Printable Version

Replacing Cells to Fight Disease

A promising treatment for type I diabetes is to encapsulate insulin-producing cells and transplant them into patients -- the challenge is to formulate capsules that prevent immunorejection

by Niraj Vasisht, Ph.D.

Dr. Niraj Vasisht is a senior research engineer in the Microencapsulation and Process Research Section, Chemistry and Chemical Engineering Division. Using physical and chemical techniques, he has encapsulated aromas, antifouling agents, flame retardants, flavors, food oils, monomers, corrosion inhibitors, and light-sensitive material. His current research includes the encapsulation of living cells.

Scientists at Southwest Research Institute (SwRI), in collaboration with researchers at the Thomas E. Starzl Transplantation Institute (TESTI) at the University of Pittsburgh Medical Center, have completed initial work to develop a membrane system, or biomatrix, that may prove highly suitable for human islet cell encapsulation.

TESTI is internationally known for its pioneering work in organ preservation, procurement, and transplantation. TESTI's surgeons, physicians, and researchers are involved in hundreds of transplants annually of hearts, livers, kidneys, pancreases, lungs, and other organs. They are also exploring alternatives that will enhance the quality of life for diabetic patients who now rely on insulin injections.

According to the American Diabetes Association, diabetes is the fourth leading cause of death by disease in this country. At least three quarters of a million people suffer from type I, or early-onset diabetes, characterized by a complete lack of insulin production, and 10 million more have late-onset diabetes. [1]

The illness results from the death of insulin-producing beta cells within the pancreas. Beta cells are clustered with other cells in groups known as islets of Langerhans. Healthy beta cells produce a regulated flow of the hormone insulin to control the body's glucose levels, a process that cannot be matched by the injection of insulin. In fact, uneven glucose levels produced by insulin injections can, over time, result in tissue damage and a number of other complications, including kidney failure.

The work conducted by SwRI and TESTI addresses a therapy for insulin-dependent diabetic patients that is being sought by several researchers, one that mimics normally functioning pancreatic beta cells with little or no need for immunosuppressants.

For more than 100 years, researchers have explored various treatments for diabetes. However, only in the past two and a half decades has dramatic progress been made toward developing a bioartificial pancreas or conducting allogenic (human to human) or xenogenic (across species barriers) islet transplants to recreate the physiological monitoring and release of insulin. One of the most promising approaches for the treatment of type I diabetes is encapsulation of the islets of Langerhans within a biocompatible, semipermeable membrane and subsequent transplantation of the islets into the peritoneum of a diabetic patient.

Despite recent breakthroughs in islet encapsulation and transplantation, long-term, repeated success in human trials has not been achieved. Several obstacles must be overcome to successfully encapsulate and transplant islets. In the case of allogenic transplants, pancreases are hard to obtain; the demand for human cadaver organs far exceeds the supply. In addition, more than one organ is generally needed to yield enough beta cells for a single patient. For this reason, xenogenic transplants of pig islets are being explored, and initial work in this area is promising. [2]

Once the pancreases are obtained, the islets must be isolated and kept alive. Although considerable progress has been made in improving the yield and viability of islets, rejection still poses a formidable challenge to the successful implantation of encapsulated islets. The most effective encapsulation biomaterials, critical to long-term survival and proper functioning of beta cells in the body, have eluded researchers. Another consideration is that the procedure used to coat the islets must do so uniformly to ensure maximum protection from antibodies and other agents of immunorejection.


Several researchers have attempted to transplant encapsulated islets, with varying degrees of success. The following is a small sample of notable foundation work in biomatrix formulation.

In 1979, Lim and Sun developed a procedure to encapsulate islets within a protectivesemipermeable membrane. [3] The cells were suspended in a solution of sodium alginate, a polysaccharide derived from marine plants that precipitates in the presence of calcium ions to form a hydrogel. This highly biocompatible hydrogel, which holds 93 percent water, allows water and dissolved oxygen to permeate readily. Lim and Sun crosslinked the alginate with polylysine, a form of the amino acid lysine, to create a "permeaselective" membrane that would screen out high-molecular weight components. Using this approach, they restored blood glucose levels in rats; however, the same results in larger animals could not be achieved, because the encapsulated islets died when fibroblasts clogged the micropores of the permeable membranes.

The "sandwiched" biomatrix used by Lim and Sun in 1979 consisted of an inner layer of alginate, a middle layer of polylysine, and an outer layer of polyethyleneimine. This biomatrix was modified in 1984 by O'Shea, Goosen, and Sun, who replaced the polyethyleneimine with another layer of alginate to enhance biocompatibility. [4]

Biocompatibility problems persisted, however, which led some researchers to examine the mechanical stability of alginates containing high concentrations of manuronic acid (one of the building blocks of the alginate molecule). Fan and colleagues discovered through in vivo trials in rats that using an alginate-lysine-alginate biomatrix with a large amount of manuronic acid resulted in a breakdown of the outer alginate layer, thus exposing pendant lysine molecules that could in turn stimulate fibrosis. [5] Their conclusions were affirmed by Soon-Shiong and coworkers in 1991. [6]

An ultra-purified alginate containing high guluronic acid content can be crosslinked with a tailored molecular length of chitosan to produce a permeable biomatrix that can both nourish and protect a living cell. The pore size allows smaller molecules such as oxygen, water, glucose, and insulin to diffuse readily through the biomatrix, but should significantly slow or prevent penetration by antibody-size molecules such as lymphocytes and immunoglobin G (IgG).

To produce a more durable alginate layer that would prevent pendant lysine molecules from contacting surrounding tissue, Soon-Shiong's group formulated microcapsules with purified alginate containing a high guluronic acid content. These microbeads exhibited improved biocompatibility, mechanical durability, and adequate porosity, and they prevented the formation of fibroblasts for a significant period of time. In 1994, using this microencapsulation technique, Soon-Shiong demonstrated restoration of normoglycemia in one individual who received transplanted human islets as well as immunosuppressants. [7] This patient has realized relatively long-term benefits from the therapy. The transplanted islets functioned well for 10 months, and the patient has since received multiple intraperitoneal islet transplants with accompanying immunotherapy. Other patients involved in the study did not respond as favorably, however, and efforts to perfect encapsulation materials and processes continue.

Research Hypothesis

As a first step in addressing the complex problems associated with islet transplantation, TESTI and SwRI identified five characteristics as most important in the selection of an encapsulation biomatrix -- biocompatibility with islet cells and neighboring tissues, sufficient permeability to allow passage of cell nutrients, impermeability to antibody-size molecules, a high degree of smoothness and low interfacial tension to reduce frictional irritation of surrounding tissues, and durability.

Purified alginate containing a high concentration of guluronic acid demonstrates excellent biocompatibility, so it is a good choice for a biomatrix. However, antibody-size molecules are capable of penetrating it easily. Combining it with a crosslinking agent is necessary to alter alginate's molecular porosity so antibodies cannot pass through. The outer and inner layers of the matrix are alginate -- sandwiched between them is a diffusion controlling substance.

The research groups led by Lim and Sun, O'Shea, and Soon-Shiong used polylysine to control diffusion. Though it does provide excellent crosslinking with alginate molecules, there is some evidence that, even when coated with a highly purified, high-guluronic-content alginate, polylysine can eventually penetrate the outer alginate layer and encourage fibrosis.

Because lysine elicits this inflammatory response, SwRI researchers decided to replace it with chitosan, a substance extracted from the chitin in crustacean shells and known for its biocompatibility. Chitosan has been used as an ultrapurified natural polysaccharide for treating burns, healing wounds, and lowering cholesterol, and as a drug-delivery suspension material such as that in vitamin gel-caps.

To achieve the desired molecular porosity, SwRI scientists tailored the chitosan to a specific molecular length before sandwiching it between two layers of ultrapurified alginate. The success of this biomatrix would be established by two necessary criteria -- it must elicit no foreign body reaction, and it must allow mass transport of insulin as well as essential nutrients such as oxygen, water, glucose, and vitamins.

Experimental Approach

Chemically, chitosan is partially deacetylated chitin, a linear biopolymer [poly beta (1-4) N-acetyl-D glucosamine]. Because of its glucosamine units, chitosan easily crosslinks with polyanions, such as alginates, to form a hydrogel. This property makes chitosan ideal for use in conjunction with alginate to encapsulate islets. To demonstrate proof of concept, biocompatibility tests were conducted with placebo alginate-chitosan-alginate (ACA) microbeads, and permeability tests were conducted using thin films of the same formulation.

Biocompatibility Studies

A purified chitosan solution was prepared that contained less than one percent acetic acid. Two equal portions of the solution were cleaved to yield two molecular weights of chitosan. The two molecular weights defined the lower and upper limits of optimal nutrient diffusion.

Placebo microcapsules were prepared with an ultrapurified alginate with more than 99 percent guluronic acid content (alginate solution concentration is also critical to the ultimate smoothness of the microcapsule). The alginate solution was extruded with an air stream using a stationary-head coextrusion nozzle, forming capsules 150 to 250 micrometers in size. The microcapsules were collected in a calcium chloride bath, allowing the alginate to react with calcium ions to form a calcium alginate hydrogel. These capsules were combined with the tailored chitosan.

At this point, the alginate inside the microcapsules was reliquified by the addition of sodium citrate. During actual islet encapsulation, the capsule's liquid interior will serve as a protective buffer for the islet, allowing it mobility while not restricting the diffusion of nutrients or insulin in any way. The microcapsules were then recoated with alginate to ensure biocompatibility with surrounding tissue, and the microbeads were intraperitoneally injected into 48 non-immunosuppressed rats at TESTI. A maximum of 2,600 microbeads was transplanted into each rat.

Three sets of microbead formulations, one of alginate only and one of each ACA biomatrix containing different molecular weights of chitosan, were tested for a maximum period of two months. No fibrosis or inflammation was observed in the animals' tissues at the conclusion of the tests.

Permeation Studies

For the permeability studies, thin films (40-100 micrometers) of pure alginate and two ACA biomatrices using two different molecular weights of chitosan were prepared and placed in a diffusion cell, or Payne's cup. The cup separated a solution containing vitamin B12, insulin, myoglobin, and albumin from a saline solution. This system was maintained at 37 degrees Celsius to simulate body temperature, and nutrient diffusion was recorded over time using high-pressure liquid chromatography (HPLC).

Because antibodies such as immunoglobin G (IgG) and immunoglobin M (IgM) are large molecules (approximately one order of magnitude larger than albumin molecules), their diffusion rates are too small to measure by simple diffusion experiments. They are also difficult to measure with an HPLC, because they exhibit broad peak bands that overlap other peaks obtained by the ultraviolet detector in an HPLC. Therefore, albumin was used as a model component to determine effective release amount as a function of time. A comparison of the permeation of albumin (the largest molecular weight molecule used in this study) in films of pure alginate and ACA revealed that the ACA film significantly restricted the diffusion of albumin when compared to alginate film alone.

Results of the nutrient diffusion coefficient experiments confirmed that the ACA membrane provides the type of size-exclusion biomatrix necessary for successful islet encapsulation and transplantation (see table). The large albumin molecules exhibited a diffusion coefficient orders of magnitude lower than that exhibited by the smaller molecules, such as vitamin B12 and insulin. The fact that albumin is significantly smaller than the smallest antibody suggests that permeation of antibodies through the same biomatrix would be even slower. Antibody permeation can only be quantified by future tests with encapsulated islets.


The biomatrix developed for this project was selected based on the hypothesis that pendant lysine molecules may cause immunorejection. Chitosan, a more biocompatible material, was therefore used in conjunction with alginate to form placebo microcapsules for study. The chitosan was tailored to specific molecular lengths to provide the molecular porosity necessary to allow essential nutrients to permeate through the biomatrix, but to prevent antibodies such as lymphocytes from permeation.

The results of this work have consequences not only for intraperitoneal islet transplantation, but for the development of bioartificial pancreases and for the encapsulation of other living cells as well. If the biomatrix is proved effective, it could be used to encapsulate endocrine, pituitary, liver, and osteoprogenitor cells. However, the ultimate goal of the collaboration between SwRI and TESTI pertains to xenotransplantation of encapsulated islets into humans. This is an increasingly important issue in light of the shortage of human cadaver organs.

Researchers at TESTI, encouraged by the successful in vivo placebo microbead tests reported here, will proceed with encapsulation of rat islets. Their approach will be to evaluate cell viability in vitro to demonstrate insulin secretion, and then to conduct in vivo tests to determine whether tissue inflammation or fibrosis develops in rats receiving the encapsulated islets, with and without immunosuppressants. Although partial success in the transplantation of microencapsulated human pancreatic islets has been achieved in an immunosuppressed patient, non-immunosuppressed transplants are of particular interest, as they could prevent the serious complications associated with immunotherapy.

Several issues pertaining to biomatrix fabrication, encapsulation, cell viability, and transplantation remain to be addressed. SwRI and TESTI, with their respective areas of expertise in fabricating biomatrices with tailored microporosity and in islet isolation and transplantation, will continue to pursue long-term solutions for diabetic patients.

Published in the Spring 1997 issue of Technology Today®, published by Southwest Research Institute. For more information, contact Joe Fohn.

Spring 1997 Technology Today SwRI Publications SwRI Home