We present a method for creating a 3D cell culture environment, which can be used to investigate the importance of cell/matrix interactions in cancer progression. Using a simple self-assembling octapeptide, the matrix surrounding encapsulated cells can be controlled, with independent regulation of mechanical and biochemical cues.
There is a growing awareness that cells grown in 3D better model in vivo behavior than those grown in 2D. In this protocol, we describe a simple and tunable 3D hydrogel, suitable for culturing cells and tissue in a setting that matches their native environment. This is particularly important for researchers investigating the initiation, growth, and treatment of cancer where the interaction between cells and their local extracellular matrix is a fundamental part of the model. Moving to 3D culture can be challenging and is often associated with a lack of reproducibility due to high batch-to-batch variation in animal-derived 3D culture matrices. Similarly, handling issues can limit the usefulness of synthetic hydrogels. In response to this need, we have optimized a simple self-assembling peptide gel, to enable the culture of relevant cell line models of cancer and disease, as well as patient-derived tissue/cells. The gel itself is free from matrix components, apart from those added during encapsulation or deposited into the gel by the encapsulated cells. The mechanical properties of the hydrogels can also be altered independent of matrix addition. It, therefore, acts as a ‘blank slate’ allowing researchers to build a 3D culture environment that reflects the target tissue of interest and to dissect the influences of mechanical forces and/or biochemical control of cell behavior independently.
The many roles played by the extracellular environment in cancer development and progression are becoming ever clearer1. Recently, detailed proteomic-based analyses have added to an already convincing literature base, demonstrating that matrix components derived from cancer-associated stromal cells, or the cancer cells themselves, are key factors in events such as the promotion of epithelial mesenchymal transition and metastatic spread2,3,4. Given this recognized importance of the extracellular matrix (ECM), it is becoming crucial to move towards cell culture platforms that allow control over the 3D environment presented to cells. In response to this need, this protocol presents a method for cell encapsulation and culture in a 3D hydrogel, with user-defined ECM composition and mechanical properties5.
Currently, there is a poor correlation between cancer therapeutic efficacy in 2D in vitro culture, the impact of these therapeutics in current in vivo (patient-derived xenograft, PDX) models and their eventual activity in clinical trials6,7. This has led to significant failures in the drug discovery pipeline with an urgent need for improved in vitro models that allow tested therapeutics to ‘fail early, fail cheap’. Many researchers use the mouse mastocytoma-derived product e.g., Matrigel (or similar products) to create 3D matrix-rich environments to grow and observe cell behavior in vitro, including PDX-derived and other close-to-patient cells8,9,10. However, this ‘one size fits all’ approach neglects the complex role played by matrix proteins/glycans in cancer initiation and progression.
Recognition of the role of extracellular matrix (ECM) in the control of cell behavior has also encouraged the use of 3D culture in or on hydrogels composed of specific matrix components11. Whilst this is useful for investigating specific interactions, these systems suffer from the inability to separate mechanical and biochemical instructions between cells and matrix. They may also be difficult to handle and can give unclear read-outs of cell behavior. Collagen gels are a key example of this problem, since cell-mediated gel contraction can dramatically reduce the ability to visualize cells within the gel5. There are also some very elegant, multi-component gel systems, which experts have used to great effect12,13,14. These can incorporate enzyme-sensitive linkers and bio-active motifs but are significantly more complex in their formulation and application than the system described here.
This protocol describes a method for creating fully defined 3D culture models, allowing the roles of the ECM in development and disease to be modeled in vitro. The basis of the 3D model is a peptide gel, which we have previously described as an optimization of a simple self-assembling octapeptide hydrogel5,15,16. By moving away from complex, animal-derived matrices this system offers a significant benefit of improved batch-to-batch consistency and improved handling. In its simple state, the peptide contains no matrix-derived motifs and effectively provides a ‘blank slate’ onto which the user can build functionality.
We demonstrate how the mechanical properties of the peptide gel can be regulated independently, alongside incorporation of matrix proteins/glycans. The system is highly tunable, allowing the encapsulation of a range of cell types in various formats. Importantly for building a cancer model, stromal cells can also be incorporated: either in direct co-culture or separated to allow specific analysis of indirect cancer cell-stroma interactions. Most crucially, the protocol described here requires no complex knowledge of chemistry and can be reproduced in any cell culture laboratory without the need for specialized chemical knowledge or equipment.
We have optimized methods for the study of cell behavior in the peptide gels, including imaging, rheological analysis, extraction of material for PCR5 and embedding for histological assessment. A clear benefit to the simple hydrogel system is the ability to visualize and study the matrix deposited by encapsulated cells. The importance of cell-derived matrices and the benefits of better understanding of how cells re-engineer their local microenvironment was highlighted recently17 and reflects a growing awareness of the importance of trapping cell-secreted matrix components, in a similar way to that which occurs in vivo. Harnessing the ability to model such processes may be one of the fundamental drivers of the improved patient relevance of hydrogel-based disease models.
We have found the peptide gels described here to be a simple, cost-effective, and flexible solution to support 3D culture of multiple cell types. By providing full control over the concentration of peptide used and the protein or glycan additions made, this method allows the peptide gels to be carefully tailored to their application.
The crucial advantage of the peptide gels over existing methods is that matrix composition and mechanical properties may be controlled independently, using a simple method that does not require any complex chemical procedures. The mechanical properties of the peptide gel are primarily determined by the peptide concentration in the initial gel precursor. Subsequent addition of cells and/or matrix components then allows creation of a fully user-defined in vitro environment. Although matrix additions may alter the initial mechanical properties of the gel, this may readily be offset by the independent variation of peptide concentration5. This provides a tangible advantage over existing systems, for instance collagen gels, in which parameters that control stiffness also commonly result in a change in integrin binding motifs20,21.
We have demonstrated the application of the peptide gel for in vitro culture of cancer cell lines and patient-derived material5. The range of stiffness accessible with the peptide gel (in the range of 100s to 1000s of Pa) is ideally suited to replicate normal and tumor matrix environments in soft tissues such as breast. However, we recognize that other applications require considerably stiffer environments, e.g. in the 10–20 kPa range for osseous regeneration. Further modification of the protocol presented here would be required to extend the achievable stiffness into this range, which is more typical of alternative approaches such as alginate gels22. Similarly, here we have described a simple method for functionalization by physical entrapment of matrix proteins/glycans within the peptide gel. For the applications described here, this approach works well and is easily adapted for use by non-specialist groups wishing to use 3D in vitro models of disease. Like many other hydrogels11, the peptide used here can be extended to include cell-binding or other biological motifs and for some applications this approach may be preferable.
We have identified a few key points that need careful attention to ensure success. Formation of the gel precursor is a critical intermediate step that enables the user to check that the conditions used are correct before cells are incorporated. This precursor can be stored for several weeks (at 4°C) but must be incubated at 80 °C and subsequently at 37 °C prior to use. A suitable precursor will be completely liquid at 80 °C, and self-supporting at 37 °C. These checks are essential to ensure that gelation will occur correctly. Cells and/or matrix may then be incorporated under physiological conditions.
Labs already using 3D matrices will be familiar with the careful handling needed to encapsulate cells in the peptide gels. Care must be taken to limit the agitation of cells before and during the encapsulation steps. We have found that specific cell types are differentially susceptible to damage during this process and this must be carefully evaluated by the user. The concentrations of the peptide gel described here allow gelation to proceed in a time frame that, for the cells mentioned, allows cells to be encapsulated before they sink to the bottom of the casting well but slowly enough that they are not damaged by this process. It is, however, of note that some sensitive cell types may require more rapid neutralization to avoid prolonged exposure to raised pH. In this case, the addition of 10 mM HEPES to the medium surrounding the peptide gel can be beneficial.
When adopting the method described in this protocol, it is very important to carefully consider the quality of the peptide source. Rather than being used as a functional motif or coating, the peptide here is the entirety of the non-soluble portion of the hydrogel. Therefore, any contaminants or variation in the peptide structure are likely to have a significant impact on the integrity or capacity to support cell viability in the final hydrogel. When moving to a new batch of peptide, care must be taken to ensure that there is good batch-to-batch consistency from the supplier as well as checking the behavior of the peptide when forming the gel precursor.
In summary, this protocol describes a 3D culture system with a crucial focus on the independent control of mechanical and biological properties. The simplicity and adaptability of the method makes it suitable for adoption by any cell culture laboratory, and for a wide range of applications5. In the future, this protocol may be extended to allow the covalent modification of the peptide sequence. This could be combined with advanced microscopy methods to investigate the tensile forces exerted by cells on their surrounding matrix. Of key importance, however, is the ability to distinguish between artificially incorporated matrix, and matrix synthesized by the encapsulated cells themselves. This ability to control and monitor matrix changes over time will allow unprecedented insights into the roles of cell-matrix interactions in the development of cancer and other diseases.
The authors have nothing to disclose.
We would like to acknowledge funding from the National Centre for the Replacement, Refinement and Reduction of Animals in Research NC/N0015831/1 to JCA, GF and CLRM, NC/T001267/1 to RBC, CLRM, JCA, KL-S, and KS, NC/T001259/1 to JCA, KL-S and CLRM and NC/P002285/1 to AMG, SJ and CLRM. Also funding from the Engineering and Physical Sciences Research Council EP/R035563/1 to KL-S and CLRM and EP/N006615/1 to JLT and CLRM. Figure 1 was created using adapted graphics from Servier Medical Art. Servier Medical Art by Servier is licensed under a Creative Commons Attribution 3.0 Unported License.
Gel fabrication – Reagents | |||
FEFEFKFK | Pepceuticals | n/a | Polypeptide; available from various suppliers. Pepceuticals is our recommended supplier due to the quality of the product. |
PBS 10X | Gibco | 70011-036 | |
Sodium hydroxide (1 M) | Sigma-Aldrich | S2770 | NaOH; dilute to 0.5 M prior to use |
Water | Sigma-Aldrich | W3500 | |
Gel fabrication – Equipment and Consumables | |||
15 mL falcon tubes | Greiner | 188261 | If using different brand ensure the material withstands temperatures of up to 90°C |
24 well plate | Corning Costar | 3524 | Alternative brands/suppliers can be used as long as there is a gap between the insert base and the plate surface |
Centrifuge | Any | 200 x g for 3 minutes | |
Class II Microbiological Safety Cabinet | Any | ||
Fine balance | Any | Readability 0.1 mg | |
Hanging insert for 24 well plate | Millipore | MCRP24H48 | Alternative brands/suppliers can be used as long as there is a gap between the insert base and the plate surface |
Incubator | Any | 37°C, 5% CO2, humidified environment | |
Oven | Any | set to 80°C | |
P1000/200/20/10 pipette | Any | It is essential the pipettes used for the procedure are calibrated | |
P1000/200/20/10 tips | Any | ||
pH meter with microprobe | Any | ||
Spatula | Any | ||
Vortex | Any | ||
Matrix addition | |||
Collagen I (human) | Stem Cell Technologies | 07005 | |
Collagen I (rat tail) | Gibco | A10483 | |
Fibronectin | Stem Cell Technologies | 07159 | |
Hyaluronic Acid | Iduron | HA804 | |
Matrigel | Corning | 354234 | |
Cell encapsulation/culture | |||
B27 Supplement (no retinoic acid) | Gibco | 12587010 | Media additions for serum free cultures (Figure 2D) |
Cholera toxin | Sigma-Aldrich | C-8052 | Media additions for MCF10A cells (Figure 3, 5) |
DMEM | Gibco | 21969-035 | |
DMEM/F12 | Sigma-Aldrich | D8062 | Media additions for MCF10A cells (Figure 3, 5) |
DMEM/F12 Phenol Red Free | Gibco | 21041-025 | Media additions for serum free cultures (Figure 2D) |
DPBS | Gibco | 14190-094 | |
EGF | SourceBiosciences | ABC016 | Media additions for MCF10A cells (Figure 3, 5) |
Fetal Bovine Serum | Gibco | 10500-064 | |
Horse serum | Gibco | 26050-070 | Media additions for MCF10A cells (Figure 3, 5) |
Human cancer/epithelial cell lines | e.g. MCF7/tdTomato MCF7/MCF10a/HCT116-mCherry | ||
Human mammary fibroblasts | e.g. HMFU19 | ||
Hydrocortisone | Sigma-Aldrich | H-0888 | Media additions for MCF10A cells (Figure 3, 5) |
Insulin | Sigma-Aldrich | I9278 | Media additions for MCF10A cells (Figure 3, 5) |
Knockout serum replacement | Gibco | 10828-028 | Media additions for serum free cultures (Figure 2D) |
L-glutamine | Gibco | 25030-024 | |
RPMI | Gibco | 21875-034 | |
RPMI Phenol Red Free | Sigma-Aldrich | R7509 | |
Imaging and other assays | |||
4% paraformaldehyde | Polysciences | 18814 | |
Agar | SLS | CHE1070 | |
Bovine Serum Albumin (BSA) | Sigma-Aldrich | 5482 | |
Confocal and/or fluorescent microscope | Any | e.g. Leica TCS SPE confocal laser scanning microscope (Figures 2-4) | |
DAPI solution | Invitrogen | D3571 | 300 uM working solution |
DPX mounting medium | ThermoFisher Scientific | ||
Glass cover slips | Any | No1 coverslips 0.13 – 0.17 mm thickness | |
Glass-bottom dishes | MatTek | ||
Goat Anti-Rabbit IgG H&L (HRP polymer) | Abcam | ab214880 | |
Haematoxylin and Eosin | Any | ||
Histology molds (disposable, plastic) | Any | ||
Image analysis software | ImageJ | ||
Live/Dead assay kit | Invitrogen | L3224 | |
Microtome | Any | ||
Phalloidin | Life Technologies | F432/R415 | |
Pierce Peroxidase IHC Detection Kit | ThermoFisher Scientific | 36000 | |
Primary Ab Caspase 3 | Abcam | ab34710 | Shown in Figure 3C |
Primary Ab Collagen I | Cell Signalling Technology | 9661 | Shown in Figure 4B, C, F |
Primary Ab pFAK Tyr 397 | ThermoFisher Scientific | 44-624G | Shown in Figure 4E |
Prolong gold/diamond anti-fade mountant with DAPI | Molecular Probes | S36939 | |
Rheometer Physica MCR 301 | Anton Paar | ||
Scalpel | Any | ||
Secondary antibody Goat anti Rabbit AF488 | nvitrogen | a11034 | |
Secondary antibody Goat anti Rabbit AF546 | Invitrogen | a11010 | |
SuperFrost slides | ThermoFisher Scientific | Coating e.g. APES can help to retain microtome sections on slides. | |
Triton X 100 | Sigma-Aldrich | X100 | |
Trypsin-EDTA (0.25%) | Gibco | 25300054 | |
Vibratome | Leica |
.