[51] MICROENCAPSULATION OF PANCREATIC ISLET CELLS 575

[51] M i c r o e n c a p s u l a t i o n o f P a n c r e a t i c Islet Cells:

A Bioartificial E n d o c r i n e P a n c r e a s
B y ANTHONY M . SUN

Since the pioneering work of Chang and others, TM numerous methods

have been developed for immobilizing biologically active materials such
as enzymes. The next obstacle was to develop a method for the microen-
capsulation of living cells within a semipermeable membrane which would
permit the passage of nutrients and oxygen, but not of cells or high molec-
ular weight substances. However, attempts to synthesize semipermeable
microcapsules which are biocompatible with human and animal tissues
have met with very little success until recently. The in vivo survival times
of transplanted microencapsulated tissue or cells have typically been less
than 2-3 weeks. This limitation severely restricts the applicability of the
microencapsulation procedure to the treatment of diseases such as diabe-
tes 5 which could benefit from cell transplantation.
It has been established that syngeneic islet transplantation can prevent
or reverse the retinal and renal complications of diabetics. However, the
immunogenicity of islet cells remains a major obstacle to the use of this
therapeutic approach. A solution to this problem may be provided by the
introduction of a physical, semipermeable barrier between the trans-
planted islets and the immune system of the host.
Studies in our laboratory have demonstrated that capsules made of an
alginate-poly(L-lysine) (PLL)-alginate membrane survive in situ for
nearly 1 year after transplantation, 6,7 indicating that this membrane is
biocompatible. When isolated rat pancreatic islets were microencapsu-
lated in this membrane and implanted into diabetic rats 7 and mice, 8
the diabetic state was reversed; in some cases, for the life span of the re-
cipient.
With our encapsulation method, greater than 90% of the islet cells
retain their viability and functional properties. However, current studies

i T. M. S. Chang, Science 146, 524 (1964).

2 K. Mosbach and R. Mosbach, Acta Chem. Scand. 20, 2807 (1966).
3 T. M. S. Chang, F. C. Macintosh, and S. G. Mason, Can. J. Physiol. Pharmacol. 44, 115
(1966).
4 V. Hackel, J. Klein, R. Megret, and F. Wagner, Eur. J. Appl. Microbiol. 1, 291 (1975).
5 F. Lim and A. M. Sun, Science 210, 908 (1980).
6 y . F. Leung, G. M. O'Shea, M. F. A. Goosen, and A. M. Sun, Artif. Organs 7, 208 (1983).
7 G. M. O'Shea, M. F. A. Goosen, and A. M. Sun, Biochim. Biophys. Acta 804, 133 (1984).
8 G. M. O'Shea and A. M. Sun, Diabetes 35, 943 (1986).

Copyright © 1988 by Academic Press, Inc.

METHODS IN ENZYMOLOGY, VOL. 137 All rights of reproduction in any form reserved.
576 MEDICAL APPLICATIONS [51]

to optimize some encapsulation parameters such as the purity and viscos-

ity of the sodium alginate, the viscosity average molecular weight (Mv) of
the PLL, the alginate-PLL reaction time, and PLL concentration are
critical in the preparation of improved microcapsules.

Materials
Pancreatic Islets. Rat or dog islets were isolated by the collagenase-
digestion technique and either hand-picked or purified through discontin-
uous Ficoll gradients as described previously.9,1° Briefly, pancreatic tissue
is dressed and perfused with Hank's balanced salt solution. The minced
tissue is then digested for 12-15 rain at 37° with collagenase (12-15 mg/4
ml of tissue). Islets can be handpicked from the digest with the aid of a
dissecting microscope, or the tissue is mixed with 25% Ficoll in a centri-
fuge tube. Three Ficoll concentrations (23, 20, and 11%) are then layered
above this suspension. Centrifugation is carried out for l0 min at 800 g.
The islets are harvested from the interface of the 23 and 20% Ficoll layers.
The cells can either be microencapsulated immediately or cultured for 1-3
days prior to encapsulation.H
Sodium Alginate Solutions. To prepare sodium alginate solutions, 3.0
g of sodium alginate (Kelco Gel L.V.) is sprinkled or sifted slowly into 100
ml of distilled water with stirring. Then 100 ml of 1.8% NaCl solution is
added and well mixed. The mixture is then centrifuged at 30,000 g for I hr
at 4°. The supernatant is sterilized by filtering through a Nalgene filtration
unit (0.2 ~m) and stored at 4°. The 0.15% sodium alginate solution is
prepared by diluting the 1.5% solution 10-fold with physiological saline.
Calcium Chloride Solutions. To prepare I. 1% calcium chloride solu-
tion, 11.0 g of CaC12 (anhydrous) is dissolved in 100 ml of distilled water.
The 0.55% and 0.28% CaCl2 solutions are made by progressive 1 : 1 dilu-
tions of the stock solution with physiological saline.
2-(N-Cyclohexylamino)ethanesulfonic Acid (CHES). To prepare the
stock solution, 2.0 g of CHES is dissolved in 100 ml of 0.6% NaCl, and the
pH is adjusted to 8.2 with 1 N NaOH. Five milliliters of the stock solution
is added to 95 ml of the 1. I% CaC12 solution to make the 0.1% (w/v) CHES
solution.
Poly(L-lysine) Solution (PLL). Fifteen milligrams of PLL (Sigma) is
dissolved in 30 ml of saline.

9 p. E. Lacy and M. Kostianovsky, Diabetes 16, 35 (1967).

10 A. M. Sun, B. J. Lin, and R. E. Haist, Can. J. Physiol. Pharmacol. 51, 175 (1973).
11 A. M. Sun, G. M. Healy, I. Vacek, and H. G. MacMorine, Biochem. Biophys. Res.
Commun. 79, 185 (1977).
[51] MICROENCAPSULATION OF PANCREATIC ISLET CELLS 577

Sodium Citrate Solutions. Trisodium citric acid (3.23 g) is dissolved in

100 ml of distilled water to which 100 ml of saline is added and mixed (55
mM sodium citrate solution).
Culture M e d i u m . Medium CMRL-1969 is supplemented with fetal calf
serum (7.5%) and gentamicin (20 mg/mL) for tissue culture.

Procedure for Microencapsulation

A syringe pump with a 10-ml syringe connected to a special air jet is
used for making sodium alginate islet droplets.
For each preparation, 2000-3000 islets in 0.2 ml of saline is mixed
gently with 2 ml of 1.5% sodium alginate, transferred to the 10-ml syringe,
and connected to the air jet. The distance from the tip of the air jet needle
to the surface of the collecting fluid is set at precisely 4 cm. The syringe
pump and air flow are then turned on to extrude sodium alginate droplets
containing islets into 50 ml of the 1.1% CaCI2 solution in a beaker. During
extrusion, the islets are kept in suspension by gently rotating a small
magnet inside the 10-ml syringe. After the extrusion process is completed,
the spherical calcium alginate gel droplets are transferred to a 50-ml poly-
styrene test tube with a conical bottom, and allowed to settle before
withdrawing the supernatant down to 5 ml using a vacuum aspirator. The
gel droplets are washed once with 30 ml of 0.55% CaC12, once with 0.28%
CaC12, and then suspended in 25 ml of 0.1% CHES solution for 3 min.
After aspirating the CHES solution, the capsules are washed with 1.1%
CaCI2 and suspended in 25 ml of 0.05% (W/V) poly(L-lysine) for 6 min.
After further washing with CHES, CaC12, and saline, the microcapsules
are incubated in 0.15% sodium alginate for 4 min and washed with saline.
The capsules are then suspended in 10 ml of 55 mM sodium citrate solu-
tion for 5 min. The final product is washed twice with saline and once with
medium CMRL-1969, and then transferred to Falcon flasks for incubation
at 37° until required for in vitro and in vivo studies.

In Vitro and in Vivo Assessment of Capsules

For gross assessment, the finished microcapsules can be examined
under an inverted microscope (see Fig. 1). A perfect spherical shape is
important to the in vivo survival time of the capsules. Evaluation of the
surface finish, wall thickness, and membrane uniformity can be done by
scanning electron microscopy.~2 The results should show smooth interior

12 M. F. A. Goosen, G. M. O'Shea, H. M. Gharapetian, S. Chou, and A. M. Sun, Biotech-

nol. Bioeng. 27, 146 (1985).
578 MEDICAL APPLICATIONS [51]

FIG. 1. Micrograph of an alginate-poly(L-lysine)-alginate capsule containing a rat pan-

creatic islet of Langerhans. The diameter of the capsule is 700/xm and the wall thickness 4 ×
5/xm.

and exterior capsular surfaces, with a membrane thickness of 4-5 /xm.

Histochemical staining of paraffin sections with aldehyde thionine is used
to evaluate the viability and integrity of the microencapsulated islets. A
normal degree of glucagon and insulin granulation should be observed in
the encapsulated islet cells.
Insulin secretion from microencapsulated islets in response to glucose
challenge can be measured by (1) long-term culture and (2) perfusion
[51] MICROENCAPSULATION OF PANCREATIC ISLET CELLS 579

experiments using dual chambers, as described by Lacy et al. ~3 In gen-

eral, the insulin secretion from the microencapsulated islets into the cul-
ture media or the perifusate is comparable in quantity with that from
control, unencapsulated islets. In the perifusion experiments, a biphasic
response of insulin release from both groups of islets is observed when the
glucose concentration is raised from 50 to 300 mg/dl.
Detailed descriptions of in vivo tests of the efficacy and biocompatibil-
ity of microencapsulated islets have been published elsewhere. 6,7,8,14
Briefly, diabetes is induced in animals with streptozotocin. Microencap-
sulated islets suspended in saline are implanted into the peritoneal cavity
of diabetic animals, using a cannula attached to a 10-ml syringe. When 4.5
× 10 3 encapsulated islets are implanted in diabetic rats normoglycemia is
restored in recipient animals within 2 days, and maintained for more than
3 months. In our laboratory, one animal remained normoglycemic until
sacrificed, 780 days posttransplantation. Xenografts of encapsulated rat
islets reversed diabetes in mice for a mean of 80 days.
Free intact capsules were recovered from transplant recipient animals
and were shown by histochemical studies to contain viable islet cells.
Furthermore, when placed into culture, the retrieved islets secreted insu-
lin in response to a glucose challenge.

Variations
Variations in membrane properties among the different preparations
are difficult to eliminate. However, the variations can be minimized by
optimizing several important microencapsulation parameters. The
strength and the permeability of the capsule membrane depend on the
viscosity average molecular weight/~v of the PLL used in the encapsula-
tion procedures; the lower the h~/v of PLL, the less permeable the cap-
sules. The alginate-PLL contact time is also a factor which affects the
strength of the capsule membrane. The second coating of sodium alginate
on alginate-PLL capsules plays a key role in the long-term durability of
the capsules during in vivo studies, although the in vitro stability is not
affected. In previous reports microcapsules prepared with alginate-PLL-
polyethyleneimine all failed to survive for a long period of time in animal
studies. ~5
13 p. E. Lacy, E. H. Finke, S. Conant, and S. Naber, Diabetes 25, 484 (1976).
14 A. M. Sun, G. M. O'Shea, and M. F. A. Goosen, in "Biocompatible Polymers, Metals and
Composites" (M. Szycher, ed.), p. 929. Technomic, Lancaster, Pennsylvania, 1983.
~5A. M. Sun, G. M. O'Shea, and M. F. A. Goosen, Appl. Biochem. Biotechnol. 10, 87
(1984).
580 MEDICAL APPLICATIONS [51]

Summary
It was about two decades ago that Chang proposed the use of microen-
capsulated islets as artificial beta cells. By using alginate-poly(L-lysine)-
alginate membranes, biocompatible, durable capsules containing viable
islet cells can be produced which are impermeable to cells and effector
molecules of the immune system, thus providing a total protection to
transplanted islets against rejection. The capsule wall contains 93% (w/w)
water and can be classified as a hydrogel. Many hydrogels have gained
general acceptance as being biocompatible materials. Microencapsulation
of pancreatic islets for use as an artificial endocrine pancreas would not
only obviate the need for immunosuppressive therapy but also has the
potential to prevent the long-term complications of diabetes. Further-
more, the microencapsulation technique can be applied to other types of
cells to produce antibodies or enzymes, and to treat a whole range of
diseases requiring endocrine replacement therapy.