Login processing...

Trial ends in Request Full Access Tell Your Colleague About Jove

Cancer Research

A 3D Spheroid Model for Glioblastoma

doi: 10.3791/60998 Published: April 9, 2020


Here, we describe an easy-to-use invasion assay for glioblastoma. This assay is suitable for glioblastoma stem-like cells. A Fiji macro for easy quantification of invasion, migration, and proliferation is also described.


Two-dimensional (2D) cell cultures do not mimic in vivo tumor growth satisfactorily. Therefore, three-dimensional (3D) culture spheroid models were developed. These models may be particularly important in the field of neuro-oncology. Indeed, brain tumors have the tendency to invade the healthy brain environment. We describe herein an ideal 3D glioblastoma spheroid-based assay that we developed to study tumor invasion. We provide all technical details and analytical tools to successfully perform this assay.


In most studies using primary or commercially available cell lines, assays are performed on cells grown on plastic surfaces as monolayer cultures. Managing cell culture in 2D represents disadvantages, as it does not mimic an in vivo 3D cell environment. In 2D cultures, the entire cell surface is directly in contact with the medium, altering cell growth and modifying drug availability. Furthermore, the nonphysiological plastic surface triggers cell differentiation1. Three-dimensional culture models have been developed to overcome these difficulties. They have the advantage of mimicking the multicellular architecture and heterogeneity of tumors2, and thus could be considered to be a more relevant model for solid tumors3. The complex morphology of spheroids contributes to better evaluate drug penetrance and resistance4. The tumor heterogeneity in the spheroid impacts the diffusion of oxygen and nutrients, and the response to pharmacological agents (Figure 1A). Diffusion of oxygen is altered when the spheroid size reaches 300 µm, inducing a hypoxic environment in the center of the spheroid (Figure 1A,C). Metabolites are also less penetrating through the cell layers and compensating metabolic reactions take place5. When the diameter of the spheroid increases, necrotic cores can be observed, further mimicking characteristics found in many solid cancers, including the aggressive brain cancer glioblastoma (GBM)6.

Several 2D or 3D invasion assays for glioblastoma have been reported in the literature7,8. Two-dimensional assays are mainly for studying invasion in a horizontal plane on a thin matrix layer or in a Boyden chamber assay9. Three-dimensional assays have been described with 3D spheroid cultures using classical glioblastoma cell lines10. More complex variants are represented by invasion of brain organoids by tumor spheroids in confrontation cultures11. However, it is still important to develop an easy-to-use and reproducible assay available to any laboratory. We have developed a protocol to generate glioblastoma stem-like cells from patient samples. The quantification of these assays is easily manageable and only requires open-access online software. Briefly, tumor pieces are cut into small pieces and enzymatically digested. Single cells derived from the digestion are cultivated in neurobasal medium. After 4-7 days, spheroid structures form spontaneously. Upon intracranial implantation in mice models, they form tumors exhibiting a necrotic core surrounded by pseudo-palisading cells12. This closely resembles the characteristics found in GBM patients.

In this article, we describe our protocol to produce spheroids from a determined number of cells to ensure reproducibility. Two complementary matrices can be used for this purpose: Matrigel and collagen type I. Matrigel is enriched in growth factors and mimics the mammalian basal membrane required for cell attachment and migration. On the other hand, collagen type I, a structural element of stroma, is the most common fibrillary extracellular matrix and is used in cell invasion assays. Herein, we illustrate our GBM spheroid model by performing migration and proliferation assays. Analysis was done not only at fixed time points but also by monitoring spheroid expansion and cell movement by live imaging. Furthermore, electron microscopy was done to visualize morphological details.

Subscription Required. Please recommend JoVE to your librarian.


Informed written consent was obtained from all patients (from the Haukeland Hospital, Bergen, Norway according to local ethics committee regulations). Our protocol follows the guidelines of our institution’s human research ethics committee. 

1. Generation of uniform size tumor spheroids

NOTE: Stem-like cells are cultured in neurobasal medium complemented with B27 supplement, heparin, FGF-2, penicillin, and streptomycin, as described in previous articles12. These cells spontaneously form spheroids in culture.

  1. Wash the tumor cells with 5 mL phosphate-buffered saline (PBS) and incubate the cells with 0.5-1 mL dissociation enzyme (see Table of Materials) for 5 minutes at 37 °C.
  2. Wash with 4-4.5 mL PBS and add 10 mL of the complete growth medium (complete neurobasal medium, cNBM).
  3. Count the cells using an automatic counting technique with trypan blue and a cell counting chamber slide.
  4. To generate 100 spheroids with 104 cells per spheroid (according to preferred size), mix 106 cells in 8 mL of NBM with 2 mL of 2% methylcellulose.
  5. Transfer the suspension to a sterile system container and dispense 100 µL/well with a multichannel pipette into a 96 well round bottom plate.
  6. Incubate the plate at 37 °C, 5% CO2 and 95% humidity. Equal sized spheroids will form and can be used after 3-4 days.

2. Three-dimensional experiments

  1. Proliferation
    1. Preparation
      1. Suspend inhibitors (e.g., rotenone as in Figure 4A) and chemicals in 100 µL of medium and add to the 100 µL of medium in each well (i.e., one spheroid per well).
      2. Incubate the plate at 37 °C, 5% CO2, and 95% humidity.
    2. Image acquisition and analysis
      1. Take pictures with a video microscope in brightfield to create a series of conditions at T0 and the following times expected.
      2. Use Fiji to analyze pictures either manually or in a semiautomated manner. To do so manually, draw a circle around the spheroid core with the freehand selection tool and measure the area of each spheroid. To analyze the images in a semiautomated manner, use the macro shown in Supplementary Document 1 with //Core Area only.
  2. Invasion
    1. Preparation
      1. Prepare the collagen matrix in a tube on ice with type I collagen at 1 mg/mL final concentration, 1x PBS, 0.023xVcollagen, 1 M sodium hydroxide, and sterile H2O. Incubate the solution on ice for 30 min.
      2. Collect spheroids from the round bottom well plate in 500 µL tubes and wash 2x with 200 µL 1x PBS.
      3. Pipette the spheroids carefully into 100 µL of the collagen matrix and insert in the center of a well in normal 96 well plates.
      4. Incubate the collagen gel for 30 min at 37 °C and then add cNBM on top of the gel. Inhibitors or activators (e.g., hydrogen chloride as shown in Figure 4B, 4C) can be added to the medium at this step.
    2. Image acquisition and analysis
      1. Take pictures sequentially with a video microscope in brightfield mode 24 h after collagen inclusion.
      2. Use Fiji to analyze pictures either manually or in a semiautomated manner. To do so manually, draw around the core and the total area of the spheroid with the freehand selection tool and measure the invasive area of each spheroid by subtract total area with core area. To analyze the images in a semiautomated manner, use the macro as indicated in Supplementary Document 1 to determine the invasive area.
  3. Migration
    1. Preparation
      1. Coat a 6 well plate with Matrigel (0.2 mg/mL) in NBM for 30 min at 37 °C, then remove the Matrigel and add 2 mL of cNBM.
      2. Transfer spheroids into 50 µL of cNBM from the round bottom well plate to the 6 well plate.
      3. Incubate the plate at 37 °C and wait 30 min for the spheroids to adhere.
      4. After 24 h of incubation, stain with 10 ng/mL of Hoechst and incubate 30 min at 37 °C.
    2. Image acquisition and analysis
      1. Obtain images using a video microscope in brightfield. A 405 nm laser is used for visualization of Hoechst staining.
      2. Use Fiji software to analyze pictures and run the macro as indicated in Supplementary Document 1.
        NOTE: Touching the bottom of the well or completely removing the supernatant damages the spheroids. For collagen type I gel handling, keep the gel on ice to avoid collagen polymerization, do not add an acidic component because the change in pH will affect the compactness of the gel, and pipette cells rapidly into the collagen to prevent cell death and the degradation of the gel.

3. Fiji Macro

NOTE: Fiji is an image analysis program developed in the public domain that allows the development of macros to speed up image analysis. Manual analysis is also possible, but this is a slow process and may introduce biases. Images can be imported by drag-and-drop in the software and quantified with the ROI Manager Tools plugin. The procedure used in this study is described below:

  1. Open the macro window: Plugins | Macros | Interactive Interpreter.
  2. Copy and paste the following adapted purple loop. Keep the purple sentences and add the green sentences of interest (Supplementary Document 1).
  3. To analyze the entire series, adjust the parameters in red for a specific quantification and run the macro using Macros | Run Macro or pressing Ctrl+R.
  4. Check and, if necessary, manually adapt the region of interest (ROI).

4. Electron microscopy of spheroids

NOTE: Most of the following steps must be done in a chemical hood.

  1. Fixation step
    1. Collect the spheroid with a cut pipette tip, put it in a 1.5 mL tube, and wash 1x with 0.1 M phosphate buffer (PB).
    2. Fix the spheroid overnight at 4 °C in 2% glutaraldehyde/2% paraformaldehyde (PFA) in 0.1 M PB.
    3. Replace the fixation solution with a solution of 1% PFA in 0.1 M PB followed by sample preparation.
  2. Sample preparation
    1. Transfer the spheroids into a strainer and put them in a glass beaker in order to avoid spheroid damage.
    2. Carefully wash 3x with 0.1 M PB.
    3. Incubate with osmium for 2 h in the dark. Dilute osmium to 4% in 1% 0.1 M PB buffer.
    4. Carefully wash 3x with 0.1 M PB.
      1. Dehydrate as follows: soak in 50% ethanol for 10 min, 70% ethanol for 10 min, 2x 90% ethanol for 15 min, 2x 100% ethanol for 20 min, and 2x acetone for 30 min.
    5. Incubate the samples in a 50/50 mixture of acetone/resin for 2 h. During this step, prepare the EPON resin (Embed-812: 11.25 g; DDSA: 9 g; NMA: 4.5 g).
    6. Discard the acetone/resin mixture, replace with freshly prepared resin and incubate overnight.
    7. Replace the resin by a new one and incubate between 2-6 h.
    8. Add the spheroids in resin into a mold at 60 °C for 48-72 h.

Subscription Required. Please recommend JoVE to your librarian.

Representative Results

Spheroids were prepared as described in the protocols section and observations were made regarding migration, invasion, proliferation, and microscopy. To measure hypoxia in distinct areas of the spherical structure, carboxic anhydrase IX staining was used for determining hypoxic activity (Figure 1A-C). More CAIX-positive cells were observed in the spheroid center (Figure 1A-C). Hypoxic cells located in the spheroid core tend to be more glycolytic than the surrounding ones. Mitochondria can be imaged for further analyses as shown by electron microscopy (Figure 1Ba-1Bb). Spheroids composed of 2.5 x 103, 5 x 103, 104, or 2 x 104 cells exhibit a spheroid diameter of about 350, 400, 500, or 650 µm respectively (Figure 2A). Spheroids may be used within 4 days after starting the experiment (Figure 2B). The quantification of each assay (i.e., proliferation, invasion, migration) is shown in Figure 3. Fiji macros were developed to quantify proliferation, invasion, or migration (Figure 3 and Supplementary Document 1).

The increase of spheroid core reflected the stimulation of cell proliferation (Figure 4A). Upon inhibition by rotenone, an established inhibitor of complex I of the mitochondrial respiratory chain, the vast majority of ATP production in the mitochondria was compromised. As a consequence, proliferation was reduced by 20% after 72 h (Figure 4A). Invasion of collagen type I was calculated by the subtraction of the total area from the core area. Acidic treatment enhanced invasion over a period of 24 h (Figure 4B). Furthermore, we found that hydrogen chloride treatment reduced the migratory area of the spheroids by 1.5 fold compared to control (Figure 4C). Spheroid dynamics was studied by by mounting for live imaging in the UniverSlide13. Spheroids had high internal dynamics and moved quickly (Movie 1 and tracking analysis in Figure 4D).

Figure 1
Figure 1: The spheroid is a relevant model to mimic solid tumors. (A) Spheroid as a round 3D structure with different areas. (a) A brightfield picture of a P3 spheroid shows a round appearance with a dense central area. Scale = 100 µm. (b) Schematic representation adapted from Hirschhaeuser et al.6 shows the O2, CO2, metabolite, and catabolite gradients in the spheroid. (c) Left panel: confocal picture of a spheroid stained with DAPI (blue) and with antibodies against carboxic anhydrase IX (green). Right panel: quantification of the fluorescence from the dashed area. Scale = 100 µm. (B) Electron microscopy images with delineated mitochondria (dashed lines). Large mitochondria are seen in the quiescentarea while they are smaller in the proliferation area. Scale = 250 nm. Please click here to view a larger version of this figure.

Figure 2
Figure 2: Overall spheroid preparation steps. (A) Generation of human P3 glioblastoma spheroids. Representative images are on the left panel and corresponding proliferation analysis on the right panel. At 24 h, the P3 cells formed dense spheroids. The initial number of cells determined the size of the spheroids. Scale bar = 250 µm. (B) Schematic illustration of the easy protocols for studying proliferation, invasion, or migration. Spheroids under various conditions: (a) In serum-free medium for tumor growth, (b) In collagen matrix for facilitating single cell invasion, and (c) On Matrigel coating for cell migration. Please click here to view a larger version of this figure.

Figure 3
Figure 3: Quantification of in vitro assays with Fiji software. Representation of the regions of interest (ROI) obtained using Fiji. The core area is represented in red and the total area, which contains the core area, in yellow. The invasive area corresponds to the subtraction of the total area from the core area. (A) Proliferation assay. (B) Invasion assay in collagen gel in brightfield acquisitions. (C) Migration assay on Matrigel coating by fluorescence acquisition. Nuclei were stained with DAPI, in blue. Please click here to view a larger version of this figure.

Figure 4
Figure 4: Glioblastoma P3 spheroid in proliferation, invasion, or migration assays. (A) Proliferation assay. Left panel: Representative pictures with DMSO as control or with 20 µM of rotenone (respiratory chain complex I inhibitor) at time 0 or 72 h. Right panel: Spheroid area quantification represented as dashed lines in the images. Scale = 250 µm. (B) Invasion assay in collagen matrix. Left panel: Representative pictures with or without 20 mM hydrogen chloride, at time 0 or 24 h. Right panel: Quantification of invasive areas. Scale = 100 µm. (C) Migration assay on Matrigel coating. Left panel: Representative images in brightfield mode at time 0 or 24 h. Magnified areas are represented in the bottom panels. Right panel: Quantification of migratory areas. Scale = 250 µm. (D) Z-stack representation of the spheroid (40 µm step) in (a) spheroid dynamic tracked over 18 h (image with 3 h interval) in (b). Cells stably expressed nuclear mCherry (orange) and cytoplasmic GFP (blue). Scale = 100 µm. Please click here to view a larger version of this figure.

Movie 1
Movie 1: P3 spheroid dynamic recorded over 18 h (imaged every 30 min). Scale bar = 100 µm. The movie represents a merged Z-stack over time with a Z-step of 5 µm for an approximate total volume of 150 µm. Cells were infected with lentivirus to express NLS:mCherry (nuclei are orange) and with cytoplasmic GFP (blue color). Please click here to view this file (Right click to download).

Supplementary Document 1: Fiji macro for analyzing invasion, proliferation, and migration of 3D spheroid. Please click here to view this file (Right click to download).

Subscription Required. Please recommend JoVE to your librarian.


Tumor spheroid assays are well adapted to study tumor characteristics including proliferation, invasion, and migration, as well as cell death and drug response. Cancer cells invade the 3D matrix forming an invasive microtumor, as seen in Figure 4B,C. During the invasive process, matrix metalloproteinases (MMP) digest matrices surrounding tumor cells13, and MMP inhibitors (e.g., GM6001 or Rebimastat) may impair cell invasion but not migration14. Migration and invasion involve overlapping but separate molecular events15, which can be studied in our spheroid assay. To do so, specific signaling pathways can be targeted either at the genetic level or through pharmacological inhibition.

Glioblastomas are known to extensively invade the surrounding tissues by different processes (e.g., co-option, white matter tract invasion, interstitial invasion)16. We recently described two novel mechanisms of glioblastoma invasion9,12,17. In particular, we studied matricellular thrombospondin-1 (TSP1) and showed that it is involved in tumor cell invasion through the activation of CD47 in tumor cells12. Furthermore, using a proteomic approach, we discovered the unexpected role of PLP1 and DNM1 in GBM invasion17. In these studies, 3D invasion assays were successfully used with or without pharmacological obstruction of TSP1, PLP1, or DNM1. Besides pharmacological obstruction, we also showed that acid treatment with hydrogen chloride impacts the invasion in the 3D assay. It is known that tumor acidosis activates a number of signaling pathways, including metabolic pathways (glycolysis), growth factors as TGFβ, and inhibits the immune response5.

Three-dimensional cultures provide a more physiological relevant environment than 2D cultures, and many molecular and metabolic parameters may be differently regulated. Thus, pharmacological modulation may have a different impact. Consequently, besides standard immunohistology, metabolic events can also be studied in 3D culture using probes such as 2-DG-IR. To corroborate these findings, the electron transport chain complex I inhibitor may also be used in this context.

Additionally, the 3D culture system is also well-suited to the study of dynamic processes using live imaging under basal conditions or in the presence of stimuli or pharmacological cues.

The following critical steps should be considered when carrying out the procedures described in this article: 1) The spheroid diameter should not exceed 400 µm in order to avoid necrosis; 2) The quantification of invasion using Fiji software must be carefully calibrated and performed as indicated in the detailed description of the procedure.; 3) The gel stiffness must be suitable to hold the spheroid in a stable configuration; 4) The pH value must be controlled, as a very acidic pH will impede the invasion process.

One limitation of the spheroid system described in this article is the lack of the complete tumor microenvironment. We acknowledge that the matrix used does not fully represent the stroma found in glioblastoma. However, collagens are part of the brain matrix and we wanted to develop a ready- and easy-to-use assay that can be used in any laboratory. Nevertheless, future experiments may also include additional matrix components as well as cellular elements, including stromal and immune cells. Another level of complexity is the inclusion of neuronal components, but these experiments must be carefully calibrated and designed.

In conclusion, we believe that our spheroid 3D system and the analytical tools we provide in this article may be useful for investigators, especially in studying brain tumor development.

Subscription Required. Please recommend JoVE to your librarian.


The authors declare that they have no competing financial interests.


This work was supported by Transcan 2017, ARC 2017, Ligue Contre le Cancer (Comité de la Gironde et de la Charente-Maritime). Joris Guyon is a recipient of fellowship from the Toulouse University Hospital (CHU Toulouse).


Name Company Catalog Number Comments
96 well round-bottom plate Falcon 08-772-212
Accutase Gibco A11105-01 Stored at 4 °C, sphere dissociation enzyme
B27 Gibco 12587 Stored at -20 °C, defrost before use
Basic Fibroblast Growth Factor Peprotech 100-18B Stored at -20 °C, defrost before use
Countess Cell Counting Chamber Slides Invitrogen C10283
DPBS 10X Pan Biotech P04-53-500 Stored at 4 °C
Fiji software ImageJ Used to analyze pictures
Flask 75 cm2 Falcon 10497302
Matrigel Corning 354230 Stored at -20 °C, diluted to a final concentration of 0.2 mg/mL in cold NBM
Methylcellulose Sigma M0512 Diluted in NBM for a 2% final concentration
NBM Gibco 21103-049 Stored at 4 °C
Neurobasal medium Gibco 21103049 Stored at 4 °C
Penicillin - Streptomycin Gibco 15140-122 Stored at 4 °C
Trypan blue 0.4% ThermoFisher T10282 Used to cell counting
Type I Collagen Corning 354236 Stored at 4 °C



  1. Pelissier, F. A., et al. Age-related dysfunction in mechanotransduction impairs differentiation of human mammary epithelial progenitors. Cell Reports. 7, 1926-1939 (2014).
  2. Ishiguro, T., et al. Tumor-derived spheroids: Relevance to cancer stem cells and clinical applications. Cancer Science. 108, 283-289 (2017).
  3. Sutherland, R. M. Cell and environment interactions in tumor microregions: the multicell spheroid model. Science. 240, 177-184 (1988).
  4. Desoize, B., Jardillier, J. Multicellular resistance: a paradigm for clinical resistance? Critical Reviews in Oncology/Hematology. 36, 193-207 (2000).
  5. Corbet, C., Feron, O. Tumour acidosis: from the passenger to the driver's seat. Nature Reviews Cancer. 17, 577-593 (2017).
  6. Hirschhaeuser, F., et al. Multicellular tumor spheroids: an underestimated tool is catching up again. Journal of Biotechnology. 148, 3-15 (2010).
  7. Berens, E. B., Holy, J. M., Riegel, A. T., Wellstein, A. A Cancer Cell Spheroid Assay to Assess Invasion in a 3D Setting. Journal of Visualized Experiments. (105), e53409 (2015).
  8. Cavaco, A. C. M., Eble, J. A. A 3D Spheroid Model as a More Physiological System for Cancer-Associated Fibroblasts Differentiation and Invasion In Vitro Studies. Journal of Visualized Experiments. (150), e60122 (2019).
  9. Boye, K., et al. The role of CXCR3/LRP1 cross-talk in the invasion of primary brain tumors. Nature Communications. 8, 1571 (2017).
  10. Dejeans, N., et al. Autocrine control of glioma cells adhesion and migration through IRE1alpha-mediated cleavage of SPARC mRNA. Journal of Cell Science. 125, 4278-4287 (2012).
  11. Golembieski, W. A., Ge, S., Nelson, K., Mikkelsen, T., Rempel, S. A. Increased SPARC expression promotes U87 glioblastoma invasion in vitro. International Journal of Developmental Neuroscience. 17, 463-472 (1999).
  12. Daubon, T., et al. Deciphering the complex role of thrombospondin-1 in glioblastoma development. Nature Communications. 10, 1146 (2019).
  13. Friedl, P., Wolf, K. Tube travel: the role of proteases in individual and collective cancer cell invasion. Cancer Research. 68, 7247-7249 (2008).
  14. Das, A., Monteiro, M., Barai, A., Kumar, S., Sen, S. MMP proteolytic activity regulates cancer invasiveness by modulating integrins. Scientific Reports. 7, 14219 (2017).
  15. Schaeffer, D., Somarelli, J. A., Hanna, G., Palmer, G. M., Garcia-Blanco, M. A. Cellular migration and invasion uncoupled: increased migration is not an inexorable consequence of epithelial-to-mesenchymal transition. Molecular and Cellular Biology. 34, 3486-3499 (2014).
  16. de Gooijer, M. C., Guillen Navarro, M., Bernards, R., Wurdinger, T., van Tellingen, O. An Experimenter's Guide to Glioblastoma Invasion Pathways. Trends in Molecular Medicine. 24, 763-780 (2018).
  17. Daubon, T., et al. The invasive proteome of glioblastoma revealed by laser-capture microdissection. Neuro-Oncology Advances. 1, (1), vdz029 (2019).
A 3D Spheroid Model for Glioblastoma
Play Video

Cite this Article

Guyon, J., Andrique, L., Pujol, N., Røsland, G. V., Recher, G., Bikfalvi, A., Daubon, T. A 3D Spheroid Model for Glioblastoma. J. Vis. Exp. (158), e60998, doi:10.3791/60998 (2020).More

Guyon, J., Andrique, L., Pujol, N., Røsland, G. V., Recher, G., Bikfalvi, A., Daubon, T. A 3D Spheroid Model for Glioblastoma. J. Vis. Exp. (158), e60998, doi:10.3791/60998 (2020).

Copy Citation Download Citation Reprints and Permissions
View Video

Get cutting-edge science videos from JoVE sent straight to your inbox every month.

Waiting X
simple hit counter