on germany

8/3/2019 on Germany

1/12

19 Microencapsulation-Based CellTherapy

ULRICH ZIMMERMANNHUBERT CRAMERANETTE JORK

FRANK THRMERWrzburg, Germany

HEIKO ZIMMERMANNGNTER FUHR

Berlin, Germany

CHRISTIAN HASSEMATTHIAS ROTHMUND

Marburg, Germany

1 Introduction 548

2 Bioencapsulation Techniques 5493 Production of Transplantation-Grade Alginates 5514 Biocompatibility Assays for Medically Approved Alginate Gels 5535 Animal and Clinical Trials with Encapsulated Tissue 5556 Conceptual Configuration of Microcapsules for Long-Term Transplantation 558

6.1 Size and Diffusion 5586.2 Swelling and Stability 5606.3 Permeability 5616.4 Topography 5626.5 Transplantation Site and Monitoring of the Functional State of Transplants 5646.6 Xenogeneic Donor Cells 565

7 Concluding Remarks 566

8 References 566

8/3/2019 on Germany

2/12

548 19 Microencapsulation-Based Cell Therapy

1 Introduction

Many diseases are closely tied with deficientor subnormal metabolic and secretory cellfunctions. Diabetes mellitus, Parkinsons dis-ease, hemophilia, hypoparathyroidism, chron-ic pain, and hepatic failure are only a few ex-amples for this kind of degenerative and dis-abling disorders. Milder forms of these diseasescan be managed by a variety of treatments.However, very frequently it is extremely diffi-cult or even impossible to imitate the mo-ment-to-moment fine regulation and thecomplex roles of the hormone, factor, or en-zyme that is not produced by the body (KH-TREIBER et al., 1999). For example, patients

who suffer from the insulin-dependent dia-betes mellitus (IDDM) must take daily insulininjections. While such treatment can restorethe average blood glucose level, true glucosehomeostasis is not achieved. This failure leadsto serious secondary side effects (such as mi-cro- and macroangiopathy, diabetic neuro-pathy, nephropathy, and retinopathy), asso-ciated with a great reduction in life quality andexpectancy. Also, the health care costs arestaggering. Similarly, patients with chronic

hypoparathyroidism show increased neuro-muscular excitability (tetanic convulsions) re-sulting from a deficiency of the parathyroidhormone (parathormone; PTH) that regulatesserum calcium. Patients with hypoparathyr-oidism are usually treated with oral calciumand vitamin D (calcitriol) when the symptomsof this disorder do not disappear and normo-calcemia in the serum is not achieved. How-ever, calcitriol lacks the complete renal cal-cium-retaining ability of parathormone. Ac-

cordingly, such patients have an increased riskof nephrolithiasis, nephrocalcinosis, and sub-sequent impairment of renal function (HASSEet al., 1999 and literature quoted there).

These two examples demonstrate the long-term failure and the high (partly unrealistic)costs of current therapies and the urgent needfor alternative therapeutic strategies. Immuno-isolated transplantation (i.e.,encapsulated-celltherapy) is one of the most promising ap-proaches to overcoming the limitations of thecurrent treatment protocols (LIM and SUN,1980; GEISEN et al., 1990; LANZA et al., 1996).

Instead of drug administration or of engineer-ing the patients own cells (somatic-gene ther-apy), non-autologous standard laboratory celllines, allogeneic (intraspecific), and xenogene-ic (interspecific) cells/tissues are used that re-lease the therapeutic substances that the bodyof the patient cannot itself produce the onlycausal therapy.To avoid a life-time of immuno-suppression therapy while excluding an im-mune response in the host, the transplantsmust be enclosed in immunoprotective cap-sules or devices (COLTON and AVGOUSTINIA-TOS, 1991).

Studies with macrocapsules (e.g., hollow fi-bers, diffusion chambers) made up of differentmaterials have shown a number of drawbacksthat stand in the way of their clinical use (LAN-ZA et al.,1996;KHTREIBER et al.,1999).Asidefrom surgery and retrieval problems, non-spe-cific fibrotic overgrowth, necrosis of the en-capsulated cells due to unfavorable (disk andtube) geometries, and thus diffusion limita-tions, breakage and other problems resultedin the early failure of the grafts. In contrast,microcapsules that are produced from hydro-gels offer potential solutions to the problemsof macrocapsules.

First, because of their spherical configura-

tion and their small size, microcapsules havemuch better surface-to-volume ratios thanmacrocapsules. Second, microcapsules allowprecise tailoring of their permeability to allowdiffusion of anabolic compounds (oxygen, glu-cose,etc.) and of cell-derived products (carbondioxide, lactate, hormones, etc.) while, simul-taneously excluding immunoglobulins. Third,microcapsules minimize the overall risk of im-munoprotection failure by using thousands ofthem instead of a single large macrocapsule.

Fourth, they can be injected directly or trans-planted with minimal-invasive surgery into themuscle, peritoneal cavity, liver, or elsewhere.

Over the past two decades, a number of mi-crocapsules made up of different hydrogels(e.g., alginate, agar, agarose, gellan gum, chito-san, synthetic polymers) have been developedand tested (DULIEU et al., 1999).This researchhas shown the feasibility of alginate-based mi-crocapsules for transplantation of laboratorycell lines as well as of allo- and xenogeneic tis-sue. Numerous technical accomplishments of

this immunoisolation method have recently

8/3/2019 on Germany

3/12

8/3/2019 on Germany

4/12

major shortcomings of alginate-poly-L-lysinecapsules are their vulnerability (including thecells) to the reliquefaction process; their rela-tively weak stability (i.e., their high colloid os-motic pressure due to the dramatic reductionin water activity by the liquid polymers; see al-so Sect. 6), and the potential for an inflamma-tory response to capsule fragments (i.e., poly-electrolyte-alginate residues) upon breakageduring transplantation (PLUMB and BRIDGMAN,1972;FRITSCHY et al.,1991;ZIMMERMANN et al.,1994; DE VOS et al., 1996, 1997a; VAN SCHILF-GAARDE and DE VOS, 1999; GSERD et al.,1999). Moreover, poly-L-lysine is stringentlycytotoxic and is used as an antineoplastic agent(ARNOLD et al., 1983).

The in vitro and in vivo long-term integrity

of the capsules is greatly improved when Ba2c

is used (TANAKA and IRIE,1988;SCHNABL andZIMMERMANN, 1989), but this divalent cation isan inhibitor of the Kc channels present in cellmembranes. Thus, careful control of the beadmanufacturing process and the subsequent re-moval of excessive Ba2c after gelation (seeSect. 6) are important to maintain high cellviability.

For bead formation, cells (or tissue pieces)are first suspended in an iso-osmolar, saline-

sodium alginate solution. The cellalginatesuspension is then forced (by using a syringe ora motor-driven piston) through a nozzle,whichforms droplets of cell-containing alginate.These droplets fall into an iso-osmolar NaClsolution containing 20 mM Ba2c (or Ca2c)that complex with the alginate, resulting in for-mation of spherical beads. The pH of the drop-ping solutions must be adjusted to pH 7 by us-ing histidine or other buffering biomoleculesof very low molecular weight (JORK et al.,

2000). Organic buffers (such as HEPES andMOPS) used by many authors in the past (e.g.,KLCK et al., 1997; DE VOS et al., 1999; ZE-KORN and BRETZEL, 1999) should be avoidedbecause these buffers can be cytotoxic whenreleased from the capsule during transplanta-tion.

Drop formation is greatly improved by ap-plication of a coaxial air jet (GRHN et al.,1994), i.e., by using a two-channel bead gener-ator (Fig.1A).Studies have shown (JORK et al.,2000) that viscosity and concentration of the

alginate, air flow rate, and the geometric prop-

erties of the channels and the nozzle are cru-cial for obtaining microcapsules that are (near-ly) spherical, small in diameter, and with a uni-form size distribution (see Sect. 6).

Optimum drop formation is also obtainedby application of a high electrostatic potentialbetween the nozzle and a stainless-steel ringplaced between the nozzle and the bath solu-tion (Fig. 1C; see also KRESTOW et al., 1991;COCHRUM et al.,1995;GOOSEN,1999).When anaxissymmetric and sinusoidal disturbance of afrequency of about 500 to 7,000 Hz is addition-ally imposed on the laminar jet flow, smalldroplets, with a very narrow size distribution,are formed (PLSS et al., 1997; BRANDENBER-GER and WIDMER,1998;HEINZEN, 1999).

There are several other (commercial) drop-

ping techniques or modifications of the abovedevices (for an excellent overview the readeris referred to a recent review article of DULIEUet al., 1999). For transplantation, the most im-portant one is the three-channel, air-jet beadgenerator (Fig. 1B; see also JORK et al., 2000).This device allows the one-step formation ofmicrocapsules of homogeneous as well as ofspatially heterogeneous composition. Theseinclude solid beads with a liquid core (e.g., oilor other hydrophobic fluids; see Sect. 6) or

solid beads composed of a core of low alginateconcentration (containing the cells) that is sur-rounded by a layer of higher alginate concen-tration (layered solid microcapsules).

Drop formation under coaxial air flow andunder electrostatic potential have been widelyused in bioencapsulation (KHTREIBER et al.,1999). For medical applications the coaxial-air-flow technique is the method of choice be-cause it allows in contrast to the electrostat-ic-potential method the use of highly vis-

cous alginate, i.e., alginate of high molecularmass. Compared to low-viscosity alginatebeads, microcapsules made up of high-viscos-ity alginate provide features that are advanta-geous for long-term transplantation (see Sect.6). A disadvantage of the technique may bethat tiny air bubbles can be included in thebeads during the gelation process (if the pro-cedure is not performed carefully). Such bub-bles may lead to diffusion limitations and/or tolong-term adverse side effects on the bead in-tegrity.


8/3/2019 on Germany

5/12

3 Production of Transplantation-Grade Alginates 551

3 Production ofTransplantation-GradeAlginates

The demands on the alginate material to beused for transplantation are stringent. Itshould be produced with reproducible charac-

teristics according to medical approval stan-dards, and it should not elicit any inflamma-tory or fibrotic response from the host, i.e., itshould not engender any cytotoxicity, andshould be biocompatible for both the host andthe cells it encloses.

Due to the harvesting and extraction pro-cess commercial alginates contain a fairly highnumber of impurities (ZIMMERMANN et al.,1992). Common contaminants are proteins,complex carbohydrates, fatty acids, phospho-lipids, lipopolysaccharides, toxins, and poly-

phenols (SKJK-BRK et al., 1989; DE VOS et

al.,1993;SUN et al., 1996).These mitogenic andinflammation-provoking impurities engenderultimately fibrotic overgrowth (see, e.g., OT-TERLEI et al.,1991;MAZAHERI et al.,1991;WIJS-MAN et al., 1992; COLE et al., 1992; CLAYTON etal., 1993; DE VOS et al., 1993; KLCK et al.,1994) with the result that transport of nutri-ents and oxygen to the encapsulated cells isgreatly impeded leading ultimately to cell ne-

crosis. Removal of the impurities from thecommercial alginate by free-flow electropho-resis or by chemical means (ZIMMERMANN etal., 1992; KLCK et al., 1994, 1997; DE VOS etal., 1997a; VAN SCHILFGAARDE and DE VOS,1999), and subsequent implantation of theempty alginate gels into rodents did not evokeany significant foreign body reaction, evenwhen alginate was implanted in diabetes-prone BB rats that exhibit elevated macro-phage activity (ROTHE et al., 1990; GOTFRED-SEN et al., 1990; WIJSMAN et al., 1992). Exten-

sive research with purified high-M and high-G

Fig. 1. Schematic diagrams of alginate capsule generators;(A) two-channel-, (B) three-channel coaxial air-jetbead generator, and (C) capsule formulation under high electrostatic potential, V. Channel 1 is fed with thealginate/cell suspension, channel 2 serves for air flow supply, and channel 3 is fed with an alginate solution,usually of higher concentration than that of channel 1 (for the formation of layered beads).The size of thebeads is controlled by the speed of the air and alginate flow (A, B), or by the electrostatic potential (C). Forfurther details, see Sect. 2.

8/3/2019 on Germany

6/12

alginates gave further clear-cut evidence (seeZIMMERMANN et al., 1999) that neither theMM nor the GG blocks of alginate polymersof high molecular mass initiate an immunosti-mulatory response (cytokine production) asdiscussed very controversially in the literature(SOON-SHIONG et al., 1991; CLAYTON et al.,1991; OTTERLEI et al., 1991, 1993; ESPEVIK etal.,1993;JAHR et al.,1997;DE VOS et al., 1997a;KULSENG et al., 1999).

Purification of crude commercial alginatehas the decisive disadvantage that many im-purities have also to be removed which are notnatural constituents of the brown algae, butrather present contaminants from the harvest-ing process (pollution by animal proteins, bac-teria products, etc.).Treatment of the raw algal

material by formaldehyde imposes furthercomplications. Purification of alginate is noteasy since high concentrations of alginates aredifficult to work with because of the high vis-cosity of the solutions. Removal of mitogenicand inflammation-provoking contaminants re-quires, therefore, multiple-step and very time-consuming procedures. Because of the largenumber of operations the risk of further con-tamination is increased. As a final result, onlysmall quantities of alginate of quite variable

purity are obtained (KLCK et al., 1994).Techniques of purification and monitoringhave been recently improved sufficiently to al-low the reproducible reduction of mitogenicand cytotoxic impurities to a negligible level.Progress was achieved by using clearly definedalgal material for the production of highlypurified alginate that fulfills the standards formedical application. Research in this directionhas shown (HILLGRTNER et al., 1999; JORK etal., 2000) that fresh stipes of brown algae har-

vested directly from the sea or sporophytes ofbrown algae grown in bioreactors are ideal in-put sources. When using such material, themanufacturing process can be simplified con-siderably. Extraction and purification stepscomprise (see flow chart in Fig. 2; for furtherinformation see JORK et al., 2000): extractionwith 50 mM EDTA, removal of all visible ag-gregates by filtration in the presence of diato-maceous material, adjustment to 0.13 M KCl,precipitation with ethanol (37.5% v/v) underinjection of air or nitrogen, manual sampling

of the alginate layer accumulated at the sur-

face of the liquid phase, redissolution in 0.5 MKCl under agitation, repetition of the precipi-tation (with an ethanol concentration of 45%v/v) and redissolution of the alginate in dis-tilled water followed by dialysis, adjustment ofthe solution to 0.13 M KCl, precipitation (50%v/v ethanol), ethanol sterilization and dryingof the snow-white alginate. All steps must beperformed at room temperature because algi-nate solutions, particularly highly viscous ones,depolymerize when the temperature is raised

(MCHUGH, 1987).


Fig. 2. Flow chart of the alginate purification pro-cess.

8/3/2019 on Germany

7/12

8/3/2019 on Germany

8/12

8/3/2019 on Germany

9/12

5 Animal and Clinical Trials with Encapsulated Tissue 555

GENER, G. ZIMMERMANN, JORK, BOHRER,MELCHER, HASSE, ROTHMUND, ZIMMERMANN,unpublished data). In this case, a slight fibrotic

reaction was observed. However, such a slightfibrotic overgrowth does not prevent nutrientand oxygen exchange between encapsulatedcells and their environment. Rather, as will bedemonstrated in the following section, such areaction is advantageous to reduce capsulebreakage and movement from the transplanta-tion site.

5 Animal and ClinicalTrials with Encapsulated

TissueImmunoisolated-islet transplantation is an

attractive therapy for insulin-dependent dia-betes mellitus (IDDM) patients. Therefore, itis not surprising that extensive animal studieshave been made with encapsulated rat, por-cine, or human islets in the last two decades. Inmost of these experiments islets were en-trapped in alginate-poly-L-lysine made up ofcommercial (non-purified), low-viscosity algi-

nate. Intraperitoneal allo- and xenografts

Fig. 4.In vivo bioassay for the detection of mitogenic impurities in al-ginates based on the induction of fibrotic overgrowth due to the im-mune response of the host. (A) Empty Ba2c alginate beads (diameter300400 m) were made from purified and highly viscous alginate(curve c in Fig.3) by using the dropping device in Fig. 1A. (B) Capsules(arrow) retrieved 3 weeks after implantation beneath the kidney cap-sule (K) of a spontaneously diabetic BB rat and (C) capsules (arrow)retrieved 4 weeks after implantation in the muscle of a baboon.(D) Control capsules (arrow) made up of commercial, unpurified al-ginate (curve b in Fig. 3) and retrieved 3 weeks after implantation be-

neath the kidney capsule (K) of a BB rat (for fixation and staining ofthe tissue/beads see KLCK et al., 1997 and HILLGRTNER et al., 1999).Note that due to the fixation process the alginate beads may collapse(B) or may become deformed (C) (for further explanations, see text);Barsp 350 m.

8/3/2019 on Germany

10/12

could normalize blood glucose of (diabetic)mice and rats for about 100150 d, occasionallysome grafts functioned up to 1 year (OSHEAand SUN,1986;WEBER et al., 1999; WANG,1999;VAN SCHILFGAARDE and DE VOS, 1999). Nor-malization of hyperglycemia by xenotrans-plantation of microencapsulated (porcine) is-lets has also been reported for spontaneouslydiabetic dogs and cynomolgus monkeys (WAR-NOCK and RAJOTTE,1988;SOON-SHIONG et al.,1992;ZHOU et al., 1994; SUN et al., 1996; LANZAet al., 1999). However, as a rule, the resultswere very variable and the success was alwaysof limited duration as expected in the light ofthe above considerations (see also below).Graft failure could occur even 2 weeks aftertransplantation due mainly to foreign body re-

actions (MAZAHERI et al., 1991; WIJSMAN et al.,1992) but sometimes in their absence (DE VOSet al., 1997b).

Ba2c microcapsules made up of alginatepurified from commercial alginates (KLCK etal., 1994) have also been successfully used forthe encapsulation of rat and porcine islets ofLangerhans (Fig. 5; ZEKORN et al., 1992a, b). Inglucose perifusion challenges, evaluation of in-sulin secretion by encapsulated rat isletsshowed the typical biphasic insulin release pat-

tern of non-encapsulated islets. During staticglucose challenge, the insulin release rangedfrom 40% to 70% as compared to the controls.

Accordingly, xenotransplantation of encapsu-lated rat and porcine islets in chemically in-duced diabetic mice demonstrated long-lastinggraft function (up to 1 year) even though fail-ure of some grafts was also observed (about30%; ZEKORN et al., 1992a, b; SIEBERS et al.,1992,1993). Histological examinations of long-term-functioning microcapsules demonstratedwell preserved islets.

Despite the promising results of animalstudies, there has been little success with theclinical allotransplantation of pancreatic isletcells into IDDM patients. This failure has gen-erally been attributed to the inability to obtainlarge numbers of viable human pancreatic is-lets for grafting.About 1 million islets must betransplanted in order to cure diabetes (see be-

low). However, the source of human organs islimited, thus only a small number of patientscould benefit from the encapsulation method.Immunoisolation of porcine islets has the po-tential to fill the gap, but concerns remainabout possible cross-species transmission ofporcine endogenous retrovirus (PATIENCE etal., 1997; but see PARADIS et al., 1999; HUNKE-LER et al., 1999).

Animal and clinical data are available forallo- and xenotransplantation of encapsulated

parathyroid glands. Parathyroid tissue excisedfrom Lewis rats, encapsulated in Ba2c alginatematrices and transplanted in parathyroid-ectomized Dark-Auita rats exhibited long-termfunction (HASSE et al., 1996, 1998). More than6 months after allotransplantation (withoutsystemic immunosuppression) nearly all ani-mals that had received microcapsules made upof amitogenic alginate were normocalcemic.These results were independent of whether thealginate used was purified from commercial al-

ginates according to the protocol of KLCK

etal. (1994) or from fresh algal material (HASSEet al., unpublished data). Throughout the stud-ies PTH and calcium concentrations were al-ways concordant. Accordingly, histology of re-trieved transplants revealed vital parathyroidtissue and intact microcapsules. The trans-plants were partly covered by a very thinfibrotic layer that apparently did not affect thefunction of the encapsulated tissue. Similar re-sults were obtained for xenotransplantation ofencapsulated human parathyroid tissue in rats

with experimental hypoparathyroidism (HAS-


Fig. 5. Uniform preparation of porcine pancreatic is-lets immunoisolated by Ba2c alginate capsules ofsmall diameter. Encapsulation of the islets was per-formed by using the dropping device in Fig. 1A. Se-paration of empty beads (not shown) was achievedby discontinuous density gradient centrifugation ac-cording to the protocol of GRHN et al., 1994.

8/3/2019 on Germany

11/12

5 Animal and Clinical Trials with Encapsulated Tissue 557

SE et al., 1997b). Evaluation of the Ca2c levelin the serum showed that about 100 d aftertransplantation 75% of the animals that re-ceived xenotransplanted human parathyroidtissue were still normocalcemic.

These and other results established the basisfor pilot randomized clinical trials performedrecently (HASSE et al., 1997a). The number ofpatients with clinical hypoparathyroidism ismuch lower than that of IDDM. Thus, thereshould be no shortage of donor human para-thyroid tissue in the future. Cultured allogene-ic parathyroid tissue immunoprotected byBa2c alginate was transplanted into the mus-cle of the right upper arm of two patients suf-fering from symptomatic persistent postopera-tive hypoparathyroidism (HASSE et al., 1997a).

Shortly after allotransplantation, the two pa-tients were normocalcemic and revealed nor-mal levels of PTH without immunosuppres-sion (Fig. 6). With ongoing transplantation

both patients reported an impressive improve-ment of symptoms and sequelae of hypopara-thyroidism. After about 90 d graft failure oc-curred. Residues of the alginate capsules andof the allogeneic tissue could not be foundwhen with permission of one of the patients the transplantation site was re-examined 10months after surgery (HASSE et al., unpublish-ed data). Studies in animal models gave evi-dence (BOHRER et al., unpublished data) thatactivated macrophages degrade encapsulatedparathyroid tissue when capsule breaks occur.However, this cannot be the only explanationbecause one year after transplantation, re-ex-amination of one of the patients showed near-ly normocalcemia and absence of symptoms,requiring much less Ca2c and vitamin D than

prior to surgery. Appropriate tests (CASANOVAet al., 1991) revealed that the PTH was re-leased from the transplantation arm (HASSE etal., unpublished data). The authors of this re-

Fig. 6. Ca2c (open squares) and parathormone (PTH; filled circles) levels in a patientwith hypoparathyroidism after allotransplantation of parathyroid tissue into themuscle of the non-dominant forearm. The allogeneic tissue pieces were encapsulatedby gelation of purified alginate (according to the protocol of KLCK et al., 1994) withBa2c.Note that graft failure occurred after about 3 months, but that functional activity of thetransplant was recorded again about 1 year after transplantation (for further details,see Sect. 5).

8/3/2019 on Germany

12/12

on germany

Documents