The goal of this protocol is production of whole extracellular matrix fibers targeted for wound repair which are suitable for preclinical evaluation as part of a regenerative scaffold implant. These fibers are produced by the culture of fibroblasts on hollow fiber membranes and extracted by dissolution of the membranes.
Engineered scaffolds derived from extracellular matrix (ECM) have driven significant interest in medicine for their potential in expediting wound closure and healing. Extraction of extracellular matrix from fibrogenic cell cultures in vitro has potential for generation of ECM from human- and potentially patient-specific cell lines, minimizing the presence of xenogeneic epitopes which has hindered the clinical success of some existing ECM products. A significant challenge in in vitro production of ECM suitable for implantation is that ECM production by cell culture is typically of relatively low yield. In this work, protocols are described for the production of ECM by cells cultured within sacrificial hollow fiber membrane scaffolds. Hollow fiber membranes are cultured with fibroblast cell lines in a conventional cell medium and dissolved after cell culture to yield continuous threads of ECM. The resulting ECM fibers produced by this method can be decellularized and lyophilized, rendering it suitable for storage and implantation.
Implantable surgical scaffolds are a fixture of wound repair,with over one million synthetic polymer meshes implanted worldwide each year for abdominal wall repair alone1. However, following implantation the synthetic materials polymers traditionally used in the fabrication of these scaffolds tend to provoke a foreign body response, resulting in inflammation deleterious to the function of the implant and scarring of tissue2. Further, as the predominant synthetic mesh materials (i.e., polypropylene) are not appreciably remodeled by the body, they are generally applicable to tissues where scarring can be tolerated, limiting their clinical usefulness toward the treatment of tissues with higher-order function such as muscle. While there are many surgical mesh products which have been applied with clinical success, recent manufacturer recalls of synthetic surgical meshes and complications from interspecies tissue implants highlight the importance of maximizing implant biocompatibility, prompting the FDA to tighten regulations on surgical mesh manufacturers3,4. Implantation of scaffolds derived from patients' own tissues reduces this immune response, but can result in significant donor-site morbidity5. Extracellular matrix (ECM) scaffolds produced in vitro are a possible alternative, as decellularized ECM scaffolds exhibit excellent biocompatibility, particularly in the case of autologous ECM implants6.
Because of the limited availability of patient tissue to harvest for autologous implantation and the risk of impeding function at the donor site, the ability to produce ECM scaffolds in vitro from the culture of human cell lines or, if possible, a patient's own cells is an attractive alternative. The primary challenges in the manufacture of substantial amounts of ECM in vitro is the sequestration of these difficult-to-capture molecules. In previous work, we have demonstrated that ECM can be produced by culturing ECM-secreting fibroblasts in sacrificial polymeric foams that are dissolved after the culture period to yield ECM which can be decellularized for implantation7,8,9,10. As ECM produced in foams tend to adopt the internal architecture of the foams, hollow fiber membranes (HFMs) were explored as a sacrificial scaffold for production of threads of ECM. Described herein are methods tasked for lab scale manufacturing of cell culture quality hollow fiber membranes and the extraction of bulk extracellular matrix fibers from the same following a period of fibroblast culture. This static culture approach is readily adoptable by laboratories containing standard mammalian cell culture equipment. ECM produced by this approach could be applied toward a variety of clinical applications.
1. Production of Extracellular Matrix Using Sacrificial Hollow Fiber Membranes
CAUTION: N-methyl-2-pyrrolidone is an irritating solvent and reproductive toxicant. Exposure to NMP may cause irritation to the skin, eyes, nose, and throat. Solvent-resistant personal protective equipment should be used when handling NMP. Use of NMP should be performed within a fume hood.
2. Decellularization of Extracellular Matrix
Successful production of extracellular matrix from sacrificial scaffolds is contingent on appropriate scaffold fabrication, cell culture, and solvent rinse procedures. Fabrication of the hollow fiber membranes is performed using a dry-jet wet-spinning system assembled from commercially available components (Figure 1) which uses extrusion of polymer solution through the annulus of a commercially available steel spinneret (inner diameter = 0.8 mm, outer diameter = 1.6 mm) to generate a nascent tube of polymer solution which precipitates into a hollow fiber membrane upon contact with a water bath.
An example process for ECM extraction from HFMs is illustrated in Figure 2. Figure 3A shows a transverse cross-section of polysulfone HFMs fabricated under this protocol, exhibiting outer and inner layers of finger-like pores characteristic of an asymmetric membrane. In this protocol, cells are seeded specifically in the inner lumen of the membrane and cultured in 6 cm diameter Petri dishes (Figure 3B), with cells tending to proliferate on all surfaces of the membrane. Cultured membranes can then be subjected to batch NMP and deionized water rinsing in standard glass scintillation vials (Figure 3C), producing translucent threads of ECM (Figure 3D). ECM-producing cells remain viable inside HFMs throughout the 3-week period of culture (Figure 4).
HFMs cultured for three weeks with primary rat skeletal muscle fibroblasts (RSMF) were dissolved via three exchanges of N-methyl-2-pyrrolidone, after which they were rinsed three times in deionized water. The extracted matrix, normally being translucent in appearance when hydrated (Figure 5A), will tend to cloud upon hydration if not subjected to appropriate solvent rinsing due to the presence of residual polymer. It should also be noted that the ECM remaining after dissolution of the membrane is somewhat fragile, requiring care in handling with fine forceps. ECM fibers assembled into meshes and then lyophilized exhibit an off-white color and fibrous appearance with a gross longitudinal alignment (Figure 5B).
Figure 1: Illustrated process flow for casting hollow fiber membranes. Hollow fiber membranes are manufactured using the ubiquitous non-solvent induced phase separation method (NIPS) using a system prepared from commercially available components listed in the table of specific materials. Polysulfone in NMP (17.8 w/w%) and NMP bore solutions (15 w/w% NMP in water) are extruded from separate stainless steel vessels by pressurized N2 into a spinneret, generating a nascent hollow fiber membrane at the spinneret outlet which fully precipitates upon contact with the tap water precipitation bath. Nascent hollow fiber is manually guided under and over guides and allowed to collect on a rotating motorized take-up wheel at a rate of 2.3 meters per minute. Please click here to view a larger version of this figure.
Figure 2: Illustrated production of ECM from cultured hollow fiber membranes. Hollow fiber membranes are seeded with fibroblasts and cultured for three weeks, followed by exchange of NMP 3x and exchange of deionized water 3x. Please click here to view a larger version of this figure.
Figure 3: Culture and extraction of ECM. Asymmetric mesoporous hollow fiber membranes (A) were cultured for three weeks with RSMF cells in DMEM/F-12 with supplemented ascorbic acid and TGF-β (B). Cultured hollow fibers (C, D top) were dissolved via 3 exchanges in n-methyl-2-pyrrolidone then rinsed three times in deionized water, resulting in continuous threads of ECM (D, bottom). High magnification scanning electron microscopy micrograph of the ECM fiber surface (E). Please click here to view a larger version of this figure.
Figure 4: Cell viability on hollow fiber membranes. Representative hollow fiber membranes cultured with RSMF cells for three weeks were longitudinally sectioned using a fine razor to reveal the luminal surface of the HFMs and subjected to live-dead staining with Calcein AM and EthD-1. Viability staining revealed a confluent layer of viable cells with negligible EthD-1 fluorescence. Please click here to view a larger version of this figure.
Figure 5: Assembly of ECM implant. Individual extracellular matrix threads (n = 30) derived from culture of RSMF cells were placed lengthwise into a silicone mold (A) and decellularized by 1% SDS followed by treatment with DNAse I, RNAse A, and penicillin-streptomycin. Decellularized ECM was then lyophilized, yielding an off-white mesh with a fibrous appearance and longitudinal architecture (B). Please click here to view a larger version of this figure.
The processes described enable the production of bulk ECM biomaterials in vitro using hollow fiber membranes cast by a dry-jet wet spinning system allowing for inexpensive bulk production of membranes as well as standard cell culture equipment. While the membranes fabricated in this protocol are intended for use in cell culture, the system described can also be adapted for the production of membranes for separation purposes, with pore size distribution and hollow fiber dimensions tunable by varying spinneret dimensions, polymer used, dope and bore flow rate, take-up speed, and environmental conditions.13 Though the protocol detailed employs hollow fiber membranes, in principle any dissolvable cell culture scaffold with appropriate transport properties such as open cell foams could be used, as demonstrated in previous work7,8,9. This general approach appears useful for the production of ECM scaffolds by sacrificial scaffolds which can more faithfully mimic the internal architecture of tissues. Implants produced by this protocol in particular exhibit a gross alignment (Figure 5B), which may be of particular benefit toward the reconstruction of highly aligned tissues such as tendons, ligaments and skeletal muscle.
Regular exchange of medium and, in particular, supplementation of medium with ascorbic acid and TGF- β is crucial to production of ECM, as ascorbic acid is an essential enzyme in collagen biosynthesis, and TGF- β induces the synthesis of several ECM proteins10. Additionally, thorough rinsing of ECM with NMP must be performed, otherwise residual polymer will remain in the extracted ECM, appearing as a white film during water rinses. Care must be taken to not overly agitate ECM during membrane dissolution, as it is relatively fragile. Collected ECM scaffolds intended for implantation must be subjected to a decellularization step as described to remove xenogeneic epitopes to minimize a potential host foreign body response.
The significance of this technique lies in its production of a biocomplex scaffold of whole extracellular matrix which can be remodeled by the body’s own wound-healing processes. By using this approach to produce ECM from cells specific to a target species, it may be possible to minimize the foreign body response which hinders the clinical effectiveness of this class of biomaterials; by using cells specific to an individual, the foreign body response may be lessened further. This approach also allows for the production of ECM targeted toward particular tissues, with recent reports suggesting that tissue-specific ECM may be particularly effective in certain applications12. As this protocol allows for production of ECM across various cell lines, implants combining ECM from several cell types (e.g. muscle, nervous, endothelial) could be used to tailor implants to more faithfully approximate the structure and chemical complexity of target tissues. While the ECM meshes presented here were originally intended as scaffolds for wound repair, they may also have use as platforms for investigations into cell-ECM interactions, durotaxis, and biosensing. In particular, this approach lends the investigator the ability to produce ECM from specific cell types of interest which may allow for new insights into the biological significance of tissue-specific ECM structure, composition and function.
Potential improvements to the ECM production techniques presented here could include scale-up via the use of dynamic and pre-conditioning bioreactors as well as exploration of alternative sacrificial scaffold materials and architectures. In particular, transition to a solventless scaffold removal process would improve the safety of the extraction process; sacrificial scaffolds composed of materials which are degradable by enzymes, such as regenerated cellulose, may allow for solventless ECM extraction14. While the tensile strength of these materials is below those of synthetic materials, there exist several avenues for improvement of these scaffolds, including exploration of various culture conditions, media formulations, and sacrificial scaffold geometries. Recent genetic engineering advances could be leveraged to produce pro-fibrotic cell lines, which in combination with platforms for ECM capture could enable the production of scaffolds satisfying clinical demands. Further, existing dynamic culture systems such as continuous flow-loop hollow fiber membrane bioreactors facilitate high rates of nutrient exchange conducive to greater and more rapid ECM production, and may be of particular interest in leveraging this technique for production of larger quantities of ECM for biological research and clinical use.
The authors have nothing to disclose.
Research reported in this publication was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under award number R15AR064481, the National Science Foundation (CMMI-1404716), as well as the Arkansas Biosciences Institute.
1/32 inch thick silicone rubber | Grainger | B01LXJULOM | |
20 mL Scintillation Vials, Borosilicate glass, Disposable – VWR | VWR | 66022-004 | With attached white urea cap and cork foil liner |
3 inch by 1 inch microscopy slides | VWR | 75799-268 | |
4C refrigerator | Thermo Fisher Scientific | FRGG2304D | Any commercial 4C refrigerator will suffice. |
50 mL tubes | VWR | 21008-178 | |
6-well cell culture plates | VWR | 10062-892 | Alternative brands may be used |
Acetone | VWR | E646 | Alternative brands may be used |
Bore vessel | McMaster-Carr | 89785K867 | 6 ft 316 steel tubing |
Bovine Plasma Fibronectin | Thermo Fisher Scientific | 33010018 | Comes as 1 mg of lyophilized protein |
CaCl2 | VWR/Amresco | 97062-590 | |
Cell Culture Incubator w/ CO2 | Any appropriate CO2-supplied mammalian cell incubator will suffice. | ||
Disposable Serological Pipets, Glass – Kimble Chase | VWR | 14673-208 | Alternative brands may be used |
DMEM/F-12, HEPES | Thermo Fisher Scientific | 11330032 | Warm in water bath at 37°C for 30 minutes prior to use |
DNase I | Sigma-Aldrich | DN25-10MG | |
Dope vessel | McMaster-Carr | 89785K867 | 6 ft 316 steel tubing |
Ethanol | VWR | BDH1160 | Dilute to 70% for sterilization |
Fetal Bovine Serum, qualified, US origin – Gibco | Thermo Fisher Scientific | 26140079 | Mix with growth media at 10% concentration (50mL in 500mL media) |
Four 1/4-inch to 1" reducing unions | Swagelok | SS-1610-6-4 | One reducing union for each inlet and outlet of each vessel |
Freeze-dryer/lyophilizer | Labconco | 117 (A65312906) | Any lyophilizer will suffice. |
Hexagonal Antistatic Polystyrene Weighing Dishes – VWR | VWR | 89106-752 | Any weigh boat will suffice |
Hollow fiber membrane immersion bath | 34L polypropylene tubs may be used or large bath containers can be fabricated from welded steel sheets | ||
Hollow Fiber Membrane Spinneret | AEI | http://www.aei-spinnerets.com/specifications.html | Made to order. Inner diameter = 0.8 mm, outer diameter = 1.6 mm |
Hot plate/stirrer | VWR | 97042-634 | |
Human TGF-β1 | PeproTech | 100-21 | |
L-Ascorbic acid | Sigma-Aldrich | A4544-25G | |
L-Ascorbic acid 2-phosphate | Sigma-Aldrich | A8960-5G | |
L-glutamine (200 mM) – Gibco | Thermo Fisher Scientific | 25030081 | Mix with growth media at 1% concentration (5mL in 500mL media) |
MgCl2 | VWR/Alfa Aesar | AA12315-A1 | |
Minus 80 Freezer | Thermo Fisher Scientific | UXF40086A | Any commercial -80C freezer will suffice. |
N2 gas cylinders (two) | |||
NIH/3T3 cells | ATCC | CRL-1658 | Alternative fibrogenic cell lines may be used. |
N-methyl-2-pyrrolidone | VWR | BDH1141 | Alternative brands may be used |
Penicillin/Streptomycin Solution – Gibco | Thermo Fisher Scientific | 15140122 | Mix with growth media at 0.1% concentration (0.5 mL in 500mL media) |
Polysulfone | Sigma-Aldrich | 428302 | Any polysulfone with an average Mw of 35,000 daltons may be used |
Portable Pipet-Aid Pipetting Device – Drummond | VWR | 53498-103 | Alternative brands may be used |
PTFE tubing (1/4-inch inner diameter) | McMaster-Carr | 52315K24 | Alternative brands may be used. |
Rat skeletal muscle fibroblasts | Independently isolated from rat skeletal muscle. Alternative fibrogenic cell lines may be used. | ||
RNase A | Sigma-Aldrich | R4642 | |
Silicone sheet | McMaster-Carr | 1460N28 | |
Take-up motor | Greartisan | B071GTTSV3 | 200 RPM DC Motor |
Tris HCl | VWR/Amresco | 97063-756 | |
Two needle valves | Swagelok | SS-1RS4 |