This model system starts from a myofibroblast-populated fibrin gel that can be used to study endogenous collagen (re)organization real-time in a nondestructive manner. The model system is very tunable, as it can be used with different cell sources, medium additives, and can be adapted easily to specific needs.
Collagen content and organization in developing collagenous tissues can be influenced by local tissue strains and tissue constraint. Tissue engineers aim to use these principles to create tissues with predefined collagen architectures. A full understanding of the exact underlying processes of collagen remodeling to control the final tissue architecture, however, is lacking. In particular, little is known about the (re)orientation of collagen fibers in response to changes in tissue mechanical loading conditions. We developed an in vitro model system, consisting of biaxially-constrained myofibroblast-seeded fibrin constructs, to further elucidate collagen (re)orientation in response to i) reverting biaxial to uniaxial static loading conditions and ii) cyclic uniaxial loading of the biaxially-constrained constructs before and after a change in loading direction, with use of the Flexcell FX4000T loading device. Time-lapse confocal imaging is used to visualize collagen (re)orientation in a nondestructive manner.
Cell and collagen organization in the constructs can be visualized in real-time, and an internal reference system allows us to relocate cells and collagen structures for time-lapse analysis. Various aspects of the model system can be adjusted, like cell source or use of healthy and diseased cells. Additives can be used to further elucidate mechanisms underlying collagen remodeling, by for example adding MMPs or blocking integrins. Shape and size of the construct can be easily adapted to specific needs, resulting in a highly tunable model system to study cell and collagen (re)organization.
Cardiovascular tissues have a prominent load-bearing function. In particular content and organization of collagen fibers in the extracellular matrix contribute to the load-bearing properties and dominate overall tissue strength1. In tissue engineering mechanical conditioning of the construct is used – typically consisting of (cyclic) straining regimens – to enhance tissue organization and mechanical properties2,3. Full understanding of strain-induced collagen organization in complex tissue geometries to create tissues with predefined collagen architecture has not yet been achieved. This is mainly due to our limited knowledge of collagen remodeling in developing tissues. Existing models mainly give information about the final net outcome of collagen remodeling with use of static strain4-6. Here we provide a highly tunable model system that allows the study of collagen (re)organization in a real-time fashion, in 3D, under influence of static or cyclic strain. The tissue constructs are fibrin-based, ensuring that all collagen in the construct is endogenous. Cell and collagen organization in the constructs is visualized, and an internal reference system allows us to relocate cells and collagen structures for time-lapse analysis. In this protocol we will describe the use of the model system for Human Vena Saphena Cells (HVSCs), since these cells are known for their enhanced extra cellular matrix production and ability to remodel the matrix and our established use in engineered cardiovascular tissues7, based on the work of de Jonge et al.8
1. Culture of Human Vena Saphena Cells
Note: Freezing of the cells is not necessary when harvesting HVSCs, but is solely used for storage.
2. Engineering of Fibrin-based Tissue Constructs
Notes: Only use the soft side of the Velcro and face this side upwards. When gluing the Velcro, only cover the Velcro with silicone glue, do not spread glue throughout the well. Since the culture plates have silicone membrane bottoms, use something underneath the well plate for reinforcing the flexible membranes, to ensure easy gluing in to the plate.
Note: To dissolve fibrinogen mix gently to prevent too much foam formation.
Note: Storage on ice is needed to prevent early gelation of thrombin and fibrinogen.
Note: The fluorescent polystyrene microspheres are used as internal reference markers for image analysis. When mixing thrombin and fibrinogen, prevent the formation of air bubbles by carefully pipetting the mixture. Air bubbles will result in holes in the fibrin gel.
Notes: Do this as quickly as possible to prevent gelation before the mixture is pipetted in the well plate. Practice before using cells and beads.
3. Applying Strain and Inducing Changes in Strain and Constraints
4. Visualizing Cells and Collagen
This model system allows for culturing myofibroblast-seeded fibrin gels. Figure 1A shows a tissue cultured first under static biaxial constraints. Tissue constraints are released by cutting the fibrin gel from two constraints, to create uniaxial static constraints, and tissue compacts and remodels afterwards (Figure 1A). For cyclic strain, the tissue is cultured under static biaxial constraints as well. After 5 days cyclic uniaxial strain can be applied (Figure 1B). To induce collagen reorientation, uniaxial strain direction can be changed to the perpendicular direction (Figure 1B). The microspheres seeded with the gel mixture, display a random pattern, which is used to relocate predefined locations for time-lapse imaging (Figure 2). Figures 2A and 2C show an image of cells, collagen and beads. These same images are scanned 3 days (Figure 2B) and 2 days (Figure 2D) later, respectively. Cells and collagen patterns have changed, but bead patterns are used to relocate the cells and collagen. Culturing samples for 12 days results in endogenous collagen production in quantities sufficient to be visualized with confocal microscopy (Figures 3 and 4). Using either static or cyclic culture results in distinctly different collagen organization, where biaxial static culture gives rise to a random collagen organization (Figure 3A) and uniaxial cyclic strain applied to a biaxially constrained tissue results in an aligned collagen structure (Figures 4A and B). This collagen organization can be studied over time, while changing the static constraints or changing the cyclic strain direction. When static constraints are changed from biaxial to uniaxial, collagen orientation changes from a random orientation (Figure 3A) to an aligned orientation (Figure 3B). Cyclic strain induces a change of orientation at the surface of the tissue (Figures 4A and C), but after 3 days, no change in the core of the tissue is observed (Figures 4B and D).
Figure 1. Constructs cultured under biaxial constraints, used to reorganize collagen by (A) releasing static constraints in one direction, or by (B) changing cyclic strain direction. Loading posts are indicated by dotted black lines. Scale bars indicate 3 mm.
Figure 2. Typical example of the use of bead patterns to relocate predefined locations in 3D, in this case in statically cultured samples. Cells are shown in red, collagen in green and beads in blue. A and C show an image of cells, collagen and beads. These images are scanned 3 days (B) and 2 days (D) later, respectively. Bead patterns are used to relocate cells and collagen. Scale bar indicates 50 μm.
Figure 3. Representative images of 12 day static strain with biaxial constraints (A) and reorientation due to uniaxial constraints within 7 days (B). Scale bar indicates 50 μm.
Figure 4. Representative images after 12 days (5 days static, followed by 7 days cyclic strain) for cyclic strain at the surface (A) and ~60 µm into the tissue (B). After a change in cyclic strain direction (after 12 days), at the surface cells and collagen reorient (C) but no changes are seen in the core of the tissue (D) after 3 days.
The described model system of cell-populated fibrin constructs has great potential for the study of cell and collagen (re)organization (de Jonge et al.15), e.g. to be used for tissue engineering purposes. By using fibrin as the initial cell carrier, after fibrin degradation, a tissue is created with cells and endogenous matrix only. In this way, cells are stimulated to react to strain, either static or cyclic in nature, by applying contractile forces16,17, sensing boundary stiffness12, or displaying strain avoidance.
Significance: Fibrin constructs have been used before to study collagen (re)organization18, but not with the ability of applying cyclic strain and/or studying the constructs in a time-lapse manner, both of which are possible in this presented method. The Velcro strips that provide the constraints in this model system have been used before19, but combining this technique with culture plates with flexible membranes allows to apply cyclic strain for prolonged periods of time. Culturing in these plates allows for easy addition and removal of medium and visualization of the constructs while still attached and sterile. This enables studying relevant remodeling processes time-lapsed or even real-time.
Modifications and future applications: Future applications of the system can be found in manipulating the 3D remodeling process, for instance by adding metalloproteinases or agents that interfere with integrin assembly or signaling. Next to the application of additives, the components of the model system can be easily modified when studying cell and collagen (re)organization. Different cell sources, healthy and diseased cells, the maturity of the collagen matrix (e.g. by varying culture time), and the shape and size of the construct can be adapted to specific needs. The system is well suited for other cell sources, instead of the currently used HVSCs. We have already used this system to culture endothelial colony forming cells (ECFCs; PlosOne, accepted for publication), where we observe that ECFCs orient their produced collagen differently upon cyclic strain than HVSCs.
Limitations of the technique: A limitation of the current model system is the relatively large size of the tissues, requiring 1.5 x 106 cells per construct when using a cell density of 15 x 106 cells/ml. This may pose a problem in cases where less available cell sources will be used. Another limitation is the current thickness of the constructs (approximately 1 mm), which is due to the (thick) Velcro attachments. This limits the confocal scanning to only a portion of the entire sample. It is possible to reach a depth of max 200 μm in the tissue, which is enough to visualize cellular responses within the core of the current tissues, but does not result in a full thickness overview.
Critical steps and troubleshooting: As the model system is based on the formation of tissue between the Velcro attachment points, this also encompasses the critical steps in the protocol. Correct cutting and gluing is essential and special attention should be paid to the careful pipetting of the gel mixture into the Velcro strips for proper attachment of the tissue, which is crucial for applying cyclic strain. Further determinative steps can be found in the choice of cell source and medium additives. Optimization will be needed when culturing with a different cell source. Currently, constructs are cultured with ACA for 7 days, to stop the fibrin from degrading. A different cell source may also result in a different time frame with regard to both collagen formation and fibrin degradation.
The authors have nothing to disclose.
This study was performed in the research program of the BioMedical Materials (BMM) institute. BMM is cofunded by the Dutch Ministry of Economic affairs, Agriculture and Innovation. The financial contribution of the Nederlandse Hartstichting is gratefully acknowledged.
Name | Company | Catalog number | Comments |
Culture plastic | Greiner | Includes culture flasks and pipettes | |
Advanced DMEM | Gibco | 12491 | |
Fetal bovine serum | Greiner | 758075 | |
Penicillin/streptomycin | Gibco | 10378016 | |
GlutaMax | Gibco | 35050-079 | |
Elastomer and curing agent | Dow Corning Corporation | 3097358-1004 | Silastic MDX 4-4210# |
Velcro | Regular store | You can buy this at a regular store, only use the soft side | |
Bioflex culture plates | Flexcell Int | BF-3001U | Untreated |
L-Ascorbic Acid 2-phosphatase | Sigma | A8960 | |
ε-Amino Caproic Acid | Sigma-Aldrich | D7754 | |
Bovine thrombin | Sigma | T4648 | |
Bovine fibrinogen | Sigma | F8630 | |
0.45 syringe filter | Whatmann (Schleicher and Scheul) | 10462100 | |
Polystyrene microspheres | Invitrogen | F-8829 | Blue fluorescent, 10 μm diameter |
Flexcell FX-4000T | Flexcell Int | Includes rectangular loading posts | |
Cell Tracker Orange | Invitrogen Molecular Probes | C2927 | |
CNA35-OG488 | Cordially provided by the Laboratory for Macromolecular and Organic Chemistry, Department of Biomedical Engineering, Eindhoven University of Technology | ||
Confocal laser scanning microscope | Carl Zeiss | LSM 510 Meta laser scanning microscope and Two-Photon-LSM mode | |
Amphotericin | Gibco | 15290-018 | Needed for cell isolation |