- Email: [email protected]
The Neuro-spheroid—A novel 3D in vitro model for peripheral nerve regeneration
Journal of Neuroscience Methods 246 (2015) 97–105
Contents lists available at ScienceDirect
Journal of Neuroscience Methods journal homepage: www.elsevier.com/locate/jneumeth
Basic Neuroscience
The Neuro-spheroid—A novel 3D in vitro model for peripheral nerve regeneration D Kraus a,∗ , V Boyle b , N Leibig c , GB Stark a , V Penna a a b c
Department of Plastic and Hand Surgery, University Medical Center Freiburg, Freiburg, Germany Clinic for Neurology, Ortenau Klinikum Lahr-Ettenheim, Lahr, Germany Department of Hand, Plastic and Reconstructive Surgery, Burn Center, BG Trauma Center Ludwigshafen, University of Heidelberg, Ludwigshafen, Germany
h i g h l i g h t s • • • •
A new 3D in vitro model with Schwann cell-neuron-spheroids in a collagen matrix. Spheroids grow interesting sprouting phenomena in this neuronal sprouting assay. Transferring a 2D into a 3D culture shows a significant increase of neurite length. It is a simple and flexible method to investigate neurite development in vitro.
a r t i c l e
i n f o
Article history: Received 21 December 2014 Received in revised form 27 January 2015 Accepted 3 March 2015 Available online 10 March 2015 Keywords: 3D in vitro model Sprouting assay Peripheral nerve regeneration Schwann cells NG108-15 cells
a b s t r a c t Background: In order to reduce in vivo animal experiments in peripheral nerve regeneration research, in vitro models are desirable. Common two dimensional (2D) co-culture models lack the complex interactions of three dimensional (3D) physiological structures. The aim of the study was to establish a neuronal 3D spheroidal sprouting assay for peripheral nerve regeneration. New method: Spheroids consisting of Schwann cells (SC, 500 cells/spheroid) and NG108-15 cells (NG, 50 cells/spheroid), a hybrid cell line, were formed in hanging drops and were embedded in a 3D collagen matrix. Spheroid sprout lengths were compared to those of the neurites of NG in a 2D co-culture with SC. Lengths were measured using phase contrast images taken every day over 10 days. Additionally we took fluorescence images to visualize the PKH26-labeled NG in both culture systems. Results: Initially thin neurites grew out in both co-cultures, over time the sprouts’ diameter in the 3D culture increased. The direct comparison of the sprout length revealed significantly longer neurites in the 3D co-culture from day 7 until day 10 (p < 0.001). Comparison with existing methods: Other co-culture models either display processes in 2D or need complex matrices to create 3D structures. Our spheroidal model is easy to establish, highly flexible and nevertheless 3D. Conclusions: The 3D-Schwann cell-neuron spheroid model shows that by simply transferring a 2D into a 3D co-culture with multiplication of cell-cell contacts, a significant increase of neurite length can be achieved. The model is a relatively simple method for the investigation of neurite development in vitro. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Bridging critical nerve gaps with nerve conduits is one possible treatment option to achieve adequate reinnervation of the target organ. To further improve the regeneration of peripheral nerve injuries bridged by conduits, defining and developing the ideal environment (i.e. conduit material, matrix, supporting cells,
∗ Corresponding author. Tel.: +49 761 27027790. E-mail address: [email protected] (D. Kraus). http://dx.doi.org/10.1016/j.jneumeth.2015.03.004 0165-0270/© 2015 Elsevier B.V. All rights reserved.
neurotrophic factors) is necessary. (Gonzalez-Perez et al., 2013, Klimaschewski et al., 2013) One approach is the use of supporting cells such as Schwann cells (SC), olfactory ensheathing cells or neurotrophins, e.g. nerve growth factor (NGF) and transforming growth factor (TGF-), that are known to positively influence nerve regeneration (Barras et al., 2002, Fine et al., 2002, Johnson et al., 2008). SC play an important role in nerve regeneration after a peripheral nerve injury. By providing a physical guide for the newly regenerating axons from the proximal stump and releasing neurotrophic factors (e.g. NGF and brain derived neurotrophic factor
98
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
Fig. 1. Possible effect on neurite length by transforming a 2D co-culture model into a 3D co-culture model, schematic representation. (A–C according to results of Armstrong et al., 2007).
(BDNF)) (Stoll and Muller, 1999). SC prepare the environment for the outgrowing neurites(Dodd and Jessell, 1988, Gundersen, 1980). In vitro, the two-dimensional (2D) co-culture of SC with neurons leads to increased outgrowth of neurites and improves neurite elongation. Armstrong et al., 2007 demonstrated, that this effect is not only achieved by secretion of neurotrophic factors by Schwann Cells, but also through the direct cell-cell-contact between SC and neurons of the NG108-15-cell line (NG). With the NG as neuroblastoma-glioma cell hybrids having the ability to differentiate toward a mature motor neuron phenotype, this cell line is frequently used to investigate neuronal characteristics and function (Armstrong et al., 2007, Klee and Nirenberg, 1974, Tojima et al., 2003). However, cells cultured in a two-dimensional setting represent an unphysiological environment. The lack of the third dimension, when growing cells on a plain surface, is not physiological and does not reflect the in vivo situation (Ahmed et al., 2006, Kofron et al., 2009). In vivo cells are organized in a complex three-dimensional (3D) network. This includes complex nutrient and signal interactions, a large number of cell-cell-contacts and interactions between cells and the extracellular matrix. All these given facts influence cell differentiation and organization (Sutherland, 1988, Dertinger and Hulser, 1984). Therefore, experiments performed in a 2D culture system can only be partially transferred to the actual in vivo situation. Thus, the conversion into a 3D culture is desirable. One approach to achieve this goal is to form spheroids. Spheroidal models are often used in basic research, e.g. in angiogenesis research (Finkenzeller et al., 2009, Korff, 1998, Wenger et al., 2005) cancer research (Ehsan et al., 2014, Lin et al., 2008, Kelm et al., 2003) or tissue engineering (Goerke et al., 2014, Haug et al., 2015, Lin et al., 2008). There are hardly any co-culture models with spheroids for peripheral nerve research. Kelm et al. (2006) described a spheroidal co-culture model of embryonic fibroblasts and ganglion cells to investigate gangliogenesis. To our knowledge there is no such spheroidal co-culture model of Schwann cells and neuronal cells published yet.
Different 3D culture systems have been numerously used to investigate the influence of different extracellular matrices on neurite formation/generation and neurite growth/development (Ahmed et al., 2006, Kofron et al., 2009, Baldwin et al., 1996, Bozkurt et al., 2007, Herbert et al., 1998, Pittier et al., 2005). One experimental set-up for three-dimensional cultivation, the spheroidal culture established 1998 by Korff and Augustin was primarily designed as an in vitro model for angiogenesis (Korff, 1998, Korff and Augustin, 1999, Finkenzeller et al., 2009, Strassburg et al., 2013, Steffens et al., 2009). This three-dimensional culture method provides gathering a lot of basic information about physiology, cell metabolism, tumor biology and toxicology as well as cell organization and development of artificial tissue (Kelm et al., 2003). It is also an established method for investigation of tumor growth, testing of pharmaceutical substances and investigation of myelination or demyelination processes in the central nervous system (Vereyken et al., 2009). Armstrong et al. (2007) could show that in NG monoculture (Fig. 1A) outgrowing neurites are not as long as in the indirect co-culture of NG and SC (Fig. 1B). Even longer neurites were seen in direct co-culture (Fig. 1C). If this effect is due to cell junctions between NG and SC, a 3D co-culture for example as in a spheroid is expected to have much longer neurites (Fig. 1D). The aim of the presented study was to investigate whether the transformation of a 2D co-culture of SC and NG into a spheroidal 3D co-culture with multiplication of cell junctions, leads to longer neurites. 2. Materials and methods 2.1. Schwann cell isolation, purification and cell culture SC were isolated and cultured as previously published (Penna et al., 2012). Neonatal rat pups (2–5 days postnatal) were anesthetized and then killed by decapitation. Sciatic nerves were exposed bilaterally and excised under sterile conditions. Enzymatic digestion was performed by incubating in Hank’s Balanced Salts
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
Solution (HBSS; Gibco BRL Life Technologies, Karlsruhe, Germany) containing 0.25% trypsin (PAA Laboratories, Coelbe, Germany) and 0.1% collagenase A (Roche Diagnostics, Penzberg, Germany) for 30 min at 37 ◦ C. The reaction was stopped by adding 5 ml culture medium (Dulbecco’s Modified Eagle Medium (DMEM), low glucose (1 g/l), with l-glutamine, (PAA Laboratories, Linz, Austria) containing 10% fetal calf serum (FCS; PAA, Linz, Austria) and 1% penicillin/streptomycin (PS; PAA Laboratories GmbH, Coelbe, Germany). The cells were counted using a cell counter and analyzer system (CASY-1® Cell counter and analyzer system, Schärfe System GmbH, Reutlingen, Germany). The suspension was centrifuged (Megafuge 3.0R, Heraeus Instruments GmbH, Hanau, Germany) for 5 min at 1000 rpm and 4 ◦ C. Higher Schwann cell purification was achieved via negative cell isolation by magnetic elimination of fibroblasts after labeling with magnetic particles (Anti-Mouse CD90.2 Magnetic Particles, BD Biosciences Europe, UK) (Assouline et al., 1983). The pellet was resuspended in Becton Dickinson (BD) IMagTM -Buffer (1 − 8 × 107 cells per ml) and transferred into an Eppendorf tube. After brief dispersion of the CD90.2-antibody, 50 l per 107 cells were added and an incubation time of 30 min at 4 ◦ C followed. After that the Eppendorf tube was placed in a magnet field of a magnetic stand (Dynal MPC-S, Magnetic Particle Concentrator, Dynal Biotech ASA, Oslo, Norway) for 6–8 min at room temperature. The supernatant containing the untouched SC was transferred into a new Eppendorf tube with the magnetically labeled fibroblasts being hold back in the magnetic field. This supernatant was further purified by incubating it with the CD90.2-antibody again. The suspension containing the SC after the second purification step was centrifuged for 5 min at 1000 rpm and 4 ◦ C, the supernatant was removed, taking care not to aspirate the cell containing pellet and the pellet was resuspended in 10 ml culture medium. The cells were seeded on a Poly-d-lysine vented flask (Biocoat 75 cm2 , BD Biosciences Europe) and cultured in a cell incubator (37 ◦ C, 5% carbon dioxide (CO2 ); Forma Scientific CO2 water jacketed incubator, model 3111, ThermoScientific, Waltham, USA). After 24 h 1 M cytosine arabinoside (Sigma Aldrich, Deisenhofen, Germany) was added for 48 h to reduce the remaining fibroblasts. After removing the cytosine arabinoside, the cells were maintained in culture medium with 5 M forskolin (Calbiochem, La Jolla, California, USA) for proliferation stimulation. The medium was changed every 72 h. In order to verify and quantify the purity grade, the SC were stained with calcium binding protein S100 primary antibody (rabbit anti-mouse, DAKO, Denmark; 1: 200 in phosphate buffered saline (PBS; Biochrom AG, Berlin, Germany) containing 1% bovine serum albumin (BSA; Sigma Aldrich, Deisenhofen, Germany) and 0.24% Triton X (Sigma Aldrich, Deisenhofen, Germany)) and rhodamine conjugated secondary antibody (goat anti-rabbit IgG-R, mouse/human adsorbed, Santa Cruz Biotechnology, USA; 1:200 in PBS containing 1% BSA and 0,24% Triton X). Sixty thousand cells per chamber were cultivated in a chamber slide® (NalgeNunc International, Rochester, USA). The cells were fixed using 500 l formaldehyde (4%; FlukaChemie, Buchs, Switzerland). After washing 3× with PBS, non-specific binding surfaces were blocked by incubation with PBS containing 1% BSA and 0.24% Triton X for 10 min at room temperature. The fixed cells were incubated with the primary antibody (200 l/chamber) for 30 min at room temperature, washed 4× 5 min with PBS, incubated with the secondary antibody (200 l/chamber) for 30 min at room temperature and washed 4× 5 min with PBS. The slides were observed under an Axioplan fluorescence microscope (Carl Zeiss Jena, Jena, Germany), photographed using a digital camera (AxioCam, Zeiss, Germany) and analyzed with Axiovision 2.0.5 Software (Carl Zeiss Vision, Munich, Germany).
99
2.2. NG108-15 cell culture The neuronal cell culture of NG, a hybrid cell line from mouse neuroblastoma- and rat gliomacells, was purchased from ECACC (European Collection of Cell Cultures, Salisbury, UK). The NG were seeded in a Costar® 75 cm2 polystyrene flask (Corning Incorporate, USA) and cultured in culture medium. The medium was replaced every 48 h. For verification NG and SC were stained with antineurofilament 200 (Monoclonal Anti-Neurofilament 200, Mouse Ascites Fluid, N 0142, Sigma Aldrich, Deisenhofen, Germany). SC were used to proof the specificity of the first antibody for NG. As secondary antibody a HRP-conjugated anti-mouse antibody was used (goat anti-mouse IgG, HRP-conjugated, EnvisionTM + System HRP Labeled Polymer. Anti-mouse, DAKO, Glastrup, Denmark). In order to identify the NG in co-culture, they were labeled with the PKH-26 red fluorescent cell linker kit (Sigma Aldrich, Deisendorf, Germany). The fluorescent particles were integrated into the cell membrane of the cells. Labeling was performed following the kit protocol using a final PKH concentration of 6.7 × 10-6 M over a 4 min incubation time. Successful labeling was confirmed microscopically using ultraviolet light (wavelength 567 nm, Axiovert 100, Zeiss, Jena, Germany). 2.3. Methocel stock solution Six grams of autoclaved carboxymethyl cellulose (Sigma, Deisenhofen, Germany) were dissolved in 250 ml prewarmed (60 ◦ C) endothelial cell basal medium (ECBM, EBM® -2, Lonza Inc., USA) by mixing with a magnetic stirrer for 20 min. After reaching room temperature again, another 250 ml ECBM with 20% FCS were added. The solution was mixed for 1 h and aliquoted into 50 ml falcons. The undissolved carboxymethyl cellulose was removed by centrifugation (2.000×g, for 4 h at room temperature). Only the clear, gel-like supernatant was used for the experiments. 2.4. Collagen type I extraction and collagen gel Collagen type I was extracted by incubating ligaments of two wistar rat-tails in 250 ml 0.1% acetic acid (Merck, Darmstadt, Germany) for 48 h at 4 ◦ C. The formed highly viscous solution was centrifuged at 24000×g and 4 ◦ C. The clear supernatant was removed to get the collagen stock solution, which was adjusted to an optical density of 0.34–0.36 by using 0.1% acetic acid. Prior to use, 290 l of collagen mix (88.8% collagen stock solution, 9.7% medium 199 10× (Sigma, Germany) and 1.5% hepes buffer 1 M (Gibco, life technologies Inc., UK) neutralized to pH 7.4 with 2 N NaOH (Merck, Germany)) and 290 l of methocel stock solution with 20% FCS were mixed to get about 500 l of clear collagen gel. 2.5. 2D Co-Culture NG were detached by gently shaking the flask, SC by using trypsin (Gibco, life technologies Inc., UK). Both were counted with the CASY-1® cell counter and mixed 1:10 (NG:SC). Cells were then seeded into a 24 well flask (Corning, USA) (66,000 cells/well). To approximate the culture conditions to the 3D experimental setup, a collagen gel layer of 250 l was added on top of the 2D co-culture after cell attachment to the flask. After polymerization of the collagen gel 250 l of culture medium were added on top (Fig. 2). The medium was changed every 48 h and the culture was kept in a cell incubator (37 ◦ C, 5% CO2 ). The 2D coculture was observed microscopically and length was measured daily for 10 days.
100
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
micro liter of the gel containing approximately 50 spheroids were rapidly transferred into a well of a pre warmed 24-well suspension plate and incubated at 37 ◦ C and 5% CO2 . After polymerization (30 min), 250 l culture medium was added on top. The medium was changed every 48 h. The 3D co-culture was observed and photographed over a period of 10 up to 20 days. 2.7. Image analysis
Fig. 2. 2D co-culture, model of experimental setup.
2.6. 3D co-culture 2.6.1. Spheroid formation A spheroidal culture assay was used in a modification of an angiogenesis assay as previously described(Korff, 1998, Finkenzeller et al., 2009). In brief, NG and SC were mixed 1:10 (NG: SC, 550 cells/spheroid), centrifuged and resuspended in culture medium with 20% methocel stock solution. 25 l portions of this suspension were seeded on plastic dishes (Greiner Bio-One, Hungary) as hanging drops (Fig. 3) to allow overnight spheroid aggregation in cell incubator (37 ◦ C, 5% CO2 ). 2.6.2. Sprouting assay After overnight incubation the spheroids were harvested by rinsing with PBS and centrifuged at 1100 rpm for 5 min (4 ◦ C). The pellet containing the spheroids was resuspended in 290 l methocel stock solution containing 20% FCS. 290 l collagen mix was added on top and the solution was mixed gently but thoroughly to achieve a good distribution of the spheroids in the gel. Five hundred
The 2D co-culture as well as the 3D co-culture were observed microscopically (Axiovert 100, Zeiss, Germany) in phase contrast and ultraviolet light and documented photographically (AxioCam, Zeiss, Germany). UV-light was used for identification of PKH-labeled NG. Phase contrast photographs were used for length measurement (AxioVision 2.0.5, Carl Zeiss Vision, Munich, Germany). Except for contrast and brightness there was no modification made subsequently. The length of all sprouts with positive PKH-staining of daily randomly chosen 30 NG of the 2D co-culture and of 15 spheroids were measured in 2D on a phase contrast photograph. 2.8. Statistical analysis The mean length and standard error of mean were calculated (Sigma STATTM 3, Systat Software, San José, California, USA). Statistical significance was calculated by t-test and Mann–Whitney Rank-Test. P-values of < 0.05 were regarded as statistically significant. 3. Results 3.1. Schwann cell culture The cultured SC expressed the S100 Antigen, as shown by positive S100 immunofluorescence (Fig. 4). Using magnetic elimination of fibroblasts, the SC cultures presented a high purity level (about 99% after the two purification steps (Assouline et al., 1983)). In cell culture SC showed a multipolar or fusiform morphology and could be effectively expanded for 12–15 weeks without losing their ability to proliferate. 3.2. NG108-15 culture The NG were highly proliferative in culture and the cells were of polymorphic, multi polar morphology. NG could be stained with neurofilament 200, the stain was located especially in the neurites. When trying to stain SC the same way, there was no positive signal, which demonstrated that Neurofilament 200 could be considered as specific marker for NG in the co-culture. In order to distinguish the NG from SC in the 2D- and 3D coculture, NG were successfully labeled with red fluorescent cell linker PKH-26 (Fig. 5). In culture and after cell separation, the cells showed a distinctive emission in ultraviolet light. No intercellular transfer or into the surrounding collagen matrix could be seen. 3.3. Co-cultures
Fig. 3. 3D co-culture, model of experimental setup. Spheroid formation in hanging drops and transfer in 3D collagen matrix.
2D co-culture with collagen layer addition could be successfully performed. There was no migration of the cells into the collagen matrix and the cells were adherent to the flask. The cultivation could be maintained easily during the 10 days of observation. After this period, evaluation became difficult because of an increase of cell density, so the 10th day was defined as cut off point. The spheroid formation (3D co-culture) in hanging drops generated round three-dimensional cell cultures, homogeneous in size and shape. Embedding the spheroids into collagen matrix with 50
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
101
Fig. 4. Characterisation of SC. Comparison of contrast image with corresponding fluorescence image of SC positive stained for S100. Scale bar 100 m.
Fig. 5. NG cell culture after labeling with PKH26. Comparison of phase contrast image with correspronding fluorescent image. Scale bar 50 m.
spheroids per 500 l gel, resulted in an even distribution in the gel. NG showed neurite outgrowth in 2D and in 3D co-culture. In 2D co-culture, the neurites formed by the NG were of filigree constitution, little elongation was found during the 10 days and there was no change in morphology (Fig. 6).
In the first four days of 3D co-culture most of the neurites had a similar filigree appearance. But then the caliber of the sprouts enlarged from the base of the spheroid and the sprouts grew (Figs. 7 and 8). Additionally to the excessive neurite elongation, there was also a growth of the entire spheroid. The spheroid
Fig. 6. 2D co-culture of NG with SC. Comparison of phase contrast image (A, C) and fluorescent image (B, D). (A) and (B) day 5, scale bar 100 m. (C) and (D) day 10, scale bar 50 m. Red stain = PKH26-labeled NG. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
102
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
Fig. 7. Development of a NG-SC cospheroid from day 3 till day 8. Comparison of phase contrast image (A, C, E, G, I, K) and fluorescent image (B, D, F, H, J, L). (A) and (B) day 3, (C) and (D) day 4, (E) and (F) day 5, (G) and (H) day 6, (I) and (J) day 7, (K) and (L) day 8, Red stain = PKH26-labeled NG, scale bar 100 m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
diameters doubled during the 10 days of observation (day 1 149.94 m ± 10.66 m, day 10 339, 16 m ± 22.5 m; p < 0.001, Fig. 9). For length comparison, the neurites of 30 NG randomly picked in 2D culture, and the neurites of 15 randomly chosen spheroids in 3D culture were measured. The direct comparison of the sprout length of both co-cultures revealed significantly higher neurite lengths in the 3D co-culture from day 7 until day 10 (p < 0.001, Fig. 10). While the neurite length in 2D culture stabilized around 65 m, the sprout length of the NG-SC-spheroids elongated from day 7 to day 10 from 115, 56 m (±8,1) to 176, 6 m (±14.26). This resulted in 2.3× longer 3D neurites than 2D neurites (mean day 7 to day 10; p < 0.001; Fig. 10).
4. Discussion Once developed for cancer research (Sutherland, 1988), spheroidal cell cultures are a common method in tissue engineering to approximate in vitro cultures to in vivo conditions (Lin et al., 2008). A lot of approaches to establish reproducible spheroidal models for different cells and purposes have been developed (Lin et al., 2008). The one by Korff and Augustin (Korff, 1998) is a highly reproducible but also modifiable spheroidal model. In combination with embedding the created spheroids in a three-dimensional matrix, it delivers an easy to establish co-culture model in a 3D environment, which can be adjusted to requirements related to the purpose. For instance different types of cells can be used for spheroid formation and spheroids can be embedded in different matrices. This so called sprouting assay is an established model
in angiogenesis research (Korff, 1998, Korff and Augustin, 1999, Finkenzeller et al., 2009, Strassburg et al., 2013, Steffens et al., 2009). Armstrong et al., 2007 showed that, in addition to soluble factors secreted by SC, the direct cell-cell-contact between SC and neuronal cells in a 2D culture does have a positive influence on neurite length. Knowing these results and the spheroidal sprouting assay established in angiogenesis research, we wanted to prove, if transforming the 2D co-culture model of SC and neuronal cells, which is a well established model in the in vitro research of nerve regeneration (Callizot et al., 2011, Sango et al., 2012), to a 3D spheroidal sprouting assay delivers a better model for nerve regeneration research. To our knowledge this is the first description of such an approach in literature. Other 3D nerve regeneration models either use monocultures of e.g. dorsal root ganglia cells (Bozkurt et al., 2007, Kofron et al., 2009) or PC12 cells (Pittier et al., 2005) or need more complex matrices to ensure that cells can interact in co-culture. Such matrices can be complex 3D fiber scaffolds (Daud et al., 2012) or collagen sponges (Gingras et al., 2008). Even more complex models embed whole tissue pieces like e.g. whole dorsal root ganglia (Allodi et al., 2011). Being a three dimensional model, the spheroidal cultivation system with increased cell-cell-contacts and three dimensional surroundings/environment is more similar to the actual in vivo situation than a two dimensional co-culture (Lu et al., 2012). Nevertheless compared to the complexity of the in vivo constitution, with the influence and interaction of different types of cells, cytokines and the immune system, the spheroidal model is abstract and highly simplified. This again, though offers the possibility to investigate different factors that influence nerve regeneration, independently and fundamentally.
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
103
Fig. 8. Change of neurite morphology within 24 h, same NG-SC co-spheroid with sprout on day 9 (A, C, E) and 10 (B, D, F), phase contrast image (A, B, C, D) and fluorescent image (E, F), Red stain = PKH26-labeled NG, scale bar 100 m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
Fig. 9. Diagram mean diameter of NG-SC co-spheroids per day. Demonstration of the spheroid growth during a cultivation period of 10 days. Significant increase in size from day 1 to day 10 (p < 0.001).
104
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
Fig. 10. Diagram comparing the mean neurite length of NG in 2D and 3D co-culture with SC per day. From day 7 significant higher neurite length in 3D experimental setup. * : p < 0.05; ** : p < 0.001.
The analysis of neurite lengths showed a significant difference between the neurites formed in 2D culture and 3D culture from day 7 on. The length of the neurites in 3D culture was approximately 2.3 (±0.3)-times greater than in 2D culture. This result emphasizes the positive influence SC have on neurite outgrowth through direct cell-cell contact. The SC and NG ratio was identical in both, the 2D and 3D experimental setup. The difference was therefore achieved by the three-dimensional setup with increase of cell-cell-contacts between both cell types. Other possibilities to increase neurite length in addition to increased cell-cell contact are differences in cultivation conditions and positive influence of the extracellular matrix on neurite growth. In order to minimize the interference of different cultivation conditions, a collagen layer was put on the 2D cell culture. We used a collagen-I-matrix, which is a well-established hydrogel for neuron cultivation (Kofron et al., 2009). This component of the physiological extracellular matrix is also employed as a material in experimental nerve conduits (Dornseifer et al., 2007, Bozkurt et al., 2007). With the collagen layer cultivation, conditions of both groups were adjusted to be more comparable. An increase of neurite length due to better cultivation conditions in 3D co-culture is therefore unlikely. Furthermore, in studies by Kofron et al., 2009 collagen did not lead to an increase of neurite length or lead of direction of neurite outgrowth. This finding can be supported in this study, where there was no neurite growth into the collagen layer in the 2D cell culture noticed. Thus it is reasonable to assume that the actual increase of cellcell contacts between SC and NG leads to the achieved length difference between 3D and 2D cell culture. In addition, due to the length measurement technique, the neurites in 3D co-culture system, which do not exactly grow in horizontal planes, were sized too short. The measurement was performed in horizontal planes, in 2D photographs, because this is a fast method that does not require a complex computer program. When measuring the neurites in 2D culture model, there was no deviation, because the neurites grew planar on the culture flask. However, in the 3D culture system the neurites grow into all possible directions. Due to the measurement in the horizontal plane, neurites in 3D set up, which grew in other planes, result in a shorter
Fig. 11. Measurement of neurites in three-dimensional set-up, schematic representation. Deviation of neurite growth from horizontal plane leads to smaller result of neurite length in 3D culture.
length (Fig. 11). The degree of this failure varies, depending on the plane deviating angle. The lengths of neurites measured in 3D culture setup are therefore longer than measured and consequently the difference between the evaluated lengths in 3D and 2D culture are greater than represented in the results. 5. Conclusion Our experiments show that by simply transferring a 2D into a 3D culture with a multiplication of cell-cell contacts, a significant increase of neurite length can be achieved. This result emphasizes the positive influence SC have on neurite elongation through direct cell-cell contact (Lopez-Verrilli and Court, 2012). As shown in this study, the cultivation of NG in spheroidal culture with subsequent sprouting assay in collagen gel, is a relatively simple method to investigate of neurite development with an interesting neuronal sprouting phenomena. Supplementary investigations, e.g. performing cryosections and immunostaining of NG-SC-spheroids in sprouting assay, are projected. Additionally, the experimental setup is flexible and therefore offers the individual investigation of the influence of different factors on nerve regeneration in a “closeto-reality” setting.
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
References Ahmed I, Liu H-Y, Mamiya PC, Ponery AS, Babu AN, Weik T, et al. Three-dimensional nanofibrillar surfaces covalently modified with tenascin-C-derived peptides enhance neuronal growth in vitro. J Biomed Mater Res 2006;A76:851–60. Allodi I, Guzmán-Lenis M-S, Hernàndez J, Navarro X, Udina E. In vitro comparison of motor and sensory neuron outgrowth in a 3D collagen matrix. J Neurosci Methods 2011;198:53–61. Armstrong SJ, Wiberg M, Terenghi G, Kingham PJ. ECM molecules mediate both Schwann cell proliferation and activation to enhance neurite outgrowth. Tissue Eng 2007;13:2863–70. Assouline JG, Bosch EP, Lim R. Purification of rat Schwann cells from cultures of peripheral nerve: an immunoselective method using surfaces coated with antiimmunoglobulin antibodies. Brain Res 1983;277:389–92. Baldwin SP, Krewson CE, Saltzman WM. PC12 cell aggregation and neurite growth in gels of collagen, laminin and fibronectin. Int J Dev Neurosci Off J Int Soc Dev Neurosci 1996;14:351–64. Barras FM, Pasche P, Bouche N, Aebischer P, Zurn AD. Glial cell line-derived neurotrophic factor released by synthetic guidance channels promotes facial nerve regeneration in the rat. J Neurosci Res 2002;70:746–55. Bozkurt A, Brook GA, Moellers S, Lassner F, Sellhaus B, Weis J, et al. In vitro assessment of axonal growth using dorsal root ganglia explants in a novel three-dimensional collagen matrix. Tissue Eng 2007;13:2971–9. Callizot N, Combes M, Steinschneider R, Poindron P. A new long term in vitro model of myelination. Exp Cell Res 2011;317:2374–83. Daud MFB, Pawar KC, Claeyssens F, Ryan AJ, Haycock JW. An aligned 3D neuronal-glial co-culture model for peripheral nerve studies. Biomaterials 2012;33:5901–13. Dertinger H, Hulser DF. Intercellular communication in spheroids Recent Results. Cancer Res 1984;95:67–83. Dodd J, Jessell TM. Axon guidance and the patterning of neuronal projections in vertebrates. Science 1988;242:692–9. Dornseifer U, Matiasek K, Fichter M, Rupp A, Henke J, Weidner N, et al. Surgical therapy of peripheral nerve lesions: current status and new perspectives. Zentralblatt Für Neurochir—Cent Eur Neurosurg 2007;68:101–10. Ehsan SM, Welch-Reardon KM, Waterman ML, Hughes CCW, George SC. A three-dimensional in vitro model of tumor cell intravasation. Integr Biol 2014;6:603–10. Fine EG, Decosterd I, Papalozos M, Zurn AD, Aebischer P. GDNF and NGF released by synthetic guidance channels support sciatic nerve regeneration across a long gap. Eur J Neurosci 2002;15:589–601. Finkenzeller G, Graner S, Kirkpatrick CJ, Fuchs S, Stark GB. Impaired in vivo vasculogenic potential of endothelial progenitor cells in comparison to human umbilical vein endothelial cells in a spheroid-based implantation model. Cell Prolif 2009;42:498–505. Gingras M, Beaulieu M-M, Gagnon V, Durham HD, Berthod F. In vitro study of axonal migration and myelination of motor neurons in a three-dimensional tissueengineered model. Glia 2008;56:354–64. Goerke SM, Obermeyer J, Plaha J, Stark GB, Finkenzeller G. Endothelial progenitor cells from peripheral blood support bone regeneration by provoking an angiogenic response. Microvasc Res 2014;98C:40–7. Gonzalez-Perez F, Udina E, Navarro X. Stefano Geuna IP, editor. Extracellular Matrix Components in Peripheral Nerve Regeneration, Int Rev Neurobiol, 108; 2013. p. 257–75 http://www.sciencedirect.com/science/article/pii/ B9780124104990000101 Gundersen RWBJN. Characterization of the turning response of dorsal root neurites toward nerve growth factor. J Cell Biol 1980;87:546–54. Haug V, Torio-Padron N, Stark GB, Finkenzeller G, Strassburg S. Comparison between endothelial progenitor cells and human umbilical vein endothelial cells on neovascularization in an adipogenesis mouse model. Microvasc Res 2015;97:159–66.
105
Herbert CB, Nagaswami C, Bittner GD, Hubbell JA, Weisel JW. Effects of fibrin micromorphology on neurite growth from dorsal root ganglia cultured in threedimensional fibrin gels. J Biomed Mater Res 1998;40:551–9. Johnson EO, Charchanti A, Soucacos PN. Nerve repair: experimental and clinical evaluation of neurotrophic factors in peripheral nerve regeneration. Injury 2008;39:37–42. Kelm JM, Ittner LM, Born W, Djonov V, Fussenegger M. Self-assembly of sensory neurons into ganglia-like microtissues. J Biotechnol 2006;121:86–101. Kelm JM, Timmins NE, Brown CJ, Fussenegger M, Nielsen LK. Method for generation of homogeneous multicellular tumor spheroids applicable to a wide variety of cell types. Biotechnol Bioeng 2003;83:173–80. Klee WA, Nirenberg M. A neuroblastoma × glioma hybrid cell line with morphine receptors. Proc Natl Acad Sci U S A 1974;71:3474–7. Klimaschewski L, Hausott B, Angelov DN. Stefano Geuna IP, editor. The Pros and Cons of Growth Factors and Cytokines in Peripheral Axon Regeneration, Int Rev Neurobiol, 108; 2013. p. 137–71 http://www.sciencedirect.com/science/article/ pii/B978012410499000006X Kofron CM, Fong VJ, Hoffman-Kim D. Neurite outgrowth at the interface of 2D and 3D growth environments. J Neural Eng 2009;6:016002. Korff T. Integration of endothelial cells in multicellular spheroids prevents apoptosis and induces differentiation. J Cell Biol 1998;143:1341–52. Korff T, Augustin HG. Tensional forces in fibrillar extracellular matrices control directional capillary sprouting. J Cell Sci 1999;112(Pt 19):3249–58. Lin R-Z, Lin R-Z, Chang H-Y. Recent advances in three-dimensional multicellular spheroid culture for biomedical research. Biotechnol J 2008;3:1172–84. Lopez-Verrilli MA, Court FA. Transfer of vesicles from Schwann cells to axons: a novel mechanism of communication in the peripheral nervous system. Front Physiol 2012:3 http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3374349/ Lu H, Searle K, Liu Y, Parker T. The effect of dimensionality on growth and differentiation of neural progenitors from different regions of fetal rat brain in vitro: 3-dimensional spheroid versus 2-dimensional monolayer culture. Cells Tissues Organs 2012;196:48–55. Penna V, Stark G-B, Wewetzer K, Radtke C, Lang EM. Comparison of Schwann cells and olfactory ensheathing cells for peripheral nerve gap bridging. Cells Tissues Organs 2012;196:534–42. Pittier R, Sauthier F, Hubbell JA, Hall H. Neurite extension and in vitro myelination within three-dimensional modified fibrin matrices. J Neurobiol 2005;63:1–14. Sango K, Kawakami E, Yanagisawa H, Takaku S, Tsukamoto M, Utsunomiya K, et al. Myelination in co-culture of established neuronal and Schwann cell lines. Histochem Cell Biol 2012;137:829–39. Steffens L, Wenger A, Stark GB, Finkenzeller G. In vivo engineering of a human vasculature for bone tissue engineering applications. J Cell Mol Med 2009;13:3380–6. Stoll G, Muller HW. Nerve injury, axonal degeneration and neural regeneration: basic insights. Brain Pathol Zurich Switz 1999;9:313–25. Strassburg S, Nienhueser H, Stark GB, Finkenzeller G, Torio-Padron N. Human adipose-derived stem cells enhance the angiogenic potential of endothelial progenitor cells, but not of human umbilical vein endothelial cells. Tissue Eng Part A 2013;19:166–74. Sutherland RM. Cell and environment interactions in tumor microregions: the multicell spheroid model. Science 1988;240:177–84. Tojima T, Kobayashi S, Ito E. Dual role of cyclic AMP-dependent protein kinase in neuritogenesis and synaptogenesis during neuronal differentiation. J Neurosci Res 2003;74:829–37. Vereyken EJF, Fluitsma DM, Bolijn MJ, Dijkstra CD, Teunissen CE. An in vitro model for de- and remyelination using lysophosphatidyl choline in rodent whole brain spheroid cultures. Glia 2009;57:1326–40. Wenger A, Kowalewski N, Stahl A, Mehlhorn AT, Schmal H, Stark GB, et al. Development and characterization of a spheroidal coculture model of endothelial cells and fibroblasts for improving angiogenesis in tissue engineering. Cells Tissues Organs 2005;181:80–8.
Contents lists available at ScienceDirect
Journal of Neuroscience Methods journal homepage: www.elsevier.com/locate/jneumeth
Basic Neuroscience
The Neuro-spheroid—A novel 3D in vitro model for peripheral nerve regeneration D Kraus a,∗ , V Boyle b , N Leibig c , GB Stark a , V Penna a a b c
Department of Plastic and Hand Surgery, University Medical Center Freiburg, Freiburg, Germany Clinic for Neurology, Ortenau Klinikum Lahr-Ettenheim, Lahr, Germany Department of Hand, Plastic and Reconstructive Surgery, Burn Center, BG Trauma Center Ludwigshafen, University of Heidelberg, Ludwigshafen, Germany
h i g h l i g h t s • • • •
A new 3D in vitro model with Schwann cell-neuron-spheroids in a collagen matrix. Spheroids grow interesting sprouting phenomena in this neuronal sprouting assay. Transferring a 2D into a 3D culture shows a significant increase of neurite length. It is a simple and flexible method to investigate neurite development in vitro.
a r t i c l e
i n f o
Article history: Received 21 December 2014 Received in revised form 27 January 2015 Accepted 3 March 2015 Available online 10 March 2015 Keywords: 3D in vitro model Sprouting assay Peripheral nerve regeneration Schwann cells NG108-15 cells
a b s t r a c t Background: In order to reduce in vivo animal experiments in peripheral nerve regeneration research, in vitro models are desirable. Common two dimensional (2D) co-culture models lack the complex interactions of three dimensional (3D) physiological structures. The aim of the study was to establish a neuronal 3D spheroidal sprouting assay for peripheral nerve regeneration. New method: Spheroids consisting of Schwann cells (SC, 500 cells/spheroid) and NG108-15 cells (NG, 50 cells/spheroid), a hybrid cell line, were formed in hanging drops and were embedded in a 3D collagen matrix. Spheroid sprout lengths were compared to those of the neurites of NG in a 2D co-culture with SC. Lengths were measured using phase contrast images taken every day over 10 days. Additionally we took fluorescence images to visualize the PKH26-labeled NG in both culture systems. Results: Initially thin neurites grew out in both co-cultures, over time the sprouts’ diameter in the 3D culture increased. The direct comparison of the sprout length revealed significantly longer neurites in the 3D co-culture from day 7 until day 10 (p < 0.001). Comparison with existing methods: Other co-culture models either display processes in 2D or need complex matrices to create 3D structures. Our spheroidal model is easy to establish, highly flexible and nevertheless 3D. Conclusions: The 3D-Schwann cell-neuron spheroid model shows that by simply transferring a 2D into a 3D co-culture with multiplication of cell-cell contacts, a significant increase of neurite length can be achieved. The model is a relatively simple method for the investigation of neurite development in vitro. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Bridging critical nerve gaps with nerve conduits is one possible treatment option to achieve adequate reinnervation of the target organ. To further improve the regeneration of peripheral nerve injuries bridged by conduits, defining and developing the ideal environment (i.e. conduit material, matrix, supporting cells,
∗ Corresponding author. Tel.: +49 761 27027790. E-mail address: [email protected] (D. Kraus). http://dx.doi.org/10.1016/j.jneumeth.2015.03.004 0165-0270/© 2015 Elsevier B.V. All rights reserved.
neurotrophic factors) is necessary. (Gonzalez-Perez et al., 2013, Klimaschewski et al., 2013) One approach is the use of supporting cells such as Schwann cells (SC), olfactory ensheathing cells or neurotrophins, e.g. nerve growth factor (NGF) and transforming growth factor (TGF-), that are known to positively influence nerve regeneration (Barras et al., 2002, Fine et al., 2002, Johnson et al., 2008). SC play an important role in nerve regeneration after a peripheral nerve injury. By providing a physical guide for the newly regenerating axons from the proximal stump and releasing neurotrophic factors (e.g. NGF and brain derived neurotrophic factor
98
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
Fig. 1. Possible effect on neurite length by transforming a 2D co-culture model into a 3D co-culture model, schematic representation. (A–C according to results of Armstrong et al., 2007).
(BDNF)) (Stoll and Muller, 1999). SC prepare the environment for the outgrowing neurites(Dodd and Jessell, 1988, Gundersen, 1980). In vitro, the two-dimensional (2D) co-culture of SC with neurons leads to increased outgrowth of neurites and improves neurite elongation. Armstrong et al., 2007 demonstrated, that this effect is not only achieved by secretion of neurotrophic factors by Schwann Cells, but also through the direct cell-cell-contact between SC and neurons of the NG108-15-cell line (NG). With the NG as neuroblastoma-glioma cell hybrids having the ability to differentiate toward a mature motor neuron phenotype, this cell line is frequently used to investigate neuronal characteristics and function (Armstrong et al., 2007, Klee and Nirenberg, 1974, Tojima et al., 2003). However, cells cultured in a two-dimensional setting represent an unphysiological environment. The lack of the third dimension, when growing cells on a plain surface, is not physiological and does not reflect the in vivo situation (Ahmed et al., 2006, Kofron et al., 2009). In vivo cells are organized in a complex three-dimensional (3D) network. This includes complex nutrient and signal interactions, a large number of cell-cell-contacts and interactions between cells and the extracellular matrix. All these given facts influence cell differentiation and organization (Sutherland, 1988, Dertinger and Hulser, 1984). Therefore, experiments performed in a 2D culture system can only be partially transferred to the actual in vivo situation. Thus, the conversion into a 3D culture is desirable. One approach to achieve this goal is to form spheroids. Spheroidal models are often used in basic research, e.g. in angiogenesis research (Finkenzeller et al., 2009, Korff, 1998, Wenger et al., 2005) cancer research (Ehsan et al., 2014, Lin et al., 2008, Kelm et al., 2003) or tissue engineering (Goerke et al., 2014, Haug et al., 2015, Lin et al., 2008). There are hardly any co-culture models with spheroids for peripheral nerve research. Kelm et al. (2006) described a spheroidal co-culture model of embryonic fibroblasts and ganglion cells to investigate gangliogenesis. To our knowledge there is no such spheroidal co-culture model of Schwann cells and neuronal cells published yet.
Different 3D culture systems have been numerously used to investigate the influence of different extracellular matrices on neurite formation/generation and neurite growth/development (Ahmed et al., 2006, Kofron et al., 2009, Baldwin et al., 1996, Bozkurt et al., 2007, Herbert et al., 1998, Pittier et al., 2005). One experimental set-up for three-dimensional cultivation, the spheroidal culture established 1998 by Korff and Augustin was primarily designed as an in vitro model for angiogenesis (Korff, 1998, Korff and Augustin, 1999, Finkenzeller et al., 2009, Strassburg et al., 2013, Steffens et al., 2009). This three-dimensional culture method provides gathering a lot of basic information about physiology, cell metabolism, tumor biology and toxicology as well as cell organization and development of artificial tissue (Kelm et al., 2003). It is also an established method for investigation of tumor growth, testing of pharmaceutical substances and investigation of myelination or demyelination processes in the central nervous system (Vereyken et al., 2009). Armstrong et al. (2007) could show that in NG monoculture (Fig. 1A) outgrowing neurites are not as long as in the indirect co-culture of NG and SC (Fig. 1B). Even longer neurites were seen in direct co-culture (Fig. 1C). If this effect is due to cell junctions between NG and SC, a 3D co-culture for example as in a spheroid is expected to have much longer neurites (Fig. 1D). The aim of the presented study was to investigate whether the transformation of a 2D co-culture of SC and NG into a spheroidal 3D co-culture with multiplication of cell junctions, leads to longer neurites. 2. Materials and methods 2.1. Schwann cell isolation, purification and cell culture SC were isolated and cultured as previously published (Penna et al., 2012). Neonatal rat pups (2–5 days postnatal) were anesthetized and then killed by decapitation. Sciatic nerves were exposed bilaterally and excised under sterile conditions. Enzymatic digestion was performed by incubating in Hank’s Balanced Salts
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
Solution (HBSS; Gibco BRL Life Technologies, Karlsruhe, Germany) containing 0.25% trypsin (PAA Laboratories, Coelbe, Germany) and 0.1% collagenase A (Roche Diagnostics, Penzberg, Germany) for 30 min at 37 ◦ C. The reaction was stopped by adding 5 ml culture medium (Dulbecco’s Modified Eagle Medium (DMEM), low glucose (1 g/l), with l-glutamine, (PAA Laboratories, Linz, Austria) containing 10% fetal calf serum (FCS; PAA, Linz, Austria) and 1% penicillin/streptomycin (PS; PAA Laboratories GmbH, Coelbe, Germany). The cells were counted using a cell counter and analyzer system (CASY-1® Cell counter and analyzer system, Schärfe System GmbH, Reutlingen, Germany). The suspension was centrifuged (Megafuge 3.0R, Heraeus Instruments GmbH, Hanau, Germany) for 5 min at 1000 rpm and 4 ◦ C. Higher Schwann cell purification was achieved via negative cell isolation by magnetic elimination of fibroblasts after labeling with magnetic particles (Anti-Mouse CD90.2 Magnetic Particles, BD Biosciences Europe, UK) (Assouline et al., 1983). The pellet was resuspended in Becton Dickinson (BD) IMagTM -Buffer (1 − 8 × 107 cells per ml) and transferred into an Eppendorf tube. After brief dispersion of the CD90.2-antibody, 50 l per 107 cells were added and an incubation time of 30 min at 4 ◦ C followed. After that the Eppendorf tube was placed in a magnet field of a magnetic stand (Dynal MPC-S, Magnetic Particle Concentrator, Dynal Biotech ASA, Oslo, Norway) for 6–8 min at room temperature. The supernatant containing the untouched SC was transferred into a new Eppendorf tube with the magnetically labeled fibroblasts being hold back in the magnetic field. This supernatant was further purified by incubating it with the CD90.2-antibody again. The suspension containing the SC after the second purification step was centrifuged for 5 min at 1000 rpm and 4 ◦ C, the supernatant was removed, taking care not to aspirate the cell containing pellet and the pellet was resuspended in 10 ml culture medium. The cells were seeded on a Poly-d-lysine vented flask (Biocoat 75 cm2 , BD Biosciences Europe) and cultured in a cell incubator (37 ◦ C, 5% carbon dioxide (CO2 ); Forma Scientific CO2 water jacketed incubator, model 3111, ThermoScientific, Waltham, USA). After 24 h 1 M cytosine arabinoside (Sigma Aldrich, Deisenhofen, Germany) was added for 48 h to reduce the remaining fibroblasts. After removing the cytosine arabinoside, the cells were maintained in culture medium with 5 M forskolin (Calbiochem, La Jolla, California, USA) for proliferation stimulation. The medium was changed every 72 h. In order to verify and quantify the purity grade, the SC were stained with calcium binding protein S100 primary antibody (rabbit anti-mouse, DAKO, Denmark; 1: 200 in phosphate buffered saline (PBS; Biochrom AG, Berlin, Germany) containing 1% bovine serum albumin (BSA; Sigma Aldrich, Deisenhofen, Germany) and 0.24% Triton X (Sigma Aldrich, Deisenhofen, Germany)) and rhodamine conjugated secondary antibody (goat anti-rabbit IgG-R, mouse/human adsorbed, Santa Cruz Biotechnology, USA; 1:200 in PBS containing 1% BSA and 0,24% Triton X). Sixty thousand cells per chamber were cultivated in a chamber slide® (NalgeNunc International, Rochester, USA). The cells were fixed using 500 l formaldehyde (4%; FlukaChemie, Buchs, Switzerland). After washing 3× with PBS, non-specific binding surfaces were blocked by incubation with PBS containing 1% BSA and 0.24% Triton X for 10 min at room temperature. The fixed cells were incubated with the primary antibody (200 l/chamber) for 30 min at room temperature, washed 4× 5 min with PBS, incubated with the secondary antibody (200 l/chamber) for 30 min at room temperature and washed 4× 5 min with PBS. The slides were observed under an Axioplan fluorescence microscope (Carl Zeiss Jena, Jena, Germany), photographed using a digital camera (AxioCam, Zeiss, Germany) and analyzed with Axiovision 2.0.5 Software (Carl Zeiss Vision, Munich, Germany).
99
2.2. NG108-15 cell culture The neuronal cell culture of NG, a hybrid cell line from mouse neuroblastoma- and rat gliomacells, was purchased from ECACC (European Collection of Cell Cultures, Salisbury, UK). The NG were seeded in a Costar® 75 cm2 polystyrene flask (Corning Incorporate, USA) and cultured in culture medium. The medium was replaced every 48 h. For verification NG and SC were stained with antineurofilament 200 (Monoclonal Anti-Neurofilament 200, Mouse Ascites Fluid, N 0142, Sigma Aldrich, Deisenhofen, Germany). SC were used to proof the specificity of the first antibody for NG. As secondary antibody a HRP-conjugated anti-mouse antibody was used (goat anti-mouse IgG, HRP-conjugated, EnvisionTM + System HRP Labeled Polymer. Anti-mouse, DAKO, Glastrup, Denmark). In order to identify the NG in co-culture, they were labeled with the PKH-26 red fluorescent cell linker kit (Sigma Aldrich, Deisendorf, Germany). The fluorescent particles were integrated into the cell membrane of the cells. Labeling was performed following the kit protocol using a final PKH concentration of 6.7 × 10-6 M over a 4 min incubation time. Successful labeling was confirmed microscopically using ultraviolet light (wavelength 567 nm, Axiovert 100, Zeiss, Jena, Germany). 2.3. Methocel stock solution Six grams of autoclaved carboxymethyl cellulose (Sigma, Deisenhofen, Germany) were dissolved in 250 ml prewarmed (60 ◦ C) endothelial cell basal medium (ECBM, EBM® -2, Lonza Inc., USA) by mixing with a magnetic stirrer for 20 min. After reaching room temperature again, another 250 ml ECBM with 20% FCS were added. The solution was mixed for 1 h and aliquoted into 50 ml falcons. The undissolved carboxymethyl cellulose was removed by centrifugation (2.000×g, for 4 h at room temperature). Only the clear, gel-like supernatant was used for the experiments. 2.4. Collagen type I extraction and collagen gel Collagen type I was extracted by incubating ligaments of two wistar rat-tails in 250 ml 0.1% acetic acid (Merck, Darmstadt, Germany) for 48 h at 4 ◦ C. The formed highly viscous solution was centrifuged at 24000×g and 4 ◦ C. The clear supernatant was removed to get the collagen stock solution, which was adjusted to an optical density of 0.34–0.36 by using 0.1% acetic acid. Prior to use, 290 l of collagen mix (88.8% collagen stock solution, 9.7% medium 199 10× (Sigma, Germany) and 1.5% hepes buffer 1 M (Gibco, life technologies Inc., UK) neutralized to pH 7.4 with 2 N NaOH (Merck, Germany)) and 290 l of methocel stock solution with 20% FCS were mixed to get about 500 l of clear collagen gel. 2.5. 2D Co-Culture NG were detached by gently shaking the flask, SC by using trypsin (Gibco, life technologies Inc., UK). Both were counted with the CASY-1® cell counter and mixed 1:10 (NG:SC). Cells were then seeded into a 24 well flask (Corning, USA) (66,000 cells/well). To approximate the culture conditions to the 3D experimental setup, a collagen gel layer of 250 l was added on top of the 2D co-culture after cell attachment to the flask. After polymerization of the collagen gel 250 l of culture medium were added on top (Fig. 2). The medium was changed every 48 h and the culture was kept in a cell incubator (37 ◦ C, 5% CO2 ). The 2D coculture was observed microscopically and length was measured daily for 10 days.
100
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
micro liter of the gel containing approximately 50 spheroids were rapidly transferred into a well of a pre warmed 24-well suspension plate and incubated at 37 ◦ C and 5% CO2 . After polymerization (30 min), 250 l culture medium was added on top. The medium was changed every 48 h. The 3D co-culture was observed and photographed over a period of 10 up to 20 days. 2.7. Image analysis
Fig. 2. 2D co-culture, model of experimental setup.
2.6. 3D co-culture 2.6.1. Spheroid formation A spheroidal culture assay was used in a modification of an angiogenesis assay as previously described(Korff, 1998, Finkenzeller et al., 2009). In brief, NG and SC were mixed 1:10 (NG: SC, 550 cells/spheroid), centrifuged and resuspended in culture medium with 20% methocel stock solution. 25 l portions of this suspension were seeded on plastic dishes (Greiner Bio-One, Hungary) as hanging drops (Fig. 3) to allow overnight spheroid aggregation in cell incubator (37 ◦ C, 5% CO2 ). 2.6.2. Sprouting assay After overnight incubation the spheroids were harvested by rinsing with PBS and centrifuged at 1100 rpm for 5 min (4 ◦ C). The pellet containing the spheroids was resuspended in 290 l methocel stock solution containing 20% FCS. 290 l collagen mix was added on top and the solution was mixed gently but thoroughly to achieve a good distribution of the spheroids in the gel. Five hundred
The 2D co-culture as well as the 3D co-culture were observed microscopically (Axiovert 100, Zeiss, Germany) in phase contrast and ultraviolet light and documented photographically (AxioCam, Zeiss, Germany). UV-light was used for identification of PKH-labeled NG. Phase contrast photographs were used for length measurement (AxioVision 2.0.5, Carl Zeiss Vision, Munich, Germany). Except for contrast and brightness there was no modification made subsequently. The length of all sprouts with positive PKH-staining of daily randomly chosen 30 NG of the 2D co-culture and of 15 spheroids were measured in 2D on a phase contrast photograph. 2.8. Statistical analysis The mean length and standard error of mean were calculated (Sigma STATTM 3, Systat Software, San José, California, USA). Statistical significance was calculated by t-test and Mann–Whitney Rank-Test. P-values of < 0.05 were regarded as statistically significant. 3. Results 3.1. Schwann cell culture The cultured SC expressed the S100 Antigen, as shown by positive S100 immunofluorescence (Fig. 4). Using magnetic elimination of fibroblasts, the SC cultures presented a high purity level (about 99% after the two purification steps (Assouline et al., 1983)). In cell culture SC showed a multipolar or fusiform morphology and could be effectively expanded for 12–15 weeks without losing their ability to proliferate. 3.2. NG108-15 culture The NG were highly proliferative in culture and the cells were of polymorphic, multi polar morphology. NG could be stained with neurofilament 200, the stain was located especially in the neurites. When trying to stain SC the same way, there was no positive signal, which demonstrated that Neurofilament 200 could be considered as specific marker for NG in the co-culture. In order to distinguish the NG from SC in the 2D- and 3D coculture, NG were successfully labeled with red fluorescent cell linker PKH-26 (Fig. 5). In culture and after cell separation, the cells showed a distinctive emission in ultraviolet light. No intercellular transfer or into the surrounding collagen matrix could be seen. 3.3. Co-cultures
Fig. 3. 3D co-culture, model of experimental setup. Spheroid formation in hanging drops and transfer in 3D collagen matrix.
2D co-culture with collagen layer addition could be successfully performed. There was no migration of the cells into the collagen matrix and the cells were adherent to the flask. The cultivation could be maintained easily during the 10 days of observation. After this period, evaluation became difficult because of an increase of cell density, so the 10th day was defined as cut off point. The spheroid formation (3D co-culture) in hanging drops generated round three-dimensional cell cultures, homogeneous in size and shape. Embedding the spheroids into collagen matrix with 50
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
101
Fig. 4. Characterisation of SC. Comparison of contrast image with corresponding fluorescence image of SC positive stained for S100. Scale bar 100 m.
Fig. 5. NG cell culture after labeling with PKH26. Comparison of phase contrast image with correspronding fluorescent image. Scale bar 50 m.
spheroids per 500 l gel, resulted in an even distribution in the gel. NG showed neurite outgrowth in 2D and in 3D co-culture. In 2D co-culture, the neurites formed by the NG were of filigree constitution, little elongation was found during the 10 days and there was no change in morphology (Fig. 6).
In the first four days of 3D co-culture most of the neurites had a similar filigree appearance. But then the caliber of the sprouts enlarged from the base of the spheroid and the sprouts grew (Figs. 7 and 8). Additionally to the excessive neurite elongation, there was also a growth of the entire spheroid. The spheroid
Fig. 6. 2D co-culture of NG with SC. Comparison of phase contrast image (A, C) and fluorescent image (B, D). (A) and (B) day 5, scale bar 100 m. (C) and (D) day 10, scale bar 50 m. Red stain = PKH26-labeled NG. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
102
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
Fig. 7. Development of a NG-SC cospheroid from day 3 till day 8. Comparison of phase contrast image (A, C, E, G, I, K) and fluorescent image (B, D, F, H, J, L). (A) and (B) day 3, (C) and (D) day 4, (E) and (F) day 5, (G) and (H) day 6, (I) and (J) day 7, (K) and (L) day 8, Red stain = PKH26-labeled NG, scale bar 100 m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
diameters doubled during the 10 days of observation (day 1 149.94 m ± 10.66 m, day 10 339, 16 m ± 22.5 m; p < 0.001, Fig. 9). For length comparison, the neurites of 30 NG randomly picked in 2D culture, and the neurites of 15 randomly chosen spheroids in 3D culture were measured. The direct comparison of the sprout length of both co-cultures revealed significantly higher neurite lengths in the 3D co-culture from day 7 until day 10 (p < 0.001, Fig. 10). While the neurite length in 2D culture stabilized around 65 m, the sprout length of the NG-SC-spheroids elongated from day 7 to day 10 from 115, 56 m (±8,1) to 176, 6 m (±14.26). This resulted in 2.3× longer 3D neurites than 2D neurites (mean day 7 to day 10; p < 0.001; Fig. 10).
4. Discussion Once developed for cancer research (Sutherland, 1988), spheroidal cell cultures are a common method in tissue engineering to approximate in vitro cultures to in vivo conditions (Lin et al., 2008). A lot of approaches to establish reproducible spheroidal models for different cells and purposes have been developed (Lin et al., 2008). The one by Korff and Augustin (Korff, 1998) is a highly reproducible but also modifiable spheroidal model. In combination with embedding the created spheroids in a three-dimensional matrix, it delivers an easy to establish co-culture model in a 3D environment, which can be adjusted to requirements related to the purpose. For instance different types of cells can be used for spheroid formation and spheroids can be embedded in different matrices. This so called sprouting assay is an established model
in angiogenesis research (Korff, 1998, Korff and Augustin, 1999, Finkenzeller et al., 2009, Strassburg et al., 2013, Steffens et al., 2009). Armstrong et al., 2007 showed that, in addition to soluble factors secreted by SC, the direct cell-cell-contact between SC and neuronal cells in a 2D culture does have a positive influence on neurite length. Knowing these results and the spheroidal sprouting assay established in angiogenesis research, we wanted to prove, if transforming the 2D co-culture model of SC and neuronal cells, which is a well established model in the in vitro research of nerve regeneration (Callizot et al., 2011, Sango et al., 2012), to a 3D spheroidal sprouting assay delivers a better model for nerve regeneration research. To our knowledge this is the first description of such an approach in literature. Other 3D nerve regeneration models either use monocultures of e.g. dorsal root ganglia cells (Bozkurt et al., 2007, Kofron et al., 2009) or PC12 cells (Pittier et al., 2005) or need more complex matrices to ensure that cells can interact in co-culture. Such matrices can be complex 3D fiber scaffolds (Daud et al., 2012) or collagen sponges (Gingras et al., 2008). Even more complex models embed whole tissue pieces like e.g. whole dorsal root ganglia (Allodi et al., 2011). Being a three dimensional model, the spheroidal cultivation system with increased cell-cell-contacts and three dimensional surroundings/environment is more similar to the actual in vivo situation than a two dimensional co-culture (Lu et al., 2012). Nevertheless compared to the complexity of the in vivo constitution, with the influence and interaction of different types of cells, cytokines and the immune system, the spheroidal model is abstract and highly simplified. This again, though offers the possibility to investigate different factors that influence nerve regeneration, independently and fundamentally.
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
103
Fig. 8. Change of neurite morphology within 24 h, same NG-SC co-spheroid with sprout on day 9 (A, C, E) and 10 (B, D, F), phase contrast image (A, B, C, D) and fluorescent image (E, F), Red stain = PKH26-labeled NG, scale bar 100 m. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article).
Fig. 9. Diagram mean diameter of NG-SC co-spheroids per day. Demonstration of the spheroid growth during a cultivation period of 10 days. Significant increase in size from day 1 to day 10 (p < 0.001).
104
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
Fig. 10. Diagram comparing the mean neurite length of NG in 2D and 3D co-culture with SC per day. From day 7 significant higher neurite length in 3D experimental setup. * : p < 0.05; ** : p < 0.001.
The analysis of neurite lengths showed a significant difference between the neurites formed in 2D culture and 3D culture from day 7 on. The length of the neurites in 3D culture was approximately 2.3 (±0.3)-times greater than in 2D culture. This result emphasizes the positive influence SC have on neurite outgrowth through direct cell-cell contact. The SC and NG ratio was identical in both, the 2D and 3D experimental setup. The difference was therefore achieved by the three-dimensional setup with increase of cell-cell-contacts between both cell types. Other possibilities to increase neurite length in addition to increased cell-cell contact are differences in cultivation conditions and positive influence of the extracellular matrix on neurite growth. In order to minimize the interference of different cultivation conditions, a collagen layer was put on the 2D cell culture. We used a collagen-I-matrix, which is a well-established hydrogel for neuron cultivation (Kofron et al., 2009). This component of the physiological extracellular matrix is also employed as a material in experimental nerve conduits (Dornseifer et al., 2007, Bozkurt et al., 2007). With the collagen layer cultivation, conditions of both groups were adjusted to be more comparable. An increase of neurite length due to better cultivation conditions in 3D co-culture is therefore unlikely. Furthermore, in studies by Kofron et al., 2009 collagen did not lead to an increase of neurite length or lead of direction of neurite outgrowth. This finding can be supported in this study, where there was no neurite growth into the collagen layer in the 2D cell culture noticed. Thus it is reasonable to assume that the actual increase of cellcell contacts between SC and NG leads to the achieved length difference between 3D and 2D cell culture. In addition, due to the length measurement technique, the neurites in 3D co-culture system, which do not exactly grow in horizontal planes, were sized too short. The measurement was performed in horizontal planes, in 2D photographs, because this is a fast method that does not require a complex computer program. When measuring the neurites in 2D culture model, there was no deviation, because the neurites grew planar on the culture flask. However, in the 3D culture system the neurites grow into all possible directions. Due to the measurement in the horizontal plane, neurites in 3D set up, which grew in other planes, result in a shorter
Fig. 11. Measurement of neurites in three-dimensional set-up, schematic representation. Deviation of neurite growth from horizontal plane leads to smaller result of neurite length in 3D culture.
length (Fig. 11). The degree of this failure varies, depending on the plane deviating angle. The lengths of neurites measured in 3D culture setup are therefore longer than measured and consequently the difference between the evaluated lengths in 3D and 2D culture are greater than represented in the results. 5. Conclusion Our experiments show that by simply transferring a 2D into a 3D culture with a multiplication of cell-cell contacts, a significant increase of neurite length can be achieved. This result emphasizes the positive influence SC have on neurite elongation through direct cell-cell contact (Lopez-Verrilli and Court, 2012). As shown in this study, the cultivation of NG in spheroidal culture with subsequent sprouting assay in collagen gel, is a relatively simple method to investigate of neurite development with an interesting neuronal sprouting phenomena. Supplementary investigations, e.g. performing cryosections and immunostaining of NG-SC-spheroids in sprouting assay, are projected. Additionally, the experimental setup is flexible and therefore offers the individual investigation of the influence of different factors on nerve regeneration in a “closeto-reality” setting.
D. Kraus et al. / Journal of Neuroscience Methods 246 (2015) 97–105
References Ahmed I, Liu H-Y, Mamiya PC, Ponery AS, Babu AN, Weik T, et al. Three-dimensional nanofibrillar surfaces covalently modified with tenascin-C-derived peptides enhance neuronal growth in vitro. J Biomed Mater Res 2006;A76:851–60. Allodi I, Guzmán-Lenis M-S, Hernàndez J, Navarro X, Udina E. In vitro comparison of motor and sensory neuron outgrowth in a 3D collagen matrix. J Neurosci Methods 2011;198:53–61. Armstrong SJ, Wiberg M, Terenghi G, Kingham PJ. ECM molecules mediate both Schwann cell proliferation and activation to enhance neurite outgrowth. Tissue Eng 2007;13:2863–70. Assouline JG, Bosch EP, Lim R. Purification of rat Schwann cells from cultures of peripheral nerve: an immunoselective method using surfaces coated with antiimmunoglobulin antibodies. Brain Res 1983;277:389–92. Baldwin SP, Krewson CE, Saltzman WM. PC12 cell aggregation and neurite growth in gels of collagen, laminin and fibronectin. Int J Dev Neurosci Off J Int Soc Dev Neurosci 1996;14:351–64. Barras FM, Pasche P, Bouche N, Aebischer P, Zurn AD. Glial cell line-derived neurotrophic factor released by synthetic guidance channels promotes facial nerve regeneration in the rat. J Neurosci Res 2002;70:746–55. Bozkurt A, Brook GA, Moellers S, Lassner F, Sellhaus B, Weis J, et al. In vitro assessment of axonal growth using dorsal root ganglia explants in a novel three-dimensional collagen matrix. Tissue Eng 2007;13:2971–9. Callizot N, Combes M, Steinschneider R, Poindron P. A new long term in vitro model of myelination. Exp Cell Res 2011;317:2374–83. Daud MFB, Pawar KC, Claeyssens F, Ryan AJ, Haycock JW. An aligned 3D neuronal-glial co-culture model for peripheral nerve studies. Biomaterials 2012;33:5901–13. Dertinger H, Hulser DF. Intercellular communication in spheroids Recent Results. Cancer Res 1984;95:67–83. Dodd J, Jessell TM. Axon guidance and the patterning of neuronal projections in vertebrates. Science 1988;242:692–9. Dornseifer U, Matiasek K, Fichter M, Rupp A, Henke J, Weidner N, et al. Surgical therapy of peripheral nerve lesions: current status and new perspectives. Zentralblatt Für Neurochir—Cent Eur Neurosurg 2007;68:101–10. Ehsan SM, Welch-Reardon KM, Waterman ML, Hughes CCW, George SC. A three-dimensional in vitro model of tumor cell intravasation. Integr Biol 2014;6:603–10. Fine EG, Decosterd I, Papalozos M, Zurn AD, Aebischer P. GDNF and NGF released by synthetic guidance channels support sciatic nerve regeneration across a long gap. Eur J Neurosci 2002;15:589–601. Finkenzeller G, Graner S, Kirkpatrick CJ, Fuchs S, Stark GB. Impaired in vivo vasculogenic potential of endothelial progenitor cells in comparison to human umbilical vein endothelial cells in a spheroid-based implantation model. Cell Prolif 2009;42:498–505. Gingras M, Beaulieu M-M, Gagnon V, Durham HD, Berthod F. In vitro study of axonal migration and myelination of motor neurons in a three-dimensional tissueengineered model. Glia 2008;56:354–64. Goerke SM, Obermeyer J, Plaha J, Stark GB, Finkenzeller G. Endothelial progenitor cells from peripheral blood support bone regeneration by provoking an angiogenic response. Microvasc Res 2014;98C:40–7. Gonzalez-Perez F, Udina E, Navarro X. Stefano Geuna IP, editor. Extracellular Matrix Components in Peripheral Nerve Regeneration, Int Rev Neurobiol, 108; 2013. p. 257–75 http://www.sciencedirect.com/science/article/pii/ B9780124104990000101 Gundersen RWBJN. Characterization of the turning response of dorsal root neurites toward nerve growth factor. J Cell Biol 1980;87:546–54. Haug V, Torio-Padron N, Stark GB, Finkenzeller G, Strassburg S. Comparison between endothelial progenitor cells and human umbilical vein endothelial cells on neovascularization in an adipogenesis mouse model. Microvasc Res 2015;97:159–66.
105
Herbert CB, Nagaswami C, Bittner GD, Hubbell JA, Weisel JW. Effects of fibrin micromorphology on neurite growth from dorsal root ganglia cultured in threedimensional fibrin gels. J Biomed Mater Res 1998;40:551–9. Johnson EO, Charchanti A, Soucacos PN. Nerve repair: experimental and clinical evaluation of neurotrophic factors in peripheral nerve regeneration. Injury 2008;39:37–42. Kelm JM, Ittner LM, Born W, Djonov V, Fussenegger M. Self-assembly of sensory neurons into ganglia-like microtissues. J Biotechnol 2006;121:86–101. Kelm JM, Timmins NE, Brown CJ, Fussenegger M, Nielsen LK. Method for generation of homogeneous multicellular tumor spheroids applicable to a wide variety of cell types. Biotechnol Bioeng 2003;83:173–80. Klee WA, Nirenberg M. A neuroblastoma × glioma hybrid cell line with morphine receptors. Proc Natl Acad Sci U S A 1974;71:3474–7. Klimaschewski L, Hausott B, Angelov DN. Stefano Geuna IP, editor. The Pros and Cons of Growth Factors and Cytokines in Peripheral Axon Regeneration, Int Rev Neurobiol, 108; 2013. p. 137–71 http://www.sciencedirect.com/science/article/ pii/B978012410499000006X Kofron CM, Fong VJ, Hoffman-Kim D. Neurite outgrowth at the interface of 2D and 3D growth environments. J Neural Eng 2009;6:016002. Korff T. Integration of endothelial cells in multicellular spheroids prevents apoptosis and induces differentiation. J Cell Biol 1998;143:1341–52. Korff T, Augustin HG. Tensional forces in fibrillar extracellular matrices control directional capillary sprouting. J Cell Sci 1999;112(Pt 19):3249–58. Lin R-Z, Lin R-Z, Chang H-Y. Recent advances in three-dimensional multicellular spheroid culture for biomedical research. Biotechnol J 2008;3:1172–84. Lopez-Verrilli MA, Court FA. Transfer of vesicles from Schwann cells to axons: a novel mechanism of communication in the peripheral nervous system. Front Physiol 2012:3 http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3374349/ Lu H, Searle K, Liu Y, Parker T. The effect of dimensionality on growth and differentiation of neural progenitors from different regions of fetal rat brain in vitro: 3-dimensional spheroid versus 2-dimensional monolayer culture. Cells Tissues Organs 2012;196:48–55. Penna V, Stark G-B, Wewetzer K, Radtke C, Lang EM. Comparison of Schwann cells and olfactory ensheathing cells for peripheral nerve gap bridging. Cells Tissues Organs 2012;196:534–42. Pittier R, Sauthier F, Hubbell JA, Hall H. Neurite extension and in vitro myelination within three-dimensional modified fibrin matrices. J Neurobiol 2005;63:1–14. Sango K, Kawakami E, Yanagisawa H, Takaku S, Tsukamoto M, Utsunomiya K, et al. Myelination in co-culture of established neuronal and Schwann cell lines. Histochem Cell Biol 2012;137:829–39. Steffens L, Wenger A, Stark GB, Finkenzeller G. In vivo engineering of a human vasculature for bone tissue engineering applications. J Cell Mol Med 2009;13:3380–6. Stoll G, Muller HW. Nerve injury, axonal degeneration and neural regeneration: basic insights. Brain Pathol Zurich Switz 1999;9:313–25. Strassburg S, Nienhueser H, Stark GB, Finkenzeller G, Torio-Padron N. Human adipose-derived stem cells enhance the angiogenic potential of endothelial progenitor cells, but not of human umbilical vein endothelial cells. Tissue Eng Part A 2013;19:166–74. Sutherland RM. Cell and environment interactions in tumor microregions: the multicell spheroid model. Science 1988;240:177–84. Tojima T, Kobayashi S, Ito E. Dual role of cyclic AMP-dependent protein kinase in neuritogenesis and synaptogenesis during neuronal differentiation. J Neurosci Res 2003;74:829–37. Vereyken EJF, Fluitsma DM, Bolijn MJ, Dijkstra CD, Teunissen CE. An in vitro model for de- and remyelination using lysophosphatidyl choline in rodent whole brain spheroid cultures. Glia 2009;57:1326–40. Wenger A, Kowalewski N, Stahl A, Mehlhorn AT, Schmal H, Stark GB, et al. Development and characterization of a spheroidal coculture model of endothelial cells and fibroblasts for improving angiogenesis in tissue engineering. Cells Tissues Organs 2005;181:80–8.