Chapter 19

Isolation and Culture of Primary Glioblastoma

Cells from Human Tumor Specimens
Sascha Seidel, Boyan K. Garvalov, and Till Acker

Abstract
Cultured tumor cells are a central tool in cancer research and have provided fundamental insights in tumor
biology. Recent evidence, however, indicates that classically established cell lines from different tumors,
including glioblastoma, do not fully reflect the genotypes and phenotypes of the respective primary tumors.
By contrast, primary cells, isolated from human tumor samples and maintained in serum-free spheroid
cultures at low passage under defined growth factor conditions, reproduce key aspects of tumor cell physi-
ology much more faithfully. Among the tumor cell characteristics that are better represented in primary
glioblastoma cell cultures is the self-renewal and differentiation potential of the tumor cells. Indeed, a large
body of evidence from the past decade indicates that glioblastomas and other tumors are composed of a
hierarchy of heterogeneous types of cells, which are generated and maintained by cells that share charac-
teristics of stem cells. This cancer stem cell/tumor initiating cell population is optimally preserved and
maintained in primary glioblastoma cultures. Here, we describe a method for the isolation and culture of
primary tumor cells from human glioblastomas in serum-free conditions, which allows the routine genera-
tion and proper maintenance of tumor cells as spheroid cultures. Such primary tumor cultures can serve as
a model of choice for the study of the mechanisms behind key aspects of glioblastoma biology, including
tumorigenicity, stem cell hierarchy, invasion, and therapeutic resistance.

Key words Glioblastoma, Brain tumor, Primary glioblastoma cells, Brain tumor initiating cells, Brain
tumor stem cells, Glioblastoma stem cells, Cancer stem cells, Tumor spheres, Primary culture

1 Introduction

Glioblastomas are the most common and most malignant brain

tumors in adults, with a median patient survival time of less than 15
months [1]. Therefore, an improved understanding of the cellular
and molecular mechanisms controlling glioblastoma growth and
invasion has been the focus of intense research. Among the most
important conceptual advances in the study of glioblastomas over
the last decade has been the realization that these tumors are orga-
nized in a hierarchy of heterogeneous types of tumor cells that
derive from neoplastic cells with stem cell characteristics [2, 3].
Glioblastoma is among the first solid tumor types for which the

Ivan N. Rich (ed.), Stem Cell Protocols, Methods in Molecular Biology, vol. 1235,
DOI 10.1007/978-1-4939-1785-3_19, © Springer Science+Business Media New York 2015

263
264 Sascha Seidel et al.

hierarchy (cancer stem cell) model was demonstrated and has

remained a prototype system for the study of this model [4], which
is now thought to be applicable to various other solid tumors [5, 6].
Traditionally, the cell biology of glioblastoma has been studied
on the basis of cultured tumor cells that were extensively passaged
and grown in adherent monolayers in the presence of serum, an
approach commonly used also for other tumors. While such tumor
cells can be tumorigenic and are convenient to handle and manipu-
late, they usually do not fully reproduce the characteristic invasive
growth phenotype of glioblastomas in xenograft models. Moreover,
they show drastically altered gene expression patterns in vitro com-
pared to the original tumor [7, 8]. Importantly, they are not well
suited for the study of the self-renewal (stem cell) capacity of glio-
blastoma cells, as serum induces profound differentiation, which
may lead to a loss of the cancer stem cell fraction. Therefore, much
recent research in glioma biology has focused on a cell culture sys-
tem of primary tumor cells derived from patient tumors and cul-
tured for a limited number of passages as spheroids in suspension,
in the presence of defined growth factors but in the absence of
serum [2, 9, 10], conditions that were originally established for the
cultivation of neural stem cells [11]. Such cultures retain much
greater similarity to the primary tumor from which they are derived
[7] and their use has enabled the discovery and characterization of
the key features of glioblastoma stem cells. For example, the pri-
mary glioblastoma cell culture system was applied to examine the
role of microenvironmental stimuli on glioblastoma cells. It was
shown that the cancer stem cell phenotype of glioblastoma cells is
induced and maintained in specific niches, including a hypoxic
niche [12–15] and a perivascular niche [10, 14, 16]. It was further
shown that glioblastoma stem cells possess enhanced resistance to
chemotherapy and radiotherapy [17, 18], a property that likely
underlies their ability to drive tumor relapse after treatment.
Interestingly, glioblastoma stem cells derived from primary cul-
tures revealed an unexpectedly broad differentiation potential
in vitro and in vivo, being able to contribute not only to the gen-
eration of the various types of tumor cells but also to the formation
of the tumor vasculature, through transdifferentiation into endo-
thelial cells and pericytes [19–21]. Furthermore, primary glioblas-
toma cells were employed to elucidate the role of inflammatory
responses on tumor growth and the interaction between glioblas-
toma stem cells and immune cells [22–24].
In this chapter, we describe a method that has been routinely
used in our laboratory for isolating primary tumor cells from
human glioblastomas and their cultivation in vitro. The cells gen-
erated in this manner are maintained as spheroids in serum-free
suspension cultures with defined growth factors. These conditions,
which are analogous to the ones used for the culture of physiological
stem cells, have also proven to be particularly suitable for the study
 Primary Glioblastoma Cell Culture 265

of the cancer stem cell and brain tumor initiating phenotype of

glioblastoma cells, as well as for many other features of glioblas-
toma biology.

2 Materials

2.1 Tumor Dissection 1. Phosphate buffered saline (PBS) without calcium and
magnesium, sterile (PAA).
2.1.1 Preparation
of Buffers and Solutions 2. 10× Hanks Balanced Salt Solution (HBSS) without calcium
and magnesium, sterile (Gibco).
3. Eagle’s Balanced Salt Solution (EBSS) without calcium and
magnesium, sterile (Gibco).
4. D-Glucose solution (300 mg/mL), sterile filtered.
5. Sodium chloride solution (0.15 M), sterile filtered.
6. Bovine serum albumin (BSA), cell culture tested (Sigma).
7. Amphotericin B (250 μg/mL) (Sigma).
8. Gentamicin (50 mg/mL) (PAA).
9. HEPES buffer, 1 M, sterile (PAA).
10. Ultrapure water (18.2 MΩ).
11. Papain (Worthington).
12. DNase I (Sigma).
13. Trypsin (Sigma).
14. Collagenase I (Worthington).
15. Hyaluronidase (Sigma).
16. Red Blood Cell Lysis Buffer (Roche Diagnostics).
17. 0.22 μm sterile syringe filters.
18. Sterile syringes.
19. 0.22 μm Stericup-GV filters, 150 and 500 mL (Millipore).
20. Autoclaved glass bottles 500 mL.
21. Sterile conical tubes.

2.1.2 Tissue Dissociation 1. Sterile 10 cm petri dishes (Greiner Bio-One; see Note 1).
2. Sterile scalpel with a #21 blade.
3. Autoclaved single edged razor blades.
4. Autoclaved forceps.
5. Sterile cell strainers, 70 and 100 μm.
6. Sterile 15 and 50 mL conical tubes.
7. Sterile plastic pipettes.
8. Water bath at 37 °C.
266 Sascha Seidel et al.

2.2 Cell Culture 1. Dulbecco’s Modified Eagle’s Medium (DMEM)/F-12 +

GlutaMAX (Gibco).
2. B27 Supplement minus vitamin A (Gibco; see Note 2).
3. Recombinant human EGF (PeproTech).
4. Recombinant human bFGF (PeproTech).
5. Accutase solution (PAA).
6. Cryo-SFM cryopreservation medium (Promo Cell).
7. Sterile 94 mm petri dishes (Greiner Bio-One).
8. Cell strainer 40 μm.
9. Freezing container (e.g., Mr. Frosty).

3 Methods

3.1 Tissue In the following section, we describe a method for the isolation
Preparation and the establishment of primary glioblastoma cell lines from
human tumor specimens. All steps of tissue dissociation should be
carried out under sterile conditions in a laminar flow cell culture
hood to reduce the risk of contamination of the primary culture.

3.1.1 Preparation Wash Buffer:

of Buffers and Solutions
1. Add 450 mL PBS without calcium and magnesium to an auto-
claved Erlenmeyer flask.
2. Dissolve 25 g of cell culture tested BSA in the PBS.
3. Add 5 mL amphotericin B (250 μg/mL).
4. Add 500 μL gentamicin (50 mg/mL) (see Note 3).
5. Fill up to 500 mL with PBS without calcium and magnesium.
6. Mix well and filter-sterilize with a 0.22 μm Stericup Filter into
an autoclaved glass bottle, store the Wash Buffer at 4 °C.
Papain stock solution:
1. Dissolve the Papain Suspension (100 mg Papain, Worthington)
in 5 mL EBSS without calcium and magnesium (final concen-
tration 20 mg/mL = 500 U/mL).
2. Filter-sterilize with a 0.22 μm Syringe Filter.
3. Store the Suspension at 4 °C.
DNase I Stock Solution:
1. Dissolve 100 mg (500 U/mg) DNase I in 5 mL sterile 0.15 M
NaCl.
2. Prepare 50 μl Aliquots and store them at −20 °C.
 Primary Glioblastoma Cell Culture 267

Dissociation solution:
1. Add 400 mL ultrapure water to an autoclaved Erlenmeyer
flask.
2. Add 50 mL 10× HBSS without calcium and magnesium.
3. Add 9 mL D-glucose solution (300 mg/mL).
4. Add 7.5 mL 1 M HEPES buffer.
5. Mix carefully and adjust the pH to 7.4.
6. Add ultrapure water to a final volume of 500 mL.
7. Filter-sterilize with a 500 mL 0.22 μm Stericup filter into an
autoclaved glass bottle.
8. Prepare 5 mL aliquots in sterile 15 mL conical tubes and 50 mL
aliquots in sterile 50 mL tubes and store them at −20 °C.
Attention: 100 mL of dissociation solution is required for the
Dissociation Medium 2. If the dissociation solution is prepared
freshly do not freeze the whole amount.
Dissociation Medium 1:
Prepare freshly for every tumor dissection!
1. Thaw one 5 mL aliquot of Dissociation solution.
2. Transfer 4.75 mL of Dissociation solution to a new sterile
conical 15 mL tube.
3. Add 200 μL of the papain stock solution.
4. Activate the above mixture at 37 °C for 30 min prior to use.
5. Thaw one 50 μL aliquot of the DNase I stock solution on ice.
6. Add the DNase I aliquot and mix carefully.
Dissociation Medium 2:
1. Pour 100 mL of pre-cooled Dissociation solution into an auto-
claved Erlenmeyer flask on ice.
2. Add 70 mg of collagenase I to a final concentration of 0.7 mg/mL.
3. Add 70 mg of hyaluronidase to a final concentration of
0.7 mg/mL.
4. Add 100 mg of trypsin to a final concentration of 1 mg/mL.
5. Mix carefully until all the enzymes are completely dissolved.
6. Filter-sterilize with a 150 mL 0.22 μm Stericup filter into an
autoclaved glass bottle.
7. Prepare 5 mL aliquots in sterile 15 mL conical tubes and store
them at −20 °C.
Prior to Tumor dissociation:
1. Thaw one 5 mL aliquot of the above mixture.
2. Thaw one 50 μL aliquot of the DNase I stock solution on ice.
268 Sascha Seidel et al.

3. Add the DNase I aliquot and mix carefully to obtain

Dissociation Medium 2.
Growth Factor Stock Solutions:
1. Reconstitute lyophilized recombinant human EGF and recom-
binant human bFGF following the supplier’s recommenda-
tions in 5 mM Tris, 0.1 % BSA, pH 7.6 to a final concentration
of 20 μg/mL.
2. Prepare 500 μL aliquots in sterile 1.5 mL tubes.
3. For long term storage keep the aliquots at −20 °C, the aliquot
in use can be stored at 4 °C for up to 1 week.
Tumor Sphere Culture Medium
1. Take 480 mL DMEM/F-12 + GlutaMAX.
2. Add 5 mL amphotericin B (250 μg/mL).
3. Add 500 μL gentamicin (50 mg/mL) (see Note 3),
4. Add 2.5 mL HEPES Buffer (1 M).
5. Add 10 mL B27 Supplement minus vitamin A (see Notes 2
and 4).
6. Mix well, protect the bottle from light by wrapping it in alumi-
num foil.
7. Store the medium at 4 °C.

3.1.2 Tissue Dissociation 1. Transfer the tumor specimen into a sterile petri dish using
autoclaved forceps (see Notes 5 and 6).
2. Wash the tumor specimen two times with wash buffer and
discard the buffer by sucking it off.
3. Cut the tissue into small fragments using a sterile scalpel with
a #21 blade (see Note 7).
4. Chop/cut the tumor fragments with a sterile single edged
razor blade until it is well minced (see Note 7).
5. Transfer the tumor tissue to a sterile 15 mL conical tube by
flushing the dish with the 5 mL activated Dissociation Medium
1 and pipetting it with a sterile pipette (see Note 8).
6. Incubate the tube in a 37 °C water bath for 30 min. During
the incubation, take the tube out of the water bath every 5 min
and mix gently with a 5 mL pipette (around 20 times). Avoid
bubbling (see Note 8).
7. Centrifuge the tube for 5 min at 300 × g.
8. Suck off the supernatant carefully (the pellet is not solid!)
(see Note 9).
9. Resuspend the pellet in 5 mL Dissociation Medium 2.
(see Note 8).
 Primary Glioblastoma Cell Culture 269

10. Incubate the tube in a 37 °C water bath for 30 min. During

the incubation take the tube out of the water bath every 5 min
and mix gently with a 5 mL pipette (around 20 times). Avoid
bubbling (see Note 8).
11. Centrifuge the tube for 5 min at 300 × g.
12. Suck off the supernatant carefully (the pellet is not solid!)
(see Note 9).
13. Resuspend the pellet in 2 mL of Red Blood Cell Lysis Buffer
and incubate the tube for 10 min at room temperature
(see Note 10).
14. Add 8 mL of wash buffer.
15. Centrifuge the tube for 5 min at 300 × g, and prepare a conical
50 mL tube with a 100 μm cell strainer.
16. Carefully remove the supernatant and resuspend in 10 mL
wash buffer.
17. Filter the suspension through the cell strainer (see Note 11).
18. Wash the filter with 10 mL wash buffer.
19. Centrifuge the tube for 5 min at 300 × g, in the meantime
prepare a conical 50 mL tube with a 70 μm cell strainer.
20. Filter the suspension through the cell strainer (see Note 11).
21. Centrifuge the tube for 5 min at 300 × g.
22. Resuspend the cell pellet in Tumor Sphere Culture Medium.

3.2 Cultivation This section describes how to establish and maintain the primary
of Primary tumor cells, obtained and plated during the previous procedure.
Glioblastoma Cells
1. Count the cells from step 22 in the previous section
(Subheading 3.1.2).
2. Plate them at a density of 2 × 106 cells in 10 mL of Tumor
Sphere Culture Medium in a 10 cm petri dish.
3. Add 10 μL (1:1,000) of the EGF and bFGF stock solutions to
a final concentration of 20 ng/mL.
4. After 24 h, collect the cells with a plastic pipette, transfer them
to a 15 mL conical tube, and centrifuge at 190 × g for 3 min
(see Note 12 and Fig. 1a).
5. Resuspend the pellet in 10 mL Tumor Sphere Culture Medium
and transfer them to the same petri dish from which they were
collected.
6. Add 10 μL (1:1,000) of the EGF and bFGF stock solutions to
a final concentration of 20 ng/mL.
7. The culture medium should be refreshed twice a week by
transferring the cells with a plastic pipette to a 15 mL conical
tube and centrifuging at 190 × g for 3 min.
270 Sascha Seidel et al.

Fig. 1 Primary glioblastoma cells in culture. (a) Isolated glioblastoma cells 12 h after tumor dissection. The
culture consists of tumor cells, which partly already start to form tumor spheres (arrows), some cell debris
(asterisks), and non-tumor cells that adhere to the petri dish (arrowheads). (b) Tumor Spheres of primary glio-
blastoma cells 5 days after splitting. Scale bars 100 μm

When refreshing the medium for the first time aspirate half of
the old medium and replace it with an equal volume of fresh
medium (i.e., refresh half of the medium). Afterwards re-plate the
cells in the original dish.
When refreshing the medium for the second time, aspirate all
the medium and replace it with fresh medium. Thereafter, re-plate
the cells in a new dish. Always add 10 μL (1:1,000) of the EGF and
bFGF stock solutions to a final concentration of 20 ng/mL after
changing the medium.
The obtained culture of primary glioblastoma cells should form
three-dimensional spheroids suspended in the medium (tumor
spheres) around 24–48 h after the dissociation (Fig. 1). The frequency
of sphere formation depends on the characteristics of the primary cell
line, e.g., the number of brain tumor initiating cells in the original
tumor specimen and their self-renewal capacity (see Notes 13–15).

3.2.1 Splitting of Tumor Depending on their size, tumor spheres of primary glioblastoma
Spheres cells should be split every 7–10 days.
1. Centrifuge the tumor spheres for 3 min at 190 × g.
2. Aspirate the medium and resuspend the pellet in 1 mL of
Accutase solution.
3. Incubate the spheres for 15 min at 37 °C.
4. In the meantime prepare a sterile 50 mL conical tube with a
40 μm cell strainer.
5. Mix the cell suspension with a 1,000 μL pipette tip until there
are no visible spheres left.
6. Add 9 mL of Tumor Sphere Culture Medium and filter
through the cell strainer.
 Primary Glioblastoma Cell Culture 271

7. Centrifuge the cells for 3 min at 190 × g.

8. Remove the supernatant and resuspend the pellet in 10 mL of
Tumor Sphere Culture Medium.
9. Plate the cells at a density of 2 × 106 cells in 10 mL of Tumor
Sphere Culture Medium in 10 cm petri dishes.
10. Add 10 μL (1:1,000) of the EGF and bFGF stock solutions to
a final concentration of 20 ng/mL.

3.2.2 Cryopreservation The primary glioblastoma cells should be used at low passage num-
of Primary ber, as there is a risk of genetic alterations and phenotypic shifts
Glioblastoma Cells over the course of prolonged culture with multiple passaging.
Therefore, a substantial number of aliquots should be cryopre-
served at the earliest passage when they can be sufficiently expanded.
Subsequently, fresh frozen aliquots should be periodically thawed
and cultured for a limited number of passages and new fresh ali-
quots at low passages should be frozen.
1. Split and seed the cells as described in Subheading 3.2.1.
2. Spin down the spheres 24 h after the re-seeding for 3 min at
190 × g.
3. Resuspend the pellet of one petri dish in 2 mL of Cryo-SFM
cryopreservation medium.
4. Transfer the cells suspension to cryo vials at 1 mL per vial.
5. Transfer the vials to a freezing container at room
temperature.
6. Freeze the cells slowly in the freezing container for 24 h
at −80 °C.
7. Transfer the cells to liquid nitrogen for long-term storage.

4 Notes

1. Petri dishes used must not have a specially treated surface for
cell culture. As the primary glioblastoma cells are maintained
as spheroids in suspension, the surface of the dish should pre-
vent, rather than promote the attachment of the cells.
2. The B27 Supplement is light sensitive. Bottles containing
medium with B27 should be wrapped in aluminum foil to pre-
vent exposure to light.
3. The addition of antibiotics and antimycotics to the buffers and
cell culture media is optional, but can reduce the risk of con-
tamination, especially if it is not clear whether the tissue sam-
ples were collected under fully sterile conditions.
4. The stability of the B27 Supplement can be affected by long-
term storage at 4 °C. Therefore, if the throughput of Tumor
272 Sascha Seidel et al.

Sphere Culture Medium is low, add B27 Supplement minus

vitamin A from frozen aliquots stored at -20 ºC directly to the
medium in the culture dish. In this case, the B27 Supplement
should be diluted 1:50.
5. The yield and the quality of the isolated primary cells strongly
depend on the quality of the provided tumor material. We rec-
ommend to contact your department of neurosurgery and dis-
cuss with them an optimal way for the collection and transfer
of the tumor material. At the same time we recommend to
obtain the clinical information about the patient regarding
potential viral infections (e.g., HIV, Hepatitis A/B/C), which
would preclude using the tumor specimen for research pur-
poses in a routine lab environment. The tumor material should
be processed as fast as possible after the resection. We suggest
that the material should be collected either in a sterile 50 mL
conical tube filled with 20 mL cold PBS, or in a sterile screw
lid sample container on sterile gauze soaked with PBS to pre-
vent the tissue from drying. Until the cell isolation procedure
is performed, the tumor material should be kept on ice. If you
process more than one tumor sample at a time, we recommend
using a separate set of instruments for tissue dissection to pre-
vent cross-contamination.
6. It is important to obtain ethical approval from the responsible
Institutional Review Board (IRB) before working with human
tissue samples. We strongly recommend the wearing of suitable
personal protective equipment (laboratory coat, protective
goggles, gloves) during the tissue dissociation, as all human
tissue is potentially infectious and sometimes insufficiently
tested for infectious agents such as hepatitis virus or HIV. For
the same reasons it is recommended that the laboratory per-
sonnel performing the primary cell isolation undergo immuni-
zation against hepatitis A/B.
7. Mincing the tissue well can take time. Take particular care dur-
ing this step when handling scalpels and razor blades to pre-
vent injuries, as the tumor material has to be seen as potentially
infectious.
8. The dissociation procedure described here is suitable for most
tumor samples with a size of up to 0.5 cm3. Nevertheless, dif-
ferent tumors samples may have different textures, cell density,
amounts of necrosis and may originate from different brain
regions. Therefore, for large tumor samples, it might be neces-
sary to increase the amount of Dissociation Media 1 and 2, or
the incubation time for the cell dissociation.
9. During the dissociation procedure care should be taken while
removing the supernatant during the centrifugation steps,
especially when aspirating the supernatant with a pump. The
pellet usually is not solid and homogenous. The cells are
 Primary Glioblastoma Cell Culture 273

located at the very bottom of the conical tube and are overlaid
with debris. The latter appears as the upper layer of the pellet
and, especially during the first dissociation step, can be quite
viscous and sticky due to released DNA and extracellular matrix
proteins. Hence, there is the danger of aspirating the cell pellet
together with the supernatant and the debris. Therefore, the
supernatant must be aspirated slowly and carefully.
10. Depending on the tumor sample it could be necessary to repeat
the Red Blood Cell Lysis step. If the cell pellet appears to be
still reddish after the washing step, repeat the incubation with
Red Blood Cell Lysis Buffer for 5 min.
11. Depending on the amount and density of the tumor, the cell
suspension may be quite thick. In such cases, pipette directly
onto the cell strainer with increased pressure or press the sus-
pension through the cell strainer with the pipette tip.
12. At this stage the culture could still contain a small amount of
debris and non-tumor cells, which usually adhere to the sur-
face of the petri dish, whereas the tumor cells float in suspen-
sion, some of them already forming small spheres (Fig. 1a).
The non-tumor cells will not proliferate under the culture con-
ditions used for the primary tumor cells and are going to die
out. The debris will usually be washed out during the medium
change the day after tumor dissection.
13. The primary glioblastoma cells isolated using the described
protocol show self-renewal capacity and stem cell characteris-
tics. This is reflected, for example, by their ability to form
spheres in culture (Fig. 1), which express neural stem cell
markers such as nestin (Fig. 2a). In addition, they express
established brain tumor initiating cell markers such as CD133,
Oct4, and others (Fig. 2b; [14]). The expression of such stem
cell markers is enhanced under tumor sphere conditions,
whereas addition of serum downregulates the stem cell genes
and upregulates differentiation markers such as the astrocytic
protein GFAP (Fig. 2b).
14. In addition to the isolation of primary glioblastoma cells from
human tumor specimens, our method could be used to re-
isolate glioblastoma cells from orthotopic or subcutaneous
tumor xenografts. In this case, the cultures will initially be a
mix of murine and human cells, but usually after two passages
the culture becomes homogeneous and only consists of human
tumor cells, which outgrow the mouse cells.
15. The described protocol has been tested on numerous brain
tumor samples and in the majority of cases has led to the suc-
cessful isolation of primary glioblastoma cells. However, not all
cultures will start growing, or the cells could stop proliferating
after one or two passages. This usually depends on the quality
274 Sascha Seidel et al.

Nestin DAPI Merge

b
40.0
TSM
30.0
FCS
mRNA expression

20.0
2.0

1.5

1.0

0.5

0.0
CD133 Oct4 GFAP

Fig. 2 Primary glioblastoma cells express cancer stem cell markers and retain their differentiation capacity.
(a) Tumor spheres of primary glioblastoma cells were spun on a microscope slide using a Cytospin centrifuge
and stained for the neural stem cell marker nestin (green) and DAPI to visualize the nucleus (blue). Scale bar
20 μm. (b) Quantitative real time RT-PCR of primary glioblastoma cells grown under sphere conditions (in
tumor sphere culture medium, TSM) or as adherent cultures supplemented with 10 % fetal calf serum (FCS).
Under sphere conditions the cancer stem cell marker CD133 and the embryonic stem cell marker Oct4 are
highly expressed, whereas their expression decreases when cells undergo differentiation in the presence of
FCS. By contrast, expression of the glial lineage differentiation marker GFAP is highly upregulated by FCS
addition

of the tumor sample, as well as the fraction of brain tumor

initiating cells in the tumor specimen and their self-renewal
capacity. We therefore encourage the researcher to process
several new tumor samples if the method does not work at
first try.

Acknowledgements

This work was supported by grants from the DFG (AC 110/4
(KFO210); AC110/3-1, 3-2 (SPP1190); the ECCPS (EXC 147)),
the Deutsche Krebshilfe (107231), the von Behring-Röntgen
 Primary Glioblastoma Cell Culture 275

Foundation, LOEWE (OSF), and the German Ministry of

Education and Research (BMBF) within the National Genome
Network (NGFNplus) and Brain Tumor Network.

References
1. Stupp R, Hegi ME, Mason WP et al (2009) glioblastoma stem cells and promotes repro-
Effects of radiotherapy with concomitant and gramming towards a cancer stem cell pheno-
adjuvant temozolomide versus radiotherapy type. Cell Cycle 8:3274–3284
alone on survival in glioblastoma in a randomised 13. Li Z, Bao S, Wu Q et al (2009) Hypoxia-
phase III study: 5-year analysis of the EORTC- inducible factors regulate tumorigenic capacity
NCIC trial. Lancet Oncol 10:459–466 of glioma stem cells. Cancer Cell 15:501–513
2. Singh SK, Clarke ID, Terasaki M et al (2003) 14. Seidel S, Garvalov BK, Wirta V et al (2010) A
Identification of a cancer stem cell in human hypoxic niche regulates glioblastoma stem cells
brain tumors. Cancer Res 63:5821–5828 through hypoxia inducible factor 2α. Brain
3. Singh SK, Hawkins C, Clarke ID et al (2004) 133:983–995
Identification of human brain tumour initiat- 15. Soeda A, Park M, Lee D et al (2009) Hypoxia
ing cells. Nature 432:396–401 promotes expansion of the CD133-positive
4. Filatova A, Acker T, Garvalov BK (2013) The glioma stem cells through activation of HIF-
cancer stem cell niche(s): the crosstalk between 1alpha. Oncogene 28:3949–3959
glioma stem cells and their microenvironment. 16. Calabrese C, Poppleton H, Kocak M et al
Biochim Biophys Acta 1830:2496–2508 (2007) A perivascular niche for brain tumor
5. Garvalov BK, Acker T (2011) Cancer stem stem cells. Cancer Cell 11:69–82
cells: a new framework for the design of tumor 17. Bao S, Wu Q, McLendon RE et al (2006)
therapies. J Mol Med (Berl) 89:95–107 Glioma stem cells promote radioresistance by
6. Visvader JE, Lindeman GJ (2012) Cancer stem preferential activation of the DNA damage
cells: current status and evolving complexities. response. Nature 444:756–760
Cell Stem Cell 10:717–728 18. Eramo A, Ricci-Vitiani L, Zeuner A et al (2006)
7. Lee J, Kotliarova S, Kotliarov Y et al (2006) Chemotherapy resistance of glioblastoma stem
Tumor stem cells derived from glioblastomas cells. Cell Death Differ 13:1238–1241
cultured in bFGF and EGF more closely mirror 19. Cheng L, Huang Z, Zhou W et al (2013)
the phenotype and genotype of primary tumors Glioblastoma stem cells generate vascular peri-
than do serum-cultured cell lines. Cancer Cell cytes to support vessel function and tumor
9:391–403 growth. Cell 153:139–152
8. Li A, Walling J, Kotliarov Y et al (2008) 20. Soda Y, Marumoto T, Friedmann-Morvinski D
Genomic changes and gene expression profiles et al (2011) Transdifferentiation of glioblas-
reveal that established glioma cell lines are toma cells into vascular endothelial cells. Proc
poorly representative of primary human glio- Natl Acad Sci U S A 108:4274–4280
mas. Mol Cancer Res 6:21–30 21. Wang R, Chadalavada K, Wilshire J et al (2010)
9. Galli R, Binda E, Orfanelli U et al (2004) Glioblastoma stem-like cells give rise to tumour
Isolation and characterization of tumorigenic, endothelium. Nature 468:829–833
stem-like neural precursors from human glio- 22. Tafani M, Di Vito M, Frati A et al (2011)
blastoma. Cancer Res 64:7011–7021 Pro-inflammatory gene expression in solid
10. Bao S, Wu Q, Sathornsumetee S et al (2006) glioblastoma microenvironment and in
Stem cell-like glioma cells promote tumor hypoxic stem cells from human glioblastoma.
angiogenesis through vascular endothelial J Neuroinflammation 8:32
growth factor. Cancer Res 66:7843–7848 23. Wu A, Wei J, Kong LY et al (2010) Glioma can-
11. Reynolds BA, Weiss S (1992) Generation of cer stem cells induce immunosuppressive macro-
neurons and astrocytes from isolated cells of phages/microglia. Neuro Oncol 12:1113–1125
the adult mammalian central nervous system. 24. Yi L, Xiao H, Xu M et al (2011) Glioma-
Science 255:1707–1710 initiating cells: a predominant role in microglia/
12. Heddleston JM, Li Z, McLendon RE (2009) macrophages tropism to glioma. J Neuroimmunol
The hypoxic microenvironment maintains 232:75–82