Vol. 12
TISSUE ENGINEERING
261
TISSUE ENGINEERING Introduction Accidents and diseases lead to devastating tissue losses and organ failures, which result in more than 8 million surgical operations each year in the United States alone (1). These problems convert to a national annual healthcare cost of approximately half trillion U.S. dollars. The state-of-the-art clinical therapies to tissue losses and organ failures can be categorized into three approaches, ie, transplantation, surgical reconstruction, and the use of prostheses. Each of these approaches has contributed to solving or alleviating the severity of these clinical problems, while all of them have serious limitations. Organ transplantation became successful in the early 1960s because of the success of immunologic suppression in the clinical setting (2). Transplantation has saved, and is continuing to save, countless lives. This approach, however, is severely limited by the dearth of donor organs. For example, fewer than 3000 available donors are way shorter than the needs of approximately 30,000 Americans for liver transplants each year (3). Surgical reconstruction utilizes tissues harvested from the patient to rebuild a critically needed body part. The use of the patient’s own tissue is advantageous in that it has a higher success rate resulting from the avoidance of immune rejection than using tissues from other sources. However, the need for second site of surgery, limited supply, inadequate size and shape, and the morbidity associated with donor site are all major concerns (4,5). In response to the shortages of needed tissues and organs, prostheses are developed to replace certain body parts for their structural and mechanical functions. There are approximately 100,000 people in the United States with transplants, while there are more than 10,000,000 with biomedical implants (6). However, the prostheses are made from artificial materials, are not biologically functional, and are therefore subject to long-term complications and rejections. Tissue engineering is a new approach to resolve the missing tissue and organ problems. Tissue engineering has been defined as an interdisciplinary field that applies the principles of engineering and the life sciences toward the development of biological substitutes that restore, maintain, or improve tissue function (3). There are three strategies in tissue engineering (3,7): (1) The use of isolated cells or cell substitutes to replace those cells that supply the needed function, including genetic or other manipulations before the cell infusion (8); (2) The delivery of tissue-inducing substances, such as growth and differentiation factors, to targeted locations (9,10); (3) Growing cells in three-dimensional (3-D) matrices (scaffolds) or devices, where cells can be either recruited from the host tissues in vivo or seeded (encapsulated) in vitro (3). The advantage of the use of isolated cells is the simplicity. Cells are often directly injected into the targeted locations to avoid complex procedures and associated complications. The disadvantages include cell death and loss of function Encyclopedia of Polymer Science and Technology. Copyright John Wiley & Sons, Inc. All rights reserved.
262
TISSUE ENGINEERING
Vol. 12
because cells need the right nutritional and substrate environment to grow and function. The success of using tissue-inducing substances is dependent on the economical large-scale production and purification of the bioactive molecules, and on the development of delivery systems that can deliver these molecules with the desired profiles. Certain gene therapy techniques can also be used for this approach. The above two approaches, ie, the use of isolated cells or tissue-inducing substances, are considered when the defects are very small and well contained. For engineering tissues of practical size scale and with predetermined 3-D structures, these two approaches are seriously limited. Therefore, the third approach, ie, growing cells in 3-D matrices (scaffolds) or devices, has become increasingly active. Some of these matrices or devices are used to develop immobilized cell systems, either as implants or extracorporeal devices (11,12). The core of this technology is semipermeable membranes or matrices that have well-defined molecular weight or size cutoff (13,14). They serve as immunoprotective barriers to support cell growth and function, which allow nutrients, metabolic products, and wastes to diffuse through, but not immune cells or antibodies. These devices offer certain biological functions but are not living tissue/organ replacements (see MEMBRANE TECHNOLOGY). Some other 3-D matrices work as templates (called scaffolds) to guide cells to grow, synthesize biological molecules and extracellular matrix components, and facilitate the organization and formation of functional tissues and organs (15–17). After fulfilling the templating function, the scaffolds degrade and disappear, leaving nothing foreign to the biological system (Fig. 1). There are many additional advantages in this approach. Patient-derived cells (stem cells or differentiated cells) or future universal cell sources (nonimmunogenic) can be used so that there will be minimum complication associated with immune rejections. These cells can be expanded in vitro to solve the donor shortage limitations. Any tissue/organ structure can be potentially mimicked by the scaffolding design. The engineered tissues will have the capacity of growing, modeling, and remodeling in concert with dynamic changes in physiological environment of the body. In this approach, biodegradable polymers (natural or synthetic) are the materials of choice. Polymers (or macromolecules) are currently used as scaffolds for nearly every tissue type including bone and other mineralized tissues. Besides polymers, only limited inorganic materials are used for certain mineralized tissue engineering research. Although growing cells on two-dimensional (2-D) substrates (such as petri dishes and culture plates) dates back centuries, designing 3-D scaffolds for tissue engineering is a new field, where polymer science and engineering play pivotal roles. There are a few basic requirements that have been widely accepted for designing polymer scaffolds. First, the scaffold has to have high porosity and proper pore size. These pores allow cell seeding and migration to achieve the needed relative uniform distribution. The pores also provide the space for cell proliferation and neo tissue deposition. The pores also satisfy the needed mass transport requirements for nutrients, signaling molecules, metabolic products, and wastes. The porous structure should also allow for vascularization and innervation for sustained function of the regenerated tissues when implanted in vivo. High porosity is also beneficial because of the reduced polymer amount and its degradation
Vol. 12
TISSUE ENGINEERING
263
Fig. 1. Schematic diagram showing the tissue engineering concept. Scaffolding materials (temporary synthetic extracellular matrices) are designed, on which mammalian cells can grow to regenerate the needed tissue or organ in three dimensions (3-D). Because the scaffolds are biodegradable, they will resorb after fulfilling the template function and leave nothing foreign in the body.
products in the neo tissues, which can provoke unwanted inflammatory response from the host. Second, a high surface area (a.k.a. surface-to-volume ratio) is needed, which is the 3-D surface area of the porous scaffolds (not just the outside surface area). Many cell types are anchorage-dependent, ie, they can survive, grow, and function only when they are attached to an appropriate substrate. The high surface area provides cells with the sufficient area to attach, grow, and deposit neo tissue components. Third, biodegradability is generally required, and a proper degradation rate is needed to match the neo tissue formation. If the scaffold degrades too fast, it can collapse before the new tissue is formed so that it fails to serve as a 3-D guidance for the neo tissue organization. If the scaffold degrades too slowly, it remains for a prolonged time period after the neo tissue is formed and stabilized. It may hinder the new tissue replacement of the scaffold space, and may cause complications associated with long-term foreign body reactions. Fourth, the scaffold must have the needed mechanical integrity to maintain the predesigned tissue structure to serve the 3-D guidance. However, scaffolds are often not as mechanically strong as the tissues to be replaced because of the required high porosity for scaffolds. They usually serve the scaffolding purpose well as long as they can maintain the structural integrity under the cultivation or implantation conditions, while the goals are that the engineered tissues from the scaffolds are mechanically and biologically functional as their natural counterparts. Fifth, the scaffold should not be toxic to the cells (biocompatible). In addition to the polymer, the degradation products of the polymer should not be toxic to the cells, which is usually a more restricting requirement than for the polymer. Sixth, ideally the scaffold should positively interact with cells, including enhanced cell adhesion, growth, migration, and differentiated function. To achieve these positive cell–scaffold interactions, surface or bulk modifications of the
264
TISSUE ENGINEERING
Vol. 12
polymer structure are often employed (18–21). Even drug, protein, or gene delivery techniques are often considered in the scaffold design (22–25).
Polymers in Tissue Engineering As discussed above, polymers play a pivotal role in tissue engineering. To fulfill the diverse needs in tissue engineering, various polymers have been exploited in tissue engineering research, including natural polymers (macromolecules), natural polymer-derived materials, synthetic polymers, and synthetic polymers made of natural monomers or modified with natural moieties. Various copolymers, polymer blends, or polymeric composite materials are also used. This section is not intended to be a complete and exhaustive review of all the polymers used in tissue engineering. Instead, some of the most frequently used polymers (macromolecules) in tissue engineering are briefly reviewed. Polymers for Porous Scaffolds. The polymers to be discussed in this category can form stable porous structures in the solid form to serve as pre-designed 3-D scaffolds. They generally do not dissolve or melt under in vitro tissue culture conditions (in an aqueous tissue culture medium) or when implanted in vivo. Linear Aliphatic Polyesters. Linear aliphatic polyesters are the most frequently used synthetic biodegradable polymers in tissue engineering and many other biomedical applications (26–28). These polymers degrade through hydrolysis of the ester bonds in the polymer backbone. The degradation rates and profiles differ between these polymers owing to their compositional, structural, and molecular weight differences. Polyglycolide, also called poly(glycolic acid) (PGA), polylactide, also called poly(lactic acid) (PLA), and their copolymers, poly(lactide-co-glycolide), also called poly(lactic acid-co-glycolic acid) (PLGA), are a family of linear aliphatic polyesters called poly(α-hydroxy acids) or poly(α-hydroxy esters) (Fig. 2). These polymers can be synthesized by direct condensation of the hydroxy acid monomers (resulting in low molecular weight polymers, such as lower than 10,000) or more commonly by a ring-opening polymerization of the cyclic dimers (to achieve a higher molecular weight), from where the names of polyglycolide and polylactide stem. Among the family of glycolic acid and lactic acid homopolymers and copolymers, PGA is the simplest in chemical structure, has many advantageous properties, and is therefore one of the most widely used scaffolding polymers (17). Because of the chain structural regularity, PGA is highly crystalline and has a high melting point of around 220◦ C (17). It does not dissolve in most common organic solvents. Because of its hydrophilic nature, PGA degrades rapidly in aqueous solutions or in vivo, and loses mechanical integrity between 2 and 4 weeks depending on the molecular weight and physical structure of the scaffolds or implants, and in vitro or in vivo conditions (17,29). It was the material used to develop the first synthetic absorbable suture, and has been processed into nonwoven fibrous fabrics as one of the most widely used scaffolds in tissue engineering today. PLA is also widely used for scaffold fabrication because of its biodegradability. Because of the extra methyl group in PLA repeating unit in comparison to PGA, PLA is more hydrophobic. The hydrophobic methyl group reduces the molecular affinity to water and leads to a slower hydrolysis rate of PLA.
Fig. 2. Polymers frequently used as scaffolds for tissue engineering. 265
266
TISSUE ENGINEERING
Vol. 12
It takes months to years for a PLA scaffold or implant to lose mechanical integrity in vitro or in vivo (30,31). Lactic acid is a chiral molecule and exists in two stereoisomeric forms, ie, the D-lactic acid and L-lactic acid. The two stereoisomers lead to four different types of PLA: poly(D-lactic acid) (PDLA), poly(L-lactic acid) (PLLA), racemic poly(D,L-lactic acid) (PDLLA) and meso poly(D,L-lactic acid) (meso-PLA). Because of the stereoregularity, PDLA and PLLA are semicrystalline, and are less susceptible to water attack (hydrolysis), resulting in slower degradation rates. The racemic copolymer PDLLA is amorphous, and therefore more susceptible to hydrolysis, resulting in a faster degradation rate. Since L-lactic acid is the natural stereoisomer of the lactic acid in the body, PLLA is the frequently used PLA. To achieve intermediate degradation rates between PGA and PLLA, various lactic acid and glycolic acid ratios are used to synthesize PLGAs. However, there is not a linear relationship between degradation rate and lactic acid content. When lactide is copolymerized with glycolide, two distinctly different effects with regard to degradation are introduced. The first is that the lactic acid unit is more hydrophobic, contributing to a slower degradation rate. The second effect is that stereoregularity is disrupted by the different repeating units, resulting in a lower crystallinity or amorphous structure, contributing to a faster degradation rate. The resulting degradation rate is a compromise of these two factors. Theoretically, when the LA/GA ratio is higher than 50:50, these two factors are in the same direction when the amount of lactide is increased. When the LA/GA ratio is lower than 50:50, these two factors are against each other when the amount of lactide is increased. However, in a large middle range of the composition, the copolymer is amorphous and the crystallinity does not play a role. In both high and low ends of the LA/GA ratio, the crystallinity of the specific scaffold or device plays an important role on the degradation rate. In addition, high crystallinity increases modulus, yield strength, and ultimate strength of the polymers. In addition to the biodegradability and biocompatibility, these polymers (PLA, PGA, and PLGAs) are among the few synthetic polymers approved by the Food and Drug Administration (FDA) for certain human clinical applications such as surgical sutures and some implantable devices. There is, however, some controversy around the use of these polymers for orthopedic applications. Based on a study of more than 2500 patients operated on using pins, rods, bolts, and screws made of PGA or PLA, 4.3% were affected by a clinically significant local inflammatory tissue reaction (32). Similarly, other biodegradable polyesters suited for manufacturing absorbable fixation implants appear to also elicit adverse tissue responses. These adverse tissue reactions are generally considered to result from the released acid during degradation. For polymer scaffolds with extremely high porosity used in tissue engineering applications (usually ≥90%) as opposed to solid implant devices used in orthopedics, the acid release (pH variation) is a less severe problem (17,30). Polymer degradation is often categorized into bulk degradation (bulk erosion) and surface degradation (surface erosion) in the fields of biomedical engineering and biotechnology (33,34). There are actually differences between erosion and degradation. Degradation is the bond cleavage of the polymer, resulting in reduction in molecular weight of the polymer. Erosion is the mass loss process. For ideal bulk degradation, hydrolysis (or other forms of degradation) occurs in the entire polymer volume simultaneously. The length of time that polymer persists
Vol. 12
TISSUE ENGINEERING
267
can be altered by changing chemical composition but not by the material’s size and shape. For ideal surface degradation, hydrolysis (or other forms of degradation) only occurs on the external surface of the material, ie, at the interface between the material and its immediate surrounding environments, such as medium or body fluids. Erosion rate is directly proportional to the external surface area. The length of time that polymer persists can be altered by varying the sample volume (thickness). Surface erosion is often advantageous for controlled drug delivery applications. For example, a thin slab erodes by reducing its thickness and maintaining a nearly constant surface area to achieve constant release rate. PLA, PGA, and PLGA polymers degrade via bulk hydrolysis. However, the length of time that a polymer persists is affected by the shape and size of the material because the degradation does not occur homogeneously. Common sense would lead people to think that the surface layer degrades faster since this area is exposed to the most abundant medium or body fluids, which hydrolyze ester bonds in the polymer chains. The reality is just the opposite (35). The cleavage of ester bonds yields carboxyl and hydroxyl end groups. The produced carboxyl end groups are capable of catalyzing the hydrolysis of remaining ester bonds in the polymer chains. In the surface layer, the chains containing carboxyl end groups or the generated small acid molecules diffuse out of the device and the acidity is reduced. This autocatalysis phenomenon is enhanced by the diffusion limitations in the thick devices. In the interior, the acidity increases as ester bonds are hydrolyzed while the acidic by-products are trapped. The increased acidity further catalyzes hydrolysis, leading to heterogeneous degradation, ie, a faster degradation in the interior than on the surface. For semicrystalline polymers such as PGA and PLLA, the degradation behavior is also related to the crystallinity and domain structures of the polymers (17). The amorphous domains are easily attacked by water molecules and degrade first. Initial mass loss occurs in the amorphous domains and causes an apparent crystallinity increase (17). As the polymer further degrades, the crystalline domains also degrade, which leads to drastic decrease in melting point and mechanical properties, and finally leads to the disintegration of the entire material (see BIODEGRADABLE POLYMERS, MEDICAL APPLICATIONS). There are other linear aliphatic polyesters, such as poly(ε-caprolactone) (PCL) and poly(hydroxy butyrate) (PHB) (Fig. 2), which are also used in tissue engineering research and other biomedical applications. PCL has been investigated for a variety of biomedical applications. It can degrade by microorganisms, hydrolytic, enzymatic, or intracellular mechanisms under physiological conditions (36,37). PCL is a semicrystalline polymer. It has a very low glass-transition temperature of around −62◦ C, and thus is always in the rubbery state and has high material permeability under physiological conditions. For highly porous scaffolds, material permeability however does not contribute significantly to the scaffold permeability. Compared with PLA, PGA, and PLGA, PCL degrades at a significantly slower rate (38), which makes PCL less attractive for general tissue engineering applications but more attractive for long-term implants and controlled release applications. Poly(hydroxy butyrate) (PHB), also called poly(β-hydroxy butyrate) or poly(hydroxy butyric acid), is the simplest member of the polyhydroxyalkanoate (PHA) polyesters (see POLY(3-HYDROXYALKANOATES)). PHB is made by
268
TISSUE ENGINEERING
Vol. 12
microorganisms via fermentation. It is biocompatible and biodegradable, but is highly crystalline and therefore very brittle. PHB degrades very slowly (years in vivo) because of the high crystallinity and hydrophobicity (39,40). PHB-based copolymers are less crystalline and mechanically more flexible, but degrade also very slowly because of the hydrophobic nature, and therefore are less popular compared to PGA, PLA, and PLGA for tissue engineering applications. Other Important Synthetic Biodegradable Polymers. Poly(phosphoesters) (PPE) are a class of biodegradable polymers synthesized in the Leong group (Fig. 2) (41). PPE contains P(O) O C linkage in the polymer backbone and the pentavalent phosphorus atom allows for attachment of a pendant component. PPE degrades through hydrolysis of phosphoester bond in the PPE backbone. The variation in backbone structure and pendant side group allows for variation in structure and properties of these polymers. They have been used for controlled drug delivery and gene delivery applications (42), and are also explored for certain tissue engineering applications (43) (see CONTROLLED RELEASE TECHNOLOGY; GENE-DELIVERY POLYMERS). Poly(propylene fumarate) (PPF) (Fig. 2) has been synthesized as a biodegradable polymer that degrades through hydrolysis of the ester bonds similar to glycolide and lactide polymers. PPF is an amorphous unsaturated polymer and has been extensively studied in the Mikos group as either a prefabricated (44) or an injectable biomaterial (45) for bone tissue engineering. Its copolymers with poly(ethylene glycol) (PEG), poly(propylene fumarate-co-ethylene glycol) polymers, can form hydrogels (46). Polyphosphazenes (qv) contain a long-chain backbone of alternating phosphorus and nitrogen, with two side groups attached to each phosphorus (Fig. 2) (47). Various synthesis routes are used to prepare these polymers, allowing different side groups (R) to be attached to the backbone. These polymers can have very different properties, from degradable to nondegradable, by changes in the side groups. Polyphosphazenes are finding more and more applications in controlled drug delivery applications (48,49), and are also finding ways into the tissue engineering research (50,51). Although polyanhydrides (qv) are attempted for use as tissue engineering scaffold (52), they are designed primarily for controlled drug delivery purposes in the Langer group (53). These polymers are very hydrophobic, and degrade through surface erosion. The surface erosion characteristic allows easy accomplishment of constant release rate for sustained drug delivery purpose. Drugs are also well protected when embedded in such polymers owing to the nearly no water penetration before the polymer erodes. Similarly, poly(ortho esters) are primarily designed for controlled drug delivery applications because of their surface erosion properties (54). Nevertheless, they have been explored for tissue engineering scaffolding applications as well (55). Amino acids are natural nutrients. Polymers derived from amino acids may have advantageous biocompatibility, as demonstrated by PLA. However, amino acid based polymers often do not have good material properties. They are often difficult to process into 3-D structure, and possess poor mechanical properties. Tyrosine-derived polymers have been developed in the Kohn group as biodegradable polymers, which have shown certain promising properties (56), and have been used for bone tissue engineering research (57).
Vol. 12
TISSUE ENGINEERING
269
Nondegradable polyurethanes (qv) have been widely used for biomedical applications because of their elastic properties (58). The segmented polyurethanes allow the structural variations to achieve a range of mechanical properties. A major limitation of polyurethanes for biomedical applications is the involvement of toxic precursors (such as toluene diisocyanates) in the synthesis. Recent efforts have been on development of biodegradable polyurethanes or urethane-based polymers using less toxic diisocyanates (59). These polymers have good mechanical properties and have been explored for vascular and other tissue engineering applications (60,61). Considering the general audience in the field of polymer science and engineering, synthetic polymers and 3-D scaffold fabrications are the main foci of this article. However, it should be noted that natural polymers, such as proteins and polysaccharides, have also been used for tissue engineering applications. Collagen (qv) is a major natural extracellular matrix component (62). Collagen proteins possess triple-helix structure over a large portion of the molecules. Among the various collagen proteins the most abundant is type I collagen. Collagen has been used for various tissue regeneration applications especially for soft tissue repair (63,64). The use of collagen as a scaffolding material, however, remains challenging because of the potential pathogen transmission and immune reactions, poor handling and mechanical properties, and less controlled biodegradability (enzymatic) of the natural material from biological sources. On the other hand, collagen as a natural extracellular component has useful biological properties desired for tissue engineering applications. Yannas and colleagues have developed collagenglycosaminoglycan (GAG) copolymers and fabricated them into scaffolds for tissue engineering (65,66). Denatured collagen (gelatin) is also processed into porous materials for tissue repair (67). Polysaccharides (qv) are another class of natural polymers, eg, alginate (68) and chitosan (69), which have been explored as tissue engineering scaffolds. Polymers for Hydrogel Scaffolds. Hydrogels (qv) are cross-linked hydrophilic polymers that contain large amounts of water without dissolution. Hydrogels are attractive candidates for certain tissue engineering applications because of the ability to fill irregularly shaped tissue defects, allowance of minimally invasive procedures such as arthroscopic surgeries, and the ease of incorporation of cells or bioactive agents (70–75). Among the synthetic hydrogels, PEG (Fig. 2) has been more frequently studied for tissue engineering research than other synthetic hydrogels. PEG has been extensively used in biomedical applications often to prevent protein and cell adhesion (76). This property also made PEG useful for prevention of tissue adhesions after surgery (77,78). In the tissue engineering field, PEG has been used as a model system for studies related to adhesion peptides. In such a case, the inertness of this polymer to cells and proteins is ideally utilized. PEG–diacrylamide was synthesized by the Hubbell group (79). A photopolymerization step can be performed in contact with cells, providing a means to produce scaffolds for tissue engineering. A PEG-based interpenetration network has been used for cartilage tissue engineering by Elisseeff and colleagues (80). Peptide-modified PEG materials have been used by Griffith and West groups to study relationships between cell behavior (such as adhesion, spreading, and migration) and peptide density and spacing (81–83). A major limitation of PEG hydrogel for tissue engineering scaffolds is its lack of degradability. Efforts are made to impart degradability to PEG by either
270
TISSUE ENGINEERING
Vol. 12
formulating copolymers with PLA or PGA, or by introducing enzyme-degradable linkages into the PEG backbone (84–86). Alginate is the salt form of alginic acid. Alginic acid, a polysaccharide from seaweed, is a family of natural copolymers of β-D-mannuronic acid (M) and αL-guluronic acid (G) (Fig. 2) (87,88). They have been processed into gel beads encapsulating living cells as a means of immunoprotection (89). Alginate hydrogels cross-linked using calcium sulfate (CaSO4 ) have recently been used as cell delivery vehicles for in vivo tissue engineering research (74,90,91). Peptidemodified alginate has been studied to improve cell attachment (74). However, one of the major disadvantages of the use of CaSO4 is that gelation kinetics is difficult to control, and the resulting structure is not uniform. Structural uniformity in tissue engineering scaffolds is necessary not only for uniform cell distribution, but also for well-controlled material properties. Mechanical properties are more consistent throughout the gel and between batches if structural uniformity can be achieved. We have developed methods to control the gelation rate so that cross-linking is allowed to take place uniformly throughout the gel to form a structurally uniform and mechanically strong alginate gel with predesigned 3-D shapes (Fig. 3a) (70). Cells can be uniformly incorporated into such gels (Fig. 3b). Collagen, in addition to porous foam, has also been used as hydrogels for a variety of tissue repair and regeneration studies (92,93). Collagen gels have been studied as a therapeutic option for the treatment of burn patients or chronic wounds (94). Cartilage defect repair has been studied using chondrocyte–collagen gel constructs (95). Studies have also been conducted with collagen gels to deliver mesenchymal stem cells and biological agents to improve healing of ligaments and tendons (96).
Polymer Processing and 3-D Scaffolds for Tissue Engineering In the body, tissues are organized into 3-D structures as functional organs and organ systems. Each tissue or organ has its own characteristic architecture depending on its physiological function. This architecture is also believed to provide the necessary channels for mass transport and spatial cell organization (15). Mass transport includes signaling molecules, nutritional supplies, and metabolic waste removal. Spatial cellular organization allows for cell–cell and cell–extracellular matrix interactions to occur, which are critical to physiological functions of a tissue or organ. To successfully engineer functional tissues and organs, the scaffolds have to be designed to facilitate the desired cell distribution and to guide tissue regeneration in 3-D. This section is, therefore, devoted to the 3-D pore structure and pore wall architecture of scaffolds for tissue engineering. Textile Technologies and Traditional Fibrous Scaffolds. The textile technologies have developed over thousands of years to produce various fibrous fabrics for clothing, filtration, packaging, and many other industrial and household applications. Fibrous fabrics have excellent mechanical properties. It was natural that fibrous materials found many biomedical applications in recent years including sutures from biodegradable polymers. It was also not surprising that earlier tissue engineering scaffolds were fibrous fabrics of biodegradable polymers
Vol. 12
TISSUE ENGINEERING
271
Fig. 3. Alginate hydrogels used as scaffolds for tissue engineering: (a) alginate gels of various molded shapes made from a slow gelation process; (b) MC3T3-E1 osteoblasts incorporated into slow-gelling alginate gels and cultured in vitro for 3 weeks, demonstrating uniform cell distribution (H&E staining, original magnification: 100×). Reprinted from c 2001, by permission of Elsevier. Ref. 70.
fabricated using textile technologies. PGA, PLA, and many other semicrystalline polymers can be processed into fibers using an extruder and these fibers can be further processed into woven, knit, or nonwoven fabrics using the textile technologies (see NONWOVENS, STAPLE FIBER). One of such scaffolds widely used in tissue engineering research is PGA nonwoven scaffolds (Fig. 4). PGA is typically melt-spun into fibers with a diameter of around 15 µm. These fibers are then processed into nonwoven fabrics using textile technologies such as carding, needling, heat pressing, and so forth. Porosity higher than 90% can be easily achieved. These PGA nonwoven scaffolds degrade via hydrolysis both in vitro and in vivo. The mass of PGA scaffolds decrease exponentially in vitro, fitting well to a first-order degradation kinetics (17). The amorphous regions in the fibers degrade first presumably because of the easy access by water
272
TISSUE ENGINEERING
Vol. 12
Fig. 4. SEM micrograph of a PGA nonwoven scaffold with a porosity of approximately 95% and a fiber diameter of approximately 15 µm.
molecules. The crystalline regions start to degrade soon after, leading to loss of molecular orientation in the fibers and decrease in mechanical properties. These PGA scaffolds have been used either alone or combined with other biodegradable polymers for the engineering of cartilage (97–99), tendon (100), ureter (101), intestine (102), blood vessels (103,104), heart valves (105), and other tissues. Although PGA nonwoven scaffolds have stimulated a large wave of tissue engineering research and generated considerable excitement in the field, there are several limitations of PGA nonwoven scaffolds, such as low mechanical strength, fast degradation rate (losing mechanical properties within 2 weeks), difficulty in pore shape control, and limited fiber diameter variations. Therefore, many other processing techniques have been developed to fabricate scaffolds for desired properties in various tissue engineering applications. Particulate-Leaching Techniques. Particulate-leaching is another technique that has been widely used to fabricate porous materials as scaffolds for tissue engineering applications. This technique was first developed by Mikos and colleagues (16), and was improved and standardized for large batch fabrications (7). This process is schematically illustrated in Figure 5, which is often called salt-leaching technique because NaCl crystals are most often used as the particles for pore generation (porogen). Briefly, salt is first ground into small particles, and the particles are then separated into different size ranges using standard sieves. The particles of desired size are added into a mold of a needed shape. A polymer solution is prepared and cast into the salt-filled mold. After the evaporation of solvent, the salt crystals are leached away using water to form the pores of the polymer foam (Fig. 6). There is no complicated equipment needed for this technique. The process is easy to carry out and the materials used (water and salt) are economical. The technique is also versatile and can be used to fabricate foams from a variety of polymers as long as the polymer can be dissolved in a solvent that does not dissolve
Vol. 12
TISSUE ENGINEERING
273
Fig. 5. Schematic illustration of the salt-leaching technique for porous foam fabrication.
the salt. The pore size can be controlled by the size of the salt crystals, and the porosity can be controlled by the salt/polymer ratio. For example, larger particle size results in larger pore size, and higher salt/polymer ratio results in higher porosity. Therefore, this technique is widely used in the tissue engineering field. However, certain critical variables such as pore shape and interpore openings are not controlled. For example, salt crystals have the characteristic cubic shape (Fig. 6). A spherical pore shape is not achievable using salt crystals. The resulting pores are often not well-connected because the polymer solution tends to penetrate between salt particles. To overcome the shortcomings of the salt-leaching
Fig. 6. SEM micrograph of a PLLA foam fabricated using the salt-leaching technique.
274
TISSUE ENGINEERING
Vol. 12
technique as well as the textile technologies, various other techniques have been developed to have better control over the pore structure of the scaffolds. Phase Separation and Novel Architectures of Scaffolds. For a homogeneous multicomponent system, under certain conditions the system becomes thermodynamically unstable and tends to separate into a multiphase system in order to lower the system free energy. Controlled phase separation processes are desired for the generation of porous structures for tissue engineering scaffolding fabrication. For a polymer solution, phase separation can be induced in several different ways such as non-solvent-induced phase separation, chemically induced phase separation, and thermally induced phase separation. When phase separation occurs, a polymer solution separates into two phases, a polymer-rich phase (with a high polymer concentration) and a polymer-lean phase (with a low polymer concentration). After the solvent is removed by extraction, evaporation, or sublimation, the polymer-rich phase solidifies. Depending upon the system and phase separation conditions the resulting materials are different in physical form: powder, closed-pore foam, or open-pore foam (Fig. 7). Porous materials formed through phase separation have been utilized as membranes for filtration and separation (106). However, the pores formed through phase separation usually have diameters on the order of a few to tens of micrometers and are often not uniformly distributed, which are not suitable for tissue engineering applications.
Fig. 7. Schematic illustration of phase separation processes that lead to different material forms: A, powder; B, open porous structure suitable as a scaffold for tissue engineering; and C, foam with closed pores.
Vol. 12
TISSUE ENGINEERING
275
As discussed earlier, a scaffold must have a pore size big enough for accommodating cells, allowing proliferation, and having room for new tissue synthesis. They also need to have high surface area to enhance cell adhesion. The pore structures should also allow for vascularization and innervation, and should facilitate diffusion of nutrients and waste materials. Thermally induced phase separation techniques have been explored in our laboratory for fabricating scaffolds to satisfy the requirements of tissue engineering. This type of phase separation is based on the thermodynamic and kinetic behavior of polymer solutions, which is a complicated process (26,107,108). The important variables are the polymer, solvent, concentration of the polymer solution, and the phase separation temperature. These variables have been manipulated to generate a variety of porous architectures for scaffold fabrication as briefly discussed below. Solid–Liquid Phase Separation. Phase separation can be achieved by inducing solvent crystallization by decreasing the temperature of a polymer solution. This process is defined as a solid–liquid phase separation (solid-phase formation in a liquid phase). To achieve this, the crystallization temperature (freezing point) of the solvent in the polymer solution needs to be higher than the liquid–liquid phase separation temperature, or the liquid–liquid phase separation is kinetically too slow to take place before the temperature reaches the solid–liquid phase separation. When the temperature of the solution decreases and reaches the crystallization temperature, the solvent crystallizes and the polymer is expelled from the solvent crystallization front (separated from the solvent crystal lattice). After the removal of the solvent crystals (sublimation or solvent exchange), the space originally taken by the solvent crystals becomes pores. By manipulating the phase separation variables, foams with a variety of pore morphologies can be obtained. For example, PLLA and PLGA solutions have been used to fabricate interconnected pore structures as scaffolds using solid–liquid phase separation techniques (Fig. 8) (5,109). This technique can be used to fabricate scaffolds of more than one type of materials, including composite scaffold fabrication (Fig. 8c) (109). These interconnected pore structures also allow further modification of the pore wall surfaces after the scaffold fabrication. Bone-like apatite coating on such polymer scaffolds via a biomimetic process is such an example (20). Proteins and other bioactive molecules can also be used for such internal pore surface modifications (110). By manipulating the phase separation conditions, various pore structures can be achieved. For example, many tissues (such as nerve, muscle, tendon, ligament, dentin, and so on) have oriented tubular or fibrous bundle architectures. To facilitate the organization and regeneration of such tissue types, a scaffold with a high porosity and an oriented array of open microtubules may be desirable. To achieve this goal, a novel phase separation technique has been developed for the creation of a parallel array of microtubules (Fig. 9) (111). In such a process, nucleation of the solvent is induced at one side of the polymer solution. The nuclei will grow in all directions until they contact with each other on this side. Then they cannot grow further in this plane, and are forced to grow along the direction perpendicular to this plane, leading to the formation of an array of parallel rodlike solvent crystals. The polymers are expelled from the solution and squeezed into thin walls surrounding these parallel solvent rods. After the removal of these rods, a parallel array of microtubules is formed as demonstrated. This oriented
Fig. 8. SEM micrographs of polymer scaffolds fabricated using solid–liquid phase separation: (a) PLLA scaffold fabricated from 5% PLLA/dioxane solution (with local regular pore structure), (b) PLLA scaffold fabricated from 2.5% PLLA/dioxane solution (with less regular structure), (c) PLLA/HAP (hydroxyapatite) composite scaffold (PLLA/HAP: 50/50) c 2001, fabricated from a 2.5% PLLA/dioxane solution. Reprinted from Refs. 5 and 109. by permission of John Wiley & Sons. 276
Vol. 12
TISSUE ENGINEERING
277
Fig. 9. SEM micrographs of a PLLA scaffold with oriented microtubular architecture prepared from 5% PLLA solution: (a) longitudinal section, and (b) cross section. From Ma c 2001, by permission of John Wiley & Sons. and Reprinted from Ref. 111.
tubular scaffold has anisotropic mechanical properties similar to fibrillar and tubular tissues, and has been shown to facilitate cell organization into oriented fibrillar or tubular tissues (Fig. 10) (111). Liquid–Liquid Phase Separation. In contrast to solid–liquid phase separation, lowering temperature can induce liquid–liquid phase separation of a polymer solution with an upper critical solution temperature and when the crystallization temperature of the solvent is sufficiently lower than the phase separation temperature. In an equilibrium phase diagram of a polymer solution, the spinodal curve divides the liquid–liquid phase separation region into two regions: a thermodynamically metastable region (between the binodal and spinodal) and a thermodynamically unstable region (enclosed by the spinodal) (Fig. 11). Above the
278
TISSUE ENGINEERING
Vol. 12
Fig. 10. MC3T3-E1 cells cultured on a PLLA scaffold with parallel open microtubular architecture for 4 weeks in vitro, von Kossa’s silver nitrate staining. Reprinted from c 2001, by permission of John Wiley & Sons. Ref. 111.
binodal curve, the solution is homogeneous. When the temperature of the polymer solution is lowered to a point below the binodal curve, the solution is not thermodynamically stable and tends to separate into two phases, a polymer-rich phase and a polymer-lean phase, to lower the system free energy. In a solution of very low polymer concentration, when the temperature is lowered to a point in the metastable region (between binodal and spinodal), the phase separation occurs via a nucleation and growth mechanism. This process can lead to the formation of small polymer-rich domains (droplets) in a polymer-lean matrix. After the removal of the solvent, polymer powder is formed (Fig. 7a). When the temperature is lowered into an unstable region (spinodal region), the phase separation occurs via a spinodal decomposition mechanism. This process leads to the formation of a
Fig. 11. Schematic equilibrium phase diagram for a polymer solution with an upper critical solution temperature.
Vol. 12
TISSUE ENGINEERING
279
bicontinuous structure, in which the polymer-rich and polymer-lean phases are both continuous phases. After the removal of the solvent, foam with open-pore structure is formed (Fig. 7b). In a solution of very high polymer concentration, when the temperature is lowered to a point in the metastable region, the phase separation leads to the formation of small solvent (polymer-lean) domains in a polymer-rich matrix. After the removal of the solvent, polymer foam of closed-pore structure is formed (Fig. 7c). If the polymer is crystalline, the phase separation process is more complicated (26). There is a competition between the liquid–liquid phase separation and polymer crystallization under certain conditions, making the final pore structure more complex (26,112). Thermally induced liquid–liquid phase separation can be utilized to fabricate scaffolds for tissue engineering. For example, a solvent can be selected, where the crystallization temperature is sufficiently lower than the liquid–liquid phase separation temperature of an amorphous polymer solution. A liquid–liquid phase separation can be induced by lowering the temperature into the unstable region on the phase diagram but above the solvent crystallization temperature. For the PLA and PLGA family, a mixture of dioxane and water has been used for liquid– liquid phase separation to fabricate polymer scaffolds with interconnected pore structure (Fig. 12) (112,113). Nanofibrous Matrix. Collagen (qv) is a major natural extracellular matrix component and possesses a fibrous structure with fiber bundles varying in diameter from 50 to 500 nm (62,114). To mimic the nanofibrous architecture of collagen and to overcome the concerns over materials from a natural source such as pathogen transmission and immune rejection, a novel phase separation technique has been developed in our laboratory to fabricate nanofibrous matrices from synthetic biodegradable polymers (112). For example, PLLA solutions are cooled to induce phase separation and gelation. The solvent is directly sublimated or first exchanged with a different solvent and then sublimated. Several solvents and solvent mixtures have been utilized to fabricate the desired nanofibrous matrices in our laboratory (Fig. 13) (112). Three-Dimensional Pore Architecture Design. As discussed above various novel scaffolds have been developed for tissue engineering to address the critical shortage of donor tissues/organs. These scaffolds have facilitated the tissue engineering research in demonstrating the feasibility and tremendous potential of tissue engineering. However, these scaffolds are not perfect. One of the common shortcomings of these fabrication technologies is the lack of precise control of 3-D pore architecture of the scaffolds. To tackle this problem, computer-assisteddesign and computer-assisted-manufacture (CAD and CAM) techniques that have been widely used in modern manufacture industry are finding ways into the field of tissue engineering (115). These techniques were early explored by a group of materials scientists and chemical engineers at MIT (116,117). They used one of the solid free-form fabrication (a.k.a. rapid prototyping) techniques, called 3-D printing (3DP). With such a technique, complex-shaped objects can be designed using CAD softwares and directly fabricated from the generated CAD models. For example, they fabricated various structures from biodegradable polymers by ink-jet printing a binder onto sequentially laid polymer powder layers. The advantages of such a technique include the precise control of geometry and the feasibility for repeated fabrication
280
TISSUE ENGINEERING
Vol. 12
Fig. 12. SEM micrographs of a porous scaffold prepared from a 10% solution of PLGA (85/15) in a mixture of dioxane/H2 O (80/20) at two different magnifications: (a) low magc 1999, by permission of nification and (b) high magnification. Reprinted from Ref. 112. John Wiley & Sons.
of the same structure (118). However, the smallness of the powder particles and the binder drops (pixels) are limited (a few hundred micrometers). The accuracy of positioning the printing nozzle is also limited. The size of a feature (shape of a pore or architectural component) is dependent on the resolution of the pixel size and positioning control. Therefore the preciseness of the technology is seriously limited. The anticipation of the future improvement of the resolution of such technologies has, nevertheless, maintained the research momentum in this direction. Similarly, another rapid prototyping technique, fused deposition modeling, has been used to fabricate 3-D scaffolds (119). This technique is suitable for processing thermoplastic polymers. A heated nozzle is utilized to extrude polymer
Vol. 12
TISSUE ENGINEERING
281
Fig. 13. SEM micrographs of a PLLA nanofibrous matrix prepared from 2.5% (w/v) PLLA/THF solution at a phase separation temperature of 8◦ C: (a) 500×; (b) 20,000×. c 1999, by permission of John Wiley & Sons. Reprinted from Ref. 112.
filament on to a platform. A layer is patterned by depositing the continuous polymer filament and raster in the x–y directions. The motion along z direction allows layer-by-layer integration of the 2-D patterns into 3-D structures. In our laboratory, a similar process is used to fabricate the negative replica of the scaffold. A polymer solution is cast into such a mold and solidified after the removal of the solvent. The mold is then dissolved away to form the polymer scaffold with the designed 3-D pore network (Fig. 14). Lithography is another processing technique that is actively explored for 3-D scaffold fabrication. Stereolithographic models derived from X-ray computed tomography and CAD software were used to recreate complex anatomic structure of a human pulmonary and aortic graft. These stereolithographic models were used to generate biodegradable heart valve scaffolds by a thermal processing technique (120). Hydroxyapatite scaffolds with various pore shapes have also been fabricated
282
TISSUE ENGINEERING
Vol. 12
Fig. 14. Scaffold fabrication using rapid prototyped negative replica: (a) A computergenerated 3-D negative replica of a scaffold with a cubical pore shape; (b) the negative replica of the scaffold fabricated using a rapid prototyping machine; (c) the polymer scaffold fabricated using the negative replica; (d) SEM micrograph of the internal pore structure of the generated scaffold (V. J. Chen and P. X. Ma, unpublished data, 2003).
by casting ceramic slurry into molds generated using stereolithography and a subsequent sintering process (Fig. 15) (121). Lithography techniques have also been used to fabricate devices to localize cell populations in patterned configurations on rigid substrates as a potential artificial liver (122). A complex silicon structure of branched vascular channels as a model for liver fabrication has also been developed using stereolithography (123). However, in addition to the special equipment requirements all these fabrication techniques have their inherent shortcomings such as limited material selections and inadequate resolution like the 3DP techniques discussed earlier. Furthermore, the resulting constructs have structural heterogeneity due to the “pixel assembly” nature of these fabrication processes. In our laboratory, reversed fabrication processes have been developed to overcome these shortcomings for tissue engineering scaffold fabrications (Fig. 16)
Vol. 12
TISSUE ENGINEERING
283
Fig. 15. Hydroxyapatite scaffolds fabricated by casting ceramic slurry into molds generated using stereolithography and a subsequent sintering process. Reprinted from Ref. 121. c 2001, by permission of Kluwer Academic Publishers.
(124). To fabricate 3-D biodegradable polymer scaffolds with well-controlled interconnected spherical pores, paraffin spheres are fabricated with a dispersion method. These paraffin spheres are then transferred into a 3-D mold of a designed shape or anatomical shape of a body part. The spheres are bonded together in the mold through a heat treatment process. A polymer solution is cast into the paraffin assembly in the mold (Note: the solvent used should not be a solvent of the paraffin spheres). After removal of the solvent through evaporation or other means, the polymer–paraffin sphere assembly is immersed in a solvent of the paraffin but not a solvent for the polymer (such as hexane for PLA or PLGA scaffolds) to dissolve the paraffin sphere assembly. In this way, an interconnected spherical pore structure is created in a predetermined shape by the mold (Fig. 17). Importantly, the generated scaffolds have homogeneous foam skeleton (platelet-like, continuous,
Fig. 16. Schematic fabrication processes for polymer scaffolds with interconnected spherical pores.
284
TISSUE ENGINEERING
Vol. 12
Fig. 17. SEM micrographs of polymer scaffolds [(a)–(d) PLLA; (e) and (f) PLGA85/15] with an interconnected spherical pore structure prepared using paraffin spheres (porogen): (a)100×, paraffin spheres: 250–420 µm; (b) 250×, paraffin spheres: 250–420 µm; (c)1000×, paraffin spheres: 250–420 µm; (d) 3000×, paraffin spheres: 250–420 µm; (e) 50×, paraffin c spheres: 420–500 µm; (f) 100×, paraffin spheres: 420–500 µm. Reprinted from Ref. 124. 2001, by permission of Mary Ann Liebert.
or other complex features, depending on the polymer and phase separation conditions) with homogeneous material properties, which are not easily achievable with free-forming, 3DP or lithography because of the limitation of pixel-by-pixel construction. Equally importantly, the spherical pore size is determined by the
Fig. 18. PLLA nanofibrous scaffold with helicoidal tubular macropore network prepared from PLLA/THF solution and a helicoidal sugar fiber of the scaffold assembly: (a) schematic illustration of helicoidal sugar fiber assembly; (b) SEM micrograph of the scaffold at an original magnification of 35×, and (c) original magnification of 250×. Reprinted from c 2000, by permission of John Wiley & Sons. Ref. 15. 285
286
TISSUE ENGINEERING
Vol. 12
paraffin sphere size used, which can be controlled through paraffin sphere preparation. The features generated are significatly better than those achievable with the resolutions of the current 3-D fabrication machines. Furthermore, the interpore openings are circular and the size of these openings is adjustable through manipulating the heat treatment conditions. For example, longer heat treatment time leads to larger openings between the pores (124). Finally there is no expensive equipment investment compared to the textile technologies or rapid prototyping techniques, which allows this technology to be easily adapted in a research as well as an industrial setting. To improve the 3-D structure of nanofibrous scaffolds for cell seeding and distribution, mass transport, vascular invasion, and tissue organization, techniques have also been developed in our laboratory to build predesigned macropore networks in nanofibrous matrices (15). For example, larger water-soluble fibers (diameters from 100 µm to 1 mm) are prepared from sugar as a geometrical porogen element, and are assembled into various 3-D structures (such as helicoidal) (Fig. 18a). PLLA solution is then cast into this 3-D assembly and is thermally induced to phase-separate for nanofibrous matrix formation. After the solvent removal, the sugar fiber assembly is dissolved away using water to achieve nanofibrous scaffolds with predesigned helicoidal tubular pore network (Figs. 18b and 18c). Similarly, we have combined the interconnected spherical pore network described earlier with nanofiber techniques to generate nanofibrous scaffolds with interconnected spherical macropores (Fig. 19) (125). As discussed above, tissue engineering scaffold fabrication is a fast evolving area. Many other polymer processing techniques such as gas-foaming, emulsion freeze-drying, and so forth have also been explored (126–128). Current efforts for these techniques include overcoming the disadvantages of closed-pore structures and undesired pore size ranges. Again this article is intended to introduce the general concepts of tissue engineering in relation to polymer science and engineering, but is not intended to be a complete and exhaustive review of the field of tissue engineering. There are
Fig. 19. SEM micrographs of nanofibrous scaffolds with interconnected spherical pore network: (a) original magnification 30×, (b) original magnification 4000×. Reprinted from c 2003, by permission of John Wiley & Sons. Ref. 125.
Vol. 12
TISSUE ENGINEERING
287
recently developed comprehensive books (129–131) available for further reading on various topics in tissue engineering.
BIBLIOGRAPHY 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.
18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
28.
J. P. Vacanti and R. Langer, Lancet 354 (Suppl 1), SI32–SI34 (1999). N. Couch, R. Wilson, E. Hager, and J. Murray, Surgery 59(2), 183–188 (1966). R. Langer and J. P. Vacanti, Science 260, 920–926 (1993). V. M. Goldberg and S. Stevenson, Orthopedics 17, 809–810, 821 (1994). P. X. Ma, R. Zhang, G. Xiao, and R. Franceschi, J. Biomed. Mater. Res. 54, 284–293 (2001). R. M. Nerem and A. Sambanis, Tissue Eng. 1(1), 3–13 (1995). P. X. Ma and R. Langer, in J. Morgan and M. Yarmush, eds., Tissue Engineering Methods and Protocols, Humana Press, Inc., Totowa, N.J., 1999, pp. 47–56. L. Peterson, M. Brittberg, I. Kiviranta, E. Akerlund, and A. Lindahl, Am. J. Sports Med. 30(1), 2–12 (2002). J. Bloch, E. Fine, N. Bouche, A. Zurn, and P. Aebischer, Exp. Neurol. 172, 425–432 (2001). G. N. King, Curr. Pharm. Biotechnol. 2(2), 131–142 (2001). C. K. Colton, Nat. Biotechnol. 17, 421–422 (1999). E. S. Tzanakakis, D. J. Hess, T. D. Sielaff, and W. S. Hu, Annu. Rev. Biomed. Eng. 2, 607–632 (2000). H. D. Humes, W. H. Fissell, W. F. Weitzel, D. A. Buffington, A. J. Westover, S. M. MacKay, and J. M. Gutierrez,Am. J. Kidney Dis. 39, 1078–1087 (2002). S. Del Guerra, C. Bracci, K. Nilsson, A. Belcourt, L. Kessler, R. Lupi, L. Marselli, P. De Vos, and P. Marchetti, Biotechnol. Bioeng. 75, 741–744 (2001). R. Zhang and P. X. Ma, J. Biomed. Mater. Res. 52, 430–438 (2000). A. G. Mikos, A. J. Thorsen, L. A. Czerwonka, Y. Bao, R. Langer, D. N. Winslow, and J. P. Vacanti, Polymer 35, 1068–1077 (1994). P. X. Ma and R. Langer, in A. G. Mikos, K. W. Leong, M. L. Radomsky, J. A. Tamada, and M. J. Yaszemski, eds., Polymers in Medicine and Pharmacy, MRS, Pittsburgh, Pa., 1995 pp. 99–104. J. A. Hubbell, Curr. Opin. Biotechnol. 10(2), 123–129 (1999). D. Barrera, E. Zylstra, P. Lansbury, and R. Langer, J. Am. Chem. Soc. 115, 11010–11011 (1993). R. Zhang and P. X. Ma, J. Biomed. Mater. Res. 45, 285–293 (1999). H. Shi and B. D. Ratner, J. Biomed. Mater. Res. 49(1), 1–11 (2000). H. Lee, R. A. Cusick, F. Browne, T. Ho Kim, P. X. Ma, H. Utsunomiya, R. Langer, and J. P. Vacanti, Transplantation 73, 1589–1593 (2002). J. Bonadio, E. Smiley, P. Patil, and S. Goldstein, Nat. Med. 5, 753–759 (1999). H. Lo, S. Kadiyala, S. E. Guggino, and K. W. Leong, J. Biomed. Mater. Res. 30, 475–484 (1996). M. H. Sheridan, L. D. Shea, M. C. Peters, and D. J. Mooney, J. Controlled Release 64(1–3), 91–102 (2000). R. Zhang and P. X. Ma, in A. Atala and R. Lanza, eds., Methods of Tissue Engineering, Academic Press, San Diego, Calif., 2001, pp. 715–724. J. H. Elisseeff and Y. Yamada, in K.-U. Lewandrowski, D. L. Wise, D. J. Trantolo, J. D. Gresser, M. J. Yaszemski, and D. E. Altobelli, eds., Tissue Engineering and Biodegradable Equivalents: Scientific and Clinical Applications, Marcel Dekker, Inc., New York, 2002, pp. 1–23. J. M. Pachence and J. Kohn, in R. P. Lanza, R. Langer, and J. Vacanti, eds., Principles of Tissue Engineering, Academic Press, San Diego, Calif., 2000, pp. 263–277.
288 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62.
TISSUE ENGINEERING
Vol. 12
A. Reed and D. Gilding, Polymer 22, 494–498 (1981). R. Zhang and P. X. Ma, Polym. Prepr. 41, 1618–1619 (2000). B. Eling, S. Gogolewski, and A. J. Pennings, Polymer 23, 1587–1593 (1982). O. M. Bostman and H. K. Pihlajamaki, Clin. Orthop. 371, 216–227 (2000). J. A. Tamada and R. Langer, Proc. Nat. Acad. Sci. U.S.A. 90, 552–556 (1993). A. Gopferich, Biomaterials 17(2), 103–114 (1996). S. Li, J. Biomed. Mater. Res. 48, 342–353 (1999). C. G. Pitt, M. M. Gratzl, G. L. Kimmel, J. Surles, and A. Schindler, Biomaterials 2, 215–220 (1981). S. C. Woodward, P. S. Brewer, F. Moatamed, A. Schindler, and C. G. Pitt, J. Biomed. Mater. Res. 19, 437–444 (1985). C. G. Pitt, in M. Chasin and R. Langer, eds., Biodegradable Polymers as Drug Delivery Systems, Marcel Dekker, Inc., New York, 1990, pp. 71–119. S. J. Holland, A. M. Jolly, M. Yasin, and B. J. Tighe, Biomaterials 8, 289–295 (1987). N. D. Miller and D. F. Williams, Biomaterials 8(2), 129–137 (1987). M. Richards, B. I. Dahiyat, D. M. Arm, P. R. Brown, and K. W. Leong, J. Biomed. Mater. Res. 25, 1151–1167 (1991). J. Wang, H. Q. Mao, and K. W. Leong, J. Am. Chem. Soc. 123, 9480–9481 (2001). S. Wang, A. C. Wan, X. Xu, S. Gao, H. Q. Mao, K. W. Leong, and H. Yu, Biomaterials. 22, 1157–1169 (2001). M. S. Wolfe, D. Dean, J. E. Chen, J. P. Fisher, S. Han, C. M. Rimnac, and A. G. Mikos, J. Biomed. Mater. Res. 61(1), 159–164 (2002). S. J. Peter, S. T. Miller, G. Zhu, A. W. Yasko, and A. G. Mikos, J. Biomed. Mater. Res. 41(1), 1–7 (1998). E. Behravesh, S. Jo, K. Zygourakis, and A. G. Mikos, Biomacromolecules 3, 374–381 (2002). H. R. Allcock, in M. Chasin and R. Langer, eds., Biodegradable Polymers as Drug Delivery Systems, Marcel Dekker, Inc., New York, 1990, pp. 163–193. P. Caliceti, F. M. Veronese, and S. Lora, Int. J. Pharm. 211(1/2), 57–65 (2000). N. N. Aldini, P. Caliceti, S. Lora, M. Fini, G. Giavaresi, M. Rocca, P. Torricelli, R. Giardino, and F. M. Veronese, J. Orthop. Res. 19, 955–961 (2001). A. M. Ambrosio, H. R. Allcock, D. S. Katti, and C. T. Laurencin, Biomaterials 23, 1667–1672 (2002). S. Cohen, M. C. Bano, L. G. Cima, H. R. Allcock, J. P. Vacanti, C. A. Vacanti, and R. Langer, Clin. Mater. 13(1–4), 3–10 (1993). A. K. Burkoth, J. Burdick, and K. S. Anseth, J. Biomed. Mater. Res. 51, 352–359 (2000). K. W. Leong, J. Kost, E. Mathiowitz, and R. Langer, Biomaterials 7, 364–371 (1986). U.S. Pat. 04093709 (1978), N. S. Choi and J. Heller. K. P. Andriano, Y. Tabata, Y. Ikada, and J. Heller, J. Biomed. Mater. Res. 48, 602–612 (1999). K. James, H. Levene, J. R. Parsons, and J. Kohn, Biomaterials 20, 2203–2212 (1999). K. A. Hooper, N. D. Macon, and J. Kohn, J. Biomed. Mater. Res. 41, 443–454 (1998). H. B. Lin, W. Sun, D. F. Mosher, C. Garcia-Echeverria, K. Schaufelberger, P. I. Lelkes, and S. L. Cooper, J. Biomed. Mater. Res. 28, 329–342 (1994). G. A. Skarja and K. A. Woodhouse, J. Biomater. Sci., Polym. Ed. 12, 851–873 (2001). J. J. Guan, M. S. Sacks, E. J. Beckman, and W. R. Wagner, J. Biomed. Mater. Res. 61, 493–503 (2002). J. Y. Zhang, E. J. Beckman, N. P. Piesco, and S. Agarwal, Biomaterials 21, 1247–1258 (2000). E. D. Hay, Cell Biology of Extracellular Matrix, 2nd ed., Plenum Press, New York, 1991, pp. 7–44.
Vol. 12
TISSUE ENGINEERING
289
63. E. Bell, M. Rosenberg, P. Kemp, R. Gay, G. Green, N. Muthukumaran, and C. Nolte, J. Biomech. Eng. 113(2), 113–119 (1991). 64. J. M. Pachence, J. Biomed. Mater. Res. 33(1), 35–40 (1996). 65. I. V. Yannas, J. Cell. Biochem. 56(2), 188–191 (1994). 66. S. M. Mueller, S. Shortkroff, T. O. Schneider, H. A. Breinan, I. V. Yannas, and M. Spector, Biomaterials 20, 701–709 (1999). 67. Y. S. Choi, S. R. Hong, Y. M. Lee, K. W. Song, M. H. Park, and Y. S. Nam, J. Biomed. Mater. Res. 48, 631–639 (1999). 68. L. Shapiro and S. Cohen, Biomaterials 18, 583–590 (1997). 69. S. V. Madihally and H. W. Matthew, Biomaterials 20, 1133–1142 (1999). 70. C. K. Kuo and P. X. Ma, Biomaterials 22, 511–521 (2001). 71. J. P. Fisher, T. A. Holland, D. Dean, P. S. Engel, and A. G. Mikos, J. Biomater. Sci., Polym. Ed. 12, 673–687 (2001). 72. J. H. Elisseeff, A. Lee, H. K. Kleinman, and Y. Yamada, Ann. N.Y. Acad. Sci. 961, 118–122 (2002). 73. J. A. Burdick, A. J. Peterson, and K. S. Anseth, Biomaterials 22, 1779–1786 (2001). 74. E. Alsberg, K. W. Anderson, A. Albeiruti, R. T. Franceschi, and D. J. Mooney, J. Dent. Res. 80, 2025–2029 (2001). 75. B. K. Mann, A. S. Gobin, A. T. Tsai, R. H. Schmedlen, and J. L. West, Biomaterials 22, 3045–3051 (2001). 76. N. A. Alcantar, E. S. Aydil, and J. N. Israelachvili, J. Biomed. Mater. Res. 51, 343–351 (2000). 77. N. D. Winblade, I. D. Nikolic, A. S. Hoffman, and J. A. Hubbell, Biomacromolecules 1, 523–533 (2000). 78. N. K. Mongia, K. S. Anseth, and N. A. Peppas, J. Biomater. Sci., Polym. Ed. 1055, 7, 1055–1064. 79. D. L. Elbert and J. A. Hubbell, Biomacromolecules 2, 430–441 (2001). 80. J. Elisseeff, K. Anseth, D. Sims, W. McIntosh, M. Randolph, M. Yaremchuk, and R. Langer, Plast. Reconstr. Surg. 104, 1014–1022 (1999). 81. L. G. Griffith and S. Lopina, Biomaterials 19, 979–986 (1998). 82. D. J. Irvine, A. M. Mayes, and L. G. Griffith, Biomacromolecules 2(1), 85–94 (2001). 83. B. K. Mann and J. L. West, J. Biomed. Mater. Res. 60(1), 86–93 (2002). 84. D. K. Han and J. A. Hubbell, Macromolecules 30, 6077–6083 (1997). 85. A. S. Sawhney, C. P. Pathak, and J. A. Hubbell, Macromolecules 26, 581–587 (1993). 86. J. L. West and J. A. Hubbell, Macromolecules 32, 241–244 (1999). ˚ 87. A. Martinsen, G. Skjak-Bræk, and O. Smidsrød, Biotechnol. and Bioeng. 33, 79–89 (1989). 88. K. I. Draget, K. Ostgaard, and O. Smidsrod, Carbohydr. Polym. 14, 159–178 (1991). 89. P. Aebischer, E. Buchser, J. Joseph, J. Favre, T. N. de, M. Lysaght, S. Rudnick, and M. Goddard, Transplantation 58, 1275–1277 (1994). 90. A. Atala, W. Kim, K. Paige, C. Vacanti, and A. Retik, J. Urol. 152(2, Ptart 2), 641–643 (1994) (discussion 644). 91. K. Paige, L. Cima, M. Yaremchuk, B. Schloo, J. Vacanti, and C. Vacanti, Plast. Reconstr. Surg. 97(1), 168–178 (1996) (discussion 179–180). 92. S. Wakitani, T. Goto, S. J. Pineda, R. G. Young, J. M. Mansour, A. I. Caplan, and V. M. Goldberg, J. Bone Joint Surg. 76, 579–592 (1994). 93. D. Ceballos, X. Navarro, N. Dubey, G. Wendelschafer-Crabb, W. R. Kennedy, and R. T. Tranquillo, Exp. Neurol. 158, 290–300 (1999). 94. F. A. Auger, M. Rouabhia, F. Goulet, F. Berthod, V. Moulin, and L. Germain, Med. Biol. Eng. Comput. 36(6), 801–812 (1998). 95. S. Kawamura, S. Wakitani, T. Kimura, A. Maeda, A. I. Caplan, K. Shino, and T. Ochi, Acta Orthop. Scand. 69(1), 56–62 (1998).
290
TISSUE ENGINEERING
Vol. 12
96. K. A. Hildebrand, F. Jia, and S. L. Woo, Microsc. Res. Tech. 58(1), 34–38 (2002). 97. C. Vacanti, R. Langer, B. Schloo, and J. Vacanti, Plast. Reconstr. Surg 88, 753–759 (1991). 98. P. X. Ma, B. Schloo, D. Mooney, and R. Langer, J. Biomed. Mater. Res. 29, 1587–1595 (1995). 99. L. E. Freed, R. Langer, I. Martin, N. R. Pellis, and G. Vunjak-Novakovic, Proc. Natl. Acad. Sci. U.S.A. 94, 13885–13890 (1997). 100. Y. Cao, J. P. Vacanti, X. Ma, K. T. Paige, J. Upton, Z. Chowanski, B. Schloo, R. Langer, and C. A. Vacanti, Transplant. Proc. 26, 3390–3392 (1994). 101. A. Atala, J. Vacanti, C. Peters, J. Mandell, A. Retik, and M. Freeman, J. Urol. 148(2, Part 2), 658–662 (1992). 102. G. Organ, D. Mooney, L. Hansen, B. Schloo, and J. Vacanti, Transplant. Proc. 25(1, Part 2), 998–1001 (1993). 103. T. Shinoka, D. Shum-Tim, P. X. Ma, R. E. Tanel, N. Isogai, R. Langer, J. P. Vacanti, and J. E. Mayer Jr., J. Thorac. Cardiovasc. Surg. 115, 536–545 (1998) (discussion 545–536). 104. L. E. Niklason, J. Gao, W. M. Abbott, K. K. Hirschi, S. Houser, R. Marini, and R. Langer, Science 284, 489–493 (1999). 105. T. Shinoka, P. X. Ma, D. Shum-Tim, C. K. Breuer, R. A. Cusick, G. Zund, R. Langer, J. P. Vacanti, and J. E. Mayer Jr., Circulation 94(9, Suppl.), II164–II168 (1996). 106. R. W. Baker, Membrane Technology and Applications, McGraw-Hill, New York, 2000. 107. P. vandeWitte, P. J. Dijkstra, J. W. A. vandenBerg, and J. Feijen, J. Membr. Sci. 117(1/2), 1–31 (1996). 108. R. Bansil, G. D. Liao, and P. Falus, Physica A 231, 346–358 (1996). 109. R. Zhang and P. X. Ma, J. Biomed. Mater. Res. 44, 446–455 (1999). 110. K. M. Woo, R. Zhang, H. Deng, and P. X. Ma, Transaction of the 28th Annual Meeting of the Society for Biomaterial, 2002, p. 605. 111. P. X. Ma and R. Zhang, J. Biomed. Mater. Res. 56, 469–477 (2001). 112. P. X. Ma and R. Zhang, J. Biomed. Mater. Res. 46(1), 60–72 (1999). 113. Y. S. Nam and T. G. Park, J. Biomed. Mater. Res. 47(1), 8–17 (1999). 114. T. Elsdale and J. Bard, J. Cell Biol. 54, 626–637 (1972). 115. S. Yang, K. F. Leong, Z. Du, and C. K. Chua, Tissue Eng. 8(1), 1–11 (2002). 116. R. A. Giordano, B. M. Wu, S. W. Borland, L. G. Cima, E. M. Sachs, and M. J. Cima, J. Biomater. Sci., Polym. Ed. 8(1), 63–75 (1996). 117. A. Park, B. Wu, and L. G. Griffith, J. Biomater. Sci., Polym. Ed. 9(2), 89–110 (1998). 118. J. Zeltinger, J. K. Sherwood, D. A. Graham, R. Mueller, and L. G. Griffith, Tissue Eng. 7, 557–572 (2001). 119. D. W. Hutmacher, T. Schantz, I. Zein, K. W. Ng, S. H. Teoh, and K. C. Tan, J. Biomed. Mater. Res. 55, 203–216 (2001). 120. R. Sodian, M. Loebe, A. Hein, D. P. Martin, S. P. Hoerstrup, E. V. Potapov, H. Hausmann, T. Lueth, and R. Hetzer, ASAIO J. 48(1), 12–16 (2002). 121. T.-M. G. Chu, J. W. Halloran, S. J. Hollister, and S. E. Feinberg, J. Mater. Sci. Mater. Med. 12, 471–478 (2001). 122. S. N. Bhatia, U. J. Balis, M. L. Yarmush, and M. Toner, Biotechnol. Prog. 14, 378–387 (1998). 123. S. Kaihara, J. Borenstein, R. Koka, S. Lalan, E. R. Ochoa, M. Ravens, H. Pien, B. Cunningham, and J. P. Vacanti, Tissue Eng. 6(2), 105–117 (2000). 124. P. X. Ma and J. Choi, Tissue Eng. 7(1), 23–33 (2001). 125. K. M. Woo, V. J. Chen, and P. X. Ma, J. Biomed. Mater. Res. (in press). 126. D. J. Mooney, D. F. Baldwin, N. P. Suh, J. P. Vacanti, and R. Langer, Biomaterials 17, 1417–1422 (1996). 127. K. Whang, C. H. Thomas, K. E. Healy, and G. Nuber, Polymer 36, 837–842 (1995).
Vol. 12
TRANSPORT PROPERTIES
291
128. C. M. Agrawal and R. B. Ray, J. Biomed. Mater. Res. 55(2), 141–150 (2001). 129. R. P. Lanza, R. Langer, and J. Vacanti, Principles of Tissue Engineering, 2nd ed., Academic Press, San Diego, Calif., 2000. 130. A. Atala and R. P. Lanza, Methods of Tissue Engineering, Academic Press, San Diego, Calif., 2002. 131. J. R. Morgan and M. L. Yarmush, in J. M. Walker, ed., Tissue Engineering Methods and Protocols (Methods in Molecular Medicine), Vol. 18. Humana Press, Inc., Totowa, N.J., 1999.
PETER X. MA University of Michigan
TRANSITIONS AND RELAXATIONS.
See VISCOELASTICITY.
