical tubers (d). a, c The
adjacent section stained for b1
integrin. A small proportion of
balloon cells in cortical tubers
express aV integrin (5 out of
8 cases) but no balloon cells
express aV integrin in FCD
(0 out of 10 cases) (P \ 0.001
Fisher’s exact test). Scale bar
50 lm
identifies different subtypes of balloon cell in the two
diseases.
aV integrin expression varied within cases and between
cases but there was no clear anatomical distribution either
within the resections or by cerebral lobe. In the aV-rich
areas, the proportion of balloon cells that were aV-positive
varied from 7.2 to 54.2% of the balloon cells (n = 3 cases,
mean = 26.7%, mean ratio of aV-positive to b1-positive
cells = 0.35). Therefore, it is likely that there is a wide
distribution of aV expression rather distinct subtypes of
TSC.
Balloon cells can be isolated in culture
We hypothesised that if there is an undifferentiated balloon
cell population then it may be possible to isolate them in
culture in a manner analogous to normal neural progenitor
or stem cells. To test this hypothesis, we grew cells from a
range of epilepsy-associated lesions in culture conditions
that promote the survival of progenitor or stem cells. A
population of large (typically in excess of 25 lm) freefloating cells could be isolated from cases of FCD with
balloon cells (FCDIIb) and from cortical tubers (TSC) but
not from normally formed cortex, polymicrogyria or cortical dysplasia lacking balloon cells (Fig. 4; Supplementary
materials 4, 5). Although occasional populations of small
adherent cells were present in some FCD cultures, these
were not a substantial or reproducible feature and when
123
they were present they could be separated from the large
cells, as the latter were free floating in the media. These
large cells survived in culture for up to 4 weeks. By timelapse microscopy, the cells appeared to be highly motile
and dynamic—undergoing cycles of attachment and
detachment from each other and occasionally extending
processes onto the underlying substrate (Supplementary
material 5).
Such large cells were only grown from cases with a
confirmed histological diagnosis of FCDIIb or TSC
(n = 10 out of 12 cases) but were never grown from cortex
or white matter with other diagnoses (n = 43) or from
forms of cortical dysplasia lacking balloon cells (n = 10).
We have not noted a difference in the cultures between
balloon cells derived from FCD and from TSC. When the
tissue was sampled for culture, a small portion of the
adjacent tissue was processed for paraffin histology. There
was a very close correlation between the presence of balloon cells in the adjacent paraffin histology and the
presence of the large cells in the cultures (Fisher’s exact
test P \ 0.001) (Supplementary material 6).
For this technique to be a useful experimental protocol,
we needed to be able to store cultured cells for future
experiments. Therefore, isolated cells were frozen at the
time of dissociation. Notably, these cells were viable following thawing and could be maintained in culture for at
least 1 week (Fig. 4b). The viability of thawed cells after a
week in culture ranged from 43 to 61% (Fig. 4c), and was
Acta Neuropathol (2010) 120:85–96
91
that the mean number of balloon cells obtained per case
was 527 [mean ± 246 (95% confidence interval) at 1 day].
The cultured balloon cells are the b1-positive
undifferentiated cell population
In order to confirm the identity of the isolated cells, we
investigated the immunophenotype of the thawed isolated
cells using a panel of markers. We found that following
1 day in culture, all isolated balloon cells expressed b1
integrin, Nestin, SOX2, Vimentin and Musashi-1, while
the majority of cells expressed CD133 (88%) (total number
of cells stained: 125 cells from 4 cases; Fig. 5; Supplementary material 7). Very few cells, however, stained for
GFAP (15%) and none stained for markers of mature
neurones (Neurofilament). A similar progenitor cell phenotype was retained following a week in culture; all cells
expressed Nestin, Vimentin and CD133, 80% of cells
expressed SOX2 and 71% of cells expressed Musashi-1.
Interestingly, the number of balloon cells expressing b1
integrin was reduced at this stage to just 27% (total number
of stained cells: 94 from 3 cases; Supplementary materials
7, 8). In preliminary experiments, the cells remain undifferentiated under a range of culture conditions (including
serum and rapamycin) (data not shown). These data confirm that the large cultured cells have the same phenotype
as a balloon cell and suggest that the cultured cells are
enriched for the undifferentiated b1-positive balloon cells
identified in vivo.
Cultured balloon cells express cell cycle markers
Fig. 4 Balloon cells can be isolated in culture. a Phase image of
cultured balloon cell from a primary culture. b Cultured balloon cells
after 7 days in vitro after thawing. The cell is stained with Hoechst
(blue) but excludes propidium iodide (red). There is no other
significant cellular component to the culture. The debris between the
cells is acellular as demonstrated by the lack of staining for either
nuclear stain and in electron microscopy has the features of myelin
fragments. Scale bar 50 lm. c Quantification of the number of viable
balloon cells from three separate cases 1 and 7 days in vitro (DIV)
after thawing. The y-axis represents the number of cells/well (2 ml)
demonstrated by their ability to exclude the dye, propidium
iodide (PI) and by their nuclear staining for Hoechst 33342.
In addition, using these assays, we were able to estimate
The phenotype of balloon cells in culture is that of a progenitor or stem cell and this raises the possibility that they
undergo cell proliferation. Several previous studies have
indicated that balloon cells in tissue sections express a
large range of markers of cell proliferation [21, 22]. In
keeping with these previous reports, we found that many
balloon cells in vivo express the cell cycle marker, MCM2
(n = 5 cases; Fig. 6d). We also found that a proportion of
balloon cells in vitro express MCM2 (28%) (Fig. 6a–c).
MCM2 is necessary to initiate DNA replication at the
beginning of S-phase and has been suggested to be a sensitive marker of the cell cycle in balloon cells [22].
However, notably the cultures did not expand significantly
over time. In principle, this may either indicate a very slow
rate of cell cycle progression (as has been described in
normal and tumour stem cells [13]) or may indicate that
there is an abnormality of cell cycle progression. Based on
the pattern of cell cycle markers expressed in vivo, Thom
et al. [21] have hypothesised that balloon cells may be
arrested in G1. In our cell culture model, we tested this
possibility by investigating whether the cells can progress
123
92
Acta Neuropathol (2010) 120:85–96
Fig. 5 a Cultured balloon cells
express markers of stem cells or
progenitor cells. Each row
represents a single balloon cell
stained by immunofluorescence
(IF) with a single marker. All
images are confocal images
except for vimentin, which
is a projected z-stack.
b Quantification of the
percentage of balloon cells
positive for a range of markers
(data represent three separate
cases). Scale bar 25 lm
through S-phase. We pulsed cultured balloon cells with
BrdU every 2 days for 9–10 days in the presence of EGF
and bFGF (known stem cell mitogens) and also in the
presence of serum or the mTOR inhibitor, rapamycin. No
balloon cells (0/60) were labelled with BrdU under these
123
circumstances suggesting that although the cells retain
immunophenotypic evidence of being in the cell cycle,
they do not undergo significant DNA synthesis. This is in
keeping with the hypothesis that they do not pass from G1
into S-phase. While we cannot exclude the possibility that
Acta Neuropathol (2010) 120:85–96
93
Fig. 6 a–c Balloon cells in culture express the cell cycle maker MCM2. The image shows a single balloon cell in culture stained with MCM2
and DAPI. Scale bar 10 lm. d Expression of MCM2 in balloon cells in vivo. Scale bar 20 lm
Fig. 7 Balloon cells show
activation of mTOR.
Immunohistochemistry for
phospho-specific epitopes of S6
(a, b) and 4E-BP-1 (c, d) shows
reactivity in balloon cells in
both FCDIIb (a, c) and TSC
(b, d). In FCDIIb, p4E-BP1
expression is more focal and
more variable than TSC.
Scale bar 50 lm
an unidentified additional growth factor is required to drive
proliferation, this seems less likely given the range of
conditions tested and the in vivo data that support a cell
cycle arrest.
Cultured balloon cells are characterised
by accumulation of intermediate filaments
and mitochondria
While all the cells were positive for progenitor cell
markers, some of the cytoplasmic markers (e.g. Nestin)
were more concentrated at the periphery of the cells suggesting that other structures are present within the cells.
Given that several studies have indicated that FCD is
associated with activation of mTOR, a key regulator of
mitochondrial synthesis and inhibitor of mitophagy, we
considered the possibility that the cells had accumulated
mitochondria.
To confirm that mTOR is active in balloon cells, we
stained for two downstream markers of mTOR activity
phosphorylated S6 (pS6) and 4E-BP1 (p4E-BP1). In
keeping with previous studies, we found that balloon cells
in FCD and tuberous sclerosis showed extensive staining
for pS6 (n = 5 cases of FCDIIb and 5 cases of TSC) [1, 8,
12, 18]. We found that p4E-BP1 was present in balloon
cells in both diseases but that was more focal and less
intense in FCDIIb than TSC (Fig. 7). This is in keeping
with previous studies showing different patterns of mTOR
activity in FCDIIb and TSC including one study that failed
to demonstrate 4E-BP1 in FCDIIb [1]. Taken together with
the published literature, our data support a role for mTOR
activity in FCDIIb and TSC.
123
94
Acta Neuropathol (2010) 120:85–96
mitochondria (e.g. do they have features of mitochondrial
cytopathy, e.g. enlarged, disorganised cristae or inclusions)
and to confirm the undifferentiated phenotype of the cells.
The cells contained abundant intermediate filaments (in
keeping with the results of the immunohistochemistry) and
frequent mitochondria. In addition, there were scattered
vacuoles. The mitochondria lacked features of pathological
mitochondria (i.e. no inclusions or abnormal architecture).
Importantly, the cultured cells lacked the ultrastructure
features of differentiated cells, e.g. neuro-secretory granules
(or other evidence of synaptic differentiation).
Discussion
Fig. 8 Balloon cells contain numerous mitochondria and intermediate filaments. a, b Cultured balloon cell labelled with mitotracker (a
phase image, b mitotracker). c, d Tissue sections of balloon cells
following histochemistry for mitochondrial enzymes, cytochrome
oxidase (c) or NADH-TR (d). In the tissue sections, the balloon cells
showed intense staining above the background glial tissue and
neuropil. e, f Transmission electron microscopy of cultured balloon
cells confirms the presence of normally formed mitochondria (e) and
abundant intermediate filaments (f). Scale bar 25 lm
To test the hypothesis that balloon cells accumulate
mitochondria, we undertook live labelling of cultured cells
with the mitochondrial marker MitoTracker. We found
intense central staining in the majority of cultured cells with
the typical morphological features of balloon cells
(including enlarged cytoplasm and occasional binucleation;
Fig. 8a, b; Supplementary material 9). To confirm that
balloon cells in vivo showed a similar accumulation, we
undertook histochemistry for the mitochondrial enzymes
NADH-TR reductase and cytochrome oxidase on histological sections of intact tissue. We found intense staining of
balloon cells in vivo in contrast to the surrounding glial cells
or neuropil, confirming the balloon cells contain abundant
mitochondria (Fig. 8c, d). Finally, we undertook electron
microscopy on cultured cells (Fig. 8e, f). The purpose of the
electron microscopy was to confirm the accumulation
of mitochondria, to test whether these were normal
123
We have described the identification, isolation and characterisation of a pathological cell with the phenotype of a
progenitor cell/stem cell from a common malformation of
cortical development in children. This extends the observation of normal stem cells isolated from post-natal brains
and the isolation of tumour stem cells from a wide range of
central nervous system tumours [14, 20]. While a number
of previous studies have suggested that there may be
pathological progenitor cells in the malformed brain [22,
23], one limit to progress in this field has been the absence
of a way to isolate these cells in culture. Furthermore, no
animal models have been described that recapitulate balloon cells [25]. Therefore, there have been no models in
which the biology of these cells can be dissected in a
mechanistic way. Our protocol indicates that this balloon
cell can be isolated and maintained in an undifferentiated
state in culture and can be stored (frozen) and recovered.
This provides an opportunity to study the biology of these
cells and their contribution to the phenotype of the disease
using a reductive experimental model.
We believe that the cells that we have isolated are a
subpopulation of balloon cells. The cells could only be
isolated from cases containing balloon cells and there was a
close relationship between the presence of balloon cells in
the adjacent tissue and balloon cells seen in culture. The
morphology of the cultured cells is characterised by large
amounts of cytoplasm and eccentrically placed often
multiple nuclei, features identical to those of balloon cells
in vivo. The immunophenotype, particularly taken in conjunction with the ultrastructural findings, argues very
strongly against a more differentiated cell. These features
taken together argue that the cells isolated in culture are a
population of balloon cells and effectively exclude the
possibility that they are one of the other populations of
large cells that might be seen in such cases (e.g. hyperplastic astrocytes or neurones).
Several lines of evidence indicate that these cultured
cells are closely related to progenitor cells or stem cells.
Acta Neuropathol (2010) 120:85–96
First, the cells express five different markers of progenitor
cells and stem cells including a number (e.g. CD133)
considered of high specificity for stem cells. Second, we
have excluded the possibility that these cells are related
to an alternative (differentiated) cell by both immunophenotyping and ultrastructural examination. Finally, as
predicted by previous studies, these cells remain in the cell
cycle as indicated by the expression of MCM2 in our study
and a much larger range of markers in previous histological
studies [21, 22].
Our data indicate that b1 integrin is a marker of a relatively homogeneous population of balloon cells with an
undifferentiated (progenitor-like) phenotype that can be
isolated in culture. Many previous studies have emphasised
that balloon cells may show a differentiated phenotype
characterised by mature markers such as Neurofilament,
NeuN and TuJ1 and many others have regarded them as
aberrant glial cells [22, 26]. In contrast, there are a number
of expression studies that have found the expression of
progenitor cell/stem cell markers in a population of balloon
cells [22, 23, 26]. Much of this variability in the published
literature is likely to arise due to an intrinsic heterogeneity
in the balloon cell population and better markers of subtypes of balloon cells are required to explore this
possibility. The selectivity of our culture system for
undifferentiated balloon cells may be a result of using
culture conditions known to promote the survival of progenitor and stem cells [5].
b1 integrin is an intriguing marker of these undifferentiated balloon cells for a number of reasons. It has been
shown to mark a population of stem cells in mice and in
humans and is necessary for proliferation of stem cells
in culture [4, 10, 12, 16]. It is also expressed in other
progenitors such as glial progenitors [11]. Furthermore,
deletion of b1 integrin in the developing mouse cortex
produces a malformation of cortical development [9] raising the possibility that b1 integrin expression in FCD may
be important in the pathogenesis of this malformation.
Although previous reports and our data indicate that
balloon cells remain in the cell cycle, we were unable to
demonstrate significant DNA synthesis despite labelling for
up to 10 days with BrdU. This could be explained in two
ways. The first is that the cells do undergo proliferation but
at an extremely low rate. This would be in keeping with
normal stem cells and tumour stem cells, both populations
of which are distinguished from more differentiated progenitor cells by a relatively slow rate of proliferation (see,
e.g. [13]). The alternative explanation, which we favour, is
that there is a specific defect in the cells that prevents
passage from G1 to S-phase of the cell cycle. This possibility is supported both by our data that indicates that none
of the cultured cells passed through S-phase after 9 days in
culture and by the observations of Thom et al. [21] who
95
Fig. 9 Four possible/hypothesised models of how balloon cells (BC)
contribute to the pathology of focal cortical dysplasia (FCD). The
models in the upper part of the panel represent scenarios in which
BCs are directly contributing or casual to the development of the
dysplastic cortex of FCD lesions. In the top left, neuronal stem cells
(NSC) or progenitor cells give rise to BC, which directly lead to the
development of the dysplastic cortex of FCD. The top right differs in
that BCs directly influence NSC or progenitor cells in a non-cell
autonomous fashion leading to the development of dysplastic cortex.
The models in the lower part of the figure represent scenarios in
which BC are an independent associated feature of FCD. In the
bottom left, the dysplastic cortex has arisen independently and prior to
the advent of BCs but then directly induces NSC or progenitor cells to
form BCs. In the bottom right, NSC or progenitor cells directly form
both the dysplastic cortex and BCs but these two phenomena are
independent of each other
found that balloon cells in tissue sections frequently
express markers of G1 but rarely express markers of later
phases of the cell cycle (in particular G2). This is an
intriguing hypothesis as it raises the possibility that the
development of the disease is caused by a specific cell
cycle defect arising in a progenitor cell population. It is
notable that so far we have not been able to drive differentiation in these cells. However, this is not surprising as
differentiation in stem cell populations is closely dependent
on cell division [3].
It is notable that the phenotype of these balloon cells is
characterised by the accumulation of large amounts of
123
96
Acta Neuropathol (2010) 120:85–96
intermediate filament and mitochondria. The best characterised signalling abnormality in cortical dysplasia is the
over activation of mTOR [1, 8, 12, 19]. Amongst mTOR’s
major functions is inhibition of autophagy, promoting
protein translation and promoting mitochondrial biosynthesis [2, 10, 17]. A tempting hypothesis is that failure to
inhibit mTOR activity in normal neural stem cells leads
to accumulation of filaments and mitochondria and this
may lead to the abnormalities of cell size, division and
differentiation.
It has not been possible previously to explore the contribution of balloon cells to the development of FCD due to
the lack of an animal model containing balloon cells and no
method to isolate the cells from humans. We put forward
four possible models by which balloon cells are implicated
in the pathologies of FCDIIb and TS (Fig. 9). In the upper
part of the panel, balloon cells are directly causative in the
development of a dysplastic cortex either by differentiation
or in a non-cell autonomous manner. In the lower panel,
balloon cells arise either directly as abnormalities in a
normal stem cell or may be induced from a normal stem
cell by factors in the dysplastic cortex. The in vitro isolation of balloon cells represents the first reductive model in
which such a mechanism can be tested. In particular, we
now have the potential to co-culture these cells with other
key cells types (glial cells, normal progenitor cells or
oragnotypic cultures of dysplastic or normal cortex) to
determine their potential contribution to the disease.
Acknowledgments The Great Ormond Street Hospital Children’s
Charity and the Pathological Society of Great Britain have funded this
research. We are grateful to Nigel Weaving, Lillian Martinen and
Kerrie Venner for technical assistance and to Janette Gardener for
administrative assistance.
Conflict of interest statement
None.
References
1. Baybis M, Yu J, Lee A et al (2004) mTOR cascade activation
distinguishes tubers from focal cortical dysplasia. Ann Neurol
56(4):478–487
2. D’Souza AD, Parikh N, Kaech SM, Shadel GS (2007) Convergence of multiple signaling pathways is required to coordinately
up-regulate mtDNA and mitochondrial biogenesis during T cell
activation. Mitochondrion 7(6):374–385
3. Farkas LM, Huttner WB (2008) The cell biology of neural stem
and progenitor cells and its significance for their proliferation
versus differentiation during mammalian brain development.
Curr Opin Cell Biol 20(6):707–715
4. Harvey AS, Cross JH, Shinnar S, Mathern BW, Taskforce IPESS
(2008) Defining the spectrum of international practice in pediatric
epilepsy surgery patients. Epilepsia 49(1):146–155
5. Jacques TS, Relvas JB, Nishimura S et al (1998) Neural precursor
cell chain migration and division are regulated through different
beta1 integrins. Development 125(16):3167–3177
123
6. Jacques TS, Skepper JN, Navaratnam V (1999) Fibroblast growth
factor-1 improves the survival and regeneration of rat vagal
preganglionic neurones following axon injury. Neurosci Lett
276(3):197–200
7. Jacques TS, Swales A, Brzozowski MJ et al (2010) Combinations
of genetic mutations in the adult neural stem cell compartment
determine brain tumour phenotypes. EMBO J 29(1):222–235
8. Ljungberg MC, Bhattacharjee MB, Lu Y et al (2006) Activation
of mammalian target of rapamycin in cytomegalic neurons of
human cortical dysplasia. Ann Neurol 60(4):420–429
9. Luo B-H, Carman CV, Springer TA (2007) Structural basis of
integrin regulation and signaling. Annu Rev Immunol 25:619–
647
10. Ma XM, Blenis J (2009) Molecular mechanisms of mTORmediated translational control. Nat Rev Mol Cell Biol 10(5):307–
318
11. Milner R, Ffrench-Constant C (1994) A developmental analysis
of oligodendroglial integrins in primary cells: changes in alpha vassociated beta subunits during differentiation. Development
120(12):3497–3506
12. Miyata H, Chiang ACY, Vinters HV (2004) Insulin signaling
pathways in cortical dysplasia and TSC-tubers: tissue microarray
analysis. Ann Neurol 56(4):510–519
13. Morshead CM, van der Kooy D (2004) Disguising adult neural
stem cells. Curr Opin Neurobiol 14(1):125–131
14. Okano H, Sawamoto K (2008) Neural stem cells: involvement in
adult neurogenesis and CNS repair. Philos Trans R Soc Lond B
Biol Sci 363(1500):2111–2122
15. Palmini A, Najm I, Avanzini G et al (2004) Terminology and
classification of the cortical dysplasias. Neurology 62(6 Suppl
3):S2–S8
16. Persad S, Attwell S, Gray V et al (2001) Regulation of protein
kinase B/Akt-serine 473 phosphorylation by integrin-linked
kinase: critical roles for kinase activity and amino acids arginine
211 and serine 343. J Biol Chem 276(29):27462–27469
17. Sarbassov DD, Ali SM, Sabatini DM (2005) Growing roles for
the mTOR pathway. Curr Opin Cell Biol 17(6):596–603
18. Schick V, Majores M, Engels G et al (2007) Differential Pi3 Kpathway activation in cortical tubers and focal cortical dysplasias
with balloon cells. Brain Pathol 17(2):165–173
19. Schick V, Majores M, Koch A et al (2007) Alterations of phosphatidylinositol 3-kinase pathway components in epilepsyassociated glioneuronal lesions. Epilepsia 48(Suppl 5):65–73
20. Stiles CD, Rowitch DH (2008) Glioma stem cells: a midterm
exam. Neuron 58(6):832–846
21. Thom M, Martinian L, Sen A et al (2007) An investigation of the
expression of G1-phase cell cycle proteins in focal cortical dysplasia type IIB. J Neuropathol Exp Neurol 66(11):1045–1055
22. Thom M, Martinian L, Sisodiya SM et al (2005) Mcm2 labelling
of balloon cells in focal cortical dysplasia. Neuropathol Appl
Neurobiol 31(6):580–588
23. Urbach H, Scheffler B, Heinrichsmeier T et al (2002) Focal cortical
dysplasia of Taylor’s balloon cell type: a clinicopathological
entity with characteristic neuroimaging and histopathological
features, and favorable postsurgical outcome. Epilepsia
43(1):33–40
24. Vinters HV, Miyata H (2004) Tuberous Sclerosis. In: Golden JA,
Harding BN (eds) Developmental Neuropathology. International
Society of Neuropathology, Basel, pp 79–87
25. Wong M (2009) Animal models of focal cortical dysplasia and
tuberous sclerosis complex: recent progress toward clinical
applications. Epilepsia 50(Suppl 9):34–44
26. Ying Z, Gonzalez-Martinez J, Tilelli C, Bingaman W, Najm I
(2005) Expression of neural stem cell surface marker CD133 in
balloon cells of human focal cortical dysplasia. Epilepsia
46(11):1716–1723