Portal de Periódicos da CAPES

The role of capsule composition and biologic responses in the function of transplanted microencapsulated islets of langerhans1

2003; Wolters Kluwer; Volume: 76; Issue: 2 Linguagem: Inglês

10.1097/01.tp.0000078625.29988.0a

1534-6080

Aileen King, Arne Andersson, Berit L. Strand, Joey Lau, Gudmund Skjåk‐Bræk, Stellan Sandler,

Diabetes and associated disorders

Type 1 diabetes patients are dependent on exogenous insulin to control their glucose metabolism. Insulin therapy usually cannot completely normalize hyperglycemia, and patients face long-term complications such as nephropathy, neuropathy, and cardiovascular disease. Moreover, diabetes patients have to monitor their blood glucose and administer insulin by subcutaneous injection on a daily basis. Despite these drawbacks, it should be noted that this disease can be reasonably well controlled by insulin therapy, and any alternative treatment should be superior with respect to the risks and side effects. Indeed, although the Edmonton protocol (1) has been successful in reversing hyperglycemia by the transplantation of islets of Langerhans, the side effects associated with the immune suppression render this treatment at present inappropriate for the majority of type 1 diabetes patients (2). The results of the Edmonton protocol, however, do show that normoglycemia can be achieved by the transplantation of islets and therefore could be a possible treatment of type 1 diabetes if the problems of donor shortage and risks of immunosuppression could be overcome. Cells may be transplanted in the absence of immunosuppression by use of encapsulation (3,4). This technique may even allow transplantation of a xenogeneic source of islet tissue, thus overcoming the problem of islet supply. The principle of encapsulation is that the transplanted cells are physically isolated from the immune system, and thus cell-mediated immune rejection is avoided (5). Encapsulation of islets can be divided into two main groups: macroencapsulation and microencapsulation. In macroencapsulation, the whole graft is placed into one capsule, whereas in microencapsulation, each capsule contains only one or two islets. One major disadvantage of macroencapsulation is that if the capsule breaks, the whole graft may be lost to immune rejection. However, in the case of microencapsulation, the risk of breakage is spread over a large number of capsules and the breakage of one capsule will not lead to complete graft failure. There are two types of macroencapsulation: one is a vascular shunt system and the other is a nonvascularized capsule. In the vascular shunt system, the graft is in close proximity to the blood supply and thus nutrients are more readily available. There are concerns about the safety of such a device, with regard to the operative procedure required for implantation and the risks of thrombosis. Macroencapsulation of islets has been investigated extensively. A problem was that these devices were often associated with a fibrotic overgrowth and the islets in the center of the capsule are prone to necrosis because of poor nutrient diffusion. Microencapsulation of islets may be a good alternative to macroencapsulation. The diffusion properties are much improved, as islets are not packed within the capsule and there is a larger surface area-to-volume ratio. The biocompatibility of microcapsules may be superior because of the circular cross-section of the capsules, which reduces the foreign body response because of the lack of corners. Several different materials have been investigated with regard to microencapsulation of islets of Langerhans. Alginate has been favored by many investigators because of its ability to gel under conditions that are mild to the islet. Moreover, this material has been shown to be nontoxic and indeed is used in many food types and in medications (6). Alginate is a polysaccharide that is found in brown algae and is produced as an extracellular coating by some types of bacteria. It is composed of mannuronic acid (M) and guluronic acid (G) and the ratio and sequence of these uronic acids affect the properties of the alginate gel (6). Alginate forms the core of the capsule and a matrix for the encapsulated cells. It is gelled around the islet in the presence of divalent cations such as calcium (Ca) or barium (Ba), forming uniformly sized beads of approximately 500 μm, provided suitable techniques for capsule formation are used (7). Poly-l-lysine (PLL) is often added to the alginate bead to form a polyanion-polycation complex membrane, which stabilizes the capsule and reduces the porosity of the gel, and thus provides an immunoprotective barrier (8). However, the positive charge of the PLL facilitates cellular adhesion of host macrophages and fibroblasts, and thus a final layer of alginate is added to shield the PLL and improve the capsule’s biocompatibility (9). The function of the microencapsulated islet can be affected by biologic responses, such as the host response to the microencapsulated islets and the ability of the islet to adapt to survive and function within the capsule. The long-term function of the graft is also dependent on the properties of the capsules. Indeed, the alginate and its interaction with PLL can affect the mechanical strength, porosity, and biocompatibility of the capsule (10–12), which in turn affect the function on the microencapsulated islet. In this review, we discuss the biologic responses and capsule properties influencing the function of microencapsulated islets. BIOLOGIC RESPONSES AFFECTING MICROENCAPSULATED ISLETS OF LANGERHANS Insulin Secretion by Microencapsulated Islets For successful transplantation of encapsulated islets, it is essential that the islets secrete insulin properly in response to nutrient stimulation, despite the presence of the capsule. To achieve this, it is essential that the microencapsulated islets are well nourished. Islets are well vascularized in vivo (13), and it is therefore of interest to show that islets can function in the absence of blood vessels. Long-term culture experiments have previously shown that islets can be maintained for prolonged periods in vitro and thus a direct blood supply is not essential to their survival (14). Moreover, insulin release from isolated islets has been shown in stagnant and perfusion systems, demonstrating that insulin release in response to nutrient stimulation can occur by diffusion (15). Indeed, studies of microencapsulated islets have shown that glucose oxidation rates and glucose-induced insulin release in microencapsulated islets is largely unaffected by the presence of the capsule (7,16). Implantation Site The implantation site may also play a role in the nutrient availability to implanted microencapsulated islets. The peritoneal cavity is currently the implantation site of choice because of the large volume of the microencapsulated islet graft. However, this site has a relatively low oxygen tension (17) compared with endogenous islets (18). In this context, it should be noted that the traditional sites for islet transplantation (i.e., the liver and under the capsule of the kidney) also have lower oxygen tension than the pancreas (19). Another disadvantage of the peritoneal cavity is the large number of macrophages present (20), which is the most common cell type in the host response to capsules (21). In addition, the peritoneal cavity is not the physiologic route for insulin delivery, although studies using insulin pumps have shown that blood glucose concentrations are lowered by intraperitoneal delivery of insulin (22). It is therefore not surprising that an increased islet mass is required to reverse hyperglycemia when islets are transplanted intraperitoneally (23). However, unless the size of capsules can be reduced to enable safe transplantation to an alternative site (24), the peritoneal cavity will remain the site of choice for experimental transplantation of microencapsulated islets. Host Response to Microencapsulated Islets A potential barrier to effective nutrient diffusion to microencapsulated islets is the host response to microencapsulated islets. The host response may be directed at the capsule itself (21) or the islet tissue within the capsule (25) and can lead to immune cell attachment on the surface of the capsule. Not only may these cells form a physical barrier to nutrient diffusion but the consumption of glucose by these cells (7) may lead to less nutrients being available to the islets within the capsule. Furthermore, proinflammatory cytokine production by these cells may also inhibit β-cell function, as it has been shown that cytokines cross the membrane (26–28). In normal circumstances, allogeneic tissue rejection occurs through mechanisms dependent on host cell interaction with the donor tissue. Because of the lack of such interaction, rejection of microencapsulated islets is dependent on nondirect mechanisms because of the shedding of antigens. Although it is unlikely that such a reaction would occur in an allogeneic transplantation, as the shed antigens would be too similar to host proteins to induce such a response, this may be a problem in xenogeneic transplantation and possibly in recurrence of the autoimmune disease (29,30). It is likely that in xenogeneic islet transplantation, the capsule will also have to be impermeable to antibodies or components of the complement system. It is thus evident that the porosity of the capsule may have to be smaller for xenogeneic compared with allogeneic transplantation. However, in all transplant types, damaged or hypoxic islets may lead to the production of stress proteins that could egress through the capsule and induce an immune reaction (31). The host response to the capsule can vary extensively between species and even within species and individuals (32). The disparity in results reported may depend in many cases on the animal model used. Cole et al. showed that BB rats had an increased cellular response to implanted capsules, and it was suggested that this may be because of cytokines (27). Indeed, it has been shown that macrophages that are in contact with capsules can produce cytokines (33). We were able to see large differences in cellular response to the capsules when implanted into BALB/c versus C57BL/6 mice, with the latter mouse strain showing a more pronounced cellular response toward the capsules (32). When the cytokine expression of the peritoneal macrophages of these mice was investigated after implantation of capsules, the C57BL/6 mice were shown to have a higher expression of interleukin-1β and tumor necrosis factor-α mRNA, both of which have been associated with the foreign body response (34,35). The host response to the capsules occurs immediately after implantation, with the reaction not increasing after the first week of implantation (32,36). The process is initiated by the inflammation caused by the implantation of the capsules and is difficult to prevent. It is evident that biologic responses can play an important role in the function of microencapsulated islets. Because such responses are difficult to prevent, especially without the use of pharmacologic agents, it is thus apparent that if the capsule properties can be manipulated to reduce host reactions and enhance the function of the microencapsulated islet graft, this may be a favorable course of action. CAPSULE PROPERTIES AFFECTING MICROENCAPSULATED ISLETS OF LANGERHANS Mechanical Strength. It is important that the capsules can withstand a certain amount of physical and osmotic stress, as breakage of the capsules will lead to immune rejection of the exposed encapsulated cells in an allogeneic or xenogeneic transplantation. Osmosis causes the capsules to swell, mainly because of the exchange of Na+ with Ca2+ in the alginate. PLL strengthens the capsule and therefore prevents excessive swelling of the capsule (12). This is achieved by the elasticity of the polyanion-polycation membrane stabilizing the capsule. The interaction of alginate with PLL also leads to a discharge of the alginate; therefore, the amount of osmotically active counterions decreases. The strength of the capsules also depends on the cross-linking of the G blocks with Ca2+. Thus, alginates with a high percentage of G have a greater mechanical strength. Gels can be further strengthened by replacing Ca2+ with Ba2+. If the alginate gel is formed in the presence of a noncharged osmolyte such as mannitol, inhomogeneous beads are formed, which have a higher percentage of alginate at the surface than in the center of the bead. These capsules are stronger than homogenous capsules and capsules with dissolved centers (11). Permeability. The optimal capsule should allow free diffusion of glucose and insulin but prevent cell-mediated and humoral immunity. Indeed, it has been shown that alginate-PLL-alginate capsules prevent cell-mediated cytotoxicity (5) and are impermeable to antibodies (37). Several variables can influence the porosity of an alginate capsule. One important factor is the interaction of alginate and PLL, as the porosity can be decreased by increasing the number of alginate-PLL ionic interactions. This can be achieved by increasing the alginate-PLL reaction time, increasing the concentration of the PLL solution, or decreasing the PLL molecular weight. Also, the composition of the alginate plays a role in the porosity. Not only does it determine the efficiency of the binding of PLL to the alginate bead but it also determines how open the structure of the gel is. For example, alginates with a higher content of the long and inflexible G blocks form a more open Ca-alginate network than alginates with a higher proportion of M (38). Moreover, G (in comparison with M and MG sequences) binds less efficiently to PLL (12), and this together with the more open gel network forms a capsule with a higher porosity. Biocompatibility. The composition of capsules can affect their biocompatibility, which in turn could depend on the alginate per se (39), impurities in the alginate (40,41), or the interaction between alginate and PLL (21). PLL is positively charged and therefore favors the adherence of cells. Furthermore, it has been suggested that PLL is toxic to encapsulated cells, and the necrosis that is induced increases the inflammatory response (42). As described above, the composition of the alginate influences its interaction with PLL, with alginates with high ratios of M or MG sequences binding PLL more efficiently. A strong interaction of the alginate and PLL leads to more effective neutralization of the PLL, preventing its interaction with host cells and thus increasing biocompatibility of the capsule. Indeed, we have seen that empty capsules with a higher M content and lower PLL exposure had improved biocompatibility, and capsules without any PLL induced no or little response (32,42). By using Ba ions to gel the alginate, PLL-free capsules containing islets of Langerhans or other cell types have been shown to function in vivo for several weeks after transplantation (43,44). We were able to reproduce these results, thus showing that PLL-free capsules are able to prevent cell-mediated immune destruction of allogeneically transplanted islets (45). Size. Capsules should be small to allow rapid diffusion of nutrients. In addition, the graft volume should be kept to a minimum. It should be noted that the volume of a capsule is a function of the radius to the power of 3, and therefore if the diameter of the capsule is halved, the volume will be decreased to an eighth. The size of the capsule can affect other properties of the capsule such as biocompatibility (46). Smaller capsules are more prone to osmotic stress and subsequent swelling than larger capsules, because of their increased surface area-to-volume ratio. Such a swelling of the capsule increases the porosity of the alginate (8). Excessive binding of PLL to small beads can cause them to collapse (47) and, to avoid this, the protocol for producing small capsules requires quicker washes with saline or ion-free solution before PLL administration. This leads to less PLL being bound and a consequent increase in porosity. The size of the capsule can be controlled using a high-voltage electrostatic bead generator, by changing the voltage, needle diameter, distance between the needle and gelling solution, or pump flow rate (47). The composition of the alginate can also affect the size of the capsule, depending on the interaction between polymers within the alginate and the binding of alginate to PLL. Thus, capsules with a higher proportion of M and MG sequences tend to be smaller, as the alginate polymers are brought closer together by these flexible sequences. It should be noted that small alginate beads containing a high proportion of M and MG sequences are less resistant to swelling and thus their porosity is not affected because of this (38). ENZYMATIC TAILORING OF ALGINATE With the choice of natural alginates presently available, it is difficult to create a capsule that is strong, small, and biocompatible and that has the desired porosity. Often, the choice of a particular alginate for a specific property is at the expense of another. C5-epimerases are enzymes that convert M into G in the polymer chain of alginate and can be used to create novel alginates with improved properties (48,49). The alginate-producing bacterium Azotobacter vinelandii encodes seven different epimerases, all of which differ in their epimerization pattern. Some of these enzymes (AlgE1 and AlgE6) generate long G blocks, thus increasing the gel-forming capacity of the polymers. In contrast to the G-block enzymes, AlgE4 converts the relatively stiff M blocks into more flexible MG blocks in the alginate polymer (Fig. 1) (48). Because the enzymes prefer to act on long M blocks, epimerization leads to a compositional homogeneity on the otherwise heterogeneous material. The flexibility of the MG blocks interspacing the G-rich junction zone in the polymer network of AlgE4-epimerized alginate leads to closer interaction of alginate polymers in the gel. This tightens the alginate network and therefore increases the strength and decreases the porosity and size of simple alginate beads (50). Furthermore, alginates epimerized in this manner bind PLL more efficiently (50), as the flexibility of the MG blocks allows increased interaction between the alginate and PLL. The alginate to be used as an outer coating of a capsule does not need to gel and therefore G blocks are unnecessary. We have therefore produced alginate by epimerizing a homopolymer mannuronan (poly-M) to form an alginate consisting almost entirely of MG blocks (50). The epimerized alginate shields the positively charged PLL more effectively than alginates containing G blocks and thus the biocompatibility of alginate-PLL-alginate capsules is improved compared to when nonepimerized alginate is used as an outer coating (51). This may enable alginates to be engineered so alginate beads or alginate-PLL-alginate capsules can be produced that have all the desirable properties for the successful transplantation of encapsulated pancreatic islets. Figure 1: Conversion of M blocks to MG sequences by the epimerase AlgE4.CONCLUSION Two major obstacles in the transplantation of microencapsulated islets of Langerhans are the size of the graft and the host response to the capsule. The porosity and strength of capsules is also of utmost importance for successful transplantation of microencapsulated islets. All of these parameters can be affected by the composition of the alginate used to produce the capsule. Alginates can be manipulated by use of enzymes known as epimerases, and therefore alginates can be formed that have a composition not found in nature. Thus, enzymatically engineered alginates may be produced that have all the necessary characteristics to form smaller, stronger, and more biocompatible alginate capsules. Acknowledgments: We thank Ing-Britt Hallgren, Astrid Nordin, and Eva Törnelius for their skillful technical assistance.