BIOMATERIALS FOR CELL IMMOBILIZATION A look at carrier design KATHRYN W. RIDDLE AND DAVID J. MOONEY University of Michigan, Chemical Engineering, 2300 Hayward, 3074 H.H. Dow Ann Arbor, MI48105 - Fax: 734-763-0459 Email: [email protected] 1. Introdnction There are several major biomedical applications where the transplantation of immobilized cells is being employed to restore, maintain or improve tissue function. These strategies can be split into two main categories: the replacement of biochemical function only or the replacement of structurally functional tissue. As only chemical communication (e.g., diffusion of molecules) is required in the former, it is possible to deliver cells encapsulated in a nanoporous, immunoisolatory polymer membrane. The membranes is constructed such that there are pores large enough to allow for nutrients, waste and the bioactive factor to diffuse but not large enough as to allow immune cells to attack the cells within [1]. This strategy has mainly been employed to temporarily or permanently replace biochemical functions of the liver [2,3], pancreas [4,5], and provide local protein delivery in neurological disorders [6]. The second major strategy involves entrapping cells on a micro or macroporous polymer scaffold and promoting the formation of a new tissue that is structurally and functionally integrated with the surrounding tissue. The scaffold is constructed with a biocompatible material that degrades over time to leave only the integrated tissue in its place. Researchers have attempted to use this strategy with a variety of tissues, including skin [7,8,9], liver [10,11,12], pancreas [13], cornea [14], blood vessels [15,16], cartilage [17,18], heart [19], and bone [20,21]. The biomaterial component of these therapies must provide the appropriate mass transport properties, membrane or scaffold stability and desirable cellular interactions depending on the location and desired function of the implant. In the case of nanoporous immunoisolated strategies, the main goal of the biomaterial is to provide a barrier to the host defences while allowing the essential molecules to diffuse through (Figure 1). In cases where cells are seeded onto a micro or macroporous biomaterial and then implanted, the main function of the biomaterial is to provide cells with a synthetic 15 V. Nedovic and R. Willaert (eds.), Fundamentals ofC ell Immobilisation Biotechnology, 15-32. © 2004 Kluwer Academic Publishers. Kathryn W. Riddle and David J. Mooney extracellular matrix (ECM) that directs multiple cell functions, including cell proliferation, migration and gene expression (Figure 2). This chapter will provide an overview of carrier design of biomaterials for cell immobilization. For the purposes of this chapter, only applications involving mammalian cells will be discussed. A brief overview of cell sourcing is followed by a section on applications of nano-, micro- and macroporous materials, and the common materials used for these applications are described in the last section. Q-.I.)'gen ('ndient. Memtl'ln. I' 0$.\ Ti$sue/ Blood Figure 1. Cartoon ofa nanoporous immunoisolated device. Polymer Figure 2. Cartoon ofa macroporous scaffold seeded with cells. 2. Cell sourcing There are three main sources of cells for cell immobilization biotechnology: autologous (obtained from the same individual for whom the cells are intended), allogeneic (obtained from an individual of the same species for whom the cells are intended), and 16 Biomaterials for cell immobilization xenogeneic (obtained from a species different from which the cells are intended). The appropriate source of cells for a specific application depends on many factors, including functional requirements, availability, ease of collection, processing and storage, and economic factors. Autologous cells potentially offer the most immunocompatible source of cells, and for this reason are frequently used for reconstruction of structural tissues [22]. However, the time required and costs incurred to generate sufficient numbers of these cells from a tissue biopsy is often a deterrent to their application. Specifically, the cells must typically be significantly expanded within a short (e.g., days to weeks) time frame before returning to the patient [23]. In addition, the scarcity of healthy tissue from which to harvest the cells or an inability to expand specific cell types in culture, as with hepatocytes or ~-islets, can make other cell sourcing options necessary. An attractive alternative to autologous cells in many situations is to instead use allogeneic cells. These cells can potentially be expanded in large batches for treating many patients and stored before they are needed [24]. Each batch can be screened for safety and function, decreasing the cost associated with patient-by-patient screening of autologous cells. In addition, genetically modified immortalized cells or cell lines can be extensively studied and characterized for use in cell immobilization strategies [25]. Standardized allogeneic cell sources may also allow the construction of complex tissues that could otherwise be prohibitively expensive in time, money, and effort if they were to be produced by custom manufacture from autologous sources [24]. However, the use of allogeneic cell therapy may require some form of immunoprotection to prevent rejection by the host. For this reason, allogeneic cells have mainly been used in applications where they will eventually be replaced by native cells (e.g., skin replacement), or in immunoisolatory devices that are intended to provide a solely biochemical function. In addition, the limited availability of certain cell types and/or an inability to expand in culture, limits this cell source in certain applications (e.g., treating diabetes with ~-islets). Xenogeneic cells are highly attractive for replacement of biochemical functions in certain situations, as they are available in unlimited supply. Two of the most widely researched applications for xenogeneic cell transplantation are the replacement of liver and pancreatic function [3,4,26,27]. Autologous and allogeneic cells from these organs are scarce and difficult to expand in vitro, leading to great difficulty in identifying good human sources for these strategies [28]. A main challenge in the application of xenogeneic cells is rejection by the host immune system. To overcome these difficulties, these cells must be encapsulated within nanoporous membranes that protect them from rejection. Xenogeneic cells tend to elicit a greater immune response from the host than allogeneic cells in these applications, possibly due to the release of antigenic products from deteriorated and dying cells within the device in both cases [29]. Immunoisolated xenogeneic cells may be useful for both implantation in the body, and as a component of extracorporeal tissue support systems [30]. 3. Material applications Applications for nano, micro and macroporous materials are widely varied from temporary replacement of biochemical function to formation of structurally functional 17 Kathryn W. Riddle and David J. Mooney tissues that are completely integrated with the surrounding host tissue. Nanoporous materials are utilized when it is desired to replace a biochemical function of a tissue by entrapping viable cells within a semi-permeable membrane that prevents cells from being recognized and therefore destroyed by the host immune system [31]. This characteristic is especially useful when allogeneic or xenogeneic cells are used in this strategy. Microporous and macroporous scaffolds provide a synthetic extracellular matrix (ECM) for attachment and proliferation of anchorage dependant cells. These types of scaffolds can be formed using natural or synthetic biomaterials, and have been investigated for the replacement and repair of many tissue types. In all cell encapsulation strategies, it is generally accepted that an appropriate biocompatible material must be isolated or synthesized, and manufactured into the desired shape and dimensions [32]. 3.1. NANOPOROUS Potential applications of nanoencapsulation technology include replacement of the biochemical functions of major organs and transplantation of engineered cells for gene therapy [29]. In many of these situations, the scarcity of healthy autologous tissue from which to harvest cells requires the use of xenogeneic and allogeneic cells requiring immunoisolation. Pancreatic function may potentially be replaced with transplanted porcine islets of Langerhans [5,26] and have been widely investigated as a treatment for diabetes. Similarly, porcine hepatocytes [2,27] have been investigated for temporary replacement of liver function. Other disorders that have been addressed by this method include chronic pain [33,34], hypocalcaemia [35], dwarfism [36], anaemia [37,38], and haemophilia [39,40]. A major limitation to this form of therapy is the limited number of cells that can be delivered, and thus a limited dose of the therapeutic factor secreted by the cells. In some cases, the sheer number of cells needed to replace the function of the damaged organ makes this form of therapy impractical. For example, it has been estimated that approximately 7x109 hepatocytes are required to adequately replace the function of the liver [41]. Considering that the average hepatocyte is approximately 25 microns in diameter [42] and that the average hollow fibre nanoporous device has a diameter of approximately three millimetres [43], it would take an implant 8 meters in length to accommodate the number of cells necessary to adequately replace liver function. In many of these cases implantation of the device is impractical and it is preferable to pursue extracorporeal strategies reviewed elsewhere [41]. 3.1.1. Design considerations Key considerations that must be addressed for this approach to be successful include the required mass transport properties, membrane stability, and cellular distribution with the encapsulation material. 3.1.1.1. Mass transport. Survival of the entrapped cells as well as efficacy of the therapeutic agent will be greatly affected by the mass transport properties of the encapsulation membrane. Ideally, the pore size in the membrane should be carefully controlled to allow bi-directional diffusion of the bioactive factor and molecules 18 Biomaterials for cell immobilization essential for cell survival (e.g., oxygen and glucose), while staying impervious to larger molecules (host complement and antibodies) and immunogeneic cells. One of the major hurdles of this technology is the fact that it is difficult to make a polymer membrane with uniform pore sizes [44]. A distribution of pore sizes leads to a gradient of diffusional resistances to molecules of varying molecular weight, instead of an absolute cut-off. Another limitation to transport can be the fibrous capsule that often forms around the implant due to the foreign body response. For example, a fibrous capsule thickness of 160 microns is sufficient to cause hypoxia for transplanted islets of Langerhans [45]. Due to this issue, materials that don't elicit a significant foreign body response are required for these applications. Other factors that influence the mass transport properties of a nanoporous membrane are the size and shape of the device, as the surface area:volume ratio is a key parameter for diffusional transport [39]. Since the metabolic requirements of each cell type are different, optimal membrane permeability will depend on choice of cells [46]. A variety of structures including, intravascular devices, spherical microcapsules, cylindrical hollow-fibre macrocapsules, flat sheets [47] and planar macrocapsules [48] have been utilized. The spherical shape provides the most surface area to volume ratio, which provides for efficient transfer of nutrients, bioactive factors and waste. However, micro spheres can be difficult to retrieve at a later time due to their small size. If retrieval of the device is desired or large numbers of cells are required, cylindrical and plate designs have been employed, in spite of their inferior mass transport properties [31]. 3.1.1.2. Membrane stability. The need to periodically replenish encapsulated cells, due to limitations on cell longevity, leads to specific requirements for membrane biostability. Following cell death, the implant may be retrieved, or the implant may be left in place and allowed to degrade over time. When short term drug delivery or temporary replacement of biochemical function is desired, it may be desirable for the membrane to slowly erode and allow the body's defences to destroy the implant. For example, a microcapsule system has been designed to degrade and allow clearance of the encapsulated cells following their death or loss of function [49]. These hydrogel based microcapsules, made from alginate and poly (L-lactide), can be made to degrade over several weeks or months and can eliminate the need for surgery to remove old capsules. If retrieval of the implant is necessary or the implant is to be replenished in vivo, a non-biodegradable membrane would be preferable. Having a non-degradable membrane is also particularly useful in applications where genetically engineered cells and cell lines are encapsulated, as the breakdown of the membrane could lead to escape ofthe cells and potential tumour development [50]. 3.1.1.3. Cellular distribution. Nanoporous materials must be designed to optimally organize the encapsulated cells, as a common observation is the appearance of dead or necrotic cells within biomaterial membranes (e.g. hollow fibre devices) [46]. Cells loaded as dilute suspensions in aqueous growth media into these materials presumably settle to create high-density aggregates that hinder diffusional transport of essential nutrients. To overcome this problem, a variety of cell immobilization matrices, including collagen, chitosan and alginate have been examined to more uniformly distribute the cells within the membrane [51]. 19 Kathryn W. Riddle and David J. Mooney 3.2. MICROPOROUS AND MACROPOROUS Potential applications for micro and macroporous cell immobilization include the replacement of tissues with structural and mechanical functions, including heart, blood vessels, bone and cartilage [52,53,54,55]. This contrasts with nanoporous cell encapsulation technologies, which are typically utilized to replace only biochemical functions of tissues. Micro and macroporous matrices, or scaffolds, can be fabricated from polymers or other materials, and act as a temporary ECM for the forming tissue. Example polymers used in these applications include both natural materials (e.g., hyaluronate), and synthetic polymers such as the polyanhydrides and polylactides. These scaffolds mechanically support the formation of tissue, and serve to either transplant cells and/or allow for cells from the surrounding tissue to migrate into this space. 3.2.1. Design considerations Ideally, a micro or macroporous cell immobilization scaffold should have the following characteristics: • Adequate porosity and pore sizes to control mass transport properties • Appropriate mechanical properties for the desired application • Controllable degradation and resorption to match tissue replacement. • Suitable cell-interactions to allow for appropriate tissue development. 3.2.1.1. Mass transport properties. The pore size in micro or macroporous scaffolds modulates the ability of cells and biological molecules to pass into and out of the scaffold. In general, very small pores (e.g., d < II-I.m), while allowing free diffusion of molecules, will not allow cellular migration, while pores in the 10 - 100's of microns readily allow transplanted or host cells to migrate through the scaffold volume. In addition, the rate of tissue ingrowth is dependent on the pore size and overall porosity of the scaffold [56]. Other characteristics that can control transport properties with an implanted scaffold include surface area and volume. A large surface area allows significant cell attachment and growth on the material, a high porosity allows significant host tissue infiltration. In general, it is believed that a porosity of greater than 90% will provide an appropriate surface area for cell-polymer interactions, sufficient space for ECM regeneration, and minimal diffusion constraints [57]. 3.2.1.2. Mechanical design issues. Whereas nanoporous cell encapsulation aims at only replacing a biochemical function, micro and macroporous scaffolds often provide temporary structural support during tissue repair, and therefore require significant mechanical stability. In soft tissue and non-load bearing applications, the mechanical properties (e.g. elastic modulus) of the scaffold are often not required to be extremely high, and a variety of natural polymers (e.g., collagen) and synthetic materials (e.g., poly(lactide-co-glycolide)) have been employed. In contrast, in load bearing applications high strength synthetic materials or reinforced natural polymers (e.g. chitosanlcalcium phosphate composites) are likely required. 20 Biomaterials for cell immobilization Mechanical stimuli conveyed to cells via the scaffold can also regulate the development of many tissues. For example, under cyclical mechanical strain of the scaffold, smooth muscle cells (SMCs) show increased proliferation, and elastin and collagen synthesis [58]. This phenotypic control over SMCs has also been demonstrated using collagen gels [59,60] and polyglycolide scaffolds [61]. Similarly, the development of engineered cartilage has also been found to depend on the mechanical environment [62]. These results indicate scaffolds must be designed to appropriately convey mechanical signals to interacting cells. 3.2.1.3. Scaffold degradation. For each application, the desired rate and method of biodegradation will playa large role in the choice of material. Generally, there are two methods by which biodegradable polymers erode in the body: non-specific hydrolysis and enzymatic degradation. These methods of erosion lead to varying degrees of local vs. pre-programmed control over resorption of the biopolymer. For example, hydrolysis of the ester linkage in poly(lactide-co-glycolide) polymers leads to bulk degradation following the cleavage of ester bonds at random sites along the polymer chains until eventually, lactic acid and/or glycolic acid is produced [63]. This bulk degradation can bring about a rapid loss in mechanical properties and mass loss after prolonged periods of reactive hydrolysis. In contrast, polymers that degrade by surface erosion via enzymatic activity provide a more gradual resorption rate. Enzymatic degradation relies on certain classes of enzymes (e.g. collagenase), secreted by cells in the body, which have the ability to cleave the biopolymer [64]. Since polymer degradation in this case is proportional to the amount of enzyme present, the polymer may erode at a rate controlled by the local cell activity. One interesting approach to controlling polymer degradation is crosslinking synthetic polymers with enzymatically degradable molecules [65]. This concept combines the benefits of the highly controllable properties of synthetic materials with local cellular control over degradation provided by the enzymatically labile cross-links. Regardless of the material, scaffolds should ideally biodegrade at a rate that is comparable to the rate of tissue regeneration so that the scaffold is completely replaced and integrated into the host tissue. 3.2.1.4. Cellular interaction. The surface chemistry of scaffolds can be modified to affect specific cellular responses. These responses may include the rate of host cell invasion, the ingrowth of specific cell types, and cell differentiation. In the past, cell adhesion to biomaterials was frequently based on the cells binding to non-specifically adsorbed proteins (e.g., vitronectin) present in body fluids [65]. A more controlled approach is to covalently, or physiochemically incorporate adhesion-promoting oligopeptides and oligosaccharides on the biomaterial surface. One of the most extensively studied adhesion-promoting peptide sequences is the arginine-glycine aspartic acid (RGD) sequence, which has been shown to promote the adhesion and spreading of a variety of anchorage dependant cell types [66,67,68]. Specific cell types may also be targeted by utilizing peptides, which only those cells can recognize and bind [69]. In addition to cell adhesion peptides, growth factors [70] and plasmid DNA [71] may also be immobilized on the surface of a biomaterial to modify cellular behaviour within and surrounding the implant. 21 Kathryn W. Riddle and David J. Mooney 4. Material chemistry In order to provide the varying requirements for each specific application, a variety of naturally derived and synthetic biomaterials are used for cell encapsulation. These materials can be processed into many different physical forms and geometries. A representative group of these materials, many of the most commonly used, are discussed in this section. 4.1. NATURAL POLYMERS Natural polymers are attractive for cell immobilization due to their abundance and apparent biocompatibility. These polymers can provide a wide range of physical properties that offer unique characteristics for cell encapsulation technologies. 4.1.1. Collagen Collagen is the major component of mammalian connective tissue and has been used in cell immobilization due to its biocompatibility, biodegradability, abundance in nature, and natural ability to bind cells. It is found in high concentrations in tendon, skin, bone, cartilage and, ligament, and these tissues are convenient and abundant sources for isolation of this natural polymer. Collagen can be readily processed into porous sponges, films and injectable cell immobilization carriers. Collagen may be gelled utilizing changes in pH, allowing cell encapsulation in a minimally traumatic manner [72,73]. It may also be processed into fibres and macroporous scaffolds [74,75]. Its natural ability to bind cells makes it a promising material for controlling cellular distribution within immunoisolated devices, and its enzymatic degradation can provide appropriate degradation kinetics for tissue regeneration in micro and macroporous scaffolds. Challenges to using collagen as a material for cell immobilization includes its high cost to purify, the natural variability of isolated collagen, and the variation in enzymatic degradation depending on the location and state of the implant site. [76]. Collagen has been used to engineer a variety of tissues, including skin [77,78], bone [79,80], heart valves [81], and ligaments [82]. 4.1.2. Alginate Alginate is a polysaccharide extracted from seaweed, and has widely been used for cell immobilization due to its biocompatibility and simple gelation with divalent cations (e.g. calcium ions). The polymer is made of the two sugars: D-mannuronate (M) and L guluronate (G) (Figure 3). The M to G ratio in alginate and their distribution will dictate the gelling and mechanical properties ofthe resulting gel [83,84,85]. Alginate can also be covalently crosslinked using bi-functional molecules such as adipic dihydrazide, methyl ester L-lysine, and polyethylene glycol (PEG) [86]. A potential limitation to the use of alginate is its often uncontrollable and unpredictable dissolution, which occurs by a process involving the loss of divalent ions into surrounding fluids [87]. In addition, alginate gels do not promote high levels of protein adsorption [88], and have therefore been modified with lectin or RGD-containing cell adhesion ligands to control cell adhesion [89]. When used as a biomaterial for nanoporous immunoisolated devices, alginate beads are typically coated with a polyamino acid such as poly-L-lysine (PLL) 22 Biomaterials for cell immobilization or poly-L-ornithine (PLO) to establish selective permeability and maintain the capsule durability and biocompatibility in vivo [46]. Alginate has also been examined as a carrier for a variety of cell types, including chondrocytes [85,90], for engineering structurally integrated tissues. Its ability to be inje.=cte d ipn a m:in:im:allty invasive manner into tissues is a significant attraction in this latter application . HO~\ HO~OH 0;; M o Figure 3. Structure of alginate. 4.1.3. Hyaluronic acid Hyaluronic acid, which is the largest glycosaminoglycan (GAG) found in nature, has many attractive features for cell encapsulation (Figure 4). These include ready isolation from abundant natural sources and the minimal inflammatory or foreign body reaction it elicits following implantation [91]. Hyaluronic acid gels can be formed by covalently cross-linking with various hydrazide derivatives [92], and these gels are enzymatically degraded by hyaluronidase [93]. Potential limitations to the use of hyaluronic acid are its long residence time in the body, and the limited range of mechanical properties available from its gels. Moreover, hyaluronic acid requires thorough purification prior to use to remove impurities and endotoxins that may potentially transmit disease or elicit an immune response [87]. Hyaluronic acid has been used in micro and macroporous cell encapsulation as a delivery vehicle for bone-marrow-derived mesenchymal progenitors [94], or as an injectable microporous cell carrier for soft tissue augmentation [91]. It has also been combined with collagen to create an osteoconductive scaffold for bone regeneration [95]. 23 Kathryn W. Riddle and David J. Mooney HO~\ o~o 0=6 I eli> Figure 4. Structure of hyaluronic acid. 4.1.4. Chitosan Chitosan is a deacetylated derivative of chitin, which is widely found in crustacean shells, fungi, insects, and molluscs (Figure 5). Chitosan forms hydrogels by ionic [96] or chemical cross-linking with glutaraldehyde [97], and degrades via enzymatic hydrolysis [98]. Chitosan and some of its complexes have been employed in a number of biological applications including wound-dressings [8], drug delivery systems [99] and space-filling implants [100]. Due to its weak mechanical properties and lack of bioactivity, chitosan is often combined with other materials to achieve more desirable mechanical properties. Specifically, chitosan has been combined with calcium phosphate to increase its mechanical strength for micro and macroporous scaffold applications [100], and has been combined with collagen to provide a more biomimetic microenvironment in nanoporous cell encapsulation applications [l01]. I;;:7i- , ,",'<' ') Figure 5. Structure of chitosan. Figure 6. Structure of agarose. 4.1.5. Agarose Agarose, similar to alginate, is a seaweed derived polysaccharide, but one that has the ability to form thermally reversible gels [46] (Figure 6). Mainly used for 24