Strategy for cell therapy: polymers for live cell encapsulation and delivery.
Subject: Cellular therapy (Equipment and supplies)
Cellular therapy (Innovations)
Polymers (Varieties)
Polymers (Usage)
Biomedical materials (Research)
Authors: Prakash, Satya
Soe-Lin, Hahn
Pub Date: 07/01/2004
Publication: Name: Trends in Biomaterials and Artificial Organs Publisher: Society for Biomaterials and Artificial Organs Audience: Academic Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2004 Society for Biomaterials and Artificial Organs ISSN: 0971-1198
Issue: Date: July, 2004 Source Volume: 18 Source Issue: 1
Topic: Event Code: 440 Facilities & equipment; 310 Science & research
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 165431727
Full Text: The potential to solve an endemic shortage of donor organs using isolated cells has been cause for widespread optimism in medicine. To overcome limitations of current cell therapy procedures, various polymers have been successfully used. For example, polymer cell encapsulation promises immunoisolation, which has initiated a flurry of research into bioartificial organs. Prospect of polymer encapsulation increasing long-term in vivo cell survivability has opened new avenues for both targeted and recurrent therapeutic drug delivery systems. The current array of polymer research has yielded promising but conflicting results in the search for a viable therapeutic product. Successful capsule designs must fulfill a number of often-conflicting criteria such as mechanical stability, permeability, and biocompatibility. Further compounding the problem is that these criteria can change depending on the nature of the application, which can range from bioartificial organs to targeted chemotherapeutic drug delivery. The resultant conflicting data complicates the development of new capsule designs, as the most appropriate polymer for a given application is frequently unclear. This review summarizes contemporary polymers currently under investigation for live cell encapsulation and therapy. In particular, this review summarizes recent advancement in polymer use for live cell delivery, discusses the principles of using specific polymer type, outlines features of various polymers currently used and addresses limitations for use of polymers in cell therapy, and provides insight for the future direction of this emergent and highly prospective technology.

Polymers for the Delivery of Live Cells

The encapsulation of live cells has long been proposed as a solution to a number of therapeutic problems in modern medicine. The potential to solve problems of donor shortage of organs as well as to provide treatments to areas of immune privilege or poor vascular supply has been cause for widespread optimism in the field. Cell encapsulation promises immunoisolation, which has initiated a flurry of research into bioartificial organs and tissue engineering, while the prospect of encapsulation increasing long-term in vivo cell survivability has opened new avenues for both targeted and recurrent therapeutic drug delivery systems. Accompanying these myriad avenues of research has been the development of a multitude of polymers from which to choose when encapsulating a given cell species. In order to successfully encapsulate a cell while maintaining its function, a polymer with appropriate mechanical strength, pore size, capsule uniformity, degradation characteristics, biocompatibility and low immunogenicity must be chosen. Conflicting reports of success illustrate the complexity of finding an appropriate polymer; variables include not only the type of cell to be encapsulated, but also the environment and duration to which that cell will be exposed. Recognizing the complexities inherent to these considerations, the following is an attempt to provide a comprehensive review of the major polymers currently under investigation for the encapsulation of live cells. These categories include alginate, HEMA-MMA and other hydrogels, PAN-PVC, (HFMs), siliceous encapsulates, cellulose membranes, and molecular variations on each of the above.


Alginate is a natural polymer derived from kelp. It is the most widely used polymer for cell encapsulation (1), possessing unique characteristics extensively reviewed elsewhere (2) that renders it especially suitable in combination with other polymers.

Mechanical Strength/Pore Size

The molecular structure of alginate consists of [beta]-d-mannuronic acid and a-L-guluronic acid residues (Fig. 1). Variations in the ratios of the two molecules (2) as well as the kinetics of the synthesis reaction (3) facilitate the control of polymer properties such as pore size and degradation characteristics (2). Alginates high in guluronic acids tend toward more rigid structure and higher porosity, while the reverse is observed for alginates rich in mannuronic acids (4) (Fig. 2). This structural adjustability is a direct result of [Ca.sup.+2] having a higher affinity for guluronic acids (5). Divalent cations such as [Ca.sup.+2], [Ba.sup.+2], and [Sr.sup.+2] have typically been cross-linked with alginate to induce gel formation (6) via a sol-gel transformation process described elsewhere (2). While barium alginate has been reported to exhibit better mechanical stability, the aforementioned design flexibility that calcium alginate affords to the final polymer structure has resulted in its more frequent employment in encapsulation systems (5).


Complexing poly-L-lysine (PLL) with polycation-linked alginates provides a membrane around the alginate beads (7), resulting in either alginate-PLL (AP) or alginate-PLL-alginate (APA) membranes that have been used in numerous encapsulation applications (8). While APA capsules have been found to be suitable for protecting live cells in a variety of harsh conditions--most notably low pH (9)--over a short term, serious questions regarding mechanical viability arise over long-term in vivo conditions (10). Possible mechanisms for a loss of long-term mechanical integrity include membrane calcium loss (11) and host immune responses (12). This long-term mechanical instability and possible immunogenicity, coupled with high costs, the inherent cytotoxicity of PLL (5), and the tendency for the capsule to rupture instead of deform in response to a critical mechanical load (13) has provided the impetus for research into polymers with better long-term clinical and economic viability. Initial characterization of poly (methylene-co-guanidine) (PMCG) has recently been reported by Orive et al. as an alternative to PLL. Optimal PMCG concentrations were determined and the resultant A-PMCG-A beads were reported to possess increased mechanical stability over APA, [Ca.sup.+2]/alginate, [Ba.sup.+2]/alginate, and Barium APA (BPA) while maintaining the metabolic activity of the encapsulated cell (14). Whether this increase in mechanical stability is sufficient for long-term in vivo capsule viability has yet to be determined.


Hydroxyethyl-methacrylate methyl metha-crylate has long been investigated as an alternative to alginate/PLL polymer-based capsules. HEMA-MMA copolymers belong to a broader family of polymers, known as hydrogels, which are insoluble in water, and thus possess greater mechanical stability in aqueous--i.e. physiological--environments (15). The prominence of HEMA-MMA designs within the hydrogel family warrants a discussion separate from other hydrogels, which will be discussed later in this review. HEMA-MMA copolymers are modifiable with respect to water content and permeability by varying their hydrophobic and hydrophilic monomers (15). The polymer has thus far been employed in the encapsulation of Human Embryo Kidney (HEK) transfected cells (16), dopamine secreting PC12 cells (17), C2C12 mouse myoblast and HepG2 cells (11).

The particular properties of HEMA-MMA preclude the adhesion of cells to the polymer surface (17,11). In the absence of a polymeric adherent surface, anchorage-dependent cells such as HEK clump together, forming self-aggregates. The resulting increase in cell density prevents adequate nutrient diffusion to the aggregate center (11). The consequences are two-fold: 1) the maximum cell density of the capsule is limited below the hypothetical cell density in the absence of aggregation and 2) a central necrotic zone forms. This latter consequence may also have implications for capsule immunogenicity, to be discussed in a subsequent section. The problem is further exacerbated in HEMA-MMA copolymer designs due to the aforementioned aggregate formation. Although early reports cited the addition of ionic monomers as successful inductors of anchorage-dependent fibroblast growth (15) in HEMA-MMA capsules, more recent solutions have focused on co-encapsulation with extra-cellular matrix analogues.

For anchorage-dependent cells, a structural framework is typically a prerequisite for cell adhesion and growth. In vivo, this framework is provided for by the presence of an extra-cellular matrix (ECM). For the purposes of cell encapsulation, ECM analogues provide the framework necessary to promote anchorage-dependent cell adhesion and growth and to prevent excess aggregation and consequent necrotic core formation. Currently agarose, a naturally derived polysaccharide (16), appears to be the analogue of choice for use with HEMA-MMA capsules. In addition to fulfilling the aforementioned criteria, agarose has emerged in co-encapsulation studies as a preferable alternative to both collagen and Matrigel due to lower immunogenicity and easier implementation within the HEMA-MMA capsule synthesis protocol (16). It should be noted that these results are polymer-specific. In PAN-PVC designs--discussed below--precipitated chitosan has also shown excellent promise as a matrix in promoting cell proliferation (18). Lahooti and Sefton also reported that agarose promoted basement membrane synthesis over and above promoting an even distribution of encapsulated 293pHL9 cells (16). These are encouraging results towards the development of self-renewing encapsulated cell populations, a necessary prerequisite for long-term in vivo viability.

As mentioned previously, all capsule designs face multiple dilemmas, as optimum capsule characteristics are frequently mutually exclusive. For example, while a certain minimum pore size and maximum wall thickness are required for adequate diffusion of nutrients and effusion of wastes, an over-permeable capsule will facilitate the infiltration of deleterious immunoglobulins and/or complement components (19) leading to the eventual degradation of the capsule. Conversely, an overly thick capsule would starve the cells of both oxygen and nutrients leading to a necrotic core, with important immunogenic consequences. A fine balance is thus required between capsule thickness and permeability. A recent set of novel microcapsule designs employing HEMA-MMA complexed with both methacrylic acid (MAA) and either collagen, chitosan or alumina sol attempted to circumvent these dilemmas (20). The authors created capsules with multiple polymer shells that added flexibility beyond what was possible by adjusting the properties of the polymers themselves. Single-layered (Type I) capsules employed HEMA-MMA-MAA polymers complexed with collagen, sacrificing mechanical stability and immunoisolation in favour of higher membrane throughput. Although the authors suggested these capsules were suitable for in vitro applications--such as extracorporeal blood filtration with immobilized hepatocytes--their own results suggest that these capsules possess insufficient mechanical stability to remain viable (20). Subsequent capsule designs within the same experiment showed more promise. Multi-layered capsules employed Type I capsules as a base and then re-encapsulated them in either additional layers of the same polymer (Type II) or exoskeletons comprised of either chitosan or alumina sol (Type III). The latter capsule type's outer shell had a positive charge unsuitable for blood-contact environments, thus necessitating the development of a Type III capsule with an extra negatively-charged shell layer, dubbed Type IV. Capsules were then characterized for permeability and cell function, with immunogenicity yet to be determined.


The challenge of preventing cell aggregation and maintaining intracapsular cell viability, as discussed previously with HEMA-MMA capsule designs, lies in choosing an appropriate matrix material. Chitosan, although extensively investigated as a potential encapsulation polymer (5,21,22) is, within the context of live cell encapsulation, more suitable as a matrix material. Recent studies have demonstrated precipitated chitosan within the lumen of PAN-PVC hollow-fibre membranes to promote cell adhesion and growth (18), while chitosan crosslinked with collagen is currently being investigated as a candidate for matrix material in bioartificial livers (23).

The development of bioartificial organs has been facilitated primarily by the ability of cell encapsulation technology to isolate foreign cells from a host immune response. Existing capsular designs, however, have managed to encapsulate a relatively small number of cells compared to the amount needed to replace whole organs (24). A different approach for cell encapsulation design is thus required, as polymers may not retain their mechanical stability or permeability as capsules are scaled up in size. Although hollow-fibre membranes also have their size limits (11), the impetus for larger devices has nonetheless driven the development of polymers especially suitable for macroencapsulation. Recent developments with AN-69 copolymers have attempted to increase existing encapsulation size limits. Existing hand-loading techniques, whereby cells are loaded into hollow fibres via a syringe (24), were insufficient for large-scale encapsulation, and thus a novel procedure was required to encapsulate large quantities of cells within a shorter period of time. Honiger, Sarkis et al. (24) developed a semiautomatic coextrusion device capable of encapsulating 50 million hepatocytes per minute. Cells were still viable after 7 days; however, long-term studies have yet to be conducted. This study represents the first attempt to encapsulate cells in sufficient quantities approaching whole-organ functional replacement.

Other HFMs have emerged from their roots in ultrafiltration applications (25) to provide for competing designs. Thermoplastics, including poly(sulfone), poly(urethane), poly(vinyl alcohol) and poly(amide) (26) have been used as the structural basis for these HFMs, and among those materials investigated it has been established that poly(acrylonitrile-vinylchloride) (PAN-PVC) polymers were most suitable based on their ease of implementation, good biocompatibility (26), and long-term in vivo stability (27).


Structurally, PAN-PVC membranes consist of two layers. A macrovoid layer provides mechanical strength while either one or two skin layers determine permselectivity (28). The latter skin layer is of greater interest, as permselectivity has more complex implications with regards to biocompatibility and nutrient diffusion. The skin layer is comprised of interconnected polymer aggregates known as nodules (28).

Recent studies characterized the effects of both fabrication conditions(26) and solvent treatment on membrane structure and permeability (28). In the former study nodule size and thus permselectivity characteristics were found to be variable functions of precipitation conditions during membrane synthesis (26). In the latter study the authors reported that membrane treatment with the solvent ethanol, ostensibly due to its affinity over water for PAN-PVC, induced membrane swelling via altered membrane permselectivity. The mechanism of this alteration occurred through a series of interactions between the individual polymer units, resulting in an entropically irreversible reaction that increased nodule size (28). The authors admit that the mechanisms linking the measured phenomenon (increased nodular size) with the observed (altered permselectivity) are insufficiently understood to postulate an explicatory hypothesis. These findings, however, represent the first empirical evidence demonstrating that solvents commonly employed in hollow-fibre membrane preparations can significantly alter permselectivity of the membrane. As stated by the authors, these results call into question the legitimacy of all prior conclusions concerning PAN-PVC immunoisolation, as previous reports (29,30) characterized membrane Molecular Weight Cut-Offs--defined as the size of a membrane pore that will reject 90% of the molecules of a corresponding size (25)--exclusive of membrane solvent treatment. Reassessments of these immunoisolation standards are thus required before viable bioartificial organ research employing these membranes can continue.


With the immunoisolation characteristics of hollow-fibre membranes still in question, a discussion of other polymer membranes suitable for macroencapsulation is timely. The aforementioned discovery of PAN-PVC HFM membranes altering their structural characteristics due to swelling necessitates a discussion of polymers that are capable of maintaining their structure while absorbing large amounts of water. Hydrogels represent a long-investigated (31) class of polymeric materials that fit this criterion by definition (32). The majority of excitement has been generated by their potential applications in the field of orthopaedic cartilaginous tissue reconstruction (33), encompassed by the wider field of tissue engineering, a subject appropriate for review elsewhere (34). For the purposes of this review it should be noted that hydrogels have been investigated for the encapsulation of islet cells (35), hepatocytes (36), recombinant E. coli (37), and have also been demonstrated to possess low thrombogenicity (38), rendering them the polymeric device of choice for blood-contact applications.


A multitude of polymers have been investigated for the purposes of hydrogel synthesis, including poly(tetramethylene oxide) (PTMO), poly(dimethyl siloxane) (PDMS), and poly(ethylene oxide) (PEO) (32) poly(ethylene glycol) di-[ethyl phosphatidyl (ethylene glycol) methacrylate] (PhosPEG-dMA) (39), PEG/[PD.sub.5]/PDMS (40), and poly (N, N-dimethyl acrylamide (PDMAAm) (35). Natural polymers, N-Isopropylacrylamide, and PEG/PPO and PEG/PLGA block copolymers have been reviewed elsewhere (41). As a group, hydrogels possess a hydrophilic surface which--although not experimentally correlated--has been widely believed to be responsible for their empirically-established lower protein absorption and cell adhesion (32), the immunogenic consequences of which are discussed subsequently. Park and Bae further reported that this hydrophilicity was variable depending on both polymer composition and topology. Hydrogels are also advantageous as their high water content imparts a softness that minimizes local tissue irritation upon implantation (42). While excellent for biocompatibility, this hydrophilicity may leave hydrogels susceptible to long-term membrane degradation in aqueous environments (37). In light of this concern, more recent hydrogel designs have investigated multi-component membranes. A recent study employed polypentamethylcyclopentasiloxane ([PD.sub.5]) to provide reinforcement to a membrane composed of poly(ethylene glycol) (PEG) and polydimethylsiloxane (PDMS) (40). Mechanisms affecting both [O.sub.2] and insulin permeability and optimal polymer concentrations for mechanical stability were characterized (40). The capsules showed both good mechanical stability and permeability. Immunogenicity for this capsule will be discussed later in this review for the purposes of comparison with other experimental results.

Mechanical Stability

As long-term successful capsule viability has not fully materialized, irrespective of the polymer used, the impetus for improving mechanical stability is obvious. Recent experiments by Isayeva et al. have attempted to improve the mechanical stability of a hydrogel already established as biocompatible with controllable permeability. The polymer in question, poly (N, N-dimethyl acrylamide) (PDMAAm), was cross-linked with multi-telechelic stars. Mechanical stability was indeed improved with increasing telechelic star cross-linkage, up to a point (35). Oxygen permeability, however, while improved over PHEMA [poly(hydroxymethyl methacrylate) hydrogels, was still lower than the established high [O.sub.2] permeability of PDMS (43,44). Although the authors state that the membranes appear sufficiently permeable to sustain long-term in vivo survivability--biocompatibility tests pending--the trade-off illustrates the difficulty with which to obtain ideal membrane conditions for all pertinent viability criteria.

Siliceous membranes

Siliceous membrane represents a much newer approach to the problem of live cell encapsulation. Encapsulation, via the Biosil method described elsewhere (45,46) involves a porous silica layer deposited directly onto the cell membrane (47). The specific silicon alkoxide precursor employed facilitates control of both membrane permeability and mechanical stability (48). As summarized by Boninsegna et al. (47), siliceous deposition occurs as gaseous silicon alkoxides are passed over the cell. Capsule thickness is thus a function of exposure time of the gas to the membrane; ostensibly finer control is possible due to the simplicity of this control mechanism. This finer control, however, comes at the cost of increased preparatory/handling difficulties, resulting in a low capsule yield. Cells must first be suspended on a scaffold before gas can be perfused over their membranes (47), imposing an inherent constraint on efficient bulk encapsulation. Toxicity issues have not yet been fully addressed. Boninsegna et al. reported Et-OH release as a byproduct of membrane formation, which was marginally toxic to encapsulated islets. Although this toxicity affected primarily those islets at the membrane interface and thus did not significantly affect overall insulin secretion, the consequences of this slight decrease in cell viability over the long term have not been established. The authors further report that, excluding immune-mediated capsule deterioration, the number of islets remains relatively constant over long periods of time (47). These results imply better suitability for intercorporeal applications--i.e. exogenous bioartificial organs--however they are inconsistent with the assertion that these siliceous envelopes represent effective immunological barriers. Although siliceous encapsulates possess promising adjustability, further refinements in both synthesis protocol and reagent are required before these polymers become viable encapsulation devices.

Cellulose Sulphate

Cellulose polymers have been investigated both as a primary polymer in dialysis membranes (49), and as a complexed sulphate in both singular (50) and multi-component microcapsules (51,52). Initial bio and mechanical stability of the latter capsules have been established (53), and investigations into their viability for encapsulating both hepatocytes and islets is proceeding. Cellulose sulphate is currently being implemented with both a sodium alginate/PMCG combination (51) and with poly[diallyldimethylammonium chloride] (PDADMAC) (52). In the latter the non-permeancy characteristic of CS provides a mechanically stable barrier while the PDADMAC confers pore adjustability (52). These initial results are promising, and in vitro encapsulation studies are forthcoming.

Sodium alginate/CS/PMCG represents a relatively more mature system. Wang et al. (54) demonstrated that mechanical strength and permeability were independently adjustable, thus overcoming a key dilemma in capsule design. Furthermore, a recent study by Canaple et al. (51) involving hepatocyte encapsulation has demonstrated the suitability of this system, pending both long-term survivability and immunogenicity studies. The study is salient in that the type of alginate used had significant effects on the cytotoxicity to the encapsulated cells (51).

Cellulose sulphate has also been used as the sole polymer in the encapsulation of allogeneic genetically modified cytochrome P450-expressing cells (50). The cells are administered to inoperable pancreatic carcinomas where they locally activate chemotherapeutic ifosfamide. Because of the short-term nature of the treatment, long-term immunogenicity, mechanical stability, and biodegradability are largely irrelevant. It is telling that, without these significant hurdles that have yet to be fully surmounted across all encapsulation technologies, patient studies have already proceeded with promising if preliminary results.


Perhaps the greatest technical hurdle in the field of cell encapsulation is that of preventing a host immune response to implanted capsules. Although many papers claim a given capsule design to exhibit low immunogenicity (16,47,55), individual experiments are limited in both scope and time, and conflicting results (56) that vary with both the type of polymer and the type of cell underscore the complexity with which a clear picture of progress can be obtained. Although the subject has been reviewed elsewhere (12), our review will focus on the specific role of polymers with respect to immunogenicity.

The three major determinants of an immunogenic response to the polymers used and an encapsulated cell are as follows: 1) fibrotic overgrowth, 2) activation of the complement system, and 3) macrophage-mediated capsular degradation.

Fibrotic Overgrowth

Fibrotic overgrowth of an implanted capsule prevents cell viability ostensibly by preventing the adequate diffusion of oxygen and nutrients (57). Early studies illustrated the proclivity of fibroblasts for PLL (58) in AP capsules. This proclivity explains the observed fibrotic overgrowth in several in vivo studies (59-62) and also provided the impetus for the addition of the second alginate layer in alginate-PLL-alginate (APA) designs. APA capsules, however, are still ineffective in completely preventing fibrotic overgrowth (63). Proposed explanations include the irregularities in the surface of the alginate capsule (64), incomplete surface coating (59), and impurities in the alginate polymer (14,63,65). While alginate polymer purity has since been addressed with the advent of commercial purification (63), other solutions were required to improve capsule surface homogeneity.

Anti-fibrotic HOE 077 has recently been investigated to limit fibrosis in pancreatic islets (66). Originally a treatment for liver fibrosis, initial results suggest that the drug is an effective reducer of pericapsular inflammation (PCI), a term encompassing both fibrotic overgrowth and the accompanying inflammatory response. Optimal doses have yet to be determined. Furthermore the mechanism of drug action occurs on fibrosis-inducing cells only in the liver, necessitating the transplantation of encapsulated islets into the liver for the drug to be efficacious (66). While this integration of islets into the liver may limit both the maximal size of implantable islets and their development as discrete bioartificial organs, the concept of administering fibrosis-inhibiting drugs to endogenous cells surrounding an implanted capsule could represent a promising future direction of research.


PEGylation of capsules has also been proposed on the hypothesis that protein adsorption, a prerequisite for an inflammatory response and consequent fibrous overgrowth, can be minimized or even prevented by steric repulsion (67) generated from the encapsulation of the membrane in a negatively charged layer. Sawhney & Hubbell (57) evaluated this hypothesis with AP capsules coated in polyethylene glycol (PEG), with demonstrably excellent anti-fibrotic effects. The improved biocompatibility, however, compromised the mechanical stability by lowering the affinity of PLL for alginate and altering membrane permselectivity. While the compromise was not terminal for the survivability of the capsules, it was clear that conditions were not ideal for long-term encapsulation. This compromise, in combination with the technical complexities of adding PEG to the capsules (55), impelled both a better performing and more easily implemented solution to be sought. Chen et al. (55) recently PEGylated the more commonly used APA capsules, resulting in encapsulated islets that were not only mechanically viable for 120 days but also demonstrated sufficiently low fibrotic growth to maintain cell viability over the same time period.

Fibrotic overgrowth was also observed in HEMA-MMA copolymer designs (17). With respect to the PEG/[PD.sub.5]/PDMS hydrogel combination described previously (40), fibrotic growth was assessed as being present but insufficient to impede biocompatibility. Although the authors conducted the trial for 9 weeks, it is unclear whether the fibrotic growth would have increased--albeit at a slow pace--past the duration of the trial. The authors furthermore cited the minimal fibrotic growth as an improvement over polymers used by Clark et al. (68) and Padera et al. (69), however in light of reports by Chen et al. (55) that most capsules exhibited zero fibrotic growth, the results of Kurian et al. (40) should only be considered promising but not ideal.

Complement Activation

Complement activation requires a cascade of reactions eventually resulting in the formation of a Membrane Attack Complex (MAC), disrupting the osmotic selectivity of the membrane and ultimately limiting long-term cell viability. Prevention of this cascading pathway requires the prevention of complement elements from diffusing through the capsule membrane. Initial experiments focused on developing a membrane with a molecular weight cut-off that would exclude larger elements such as C1q (70), because of the cascading nature of the reaction, the exclusion of a large but singular component could shut down the entire pathway, preventing complement-mediated cell lysis without compromising inward nutrient diffusion. However, many conflicting reports emerged in which membranes impermeable to one molecule would be permeable to a second molecule of lower molecular weight, as summarized by Rihova (12). Both molecular shape and charge were proposed as elements explaining these inconsistent results (71,72).

More recently, Wang (73) illustrated the inherent inconsistency of the molecular weight exclusion model altogether. Because the molecular weight cut-off represents an average, and in reality pore sizes are inhomogeneous, they cannot be counted upon to provide a reliable barrier against immune penetration. This revelation spurred the development of multi-component capsules as described previously (20,73) using a new model of immune entrapment. Whereas the previous barrier model upon which the definition of a molecular weight cut-off is based sought to prevent any immune molecular penetration, the entrapment model pioneered by Wang (73) seeks to let immune molecules through a large, porous outer layer and then trap the molecules between this layer and an inner layer possessing much smaller pores. Given the lack of success in developing a full immunological barrier, this new model may represent a promising paradigm shift towards more effective immunoprotective capsules.

Macrophage-Mediated Capsular Degradation

Rihova (12) stated that host macrophages are perhaps the immune cells that ultimately determine long-term encapsulated cell viability. Macrophages not only release cytokines and other substances that are directly toxic to cells, but they also instigate longer-term reactions such as chronic delayed type hypersensitivity (DTH), ultimately resulting in capsular fibrosis (12). The tendency for a polymeric material to activate macrophages is thus an important concern in biocompatibility studies. Babensee & Sefton (74) recently speculated that HEMA-MMA membranes could act as adjuvant for macrophage activation and recruitment based on recent studies. The propensity of newer designs--such as siliceous encapsulates and membranes employing cellulose sulphate--to activate macrophages has not yet been established. This propensity must be considered in tandem with fibrogenicity, as central necrosis involved in fibrotic capsules results in the escape of antigenic cell components that macrophages can absorb and present as antigens to other components of the immune system.

Risbud, Bhargava et al. (75) recently speculated that internal proteins cyclically digested by cells and displayed as superficial MHC molecules share the same fate as necrotic cell products leaked out to the surrounding environment. Thus while the prevention of antigenic molecules leaking out of the capsule may be an impossible task, the absence of macrophages already proliferating around the capsule due to the low immunogenicity of the membrane should limit both the scale and speed of a full immune response. Inactivating local macrophages altogether, perhaps via endogenous cytokines secreted by the encapsulated cells, should represent a direction in future research towards achieving a fully biocompatible membrane.

Conclusion and Future Direction

Cell encapsulation, although already several decades in development, is ultimately a technology still searching for success. If one picture should emerge from the above discussion, it is that the myriad avenues of research currently being pursued are illustrative of a lack of a definitive emergent encapsulation solution. Although there are many promising new polymers in early development, recurrent difficulties emerge with respect to simultaneously optimizing the characteristics of mechanical strength, pore size, capsule uniformity, degradation characteristics, biocompatibility and low immunogenicity. Promising results with more established polymers have made this last criterion, immunogenicity, perhaps the greatest remaining hurdle to successful therapeutic development. Designs employing alginate-PLL-alginate and multi-layered HEMA-MMA capsules have already solved many of the difficulties regarding optimizing mechanical durability and permselectivity; while newer designs aim to achieve similar goals, albeit on different scales and in different environments, their maturation will take several years. What remains now for the further-developed polymers is the establishment of tangible and reproducible reductions in immunogenicity with subsequent human clinical trials.

Conversely, the scope of the problems cell encapsulation aims to solve also accounts for the wide variety of polymers currently being explored. Environments as diverse as the gastrointestinal tract, the peritoneal cavity, and the brain are all locations for either bioartificial organs, the controlled release of therapeutic drugs, or the perpetual production of substances to treat numerous endogenous deficiencies. Given this enormous variance, it is unlikely that a single polymer will emerge that will be equally suitable in all environments. Although certain polymer combinations such as APA have the potential to be effective in a variety of cell capsule designs, other polymers, such as cellulose sulphate, are finding niche applications (50). In the long-term this development framework, in which multiple polymers fulfill specific roles in varying applications, could result in a more effective implementation of cell encapsulation, much in the way that targeted drug delivery aims to significantly increase the efficacy of previously systemic therapies. In the short-term, however, data resultant from this divergent research generates the excitement for new approaches in polymer design and continues to simulate current polymer research for live cell therapy applications.


This work was supported by Canadian Institute of Health Research (CIHR), Natural Science & Engineering Research Council (NSERC) of Canada, Fonds qudbdcois de la recherche sur la nature et les technologies (FQRNT).


(1.) Gerbsch,N. & Buchholz,R. FEMS Microbiol Rev 16, 259-269. 1995.

(2.) Wee,S. & Gombotz,W.R. Adv. Drug Deliv. Rev 31, 267-285 (1998).

(3.) Thu,B. et al. Biomaterials 17, 1031-1040 (1996).

(4.) Smidsrod,O. & Skjak-Braek,G. Trends Biotechnol. 8, 71-78 (1990).

(5.) De,S. & Robinson,D. J. Control Release 89, 101-112 (2003).

(6.) Stokke,B.T., Smidsrod,O., Bruheim,P. & Skjak-Braek,G. Macromolecules 24, 4637-4645 (1991).

(7.) Uludag,H., De Vos,P. & Tresco,P.A. Adv. Drug Deliv. Rev 42, 29-64 (2000).

(8.) Onsoyen,E. Carbohydrates in Europe 14, 26-31 (1996).

(9.) CuiAH., GohAS., Kim,P.H., Choi,S.H. & Lee,B.J. Int. J. Pharm. 210, 51-59 (2000).

(10.) Strand,B.L. et al. Cell Transplant 10, 263-275 (2001).

(11.) Sefton,M.V., May,M.H., Lahooti,S. & Babensee,J.E. J. Control Release 65,173-186 (2000).

(12.) Rihova,B. Adv. Drug Deliv. Rev 42, 65-80 (2000).

(13.) Bartkowiak,A. et al. Ann. N. Y. Acad. Sci. 875,135-145 (1999).

(14.) Orive,G. et al. Biomaterials 23, 3825-3831 (2002).

(15.) Dawson,R.M., Broughton,R.L., Stevenson,W.T. & Sefton,M.V. Biomaterials 8, 360-366 (1987).

(16.) Lahooti,S. & Sefton,M.V. Biomaterials 21, 987-995 (2000).

(17.) Roberts,T., De Boni,U. & Sefton,M.V. Biomaterials 17, 267-275 (1996).

(18.) Zielinski, B.A. & Aebischer,P. Biomaterials 15, 1049-1056 (1994).

(19.) Babensee,J.E., Cornelius,R.M., Brash,J.L. & Sefton,M.V. Biomaterials 19, 839-849 (1998).

(20.) Chia,S.M. et al. Biomaterials 23, 849-856 (2002).

(21.) Kas,H.S. J. Microencapsul. 14, 689-711 (1997).

(22.) Li,S. et al. J. Control Release 84, 87-98 (2002).

(23.) Wang,X.H. et al. Biomaterials 24, 3213-3220 (2003).

(24.) Honiger,J. et al. Biomaterials 21, 1269-1274 (2000).

(25.) Broadhead,K.W., Biran,R. & Tresco,P.A. Biomaterials 23, 4689-4699 (2002).

(26.) Broadhead,K.W. & Tresco,P.A. Journal of Membrane Science 147, 235-245 (1998).

(27.) Shoichet,M.S. & Rein,D.H. Biomaterials 17, 285-290 (1996).

(28.) Bridge,M.J., Broadhead,K.W., Hlady,V. & Tresco,P.A. Journal of Membrane Science 195, 51-64 (2002).

(29.) Gallefti,P.M., Aebischer,P. & Lysaght,M.J. ASAIO J. 41, 49-57 (1995).

(30.) Uludag,H., De Vos,P. & Tresco,P.A. Adv. Drug Deliv. Rev 42, 29-64 (2000).

(31.) Wichterle,0. & Lim,D. Nature 185, 117 (1960).

(32.) Park,J.H. & Bae,Y.H. Biomaterials 23, 1797-1808 (2002).

(33.) Sechriest,V.F. et al. J. Biomed. Mater Res. 49, 534-541 (1999).

(34.) Shin,H., Jo,S. & MikosAG. Biomaterials. 2003.

(35.) Isayeva,I.S., Kasibhatla,B.T., Rosenthal,K.S. & Kennedy,J.P. Biomaterials 24, 3483-3491 (2003).

(36.) Honiger,J. et al. Biomaterials 16, 753-759 (1995).

(37.) Premkumar,J.R., Rosen,R., Belkin,S. & Lev,O. Analytica Chimica Acta 462, 11-23 (2002).

(38.) Griffith,L.G. Acta Mater 48, 263-277 (2000).

(39.) Wang,D.A., Williams,C.G., Li,Q.A., Sharma,B. & Elisseeff,J.H. Biomaterials 24, 3969-3980 (2003).

(40.) Kurian,P. et al. Biomaterials 24, 3493-3503 (2003).

(41.) Jeong,B., Kim,S.W. & Bae,Y.H. Adv. Drug Deliv. Rev 54,37-51 (2002).

(42.) Anderson,J.M. Eur. J. Pharm. Biopharm. 40, 1-8 (1994).

(43.) Galletti,P.M. & Colton,C.K. Biomedical Engineering Handbook. Bronzino,J.D. (ad.), pp. 129 (CRC Press, Boca Raton,2000).

(44.) Kunzler,J.F. Polymeric Materials Encyclopaedia. Salomone,J.C. (ad.), pp. 1497 (CRC Press, Boca Raton,1996).

(45.) Cappelletti,E.M., Carturan,G. & Piovan,A. Production of secondary metabolites with plant cells immobilized in a porous inorganic support. (US Patent 5,998,162). 1999. USA.

(46.) Carturan,G., Muraca,M. & Dal Monte,R. Encapsulation of supported animal cells using gas-phase inorganic alkoxides. (US Patent 6,214,593). 2001. USA.

(47.) Boninsegna,S. et al. J. Biotechnol. 100, 277-286 (2003).

(48.) SgIavo,V.M., Carturan,G., Dal Monte,R. & Muraca,M. Journal of Materials Science 34, 3587-3590 (1999).

(49.) Risbud,M.V. & Bhonde,R.R. J. Biomed. Mater Res. 54, 436-444 (2001).

(50.) Lohr,M. et al. Lancet 357, 1591-1592 (2001).

(51.) Canaple,L. et al. J. Hepatol. 34, 11-18 (2001).

(52.) Dautzenberg,H., Schuldt,U., Lerche,D., Woehlecke,H. & Ehwald,R. J. Membr. Sci. 162, 165-171 (1999).

(53.) Miyamoto,T., Takahashi,S., Ito,H., Inagaki,H. & Noishiki,Y. J. Biomed. Mater Res. 23, 125-133 (1989).

(54.) Wang,T. et al. Nat. Biotechnol. 15, 358-362 (1997).

(55.) ChenAP., Chu,I.M., Shiao,M.Y., Hsu,B.R.S. & Fu,S.H. Journal Of Fermentation And Bioengineering 86, 185-190 (1998).

(56.) Schneider,B.L., Schwenter,F., Pralong,W.F. & Aebischer,P. Mol. Ther. 7, 506-514 (2003).

(57.) Sawhney,A.S., Pathak,C.P. & Hubbell,J.A. Biomaterials 14,1008-1016 (1993).

(58.) De Vos,P., De Haan,B. & Van Schilfgaarde,R. Biomaterials 18, 273-278 (1997).

(59.) Fritschy,W.M. et al. Transpl. Int. 7, 264-271 (1994).

(60.) Mazaheri,R. et al. Transplantation 51, 750-754 (1991).

(61.) O'Shea,G.M., Goosen,M.F. & Sun,A.M. Biochim. Biophys. Acta 804, 133-136 (1984).

(62.) O'Shea,G.M. & Sun,A.M. Diabetes 35, 943-946 (1986).

(63.) De Vos,P., De Haan,B., Wolters,G.H., Strubbe,J. & Van Schilfgaarde,R. Purification of alginate for microencapsulation of pancreatic islets: effect on biocompatibility and graft function. 2003 (Unpublished)

(64.) De Vos,P., De Haan,B., Wolters,G.H. & Van Schilfgaarde,R. Transplantation 62, 888-893 (1996).

(65.) Zhang,W.J., Laue,C., Hyder,A. & Schrezenmeir,J. Transplant Proc. 33, 3517-3519 (2001).

(66.) Zhang,W.J. et al. Transplant Proc. 32, 206-209 (2000).

(67.) Klein,J. & Luckham,P.F. Colloids and Surfaces 10, 65-76 (1984).

(68.) Clark,H., Barbari,T.A., Stump,K. & Rao,G. J. Biomed. Mater Res. 52,183-192 (2000).

(69.) Padera,R.F. & Colton,C.K. Biomaterials 17, 277-284 (1996).

(70.) Iwata,H. et al. J. Biomed. Mater Res. 28,1201-1207 (1994).

(71.) Iwata,H. et al. J. Biomed. Mater Res. 28,1003-1011 (1994).

(72.) Kulseng,B., Thu,B., Espevik,T. & Skjak-Braek,G. Cell Transplant 6, 387-394 (1997).

(73.) Wang,T.G. Artif. Organs 22, 68-74 (1998).

(74.) Babensee,J.E. & Sefton,M.V. Cell Transplantation 5, 56 (1996).

(75.) Risbud,M.V., Bhargava,S. & Bhonde,R.R. J. Biomed. Mater Res. 66A, 86-92 (2003).

(76.) Hasse,C. et al. World J. Surg. 22, 659-665 (1998).

(77.) Kino,Y., Sawa,M., Kasai,S. & Mito,M. J. Surg. Res. 79, 71-76 (1998).

Satya Prakash * and Hahn Soe-Lin

Biomedical Technology and Cell Therapy Research Laboratory

Department of Biomedical Engineering and Artificial Cells and Organs Research Centre

Faculty of Medicine, McGill University, 3775 University Street, Montreal, Quebec, H3A 2B4, Canada

* Corresponding Author, E-mail:
Table 1: Features of currently established and promising available
polymers for Live Cell Encapsulation

Polymer                     Features                         References

APA [Alginate-Poly          Strengths: Short/Medium-term     2-6,
(I-Lysine)-Alginate]        Mechanical Stability, flexible   9, 10,14,
+ Variants (/PEG,           permselectivity, established     55, 58,
/[Ba.sup.+2]                synthesis protocols, low         62, 63,
/[Ca.sup.+2]                immunogenicity when PEGylated.   65, 66,
                            Weaknesses: Susceptible to       72, 76
                            long-term [Ca.sup.+2] loss,
                            consequent mechanical
                            instability, structurally
                            rigid, must be PEGylated to
                            prevent fibrotic overgrowth.

A-PMCG-A [Alginate-         Strengths: Better mechanical     14
poly(methylene-co-          stability than [Ca.sup.+2]/A,
guanidine)-Alginate]        [Ba.sup.+2]/A, BAPA, PMCG
                            cheaper than PLL, capsule
                            size / permeability
                            independently adjustable.
                            Weaknesses: Immunogenicity,
                            long-term mechanical stability
                            yet to be determined.
HEMA-MMA                    Strengths: Insolubility in       11, 15,
(Hydroxymethylacrylate-     aqueous solutions confers        16, 17,
Methyl Methacrylate)        greater mechanical stability.    19, 74
                            Weaknesses: non-adherent
                            membrane properties requires
                            co-encapsulation with matrix
                            to facilitate anchorage-depen-
                            dent cell adhesion/growth.
Multi-layered HEMA-MMA-     Strengths: Exceptional design    20
MAA                         flexibility, independent
                            adjustment of mechanical
                            stability, permselectivity,
                            promising compatibility
                            with blood-contact
                            Weaknesses: Single-layered
                            capsules possess insufficient
                            mechanical stability,
                            immunogenicity yet to be
                            determined, synthesis protocol
                            more complex than other
PAN-PVC [poly(acryloni-     Strengths: established           25, 26,
trile-vinylchloride)]       mechanical stability,            27, 28,
                            permselectivity, good            56, 68
                            Weaknesses: Molecular-weight
                            cut-offs currently in
                            question, long-term immuno-
                            genicity not yet established.
AN-69 (Acrylonitrile/       Strengths: good mechanical       24
Sodium Methallylsulfonate)  stability, permselectivity,
                            amitogenic, large-scale
                            encapsulation (~50 million
                            cells/minute) now possible.
                            Weaknesses: Immunogenicity not
                            well established.
PEG/[PD.sub.5]/PDMS         Strengths: good mechanical       37, 40
[poly(ethylene glycol) /    stability, PDMS confers
poly(pentamethylcyclopen-   excellent oxygen permeability
tasiloxane)/poly            Weaknesses: long-term
(dimethylsiloxane)]         fibrogenicity not ideal for
PDMAAm [Poly (N,N-          cell encapsulation.
dimethyl acrylamide)]       Strengths: improved mechanical
                            stability when cross-linked
                            with telechelic stars.
                            Weaknesses: oxygen permeabi-     35
                            lity inferior to copolymers
                            with PDMS
Siliceous Encapsulates      Strengths: Simple synthesis      37, 46,
                            mechanism confers high design    48, 47
                            Weaknesses: Questionable
                            toxicity, immunogenicity.
CS/A/PMCG [Cellulose        Strengths: Short-term            49, 52,
Sulphate/Sodium             applications negate long-term    54, 75,
Alginate/Poly(Methylene-    mechanical stability and         50, 77
Co-Guanidine)]              biocompatibility concerns.
                            Weaknesses: Encapsulated cells
                            sensitive to alginate purity
Gale Copyright: Copyright 2004 Gale, Cengage Learning. All rights reserved.