This paper lists the steps required to seed neuroblastoma cell lines on previously described three-dimensional collagen-based scaffolds, maintain cell growth for a predetermined timeframe, and retrieve scaffolds for several cell growth and cell behavior analyses and downstream applications, adaptable to satisfy a range of experimental aims.
Neuroblastoma is the most common extracranial solid tumor in children, accounting for 15% of overall pediatric cancer deaths. The native tumor tissue is a complex three-dimensional (3D) microenvironment involving layers of cancerous and non-cancerous cells surrounded by an extracellular matrix (ECM). The ECM provides physical and biological support and contributes to disease progression, patient prognosis, and therapeutic response.
This paper describes a protocol for assembling a 3D scaffold-based system to mimic the neuroblastoma microenvironment using neuroblastoma cell lines and collagen-based scaffolds. The scaffolds are supplemented with either nanohydroxyapatite (nHA) or glycosaminoglycans (GAGs), naturally found at high concentrations in the bone and bone marrow, the most common metastatic sites of neuroblastoma. The 3D porous structure of these scaffolds allows neuroblastoma cell attachment, proliferation and migration, and the formation of cell clusters. In this 3D matrix, the cell response to therapeutics is more reflective of the in vivo situation.
The scaffold-based culture system can maintain higher cell densities than conventional two-dimensional (2D) cell culture. Therefore, optimization protocols for initial seeding cell numbers are dependent on the desired experimental timeframes. The model is monitored by assessing cell growth via DNA quantification, cell viability via metabolic assays, and cell distribution within the scaffolds via histological staining.
This model's applications include the assessment of gene and protein expression profiles as well as cytotoxicity testing using conventional drugs and miRNAs. The 3D culture system allows for the precise manipulation of cell and ECM components, creating an environment more physiologically similar to native tumor tissue. Therefore, this 3D in vitro model will advance the understanding of the disease pathogenesis and improve the correlation between results obtained in vitro, in vivo in animal models, and human subjects.
Neuroblastoma is a pediatric cancer of the sympathetic nervous system arising during embryonic development or early post-natal life due to the transformation of neural crest cells1. It is the most common solid extracranial tumor in children, representing 8% of the malignancies diagnosed in patients under 15 years and is responsible for 15% of all childhood cancer deaths. The disease displays highly heterogeneous clinical behaviors due to specific chromosomal, genetic and epigenetic alterations, and histopathology features.
These alterations contribute to the aggressiveness of neuroblastoma and poor outcomes in pediatric patients. Hence, current therapies prove ineffective in the long term for almost 80% of patients with the clinically aggressive disease2, highlighting the fact that treatment for this group of patients remains challenging. This is likely due to the mechanisms of neuroblastoma heterogeneity and metastases still not being fully understood. However, the tumor microenvironment (TME) is now widely believed to play a role in the progression of many cancers; yet it remains understudied in neuroblastoma3,4.
The native TME is a complex 3D microenvironment involving cancerous and non-cancerous cells surrounded by an ECM. The ECM refers to the acellular component of a tissue that provides structural and biochemical support to its cellular residents and contributes to disease progression, patient prognosis, and therapeutic response5. This promotion of disease progression is due to "dynamic reciprocity" or ongoing bidirectional communication between cells and the ECM6,7,8. As cancer progresses, stromal collagen is reorganized often in linear patterns perpendicular to the stroma-cancer interface, which cancer cells use as a migratory route to metastasis9,10,11.
The main components of this native functional biological scaffold include a fibrous network of collagens type I and II and other proteins, including elastin, glycoproteins such as laminin, as well as a range of proteoglycans and other soluble components12,13. These proteins of the native ECM have now become attractive natural biomolecules for developing 3D in vitro models3. The application of 3D scaffolds for in vitro cell culture is increasing in popularity owing to its greater physiological representation of the TME compared to traditional 2D monolayer culture. The manufactured 3D scaffolds assist cell attachment, proliferation, migration, metabolism, and response to stimuli seen in in vivo biological systems.
The principal component of these 3D scaffolds is collagen, which is a key player in many normal biological processes including tissue repair, angiogenesis, tissue morphogenesis, cell adhesion, and migration11. Collagen-based 3D matrices have shown their robust functionality to model ECM, serving as an in vitro biomimetic microenvironment while enabling cell-ECM interactions as well as cell migration and invasion. These 3D matrices also provide a more accurate analysis of cell response to chemotherapeutic drugs than traditional 2D or "flat" culture in many cancer models14,15,16, including neuroblastoma17,18. Genetic analysis of 3D cell cultures has reported a higher correlation with the human tissue profile even when compared to animal models19. Overall, the cornerstone of these 3D scaffolds is to provide cells a suitable in vitro environment, which recapitulates the native tissue architecture and facilitates bidirectional molecular crosstalk8.
To increase the complexity of collagen-based models, other common ECM components are incorporated in the tissue engineering process, thus creating more physiologically relevant models to reflect niche TMEs of different tissues. For example, GAGs, negatively charged polysaccharides present in all mammalian tissues20, facilitate cell attachment, migration, proliferation, and differentiation. Chondroitin sulfate is a specific type of GAG found in the bone and cartilage, which has been previously used in tissue engineering applications for bone repair21,22,23,24,25. Nano-hydroxyapatite (nHA) is the main inorganic constituent of the mineral composition of human bone tissues, constituting up to 65% of bone by weight26 and is therefore widely used for bone replacement and regeneration27. Thus, GAGs and nHA are attractive composites for reconstructing the primary neuroblastoma ECM and modeling the most common metastatic sites of neuroblastoma, bone marrow (70.5%), and bone (55.7%)28.
Scaffolds incorporating these ECM components were originally developed for bone tissue engineering applications with extensive analysis of their biocompatibility, toxicity, and osteoconductive and osteoinductive features29,30. They are porous, collagen-based matrices produced using freeze-drying techniques to control their physical and biological properties. The collagen scaffolds supplemented with either nHA (Coll-I-nHA) or chondroitin-6-sulfate (Coll-I-GAG) demonstrated success in mimicking the primary TME in breast cancer31 and metastasis to bone in prostate cancer15 as well as neuroblastoma17. The freeze-drying technique used to manufacture these composite scaffolds yields reproducible homogeneity in pore size and porosity within the scaffolds22,23,24. Briefly, a collagen slurry (0.5 wt%) is fabricated by blending fibrillar collagen with 0.05 M acetic acid. For Coll-I-GAG, 0.05 wt% of chrondoitin-6-sulfate isolated from shark cartilage is added to the collagen slurry while blending. For the composite Coll-I-nHA scaffolds, nano-sized hydroxyapatite particles are synthesized as previously described27 and added to the collagen slurry at a 2:1 ratio to the weight of the collagen during the blending process. All scaffolds are physically crosslinked and sterilized using a dehydrothermal treatment at 105 °C for 24 h25. Cylindrical scaffolds (6 mm diameter, 4 mm height) are obtained using a biopsy punch and can be chemically crosslinked with 3 mM N-(3-dimethylaminopropyl)-N'-ethylcarbodiimide hydrochloride and 5.5 mM N-hydroxysuccinimide (EDAC/NHS) in distilled water (dH2O) to improve the mechanical properties of the constructs30. This well-optimized manufacturing process of two collagen scaffolds creates scaffolds with reproducible mechanical properties, including pore size, porosity, and stiffness (kPa). Both Coll-I-GAG and Coll-I-nHA scaffolds have varying physical properties, creating different environmental conditions. The properties of each scaffold are displayed in Table 1.
Coll-I-GAG | Coll-I-nHA | |
Scaffold Size (diameter [mm] x height [mm]) |
6 x 4 17 | 6 x 4 17 |
Collagen Concentration (wt. %) | 0.5 17 | 0.5 17 |
Substrate Concentration (wt. %) [based off weight of collagen] |
0.05 15,17 | 200 17 |
Mean pore size (mm) | 96 22 | 96 – 120 29 |
Porosity (%) | 99.5 23 | 98.9 – 99.4 27 |
Stiffness (kPa) | 1.5 27 | 5.5 – 8.63 29 |
Table 1: Overview of the mechanical properties of the two scaffolds adopted for studying neuroblastoma biology.
This paper describes a protocol of assembling a 3D scaffold-based system to better mimic the neuroblastoma microenvironment using neuroblastoma cell lines and previously described collagen-based scaffolds supplemented with either nHA (Coll-I-nHA) or chondroitin-6-sulphate (Coll-I-GAG). The protocol includes downstream methods to analyze the growth mechanisms of the neuroblastoma cells in a more physiologically relevant environment using previously optimized inexpensive methods adapted from 2D monolayer culture Figure 1.
Figure 1: Overall protocol workflow. (A) Cells are grown to sufficient numbers, split, counted, and resuspended in an appropriate volume of medium. (B) This cell stock then undergoes serial dilution to prepare a total of 4 cell suspensions of different densities. (C) Collagen-based scaffolds are sterilely plated in non-adherent 24-well plates, and (D) 20 µL of cell suspension is added to the center of each scaffold and left to incubate at 37 °C, 5% CO2, and 95% humidity for 3-5 h. (E) Complete growth medium (1 mL) is then slowly added to each scaffold, and the plates are placed back into the incubator to allow cell growth for the desired timeframe. (F) At each predetermined time point, several scaffolds are retrieved for cell viability and growth assessment, gene expression analysis, and histological staining. Please click here to view a larger version of this figure.
1. Experimental design
NOTE: The number of scaffolds and cells needed for each experiment will be dependent on the scale of the experiment and can be calculated using the tools in this section on experimental design.
2. Preparation of collagen-based scaffolds
NOTE: Coll-I-nHA and Coll-I-GAG cylindrical scaffolds (diameter 6 mm, height 4 mm) are prepared using established methods15,21,27. Once chemically crosslinked as per previously published methods17, the scaffolds must be used within 1 week.
3. Propagate neuroblastoma cells in a multilayer cell culture flask
NOTE: The optimal seeding density for the multilayer flask will vary. For the flask used in this experiment, the optimal density as per the manufacturer's instructions is 1 × 107 cells. Before seeding the multilayer flask, propagate cells to a density of 1 × 107 cells or higher in an appropriate tissue culture flask (e.g., a T175 cm2 tissue culture flask). To seed cells into the multilayer flask (section 3.1), grow them until 70-80% confluent, harvest, and count the numbers of cells per mL, referring to steps 3.2.16-3.2.20 for performing the cell count. Once the cell suspension is counted, proceed immediately to the seeding of the multilayer flask. Cell culture work must be carried out in a laminar flow hood to maintain sterility.
Figure 2: Cell counting using a hemocytometer. Ten microliters of cell suspension are added to the hemocytometer beneath the coverslip. The chamber is then placed under the 4x objective lens of a microscope, and the number of cells in the four outer corners of the grid are counted. Please click here to view a larger version of this figure.
4. Seed neuroblastoma cells on scaffolds
Figure 3: Serial dilution of cell stock to prepare 4 suspensions for 4 different scaffold seeding densities. (A) Numbers can be adjusted to suit the desired seeding density per scaffold and (B) multiplied for the total number of scaffolds per density, with each scaffold receiving 20 µL of cell suspension. In this example, Density 1 requires 6 × 105 cells per scaffold, equivalent to 1.8 × 107 cells in 600 µL for 30 scaffolds. This number is doubled to begin the serial dilution, as 600 µL is then transferred and diluted in 600 µL of growth medium in the next tube. This process continues until there are 4 cell suspensions with a factor of 2 between each. A negative control is made by adding 600 µL of medium only to a tube. Please click here to view a larger version of this figure.
5. Scaffold retrieval and applications
NOTE: At each time point, several applications can be used to monitor cell growth on the scaffolds or assess gene and protein expression profiles. The conditions of scaffold retrieval will depend on the analysis to be performed, with multiple retrieval methods outlined in the following subsections and demonstrated in Figure 4.
Figure 4: Retrieval of scaffolds for different analyses at each time point. (A) Three scaffold replicates are retrieved for cell viability analysis. (B) These scaffolds can then be washed in PBS, placed in 1% Triton X-100 in 0.1 M NaHCO3, and stored at -80 °C for DNA quantification. (C) Three more replicates are fixed in 10% PFA for 15 min, neutralized in PBS, and stored at 4 °C for histological staining and imaging. (D) Finally, 3 replicates are added to a phenol/guanidine-based cell lysis reagent and stored at -20 °C for gene expression analysis. Abbreviations: PBS = phosphate-buffered saline; PFA = paraformaldehyde. Please click here to view a larger version of this figure.
Figure 5: Preparation of eight DNA standards for the generation of a standard curve. A stock solution of λDNA is provided at 100 µg/mL. This is diluted 50-fold in TE buffer to create standard A at 2000 ng/mL; 400 µL of A is then transferred to tube B, containing 400 µL TE buffer; 400 µL of B is then transferred and diluted 2-fold in C, and so on until G. Standard H is composed of only TE buffer and therefore has a DNA concentration of 0 ng/mL. Abbreviation: TE = Tris-EDTA. Please click here to view a larger version of this figure.
The collagen-based scaffold model described here has many applications ranging from studying neuroblastoma biology to the screening of anticancer therapeutics in an environment that is more physiologically similar to native tumors than conventional 2D cell culture. Before testing a given research question, it is crucial to obtain a complete characterization of cell attachment, proliferation, and infiltration within the desired experimental timeframe. The growth conditions will depend on the biology of each specific cell line. Importantly, several methods of cell growth assessment must be implemented to determine optimal conditions and robust performance.
Here, the viability of neuroblastoma cells grown on scaffolds was assessed using a colorimetric cell viability assay. This assay can be performed as frequently as desired throughout the experimental timeframe. For the described experiment, cell viability assessment was performed on days 1, 7, and 14 for two neuroblastoma cell lines, KellyLuc and IMR32, grown on Coll-I-nHA scaffolds at 4 different densities (Figure 6). Viability on Day 1 was set as a baseline to compare all subsequent measurements. The rate of reduction of the cell viability reagent is reflective of the cell biology and growth characteristics of individual cell lines, including their proliferation rates and metabolism. A correlation between the number of cells seeded on the scaffolds and the level of reduction was expected. In this experiment, the reduction of the cell viability reagent generally increased with each time point for both cell lines at all densities, as expected.
Each density was then assessed individually for both cell lines to compare the reduction across time points. One-way ANOVA with Tukey's multiple comparisons test was performed to detect significant differences in reduction between time points (Figure 7). For both cell lines and all seeding densities, there was a significant increase (P<0.05) in the reduction of the cell viability reagent when comparing day 1 and day 14. This indicated a significant increase in metabolically active cells present on the scaffolds. This increase was not significant in all cases when assessing the 7-day intervals (day 1 vs. day 7, day 7 vs. day 14), demonstrating the importance of the optimization of the seeding density to achieve the desired growth window.
To support the results of the cell viability assay, cell growth on scaffolds can also be indirectly measured via the quantification of dsDNA extracted from scaffolds using a fluorescent dsDNA stain (Figure 8A). Like cell viability, DNA quantification can be done as frequently as desired within the experimental timeline. However, this analysis requires the complete retrieval of scaffolds and termination of cell growth and so must be factored into experimental planning as discussed in section 1. For this experiment, DNA was quantified on days 1, 7, and 14 for two neuroblastoma cell lines, KellyLuc and IMR32, grown on Coll-I-nHA scaffolds at 4 different densities. As the average concentration of dsDNA per cell is known for these cell lines, it was possible to derive the number of cells per sample from the quantified DNA (Figure 8B).
DNA quantification gave rise to higher variability between biological replicates than cell viability assessment but generally increased for each time point, with the highest levels quantified on day 14. IMR32 cells appear to reach higher cell numbers on Coll-I-nHA scaffolds, as indicated by DNA concentration, than KellyLuc cells. Each density was then assessed individually for the two cell lines to compare the reduction across time points. One-way ANOVA with Tukey's multiple comparisons test was performed to detect significant differences in reduction between time points (Figure 8B).
For both cell lines and all seeding densities, there was a significant increase (P<0.05) in cell numbers when comparing day 1 and day 14, with the exception of KellyLuc at seeding density 4 (1 × 105 cells/scaffold), which did not yield significant increases across any of the time points. Similar to the cell viability results, the increases were not significant in all cases when assessing the 7-day intervals (day 1 vs. day 7, day 7 vs. day 14). When comparing the time point trends for cell viability and DNA quantification, there were some slight differences between the two analyses. However, overall similar trends were observed, with mean values increasing between 7-day intervals for most densities. This demonstrates the importance of monitoring cell growth using more than one method.
A visual assessment of cell growth morphology and distribution on the scaffolds was next implemented, encompassing traditional hematoxylin and eosin (H&E) staining as well as IHC. It is expected that the different growth patterns of individual cell lines will lead to varied spatial arrangements on scaffolds, including different degrees of penetration into the scaffold and cell clustering. Scaffolds were formalin-fixed, paraffin-embedded, and cut into 5 mm sections (Figure 9A), preparing the scaffolds for multiple visualization techniques, including histological staining and IHC.
Routine H&E staining was applied to Kelly, KellyCis83, and IMR32 cells grown on collagen-based scaffolds on days 1, 7, and 14 (Figure 9B). This allowed visualization of the cells' spatial orientation on two collagen-based scaffolds over a 14-day period. Cisplatin-sensitive Kelly cells and resistant KellyCis83 cells were grown on both Coll-I-nHA scaffolds (Figure 9B, i) and Coll-I-GAG scaffolds (Figure 9B, ii). Consistent with previously published data, KellyCis83 cells grew at a higher rate and infiltrated deeper into both scaffold compositions than the less invasive Kelly cell line. The H&E stain of another neuroblastoma cell line, IMR32, grown on Coll-I-nHA demonstrates a contrasting growth pattern (Figure 9B, iii). This cell line grew in large, densely packed clusters on the collagen scaffolds over the 14-day period. Brightfield confocal microscopy can be used to visualize the porous architecture of collagen-based scaffolds (Figure 9C) owing to the autofluorescence of collagen fibers.
We stained cells with phalloidin targeting cytoskeletal actin and the nuclear counterstain, 4′,6-diamidino-2-phenylindole (DAPI), to monitor specific cell traits throughout the experimental timeline. An abundance of actin was observed in Kelly and KellyCis83 cells on Coll-I-GAG scaffolds using this technique (Figure 9D). These results demonstrate how multiple imaging techniques can be used to derive spatially resolved information from neuroblastoma cells grown on scaffolds using this protocol. This characterization of cell growth patterns on collagen-based scaffolds over a given period will improve the understanding and interpretation of any downstream biochemical assays.
Protein expression by cells grown on collagen-based scaffolds can be analyzed to compare cellular activity to in vivo scenarios. Previously published data examined the expression of chromogranin A (CgA) as a surrogate secreted marker of neuroblastoma by KellyLuc and KellyCis83Luc cells grown in cell monolayers as well as on Coll-I-nHA and Coll-I-GAG scaffolds (Figure 10). CgA was assessed in the conditioned media using an enzyme-linked immunosorbent assay (ELISA) (Figure 10A). CgA is secreted at a higher rate in the more aggressive chemo-resistant KellyCis83 cell line than in Kelly (Figure 10B,C). This was significant on day 7 on both Coll-I-GAG and Coll-I-nHA scaffolds (P<0.05), whereas there was no significant difference at this time point for cells grown as a monolayer by conventional 2D culture.
These results also highlight the restricted experimental timeline when growing cells in a monolayer, with only 7 days of growth proving feasible before cells reach confluency. The growth of cells on scaffolds overcomes this limitation as they can be maintained over a longer period in more physiologically relevant conditions. The above combination of techniques to acquire information on cell viability, DNA content, cellular morphology and spatial arrangement, and expression profiles facilitates the assessment of the growth of neuroblastoma cells on a range of collagen-based scaffolds. This protocol can also be easily adapted to satisfy specific experimental requirements and desired applications.
Figure 6: Cell viability analysis. (A) General procedure for measuring the viability of neuroblastoma cells on collagen-based scaffolds using a colorimetric cell viability assay. The incubation period must be optimized for each new cell line, referring to the manufacturer's guidelines. (B) Percentage reduction of cell viability reagent by KellyLuc and IMR32 cells grown on Coll-I-nHA scaffolds at four different initial seeding densities, measured on days 1, 7, and 14. Samples were assessed in biological triplicate with error bars representing the standard deviation. Abbreviations: nHA = nanohydroxyapatite; Coll-I-nHA = collagen scaffolds supplemented with nHA. Please click here to view a larger version of this figure.
Figure 7: Cell viability by seeding density for cells grown on Coll-I-nHA over a 14-day period. (A) KellyLuc;(B) IMR32.Titled cell numbers refer to the initial cell seeding density on the scaffolds on Day 0. Samples were assessed in biological triplicate, indicated by triplicate points, with bars representing the mean. One-Way ANOVA with Multiple Comparisons was used to detect significant differences in % cell viability reagent reduction across the three time points, noted on the graphs (ns P > 0.05, * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001). Abbreviations: nHA = nanohydroxyapatite; Coll-I-nHA = collagen scaffolds supplemented with nHA; ANOVA = analysis of variance; ns = not significant. Please click here to view a larger version of this figure.
Figure 8: Quantification of DNA extracted from cells in scaffolds. (A) Process of quantifying dsDNA from cells grown on collagen-based scaffolds using a fluorescent dsDNA stain. (B) Cell numbers from DNA quantification analysis by seeding density for KellyLuc and IMR32 cells grown on Coll-I-nHA over a 14-day period. Titled cell numbers refer to the initial cell seeding density onto scaffolds on Day 0. Samples were assessed in biological triplicate, indicated by triplicate points, with bars representing the mean. One-Way ANOVA with Multiple Comparisons was used to detect significant differences in cell numbers across the three time points, noted on the graphs (ns P > 0.05, * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001). Abbreviations: nHA = nanohydroxyapatite; Coll-I-nHA = collagen scaffolds supplemented with nHA; dsDNA = double-stranded DNA; TE = Tris-EDTA; ANOVA = analysis of variance; ns = not significant. Please click here to view a larger version of this figure.
Figure 9: Tissue processing steps for immunohistochemistry analysis of scaffolds. (A) Schematic representation of the protocol for processing scaffolds for image analysis. This process allows routine histological staining and specific antibody probing using primary antibodies and fluorescently labeled secondary antibodies. (B) Representative images of three neuroblastoma cell lines subjected to H&E staining. H&E images are taken on Days 1, 7, 14 to monitor growth patterns over the time course of the experiment. Scale bar = 200 µm. Dashed squares represent the area that was chosen for zoomed-in 20x images at the lower left edge. Scale bar = 20 µm. (i and ii) H&E of Kelly and KellyCis83 neuroblastoma cell lines (upper and lower panels, respectively) on two types of collagen-based scaffolds. (iii) H&E of IMR32 cell line, representing clustered cellular growth on the Coll-I-nHA scaffold. (C) Representative image of the Kelly cell line, subjected to brightfield confocal microscopy. The collagen autofluorescence allows visualization of the porous scaffold. 10x Scale bar = 200 µm, 20x scale bar = 20 µm. (D) Representative image of embedded scaffolds followed by analysis by IHC with phalloidin and DAPI at 10x magnification, Scale bar = 200 µm. Smaller inside squares represent zoomed-in images (20x), scale bar = 20 µm. Abbreviations: nHA = nanohydroxyapatite; Coll-I-nHA = collagen scaffolds supplemented with nHA; GAG = glycosaminoglycan; Coll-I-GAG = collagen scaffolds supplemented with chondroitin-6-sulfate; H&E = hematoxylin and eosin; IHC = immunohistochemistry; DAPI = 4′,6-diamidino-2-phenylindole. Please click here to view a larger version of this figure.
Figure 10: Protein expression by neuroblastoma cells grown on 3D collagen-based scaffolds compared to 2D plastic. (A) A schematic of how the CgA ELISA was performed on conditioned media of cells grown on 2D plastic or 3D collagen-based scaffolds. (B) CgA protein expression levels taken from conditioned media of cells grown on a 2D plastic monolayer. As the cells reached confluency after 7 days, the 14-day time point was not readable. By day 7 on plastic, there was no significant difference in CgA levels between Kelly and KellyCis83 cell lines. (C) CgA ELISA performed using conditioned media of cells grown on collagen-based scaffolds for 14 consecutive days. On day 7, on both collagen scaffolds, CgA levels are higher in the more aggressive KellyCis83 cell line, highlighting more physiological relevant levels of CgA in 3D matrix compared to 2D monolayer. This figure has been modified from Curtin et al.17. Abbreviations: 3D = three-dimensional; 2D = two-dimensional; CgA = chromogranin A; ELISA = enzyme-linked immunosorbent assay; nHA = nanohydroxyapatite; Coll-I-nHA = collagen scaffolds supplemented with nHA; GAG = glycosaminoglycan; Coll-I-GAG = collagen scaffolds supplemented with chondroitin-6-sulfate; TMB = 3,3',5,5'-tetramethylbenzidine; HRP = horseradish peroxidase. Please click here to view a larger version of this figure.
The 3D scaffold-cancer cell model has been proven as a valuable and versatile tool for gaining mechanistic insight into neuroblastoma cell growth, viability, and infiltration of cells in a simplified TME32. The 3D neuroblastoma model described here mimics the minimal TME and provides more physiologically relevant data than a 2D monolayer culture. A major drawback of 3D cell culture is increased experimental complexity and longer timeframes. Described here is an optimized protocol for seeding, growth, and maintenance of neuroblastoma cells on collagen-based scaffolds followed by downstream analyses and applications, yielding robust characterization of cell growth. We aimed to gain insights into the optimal cell seeding density for the scaffolds to create a predictable and controllable environment for assessing anticancer drug treatments in an expeditious 14-day experimental window. The combination of all these described simple protocols provides a well-rounded assessment of neuroblastoma cell growth in the scaffold-based in vitro culture system.
The critical points in the protocol setup have been emphasized to allow scientists to establish the same in their laboratories quickly. For example, the indicated incubation times for better performance of the colorimetric cell viability assay allow deeper penetration of the reagent into the scaffold pores to reach all cells. Moreover, the fluorescent dsDNA staining technique is robust and straightforward; however, DNA release from the scaffolds requires vigorous cell lysis as the cells are 'trapped' within collagen fibers.
Using the simple DNA quantification assay described, we can identify the log growth phase on collagen-based scaffolds for anticancer drug screening using this model. In the described experimental setting, 4 initial cell seeding densities were used with an overall 14-day period and analysis time points on Days 1, 7, and 14. We identified that KellyLuc cells seeded at 4 × 105 cells/scaffold have the most significantly active proliferation window between Days 7 and 14. This log phase growth data will allow for reliable interpretation of various cell cytotoxicity experiments. It eliminates speculation about any decline in growth or cell death resulting from suppressed growth on the 3D porous platform rather than from drug toxicities. Cell viability is also a widely used assessment for the suitability of 3D platforms to support the growth of different cell types33,34. While there are many assays to measure cell viability, including live/dead staining, ATP measurement, proliferation assays, we found the use of the Alamar Blue colorimetric cell viability assay to be a simple and effective technique to support DNA quantification data.
The combined use of DNA quantification and cell viability provided complementary evidence that, on average, the optimal density to seed cells on the scaffold to achieve continued growth over a 14-day period is 2-4 × 105 cells/scaffold. However, this protocol can easily be adapted to satisfy different experimental timeframes, analysis time points, and downstream applications. Although this protocol describes the evaluation of monoculture cell growth of neuroblastoma cells on scaffolds, the scaffolds are easily amendable for use as a platform for co-culture, described by do Amaral et al., who utilized collagen-GAG scaffolds to co-culture keratinocytes and fibroblasts in an investigation of wound healing35.
The described 3D model enables the visualization of cell growth and infiltration using different well-known techniques, such as immunofluorescence and standard H&E. It is important to visualize the cells along with the characterization of growth using biochemical assays due to the diversity of cell morphology and growth patterns on scaffolds. Understanding the growth pattern can yield insights into growth behavior and future response to anticancer drugs. For example, IMR32 growth using DNA quantification yields similar patterns to Kelly, although upon visualization using H&E, IMR32 grows in larger clusters than Kelly, which displayed more dispersed growth (Figure 9). These varied growth patterns of cell lines in scaffolds reflect the clinical scenario of tumor heterogeneity. Examining anticancer drug response using a panel of cell lines with different morphologies in 3D scaffolds will increase the predictive value for patient response to the same drugs.
Detection of gene or protein expression can also be performed using other approaches such as RT-qPCR or ELISA if the protein of interest is secreted. A surrogate marker of neuroblastoma progression, chromogranin A (CgA)36, was used to additionally characterize neuroblastoma cell growth in 3D. As described in previous work17, CgA secretion increased as cells proliferated (Figure 10). While monolayer cell culture could not capture this increase, as proliferation meant cells reached full confluency in the culture dishes, the use of the 3D collagen scaffolds allowed prolonged assessment of CgA secretion.
This 3D in vitro model may not be suitable for all research questions to study neuroblastoma biology and response to therapeutics. One of the limitations is uneven cell penetration within scaffolds and the formation of cell clusters of varying size, which depends on a given cell line and may lead to uncontrollable diffusion of nutrients and test drugs. This feature affects the robustness in therapeutic screening. However, despite this limitation, it is important to consider that native tumors are also heterogeneous in size and cancer cell distribution and contain many other cell types within the tumor tissue. To overcome this limitation, we propose the use of each cell-populated scaffold as a single microtissue for which the following parameters will be optimized: (a) incubation times for the cell viability reagent to reach the cells and cell clusters, and (b) lysing of the cells in Triton X-100 buffer by preprocessing of cells on scaffolds with a tissue lyser to release the DNA of the cells contained deep in the scaffold.
Another technical limitation of this protocol is the lack of mechanical testing of each batch of newly manufactured scaffolds for this model. However, using the robust manufacturing process of the scaffolds, which have been extensively characterized in relation to physical and chemical properties of the scaffolds, such as compressive and tensile modulus, porosity and visual pore structure, and homogeneity, ensures that scaffold qualities are maintained through batches21,24,27,30,37.
In summary, this paper presents a series of simple methods for the analysis of cellular growth on collagen-based scaffolds. Both the experimental timeline and analysis points can be interchanged depending on the specific research questions. This protocol is also adaptable to other cell types. The results shown above provide evidence on how this compilation of methods gave insight into the optimal seeding density for various neuroblastoma cell lines to create continuous growth over 14 days. The amalgamation of results obtained from all the methods in this protocol yields a superior understanding of cell growth within the 3D collagen matrix. Future utilization of this model will likely involve co-culture systems specific to the neuroblastoma TME and the testing of various novel anticancer drugs.
The authors have nothing to disclose.
This work was supported by the National Children's Research Centre (NCRC), Irish Research Council (IRC), and Neuroblastoma UK. The illustrations were created using BioRender.
Cells | |||
IMR-32 | ATCC | CCL-127 | |
Kelly | ECACC | 82110411 | |
KellyCis83 | Made in lab – derived from Kelly (Piskareva et al., 2015) | – | Increasing exposure to cisplatin. Cross resistance acquired |
SH-SY5Y | ATCC | CRL-2266 | |
Disposable | |||
0.22 µm syringe filter | Millex | SLHP033RS | |
1.5 mL Eppendorf tube | Eppendorf | 0030 120.086 | |
100 mL sterile Pot | Starstedt | – | |
10 mL plastic pipette | Cellstar | 607 180 | |
15 mL Falcon tube | Starstedt | 62.554.502 | |
25 mL plastic pipette | Cellstar | 760 180 | |
50 mL Falcon tube | Starstedt | 62.547.254 | |
5 mL plastic pipette | Cellstar | 606 180 | |
6 mm Biopsy punches | Kai Medical | BP-60F | |
Aluminium foil | – | – | |
Cover Slip | Menzel-Glaser | – | |
HYPERflask | Corning | CLS10030 | |
Microscope slides | Thermo Scientific | J1840AMNT | |
Opaque black 96-well plate | Costar | 3915 | |
Sterile P10 tips | Starlab | S1121-3810 | |
Sterile P1000 tips | Starlab | S1122-1830 | |
Sterile P20 tips | Starlab | S1123-1810 | |
Sterile P200 tips | Starlab | S1120-8810 | |
T-175 (175 cm2 flask) | Sarstedt | 83.3912 | |
T-75 (75 cm2 flask) | Sarstedt | 83.3911.302 | |
Translucent clear 96 well plate | Cellstar | 655180 | |
Translucent non-adherent 24 well plates | Cellstar | 83.3922.500 | |
Equipment | |||
Autoclave | Astell | – | |
Automatic tissue processor | Leica | TP1020 | |
Centrifuge 5804 | Eppendorf | – | |
Hemocytometer | Hausser Scientific | – | |
Incubator | ThermoScientific | – | |
Microtome | Leica | RM2255 | |
Oven | Memmert | Calibrated by: Cruinn diagnostics Ltd | |
P10 pipette | Gilson | ||
P100 pipette | Gilson | ||
P1000 pipette | Gilson | ||
P20 pipette | Gilson | Calibrated by: Cruinn diagnostics Ltd | |
P200 pipette | Gilson | Calibrated by: Cruinn diagnostics Ltd | |
Paraffin section flotation bath | Electrothermal | MH8517 | Calibrated by: Cruinn diagnostics Ltd |
Pipette electronic dispenser | Corning | StripipetterUltra | Calibrated by: Cruinn diagnostics Ltd |
Plate cooler | Leica | EG1140C | Calibrated by: Cruinn diagnostics Ltd |
Refrigerator -20 °C | Liebherr | – | |
Refrigerator -80 °C | Liebherr | – | |
Refrigerator 4 °C | Liebherr | – | |
Seesaw Rocker | DLAb | SK-D1807-E | |
Spectrophotometer – Victor3V Platereader | PerkinElmer | 1420 | |
Tissue culture hood/Laminar flow hood | GMI | 8038-30-1044 | |
Tissue Lyser | Qiagen | TissueLyser LT | |
Tweezers | – | – | |
Water bath | Grant | – | |
Wax embedder | Leica | EG1140H | |
Materials | |||
1 L Water | Adrona – Biosciences | 568 | |
1% Triton-X | Sigma Aldrich | 9002-93-1 | |
10x PBS tablets | Sigma Aldrich | P4417-100TAB | |
37% paraformaldehyde | Sigma-Aldrich | F8775 | |
Alamar Blue Cell Viability Reagent | Invitrogen | DAL1100 | |
Collagen- glycosaminoglycan scaffold | Tissue engineering research group (TERG) | ||
Collagen-nanohydroxyapatite scaffold | Tissue engineering research group (TERG) | ||
dH20 | Adrona – Biosciences | 568 | |
Eosin | Sigma-Aldrich | E4009 | Made as per: (Cunniffe et al., 2010, Fitzgerald et al., 2015; O’Brien et al., 2005) |
EtOH | Sigma-Aldrich | 1.00983.2500 | Made as per: (Cunniffe et al., 2010, Fitzgerald et al., 2015; O’Brien et al., 2005) |
F12 | Gibco | 21765-029 | |
FBS | Gibco | 10270-106 | |
Hemaytoxylin | Sigma-Aldrich | HHS32-1L | |
L-Glutamine | Gibco | 25030-024 | |
MEM | Gibco | 21090-022 | |
miRNA easy Kit | Qiagen | 217004 | |
MNEAA’s | Gibco | 11140-035 | |
Penicillin/streptomycin | Gibco | 015140-122 | |
Qiazol | Qiagen | 79306 | |
Quant-iT PicoGreen dsDNA Assay Kit | Invitrogen | P11496 | |
RPMI | Gibco | 21875-034 | |
Sodium bicarbonate | Sigma Aldrich | S7795-500G | |
Tissue embedding Medium | Sigma | A6330-4LB | |
Trypsin-EDTA | Gibco | 25300-054 | |
Software | |||
Excel | – | Excel 2016 | |
ImageJ | – | – | |
Prism | – | Version 9 |