Herein we present a rapid, facile, and low-cost method for fabricating custom polydimethylsiloxane molds that can be used for producing hydrogel-based engineered tissues with complex geometries. We additionally describe results from mechanical and histological assessments conducted on engineered cardiac tissues produced using this technique.
As the field of tissue engineering has continued to mature, there has been increased interest in a wide range of tissue parameters, including tissue shape. Manipulating tissue shape on the micrometer to centimeter scale can direct cell alignment, alter effective mechanical properties, and address limitations related to nutrient diffusion. In addition, the vessel in which a tissue is prepared can impart mechanical constraints on the tissue, resulting in stress fields that can further influence both the cell and matrix structure. Shaped tissues with highly reproducible dimensions also have utility for in vitro assays in which sample dimensions are critical, such as whole tissue mechanical analysis.
This manuscript describes an alternative fabrication method utilizing negative master molds prepared from laser etched acrylic: these molds perform well with polydimethylsiloxane (PDMS), permit designs with dimensions on the centimeter scale and feature sizes smaller than 25 µm, and can be rapidly designed and fabricated at a low cost and with minimal expertise. The minimal time and cost requirements allow for laser etched molds to be rapidly iterated upon until an optimal design is determined, and to be easily adapted to suit any assay of interest, including those beyond the field of tissue engineering.
Over the past two decades, soft lithography has been used extensively as a fabrication technique to support scientific research, particularly in the fields of microfluidics, materials research, and tissue engineering1,2,3. Replica molding, in which an object with a desired shape is created from a negative master mold, offers a convenient and low-cost method of producing positive PDMS replicates that can be used for casting shaped hydrogels. However, the required negative master molds are typically produced using microfabrication techniques that are expensive, time-consuming, limited in size, and require clean room space and sophisticated equipment. While 3D printing offers a potential alternative, its utility is somewhat limited due to the resolution limits of lower-cost printers and the chemical interactions between common 3D printer polymers and PDMS that can inhibit curing.
Laser cutter systems capable of both cutting and etching materials such as plastic, wood, glass, and metal have recently become drastically less expensive and therefore more accessible for fabricating research tools. Commercial grade laser cutters are capable of fabricating objects on the centimeter scale with minimum features smaller than 25 µm, and further require minimal training, expertise, and time to use. While laser ablation of PDMS has been previously used in the fabrication of microfluidics devices, to our knowledge no manuscript has described a process by which millimeter and centimeter scale molds can be fabricated from laser cut negative master molds4.
We have used this technique primarily to manipulate the shape of engineered tissues in order to improve nutrient diffusion, cellular alignment, and mechanical properties5,6,7. However, versatility of this technique allows for the utilization in any field where molded hydrogels are of interest, such as drug delivery and material science research8. With access to a laser cutter, PDMS mold replicates can be made for nearly any geometry without overhangs (that would inhibit removal without a multi-part mold, which is beyond the scope of this manuscript) and that fits within the dimensions of the laser bed.
1. Create the Vector Format Master Mold Designs
2. Laser Cut the Acrylic Master Molds
3. Prepare the PDMS Molds for Cell or Tissue Culture
4. Cast the Collagen and Fibrin Hydrogel Tissues
NOTE: Use a proper aseptic procedure to maintain sterility.
5. Analysis Techniques: Tissue Compaction
NOTE: Compaction resulting from matrix remodeling is an indicator of tissue viability and development that can be easily measured through optical microscopy and image analysis.
6. Analysis Techniques: Tensile Testing
Note: Both active mechanics (forces or strains generated by an engineered tissue because of cell activity) and passive mechanics (forces or strains generated in response to applied strains or forces) are critical functional characteristics of many engineered tissues, and this is particularly true for engineered cardiac tissues. The micromechanical analyzer used for the analyses is described in the Table of Materials. Other mechanical testing apparatuses could be similarly applied assuming they allow for hydrated testing and are capable of length control and force measurements over ranges and resolutions relevant for the tissue. For tissues with cross-sectional areas on the order of single square millimeters and stiffnesses on the order of tens of kPa, a 5 mN load cell is a good fit. Larger and stiffer materials would require a larger load cell. Prior to testing, ensure that both the force transducer and length controller are properly calibrated.
7. Analysis Technique: Paraffin Histology and Immunohistochemistry
Note: We have had the greatest success in imaging engineered tissue sections using paraffin blocks so that tissue morphology is best preserved. All steps of the process must be carefully considered and tailored to the engineered tissue, including processing samples without vacuum or pressure, empirically determining the appropriate antigen retrieval methods, and titrating the primary antibody concentration. Other techniques, such as using frozen blocks for preparing slides, may require less time and expense while yielding sufficient results depending on the intended application.
8. Analysis Technique: Cell Alignment
Note: Manipulating the tissue shape and internal stress fields can modulate cell alignment, a defining feature of many native tissues.
The optics of the laser cutter will cause etched areas to have very slightly decreased dimensions as etching depth increases, and results in mold walls with a very subtle bevel, due to tapering of the laser beam. This will help facilitate the removal of the cast PDMS molds, but should be carefully considered if very deeply etched negative master molds (>6 mm) are required (Figure 1).
Over time in culture, cellularized constructs will compact due to matrix remodeling, though the rate and extent to which this occurs will depend on the scaffold composition, cell load, and culturing conditions.
Matrix remodeling can occur through both reorganization and degradation of the surrounding matrix (as well as deposition of new matrix), but is typically associated with an increase in mechanical stiffness due to the decrease in cross-sectional area. With collagen-only constructs composed of 1.6 mg/mL collagen and 12 x 106 cardiomyocytes/mL, we see constructs compact to 19.7 ± 2.8% of their initial width over the four days following casting (Figure 2) via the compaction assay. While this assay yields a representative 2D approximation of a 3D process, the ease of data collection and non-destructive nature make it a powerful tool for studying the construct development process. Note that under cell culture conditions, even in the absence of cells, collagen mechanics can change over time due to both self-assembly and cross-linking11. Fibrin can be rapidly degraded by fibrinolysis both in vivo and in vitro if not in the presence of an antifibrinolytic, such as aprotinin or aminocaproic acid12. Therefore, the impact of scaffold components on long term tissue development, and not just tissue formation, should be considered when selecting a construct formulation. If final tissue size is important for a specific application, compaction must also be considered in mold design and empirically determined based on the cell type(s) and matrix composition. Note that tissue compaction can also induce stress fields within the tissue, which can be manipulated in cellularized constructs to encourage cell alignment (Figure 4).
A wide range of scaffold polymer concentrations and initial cell seeding densities have been used to create engineered tissues in the literature, and this can be attributed primarily to differing requirements for various cell types, cell lines, and applications. Based on our own work with hiPSC-derived cardiomyocytes, we believe that a collagen concentration of 1.25 mg/mL and a seeding density of ~ 15 million cells/mL is a good starting point13. Alternatively, fibrin is widely used as a cardiac tissue scaffold material as well, typically in the range of 3 – 4 mg/mL14. Cell seeding density may be selected based on a number of factors depending on the application, but the cell densities of native tissues provide a good reference point. Also consider that highly concentrated cell solutions can become challenging to work with, especially for small volumes. For a given cell population, the scaffold formulation can be tuned; generally by increasing the polymer concentration when tissues are too fragile or break upon compaction, and increasing the polymer concentration when tissues are too stiff or fail to compact15.
Prior to performing passive mechanical analysis at any time point during culture, it may be appropriate to mechanically precondition the construct sample. Preconditioning of natural polymer hydrogels and engineered tissues will increase the reproducibility of the testing result due to material viscoelasticity and provide a better indication of the properties that the construct will exhibit in a clinical application. We use 8 cycles of 10% strain in a triangular waveform at a rate of 10% strain/min prior to starting mechanical assessment (Figure 3).
Tissue and cell-type specific morphology can be assessed through histology and immunohistochemistry with traditional methods. However, we have found that optimization of nearly all steps of the paraffin processing, embedding, sectioning, antigen retrieval, and staining have been necessary for the engineered cardiac tissue compared to plated cells or sectioned native tissue (Figure 4).
Figure 1: Outline of the process for designing and preparing PDMS molds from laser cut acrylic masters. (A) Mold design prepared in vector graphics format. (B) Cleaned laser etched acrylic negative master. (C) PDMS cast on the surface of the taped acrylic master mold. (D) Resulting PDMS mold ready for sterilization prior to tissue culture. Inset: top view, same scale. Please click here to view a larger version of this figure.
Figure 2: Construct compaction over time in culture. (A) Images of rectangular constructs prepared in triplicate compacting over time in culture. Green overlays represent masks used to calculate visible construct area for image analysis. (B) Plot of construct area (a two-dimensional metric of construct compaction) over time. Horizontal lines represent the mean values and error bars indicate the standard deviation. For all groups, n = 3 and * indicates p <0.05 as evaluated by ANOVA. Please click here to view a larger version of this figure.
Figure 3: Raw traces for mechanical characterization of engineered cardiac tissues. Insets display a single representative twitch contraction trace (same axes as main plot). (A) Active mechanical response resulting from rapid steps followed by holds at 5% strain increments. (B) Passive mechanical response resulting from a pull-to-break test at a rate of 10% strain/min. All samples were analyzed in a 37 °C bath of Tyrode's solution. Please click here to view a larger version of this figure.
Figure 4: Paraffin block histology images for engineered cardiac tissue constructs of various designs. Paraffin block histology images for engineered cardiac tissue constructs of various designs stained with (A) hematoxylin and eosin, (B) diaminobenzidine (anti-cardiac troponin T, brown), and hematoxylin nuclear counterstain, (C) picrosirius red stain for collagen with fast green cytoplasmic counterstain, and (D) mouse anti-α-actinin antibody (green) and 4',6-diamidino-2-phenylindole (DAPI) nuclear counterstain (blue). Cell alignment differs as a result of the construct design and tissue compaction. Scale bar in A applies to B and C as well. Inset in D shows striated cardiomyocytes. Please click here to view a larger version of this figure.
Customized PDMS mold geometries that are compatible with tissue culture have great utility in tuning important engineered tissue properties, such as cell alignment, diffusion rate, and effective stiffness. Additionally, these molds are very useful for preparing tissues for analysis applications in which geometry is important, such as mechanical testing16,17. Preparing these devices from laser cut negative master molds offers a rapid, facile, and low-cost method of utilizing these tools, especially when compared to the time and cost associated with traditional microfabrication. Laser cutting also permits a larger maximum mold size, limited only by the bed size of the cutter. We have successfully used these versatile molds to execute a wide variety of studies with engineered cardiac tissues, including optical mapping of action potential propagation, assessment of force production and passive mechanical properties, and implantation in a rat model of myocardial infarction13,18. We recognize that beyond the research niche of cardiovascular regenerative engineering, the applications for the fields tissue engineering, drug delivery, and materials research are vast.
While there are few technically challenging steps in the fabrication of the molds themselves, there are a number of critical steps involved in creating functional tissues. If constructs fail to compact the surrounding matrix after 24 h, first confirm the cell viability through viability staining of the engineered tissues. If cell viability is high, consider altering the construct composition for the next set of tissues. While outcomes will vary greatly depending on the cell population, we have observed increased compaction associated with higher seeding densities and lower collagen concentrations. Finally, it may also be useful to supplement the seeded cell population with a cell type well suited for matrix remodeling, such as fibroblasts, to encourage compaction.
One limitation of these molds is the potential for PDMS to adsorb small hydrophobic molecules. While for our applications this has not been problematic, it may be of concern in assays very sensitive to the loss of these molecules. In these cases, PDMS protein adsorption can be mitigated through treatment with an antifouling agent such as polyhydrophilic or polyzwitterionic materials19. Alternatively, a sterilized PDMS mold could be prepared as a negative master (from a laser-etched positive mold) for a culture mold to be cast in another, non-adsorbent material, such as agarose.
The authors have nothing to disclose.
The authors acknowledge funding from NIH R00 HL115123 and Brown University School of Engineering. They are also grateful to the Brown Design Workshop and Chris Bull for training and support with the laser cutter.
Item | |||
Bovine fibrinogen | Sigma | F8630-5G | Constructs |
Bovine thrombin | Sigma | T6634-250UN | Constructs |
Bovine aprotinin | Sigma | 10820-25MG | Constructs |
Rat tail collagen I, 4 mg/mL | Advanced Biomatrix | 5153-100MG | Constructs |
Sodim chloride | Fisher | BP358-10 | Constructs |
PBS | Life Technologies | 14190-250 | Constructs |
Fine forceps | Fine Science Tools | 11252-20 | Constructs |
Sylgard 184 silicone elastomer | Corning | 4019862 | PDMS Molds |
Lab tape | Fisher | 15-901-5R | PDMS Molds |
Acrylic, 1/4" thick | McMaster-Carr | 8560K356 | PDMS Molds |
HEPES Buffer, 1 M | Sigma | H3537-100ML | Constructs |
RPMI 1640 medium, powder | Fisher | 31800-089 | Constructs |
Calcium chloride dihydrate | Fisher | AC423520250 | Constructs |
Magnesium chloride hexahydrate | Fisher | M33 500 | Constructs |
Potassium chloride | Sigma | P9541-500G | Constructs |
Sodium phosphate dibasic heptahydrate | Sigma | S9390-500G | Constructs |
Glucose | Sigma | G5767-25G | Constructs |
OCT | VWR | 25608-930 | Histology |
Frozen block molds | VWR | 25608-916 | Histology |
Hematoxylin | Fisher | 3530 1 | Histology |
Eosin Y | Fisher | AC152880250 | Histology |
Fast green FCF | Fisher | AC410530250 | Histology |
Software | |||
Illustrator | Adobe Systems | Vector Graphics | |
Inkscape | (Open Source) | Vector Graphics | |
UCP (Universal Control Panel) | Universal Laser Systems | Laser Cutter Interface | |
Equipment | |||
PLS6.75 Laser Cutter | Universal Laser Systems | Laser Cutter | |
Micromechanical Analyzer | Aurora Scientific | 1530A with 5 mN load cell | Mechanical Analysis |