Chapter 1: Microtechnologies in the Fabrication of Fibers for Tissue Engineering
-
Published:19 Nov 2014
-
Series: Nanoscience & Nanotechnology
M. Akbari, A. Tamayol, N. Annabi, D. Juncker, and A. Khademhosseini, in Microfluidics for Medical Applications, ed. A. van den Berg and L. Segerink, The Royal Society of Chemistry, 2014, ch. 1, pp. 1-18.
Download citation file:
Engineering tissues and organs for implantation in the human body or research require the fabrication of constructs that reproduce a physiological environment. Moreover, the construction of complex and sizable three-dimensional tissues requires a precise control over cell distribution and an effective vasculature network to supply oxygen and nutrients, and remove waste. Fiber-based tissue engineering that forms 3D structures using fibers can address many of these challenges, but depends on the quality of the fibers. Recent progresses in microtechnologies have enabled researchers to fabricate biocompatible fibers with advanced biochemical and physical properties, including cell-laden fibers that are pre-seeded with cells. In this chapter, we discuss fiber fabrication techniques including co-axial flow spinning, wetspinning, meltspinning, and electrospinning, which have leveraged microtechnologies to improve their performance. We compare the properties of the fibers fabricated with these methods and discuss their strengths and weaknesses in the context of tissue engineering.
1.1 Introduction
Tissue engineering is a multidisciplinary field that brings together researchers with backgrounds in engineering, biology, medicine, and chemistry to build tissue-like constructs for patient treatment or research. The ultimate goal of many research efforts in tissue engineering is to create biological replacements for diseased and damaged organs in the human body. Such constructs should mimic the physiological environment including the structural and physicochemical features of native tissues.1 Therefore, fabrication tools that allow for the creation of biocompatible complex 3D structures with controlled internal architecture and cell distribution and an effective vascular network are required.
Fiber-based techniques, which include textile technologies (i.e. weaving, braiding, knitting, embroidering), electrospinning, and direct writing, hold great promise for engineering 3D biomimetic tissue-like constructs. These techniques enable tuning the mechanical and structural properties of the fabricated constructs with interconnected pores and controlling the distribution of different cell lines in the constructs.3 Creating biopolymeric fibers with topographical properties that vary spatiotemporally on the micro- or nanoscale is the initial step for any fiber-based tissue engineering approach. In addition, fibers can serve as carriers for biomolecules and microorganisms. The biological and mechanical properties of the fabricated fibers are essential for the functionality of the resultant tissue constructs.2 Surface topology of the fibers also plays an important role in directing cell growth within the tissue construct.
Recent developments in microtechnologies along with the fast pace of growth of biopolymer science have allowed for the fabrication of fibers with amenable biomechanical properties for tissue engineering. In this chapter, we describe fiber fabrication techniques used in tissue engineering while emphasizing the role of microfluidics and microtechnologies. We categorize current fiber formation techniques into four methods: i) co-axial flow systems, ii) wetspinning, iii) meltspinning (extrusion), and iv) electrospinning. These methods are popular and have been enhanced by microtechnologies. We discuss the operational principles of these techniques and explore their advantages and limitations in tissue engineering.
1.2 Fiber Formation Techniques
1.2.1 Co-axial Flow Systems
Co-axial flow in microsystems is achieved by creating two or more flow streams in parallel. Due to the laminar nature of the flow in micro-channels, the interface of fluids remains stable and mixing only occurs due to molecular diffusion across the interface between the fluids. As a result, fibers with uniform cross-section can be fabricated. Co-axial flow-based microfluidic systems have been recently used for creating micron-size fibers featuring different shapes and sizes and containing different cell types and chemicals. This section describes the principle and theory of co-axial fiber fabrication and explores the current state-of-the-art in creating hydrogel fibers using microfluidic systems.
The fabrication of single layer hydrogel fibers in a co-axial flow format is shown in Figure 1.1a. The microfluidic system contains a central channel that delivers a pre-polymer solution (core) into a main channel. The delivered solution from two side-channels forms a sheath flow around the core stream. Polymerization of the core solution (hydrogel formation) occurs downstream of the flow either by cross-linkers directly from the neighboring fluids or by light irradiation. The core solution can be loaded with cells or chemicals for different biomedical applications.4 The sheath flow acts as a lubricant and facilitates fiber formation by preventing channel clogging during the hydrogel formation. Moreover, due to the short length of micro-channels containing the co-axial flow, cells are only exposed to a high shear stress and cross-linking reagents for a short time; this property helps the formation of hydrogel fibers containing viable and functional cells.
Fiber dimensions can be tuned by changing the ratio between the core and sheath flow rates and their relative viscosities. The fiber diameter, when the sheath and core viscosities are identical, is obtained from the following relationship:5
where Dfiber is the fiber diameter, Dchannel is the main channel diameter, and Qsh and Qcore are the sheath and core flow rates, respectively. Jeong et al. for the first time fabricated fibers using co-axial flow in a microfluidic system.5 Their microfluidic device was similar to the schematic shown in Figure 1.1a and comprised of a pulled glass capillary inserted in a polydimethylsiloxane (PDMS) substrate with feeding tubes, which were connected to syringe pumps. They used a photopolymerizable pre-polymer (4-hydroxybutyl acrylate (4-HBA)) as the core fluid and a mixture of 50% (v/v) polyvinyl alcohol (PVA) and 50% (v/v) deionized water (DI). They exposed the outlet channel to ultraviolet (UV) light in order to photopolymerize the core solution “on-the-fly”. They showed that eqn (1) can be used for predicting the diameter of the fabricated fibers within ±8%. In an attempt to create a glucose sensing microfiber, they mixed two enzymes, i.e. glucose oxidase (GOX) and horseradish peroxidase (HRP) in the core solution. Fibers containing enzymes responded to the glucose existing in the solution by emitting a fluorescent signal. No response was detected in the fibers without enzymes.5
In another study, Hwang et al. used a similar microfluidic device and created microfibers from poly(lactic-co-glycolic acid) (PLGA) to investigate the effects of the microfibers’ diameter on the orientation of mouse fibroblasts of L929 cells.6 The core solution was 10% (w/v) PLGA dissolved in dimethyl sulfoxide (DMSO) and the sheath solution was a mixture of 50% (v/v) glycerine in water. The exchange of DMSO and water at the interface between the core and sheath solution solidified the PLGA and fibers in the range of 10–242 μm were collected at the device outlet on a motorized rotating glass slide. Hwang et al. showed that cells tend to orient themselves along the long axis of these fibers more as the fiber diameter decreases.6
Shin et al. created alginate fibers using a microfluidic device, schematically shown in Figure 1.1a.7 They used sodium alginate as the core and calcium chloride (CaCl2) as the sheath solutions. Consequently, Ca2+ ions diffused from CaCl2 into the alginate central stream along the flow direction, forming calcium alginate cell-laden fibers before exiting the microfluidic device. They showed that for a constant sheath flow rate, the fiber diameters increased as the core flow rate was increased. However, as the core flow rate increased, instability in the flow occurred and spiral curls were formed. To assess the potential use of the alginate fibers fabricated by co-axial flow configuration, Shin et al. loaded the fibers with protein and mammalian cells by mixing bovine serum albumin (BSA) and human fibroblast cells (L292) in sodium alginate, respectively.7 The in vitro cell viability assay confirmed that the fabrication process was not harmful to the cells.
Inspired by the work of Shin et al.,7 Ghorbanian et al.8 developed a microfluidic direct writer (MFDW) to construct 3D cell-laden alginate structures containing interconnected pores. The MFDW was mounted on a motorized stage and was automatically controlled and moved at a speed synchronized with the speed of fiber fabrication. To avoid channel blockage, they designed a declogging mechanism that injected a degelling agent (e.g. ethylenediaminetetraacetic acid) to dissolve the clogged gel. They formed a simple 3D construct by layer-by-layer deposition of these cell-laden alginate fibers on a glass slide. Using a standard live/dead assay, they showed that the writing process was not harmful to mammalian cells.
To improve cell adhesion properties of alginate for tissue engineering applications, Lee et al. used a co-axial flow microfluidic system to create chitosan-alginate composite fibers.9 They used water-soluble chitosan and sodium alginate mixture as the core and CaCl2 as the sheath streams. It was found that the bi-component fibers offer a superior cell viability over pure alginate fibers, evidenced by their live/dead assay results for human hepatocellular carcinoma (HepG2) cells over 7 days of incubation.9 Although water-soluble chitosan is mechanically weak, the mechanical strength of the composite fibers did not change significantly.
Multiple-layer fibers can be fabricated by adding more streams to the microfluidic device. For example, Figure 1.1b shows the process of creating a two-layer composite fiber. First pre-polymer (core 1) surrounded by the second pre-polymer (core 2) enters the main channel and a sheath solution is formed around them. Cross-linking of the pre-polymers occurs downstream of the main channel either chemically or optically. The diameter of the created two-layer composite fiber can be estimated using the following relationship:10
where Qcore1and Qcore2 are the volumetric flow rates of the first and second pre-polymer solutions, respectively.
Lee et al. fabricated a microfluidic device, similar to the schematic shown in Figure 1.1b, to create hollow alginate fibers.11 They used CaCl2 as core 1, sodium alginate as core 2, and another stream of CaCl2 as the sheath stream. They encapsulated HIVE-25 cells in the hollow fibers and implanted them in a composite hydrogel (mixture of agar, gelatin, and fibronectin).11 The composite hydrogel was loaded with human smooth muscle cells (HIVS-125) to closely mimic a tissue with a stable microvessel network.
In another study, Hu et al. devised a triple-orifice spinneret to create co-axial flow of three different types of hydrogels.10 They used a wide range of hydrogel materials including enzymatically cross-linking gelatin-hydroxyphenylpropionic acid (Gtn-HPA), alginate, poly-(N-isopropyl acrylamide) (poly(NIPAAM)), and polysulfone. With their triple-orifice spinneret, Hu et al. fabricated hollow and multilayer composite cell-laden fibers. The ability to change the flow rate of each stream during the fiber fabrication process enabled them to change the total fiber diameter and thickness of each layer “on the fly”.10
The morphology of the fabricated fibers can be determined by changing the cross-sectional shape of the main channel. For example, Kang et al. used a grooved round channel to fabricate artificial tubuliform fibers with grooves on their surfaces (Figure 1.1c).4 They showed that the number of grooves and their sizes could be tuned by changing the shape of the channel and adjusting the flow rates. The grooved fibers were then used to investigate the effect of mechanical cues on rat embryonic neurons (Figure 1.1d). It was shown that the neural cells aligned themselves along the ridges of the nanogrooved fibers, indicating that the surface morphology of fibers played an important role on the cellular behaviour of the neurons.
Hydrogel fibers that are heterogeneous in chemical and physical microstructure are of interest to better mimic the microenvironment of natural tissues. Yamada et al. fabricated a PDMS-based microfluidic system to continuously synthesize chemically and physically anisotropic calcium–alginate fibers.12 Their device comprised a main core stream (propylene glycol+alginate+cells) co-axed with three streams of propylene glycol+alginate, buffer solution, and CaCl2 solution. Cell-laden fibers were collected by a rotating roller that was partially dipped in a bath of CaCl2. They added polyglycolic acid (PGA) to the alginate solution to adjust the stiffness of the local region of the alginate fibers and form sandwich-type solid-soft-solid structures. They showed that these structures provided better control of the cellular proliferation and networking. Inspired by the silk-spinning process in spiders, Kang et al. developed a microfluidic chip consisting of six hydrogel streams and one sheath flow (Figure 1.2a).4 They used sodium alginate solutions loaded with different chemicals and cell types and CaCl2 as the curing agent. The composition of the fabricated fibers was controlled using pneumatic valves. With this configuration, they controlled the spatial topography and chemical composition along the fibers. For example, they coded their fibers in serial (Figure 1.2b) and parallel (Figure 1.2c) configurations. They used the parallel coding feature to co-culture hepatocytes and fibroblasts on a single fiber. This shows the robustness of using microfluidic platforms for fabrication of cell-laden fibers for tissue engineering applications.
Fiber fabrication using co-axial flow in microfluidic systems holds great promise as it enables continuous fabrication of fibers with tunable morphological, structural, and chemical features. Various hydrogels including chemically and optically cross-linkable materials can be used in the co-axial format. Moreover, the incorporation of cells and chemicals in single- and multilayer fibers during the manufacturing process is possible. However, the biopolymers that are currently in use cannot form mechanically strong fibers to be easily manipulated.2 Furthermore, the process of fiber fabrication is relatively slow, which makes creating 3D structures a time-consuming process.
1.3 Wetspinning
In wetspinning, polymer fibers are formed by continuously injecting a pre-polymer solution into one or multiple coagulation baths to polymerize and form long fibers. The system is usually simple and includes a reservoir for the pre-polymer, a spinneret, and a coagulation bath (Figure 1.3a). Pre-polymers can be injected manually,13,14 gravitationally,15 using a syringe pump,16,17 or using pressurized air.18–20 To achieve chemically and mechanically stable fibers, the pre-polymer and the final polymer should be insoluble in the coagulation solution.16 Fibers fabricated by this method can be deposited randomly or in a predefined pattern in the coagulation bath to form a porous scaffold or can be collected on a motorized spool and assembled in a consequent process. Wetspinning has been widely used for fabrication of fibers from various biocompatible materials including alginate, collagen-alginate composite,9 collagen,13 chitosan,21 poly ε-caprolactone (PCL),16 starch-PCL composite,22 chitosan-tripolyphosphate composite,23 and calcium phosphate cement-alginate composite.9 Fiber diameter can be adjusted by controlling the polymer composition and viscosity, the injection flow rate, and the diameter of the spinneret.9,24 In general, wetspun fibers have been fabricated with a wide a range of diameters from ∼30 to 600 μm.25 In addition, the velocity of the pre-polymer jet entering the coagulation bath and the relative viscosity of both solutions should be optimized to prevent instability in the fiber (polymer) stream, which can affect the quality of the fabricated fibers.
For example, Tuzlakoglu et al. created chitosan fibers using the wetspinning process for bone tissue engineering.21 They dissolved chitosan in acetic acid and injected the solution in a bath of 30% 0.5 M Na2SO4, 10% 1 MNaOH, and 60% water and created randomly deposited fibers.21 They kept fibers in the coagulation bath for 1 day to fully cross-link. Tuzlakoglu et al. seeded the scaffolds with osteoblast-like cells, which proliferated (Figure 1.3b,c). In another work, Neves et al. created chitosan/PCL composite fibers by injecting the pre-polymer solution into a methanol bath, which formed fibers.27 They formed randomly deposited 3D scaffolds with porosity in the range of 64% to 83%. The scaffolds were seeded with bovine chondrocytes for cartilage tissue engineering. They showed that the presence of PCL resulted in rough fiber surface, which led to better cell attachment. Pati et al. formed fibers by adding chitosan to two different coagulation baths containing sodium triphosphate and NaOH to form chitosan-TPP and regular chitosan fibers, respectively.23 Fibers were deposited over each other to form a scaffold with porosities up to 89%. Chitosan-TPP scaffolds had a higher degradation rate than chitosan scaffolds. Chitosan-TPP scaffolds also offered a better cell viability and a faster degradation rate.
DeRosa et al. cross-linked type I collagen solution in acetic acid by a bath of acetone and ammonium hydroxide at a pH of 9.28 They coated the fibers with poly-l-lysine (PLL) and then seeded with Schwann cells for neural tissue engineering. Cell attachment and proliferation was much higher on the fiber with surface treatment. Puppi et al. used wetspinning and cross-linked star PCL using an ethanol bath.16 The fabricated fibers were porous and had diameters in the range of 100–300 µm. The fibers were randomly deposited to form a 3D structure, which was then seeded with murine pre-osteoblasts for bone tissue engineering. They also encapsulated levofloxacin and enrofloxacin within the fibers as antimicrobial drugs.
Landers et al. injected a mixture of sodium alginate 5% (w/v) and gelatin 1% (w/v) into a CaCl2 bath to form solid fibers.26 They also fabricated alginate-fibrin fibers by a adding mixture of alginic acid with fibrinogen to a solution of CaCl2, thrombin, and sodium chloride. These fibers were deposited using a bioplotter to form a 3D fibrous mesh with a predefined fiber arrangement (Figure 1.3d). They seeded the scaffolds with mouse connective tissue fibroblasts. Lee et al. cross-linked α-tricalcium phosphate/alginate fibers with CaCl2.25 By changing the needle tip size they fabricated fibers with diameters in the range of 200–600 µm. The fibers were seeded with mesenchymal stem cells (MSCs) for bone tissue engineering. The scaffolds supported the differentiation of MSCs along the osteogenic lineage.
To improve the mechanical properties of wetspun hydrogel fibers carbon nanotubes (CNTs) or graphene oxide have been added to the pre-polymer solution. He et al. injected sodium alginate/graphene oxide (GO) into a bath of CaCl2.29 The maximum tensile strength and Young's modulus were increased by several fold at 4 wt% GO loading (from 0.32 and 1.9 to 0.62 and 4.3 GPa, respectively). Fibers were seeded with rabbit cartilage cells and showed a good attachment and proliferation. Spinks et al. added CNT to wetspun chitosan fibers to improve their mechanical properties.17 They showed that the Young's moduli of fibers were improved from 4.25 GPa for pure chitosan fibers to 10.25 GPa for those containing 5% CNT. CNTs can also enhance the electrical conductivity of hydrogel fibers. Electrically conductive fibers can help stimulation of cells such as cardiomyocytes. MacDonald et al. showed that the electrical conductivity of fibroblast-laden collagen increases from 3 Ms cm−1 to 7 mS cm−1 upon addition of 4% CNTs.
Scaffolds can be made by wetspinning fibers and randomly depositing them on a substrate,9,16,22,23,25 by rolling them up,26 or by patterning the fibers in an ordered arrangement.26 Since the fibers fabricated with wetspinning are relatively thick, the pore size of the formed scaffolds is large (∼250–500 μm).27 The porosity of scaffolds fabricated by wetspinning ranges from 15% to 92%.23 Lee et al. randomly deposited wetspun α-tricalcium phosphate/alginate fibers and created scaffolds with low (13.6%), medium (34.0%), and high (53.7%) porosities.25 They showed that the mechanical properties of the fabricated scaffolds increased by decreasing their porosities. For example, the elastic moduli of the scaffolds with high, medium, and low porosities were 96±63 MPa, 398±63 MPa, and 573±87 MPa, respectively. These values fall in the biological range for trabecular bone (50–500 MPa). Reducing the porosity, however, increases the diffusional mass transfer resistance and reduces cell infiltration to the inner parts of the scaffold. Pati et al. deposited chitosan-TPP fibers to form 3D scaffolds with a high porosity (up to 89%) and seeded them with 3T3 fibroblasts. Lander et al. patterned wetspun alginate/fibrin fibers to form 3D structures (Figure 1.3d). They also showed that the porosity of the fabricated scaffolds can be determined by the following relationship:26
where d1 is the fiber diameter, d2 and d3 are the spacing between adjacent parallel fibers in horizontal and vertical directions.
Incorporation of cells within the fibers is possible using wetspinning. However, long exposure of the materials to cross-linking reagents, which are usually not cell-friendly, can limit the fabrication of cell-laden fibers from some materials. The most common material for fabrication of cell-laden fibers is sodium alginate because the pre-polymer solution and the gelation agent (calcium chloride) are both compatible with cells. For example, Arumuganathar et al. formed cell-laden fibers using a three-needle pressure-assisted system to fabricate multi-compositional structures that carried living cells in an inner layer.30 The encapsulating medium flowed in the outer needle and provided a sheath for the suspension layer.30–32
In general, wetspinning is a simple method for fabrication of many natural and synthetic polymer fibers. As a multiple spinneret can be used this technique is scalable and can be used for industrial fabrication of fibers. However, the long time required for cross-linking of wetspun fibers has prevented fabrication of cell-laden fibers from a wide range of polymers.
1.4 Meltspinning (Extrusion)
Meltspinning or extrusion is the process of creating fibers by injecting a melted polymer through a micron size spinneret to form continuous fiber strands (Figure 1.4). The fiber strands are cooled after spinning on cooled drums (cold-drawn). Another heating step can be added to the process (hot-drawn) to improve the fiber properties. Meltspun fibers exhibit relatively high mechanical properties. Yuan et al. showed that the tensile strength of meltspun poly(l-lactic acid) (PLLA) fibers was in the range of 300–600 MPa.33 They also showed that hot drawing of fibers significantly affects the mechanical properties of meltspun fibers. For example, Young's modulus of hot-drawn PLLA fibers (3.6–5.4 GPa) was significantly higher than their cold-drawn counterparts. In another study, Gomes et al. used meltspun fibers from starch with ε-polycaprolactone (SPCL) and starch with poly(lactic acid) (SPLA).34 They created scaffolds by a fiber bonding process that includes bonding of meltspun fibers in an additional annealing process. They showed that fabricated scaffolds exhibited compression modulus in the range of 1.8 MPa–9.61 MPa.
The cross-section of meltspun fibers and their texture can be tailored by using spinnerets with predefined cross-sections. This feature is particularly beneficial for applications where the fiber texture is important for orientating the cells. Lu et al. used this capability to create grooved fibers for mechanical guidance of rat skin fibroblasts (RSFs) and rat aortic smooth muscle cells (RASMCs).35 Poly(lactic acid) (PLA) and polyethylene terephtalate (PET) grooved fibers were fabricated using the process shown in Figure 1.4. They used a spinneret with grooves 5–15 μm deep and 10 μm wide to fabricate fibers with the nominal diameter of 50 μm. They showed that cultured cells attached and extended their cytoplasmic lamellapodia within the grooves after 4 weeks of incubation. Moreover, the cells aligned themselves parallel to the direction of the grooves.
Sinclair et al. showed that the dimension of the grooves played an important role in cell alignment.36 They produced meltspun PET grooved fibers with different groove widths (6–53 μm) and depths (6–35 μm). Normal human dermal fibroblast (NHDF) cells were cultured on the fibers for a period of 2 weeks. Their results indicated that all grooved fibers were capable of supporting cellular attachment and proliferation, nuclear elongation parallel to the fiber's longitudinal axis, and aligned matrix synthesis. However, when the width of the grooves was smaller than the cells size, cells were unable to fit into the grooves. As a result, cells flattened on the top of the groove and extended parallel to the fiber's longitudinal axis.
Fabrication of hollow fibers with meltspinning was also demonstrated. Hinüber et al. used a special spinneret to create hollow fibers from poly(3-hydroxybutyrate) (PHB).37 They showed that meltspinning at low process temperatures without additives led to the formation of well-defined hollow PHB fibers. These hollow fibers can be used to deliver oxygen and nutrients to cells during the culturing period.
In an interesting study, Miller et al. combined meltspinning and direct writing processes to create 3D filament networks of carbohydrate glasses.38 They used this cytocompatible network as a sacrificial template for creating 3D vascular networks in tissue constructs. Their device was comprised of a glass syringe and steel nozzles. The syringe was mounted on a custom-modified RepRap Mendel 3D printer with associated electronics. They heated carbohydrate glass up to 100 °C and printed the molten glass under nitrogen pressure with pneumatic control to create a sacrificial network. After pouring a suspension of cells in ECM pre-polymer on the network and cross-linking, the carbohydrate glass filaments were dissolved to form vessels while their interfilament fusions become intervessel junctions. They showed that the perfused vascular network created by this method sustained the metabolic function of primary rat hepatocytes in engineered tissue constructs.
Meltspinning is a relatively fast process that allows the high-throughput fabrication of micron size fibers with irregular cross-sections. For example, Yuan et al. fabricated 10 meters of PLLA fibers with diameter of 110–160 μm in 10 minutes (1 m min−1) and Gomes et al. created 100 meters of SPCL and SPLA with diameter 120–500 μm in 5 minutes (20 m min−1).33,34 However, meltspinning is a sophisticated method that requires high temperatures in the range of 150 °C to 295 °C.34,36 These high temperatures prevent the encapsulation of cells, proteins, and temperature-sensitive chemicals in the fibers. Moreover, the melted polymers are usually highly viscous (e.g. viscosity of PLA at melting point is 104 times higher than water39 ), therefore high pressures are required to push the polymer through the spinneret.40 Mass loss, rapid decrease of viscosity during the process, and thermal degradation of some materials are the other main challenges of the meltspinning process that often have considerable influence on the mechanical properties of the resultant fibers.41 In general, fibers larger than a few micrometers can be fabricated with meltspinning.
1.5 Electrospinning
Electrospinning is a fiber fabrication technique that uses an electrical field to draw a viscoelastic polymer from a spinneret and deposit the fibers on a collector plate.42,43 The essential components of an electrospinning device include a viscous polymer solution, a pumping system (injector) with a metallic tip, and a collector plate at a distance from the tip (Figure 1.5a). As a result of the applied electrical field, the polymer stream breaks into smaller branches, which in turn form fibers that deposit on the collector surface (Figure 1.5). The properties of the polymer, the applied voltage, the distance from the tip of the injector to the collector plate, and the properties of the collector plate could affect the microstructure of the final fibrous construct. Depending on the polymer physical properties (i.e. viscosity, surface tension, and electrical conductivity), fibers with diameters in the range of a few nanometers to several micrometers can be created.44 Fibrous scaffolds can be created by random deposition of fibers on a collector plate. Fibers can also be collected on a rotating collector to form mats with aligned fibers. The relative simplicity of the method as well as the ease of controlling the key process parameters such as flow rate and voltage are the main reasons for the popularity of this technique.43
Electrospinning has been widely used for tissue engineering applications as scaffolds fabricated by this method mimic the microstructure of the extracellular matrix of native tissues. Electrospun scaffolds can be fabricated from natural and synthetic polymers including silk fibroin, chitosan, collagen, gelatin, hyaluronic acid, fibrinogen, poly(ester urethane) urea (PEUU), poly(glycerol sebacate) (PGS), PCL, polyurethane (PU), PLGA, and PGA44 for applications such as cardiac graft, wound dressing, and cartilage and bone tissue engineering.3
Selection of the proper solvent to dissolve the polymer prior to electrospinning is essential. The solvent should be volatile to evaporate before the fibers hit the collector plate. Moreover, the solution should be viscoelastic. For example, it has been observed that if the solution is not viscous enough, fibers do not form and particles would form instead. Some of the solvents that have been used include 1,1,1,3,3,3-hexafluoro-2-propanol for collagen PEUU, and silk fibroin, tetrahydrofolate for chitosan, anhydrous chloroform-ethanol mixture for PCL and PGS, and water for gelatin and hyaluronic acid.
The applied voltage affects the diameter of the formed fiber and the microstructure of the electrospun scaffolds.48 Low applied voltage cannot initiate the electrospinning process. As the applied electric field (larger than the threshold) increases, the polymer stream breaks into more branches and the fiber diameter is reduced initially. Sant et al. fabricated electrospun PGS-PCL sheets for cardiac tissue engineering (Figure 1.5c). They observed that at 2:1 PGS:PCL ratio, a minimum fiber diameter was achieved by applying a 17.5 kV voltage. Their results showed that the fiber diameter decreased from 4 µm at 12.5 kV to 3 µm at 17.5 kV and then increased to 8 µm at 20 kV.46
Injection flow rate is another important parameter that affects the quality and geometry of the fabricated fibers. High flow rates (injection velocities) lead to wet and non-polymerized fibers as the solvent does not have enough time to evaporate. The unevaporated solvent could be harmful to cells. In general, fiber diameter has a direct relationship with the volumetric flow rate. The distance from the needle tip to the collector plate is also important as it should be large enough that the solvents evaporate prior to fiber attachment to the collector plate.
In regular electrospinning on aluminum foil collectors the fibers randomly distribute and orient in the scaffold (Figure 1.5a). Other types of collectors include rotating disk or mandrel, pin, mesh, and parallel wires. Also, it is possible to collect the fibers in a coagulation bath to cross-link the fibers. Rotating disks have been used to control the alignment of the fibers.49 Yang et al. electrospun PLLA fibers on a disk with a rotation of 1000 rpm and a flow rate of 1 mL hr−1 and an applied voltage of 12 kV.45 By changing the concentration of the polymer they fabricated aligned fibers with diameter in the rage of 150–3000 nm (Figure 1.5b). In another notable study, Amoroso et al. electrospun PEUU fibers on a mandrel with rotation speed of 266 rpm and rastering speed of 0.3–30 cm s−1. Simultaneously, they electrosprayed cell culture medium and cells on the scaffold during the fabrication process to cut down the time for penetration of cells between various layers of fibers. The mechanical properties of the fabricated electrospun sheets were comparable to those of native porcine pulmonary valve.
In spite of all the progresses made in electrospinning, manufacturing of thick 3D complex scaffolds is difficult.50 In addition, the high fiber packing density and the small pore sizes (∼10–15 μm) limits cell infiltration into the scaffolds.51 The latter issue has triggered an array of techniques including salt/polymer leaching,52 wet electrospinning using a bath collector53 or an ice crystal collector,54 and laser/UV irradiation55 to produce electrospun scaffolds with both large pores and high porosity.
Encapsulation of living cells in electrospun fibers is a difficult process as: i) the employed solvents are usually cytotoxic; and ii) the high electrical field could damage the cells. Despite these challenges fabrication of cell-laden fibers though electrospinning has been demonstrated in several studies47,56,57 (Figure 1.5d). In these studies, a co-axial needle configuration was used, in which each needle was connected to a separate syringe pump. The central stream contained a cell and the sheath stream contained a polymer. The encapsulation of living cells in electrospun fibers faced many challenges as the diameter of the fibers is typically smaller than a single cell.56,57 Another shortcoming of the electrospinning of cell encapsulated constructs is the lack of control over cell distribution in the volume.
1.6 Conclusions
Fiber-based tissue engineering is an emerging area that holds a great promise for creating artificial organs for transplantation. Fibers are the building blocks of these engineered constructs. Owing to recent advancements of microtechnologies, fabrication of sophisticated fibers with engineered mechanical properties, topography, and composition are now possible. These fibers are usually formed using wetspinning, co-axial flow systems, meltspinning, and electrospinning. Here, we discussed various fiber fabrication techniques that utilize microtechnologies. Among fiber fabrication techniques, electrospinning has become popular for fabricating tissue scaffolds owing to its relative simplicity, similarity of the fabricated structures to native tissues, and ease of control over the key process parameters. However, the fabrication of cell-laden fibers with precise control over cell distribution using electrospinning is still a major challenge.
Wetspinning is probably the simplest method for fiber fabrication and wetspun fibrous structures offer a tunable porosity, but larger pore sizes in comparison to electrospun structures that make them more amenable for cell penetration. However, long exposure of the fibers to harmful cross-linking reagents has limited fabrication of cell-laden fibers to alginate fibers. In addition, due to the large size of the fabricated fibers the wetspun fibers cannot mimic the extracellular matrix.
Co-axial systems for fiber spinning that encompass those using a microfluidic chip can fabricate fibers with tunable morphological, structural, and chemical features, continuously. Moreover the size of fibers fabricated by co-axial systems is usually smaller than wetspinning. The main drawback of these systems is their reliability as they are vulnerable to clogging during the manufacturing process.
Meltspinning facilitates the fabrication of fibers with irregular texture and cross-section; however, the high temperature required during the fabrication process prevents encapsulation of cells within the fibers. Meltspun fibers have the potential to be used sacrificially for forming a vasculature network within hydrogel constructs.
Financial support from NSERC, CIHR, CHRP, CFI, Genome Canada, and Genome Quebec is gratefully acknowledged. M.A and A.T. acknowledge NSERC postdoctoral fellowships. D.J. acknowledges support from a Canada Research Chair. The authors declare no conflict of interests in this work. A.K. acknowledges funding from the National Science Foundation CAREER Award (DMR 0847287), the office of Naval Research Young National Investigator Award, and the National Institutes of Health (HL092836, DE019024, EB012597, AR057837, DE021468, HL099073, EB008392).
Both authors have contributed equally to the work.